ML14063A529

Attachment 1 to NL-14-030: Akrf Report - Update of Aquatic Impact Analyses Presented in Nrc'S FSEIS (December 2010) Regarding Potential Impacts of Operation of Indian Point Units 2 and 3

ML14063A529
Person / Time
Site:	Indian Point
Issue date:	02/19/2014
From:	AKRF
To:	Entergy Nuclear Operations, Office of Nuclear Reactor Regulation
Shared Package
ML14064A326	List: ML14063A529 NL-14-030, Final Supplemental Environmental Impact Statement NL-14-030, Indian Point Nuclear Generating Unit Nos. 2 & 3 - Final Supplemental Environmental Impact Statement
References
NL-14-030
Download: ML14063A529 (263)

Text

ATTACHMENT 1 TO NL-14-030 AKRF REPORTUpdate of Aquatic Impact Analyses Presented in NRC's FSEIS (December 2010) Regarding Potential Impacts of Operation of Indian Point Units 2 and 3ENTERGY NUCLEAR OPERATIONS, INC.INDIAN POINT NUCLEAR GENERATING UNIT NOS. 2 & 3DOCKET NOS. 50-247 AND 50-286 Update of Aquatic Impact AnalysesPresented in NRC's FSEIS (December 2010)Regarding Potential Impacts ofOperation of Indian Point Units 2 and 32/19/14Prepared for:INDIAN POINT ENERGY CENTER450 Broadway, Suite 1Buchanan, NY 10511Prepared byAKRF, Inc.7250 Parkway Drive, Suite 210Hanover, MD 21076 Table of Contents-I. Introduction and O verview ....................................................................................

1A .Background

...................................................................................................

IB. Inter-annual Changes in LRS, FSS and BSS Sampling Designs ..................

2C. Potential Confounding Effects of Sampling Design Changes .......................

2D .N ew ly A vailable D ata ..................................................................................

3E. A nalysis Update ............................................................................................

4F. Conclusions

...................................................................................................

4II. A nalysis U pdate M ethods ..................................................................................

4A .Trends A nalysis M ethods .............................................................................

4B. SO C A nalysis M ethods ................................................................................

51. A pparent Typographical Errors in FSEIS .................................................

5C .Independent Q uality Control Review ..........................................................

6III. Results ............................................................................................................

7A .Updated Trends A nalyses .............................................................................

7B .Updated SO C A nalyses ................................................................................

8C .U pdated Im pact Conclusions

........................................................................

8IV .D iscussion

.....................................................................................................

9A .N RC 's Precautionary M ethodology

.............................................................

91. SO C M ethods ............................................................................................

92. Trends A nalysis M ethods ........................................................................

103. Conservative Results ................................................................................

10B .FSS G ear Change .........................................................................................

10C .Sum m ary of Changes in Im pact Conclusions

..............................................

121. H ogchoker, W eakfish and W hite Perch ...................................................

122. Striped Bass .............................................................................................

123. B lueback H erring ....................................................................................

134. Rainbow Sm elt .........................................................................................

14V .Literature C ited ................................................................................................

15V I. Figures ..........................................................................................................

16V II. Tables ..........................................................................................................

24V III. A ppendix A ................................................................................................

41IX .A ppendix B ..................................................................................................

42 I. Introduction and OverviewA. Background In December 2010, the Nuclear Regulatory Commission

("NRC") issued the"Generic Environmental Impact Statement for License Renewal of Nuclear PlantsSupplement 38 Regarding Indian Point Nuclear Generating Unit Nos. 2 and 3 FinalReport. NUREG-1437"

("FSEIS").

The FSEIS included a chapter entitled"Environmental Impacts of Operation" which described an assessment of potential impacts of entrainment and impingement at Indian Point Units 2 and 3 ("IP2 and IP3")on fish populations in the Hudson River. That impact assessment was based on two setsof analyses:

1) an assessment of trends in young of year ("YOY") fish populations, and2) and assessment of what NRC referred to as strength of connection

("SOC").

Bothassessments were conducted using data provided to NRC by Entergy, which operatesIP2 and IP3.NRC's trends assessment had two components:

1) riverwide trends in fishabundance, and 2) trends in fish abundance in the sampling region adjacent to IP2 andIP3 (referred to as River Segment 4). For both components, NRC calculated indices ofabundance using Entergy provided data from three Hudson River fish samplingprograms:

1) Long River Survey ("LRL") which collected data on eggs, larvae andjuvenile fish, 2) Fall Shoals Survey ("FSS") which collected data on juvenile and olderfish, and 3) Beach Seine Survey ("BSS") which collected data on juvenile and older fish.Separate indices of abundance were calculated for each species addressed by theassessment.

NRC's indices of abundance for the riverwide trends assessment were estimates of catch per unit effort ("CPUE"),

i.e., the number of fish collected divided by thenumber of samples taken. In addition, NRC used a riverwide index of abundance fromannual reports prepared by electric utility companies that operate power plants on theHudson River and fund and manage the LRS, FSS and BSS sampling programs.

For theRiver Segment 4 indices of abundance, NRC calculated estimates of CPUJE andestimates of density, i.e. the number of fish collected divided by the volume of watersampled.NRC's SOC assessment used estimates of density from River Segment 4 fromthe BSS and FSS to characterize long-term linear trends in abundance and interannual variability in abundance.

That information was coupled with NRC's estimates ofentrainment and impingement mortality rates. NRC's estimates of entrainment andimpingement mortality rates were based on annual estimates of total number oforganisms entrained (1981-1987) and impinged (1984-1990) and estimates of theabundance of entrainable organisms within River Segment 4 from the LRS.I B. Inter-annual Changes in LRS, FSS and BSS Sampling DesignsThe data on fish population abundance in the Hudson River that Entergyprovided to NRC in 2007 and that NRC used for the FSEIS were collected from the 27year period 1979 through 2005. Over that period of years, the data were affected byinter-annual changes in sampling designs.

Major changes in sampling designs included:

1) different sets of weeks of sampling in each year and sampling

program, and 2) achange in the sampling gear used by the Fall Shoals Survey to sample the bottomstratum of the Hudson River. That gear change, from an epibenthic sled to a beamtrawl, occurred in 1985.The BSS sampling design saw a dramatic change in 1981 when the number ofweeks of sampling was greatly curtailed (Figure 1). The number of weeks of samplingmore than doubled by the late 1980's and then remained the same. Starting in 1998,sampling by the FSS beam trawl (which began in 1985) expanded to include the late fall(Figure 2). Like the BSS, weeks of sampling by the FSS epibenthic sled (which wasterminated in 1984) were curtailed in 1981 (Figure 3). The weeks of sampling by theFSS tucker trawl have been fairly consistent although fewer weeks were sampled in theearly 1980's (Figure 4). Starting in 1991, the LRS increased the weeks of sampling toinclude much of the fall (Figures 5 and 6).C. Potential Confounding Effects of Sampling Design ChangesBecause the presence of early life stages of fish in the Hudson River is seasonal, changes in the weeks of sampling can introduce confounding effects to fish abundance data. For example, consider the hypothetical scenario of a species of fish whose larvaeare only present in May of each year. If sampling only occurred during May, then anestimate of CPUE computed as the total number of those larvae collected divided by thetotal number of samples taken would be a valid index of abundance.

Now consider theeffect of doubling the sampling effort in the later years of the sampling program byextending the period of sampling to also include June (when the larvae are no longerpresent).

Estimates of CPUE, computed as the total number of those larvae divided bythe total number of samples taken, for the later years would not be comparable to theestimates of CPUE from the earlier years of the program.

Even if the abundance oflarvae did not change, it would appear as if the abundance had declined to half becausethe estimates of CPUE in the later years would be half the estimates from the earlieryears.Included in the data files provide to NRC by Entergy were data files thatcontained total counts of each species of fish (over all life stages) collected by eachsampling program per year over all weeks of sampling.

One data file of this type wasprovided for each of three sampling programs:

LRS, FSS and BSS. An accompanying file for each sampling program was provided that listed the total number of samplescollected by each program in each year. Those data files apparently were used by NRCto compute annual riverwide catch per unit of effort ("CPUE")

indices of abundance.

2 For each sampling

program, species and year, NRC apparently divided the total numberof fish collected by the number of samples collected to compute an annual CPUE indexof abundance.

For the reasons discussed above, the historical changes in the weeks ofsampling that occurred in the LRS, FSS and BSS appear to have introduced non-trivial confounding effects into NRC's CPUE indices of riverwide abundance.

For the River Segment 4 trends analyses, NRC apparently used a more detailedset of data files provided by Entergy that listed fish density by lifestage and week. Forthe River Segment 4 trends analyses, NRC subset the data to include a more consistent set of weeks in all years, and only included YOY fish (Table 1). Therefore, the RiverSegment 4 trends analyses likely were not as confounded by the changes in weekssampled in each year or by changes in lifestage composition among years.As previously provided in comments to NRC, the change in FSS sampling gearin 1985 appears to have also introduced non-trivial confounding effects to the FSEIStrends and SOC assessments.

The gear change was substantial from the epibenthic sledwith a I m2 mouth opening and 3 mm mesh collection net to the beam trawl with a 2.7m2 mouth opening and 1.3 cm mesh collection net. The use of catch data from surveynets to address trends depends on the assumption of constant collection efficiency overall years of the survey. Collection efficiency can be thought of as the ratio of theaverage number of fish collected in a single sample to the underlying abundance of thosefish in the portion of the river subject to sampling.

For some species, like bay anchovy,the increase in mesh size of the beam trawl allowed YOY fish, which would have beenretained by the smaller mesh of the epibenthic sled, to pass through the beam trawl net.Accordingly, for bay anchovy the collection efficiency of the beam trawl was lower thanthe collection efficiency of the epibenthic sled. Similarly, for species like striped bass,the smaller epibenthic sled appeared to be more easily avoided than the larger beamtrawl. Under those circumstances the beam trawl would have a higher collection efficiency than the epibenthic sled.D. Newly Available DataAs noted above, the data that Entergy provided and NRC used for the FSEISwere collected from the 27 year period of years 1979 through 2005. Since the timeEntergy provided those data, the Hudson River Biological Monitoring Program has beencontinued, with fish data collected through the LRS, BSS and FSS. Data from thoseprograms have been published in the annual series of reports titled, "Year Class Reportfor the Hudson River Estuary Monitoring Program"

("YCR").

Accordingly, data fromthe LRS, BSS and FSS now are available for the 27 year period 1985 through 2011.During this 27 year period of years, there were no gear changes in any of the samplingprograms, thus eliminating this confounding effect and bringing the data current.

Otherregulators also have performed

analyses, which allow the dataset, if brought current, tobe more readily compared to these regulatory findings.

3 E. Analysis UpdateThis report describes an update to the trends and SOC analyses presented in theFSEIS. This update of the FSEIS analyses used the LRS, BSS and FSS data from 1985through 2011. For this analysis update, the data were subset in every year to includeonly a consistent set of weeks for each sampling program (Table 2). Furthermore, thedata used for the trends analyses were subset to include only YOY fish. These stepsremoved the confounding effects on riverwide CPUE indices of abundance due tochanges in the weeks sampled in each year and due to the inclusion of all life stagescollected.

In addition, this analysis update avoids the confounding effects of the FSSgear change that occurred in 1985.F. Conclusions In comparison to the conclusions reported in the FSEIS, results from the updatedanalyses changed the impact conclusions for seven (7) of the 18 aquatic speciesevaluated in the FSEIS:-Alewife changed from Moderate to Small-Blueback Herring changed from Large to Small-Hogchoker changed from Large to Moderate-Rainbow Smelt changed from Moderate-Large to Moderate-Striped Bass changed from Small to Moderate-Weakfish changed from Moderate to Small-White Perch changed from Large to SmallThese changes in impact conclusions were due to a combination of changes in the resultsfrom the trends analyses and from the SOC analyses.

The results from both sets ofupdated analyses were free from confounding effects due to inter-annual changes in theweeks of sampling by the LRS, BSS and FSS. The results from the updated analyses arealso free from confounding effects of the FSS gear change that occurred in 1985. Also,results from the updated riverwide trends analyses were not affected by interannual changes in lifestage composition.

The impact conclusion change for rainbow smelt isdue to newly available information on the range contraction of rainbow smelt on theAtlantic coast.II. Analysis Update MethodsA. Trends Analysis MethodsThe updated trends analyses were conducted according to the methods described in section 1.2.1 of Appendix I of the FSEIS (pages 1-2 through 1-50). Because the updateis based on data from the 27-year period 1985-2011, analysis steps that NRC used to4 address the FSS gear change that occurred in 1985 did not have to be conducted.

Stepsdescribed on the following pages of Appendix I were not performed:

1. pages 1-9 through 1-14: River Segment 4 trends in FSS density2. pages 1-23 through 1-26: River Segment 4 trends in FSS CPUE3. pages 1-34 through 1-37: Riverwide trends in FSS CPUE.Not having to address the issue of the FSS gear change in 1985 greatly simplified thetrends analyses and materially reduced the uncertainty in the results of the trendsanalyses.

B. SOC Analysis MethodsThe updated SOC analyses were conducted according to the methods described in section 1.2.2 of Appendix I of the FSEIS (pages 1-50 through 1-63). In the FSEIS, thecoefficient of variation required for the SOC analyses was calculated from the first 12years of data used in the FSEIS analyses, i.e., 1979-1990 (FSEIS Table 1-46). For theupdated analyses, the coefficient of variation was calculated from the first 12 years ofdata used in the updated analyses, i.e., 1985-1996.

The species-specific entrainment mortality rates ("EMR") and impingement mortality rates ("IMR") used in the FSEISSOC analyses were also used for the updated SOC analyses.

For spottail shiner, theEMR estimate used for the update was taken from NRC's June 2012 "NUREG-1437, Supplement 38, Volume 4, draft supplement to final -Draft Report for Comment".

As described below, some minor changes to the methods as documented weremade for the analysis update to account for apparent typographical errors in the FSEIS.1. Apparent Typographical Errors in FSEISEquation (2) on page 1-51 of the FSEIS indicates that the entrainment mortality rate (EMR) only affects the initial number of fish (No), and that the impingement mortality rate (IMR) only affects the slope parameter (r):No = No( + EMR) and r* =rucL(I-IMR)/max(1,CV))

(2)"where EMR and IMR are conditional mortality rates for entrainment andimpingement; rucL is the upper 95 percent confidence limit of the linearslope; and CV is the coefficient of variation of the annual 75th percentiles from the weekly catch density."

Because the FSEIS SOC analysis was intended to address entrainment and impingement, it appears that the omission of IMR from the definition of No, and the omission of EMRfrom the definition of r* were typographical errors. Therefore, for the analysis update,5 equation (2) was revised so that both the entrainment mortality rate and the impingement mortality rate affected the initial number of fish and the slope parameter:

No = No(I + EMR + IMR) and r* =rUCL(I-EVIR-IMiR)/max(1,CV))

(3)Table 1-46 in the FSEIS lists values for the "Upper 95% Confidence Limit of theSlope" that were used with equation (2). As shown below, the values in the columnlabeled "Upper 95% Confidence Limit of the Slope" apparently were mislabeled.

Rather than the upper 95% confidence limits they are the slope estimates plus onestandard error.The "Linear Slope (r)" and "Upper 95% Confidence Limit of the Slope" entriesin Table 1-46 were taken from Table 1-9 (for FSS) and Table 1-12 (for BSS) of theFSEIS. For each species, the entry in the column labeled "Upper 95% Confidence Limitof the Slope" in Table 1-46 is the corresponding linear regression slope estimate fromTable 1-9 or 1-12 plus the undefined value to the right of the +/- symbol in the linearregression slope column of the table. It can be shown from the "p-value",

also listed inTables 1-9 and 1-12, that the undefined value to the right of the +/- symbol is the standarderror of the linear slope estimate (Appendix A).Therefore, the entries listed in Table 1-46 are, if fact, the slopes plus one standarderror. If those entries had been upper 95% confidence limits, they would have beenapproximately equal to the slopes plus two standard errors.Based on the values listed in Table 1-46, it appears that the SOC analysespresented in the FSEIS were conducted with rUcL in equation (2) set equal to theestimated slope plus the standard error of the slope. Accordingly, to be consistent withthe SOC analyses presented in the FSEIS, the value of rucL in equation (2) was set to theestimated slope plus the standard error of the slope for this analysis update.C. Independent Quality Control ReviewThe updated analyses were conducted using data analysis programs written withSAS computer

software, and all data inputs were in SAS format data files. The fullset of computer programs and input data files used for the updated analyses weresubmitted to John Young, PhD of ASA Analysis

& Communication, Inc. for a thoroughquality control review. The purpose of the review was to determine whether thecomputer code was correctly written to accurately conduct the analyses documented inthe FSEIS. Dr. Young has decades of experience working with data files from theHudson River Biological Monitoring Program.

Dr. Young's independent review of thecomputer programs and input data files used for the updated analyses confirmed thecomputer programs accurately reflected the analysis methods documented in the FSEISand identified no computer programming errors (see Appendix B).6 III. ResultsA. Updated Trends AnalysesFollowing the trends analysis methods documented in the FSEIS, a total of ninesets of trends analyses were conducted:

three for River Segment 4 density (FSS, BSSand LRS), two for River Segment 4 CPUE (FSS and LRS), three for riverwide CPUE(FSS, BSS and LRS), and one for the YCR abundance indices.

Each set of trendsanalyses included analyses conducted using linear regression and analyses usingsegmented regression.

For each species, one type of regression was selected using thedecision rules documented in the FSEIS. Based on the results of the selected type ofregression

analysis, each species was assigned a trend score of either 1 (i.e., no declinedetected) or 4 (i.e., decline detected).

For River Segment 4 trends, comparisons of results from the two types ofregressions are summarized in the following tables:-FSS Density (Table 3)-BSS Density (Table 5)-LRS Density (Table 7)-FSS CPUE (Table 9)-LRS CPUE (Table 11)The corresponding trend conclusions (i.e., score of 1 or 4) for these five sets of analysesare summarized in the following tables:-FSS Density (Table 4)-BSS Density (Table 6)-LRS Density (Table 8)-FSS CPUE (Table 10)-LRS CPUE (Table 12)The River Segment 4 trends conclusions, based on the average score from the five setsof analyses are listed in Table 13.For riverwide trends, comparisons of results from the two types of regressions are summarized in the following tables:-FSS CPUE (Table 14)-BSS CPUE (Table 16)-LRS CPUE (Table 18)-YCR Abundance Index (Table 20)7 The corresponding trend conclusions (i.e., score of 1 or 4) for these four sets of analysesare summarized in the following tables:-FSS CPUE (Table 15)-BSS CPUE (Table 17)-LRS CPUE (Table 19)-YCR Abundance Index (Table 21)The riverwide trends conclusions, based on the average score from the four sets ofanalyses are listed in Table 22.The overall trends conclusions, which were based on weighted averages of theRiver Segment 4 scores and riverwide scores, are summarized in Table 23. Incomparison to the conclusions reported in the FSEIS, results from the updated analyseschanged the trends conclusions for 8 of the 13 species analyzed in the FSEIS:-Alewife changed from Variable to Undetected Decline-Bluefish changed from Detected Decline to Undetected Decline-Hogchoker changed from Detected Decline to Variable-Spottail Shiner changed from Detected Decline to Undetected Decline-Striped Bass changed from Undetected Decline to Variable-Weakfish changed from Variable to Undetected Decline-White Catfish changed from Variable to Undetected Decline-White Perch changed from Detected Decline to Undetected DeclineB. Updated SOC AnalysesParameter values used in the updated SOC analyses are listed in Table 24. Allparameter values except EMR and IMR (which remain the values that were used in theFSEIS) were computed using the same data files used for the updated trends analyses.

Results from the SOC Monte Carlo analyses, and corresponding SOCconclusions, are summarized in Table 25. In comparison to the conclusions reported inthe FSEIS, results from the updated analyses changed the SOC conclusions for 3 of the13 species analyzed in the FSEIS:-Alewife changed from High to Low-Blueback Herring changed from High to Low-White Perch changed from High to LowC. Updated Impact Conclusions The overall impact conclusions based on the updated analyses (Table 26) weredetermined by combining the trends conclusions and SOC conclusions as described inAppendix H of the FSEIS. In comparison to the conclusions reported in the FSEIS,8 results from the updated analyses changed the subsidiary impact conclusions for 7 of the18 species analyzed in the FSEIS:-Alewife changed from Moderate to Small-Blueback Herring changed from Large to Small-Hogchoker changed from Large to Moderate-Rainbow Smelt changed from Moderate-Large to Moderate-Striped Bass changed from Small to Moderate-Weakfish changed from Moderate to Small-White Perch changed from Large to SmallThe change for rainbow smelt was due to newly available information on the rangecontraction of Atlantic coast rainbow smelt (see Discussion

section, below).IV. Discussion A. NRC's Precautionary Methodology The methods applied by NRC to assess the magnitude of potential aquaticimpacts due to the operation of IP2 and IP3 are highly conservative in that they includeseveral components that tend to lead to conclusions of "Large" impacts.

As discussed below, both the trends in YOY abundance analyses and the SOC analyses contain suchcomponents.

1. SOC MethodsNRC's SOC analyses are based on the comparison of the magnitude ofentrainment (and impingement) mortality rates to the magnitude of interannual variability in YOY abundance.

For the purpose of the SOC analyses, NRC defined themagnitude of entrainment mortality as the difference between:

1) projected population abundance with entrainment and 2) projected population abundance withoutentrainment.

Key among the conservative components of the SOC analyses are:I. Estimates of entrainment mortality rates were based on total annualentrainment in comparison to the number of entrainable organisms found insampling River Segment 4 only, rather than to the entire Hudson Riverpopulation.

Because most entrainable organisms found in sampling RiverSegment 4 are transient, moving with tidal currents into and out of samplingRiver Segment 4, the number of fish in River Segment 4 severelyunderestimated the total number of fish from which those entrained weredrawn. Therefore, entrainment mortality rates were overstated.

9

2. Projected abundance in the absence of entrainment was based on the upperconfidence limit of the estimated historical trend in abundance; whereas,projected abundance with entrainment was based on the estimated trenditself. Therefore, even in the absence of entrainment, the method wouldshow a purported reduction in abundance due to entrainment.

2. Trends Analysis MethodsThe methods applied by NRC to assess trends in abundance also contained conservative components.

For each species, the trends assessment included a linearregression analysis and segmented regression analysis.

If the residual error from thesegmented regression was lower than the residual error from the linear regression

analysis, NRC selected the segmented regression analysis for the trends assessment.

Because the segmented regression included four parameters (compared to twoparameters for the linear regression) it was able to fit the data more closely and therefore was often selected over the linear regression.

The segmented regression analysis resulted in two connected line segments beingfit to the data, each with an estimated

duration, and each with an estimated slope. Ifeither slope was negative (and statistically significant)

NRC's method was to conclude adetected decline in abundance, regardless of the duration.

For example, a short-term decline followed by a long term increase would be recorded as a detected decline.3. Conservative ResultsFor the reasons discussed above, the results from analyses conducted usingNRC's methods can be viewed as being highly conservative.

Therefore, even allowingfor inherent uncertainties, the results from the updated analyses support the conclusion that the continued operation of IP2 and IP3 will not pose any meaningful risks ofadverse impacts to fish populations in the Hudson River.B. FSS Gear ChangeAs noted above, in addition to using datasets with a consistent set of weeks ofsampling in all years, this analysis update also used data from the 27-year period 1985-2011, whereas the FSEIS used data from the 27-year period 1979-2005.

Data from theperiod 1985-2011 do not suffer from the potential confounding effects of the FSS gearchange in 1985, which may be substantial.

The FSS gear for sampling the bottomstratum changed from the epibenthic sled with a 1 m2 mouth opening and 3 mm meshcollection net to the beam trawl with a 2.7 m2 mouth opening and 1.3 cm meshcollection net (i.e., 13 mm mesh). Those changes in gear specifications materially altered the collection efficiency of samples from the FSS, which necessarily affectedestimates of CPUE and density.10 In 1984, a gear comparison study was conducted that deployed over 250 pairedepibenthic sled and beam trawl samples in the Tappan Zee, Croton-Haverstraw, andIndian Point regions of the Hudson River, during four alternate weeks of sampling inAugust and September (Normandeau Associates, Inc. 1986). Density estimates forstriped bass young of the year ("YOY") were 4 times higher for the beam trawl than forthe epibenthic sled, and the density estimates for YOY striped bass were higher for thebeam trawl in all four weeks sampled.

Density estimates for bay anchovy YOY were 46times higher for the epibenthic sled, and the density estimates for YOY bay anchovywere higher for the epibenthic sled in all four weeks sampled.

Comparisons for otherspecies were not presented in the report.The 1984 gear comparison study clearly demonstrated that the beam trawl andepibenthic sled had materially different collection efficiencies, and that the differences were species-specific.

For that reason, data collected by the two gear types are notdirectly comparable and cannot be used together in a valid trends assessment withoutaccounting for the species-specific differences in collection efficiencies.

NRC addressed the FSS gear change by conducting a series of statistical analysisthat compared FSS densities to BSS densities before and after 1985, and FSS CPUE toBSS CPUE before and after 1985. Based on those analyses of densities in RiverSegment 4, NRC concluded that for the 12 species considered, the gear change onlycaused a biological difference to bay anchovy.

For CPUE in River Segment 4, NRCconcluded that for the 11 species considered, the gear change only caused a biological difference to bay anchovy and blueback herring.

For riverwide CPUE, NRC concluded that for the 10 species considered, the gear change caused a biological difference toalewife, American shad, bay anchovy, blueback

herring, and bluefish.

For these 8 out of 33 combinations of species and abundance

indices, NRCconducted separate trends analyses for the period of year 1979-1984 and the period ofyears 1985-2005.

For all other combinations of species (including striped bass) andabundance

indices, NRC made no adjustments for the gear change. Furthermore, for thetrends component of the SOC analyses (based on density estimates in River Segment 4),no adjustments for the gear change were made for any species.Although NRC made efforts to address the FSS gear change, the resultspresented in the FSEIS still contain uncertainties due to the FSS gear change thatoccurred in 1985. Because the updated analyses were based on a 27 years of data thatwere not affected by any gear changes, the results from the updated analyses do notcontain that layer of uncertainty.

II C. Summary of Changes in Impact Conclusions As noted above, impact conclusions from the updated analyses differed from theimpact conclusions from the FSEIS for seven species:

alewife, blueback herring,hogchoker, rainbow smelt, striped bass, weakfish and white perch. Each change isdiscussed below.1. Hogchoker, Weakfish and White PerchFor three of these seven species the change was to a lower potential impact leveldue to revised trends conclusions:

Species FSEIS Updated AnalysesTrends Conclusion Trends Conclusion Hogchoker Detected Decline VariableWeakfish Variable Undetected DeclineWhite Perch Detected Decline Undetected DeclineThese changes in the trends conclusions were largely due to changes in Riverwide Assessment Scores:Species Riverwide Assessment River Segment 4 Assessment Score ScoreFSEIS Updated FSEIS Updated(Table H- 15) Analyses (Table H- 15) Analyses(Table 23) (Table 23)Hogchoker 3.0 1.0 4.0 4.0Weakfish 2.5 1.0 2.5 2.5White Perch 4.0 1.0 3.0 2.0This pattern of changes is consistent with the expected confounding effects of theinadvertent inclusion of all weeks of sampling in the FSEIS Riverwide Assessment.

2. Striped BassFor Striped Bass, the change in impact conclusion was to a higher level ofpotential impact (i.e., "Small" to "Variable").

This change was due to changes in boththe Riverwide and River Segment 4 Assessments:

12 Species Riverwide Assessment River Segment 4 Assessment Score ScoreFSEIS Updated FSEIS Updated(Table H-15) Analyses (Table H-15) Analyses(Table 23) (Table 23)Striped Bass 1.0 2.0 1.0 3.0Because the updated analyses were based on a more recent period of years thanthe FSEIS analyses, these changes in trend scores reflect the recent decline in stripedbass stock abundance that followed a period of abundance increases.

Beginning in themid-i 980's the Atlantic coast striped bass stock experienced a surge in abundance inresponse to reduced fishing pressure due to a coastwide fishing moratorium on stripedbass. After the moratorium was lifted in 1990, the stock continued to increase inabundance through the late 1990's after which it began to decline (Atlantic StatesMarine Fisheries Commission, 2013).3. Alewife and Blueback HerringFor alewife and blueback

herring, the change in impact conclusion from "Large"to "Small" is due to the change in the SOC conclusion from "High" (from the FSEIS) to"Low". The "Low" SOC conclusion from the updated analyses for alewife andblueback herring is consistent with the historical distribution patterns of entrainable lifestages of river herring (i.e., collectively alewife and blueback herring).

The vastmajority of entrainable lifestages of river herring inhabit portions of the Hudson Riverthat are far upstream of IP2 and IP3 (Figure 7). The documented distribution patterns ofentrainable lifestages of river herring in the Hudson River, in comparison to the locationof IP2 and IP3, are consistent with the "Low" SOC conclusion from the updatedanalyses.

The recent conclusion of the New York State Department of Environmental Conservation

("NYSDEC")

that recruitment of river herring is variable but stable,despite the upsurge in the use of river herring as bait for striped bass (NYSDEC, 2011)is consistent with a finding of "Small" potential impacts as well. NYSDEC proposed tomaintain the Hudson River and tributaries as a restricted river herring fishery because,under current conditions (including the operation of IP2 and IP3) the fishery was"sustainable" and would "not diminish potential future reproduction and recruitment ofherring stocks."

NYSDEC also noted that since the mid-1990's there has been anincreasing trend in YOY alewife abundance.

In addition, in the National Marine Fishery Service ("NMFS")

decision not to listblueback herring as threatened or endangered (Fed. Reg. Vol. 78, No. 155. August 12,2013), NMFS concluded that water withdrawals and outfalls (including pumped storage,13 irrigation, thermal discharges, industrial pollutants and atmospheric deposition) collectively posed only a "medium low" threat to blueback herring.

The number onethreat was listed as "dams and other barriers".

Behind that, "climate change,"

"waterquality (chemical)",

"incidental catch", and "predation",

ranked as medium threats.

TheNMFS's findings are consistent with the change in impact conclusion for bluebackherring of "Large" to "Small" for IP2 and IP3.4. Rainbow SmeltFor Rainbow Smelt, NRC modified the conclusion of a "Moderate" impact, thatwas determined using the impact assessment methodology of the FSEIS, to a conclusion of "Moderate to Large":"Although detectable population declines occurred in two of four riverdata sets, indicating population trend results were variable, the staffconcluded that a MODERATE to LARGE, rather than just MODERATE, impact was present based on the dramatic population declines observedfor this species over the past three decades."

(FSEIS Section 4.1.3.3,page 4-24)This position regarding rainbow smelt is not supported by recent evidenceregarding large-scale changes in the distribution of rainbow smelt. The decline inabundance of rainbow smelt in the Hudson River has been due to a coastwide contraction of the range of rainbow smelt on the Atlantic coast. Several decades ago,rainbow smelt populations were found as far south as Chesapeake Bay. Now their rangeonly includes waters north of Long Island Sound (Enterline and Chase, 2012; NationalOceanic and Atmospheric Administration, 2010).The decline in rainbow smelt abundance in the Hudson River occurredsimultaneously with the decline in abundance in coastal streams in Connecticut, whichsupports the conclusion that the decline was not due to the operation of IP2 and IP3.Because rainbow smelt is a cold water species, the cause of its range contraction may berelated to global warming."The Hudson River population of rainbow smelt is at the southernextreme of the reproductive range (Lee et al. 1980), although historically it occurred farther south (Smith 1985). The abrupt decline in rainbowsmelt early life stages in the ichthyoplankton may result from globalwarming.

Ashizawa and Cole (1994) documented the trend of slowlyincreasing water temperature in the Hudson River. The rainbow smeltruns in the coastal streams of western Connecticut have drastically declined or disappeared simultaneously with the decline in the HudsonRiver population (S. Gephard, Connecticut Department of Environmental Protection, personal communication)."

(Daniels, et al, 2005)14 For these reasons, the impact conclusion for rainbow smelt from the updatedanalyses was kept at "Moderate",

based on the results from applying the trends analysisand SOC methodologies of the FSEIS to the updated input data files.V. Literature CitedAtlantic States Marine Fisheries Commission.

2013. Striped bass stock assessment for2013; 57th SAW Assessment Report.Daniels, R.A., K.E. Limburg, R.E. Schmidt, D.L. Strayer and R.C. Chambers.

2005.Changes in fish assemblages in the tidal Hudson River, New York. AmericanFisheries Society Symposium 45:471-503.

Draper, N.R. and H. Smith. 1966. Applied Regression Analysis.

John Wiley & Sons,Inc. New York. 407 pp.Enterline C. and B. Chase. 2012. Recent range changes by rainbow smelt (Osmerusmordax) and current annual migrations.

Maine Department of MarineResources, Orno ME.National Oceanic and Atmospheric Administration.

2010. Rainbow Smelt AnImperiled Fish in a Changing World. Support provided by the MaineDepartment of Marine Resources, Massachusetts Division of Marine Fisheries, New Hampshire Department of Fish and Game, and National Oceanic andAtmospheric Administration.

Normandeau Associates.

Inc. 1986. Size selectivity and relative catch efficiency of a3m beam trawl and a I m2 epibenthic sled for sampling young of the year stripedbass and other fishes in the Hudson River estuary.NYSDEC. 2011. Sustainable Fishing Plan for New York River Herring Stocks.15 VI. Figures16 Figure 1.WEEK49-4541-37-3329-2521-17139-5-1-Number of Samples Collected per Week(proportionate to drde size)program = BSS samplinggear

= Beach Seine location

= Riverwide 000 O000000000 00 0oo 000OO0000c00000000 000 0 0000000000000000C

ý-"ý oo-"0 ,--, -00 /-0000,-O000000000C O00 0 0000 .000000000 000000 00OO000000C O000000 0000 0000000000C O 000 0000 000. 0000 O:O 000OO.OO 0.C" 00 000 00 000 000 0 O0OO00000.C O00 00 0 000 000 0 00OO 000O 00 "00 0 000 0000 O0000000O0C 1 111 111 1 11 11 11 1 11 111 1 11 1 11222222222222 99999999999999999999999999000000.000000 7777778888888888999999999900000000001 14 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 .5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1YEAR OF DATA COLLECTION 17 Figure 2.Number of Samples Collected per Week(proportionate to circle size)program=FSS sampling_gar=

Beam Trawl location=

Riverwide WEEK49-1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 2 299999999999999999999999999000000000000 7777778888888888999999999900000000001 145678901 2345678901 2345678901 2345678901 YEAR OF DATA COLLECTION 18 Figure 3.Number of Samples Collected per Week(proportionate to dirde size)program=

FSS samplinggear=

Epibenthic Sled location

=Riverwide WEEK49111111111111111111111111112222222222222 99999999999999999999999999000000000000 7777778888888888999999999900000000001 1456789012345678901 23456789012345678901 YEAR OF DATA COLLECTION 19 Figure 4.WEEKNumber of Samples Collected per Week(proportinae to circle size)program = FSS samplinggear

= Tucker Trawl location

Riverwide 1 1 1 11 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 222222222222 99999999999999999999999999000000000000 77777788888888 889999 9999 9900000 000001 145678901 2345678901 2345678901 2345678901 YEAR OF DATA COLLECTION 20 Figure 5.WEEK49-4541-37-33-29-2521-17-1395-1-Number of Samples Collected per Week(:oportiadate to drdve sze)program = LRS samplinggear

Epibenthic Sled locaton = Riverwide 19741 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 2 2999 999 9999999999999 9999990 000000000 0 077 7778888888

,88899999999 9 90000000

,0001 15678901 2345678901 2345678901 23456 78901YEAR OF DATA COLLECTION 21 Figure 6.Number of Samples Collected per Week(poportinae to drde size)program = LRS samplinggear

= Tucker Trawl locatlon

= Riverwide WEEK49- 0 041 00 0 0 00 0 a0 000 0O00O00O0QO00O 37 00 0 0000000000000

.0 0 0 0 0 0 0 0 0 0 0 0 0 0 C0 0 0 0 000 0 0 000 0 0 0 0 0 0 0 0 C33 000000 00 oo° 0-0o o 000 00 000 0025-21 w 11113 0 0 0 00 0013 o00 0060 0 00 00 0 0 00000 0000Q000 0 0 0 09- 0 0 0 0 00 05I I l I I I I I I 'l I-- , 1 1 1 11 , 1 1I IT 1 1 1 1 1 1' 'I 1 1 1 119741975197619771978197919801111111199.999999 888888881234567819891990199119921993111119999999999456781999200020012002200320042005222222000000000011678901YEAR OF DATA COLLECTION 22 Figure 7. Spatial distribution of early lifestages of river herring (Blueback herring andAlewife) in theHudson River based on LRS sampling (copy of Figure 4-46, 2011 Year ClassReport for the Hudson River Estuary Monitoring Program).

100%80%60%40%20%0%100%80%60%40%20%0%100%80%60%40%20%0%100%80%60%40%20%0%Eggs-iL201i-F]= A A = .i'%'.1YK TZ CH IP WP CW PK HP KG SG CS ALYolk Sac LarvaeT -r 2 0 1 1--,.1974_2010 YK TZ CH tP WP CW PK HP KG SG CS ALPost Yolk Sac LarvaeT- 12011---1974-2010 YK TZ CH IP WP CW PK HP KG SG CS ALYoung-of-Year 2011No young-of-year Alosa spp. were collected during the -,-172010 temporal limits (weeks 18 -26) of this index in 2011.YK TZ CH 1P WP CW PK HP KG SG CS ALRegion23 VII. Tables24 Table 1. Weeks, sampling gears and lifestages included in NRC FSEIS Trends Analyses (27 years: 1979-2005).

Annual abundance indices confounded by inter-annual changes in sampling designs.River Segment 4 Density and CPUE Riverwide CPUEBSS FSS LRS BSS FSS LRSLifestage YOY YOY YOY All All AllWeeks 22-43 27-43 20-40

• All All AllGears Beach Seine Tucker Trawl (1979-2005)

Tucker Trawl Beach Seine Tucker Trawl (1979-2005)

Tucker TrawlEpibenthic Sled (1979-1984)

Epibenthic Sled Epibenthic Sled (1979-1984)

Epibenthic SledBeam Trawl (1985-2005)

, Beam Trawl (1985-2005)

• Inferred from FSEIS Atlantic Tomcod indices of abundance.

Table 2. Weeks, sampling gears and lifestages included in Trends Analyses Update (27 years: 1985-2011).

Consistent set ofsampling conditions among years.River Segment 4 Density and CPUE RiverwideCPUE BSS FSS LRS BSS FSS LRSLifestage YOY YOY YOY YOY YOY YOYWeeks 28-42 29-42 17-27 28-42 29-42 17-27Gears Beach Seine Tucker Trawl Tucker Trawl Beach Seine Tucker Trawl Tucker TrawlBeam Trawl Epibenthic Sled Beam Trawl Epibenthic Sled25 Table 3. Competing Models Used To Characterize the Standardized River Segment 4 FSS Population Trends of YOY Fish Density Using a 3-Year Moving Average (updated FSEIS Table 1-9).Species Linear Regression Segmented Regression MSE Slope Std Err p-value MSE Slope 1 Join Slope 2of Slope PointEstimate Lower Upper Lower Upper95% CL 95% CL 95% CL 95% CLAlewife 0.877 -0.054 0.026 0.048 0.867 -0.827 1.912 1989 -0.135 -0.005American Shad 0.232 -0.120 0.013 0.000 0.186 -2.198 0.339 1988 -0.139 -0.083Atlantic Tomcod 0.678 -0.080 0.023 0.002 0.370 -1.075 -0.275 1991 -0.069 0.029ay Anchovy 0.601 -0.088 0.022 0.000 0.527 -1.453 2.818 1989 -0.158 -0.063Blueback Herring 0.878 -0.054 0.026 0.048 0.081 -5.254 -3.578 1988 -0.027 0.010Bluefish 0.925 -0.046 0.027 0.100 0.591 0.079 1.508 1990 -0.160 -0.044Hogchoker 0.434 -0.104 0.018 0.000 Failed to ConvergeRainbow Smelt 0.623 -0.086 0.022 0.001 0.535 -0.193 0.534 1992 -0.188 -0.060Striped Bass 0.776 -0.069 0.024 0.010 0.684 -0.720 4.145 1988 -0.144 -0.036Weakfish 0.811 -0.064 0.025 0.017 0.459 -0.030 0.157 2001 -0.408 -0.139White Catfish 0.945 -0.042 0.027 0.136 0.967 -0.257 0.374 1995 -0.181 0.022White Perch 0.838 -0.060 0.025 0.026 0.656 -0.542 -0.023 1995 -0.062 0.105Table 4. River Segment 4 Assessment of the Level of Potential Negative Impact Based on the Standardized FSS Density Using a 3-Year Moving Average (updated FSEIS Table 1-10).Species Best Fit Slope Slope 1 Slope 2 Finalfrom from from DecisionLinear Segmented Segmented Regression Regression Regression Alewife SR S1=0 S2<0 4American Shad SR S1=0 S2<0 4Atlantic Tomcod SR S1<0 S2=0 4Bay Anchovy SR S1=0 S2<0 4Blueback Herring SR SI<0 S2=0 4Bluefish SR S1>0 S2<0 4Hogchoker LR S<0 4Rainbow Smelt SR S1=0 S2<0 4Striped Bass SR S1=0 S2<0 4Weakfish SR S1=0 S2<0 4White Catfish LR S=0 IWhite Perch SR S1<0 S2=0 426 Table 5. Competing Models Used To Characterize the Standardized River Segment 4 BSS Population Trends of YOY Fish Density Using a 3-Year Moving Average (updated FSEIS Table 1-12).Species Linear Regression Segmented Regression MSE Slope Std Err p-value MSE Slope 1 Join Slope 2of Slope PointEstimate Lower Upper Lower Upper95% CL 95% CL 95% CL 95% CLAlewife 0.725 0.075 0.024 0.004 0.698 -0.088 0.120 2002 -0.007 0.376American Shad 0.235 -0.120 0.013 0.000 0.252 -0.149 -0.06 1 2005 -0.426 0.074Bay Anchovy 0.844 0.059 0.025 0.029 Failed to ConvergeBlueback Herring 0.726 -0.075 0.024 0.004 0,665 -0.154 0.369 1994 -0.211 -0.043Bluefish 1.034 0.013 0.028 0.646 0.915 -0.355 0.083 1997 -0.014 0.224Hogchoker 0.776 0.069 0.024 0.010 0.331 -0.251 -0.023 1998 0.152 0.310Spottail Shiner 0.989 -0.031 0.028 0.271 0.932 -0.556 2.283 1989 -0.123 0.012Striped Bass 1.016 0.022 0.028 0.436 0.454 0.107 0.342 1999 -0.284 -0.076White Perch 1.035 -0.012 0.028 0.663 0.873 -1.114 0.115 1991 -0.042 0.109Table 6. River Segment 4 Assessment of the Level of Potential Negative Impact Based on the Standardized BSS Density Using a 3-Year Moving Average (updated FSEIS Table 1-13).Species Best Fit Slope Slope 1 Slope 2 Finalfrom from from DecisionLinear Segmented Segmented Regression Regression Regression Alewife SR SI=0 S2=0 1American Shad LR S<0 4Bay Anchovy LR S>0 IBlueback Herring SR S1=0 S2<0 4Bluefish SR S1=0 S2=0 1Hogchoker SR S1<0 S2>0 4Spottail Shiner SR SI=0 S2=0 1Striped Bass SR SI>0 S2<0 4White Perch SR S1 =0 S2=0 127 Table 7. Competing Models Used To Characterize the Standardized River Segment 4 LRS Population Trends of YOY Atlantic Tomcod Density Using a 3-Year Moving Average (updated FSEIS Table I-15).Species Linear Regression Segmented Regression MSE Slope Std Err p-value MSE Slope I Join Slope 2of Slope PointEstimate Lower Upper Lower Upper95% CL 95% CL 95% CL 95% CLAtlantic Tomcod 1.030 -0.015 0.028 0.590 0.471 -2.721 -0.702 1989 -0.007 0.089Table 8. River Segment 4 Assessment of the Level of Potential Negative Impact Based on the Standardized LRS Atlantic Tomcod YOY Density Using a 3-Year Moving Average (updated FSEIS Table 1-16).Species Best Fit Slope Slope I Slope 2 Finalfrom from from DecisionLinear Segmented Segmented Regression Regression Regression Atlantic Tomcod SR S1<0 S2=0 428 Table 9. Competing Models Used To Characterize the Standardized River Segment 4, FSS Population Trends of YOY Fish CPUE (updated FSEIS Table 1-19).Species Linear Regression Segmented Regression MSE Slope Std Err p-value MSE Slope 1 Join Slope 2of Slope PointEstimate Lower Upper Lower Upper95% CL 95% CL 95% CL 95% CLAlewife 0.955 -0.036 0.024 0.149 -0.097 0.009 2010American Shad 0.753 -0.066 Q.021 0.005 0.677 -0.106 0.179 1996 -0.234 -0.030Atlantic Tomcod 0.791 -0.062 0.022 0.010 Failed to ConvergeBay Anchovy 1.039 0.003 0.025 0.905 0.880 -0.030 0.227 1999 -0.265 0.023Blueback Herring 0.936 -0.040 0.024 0.108 0.596 -2.448 -0.190 1987 -0.045 0.050Bluefish 0.861 -0.052 0.023 0.031 0.848 -0.099 0.090 2002 -0.371 0.049Hogchoker 0.832 -0.056 0.023 0.019 0.805 -1.988 3.263 1987 -0.125 -0.022Rainbow Smelt 0.839 -0.055 0.023 0.022 0.837 -0.265 0.451 1992 -0.163 -0.016Striped Bass 0.952 -0.037 0.024 0.141 0.895 -0.560 2.207 1987 -0.115 0.001Weakfish 1.014 -0.020 0.025 0.432 0.968 -0.091 0.130 2000 -0.296 0.092White Perch 0.948 -0.038 0.024 0.132 0.943 -0.318 0.065 1996 -0.091 0.127Table 10. River Segment 4 Assessment of the Level of Potential Negative Impact Based on the Standardized FSS CPUE (updated FSEIS Table 1-20).Species Best Fit Slope Slope 1 Slope 2 Finalfrom from from DecisionLinear Segmented Segmented Regression Regression Regression Alewife LR S=0 IAmerican Shad SR S1=0 S2<0 4Atlantic Tomcod LR S<0 4Bay Anchovy SR S1=0 S2=0 1Blueback Herring SR S1<0 S2=0 4Bluefish SR SI=0 S2=0 1Hogchoker SR S1=0 S2<0 4Rainbow Smelt SR S1=0 S2<0 4Striped Bass SR S1--0 S2=0 IWeakfish SR S1=0 S2=0 1White Perch SR SI=0 S2=0 I29 Table 11. Competing Models Used To Characterize the Standardized River Segment 4 LRS Population Trends of YOY Atlantic Tomcod CPUE Using a 3-Year Moving Average (updated FSEIS Table I-22).Species Linear Regression Segmented Regression MSE Slope Std Err p-value MSE Slope 1 Join Slope 2of Slope PointEstimate Lower Upper Lower Upper95% CL 95% CL 95% CL 95% CLtlantic Tomcod 1.012 -0.021 0.025 0.410 0.842 -1.609 0.089 1988 -0.044 0.076Table 12. River Segment 4 Assessment of the Level of Potential Negative Impact Based 7 on theStandardized LRS Atlantic Tomcod YOY CPUE Using a 3-Year Moving Average (updated FSEISTable 1-23).Species Best Fit Slope Slope I Slope 2 Finalfrom from from DecisionLinear Segmented Segmented Regression Regression Regression Atlantic Tomcod SR I S1=0 S2=0 130 Table 13. Assessment of Population Impacts for IP2 and IP3 River Segment 4 (updated FSEIS Table 1-24).Species Density CPUE RiverSegmentFSS BSS LRS FSS LRS Assessment Alewife 4 1 N/A 1 N/A 2.0American Shad 4 4 N/A 4 N/A 4.0Atlantic Menhaden N/A N/A N/A N/A N/A UnknownAtlantic Sturgeon N/A N/A N/A N/A N/A UnknownAtlantic Tomcod 4 N/A 4 4 1 3.3Bay Anchovy 4 1 N/A 1 N/A 2.0Blueback Herring 4 4 N/A 4 N/A 4.0Bluefish 4 1 N/A 1 N/A 2.0Gizzard Shad N/A N/A N/A N/A N/A UnknownHogchoker 4 4 N/A 4 N/A 4.0Rainbow Smelt 4 N/A N/A 4 N/A 4.0Shortnose Sturgeon N/A N/A N/A N/A N/A UnknownSpottail Shiner N/A 1 N/A N/A N/A 1.0Striped Bass 4 4 N/A 1 N/A 3.0Weakfish 4 N/A N/A 1 N/A 2.5White Catfish 1 N/A N/A N/A N/A 1.0White Perch 4 1 N/A 1 N/A 2.0Blue Crab N/A N/A N/A N/A N/A Unknown31 Table 14. Competing Models Used To Characterize the Standardized Riverwide FSS Population Trends ofYOY Fish CPUE (updated FSEIS Table 1-27).Species Linear Regression Segmented Regression MSE Slope Std Err p-value MSE Slope 1 Join Slope 2of Slope PointEstimate Lower Upper Lower Upper95% CL 95% CL 95% CL 95% CLAlewife 1.006 0.023 0.025 0.370 1.011 -0.038 0.188 2000 -0.266 0.130American Shad 0.553 -0.086 0.018 0.000 0.556 -0.380 0.596 1989 -0.151 -0.048Atlantic Tomcod 0.725 -0.069 0.021 0.003 0.768 -0.414 0.146 1993 -0.125 0.026Bay Anchovy 1.036 0.008 0.025 0.763 0.902 -0.027 0.348 1995 -0.175 0.038Blueback Herring 0.701 -0.072 0.021 0.002 0.746 -0.394 0.460 1991 -0.154 -0.025Bluefish 0.822 -0.058 0.022 0.016 0.828 -0.121 0.104 1999 -0.281 0.034Hogchoker 0.921 -0.043 0.024 0.084 0.912 -0.222 0.014 1999 -0.126 0.204Spottail Shiner 0.827 -0.057 0.022 0.018 0.880 -0.174 0.019 2002 -0.218 0.209Striped Bass 0.833 -0.056 0.023 0.020 0.614 0.088 2.381 1987 -0.135 -0.039White Perch 1.004 -0.023 0.025 0.352 0.988 -0.479 0.155 1993 -0.066 0.106Table 15. Riverwide Assessment of the Level of Potential Negative Impact Based on the Standardized FSSCPUE (updated FSEIS Table 1-28).Species Best Fit Slope Slope 1 Slope 2 Finalfrom from from DecisionLinear Segmented Segmented Regression Regression Regression Alewife LR S=0 IAmerican Shad LR S<0 4Atlantic Tomcod LR S<0 4Bay Anchovy SR S1=O S2=0 1Blueback Herring LR S<0 4Bluefish LR S<0 4Hogchoker SR S1=0 S2=0 ISpottail Shiner LR S<0 4Striped Bass SR SI>0 S2<0 4White Perch SR S 1=0 S2=0 I32 Table 16. Competing Models Used To Characterize the Standardized Riverwide BSS Population Trends ofYOY Fish CPUE (updated FSEIS Table 1-30).Species Linear Regression Segmented Regression MSE Slope Std Err p-value MSE Slope I Join Slope 2of Slope PointEstimate Lower Upper Lower Upper95% CL 95% CL 95% CL 95% CLAlewife 0.744 0.067 0.021 0.004 0.726 -0.054 0.107 2002 -0.053 0.402American Shad 0.551 -0.086 0.018 0.000 0.554 -0.285 0.451 1990 -0.162 -0.051Atlantic Tomcod 0.543 -0.087 0.018 0.000 0.341 -1.704 0.004 1988 -0.087 -0.016Bay Anchovy 0.8 13 0.059 0.022 0.014 0.607 -0.046 0.062 2006 -0.026 0.994Blueback Herring 1.036 -0.008 0.025 0.760 1.072 -0.515 0.839 1990 -0.101 0.043luefish 1.040 0.003 0.025 0.919 1.073 -0.076 0.180 1999 -0.240 0.119ogchoker 1.034 0.010 0.025 0.695 1.068 -0.204 0.113 1998 -0.076 0.208ainbow Smelt 0.972 -0.032 0.024 0.199 1.002 -0.269 0.370 1993 -0.139 0.034Spottail Shiner 0.743 0.067 0.021 0.004 0.805 -0.124 0.230 1996 -0.026 0.176Striped Bass 1.020 0.018 0.025 0.488 0.932 -0.017 0.127 2005 -0.704 0.251Weakfish 0.918 -0.043 0.024 0.081 1986 -0.055 0.024White Catfish 1.034 -0.010 0.025 0.699 1.010 -1.315 0.545 1989 -0.047 0.083White Perch 1.033 -0.010 0.025 0.691 1.015 -0.367 0.092 1994 -0.063 0.144Table 17. Riverwide Assessment of the Level of Potential Negative Impact Based on the BSS CPUE(updated FSEIS Table 1-31).Species Best Fit Slope Slope I Slope 2 Finalfrom from from DecisionLinear Segmented Segmented Regression Regression Regression Alewife SR S1=0 S2=0 1American Shad LR S<0 4Atlantic Tomcod SR S1=0 S2<0 4Bay Anchovy SR S1=0 S2=0 IBlueback Herring LR S=0 IBluefish LR S=0 1Hogchoker LR S=0 1Rainbow Smelt LR S=0 ISpottail Shiner LR S>0 1Striped Bass SR SI=0 S2=0 IWeakfish LR S=0 IWhite Catfish SR S1=0 S2=0 IWhite Perch SR S1=0 S2=0 I33 Table 18. Competing Models Used To Characterize the Standardized Riverwide LRS Population Trend ofYOY Atlantic Tomcod CPUE (updated FSEIS Table 1-33).Species Linear Regression Segmented Regression MSE Slope Std Err p-value MSE Slope 1 Join Slope 2of Slope PointEstimate Lower Upper Lower Upper95% CL 95% CL 95% CL 95% CLAtlantic Tomcod 0.938 -0.039 0.024 0.112 -0.089 0.010 2016Table 19. Riverwide Assessment of the Level of Potential Negative Impact Based on the Standardized LRSCPUE of Atlantic Tomcod (updated FSEIS Table 1-34).Species Best Fit Slope Slope I Slope 2 Finalfrom from from DecisionLinear Segmented Segmented Regression Regression Regression Atlantic Tomcod LR S=0 134 Table 20. Competing Models Used To Characterize the Standardized Riverwide YOY Abundance IndexTrends (updated FSEIS Table 1-36).Species Linear Regression Segmented Regression MSE Slope Std Err p-value MSE Slope 1 Join Slope 2of Slope PointEstimate Lower Upper Lower Upper95% CL 95% CL 95% CL 95% CLAlewife 1.017 0.019 0.025 0.458 1American Shad 0.596 -0.082 0.019 0.000 0.594 -0.588 0.838 1989 -0.150 -0.050Atlantic Tomcod 0.576 -0.084 0.019 0.000 0.547 -1.637 2.690 1987 -0.141 -0.056Bay Anchovy 0.744 -0.067 0.021 0.004 0.786 -0.194 0.005 2000 -0.198 0.152Blueback Herring 0.792 -0.062 0.022 0.010 0.794 -0.154 -0.021 2006 -0.300 0.581Bluefish 1.035 0.009 0.025 0.731 0.967 -0.038 0.205 1999 -0.259 0.081Hogchoker 0.902 -0.046 0.023 0.062 0.942 -0.165 0.017 2003 -0.219 0.299Rainbow Smelt 0.971 -0.033 0.024 0.193 0.960 -0.216 0.409 1992 -0.148 0.022Spottail Shiner 0.844 0.055 0.023 0.024 0.879 -0.008 0.168 2003 -0.273 0.227Striped Bass 1.039 0.005 0.025 0.855 0.925 -0.024 0.131 2005 -0.657 0.095Weakfish 0.647 -0.077 0.020 0.001 0.576 -0.561 0.032 1992 -0.095 0.027White Catfish 0.833 -0.056 0.023 0.020 0.863 -0.198 0.011 2001 -0.173 0.193White Perch 1.039 -0.003 0.025 0.906 1.093 -0.079 0.103 2003 -0.424 0.243Table 21. Riverwide Assessment of the Level of Potential Negative Impact Based in the Abundance Index(updated FSEIS Table 1-37).Species Best Fit Slope Slope 1 Slope 2 Finalfrom from from DecisionLinear Segmented Segmented Regression Regression Regression Alewife LR S=0 IAmerican Shad SR S 1O S2<0 4Atlantic Tomcod SR S1=0 S2<0 4Bay Anchovy LR S<0 4Blueback Herring LR S<0 4Bluefish SR SI=0 S2=0 IHogchoker LR S=0 IRainbow Smelt SR S1=0 S2=0 ISpottail Shiner LR S>0 1Striped Bass SR S1=0 S2=0 IWeakfish SR S1=0 S2=0 IWhite Catfish LR S<0 4White Perch LR S=0 I35 Table 22. Assessment of Riverwide Population Impacts (updated FSEIS Table 1-38).Species CPUE Abundance Riverwide Index Assessment FSS BSS LRSAlewife 1 1 N/A 1 1.0American Shad 4 4 N/A 4 4.0Atlantic Menhaden N/A N/A N/A N/A UnknownAtlantic Sturgeon N/A N/A N/A N/A UnknownAtlantic Tomcod 4 4 1 4 3.3Bay Anchovy 1 1 N/A 4 2.0Blueback Herring 4 1 N/A 4 3.0Bluefish 4 1 N/A 1 2.0Gizzard Shad N/A N/A N/A N/A UnknownHogchoker I I N/A 1 1.0Rainbow Smelt N/A 1 N/A 1 1.0Shortnose Sturgeon N/A N/A N/A N/A UnknownSpottail Shiner 4 1 N/A 1 2.0Striped Bass 4 1 N/A 1 2.0Weakfish N/A I N/A 1 1.0White Catfish N/A I N/A 4 2.5White Perch I I N/A 1 1.0Blue Crab N/A N/A N/A N/A Unknown36 Table 23. Weight of Evidence Results for the Population Trend Line of Evidence (updated FSEIS Table H-15).Species River Riverwide WOE ImpactSegment Assessment Score Conclusion Assessment ScoreScoreAlewife 2.0 1.0 1.6 Undetected DeclineAmerican Shad 4.0 4.0 4.0 Detected DeclineAtlantic Menhaden Unknown Unknown Unknown Unresolved Atlantic Sturgeon Unknown Unknown Unknown Unresolved Atlantic Tomcod 3.3 3.3 3.3 Detected DeclineBay Anchovy 2.0 2.0 2.0 Undetected DeclineBlueback Herring 4.0 3.0 3.6 Detected DeclineBluefish 2.0 2.0 2.0 Undetected DeclineGizzard Shad Unknown Unknown Unknown Unresolved Hogchoker 4.0 1.0 2.8 VariableRainbow Smelt 4.0 1.0 2.8 VariableShortnose Sturgeon Unknown Unknown Unknown Unresolved Spottail Shiner 1.0 2.0 1.4 Undetected DeclineStriped Bass 3.0 2.0 2.6 VariableWeakfish 2.5 1.0 1.9 Undetected DeclineWhite Catfish 1.0 2.5 1.6 Undetected DeclineWhite Perch 2.0 1.0 1.6 Undetected DeclineBlue Crab Unknown Unknown Unknown Unresolved 37 Table 24. Parameter Values Used in the Monte Carlo Simulation (updated FSEIS Table 1-46).RIS Survey Linear Slope Error CV of EMR IMRUsed Slope plus Mean Density(r) Standard Square DataError of from (1985-the Slope Regression 1996)EstimateAlewife BSS 0.075 0.099 0.725 1.294 0.095 0.0020American Shad BSS -0.120 -0.106 0.235 0.510 0.042 0.0005Atlantic Tomcod FSS -0.080 -0.058 0.678 0.794 0.036 0.0300Bay Anchovy FSS -0.088 -0.067 0.601 0.511 0.213 0.0040Blueback Herring BSS -0.075 -0.051 0.726 1.034 0.095 0.0040Bluefish BSS 0.013 0.041 1.034 0.754 0.003 0.0005Hogchoker FSS -0.104 -0.086 0.434 1.225 0.386 0.0005Rainbow Smelt FSS -0.086 -0.064 0.623 1.211 0.258 0.0005Spottail Shiner BSS -0.031 -0.004 0.989 1.182 0.031 0.0070Striped Bass BSS 0.022 0.050 1.016 0.523 0.106 0.0080Weakfish FSS -0.064 -0.039 0.811 0.698 0.544 0.0005White Catfish FSS -0.042 -0.015 0.945 2.566 0.114 0.0005White Perch BSS -0.012 0.016 1.035 1.005 0.076 0.032038 Table 25. Quartiles of the Relative Difference in Cumulative Abundance and Conclusions for the Strength-of-Connection From the Monte Carlo Simulation (updated FSEIS Table 1-47).Taxa Number No= 1000 No= 1 x 108 Strength ofof Median Q1 Q3 Median Q1 Q3 Connection Years Conclusion Alewife 20 -0.07 -1.19 1.03 -0.07 -1.17 1.01 Low27 -0.32 -1.63 1.02 -0.32 -1.69 1.04American Shad 20 0.07 -0.01 0.14 0.07 -0.01 0.14 Low27 0.06 0.00 0.11 0.06 0.00 0.11Atlantic Tomcod 20 0.15 -0.03 0.34 0.16 -0.03 0.35 Low27 0.16 0.01 0.30 0.15 0.01 0.30Bay Anchovy 20 0.29 0.13 0.44 0.29 0.13 0.44 High27 0.27 0.15 0.39 0.27 0.15 0.39Blueback Herring 20 0.21 -0.03 0.46 0.22 -0.02 0.46 Low27 0.22 0.04 0.41 0.23 0.04 0.42Bluefish 20 0.45 -0.09 0.99 0.45 -0.09 0.98 Low27 0.67 0.11 1.21 0.69 0.15 1.23Hogchoker 20 0.58 0.31 0.85 0.57 0.30 0.86 High27 0.56 0.35 0.78 0.56 0.35 0.78Rainbow Smelt 20 0.45 0.16 0.74 0.46 0.16 0.76 High27 0.45 0.23 0.68 0.45 0.23 0.68Spottail Shiner 20 0.27 -0.14 0.68 0.27 -0.13 0.69 Low27 0.34 -0.00 0.69 0.34 0.00 0.68Striped Bass 20 0.84 0.25 1.43 0.83 0.24 1.42 High27 1.27 0.64 1.93 1.28 0.64 1.92Weakfish 20 0.74 0.42 1.07 0.75 0.42 1.07 High27 0.76 0.49 1.02 0.76 0.50 1.02White Catfish 20 0.42 -0.26 1.10 0.44 -0.27 1.13 Low27 0.49 -0.07 1.06 0.47 -0.10 1.06White Perch 20 0.40 -0.08 0.90 0.40 -0.07 0.88 Low1 27 0.51 0.10 0.93 0.51 0.08 0.9339 Table 26. Impingement and Entrainment Impact Summary for Hudson River YOY RIS (updated FSEISTable H- 17).Species Population Trend Strength of Impacts of IP2 andLine of Evidence Connection IP3 Cooling SystemsLine of Evidence on YOY RISAlewife Undetected Decline Low SmallAmerican Shad Detected Decline Low SmallAtlantic Menhaden Unresolved Low(b) SmallAtlantic Sturgeon Unresolved Low(bJ SmallAtlantic Tomcod Detected Decline Low SmallBay Anchovy Undetected Decline High SmallBlueback Herring Detected Decline Low SmallBluefish Undetected Decline Low SmallGizzard Shad Unresolved Low(b) SmallHogchoker Variable High ModerateRainbow Smelt Variable High ModerateShortnose Sturgeon Unresolved Low(b) SmallSpottail Shiner Undetected Decline Low SmallStriped Bass Variable High ModerateWeakfish Undetected Decline High SmallWhite Catfish Undetected Decline Low SmallWhite Perch Undetected Decline Low SmallBlue Crab Unresolved Low(b) Small(b Strength of connection could not be established using Monte Carlo Simulation; therefore, strength of connection was based onthe rate of entrainment and impingement.

40 VIII. Appendix AThe "p-value" is the probability level for the significance test of the estimated slope (i) from thelinear regression.

It is the probability that the absolute value of a random variable from a t-distribution isgreater than the ratio:Pse(i)where se(i) is the standard error of the estimated slope (Draper and Smith, 1966). For Tables 1-9 and 1-12it is a t-distribution with 23 degrees of freedom because the time series of 3-year averages of River Segment4 density estimates contained 25 index values and the linear regression model had 2 parameters.

For each linear slope estimate listed in Table 1-46, the corresponding "p-value" listed in Table 1-9 orTable 1-12 was equal (allowing for round-off errors with 3 significant digits listed in Tables 1-9 and 1-12) tothe probability that the absolute value of a random variable with a t-distribution with 23 degrees of freedomwas greater than the ratio:estimated slopevalue to the right of the +/- symbolThis demonstrates that the undefined value to the right of the +/- symbol in the slope column of Tables 1-9and 1-12 was, in fact, the standard error of the estimated slope.41 IX. Appendix BReport on QC Review of Analysis UpdatePrepared by John Young, PhDASA Analysis

& Communication, Inc.921 Pike Street, PO Box 303Lemont, PA 16851-0303 October 18, 2013ASA reviewed the SAS programs used to analyze clean data files (1985-20 t 1) from the HudsonRiver Biological Monitoring Program with NRC's trend assessment methods.

The first step in the reviewprocess was to create SAS datasets from 1974-2011 level files for BSS, FSS, and LRS programs.

Thesedatasets contained one observation for each of the target species for each sample from each program.

Aseparate dataset was constructed for each program each year. Once the datasets had been created, the nextstep was to run the following series of SAS programs used for the analyses:

1. NRC Region 4 Indices Corrected vl02. NRC Riverwide Indices Corrected v l03. NLIN and REG NRC trends Corrected v134. NRC Trends summary Corrected vi 65. SOC input updated trends results v 16. SOC update vl0Each line of the resulting SAS log files was evaluated for error, warning and other unexpected messages.

The presence of such messages indicates an error is present in the program that may lead toinaccurate results.

All six programs ran successfully and were found to be free of errors. The draft resultspresented in table H-15, H-17, 1-24, and 1-38 were reproduced.

42 After confirming that the programs were running successfully, the methodology applied within theprograms was compared to that used by NRC to ensure that the analysis was accurately reproduced.

Thefollowing steps were taken during the methodology review:1. Reviewed all input and output datasets of the program2. Evaluated sort order of all datasets3. Evaluated all macros4. Evaluated code logicThese steps ultimately tested the accuracy and integrity of the program logic and output which it produces.

The methodology review did not identify deviances from the NRC data analysis methodology.

All inputand output datasets were accurate, sorting was not found to be an issue, macros ran without error, and codelogic mirrored that used by NRC.In summary, the programs used to apply NRC trend assessment methodology to the clean data filesfrom the Hudson River Biological Monitoring Program were found to be free of errors. The resultsproduced by the programs are considered to be accurate.

43 ATTACHMENT 2 TO NL-14-030 78 Fed. Reg. 48944 (August 12, 2013);Sustainable Fishing Plan for New York River Herring Stocks (2011);Rainbow Smelt: An Imperiled Fish in a Changing World (2010);A Regional Conservation Plan for Anadromous Smelt (2012); and,Correspondence from Mark D. Sanza, Assistant Counsel for NYSDEC to ALJsVilla and O'Connell, Administrative Law Judges for NYSDEC, re: EntergyNuclear Indian Point Units 2 and 3, CWA Section 401 WQC Application Proceeding.

ENTERGY NUCLEAR OPERATIONS, INC.INDIAN POINT NUCLEAR GENERATING UNIT NOS. 2 & 3DOCKET NOS. 50-247 AND 50-286 FEDERAL REGISTERVol. 78 Monday,No. 155 August 12, 2013Part IIDepartment of CommerceNational Oceanic and Atmospheric Administration Endangered and Threatened Wildlife and Plants; Endangered Species ActListing Determination for Alewife and Blueback Herring; Notice 48944Federal Register/Vol.

78, No. 155/Monday, August 12, 2013/Notices DEPARTMENT OF COMMERCENational Oceanic and Atmospheric Administration

[Docket No. 111024651-3630-02]

RIN 0648-XA739 Endangered and Threatened Wildlifeand Plants; Endangered Species ActListing Determination for Alewife andBlueback HerringAGENCY: National Marine Fisheries Service (NMFS), National Oceanic andAtmospheric Administration (NOAA),Commerce.

ACTION: Notice of a listingdetermination.

SUMMARY:

We, NMFS, have completed acomprehensive review of the status ofriver herring (alewife and bluebackherring) in response to a petitionsubmitted by the Natural Resources Defense Council (NRDC) requesting thatwe list alewife (Alosa pseudoharengus) and blueback herring (Alosa oestivalis) as threatened under the Endangered Species Act (ESA) throughout all or asignificant portion of their range or asspecific distinct population segments(DPS) identified in the petition.

TheAtlantic States Marine Fisheries Commission (ASMFC) completed acomprehensive stock assessment forriver herring in May 2012 which coversover 50 river specific stocks throughout the range of the species in the UnitedStates. The ASMFC stock assessment contained much of the information necessary to make an ESA listingdetermination for both species;however, any deficiencies wereaddressed through focused workshops and working group meetings and reviewof additional sources of information.

Based on the best scientific andcommercial information available, wehave determined that listing alewife asthreatened or endangered under theESA is not warranted at this time.Additionally, based on the bestscientific and commercial information available, we have determined thatlisting blueback herring as threatened orendangered under the ESA is notwarranted at this time.DATES: This finding is effective onAugust 12, 2013.ADDRESSES:

The listing determination, list of references used in the listingdetermination, and other relatedmaterials regarding this determination can be obtained via the Internet at:http://www.nero.nooo.gov/prot res/CandidateSpeciesProgram/River HerringSOC.htm or by submitting arequest to the Assistant RegionalAdministrator, Protected Resources

Division, Northeast Region, NMFS, 55Great Republic Drive, Gloucester, MA01930.FOR FURTHER INFORMATION CONTACT:

KimDamon-Randall, NMFS Northeast Regional Office, (978) 282-8485; orMarta Nammack, NMFS, Office ofProtected Resources (301) 427-8469.

SUPPLEMENTARY INFORMATION:

Background

On August 5, 2011, we, the NationalMarine Fisheries Service (NMFS),received a petition from the NaturalResources Defense Council (NRDC),requesting that we list alewife (Alosapseudoharengus) and blueback herring(Alosa aestivalis) under the ESA asthreatened throughout all or asignificant portion of their ranges. In thealternative, they requested that wedesignate DPSs of alewife and bluebackherring as specified in the petition(Central New England, Long IslandSound, Chesapeake Bay, and Carolinafor alewives, and Central New England,Long Island Sound, and Chesapeake Bayfor blueback herring).

The petitioncontained information on the twospecies, including the taxonomy, historical and current distribution, physical and biological characteristics of their habitat and ecosystem relationships, population status andtrends, and factors contributing to thespecies' decline.

The petition alsoincluded information regarding potential DPSs of alewife and bluebackherring as described above. Thefollowing five factors identified insection 4(a)(1) of the ESA wereaddressed in the petition:

(1) Present orthreatened destruction, modification, orcurtailment of habitat or range; (2) over-utilization for commercial, recreational, scientific, or educational purposes; (3)disease or predation; (4) inadequacy ofexisting regulatory mechanisms; and (5)other natural or man-made factorsaffecting the species' continued existence.

We reviewed the petition anddetermined that, based on theinformation in the petition and in ourfiles at the time we received thepetition, the petitioned action may bewarranted.

Therefore, we published apositive 90-day finding on November 2,2011, and as a result, we were requiredto review the status of the species (e.g.,anadromous alewife and bluebackherring) to determine if listing under theESA is warranted.

We formed aninternal status review team (SRT)comprised of nine NMFS staff members(Northeast Regional Office (NERO)Protected Resources Division andNortheast Fisheries Science Center staff)to compile the best commercial andscientific data available for alewife andblueback herring throughout theirranges.In May 2012, the ASMFC completed a river herring stock assessment, whichcovers over 50 river-specific stocksthroughout the ranges of the species inthe United States (ASMFC, 2012;hereafter referred to in thisdetermination as "the stockassessment").

In order to avoidduplicating this extensive effort, weworked cooperatively with ASMFC touse this information in the review of thestatus of these two species and identifyinformation not in the stock assessment that was needed for our listingdetermination.

We identified themissing required elements and heldworkshops/working group meetingsfocused on addressing information onstock structure, extinction risk analysis, and climate change.Reports from each workshop/working group meeting were compiled andindependently peer reviewed (the stockstructure and extinction risk reportswere peer reviewed by reviewers selected by the Center for Independent

Experts, and the climate change reportwas peer reviewed by 4 expertsidentified during the workshops).

Thesereports did not contain any listingadvice or reach any ESA listingconclusions-such synthesis andanalysis for river herring is solelywithin the agency's purview.

We usedthis information to determine whichextinction risk method and stockstructure analysis would best inform thelisting determination, as well asunderstand how climate change mayimpact river herring, and ultimately, weare using these reports along with thestock assessment and all other bestavailable information in this listingdetermination.

Alewife and blueback herring arecollectively referred to as "riverherring."

Due to difficulties indistinguishing between the species, theyare often harvested together incommercial and recreational fisheries, and managed together by the ASMFC.Throughout this finding, where thereare similarities, they will be collectively referred to as river herring, and wherethere are distinctions, they will beidentified by species.RangeRiver herring can be found along theAtlantic coast of North America, fromthe Southern Gulf of St. Lawrence, Canada to the southeastern UnitedStates (Mullen et al., 1986; Schultz etal., 2009). The coastal ranges of the two Federal Register/Vol.

78, No. 155/Monday, August 12, 2013 / Notices48945species overlap.

Blueback herring rangefrom Nova Scotia south to the St. John'sRiver, Florida; and alewife range fromLabrador and Newfoundland south toSouth Carolina, though their occurrence in the extreme southern range is lesscommon (Collette and Klein-MacPhee, 2002; ASMFC, 2009a; Kocik et al.,2009).In Canada, river herring (i.e.,gaspereaul are most abundant in theMiramichi,

Margaree, LaHave, Tusket,Shubenacadie and Saint John Rivers(Gaspereau Management Plan, 2001).They are proportionally less abundantin smaller coastal rivers and streams(Gaspereau Management Plan, 2001).Generally, blueback herring in Canadaoccur in fewer rivers than alewives andare less abundant in rivers where bothspecies coexist (DFO 2001).Habitat and Migration River herring are anadromous, meaning that they mature in the marineenvironment and then migrate upcoastal rivers to estuarine andfreshwater rivers, ponds, and lakehabitats to spawn (Collette and Klein-MacPhee, 2002; ASMFC, 2009a; Kociket al., 2009). In general, adult riverherring are most often found at depthsless than 328 feet (ft) (100 meters (m))in waters along the continental shelf(Neves, 1981; ASMFC, 2009a; Schultz etal., 2009). They are highly migratory,

pelagic, schooling

species, withseasonal spawning migrations that arecued by water temperature (Collette andKlein-MacPhee, 2002; Schultz et al.,2009). Depending upon temperature, blueback herring typically spawn fromlate March through mid-May.

However,they spawn in the southern parts oftheir range as early as December orJanuary, and as late as August in thenorthern portion of their range (ASMFC,2009a). Alewives have beendocumented spawning as early asFebruary in the southern portion of theirrange, and as late as August in thenorthern portion of the range (ASMFC,2009a). The river herring migration inCanada extends from late April throughearly July, with the peak occurring inlate May and early June. Bluebackherring generally make their spawningruns about 2 weeks later than alewivesdo (DFO, 2001). River herring conformto a metapopulation paradigm (e.g., agroup of spatially separated populations of the same species which interact atsome level) with adults frequently returning to their natal rivers forspawning but with some limitedstraying occurring between rivers (Jones,2006; ASMFC, 2009a).Throughout their life cycle, riverherring use many different

habitats, including the ocean, estuaries, rivers,and freshwater lakes and ponds. Thesubstrate preferred for spawning variesgreatly and can include gravel, detritus, and submerged aquatic vegetation.

Blueback herring prefer swifter movingwaters than alewives do (ASMFC,2009a). Nursery areas includefreshwater and semi-brackish waters.Little is known about their habitatpreference in the marine environment (Meadows, 2008; ASMFC, 2009a).Landlocked Populations Landlocked populations of alewivesand blueback herring also exist.Landlocked alewife populations occurin many freshwater lakes and pondsfrom Canada to North Carolina as wellas the Great Lakes (Rothschild, 1966;Boaze & Lackey, 1974). Manylandlocked populations occur as a resultof stocking to provide a forage base forgame fish species (Palkovacs et al.,2007).Landlocked blueback herring occurmostly in the southeastern United Statesand the Hudson River drainage.

Theoccurrence of landlocked bluebackherring is primarily believed to be theresult of accidental stockings inreservoirs (Prince and Barwick, 1981),unsanctioned stocking by recreational anglers to provide forage for game fish,and also through the construction oflocks, dams and canal systems that havesubsequently allowed for bluebackherring occupation of several lakes andponds along the Hudson River drainageup to, and including Lake Ontario(Limburg et al., 2001).Recent efforts to assess theevolutionary origins of landlocked alewives indicate that they rapidlydiverged from their anadromous cousinsbetween 300 and 5,000 years ago, andnow represent a discrete life historyvariant of the species, Alosapseudoharengus (Palkovacs et al., 2007).Though given their relatively recentdivergence from anadromous populations, one plausible explanation for the existence of landlocked populations may be the construction ofdams by either native Americans orearly colonial settlers that precluded thedownstream migration of juvenileherring (Palkovacs et al., 2007). Sincetheir divergence, landlocked alewiveshave evolved to a point they nowpossess significantly different mouthparts than their anadromous

cousins, including narrower gapes andsmaller gill raker spacings to takeadvantage of year round availability ofsmaller prey in freshwater lakes andponds (Palkovacs et al., 2007).Furthermore, the landlocked alewife,compared to its anadromous cousin,matures earlier, has a smaller adult bodysize, and reduced fecundity (Palkovacs et al., 2007). At this time, there is nosubstantive information that wouldsuggest that landlocked populations canor would revert back to an anadromous life history if they had the opportunity to do so (Gephard, CT DEEP, Pers.comm. 2012; Jordaan, UMASS Amherst,Pers. comm. 2012).The discrete life history andmorphological differences between thetwo life history variants (anadromous and landlocked) provide substantial evidence that upon becominglandlocked, landlocked populations become largely independent andseparate from anadromous populations and occupy largely separate ecological niches (Palkovacs and Post, 2008).There is the possibility that landlocked alewife and blueback herring may havethe opportunity to mix withanadromous river herring during highdischarge years and through damremovals which could provide passageover dams and access to historicspawning habitats restored foranadromous populations, where it didnot previously exist. The implications ofthis are not known at this time.In summary, genetics indicate thatanadromous alewife populations arediscrete from landlocked populations, and that this divergence can beestimated to have taken place from 300to 5,000 years ago. Some landlocked populations of blueback herring dooccur in the Mid-Atlantic andsoutheastern United States. Given thesimilarity in life histories betweenanadromous alewife and bluebackherring, we assume that landlocked populations of blueback herring wouldexhibit a similar divergence fromanadromous blueback

herring, as hasbeen documented with alewives.

A Memorandum of Understanding (MOU) between the U.S. Fish andWildlife Service (USFWS) and NMFS(collectively, the Services) regarding jurisdictional responsibilities and listingprocedures under the ESA was signedAugust 28, 1974. This MOU states thatNMFS shall have jurisdiction overspecies "which either (1) reside themajor portion of their lifetimes inmarine waters; or (2) are species whichspend part of their lifetimes in estuarine waters, if the major portion of theremaining time (the time which is notspent in estuarine waters) is spent inmarine waters."Given that landlocked populations ofriver herring remain in freshwater throughout their life history and aregenetically divergent from theanadromous

species, pursuant to theaforementioned MOU, we did not 48946Federal Register/Vol.

78, No. 155/Monday, August 12, 2013/Notices include the landlocked populations ofalewife and blueback herring in ourreview of the status of the species anddo not consider landlocked populations in this listing determination in responseto the petition to list these anadromous species.Listing Species Under the Endangered Species ActWe are responsible for determining whether alewife and blueback herringare threatened or endangered under theESA (16 U.S.C. 1531 et seq.).Accordingly, based on the statutory, regulatory, and policy provisions described below, the steps we followedin making our listing determination foralewife and blueback herring were to:(1) Determine how alewife and bluebackherring meet the definition of "species";

(2) determine the status of the speciesand the factors affecting them; and (3)identify and assess efforts being made toprotect the species and determine ifthese efforts are adequate to mitigateexisting threats.To be considered for listing under theESA, a group of organisms mustconstitute a "species."

Section 3 of theESA defines a "species" as "anysubspecies of fish or wildlife or plants,and any distinct population segment ofany species of vertebrate fish or wildlifewhich interbreeds when mature."Section 3 of the ESA further defines anendangered species as "any specieswhich is in danger of extinction throughout all or a significant portion ofits range" and a threatefied species asone "which is likely to become anendangered species within theforeseeable future throughout all or asignificant portion of its range." Thus,we interpret an "endangered species" tobe one that is presently in danger ofextinction.

A "threatened species,"

onthe other hand, is not presently indanger of extinction, but is likely tobecome so in the foreseeable future (thatis, at a later time). In other words, theprimary statutory difference between athreatened and endangered species isthe timing of when a species may be indanger of extinction, either presently (endangered) or in the foreseeable future(threatened).

On February 7, 1996, the Servicesadopted a policy to clarify ourinterpretation of the phrase "distinct population segment of any species ofvertebrate fish or wildlife" (61 FR 4722).The joint DPS policy describes twocriteria that must be considered whenidentifying DPSs: (1) The discreteness ofthe population segment in relation tothe remainder of the species (orsubspecies) to which it belongs; and (2)the significance of the population segment to the remainder of the species(or subspecies) to which it belongs.

Asfurther stated in the joint policy, if apopulation segment is discrete andsignificant (i.e., it meets the DPS policycriteria),

its evaluation for endangered or threatened status will be based on theESA's definitions of those terms and areview of the five factors enumerated insection 4(a)(1) of the ESA.As provided in section 4(a) of theESA, the statute requires us todetermine whether any species isendangered or threatened because ofany of the following five factors:

(1) Thepresent or threatened destruction, modification, or curtailment of itshabitat or range; (2) overutilization forcommercial, recreational, scientific, oreducational purposes; (3) disease orpredation; (4) the inadequacy of existingregulatory mechanisms; or (5) othernatural or manmade factors affecting itscontinued existence (section4(a)(1)(A)(E)).

Section 4(b)(1)(A) of theESA further requires that listingdeterminations be based solely on thebest scientific and commercial dataavailable after taking into accountefforts being made to protect thespecies.Distribution and Abundance United StatesThe stock assessment (described above) was prepared and compiled bythe River Herring Stock Assessment Subcommittee, hereafter referred to asthe 'subcommittee,'

of the ASMFC Shadand River Herring Technical Committee.

Data and reports used for thisassessment were obtained from Federaland state resource

agencies, powergenerating companies, and universities.

The subcommittee conducted itsassessment on the coastal stocks ofalewife and blueback herring byindividual rivers as well as coast-wide depending on available data. Thesubcommittee concluded that riverherring should ideally be assessed andmanaged by individual river system, butthat the marine portion of their lifehistory likely influences survivalthrough mixing in the marine portion oftheir range. However, coast-wide assessments are complicated by thecomplex life history of these species aswell, given that factors influencing population dynamics for the freshwater portion of their life history can notreadily be separated from marinefactors.

In addition, it was noted thatdata quality and availability varies byriver and is mostly dependent upon themonitoring efforts that each statededicates to these species, which furthercomplicated the assessment.

The subcommittee also noted thatmost state landings records listedalewife and blueback herring together as'river herring' rather than identifying byspecies.

These landings averaged 30.5million pounds (lbs) (13,847 metric tons(mt)) per year from 1889 to 1938, andsevere declines were noted coast-wide starting in the 1970s. Beginning in 2005,states began enacting moratoria on riverherring fisheries, and as of January2012, all directed harvest of riverherring in state waters is prohibited unless states have submitted andobtained approved sustainable fisheries management plans (FMP) underASMFC's Amendment 2 to the Shad andRiver Herring FMP.The subcommittee summarized itsfindings for trends in commercial catch-per-unit-effort (CPUE); run counts;young-of-the-year (YOY) seine surveys;juvenile-adult fisheries independent seine, gillnet and electrofishing surveys;juvenile-adult trawl surveys; meanlength; maximum age; mean length-at-age; repeat spawner frequency; totalmortality (Z) estimates; and exploitation rates. Because the stock assessment contains the most recent andcomprehensive description of thisinformation and the subcommittee's conclusions, the following sections weretaken from the stock assessment (ASMFC, 2012).Commercial CPUESince the mid-1990s, CPUE indicesfor alewives showed declining trends inthe Potomac River and James River(VA), no trend in the Rappahannock River (VA), and increasing trends in theYork River (VA) and Chowan River(NC). CPUE indices available forblueback herring showed a declining trend in the Chowan River and no trendin the Santee River (SC). Combinedspecies CPUE indices showed declining trends in Delaware Bay and theNanticoke River, but CPUE has recentlyincreased in the Hudson River (ASMFC,2012).Run CountsMajor declines in run sizes occurredin many rivers from 2001 to 2005. Thesedeclines were followed by increasing trends (2006 to 2010) in theAndroscoggin River (ME), Damaraiscotta River (ME), Nemasket River (MA),Gilbert-Stuart River (RI), and NonquitRiver (RI) for alewife and in theSebasticook River (ME), Cocheco River(NH), Lamprey River (NH), andWinnicut River (NH) for both speciescombined.

No trends in run sizes wereevident following the recent majordeclines in the Union River (ME),Mattapoisett River (MA), and Federal Register/Vol.

78, No. 155/Monday, August 12, 2013/Notices 48947Monument River (MA) for alewife andin the Exeter River (NH) for both speciescombined.

Run sizes have declined orare still declining following recent andhistorical major declines in the OysterRiver (NH) and Taylor River (NH) forboth species, in the Parker River (MA)for alewife, and in the Monument River(MA) and Connecticut River forblueback herring (ASMFC, 2012).Young-of-the-Year Seine SurveysThe young-of-the-year (YOY) seinesurveys were quite variable and showeddiffering patterns of trends amongrivers. Maine rivers showed similartrends in alewife and blueback herringYOY indices after 1991, with peaksoccurring in 1995 and 2004. YOYindices from North Carolina andConnecticut showed declines from the1980s to the present.

New York'sHudson River showed peaks in YOYindices in 1999, 2001, 2005, and 2007.New Jersey and Maryland YOY indicesshowed peaks in 1994, 1996, and 2001.Virginia YOY surveys showed peaks in1993, 1996, 2001, and 2003 (ASMFC,2012).Juvenile-Adult Fisheries-Independent Seine, Gillnet and Electrofishing SurveysThe juvenile-adult indices fromfisheries-independent seine, gillnet andelectrofishing surveys showed a varietyof trends in the available datasets for theRappahanock River (1991-2010),

JamesRiver (2000-2010),

St. John's River, FL(2001-2010),

and Narragansett Bay(1988-2010).

The gillnet indices fromthe Rappahannock River (alewife andblueback herring) showed a low andstable or decreasing trend after a majordecline after 1995 and has remained lowsince 2000 (except for a rise in alewifeCPUE during 2008). The gillnet andelectrofishing indices in the James River(alewife and blueback herring) showeda stable or increasing trend. Bluebackherring peak catch rates occurred in2004, and alewife peak catch ratesoccurred in 2005. The blueback herringindex from electrofishing in the St.John's River, FL, showed no trend aftera major decline from 2001-2002.

Theseine indices in Narragansett Bay, RI(combined species) and coastal ponds(combined species) showed no trendsover the time series. The CPUE forNarragansett Bay fluctuated withouttrend from 1988-1997, increased through 2000, declined and thenremained stable from 2001-2004.

Thepond survey CPUE increased during1993-1996, declined through 1998,increased in 1999, declined through2002, peaked in 2003 and then declinedand fluctuated without trend thereafter.

The electrofishing indices showedopposing trends and then declining trends in the Rappahannock River(alewife and blueback herring) withcatch rates of blueback herring peakingduring 2001-2003, and catch rates ofalewives lowest during the same timeperiod (ASMFC, 2012).Juvenile and Adult Trawl SurveysTrends in trawl survey indices variedgreatly with some surveys showing anincrease in recent years, some showinga decrease, and some remaining stable.Trawl survey data were available from1966-2010 (for a complete description of data see ASMFC (2012)).

Trawlsurveys in northern areas tended toshow either an increasing or stable trendin alewife indices, whereas trawlsurveys in southern areas tended toshow stable or decreasing trends.Patterns in trends across surveys wereless evident for blueback herring.

TheNMFS surveys showed a consistent increasing trend coast-wide and in thenorthern regions for alewife and thecombined river herring species group(ASMFC, 2012).Mean LengthMean sizes for male and femalealewife declined in 4 of 10 rivers, andmean sizes for female and maleblueback herring declined in 5 of 8rivers. Data were available from 1960-2010 (for a complete description of datasee ASMFC (2012)).

The common traitamong most rivers in which significant declines in mean sizes were detected isthat historical length data were available for years prior to 1990. Mean lengthsstarted to decline in the mid to late1980s; therefore, it is likely that declinesin other rivers were not detectedbecause of the shortness of their timeseries. Mean lengths for combined sexesin trawl surveys were quite variablethrough time for both alewives andblueback herring.

Despite thisvariability, alewife mean length tendedto be lowest in more recent surveys.This pattern was less apparent forblueback herring.

Trend analysis ofmean lengths indicated significant declines in mean lengths over time foralewives coast-wide and in the northernregion in both seasons, and for bluebackcoast-wide and in the northern region infall (ASMFC, 2012).Maximum AgeExcept for Maine and NewHampshire, maximum age of male andfemale alewife and blueback herringduring 2005-2007 was 1 or 2 yearslower than historical observations (ASMFC, 2012).Mean Length-at-Age Declines in mean length of at leastone age were observed in most riversexamined.

The lack of significance insome systems is likely due to theabsence of data prior to 1990 when thedecline in sizes began, similar to thepattern observed for mean length.Declines in mean lengths-at-age for mostages were observed in the north (NH)and the south (NC). There is littleindication of a general pattern of sizechanges along the Atlantic coast(ASMFC, 2012).Repeat Spawner Frequency Examination of percentage of repeatspawners in available data revealedsignificant, declining trends in theGilbert-Stuart River (RI-combined species),

Nonquit River (RI-combined species),

and the Nanticoke River(blueback herring).

There were notrends in the remaining rivers for whichdata are available, although scant datasuggest that current percentages ofrepeat spawners are lower thanhistorical percentages in the MonumentRiver (MA) and the Hudson River (NY)(ASMFC, 2012).Total Mortality (Z) Estimates With the exception of male bluebackherring from the Nanticoke River, whichshowed a slight increase over time,there were no trends in the Z estimates produced using age data (ASMFC,2012).Exploitation RatesExploitation of river herring appearsto be declining or remaining stable. In-river exploitation estimates havefluctuated, but are lower in recent years.A coast-wide index of relativeexploitation showed a decline following a peak in the 1980s, and the indexindicates that exploitation has remainedfairly stable over the past decade. Themajority of depletion-based stockreduction analysis (DB-SRA) modelruns showed declining exploitation rates coast-wide.

Exploitation ratesestimated from the statistical catch-at-age model for blueback herring in theChowan River also showed a slightdeclining trend from 1999 to 2007, atwhich time a moratorium wasinstituted.

There appears to be aconsensus among various assessment methodologies that exploitation hasdecreased in recent times. The declinein exploitation over the past decade isnot surprising because river herringpopulations are at low levels and morerestrictive regulations or moratoria havebeen enacted by states (ASMFC, 2012).

48948Federal Register/Vol.

78, No. 155/Monday, August 12, 2013/Notices Summary of Stock Assessment Conclusions Of the in-river stocks of alewife andblueback herring for which data wereavailable and were considered in thestock assessment, 22 were depleted, 1was increasing, and the status of 28stocks could not be determined becausethe time-series of available data was tooshort. In most recent years, 2 in-riverstocks were increasing, 4 weredecreasing, and 9 were stable, with 38rivers not having enough data to assessrecent trends. The coast-wide meta-complex of river herring stocks in theUnited States is depleted to nearhistorical lows. A depleted statusindicates that there was evidence fordeclines in abundance due to a numberof factors, but the relative importance ofthese factors in reducing river herringstocks could not be determined.

Commercial landings of river herringpeaked in the late 1960s, declinedrapidly through the 1970s and 1980sand have remained at levels less than 3percent of the peak over the pastdecade. Estimates of run sizes variedamong rivers, but in general, declining trends in run size were evident in manyrivers over the last decade. Fisheries-independent surveys did not showconsistent trends and were quitevariable both within and amongsurveys.

Those surveys that showeddeclines tended to be from areas southof Long Island. A problem with themajority of fisheries-independent surveys was that the length of their timeseries did not overlap the period of peakcommercial landings that occurred priorto 1970. There appears to be aconsensus among various assessment methodologies that exploitation hasdecreased in recent times. The declinein exploitation over the past decade isnot surprising because river herringpopulations are at low levels and morerestrictive regulations or moratoria havebeen enacted by states (ASMFC, 2012).CanadaThe Department of Fisheries andOceans (DFO) monitors and managesriver herring runs in Canada. Riverherring runs in the Miramichi River inNew Brunswick and the Maragree Riverin Cape Breton, Nova Scotia weremonitored intensively from 1983 to2000 (DFO, 2001). More recently (1997to 2006) the Gaspereau River alewiferun and harvest has been intensively monitored and managed partially inresponse to a 2002 fisheries management plan that had a goal ofincreasing spawning escapement to400,000 adults (DFO, 2007). Elsewhere, river herring runs have been monitored less intensively, though harvest rates aremonitored throughout Atlantic Canadathrough license sales, reporting requirements, and a logbook system thatwas enacted in 1992 (DFO, 2001).At the time DFO conducted their laststock assessment in 2001, theyidentified river herring harvest levels asbeing low (relative to historical levels)and stable, to low and decreasing acrossmost rivers where data were available (DFO, 2001). With respect to thecommercial harvest of river herring,reported landings of river herringpeaked in 1980 at slightly less than 25.5million lbs (11,600 mt) and declined toless than 11 million lbs (5,000 mt) in1996. Landings data reported throughDFO indicate that river herring harvestshave continued to decline through 2010.Consideration as a Species Under theESADistinct Population SegmentBackground According to Section 3 of the ESA, theterm "species" includes "anysubspecies of fish or wildlife or plants,and any distinct population segment ofany species of vertebrate fish or wildlifethat interbreeds when mature."Congress included the term "distinct population segment" in the 1978amendments to the ESA. On February 7,1996, the Services adopted a policy toclarify their interpretation of the phrase"distinct population segment" for thepurpose of listing, delisting, andreclassifying species (61 FR 4721). Thepolicy described two criteria apopulation segment must meet in orderto be considered a DPS (61 FR 4721): (1)It must be discrete in relation to theremainder of the species to which itbelongs; and (2) it must be significant tothe species to which it belongs.Determining if a population isdiscrete requires either one of thefollowing conditions:

(1) It is markedlyseparated from other populations of thesame taxon as a consequence ofphysical, physiological, ecological, orbehavioral factors.

Quantitative measures of genetic or morphological discontinuity may provide evidence ofthis separation; or (2) it is delimited byinternational governmental boundaries within which differences in control ofexploitation, management of habitat,conservation status, or regulatory mechanisms exist that are significant inlight of section 4(a)(1)(D) of the ESA.If a population is deemed discrete, then the population segment isevaluated in terms of significance.

Factors to consider in determining whether a discrete population segmentis significant to the species to which itbelongs include, but are not limited to,the following:

(1) Persistence of thediscrete population segment in anecological setting unusual or unique forthe taxon; (2) evidence that loss of thediscrete population segment wouldresult in a significant gap in the rangeof the taxon; (3) evidence that thediscrete population segment represents the only surviving natural occurrence ofa taxon that may be more abundantelsewhere as an introduced population outside its historic range; or (4)evidence that the discrete population segment differs markedly from otherpopulations of the species in its geneticcharacteristics.

If a population segment is deemeddiscrete and significant, then it qualifies as a DPS.Information Related to Discreteness To obtain expert opinion aboutanadromous alewife and bluebackherring stock structure, we convened aworking group in Gloucester, MA, onJune 20-21, 2012. This working groupmeeting brought together river herringexperts from state and Federal fisheries management agencies and academicinstitutions.

Participants presented information to inform the presence orabsence of stock structure such asgenetics, life history, andmorphometrics.

A public workshop washeld to present the expert workinggroup's findings on June 22, 2012, andduring this workshop, additional information on stock structure wassought from the public. Subsequently, asummary report was developed (NMFS,2012a), and a peer review of thedocument was completed by threeindependent reviewers.

The summaryreport and peer review reports areavailable on the NMFS Web site (see theADDRESSES section above).Steve Gephard of the Connecticut Department of Energy andEnvironmental Protection (CT DEP)presented a preliminary U.S. coast-wide genetic analysis of alewife and bluebackherring data (Palkovacs et al., 2012,unpublished report).

Palkovacs et al.,(2012, unpublished report) used 15novel microsatellite markers on samplescollected from Maine to Florida.

Foralewife, 778 samples were collected from spawning runs in 15 different rivers, and 1,201 blueback herringsamples were collected from 20 rivers.Bayesian analyses identified fivegenetically distinguishable stocks foralewife with similar results using bothSTRUCTURE and Bayesian Analysis ofPopulation Structure (BAPS) softwaremodels. The alewife stock complexes identified were: (1) Northern NewEngland; (2) Southern New England; (3)

Federal Register/Vol.

78, No. 155/Monday, August 12, 2013/Notices 48949Connecticut River; (4) Mid-Atlantic; and(5) North Carolina.

For bluebackherring, no optimum solution wasreached using STRUCTURE, whileBAPS suggested four genetically identifiable stock complexes.

The stockcomplexes identified for bluebackherring were: (1) Northern NewEngland; (2) Southern New England; (3)Mid Atlantic; (4) and Southern.

However, it should be noted that theseBayesian inferences of population structure provide a minimum number ofgenetically distinguishable groups. Inthe future, in order to better definepotential stock complexes, further testsexamining structure within designated stocks should be conducted usinghierarchical clustering analysis andgenetic tests.The study also examined the effects ofgeography and found a strong effect oflatitude on genetic divergence, suggesting a stepping stone model ofpopulation structure, and a strongpattern of isolation by distance, wheregene flow is most likely amongneighboring spawning populations.

Thepreliminary results from the studyfound significant differentiation amongspawning rivers for both alewife andblueback herring.

Based on the results oftheir study, the authors' preliminary management recommendations suggestthat river drainage is the appropriate level of management for both of thespecies.

This inference was alsosupported by genetic tests which wereconducted later. These tests suggest thatthere is substantial population structure at the drainage scale.The authors noted a number ofcaveats for their study including:

(1)Collection of specimens on theirupstream spawning run may poolsamples from what are truly distinctspawning populations within the majorriver drainages

sampled, thereby,underestimating genetic structure within rivers (Hasselman, 2010); (2) amore detailed analysis of population structure within the major stocksidentified (i.e., using hierarchical Bayesian clustering methods and genictest) would be useful for identifying anysubstructure within these major stocks;(3) neutral genetic markers used in thisstudy represent the effects of gene flowand historical population isolation, butnot the effects of adaptive processes, which are important to consider in thecontext of stock identification; (4) theanalysis is preliminary, and there are anumber of issues that need to be furtherinvestigated, including the effect ofdeviations in the Hardy-Weinberg Equilibrium model encountered in fouralewife loci and the failure ofSTRUCTURE to perform well on theblueback herring dataset; and (5)hybridization may be occurring betweenalewife and blueback herring and mayinfluence the results of the species-specific analyses.

Following the Stock Structure

Workshop, additional analyses were runon the alewife dataset to examine theuniqueness of the (tentatively) designated Connecticut River alewifestock complex.

Hybrids andmisidentified samples were found andsubsequently removed for this analysis, and the results were refined.

Byremoving these samples from theConnecticut River alewife dataset,Palkovacs et aJ. (2012, unpublished report) found that, for alewife, theConnecticut and Hudson Rivers belongto the Southern New England stock. Theanalyses were further refined andPalkovacs et al. (2012, unpublished report) provided an updated map of thealewife genetic stock complexes, combining the tentative North Carolinastock with the Mid-Atlantic stock. Thisinformation and analysis is completeand is currently being prepared forpublication.

Thus, the refined geneticstock complexes for alewife in thecoastal United States include NorthernNew England, Southern New England,and the Mid-Atlantic.

For bluebackherring, the identified genetic stocksinclude Northern New England,Southern New England, Mid-Atlantic and Southern (Palcovacs et al., 2012,unpublished report).Bentzen et al. (2012) implemented atwo-part genetic analysis of river herringto evaluate the genetic diversity ofalewives in Maine and MaritimeCanada, and to assess the regionaleffects of stocking on alewives andblueback herring in Maine. The geneticanalysis of alewives and bluebackherring along mid-coast Maine revealedsignificant genetic differentiation amongpopulations.

Despite significant differentiation, the patterns ofcorrelation did not closely correspond with geography or drainage affiliation.

The genetic analysis of alewives fromrivers in Maine and Atlantic Canadadetected isolation by distance, suggesting that homing behaviorindicative of alewives' metapopulation conformance does produce genetically distinguishable populations.

Furthertesting also suggested that there may beinterbreeding between alewives andblueback herring (e.g., hybrids),

especially at sample sites withimpassible dams.The unusual genetic groupings ofriver herring in Maine are likely a resultof Maine's complex stocking

history, asalewife populations in Maine have beensubject to considerable within and outof basin stocking for the purpose ofenhancement, recolonization ofextirpated populations, and stockintroduction.

Alewife stocking in Mainedates back at least to 1803 whenalewives were reportedly moved fromthe Pemaquid and St. George Rivers tocreate a run of alewives in theDamariscotta River (Atkins and Goode,1887). These efforts were largelyresponsive to considerable declines inalewife populations following theconstruction of dams, over exploitation and pollution.

Although there has beenconsiderable alewife stocking andrelocation throughout Maine, there arevery few records documenting theseefforts.

In contrast, considerably lessstocking of alewives has occurred inMaritime Canada. These geneticanalyses suggest that river herring fromCanadian waters are genetically distinctfrom Maine river herring.All of the expert opinions we receivedduring the Stock Structure Workshopsuggested evidence of regional stockstructure exists for both alewife andblueback herring as shown by the recentgenetics data (Palkovacs et al., 2012,unpublished report; Bentzen et al.,unpublished data). However, thesuggested boundaries of the regionalstock complexes differed from expert toexpert. Migration and mixing patterns ofalewives and blueback herring in theocean have not been determined, thoughregional stock mixing is suspected.

Therefore, the experts suggested that theocean phase of alewives and bluebackherring should be considered a mixedstock until further tagging and geneticdata become available.

There isevidence to support regional differences in migration

patterns, but not at a levelof river-specific stocks.In the mid-1980s, Rulifson et al.(1987) tagged and releasedapproximately 19,000 river herring inthe upper Bay of Fundy, Nova Scotiawith an overall recapture rate of 0.39percent.

Alewife tag returns were fromfreshwater locations in Nova Scotia, andmarine locations in Nova Scotia andMassachusetts.

Blueback herring tagreturns were from freshwater locations in Maryland and North Carolina andmarine locations in Nova Scotia.Rulifson et al. (1987) suspected fromrecapture data that alewives andblueback herring tagged in the Bay ofFundy were of different origins,hypothesizing that alewives were likelyregional fish from as far away as NewEngland, while the blueback herringrecaptures were likely not regional fish,but those of U.S. origin from the mid-Atlantic region. However, the low tagreturn numbers (n = 2) made it difficult to generalize about the natal rivers of 48950Federal Register/Vol.

78, No. 155/Monday, August 12, 2013/Notices blueback herring caught in the Bay ofFundy. The results of this tagging studyshow that river herring present inCanadian waters may originate fromU.S. waters and vice versa.Metapopulations of river herring arebelieved to exist, with adults frequently returning to their natal rivers forspawning and some straying occurring between rivers-straying rates havebeen estimated up to 20 percent (Jones,2006; ASMFC, 2009a; Gahagan et al.,2012). Given the available information on genetic differentiation coast-wide foralewife and blueback

herring, it appearsthat stock complexes exist for bothspecies.River herring originating fromCanadian rivers are delimited byinternational governmental boundaries.

Differences in control of exploitation, management of habitat, conservation status, or regulatory mechanisms existand, therefore, meet the discreteness criterion under the DPS policy;however, intermixing between bothalewife and blueback herring from U.S.and Canadian coastal waters occurs, andthe extent of this mixing is unknown.Given the best available information, it is possible to determine that thevarious stocks of both alewife andblueback herring are discrete.

The bestavailable information suggests that thedelineation of the stock complexes is asdescribed above; however, future workwill likely further refine thesepreliminary boundaries.

Additionally, further information is needed on theoceanic migratory patterns of bothspecies.Information Related to Significance If a population is deemed discrete, thepopulation is evaluated in terms ofsignificance.

Significance can bedetermined using the four criteria notedabove. Since the best available information indicates that the stockcomplexes identified for alewives andblueback herring are most likelydiscrete, the SRT reviewed the available information to determine if they aresignificant.

In evaluating the significance criterion, the SRT considered all of theabove criteria.

As indicated

earlier, bothalewives and blueback herring occupy alarge range spanning almost the entireEast Coast of the United States and intoCanada. They appear to migrate freelythroughout their oceanic range andreturn to freshwater habitats to spawn instreams, lakes and rivers. Therefore, they occupy many different ecological settings throughout their range.As described

earlier, the Palkovacs etal. (2012, unpublished report) studyassessed the genetic composition ofalewife and blueback herring stockswithin U.S. rivers using 15 neutral lociand documented that there are at leastthree stock complexes of alewife in theUnited States and four stock complexes of blueback herring in the United States.Palkovac et al. (2012, unpublished report) showed a strong effect of latitudeon genetic divergence, suggesting thatalthough most populations aregenetically differentiated, gene flow isgreater among neighboring runs thanamong distant runs. The genetic data areconsistent with the recent results of theASMFC stock assessment (2012), whichnoted that even among rivers within thesame state, there are differences intrends in abundance

indices, size-at-age, age structure and other metrics,indicating there are localized factorsaffecting the population dynamics ofboth species.Neutral genetic markers such asmicrosatellites have a longstanding history of utilization in stockdesignation for many anadromous fishspecies (Waples, 1998). However, thesemarkers represent the effects of geneflow and historical population isolation and not the effects of adaptiveprocesses.

The effects of adaptivegenetic and phenotypic diversity arealso extremely important to consider inthe context of stock designation, but arenot captured by the use of neutralgenetic markers.

Therefore, the available genetic data are most appropriately usedin support of the discreteness criterion, rather than to determine significance.

Determining whether a gap in therange of the taxon would be significant if a stock were extirpated is difficult todetermine with anadromous fish such asriver herring.

River herring aresuspected to migrate great distances between their natal rivers andoverwintering areas, and therefore, estuarine and marine populations arecomprised of mixed stocks.Consequently, the loss of a stockcomplex would mean the loss ofriverine spawning subpopulations, while the marine and estuarine habitatwould most likely still be occupied bymigratory river herring from other stockcomplexes.

As it has been shown thatgene flow is greater among neighboring runs than among distant runs, we mightexpect that river herring would re-colonize neighboring systems over arelatively short time frame. Thus, theloss of one stock complex in itself maynot be significant; the loss of contiguous stock complexes may be. The goal thenfor river herring stock complexes is tomaintain connectivity between geneticgroups to support propermetapopulation function (spatially separated populations of the samespecies that interact, recolonize vacanthabitats, and occupy new habitatsthrough dispersal mechanisms (Hanskiand Gilpin, 1991)).DPS Determination Evidence for genetic differentiation exists for both alewife and bluebackherring, allowing for preliminary identification of stock complexes;

however, available data are lacking onthe significance of each of theseindividual stock complexes.

Therefore, we have determined that there is notenough evidence to suggest that thestock complexes identified throughgenetics should be treated under theDPS policy as separate DPSs. The stockcomplexes may be discrete, but underthe DPS policy, they are not significant to the species as a whole. Furthermore, given the unknown level of intermixing between Canadian and U.S. riverherring in coastal waters, the Canadianstock complex should also not beconsidered separately under the DPSpolicy.Throughout the rest of thisdetermination, the species will bereferred to by species (alewife orblueback herring),

as river herringwhere information

overlaps, and by theidentified stock complexes (Palkovacs etal., 2012, unpublished report) for eachspecies as necessary.

While theindividual stock complexes do notconstitute separate DPSs, they areimportant components of the overallspecies and relevant to the evaluation ofwhether either species may bethreatened or endangered in asignificant portion of their overall range.Therefore, we have evaluated the threatsto, and extinction risk of the overallspecies and each of the individual stockcomplexes as presented below. For thisanalysis, the identified stock complexes for alewife (Figure 1) in the coastalUnited States for the purposes of thisfinding will include Northern NewEngland, Southern New England, theMid-Atlantic, and Canada; and stockcomplexes for blueback herring (Figure2) will include Northern New England,Southern New England, Mid-Atlantic, Southern

Atlantic, and Canada. Whilethe SRT concluded that there was notsufficient information at this time todetermine with any certainty whetheralewife or blueback herring stockcomplexes constitute separate DPSs,they recognized that future information on behavior, ecology and geneticpopulation structure may revealsignificant differences, showing fish tobe uniquely adapted to each stockcomplex.

We agree with this conclusion.

Thus, we are not identifying DPSs foreither species.

Federal Register/Vol.

78, No. 155/Monday, August 12, 2013/Notices 48951Figure 1. Alewife stock structure identified in Palkovacs et al., 2012, unpublished report.

48952Federal Register/Vol.

78, No. 155/Monday, August 12, 2013/Notices Stock Complex..,

• Northern New EnglandA Southern New EnglandMid-Altantic Southern0": 90 180 60 540~Kilom~eters A~Figure 2. Blueback herring stock structure identified in Palkovacs et al., 2012,unpublished report.Foreseeable Future and Significant Portion of Its RangeThe ESA defines an "endangered species" as "any species which is indanger of extinction throughout all or asignificant portion of its range," while a"threatened species" is defined as "anyspecies which is likely to become anendangered species within theforeseeable future throughout all or asignificant portion of its range." NMFSand the U.S. Fish and Wildlife Servce(USFWS) recently published a draftpolicy to clarify the interpretation of thephrase "significant portion of the range"in the ESA definitions of "threatened" and "endangered" (76 FR 76987;December 9, 2011). The draft policyprovides that: (1) If a species is foundto be endangered or threatened in onlya significant portion of its range, theentire species is listed as endangered orthreatened, respectively, and the ESA'sprotections apply across the species'entire range; (2) a portion of the rangeof a species is "significant" if its Federal Register/Vol.

78, No. 155/Monday, August 12, 2013/Notices 48953contribution to the viability of thespecies is so important that, withoutthat portion, the species would be indanger of extinction; (3) the range of aspecies is considered to be the generalgeographical area within which thatspecies can be found at the time USFWSor NMFS makes any particular statusdetermination; and (4) if the species isnot endangered or threatened throughout all of its range, but it isendangered or threatened within asignificant portion of its range, and thepopulation in that significant portion isa valid DPS, we will list the DPS ratherthan the entire taxonomic species orsubspecies.

The Services are currently reviewing public comment received on the draftpolicy. While the Services' intent is toestablish a legally binding interpretation of the term "significant portion of therange," the draft policy does not havelegal effect until such time as it may beadopted as final policy. Here, we applythe principles of this draft policy asnon-binding guidance in evaluating whether to list alewife or bluebackherring under the ESA. If the policychanges in a material way, we willrevisit the determination and assesswhether the final policy would result ina different outcome.While we have determined that DPSscannot be defined for either of thesespecies based on the available information, the stock complexes dorepresent important groupings withinthe range of both species.

Thus, in ouranalysis of extinction risk and threatsassessment below, we have evaluated whether either species is at riskrangewide and within any of theindividual stock complexes so that wecan evaluate whether either species isthreatened or endangered in asignificant ortion of its range.We estabflished that the appropriate period of time corresponding to theforeseeable future is a function of theparticular type of threats, the life-history characteristics, and the specific habitatrequirements for river herring.

Thetimeframe established for theforeseeable future takes into account thetime necessary to provide for theconservation and recovery of eachspecies and the ecosystems upon whichthey depend, but is also a function ofthe reliability of available data regarding the identified threats and extends onlyas far as the data allow for makingreasonable predictions about thespecies' response to those threats.

Asdescribed below, the SRT determined that dams and other impediments tomigration have already created a clearand present threat to river herring thatwill continue into the future. The SRTalso evaluated the threat from climatechange from 2060 to 2100 and climatevariability in the near term (as described in detail below).Highly productive species with shortgeneration times are more resilient thanless productive, long lived species, asthey are quickly able to take advantage of available habitats for reproduction (Mace et al., 2002). Species with shortergeneration times, such as river herring(4 to 6 years), experience greaterpopulation variability than species withlong generation times, because theymaintain the capacity to replenish themselves more quickly following aperiod of low survival (Mace et 0l.,2002). Given the high population variability among clupeids, projecting out further than three generations couldlead to considerable uncertainty in theprobability that the model will providean accurate representation of thepopulation trajectory for each species.Thus, a 12 to 18 year timeframe (e.g.,2024-2030),

or a three-generation timeperiod, for each species was determined by the Team to be appropriate for useas the foreseeable future for both alewifeand blueback herring.

We agree with theTeam that a three-generation timeperiod (12-18 years) is a reasonable foreseeable future for both alewife andblueback herring.Connectivity, population resilience and diversity are important whendetermining what constitutes asignificant portion of the species' range(Waples et al., 2007). Maintaining connectivity between genetic groupssupports proper metapopulation

function, in this case, anadromy.

Ensuring that river herring populations are well represented across diversehabitats helps to maintain and enhancegenetic variability and population resilience (McElhany et al., 2000).Additionally, ensuring wide geographic distribution across diverse climate andgeographic regions helps to minimizerisk from catastrophes (e.g., droughts, floods, hurricanes, etc.; McElhany et al.,2000). Furthermore, preventing isolation of genetic groups protects againstpopulation divergence (Allendorf andLuikart, 2007).Threats Evaluation As described above, Section 4(a)(1) ofthe ESA and NMFS implementing regulations (50 CFR 424) states that wemust determine whether a species isendangered or threatened because ofany one or a combination of thefollowing factors:

(A) Current orthreatened habitat destruction ormodification or curtailment of habitat orrange; (B) overutilization forcommercial, recreational, scientific, oreducational purposes; (C) disease orpredation; (D) inadequacy of existingregulatory mechanisms; and (E) othernatural or man-made factors affecting the species' continued existence.

Thissection briefly summarizes the findingsregarding these factors.A. The Present or Threatened Destruction, Modification, orCurtailment of Its Habitat or RangePast, present, and reasonably foreseeable future factors that have thepotential to affect river herring habitatinclude, but are not limited to, damsand hydropower facilities,

dredging, water quality (including land usechange, water withdrawals, discharge and contaminants),

climate change andclimate variability.

As noted above,river herring occupy a variety ofdifferent habitats including freshwater, estuarine and marine environments throughout their lives, and thus, theyare subjected to habitat impactsoccurring in all of these different habitats.

Dams and Other BarriersDams and other barriers to upstreamand downstream passage (e.g., culverts) can block or impede access to habitatsnecessary for spawning and rearing; cancause direct and indirect mortality frominjuries incurred while passing overdams, through downstream passagefacilities, or through hydropower turbines; and can degrade habitatfeatures necessary to support essential river herring life history functions.

Man-made barriers that block or impedeaccess to rivers throughout the entirehistorical range of river herring haveresulted in significant losses ofhistorical spawning habitat for riverherring.

Dams and other man-madebarriers have contributed to thehistorical and current declines inabundance of both blueback and alewifepopulations.

While estimates of habitatloss over the entire range of riverherring are not available, estimates fromstudies in Maine show that less than 5percent of lake spawning habitat and 20percent of river habitat remainsaccessible for river herring (Hall et al.,2010). As described in more detailbelow, dams are also known to impactriver herring through variousmechanisms, such as habitat alteration, fish passage delays, and entrainment and impingement (Ruggles 1980; NRC2004). River herring can undergoindirect mortality from injuries such asscale loss, lacerations,

bruising, eye orfin damage, or internal hemorrhaging when passing through turbines, overspillways, and through bypasses(Amaral et al., 2012).

48954Federal Register/Vol.

78, No. 155/Monday, August 12, 2013/Notices The following summary of the effectsof dams and other barriers on riverherring is taken from Amendment 2 tothe Interstate Fishery Management Planfor Shad and River Herring (hereafter, referred to as "Amendment 2" and citedas "ASMFC, 2009"). Because it includesa detailed description of barriers toupstream and downstream

passage, it isthe best source of comprehensive information on this topic. Please refer toAmendment 2 for more information.

Dams and spillways impeding riversalong the East Coast of the United Stateshave resulted in a considerable loss ofhistorical spawning habitat for shad andriver herring.

Permanent man-madestructures pose an ongoing barrier tofish passage unless fishways areinstalled or structures are removed.Low-head dams can also pose aproblem, as fish are unable to pass overthem except when tides or riverdischarges are exceptionally high(Loesch and Atran, 1994). Historically, major dams were often constructed atthe site of natural formations conducive to waterpower, such as natural falls.Diversion of water away from rapids atthe base of falls can reduce fish habitat,and in some cases cause rivers to rundry at the base for much of the summer(MEOEA, 2005; ASMFC, 2009).Prior to the early 1990s, it wasthought that migrating shad and riverherring suffered significant mortality going through turbines duringdownstream passage (Mathur andHeisey, 1992). Juvenile shad emigrating from rivers have been found toaccumulate in larger numbers near theforebay of hydroelectric facilities, wherethey become entrained in intake flowareas (Martin et a0., 1994). Relatively high mortality rates were reported (62percent to 82 percent) at a hydroelectric dam for juvenile American shad andblueback

herring, depending on thepower generation levels tested (Taylorand Kynard, 1984). In contrast, Mathurand Heisey (1992) reported a mortality rate of 0 percent to 3 percent forjuvenile American shad (2 to 6 in forklength (55 to 140 mm)), and 4 percentfor juvenile blueback herring (3 to 4 infork length (77 to 105 mm)) throughKaplan turbines.

Mortality rateincreased to 11 percent in passagethrough a low-head Francis turbine(Mathur and Heisey, 1992). Otherstudies reported less than 5 percentmortality when large Kaplan and fixed-blade, mixed-flow turbines were used ata facility along the Susquehanna River(RMC, 1990; RMC, 1994). At the samesite, using small Kaplan and Francisrunners, the mortality rate was as highas 22 percent (NA, 2001). At anothersite, mortality rate was about 15 percentwhere higher revolution, Francis-type runners were used (RMC, 1992; ASMFC,2009).Additional studies reported thatchanges in pressure had a morepronounced effect on juveniles withthinner and weaker tissues as theymoved through turbines (Taylor andKynard, 1984). Furthermore, some fishmay die later from stress, or becomeweakened and more susceptible topredation, and as such, losses may notbe immediately apparent to researchers (Gloss, 1982) (ASMFC, 2009).Changes to the river system, resulting in delayed migration among otherthings, were also identified inAmendment 2 as impacting riverherring.

Amendment 2 notes that whenjuvenile alosines delay out-migration, they may concentrate behind dams andbecome more susceptible to activelyfeeding predators.

They may also bemore vulnerable to anglers that targetalosines as a source of bait. Delayed out-migration can also make juvenilealosines more susceptible to marinepredators that they may have avoided ifthey had followed their naturalmigration patterns (McCord, 2005a). Inopen rivers, juvenile alosines gradually move seaward in groups that are likelyspaced according to the spatialseparation of spawning and nurserygrounds (Limburg, 1996; J. McCord,South Carolina Department of NaturalResources, personal observation).

Releasing water from dams andimpoundments (or reservoirs) may leadto flow alterations, altered sedimenttransport, disruption of nutrientavailability, changes in downstream water quality (including both reducedand increased temperatures),

streambank

erosion, concentration ofsediment and pollutants, changes inspecies composition, solubilization ofiron and manganese and their absorbedor chelated ions, and hydrogen sulfidein hypolimnetic (water at low leveloutlets) releases (Yeager, 1995; Erkan,2002; ASMFC, 2009).Many dams spill water over the top ofthe structure where water temperatures are the warmest, essentially creating aseries of warm water ponds in place ofthe natural stream channel (Erkan,2002). Conversely, water released fromdeep reservoirs may be poorlyoxygenated, at below-normal seasonalwater temperature, or both, therebycausing loss of suitable spawning ornursery habitat in otherwise habitable areas (ASMFC, 2009).Reducing minimum flows can reducethe amount of water available and causeincreased water temperature or reduceddissolved oxygen levels (ASMFC, 1985;ASMFC, 1999; USFWS et al., 2001).Such conditions have occurred alongthe Susquehanna River at theConowingo Dam, Maryland, from latespring through early fall, and havehistorically caused large fish kills belowthe dam (Krauthamer and Richkus,1987; ASMFC, 2009).Disruption of seasonal flow rates inrivers can impact upstream anddownstream migration patterns for adultand juvenile alosines (ASMFC, 1985;Limburg, 1996; ASMFC, 1999; USFWSet al., 2001). Changes to natural flowscan also disrupt natural productivity and availability of zooplankton thatlarval and early juvenile alosines feedon (Crecco and Savoy, 1987; Limburg,1996; ASMFC, 2009).Although most dams that impactdiadromous fish are located along thelengths of rivers, fish can also beaffected by hydroelectric projects at themouths of rivers, such as the large tidalhydroelectric project at the Annapolis River in the Bay of Fundy, Canada. Thisparticular basin and other surrounding waters are used as foraging areas duringsummer months by American shad fromall runs along the East Coast of theUnited States (Dadswell et al., 1983).Because the facilities are tidalhydroelectric

projects, fish may move inand out of the impacted areas with eachtidal cycle. While turbine mortality isrelatively low with each passage, therepeated passage in and out of thesefacilities may cumulatively result insubstantial overall mortalities (Scarratt and Dadswell, 1983; ASMFC, 2009).Additional man-made structures thatmay obstruct upstream passage include:tidal and amenity barrages (barriers constructed to alter tidal flow foraesthetic purposes or to harness energy);tidal flaps (used to control tidal flow);mill, gauging,

amenity, navigation, diversion, and water intake weirs; fishcounting structures; and earthen berms(Durkas, 1992; Solomon and Beach,2004). The impact of these structures issite-specific and will vary with anumber of conditions including headdrop, form of the structure, hydrodynamic conditions upstream anddownstream, condition of the structure, and presence of edge effects (Solomonand Beach, 2004). Road culverts are alsoa significant source of blockage.

Culverts are popular, low-costalternatives to bridges when roads mustcross small streams and creeks.Although the amount of habitat affectedby an individual culvert may be small,the cumulative impact of multipleculverts within a watershed can besubstantial (Collier and Odom, 1989;ASMFC, 2009).Roads and culverts can also imposesignificant changes in water quality.

Federal Register/Vol.

78, No. 155/Monday, August 12, 2013/Notices 48955Winter runoff in some states mayinclude high concentrations of road salt,while stormwater flows in the summermay cause thermal stress and bring highconcentrations of other pollutants (MEOEA, 2005; ASMFC, 2009).Sampled sites in North Carolinarevealed river herring upstream anddownstream of bridge crossings, but noherring were found in upstream sectionsof streams with culverts.

Additional study is underway to determine if riverherring are absent from these areasbecause of the culverts (NCDENR, 2000).Even structures only 8 to 12 in (20 to 30cm) above the water can block shad andriver herring migration (ASMFC, 1999;ASMFC, 2009).Rivers can also be blocked by non-anthropogenic

barriers, such as beaverdams, waterfalls, log piles, andvegetative debris. These blockages mayhinder migration, but they can alsobenefit by providing adhesion sites foreggs, protective cover, and feeding sites(Klauda et al., 1991b). Successful passage at these natural barriers oftendepends on individual stream flowcharacteristics during the fish migration season (ASMFC, 2009).DredgingWetlands provide migratory corridors and spawning habitat for river herring.The combination of incremental lossesof wetland habitat, changes inhydrology, and nutrient and chemicalinputs over time, can be extremely

harmful, resulting in diseases anddeclines in the abundance and quality.Wetland loss is a cumulative impactthat results from activities related todredging/dredge spoil placement, portdevelopment,

marinas, solid wastedisposal, ocean disposal, and marinemining. In the late 1970s and early1980s, the United States was losingwetlands at an estimated rate of 300,000acres (1,214 sq km) per year. The CleanWater Act and state wetland protection programs helped decrease wetlandlosses to 117,000 acres (473 sq km) peryear, between 1985 and 1995. Estimates of wetlands loss vary according to thedifferent agencies.

The U.S. Department of Agriculture (USDA) attributes 57percent of wetland loss to development, 20 percent to agriculture, 13 percent tothe creation of deepwater

habitat, and10 percent to forest land, rangeland, andother uses. Of the wetlands lost between1985 and 1995, the USFWS estimates that 79 percent of wetlands were lost toupland agriculture.

Urban development and other types of land use activities were responsible for 6 percent and 15percent of wetland loss, respectively.

Amendment 2 identifies channelization and dredging as a threatto river herring habitat.

The following

section, taken from Amendment 2,describes these threats.Channelization can cause significant environmental impacts (Simpson et al.,1982; Brookes, 1988), including bankerosion, elevated water velocity, reduced habitat diversity, increased

drainage, and poor water quality(Hubbard, 1993). Dredging and disposalof spoils along the shoreline can alsocreate spoil banks, which block accessto sloughs, pools, adjacent vegetated areas, and backwater swamps(Frankensteen, 1976). Dredging may alsorelease contaminants, resulting inbioaccumulation, direct toxicity toaquatic organisms, or reduced dissolved oxygen levels (Morton, 1977).Furthermore, careless land use practices may lead to erosion, which can lead tohigh concentrations of suspended solids(turbidity) and substrate (siltation) inthe water following normal and intenserainfall events. This can displace larvaeand juveniles to less desirable areasdownstream and cause osmotic stress(Klauda et al., 1991b; ASMFC, 2009).Spoil banks are often unsuitable habitat for fishes. Suitable habitat isoften lost when dredge disposal materialis placed on natural sand bars and/orpoint bars. The spoil is too unstable toprovide good habitat for the food chain.Draining and filling, or both, ofwetlands adjacent to rivers and creeksin which alosines spawn has eliminated spawning areas in North Carolina(NCDENR, 2000; ASMFC, 2009).Secondary impacts from channelformation include loss of vegetation anddebris, which can reduce habitat forinvertebrates and result in reducedquantity and diversity of prey forjuveniles (Frankensteen, 1976).Additionally, stream channelization often leads to altered substrate in theriverbed and increased sedimentation (Hubbard, 1993), which in turn canreduce the diversity,

density, andspecies richness of aquatic insects(Chutter, 1969; Gammon, 1970; Taylor,1977). Suspended sediments can reducefeeding success in larval or juvenilefishes that rely on visual cues forplankton feeding (Kortschal et al.,1991). Sediment re-suspension fromdredging can also deplete dissolved oxygen, and increase bioavailability ofany contaminants that may be bound tothe sediments (Clark and Wilber, 2000;ASMFC, 2009).Migrating adult river herring avoidchannelized areas with increased watervelocities.

Several channelized creeks inthe Neuse River basin in North Carolinahave reduced river herring distribution and spawning areas (Hawkins, 1979).Frankensteen (1976) found that thechannelization of Grindle Creek, NorthCarolina removed in-creek vegetation and woody debris, which had served assubstrate for fertilized eggs (ASMFC,2009).Channelization can also reduce theamount of pool and riffle habitat(Hubbard, 1993), which is an important food-producing area for larvae (Keller,1978; Wesche, 1985; ASMFC, 2009).Dredging can negatively affect alosinepopulations by producing suspended sediments (Reine et al., 1998), andmigrating alosines are known to avoidwaters of high sediment load (ASMFC,1985; Reine et al., 1998). Fish may alsoavoid areas that are being dredgedbecause of suspended sediment in thewater column. Filter-feeding fishes,such as alosines, can be negatively impacted by suspended sediments ongill tissues (Cronin et al., 1970).Suspended sediments can clog gills thatprovide oxygen, resulting in lethal andsub-lethal effects to fish (Sherk et al.,1974 and 1975; ASMFC, 2009).Nursery areas along the shorelines ofthe rivers in North Carolina have beenaffected by dredging and filling, as wellas by erection of bulkheads; however,the degree of impact has not beenmeasured.

In some areas, juvenilealosines were unable to enterchannelized sections of a stream due tohigh water velocities caused bydredging (ASMFC, 2000 and 2009).Water QualityNutrient enrichment has become amajor cumulative problem for manycoastal waters. Nutrient loading resultsfrom the individual activities of coastaldevelopment, marinas and recreational

boating, sewage treatment and disposal, industrial wastewater and solid wastedisposal, ocean disposal, agriculture, and aquaculture.

Excess nutrients fromland based activities accumulate in thesoil, pollute the atmosphere, polluteground water, or move into streams andcoastal waters. Nutrient inputs areknown to have a direct effect on waterquality.

For example, nutrientenrichment can stimulate growth ofphytoplankton that consumes oxygenwhen they decay, which can lead to lowdissolved oxygen that may result in fishkills (Correll, 1987; Tuttle et al., 1987;Klauda et al., 1991b); this condition isknown as eutrophication.

In addition to the direct cumulative effects incurred by development activities, inshore and coastal habitatsare also threatened by persistent increases in certain chemicaldischarges.

The combination ofincremental losses of wetland habitat,changes in hydrology, and nutrient andchemical inputs produced over time can 48956Federal Register/Vol.

78, No. 155/Monday, August 12, 2013/Notices be extremely harmful to marine andestuarine biota, including river herring,resulting in diseases and declines in theabundance and quality of the affectedresources.

Amendment 2 identified land usechanges including agriculture, logging/forestry, urbanization and non-point source pollution as threats to riverherring habitat.

The following section,taken from Amendment 2, describes these threats.The effects of land use and land coveron water quality, stream morphology, and flow regimes are numerous, andmay be the most important factorsdetermining quantity and quality ofaquatic habitats (Boger, 2002). Studieshave shown that land use influences dissolved oxygen (Limburg andSchmidt, 1990), sediments and turbidity (Comeleo et al., 1996; Basnyat et al.,1999), water temperature (Hartman etaL., 1996; Mitchell, 1999), pH (Osborneand Wiley, 1988; Schofield, 1992),nutrients (Peterjohn and Correll, 1984;Osborne and Wiley, 1988; Basnyat et aL.,1999), and flow regime (Johnston et al.,1990; Webster et al., 1992; ASMFC,2009).Siltation, caused by erosion due toland use practices, can kill submerged aquatic vegetation (SAV). SAV can beadversely affected by suspended sediment concentrations of less than 15ppm (15 mg/L) (Funderburk et al., 1991)and by deposition of excessive sediments (Valdes-Murtha and Price,1998). SAV is important because itimproves water quality (Carter et al.,1991). SAV consumes nutrients in thewater and as the plants die and decay,they slowly release the nutrients backinto the water column. Additionally, through primary production andrespiration, SAV affects the dissolved oxygen and carbon dioxideconcentrations, alkalinity, and pH of thewaterbody.

SAV beds also bindsediments to the bottom resulting inincreased water clarity, and theyprovide refuge habitat for migratory.

fishand planktonic prey items (Maldeis, 1978; Monk, 1988; Killgore et al., 1989;ASMFC, 2009).Decreased water quality fromsedimentation became a problem withthe advent of land-clearing agriculture in the late 18th century (McBride, 2006).Agricultural practices can lead tosedimentation in streams, riparianvegetation loss, influx of nutrients (e.g.,inorganic fertilizers and animal wastes),and flow modification (Fajen andLayzer, 1993). Agriculture, silviculture, and other land use practices can lead tosedimentation, which reduces theability of semi-buoyant eggs andadhesive eggs to adhere to substrates (Mansueti, 1962; ASMFC, 2009).From the 1950s to the present,increased nutrient loading has madehypoxic conditions more prevalent (Officer et al., 1984; Mackiernan, 1987;Jordan et al., 1992; Kemp et al., 1992;Cooper and Brush, 1993; Secor andGunderson, 1998). Hypoxia is mostlikely caused by eutrophication, duemostly to non-point source pollution (e.g., industrial fertilizers used inagriculture) and point source pollution (e.g., urban sewage).Logging activities can modifyhydrologic balances and in-stream flowpatterns, create obstructions, modifytemperature

regimes, and add nutrients, sediments, and toxic substances intoriver systems.

Loss of riparianvegetation can result in fewer refugeareas for fish from fallen trees, fewerinsects for fish to feed on, and reducedshade along the river, which can lead toincreased water temperatures andreduced dissolved oxygen (EDF, 2003).Threats from deforestation of swampforests include:

siltation from increased erosion and runoff; decreased dissolved oxygen (Lockaby et al., 1997); anddisturbance of food-web relationships inadjacent and downstream waterways (Batzer et al., 2005; ASMFC, 2009).Urbanization can cause elevatedconcentrations of nutrients,

organics, orsediment metals in streams (Wilber andHunter, 1977; Kelly and Hite, 1984;Lenat and Crawford, 1994). Moreresearch is needed on how urbanization affects diadromous fish populations;

however, Limburg and Schmidt (1990)found that when the percent ofurbanized land increased to about 10percent of the watershed, the number ofalewife eggs and larvae decreased significantly in tributaries of theHudson River, New York (ASMFC,2009).Water Withdrawal/Outfall Water withdrawal facilities and toxicand thermal discharges have also beenidentified as impacting river herring,and the following section is summarized from Amendment 2.Large volume water withdrawals (e.g.,drinking water, pumped-storage hydroelectric

projects, irrigation, andsnow-making) can alter local currentcharacteristics (e.g., reverse river flow),which can result in delayed movementpast a facility or entrainment in waterintakes (Layzer and O'Leary, 1978).Planktonic eggs and larvae entrained atwater withdrawal projects experience high mortality rates due to pressurechanges, shear and mechanical

stresses, and heat shock (Carlson and McCann,1969; Marcy, 1973; Morgan et al., 1976).While juvenile mortality rates aregenerally low at well-screened facilities, large numbers of juveniles can beentrained (Hauck and Edson, 1976;Robbins and Mathur, 1976; ASMFC,2009).Fish impinged against water filtration screens can die from asphyxiation, exhaustion, removal from the water forprolonged periods of time, removal ofprotective mucous, and descaling (DBC,1980). Studies conducted along theConnecticut River found that larvae andearly juveniles of alewife, bluebackherring, and American shad suffered100-percent mortality whentemperatures in the cooling system of apower plant were elevated above 82 'F(280C); 80 percent of the total mortality was caused by mechanical damage, 20percent by heat shock (Marcy, 1976).Ninety-five percent of the fish near theintake were not captured by the screen,and Marcy (1976) concluded that it didnot seem possible to screen fish larvaeeffectively (ASMFC, 2009).The physical characteristics ofstreams (e.g., stream width, depth, andcurrent velocity; substrate; andtemperature) can be altered by waterwithdrawals (Zale et al., 1993). Riverherring can experience thermal stress,direct mortality, or indirect mortality when water is not released during timesof low river flows and watertemperatures are higher than normal.Water flow disruption can also result inless freshwater input to estuaries (Rulifson, 1994), which are important nursery areas for river herring and otheranadromous species (ASMFC, 2009).Industrial discharges may containtoxic chemicals, such as heavy metalsand various organic chemicals (e.g.,insecticides,

solvents, herbicides) thatare harmful to aquatic life (ASMFC,1999). Many contaminants can haveharmful effects on fish, including reproductive impairment (Safe, 1990;Mac and Edsall, 1991; Longwell et al.,1992). Chemicals and heavy metals canmove through the food chain, producing sub-lethal effects such as behavioral andreproductive abnormalities (Matthews etal., 1980). In fish, exposure topolychlorinated biphenyls (PCBs) cancause fin erosion, epidermal lesions,blood anemia, altered immune response, and egg mortality (Post, 1987; Kennishet al., 1992). Steam power plants thatuse chlorine to prevent bacterial, fungal,and algal growth present a hazard to allaquatic life in the receiving stream, evenat low concentrations (Miller et al.,1982; ASMFC, 2009).Pulp mill effluent and other oxygen-consuming wastes discharged into riversand streams can reduce dissolved oxygen concentrations below what is Federal Register/Vol.

78, No. 155/Monday, August 12, 2013/Notices 48957required for river herring survival.

Lowdissolved oxygen resulting fromindustrial pollution and sewagedischarge can also delay or preventupstream and downstream migrations.

Everett (1983) found that during timesof low water flow when pulp milleffluent comprised a large percentage ofthe flow, river herring avoided theeffluent.

Pollution may be diluted in thefall when water flows increase, but fishthat reach the polluted watersdownriver before the water has flushedthe area will typically succumb tosuffocation (Miller et aI., 1982; ASMFC,2009).Effluent may also pose a greater threatduring times of drought.

Suchconditions were suspected of interfering with the herring migration along theChowan River, North Carolina, in 1981.In the years before 1981, the effluentfrom the pulp mill had passed prior tothe river herring run, but droughtconditions caused the effluent to remainin the system longer that year. Toxiceffects were indicated, and researchers suggested that growth and reproduction might have been disrupted as a result ofeutrophication and other factors(Winslow et al., 1983; ASMFC, 2009).Klauda et al. (1991a) provides anextensive review of temperature thresholds for alewife and blubackherring.

In summary, the spawningmigration for alewives most often occurswhen water temperatures range from50-64 0F (10-18 °C), and for bluebacks when temperatures range from 57-77 0F(14-25 'CQ. Alewife egg deposition mostoften occurs when temperatures rangebetween 50-72 OF (10 and 22 0C), andfor bluebacks when temperatures rangebetween 70-77 OF (21 and 25 0C).Alewife egg and larval development isoptimal when temperatures range from63-70 OF (17-21 °C), and for bluebacks when temperatures range from 68-75 OF(20-24 'C) (temperature ranges werealso presented and discussed at theClimate Workshop (NMFS, 2012b)).Thermal effluent from power plantsoutside these temperature ranges whenriver herring are present can disruptschooling

behavior, causedisorientation, and may result in death.Sewage can directly and indirectly affect anadromous fish. Majorphytoplankton and algal blooms thatreduced light penetration (Dixon, 1996)and ultimately reduced SAV abundance (Orth et al., 1991) in tidal freshwater areas of the Chesapeake Bay in the1960s and early 1970s may have beencaused by ineffective sewage treatment (ASMFC, 2009).Water withdrawal for irrigation cancause dewatering or reduced streamflow of freshwater

streams, which candecrease the quantity of both spawningand nursery habitat for anadromous fish. Reduced streamflow can reducewater quality by concentrating pollutants and/or increasing watertemperature (ASMFC, 1985). O'Connell and Angermeier (1999) found that insome Virginia

streams, there was aninverse relationship between theproportion of a stream's watershed thatwas agriculturally developed and theoverall tendency of the stream tosupport river herring runs. In NorthCarolina, cropland alteration alongseveral creeks and rivers significantly reduced river herring distribution andspawning areas in the Neuse River basin(Hawkins, 1979; ASMFC, 2009).Atmospheric deposition occurs whenpollutants (e.g. nitrates,

sulfates, ammonium, and mercury) aretransferred from the air to the earth'ssurface.

Pollutants can get from the airinto the water through rain and snow,falling particles, and absorption of thegas form of the pollutants into the water.Atmospheric pollutants can result inincreased eutrophication (Paerl et al.,1999) and acidification of surface waters(Haines, 1981). Atmospheric nitrogendeposition in coastal estuaries can leadto accelerated algal production (oreutrophication) and water qualitydeclines (e.g., hypoxia,

toxicity, and fishkills) (Paerl et al., 1999). Nitrate andsulfate deposition is acidic and canreduce stream pH (measure of thehydronium ion concentration) andelevate toxic forms of aluminum(Haines, 1981). When pH declines, thenormal ionic salt balance of the fish iscompromised and fish lose body salts tothe surrounding water (Southerland etal., 1997). Sensitive fish species canexperience acute mortality, reducedgrowth, skeletal deformities, andreproductive failure (Haines, 1981).Climate Change and Climate Variability Possible climate change impacts toriver herring were noted in the stockassessment (ASMFC, 2012) based onregional patterns in trends (e.g., trawlsurveys in southern regions showeddeclining trends more frequently compared to those in northern regions).

However, additional information wasneeded on this topic to inform ourlisting decision, and as noted above, weheld a workshop to obtain expertopinion on the potential impacts ofclimate change on river herring (NMFS,2012b).As discussed at the workshop, bothnatural climate variability andanthropogenic-forced climate changewill affect river herring (NMFS, 2012b).Natural climate variability includes theAtlantic Multidecadal Oscillation, theNorth Atlantic Oscillation, and the ElNifio Southern Oscillation.

During theworkshop, it was noted that impactsfrom global climate change induced byhuman activities are likely to becomemore apparent in future years(Intergovernmental Panel on ClimateChange (IPCC), 2007). Results presented from the North American RegionalClimate Change Assessment Program(NARCCAP-a group that uses fieldsfrom the global climate models toprovide boundary conditions forregional atmospheric models coveringmost of North America and extending over the adjacent oceans) suggest thattemperature will warm throughout theyears over the northeast, mid-Atlantic and Southeast United States (comparing 1968-1999 to 2038-2069; NMFS,2012b). Additionally, it was noted thatthere is an expected but less certainincrease in precipitation over thenortheast United States during fall andwinter during the same years (NMFS,2012b). In conjunction with increased evaporation from warmer temperatures, the Northeast and mid-Atlantic mayexperience decrease in runoff anddecreased stream flow in late winter andearly spring (NMFS, 2012b).Additionally, enhanced oceanstratification could be caused by greaterwarming at the ocean surface than atdepth (NMFS, 2012b).Many observed changes in riverherring biology related to environmental conditions were noted at the workshop, but few detailed analyses were available to distinguish climate change fromclimate variability.

One analysis byMassachusetts Division of MarineFisheries showed precipitation effectson spawning run recruitment atMonument River, MA (1980-2012; NMFS, 2012b). Jordaan and Kritzer(unpublished data) showed normalized run counts of alewife and bluebackherring have a stronger correlation withfisheries and predators than variousclimate variables at broad scales (NMFS,2012b). Once fine-scale (flow related tofishways and dams) data were used,results indicate that summer and fallconditions were more important.

Nye etal. (2012) investigated climate-related mechanisms in the marine habitat of theUnited States that may impact riverherring.

Their preliminary resultsindicate the following:

(1) A shift innorthern ocean distribution for bothblueback herring and alewife depending on the season; (2) decrease in oceanhabitat within the preferred temperature for alewife and blueback herring in thespring; and (3) effects of climate changeon river herring populations maydepend on the current condition (e.g.,

48958Federal Register/Vol.

78, No. 155/Monday, August 12, 2013/Notices abundance and health) of thepopulation, assumptions, andtemperature tolerances (e.g., bluebackherring have a higher temperature tolerance than alewife).

Although preliminary, Nye et a0.(2012) indicate that climate change willimpact river herring.

The results (alsosupported by Nye et al., 2009) indicatethat both blueback herring and alewifehave and will continue to shift theirdistribution to more northerly waters inthe spring, and blueback herring hasalso shifted its distribution to morenortherly waters in the fall (1975-2010)

(Nye et al., 2012). Additionally, Nye etal. (2012) found a decrease in habitat(bottom waters) within the preferred temperature for alewife and bluebackherring in the spring under futureclimate predictions (2020-2060 and2060-2100).

They concluded that anexpected decrease in optimal marinehabitat and natal spawning habitat willnegatively affect river herringpopulations at the southern extent oftheir range. Additionally, Nye et al.(2012) infer that this will have negativepopulation level effects and causepopulation declines in southern rivers,resulting in an observed shift indistribution which has already beenobserved.

Nye et al. (2012) also foundthat the effects of climate change onriver herring populations may dependon the current condition (e.g.,abundance and health) of thepopulation, assumptions, andtemperature tolerances.

Using themodel, projections of alewifedistribution and abundance can bepredicted for each year, but for ease ofinterpretation, 2 years of low and highrelative abundance were chosen toillustrate the effects of population abundance and temperature on alewifedistribution.

The low and highabundance years were objectively chosen as the years closest to -1 and+1 standard deviation from overallmean abundance.

Two years closest tothe -1 and +1 standard deviation frommean population abundance wereselected to reflect the combined effect ofwarming with low and high abundance of blueback herring.

The difference inspecies response (as noted below) mayreflect the different temperature tolerances (9-11 °C for blueback herringand 4-11 °C for alewife) as indicated bythe southern limit of their ranges.Blueback herring may be able to toleratehigher temperature as their rangeextends as far south as Florida, but thesouthern extent of the alewife's range islimited to North Carolina.

For bothspecies, the Nye et al. (2012) analysisindicates that, if robust populations ofthese species are maintained, declinesdue to the effects of climate change willbe reduced.

Their specific resultsinclude the following:

e Alewife:

At low population size,coast-wide abundance is projected todecrease with less suitable habitat andpatchy areas of high density in the Gulfof Maine and Georges Bank in 2060-2100. At high population size,abundance is projected to increaseslightly from 2020-2060

(+4.64 percent)but is projected to decrease

(- 39.14percent) and become more patchy in2060-2100.

e Blueback herring:

Abundance isprojected to increase at both high andlow population size throughout theNortheast United States, especially inthe mid-Atlantic and Georges Bank.However, at low abundance the increaseis minimal and remains at a level belowthe 40-year mean. The percentage change due to climate change (factoring only temperature) is +29.93 percent forthe time period 2020-2060 and +55.81percent from 2060-2100.

We hoped to obtain information during the workshop on potential impacts of climate change by region,including information on species, lifestage, indicators, potential

impacts, andavailable data/relevant references (NMFS, 2012b). Although we did obtaininformation on each of these categories, substantial data gaps in the speciesinformation were apparent (NMFS,2012b). For example, although nospecific information on impacts ofocean acidification on river herring waspresented, possible effects on larvaldevelopment, chemical signaling (olfaction),

and de-calcification of preywere noted (NMFS, 2012b). Additional research is needed to identify thelimiting factor(s) for river herringpopulations.

As Nye et al. (2012) noted,the links between climate and riverherring biology during freshwater stagesare unclear and will require additional time to research and thoroughly analyze.

This conclusion is supported by the results of the workshop, whichnoted numerous potential climateeffects on the freshwater stages, butlittle synthesis has been accomplished to date. The preliminary analysis of Nyeet al. (2012) indicates that watertemperatures in the rivers will bewarmer, and there will be a decrease inthe river flow in the northeast and Mid-Atlantic in late winter/early spring.Although current information indicates climate change is and willcontinue to impact river herring (e.g.,Nye et al., 2012), climate variability rather than climate change is expectedto have more of an impact on riverherring from 2024-2030.

Several studieshave shown that the climate changesignal is readily apparent by the end ofthe 21st century (Hare et al., 2010; Hareet al., 2012). At intermediate timeperiods (e.g., 2024-2030),

the signal ofnatural climate variability is likelysimilar to the signal of climate change.Thus, a large component of the climateeffect on river herring in 2024-2030 willbe composed of natural climatevariability, which could be eitherwarming or cooling.Summary and Evaluation of Factor ADams and hydropower facilities, water quality and water withdrawals from urbanization and agricultural runoff, dredging and other wetlandalterations are likely the causes ofhistorical and recent declines inabundance of alewife and bluebackherring populations.

Climate variability rather than climate change is expectedto have more of an impact on riverherring from 2024-2030 (NMFS'foreseeable future for river herring).

Nyeet al., (2012) conducted a preliminary analysis investigating climate-related mechanisms in the marine habitat of theUnited States that may impact riverherring, and found that changes in theamount of preferred habitat and apotential northward shift in distribution as a result of climate change may affectriver herring in the future (e.g., 2020-2100). Thus, the level of threat posed bythese potential stressors is evaluated further in the qualitative threatsassessment as described below.B. Overu tilization for Commercial, Recreational, Scientific, or Educational PurposesDirected Commercial HarvestThis following section on riverherring fisheries in the United States isfrom the stock assessment (ASMFC,2012).Fisheries for anadromous specieshave existed in the United States for avery long time. They not only providedsustenance for early settlers but a sourceof income as the fisheries werecommercialized.

It is difficult to fullydescribe the characteristics of theseearly fisheries because of the lack ofquantifiable data.The earliest commercial river herringdata were generally reported in stateand town reports or local newspapers.

In 1871, the U.S. Fish Commission wasfounded (later became known as theU.S. Fish and Fisheries Commission in1881). This organization collected fisheries statistics to characterize thebiological and economic aspects ofcommercial fisheries.

Data describing historical river herring fisheries were Federal Register/Vol.

78, No. 155/Monday, August 12, 2013/Notices 48959available from two of this organization's publications-the Bulletin of the U.S.Fish Commission (renamed FisheryBulletin in 1971; Collins and Smith,1890; Smith, 1891) and the U.S. FishCommission Annual Report (USFC,1888-1940).

In the stock assessment, theriver herring data were transcribed andwhen available, dollar values wereconverted to 2010 dollar values usingconversion factors based on the annualaverage consumer price index (CPI)values, which were obtained from theU.S. Bureau of Labor Statistics.

Notethat CPI values are not available foryears prior to 1913 so conversion factorscould not be calculated for years earlierthan 1913 (ASMFC, 2012).There are several caveats to using thehistorical fisheries data. There is anapparent bias in the area sampled.

Inmost cases, there was no systematic sampling of all fisheries; instead,sampling appeared to be opportunistic, concentrating on the mid-Atlantic States. It is also difficult to assess theaccuracy and precision of these data. Insome instances, the pounds werereported at a fine level of detail (e.g., atthe state/county/gear level), but detailsregarding the specific source of the datawere often not described.

The level ofdetail provided in the reports variedamong states and years. Additionally, not all states and fisheries werecanvassed in all years, so absence oflandings data does not necessarily indicate the fishery was not active as itis possible that the data just were notcollected.

For these reasons, thesehistorical river herring landings shouldnot be considered even minimum valuesbecause of the variation in detail andcoverage over the time series. Noattempt was made to estimate missingriver herring data since no benchmark or data characteristics could be found,and the stock assessment subcommittee also did not attempt to estimate missingdata in a time series at a particular location because of the bias associated with these estimates (ASMFC, 2012).During 1880 to 1938, reportedcommercial landings of river herringalong the Atlantic Coast averagedapproximately 30.5 million lbs (13,835mt) per year. The majority of riverherring landed by commercial fisheries in these early years are attributed to themid-Atlantic region (NY-VA).

Thedominance of the mid-Atlantic region is,in part, due to the apparent bias in thespatial coverage of the canvass (seeabove). From 1920 to 1938, the averageannual weight of reported commercial river herring landings was about 22.8million lbs (10,351 mt). The value of thecommercial river herring landingsduring this same time period wasapproximately 2.87 million dollars(2010 USD) (ASMFC, 2012).Domestic commercial landings ofriver herring were presented in the stockassessment by state and by gear from1887 to 2010 where available.

Landingsof alewife and blueback herring werecollectively classified as "river herring"by most states. Only a few states hadspecies-specific information recordedfor a limited range of years. Commercial landings records were available for eachstate since 1887 except for Florida andthe Potomac River Fisheries Commission (PRFC), which beganrecording landings in 1929 and 1960,respectively.

It is important to note thathistorical landings presented in thestock assessment do not include alllandings for all states over the entiretime period and are likelyunderestimated, particularly for the firstthird of the time series, since not allriver landings were reported (ASMFC,2012).Total domestic coast-wide landingsaveraged 18.5 million lb (8,399 mt) from1887 to 1928 (See table 2.2 in ASMFC(2012)).

During this early time period,landings were predominately fromMaryland, North Carolina,

Virginia, andMassachusetts (overall harvest is likelyunderestimated because landings werenot recorded consistently during thistime). Virginia made up approximately half of the commercial landings from1929 until the 1970s, and the majorityof Virginia's landings came from theChesapeake Bay, Potomac River, YorkRiver, and offshore harvest.

Coast-wide landings started increasing sharply inthe early 1940s and peaked at over 68.7million lb (31,160 mt) in 1958 (SeeTable 2.2, ASMFC, 2012). In the 1950sand 1960s, a large proportion of theharvest came from Massachusetts purseseine fisheries that operated offshore onGeorges Bank targeting Atlantic herring(G. Nelson, Massachusetts Division ofMarine Fisheries, Pers. comm., 2012).Landings from North Carolina were alsoat their highest during this time andoriginated primarily from the ChowanRiver pound net fishery.

Severe declinesin landings began coast-wide in theearly 1970s and domestic landings arenow a fraction of what they were at theirpeak, having remained at persistently low levels since the mid-1990s.

Moratoria were enacted inMassachusetts (commercial andrecreational in 2005), Rhode Island(commercial and recreational in 2006),Connecticut (commercial andrecreational in 2002), Virginia (forwaters flowing into North Carolina in2007), and North Carolina (commercial and recreational in 2007). As of January1, 2012, river herring fisheries in statesor jurisdictions without an approvedsustainable fisheries management plan,as required under ASMFC Amendment 2 to the Shad and River Herring FMP,were closed. As a result, prohibitions onharvest (commercial or recreational) were extended to the following states:New Jersey, Delaware, Pennsylvania,

Maryland, DC, Virginia (for all waters),Georgia and Florida (ASMFC, 2012).Pound nets were identified as thedominant gear type used to harvest riverherring from 1887 through 2010. Seineswere more prevalent prior to the 1960s,but by the 1980s, they were rarely used.Purse seines were used only for herringlanded in Massachusetts, but made upa large proportion of the landings in the1950s and 1960s. Historically, gill netsmade up a small percentage of theoverall harvest.

However, even thoughthe actual pounds landed continued todecline, the proportion of gill nets thatcontributed to the overall harvest hasincreased in recent years (ASMFC,2012).Foreign fleet landings of river herring(reported as alewife and blueback shad)are available through the Northwest Atlantic Fisheries Organization (NAFO).Offshore exploitation of river herringand shad (generally

<7.5 in (190 mm) inlength) by foreign fleets began in the late1960s and landings peaked at about 80million lbs (36,320 mt) in 1969(ASMFC, 2012).Total U.S. and foreign fleet harvest ofriver herring from the waters off thecoast of the United States (NAFO areas5 and 6) peaked at about 140 million lb(63,560 mt) in 1969, after whichlandings declined dramatically.

After1977 and the formation of the FisheryConservation Zone, foreign allocation ofriver herring (to both foreign vessels andjoint venture vessels) between 1977 and1980 was 1.1 million lb (499 mt). Theforeign allocation was reduced to220,000 lb (100 mt) in 1981 because ofthe condition of the river herringresource.

In 1985, a bycatch cap of nomore than 0.25 percent of total catchwas enacted for the foreign fishery.

Thecap was exceeded once in 1987, and thisshut down the foreign mackerel fishery.In 1991, area restrictions were passed toexclude foreign vessels from within 20miles (32.2 kin) of shore for two reasons:1) In response to the increased occurrence of river herring bycatchcloser to shore and 2) to promoteincreased fishing opportunities for thedomestic mackerel fleet (ASMFC, 2012).In-river Exploitation The stock assessment subcommittee calculated in-river exploitation rates ofthe spawning runs for five rivers(Damariscotta River (ME-alewife),

48960Federal Register/Vol.

78, No. 155/Monday, August 12, 2013/Notices Union River (ME-alewife),

MonumentRiver (MA-both species combined),

Mattapoisett River (MA-alewife),

andNemasket River (MA-alewife))

bydividing in-river harvest by total runsize (escapement plus harvest) for agiven year. Exploitation rates werehighest (range: 0.53 to 0.98) in theDamariscotta River and Union Riverprior to 1985, while exploitation waslowest (range: 0.26 to 0.68) in theMonument River. Exploitation declinedin all rivers through 1991 to 1992.Exploitation rates of both species in theMonument River and of alewives in theMattapoisett River and Nemasket Riverwere variable (average

= 0.16) and,except for the Nemasket River, declinedgenerally through 2005 until theMassachusetts moratorium wasimposed.

Exploitation rates of alewivesin the Damariscotta River were low(<0.05) during 1993 to 2000, but theyincreased steadily through 2004 andremained greater than 0.34 through2008. Exploitation in the Damariscotta dropped to 0.15 in 2009 to 2010.Exploitation rates of alewives in theUnion River declined through 2005 buthave remained above 0.50 since 2007(ASMFC, 2012).According to the stock assessment, exploitation of river herring appears tobe declining or remaining stable. In-river exploitation was highest in Mainerivers (Damariscotta and Union) and hasfluctuated, but it is currently lower thanlevels seen in the 1980s. Also, in-riverexploitation in Massachusetts rivers(Monument and Mattapoisett) wasdeclining at the time a moratorium wasimposed in 2005. The coast-wide indexof relative exploitation also declinedfollowing a peak in the late 1980s andhas remained fairly stable over the pastdecade. Exploitation rates declined inthe DB-SRA model runs except whenthe input biomass-to-K ratio in 2010 was0.01. Exploitation rates estimated fromthe statistical catch-at-age model forblueback herring in the Chowan River(see the NC state report in the stockassessment) also showed a slightdeclining trend from 1999 to 2007, atwhich time a moratorium wasinstituted.

There appears to be aconsensus among various assessment methodologies that exploitation hasdecreased in recent times. The stockassessment indicates that the decline inexploitation over the past decade is notsurprising because river herringpopulations are at low levels and morerestrictive regulations or moratoria havebeen enacted by states (ASMFC, 2012).Past high exploitation may also be areason for the high amount of variation and inconsistent patterns observed infisheries-independent indices ofabundance.

Fishing effort has beenshown to increase variation in fishabundance through truncation of the agestructure, and recruitment becomesprimarily governed by environmental variation (Hsieh et al., 2006; Andersonet al., 2008). When fish species are atvery low abundances, as is believed forriver herring, it is possible that the onlypopulation regulatory processes operating are stochastic fluctuations inthe environment (Shepherd andCushing, 1990) (ASMFC, 2012).Canadian HarvestFisheries in Canada for river herringare regulated through limited seasons,gears, and licenses.

Licenses may coverdifferent gear types; however, few newlicenses have been issued since 1993(DFO, 2001). River-specific management plans include closures and restrictions.

River herring used locally for bait inother fisheries are not accounted for inriver-specific management plans (DFO,2001). DFO estimated river herringlandings at just under 25.5 million lb(11,577 mt) in 1980, 23.1 million lb(10,487 mt) in 1988, and 11 million lb(4,994 mt) in 1996 (DFO, 2001). Thelargest river herring fisheries inCanadian waters occur in the Bay ofFundy, southern Gulf of Maine, NewBrunswick, and in the Saint John andMiramichi Rivers where annual harvestestimates often exceed 2.2 million lb(1,000 mt) (DFO, 2001). Recreational fisheries in Canada for river herring arelimited by regulations including area,gear and season closures with limits onthe number of fish that can be harvested per day; however, information onrecreational catch is limited.

Licensesand reporting are not required byCanadian regulations for recreational fisheries, and harvest is not welldocumented.

Incidental CatchThe following section on river herringincidental catch in the United States isfrom the stock assessment (ASMFC,2012).Three recent studies estimated riverherring discards and incidental catch(Cieri et al., 2008; Wigley et al., 2009;Lessard and Bryan, 2011). The discardand incidental catch estimates fromthese studies cannot be directlycompared as they used different ratioestimators based on data from theNortheast Fishery Observer Program(NEFOP),

as well as different raisingfactors to obtain total estimates.

Cieri etal. (2008) estimated the kept (i.e.,landed) portion of river herringincidental catch in the Atlantic herringfishery.

Cieri et al. (2008) estimated anaverage annual landed river herringcatch of approximately 71,290 lb (32.4mt) in the Atlantic herring fishery for2005-2007, and the corresponding coefficient of variation (CV) was 0.56.Cournane et al. (2010) extended thisanalysis with additional years of data.Further work is needed to elucidate howthe landed catch of river herring in thedirected Atlantic herring fisherycompares to total incidental catch acrossall fisheries.

Since this analysis onlyquantified kept river herring in theAtlantic herring fishery, itunderestimates the total catch (kept plusdiscarded) of river herring across allfishing fleets. Wigley et al. (2009)quantified river herring discards acrossfishing fleets that had sufficient observer coverage from July 2007-August 2008. Wigley et al. (2009)estimated that approximately 105,820 lb(48 mt) were discarded during the 12months (July 2007 to August 2008), andthe estimated precision was low (149percent CV). This analysis estimated only river herring discards (in contrastto total incidental catch), and noted thatmidwater trawl fleets generally retainedriver herring while otter trawls typically discarded river herring.Lessard and Bryan (2011) estimated an average incidental catch of riverherring and American shad of 3.3million lb (1,498 mt)/yr from 2000-2008. The methodology used in thisstudy differed from the Standardized Bycatch Reporting Methodology (SBRM)(the method used by NOAA's Northeast Fisheries Science Center (NEFSC) toquantify bycatch in stock assessments)

(Wigley et al., 2007; Wigley et al., 2012).Data from NEFOP were analyzed at thehaul level; however, the sampling unitfor the NEFOP database is at the triplevel. Within each gear and region, alldata, including those from high volumefisheries, appeared to be aggregated across years from 2000 through 2008.However, substantial changes in NEFOPsampling methodology for high volumefisheries were implemented in 2005,limiting the interpretability of estimates from these fleets in prior years. Totalnumber of tows from the fishing vesseltrip report (VTR) database was used asthe raising factor to estimate totalincidental catch. The use of effortwithout standardization makes theimplicit assumption that effort isconstant across all tows within a geartype, potentially resulting in a biasedeffort metric. In contrast, the total keptweight of all species is used as theraising factor in SBRM. Whenquantifying incidental catch acrossmultiple fleets, total kept weight of allspecies is an appropriate surrogate foreffective fishing power because it is Federal Register/Vol.

78, No. 155/Monday, August 12, 2013/Notices 48961likely that all trips will not exhibit thesame attributes.

Lessard and Bryan(2011) also did not provide precision estimates, which are imperative forestimation of incidental catch.The total incidental catch of riverherring was estimated as part of thework for Amendment 14 to the AtlanticMackerel, Squid and Butterfish (MSB)Fishery Management Plan, that includesmeasures to address incidental catch ofriver herring and shads, From 2005-2010, the total annual incidental catchof alewife ranged from 41,887 lb (19.0mt) to 1.04 million lb (472 mt) in NewEngland and 19,620 lb (8.9 mt) to564,818 lb (256.4 mt) in the Mid-Atlantic.

The dominant gear variedacross years between paired midwatertrawls and bottom trawls.Corresponding estimates of precision (COV) exhibited substantial interannual variation and ranged from 0.28 to 3.12across gears and regions.

Total annualblueback herring incidental catch from2005 to 2010 ranged from 30,643 lb(13.9 mt) to 389,111 lb (176.6 mt) inNew England and 2,645 lb (1.2 mt) to843,479 lb (382.9 mt} in the Mid-Atlantic.

Across years, paired and singlemidwater trawls exhibited the greatestblueback herring catches, with theexception of 2010 in the mid-Atlantic where bottom trawl was the mostdominant gear. Corresponding estimates of precision ranged from 0.27 to 3.65.The temporal distribution of incidental catches was summarized by quarter andfishing region for the most recent 6-yearperiod (2005 to 2010). River herringcatches occurred primarily in midwatertrawls (76 percent, of which 56 percentwere from paired midwater trawls andthe rest from single midwater trawls),followed by small mesh bottom trawls(24 percent).

Catches of river herring ingillnets were negligible.

Across geartypes, catches of river herring weregreater in New England (56 percent)than in the Mid-Atlantic (44 percent).

The percentages of midwater trawlcatches of river herring were similarbetween New England (37 percent) andthe Mid-Atlantic (38 percent).

However,catches in New England small meshbottom trawls were three times higher(18 percent) than those from the Mid-Atlantic (6 percent).

Overall, the highestquarterly catches of river herringoccurred in midwater trawls duringQuarter I in the Mid-Atlantic (35percent),

followed by catches in NewEngland during Quarter 4 (16 percent)and Quarter 3 (11 percent).

Quarterly catches in small mesh bottom trawlswere highest in New England duringQuarter 1 (7 percent) and totaled 3 to 4percent during each of the other threequarters.

Recreational HarvestThe Marine Recreational FisheryStatistics Survey (MRFSS) providedestimates of numbers of fish harvested and released by recreational fisheries along the Atlantic coast. The stockassessment subcommittee extracted state harvest and release estimates foralewives and blueback herring from theMRFSS catch and effort estimates filesavailable on the web (http://www.sefsc.noaa.gov/about/mrfss.htm).

Historically, there were few reports ofriver herring taken by recreational anglers for food. Most often, riverherring were taken for bait. MRFSSestimates of the numbers of river herringharvested and released by anglers arevery imprecise and show little trend.Thus, the stock assessment concluded that these data are not useful formanagement purposes.

MRFSSconcentrates their sampling strata incoastal water areas and does not captureany data on recreational fisheries thatoccur in inland waters. Few statesconduct creel surveys or otherconsistent survey instruments (diary orlog books) in their inland waters tocollect data on recreational catch ofriver herring.

Some data are reported inthe state chapters in the stockassessment; but the stock assessment committee concluded that data are toosparse to conduct any systematic comparison of trends (ASMFC, 2012).Scientific Monitoring and Educational HarvestMaine, New Hampshire, Massachusetts and Rhode Islandestimate run sizes using electronic counters or visual methods.

Variouscounting methods are used at theHolyoke Dam fish lift and fishways onthe Connecticut River. Young of year(YOY) surveys are conducted throughfixed seine surveys capturing YOYalewife and blueback herring generally during the summer and fall in Maine,Rhode Island, Connecticut, New York,New Jersey, Maryland, District ofColumbia, Virginia and North Carolina.

Rhode Island conducts surveys forjuvenile and adult river herring at largefixed seine stations.

Virginia samplesriver herring using a multi-panel gill netsurvey and electroshocking surveys.Florida conducts electroshocking surveys to sample river herring.

Maine,New Hampshire, Massachusetts, RhodeIsland, Maryland, and North Carolinacollect age data from commercial andfisheries independent samplingprograms, and length-at-age data. All ofthese scientific monitoring efforts arebelieved to have minimal impacts onriver herring populations.

Summary and Evaluation of Factor BHistorical commercial andrecreational fisheries for river herringlikely contributed to the decline inabundance of both alewife and bluebackherring populations.

Current directedcommercial and recreational alewifeand blueback herring fisheries, as wellas commercial fishery incidental catchmay continue to pose a threat to thesespecies.

Since the 1970s, regulations have been enacted in the United Stateson the directed harvest of river herringin an attempt to halt or reverse theirdecline with the most recent regulations being imposed in January 2012.Additionally, there are regulations inCanada on river herring harvest.Historical landings data and currentfishery effort is the best available information to describe the impact thatthe commercial fishery may be havingon river herring.Moratoria are in place on directedcatch of these species throughout mostof the United States; however, they aretaken as incidental catch in severalfisheries.

The extent to which incidental catch is affecting river herring has notbeen quantified and is not fullyunderstood.

Thus, the level of threatposed by directed and indirect catch isevaluated further in the qualitative threats assessment as described below.Scientific collections or collections foreducational purposes do not appear tobe significantly affecting the status ofriver herring, as they result in lowmortality.

C. Disease and Predation DiseaseLittle information exists on diseasesthat may affect river herring; however,there are reports of a variety of parasites that have been found in both alewifeand blueback herring.

The mostcomprehensive report is that of Landryet al. (1992) in which 13 species ofparasites were identified in bluebackherring and 12 species in alewives fromthe Miramichi River, New Brunswick, Canada. The parasites found includedone monogenetic trematode, fourdigenetic trematodes, one cestode, threenematodes, one acanthocephalan, oneannelid, one copepod and one mollusk.The same species were found in bothalewife and blueback herring with theexception of the acanthocephalan, which was absent from alewives.

In other studies, Sherburne (1977)reported piscine erythrocytic necrosis(PEN) in the blood of 56 percent ofprespawning and 10 percent of 48962Federal Register/Vol.

78, No. 155/Monday, August 12, 2013/Notices postspawning alewives in Maine coastalstreams.

PEN was not found in juvenilealewives from the same locations.

Coccidian parasites were found in thelivers of alewives and other finfish offthe coast of Nova Scotia (Morrison andMarryatt, 1990). Marcogliese andCompagna (1999) reported that mostfish species, including

alewife, in the St.Lawrence River become infected withtrematode metacercariae during the firstyears of life. Examination of Great Lakesfishes in Canadian waters showed larvalDiplostomum (trematode) commonly inthe eyes of alewife in Lake Superior(Dechtiar and Lawrie, 1988) and LakeOntario (Dechtiar and Christie, 1988),though intensity of infections was low(<9/host).

Heavy infections ofSaprolegnia, a fresh and brackish waterfungus, were found in 25 percent ofLake Superior alewife examined, andlight infections were found in 33percent of Lake Ontario alewife(Dechtiar and Lawrie, 1988). Larvalacanthocephala were also found in theguts of alewife from both lakes.Saprolegnia typically is a secondary infection, invading open sores andwounds, and eggs in poorenvironmental conditions, but under theright conditions it can become a primarypathogen.

Saprolegnia infections usually are lethal to the host.More recently, alewives were foundpositive for Cryptosporidium for thefirst time on record by Ziegler et al.(2007). Mycobacteria, which can resultin ulcers, emaciation, and sometimes death, have been found in manyChesapeake Bay fish, including blueback herring (Stine et al., 2010).Predation Information on predation of riverherring was compiled and published inVolume I of the River HerringBenchmark Assessment (2012) byASMFC. The following section onpredation was compiled by Dr. KatieDrew from this assessment.

Alewife and blueback herring are animportant forage fish for marine andanadromous predators, such as stripedbass, spiny dogfish,

bluefish, Atlanticcod, and pollock (Bowman et al., 2000;Smith and Link, 2010). Historically, river herring and striped bass landingshave tracked each other quite well, withhighs in the 1960s, followed by declinesthrough the 1970s and 1980s. Althoughpopulations of Atlantic cod and pollockare currently low, the populations ofstriped bass and spiny dogfish haveincreased in recent years (since the early1980s for striped bass and since 2005 forspiny dogfish),

while the landings andrun counts of river herring remain athistorical lows. This has led tospeculation that increased predation may be contributing to the decline ofriver herring and American shad(Hartman, 2003; Crecco et al., 2007;Heimbuch, 2008). Quantifying theimpacts of predation on alewife andblueback herring is difficult.

The diet ofstriped bass has been studiedextensively, and the prevalence ofalosines varies greatly depending onlocation, season, and predator size(Walter et al., 2003). Studies from thenortheast U.S. continental shelf showlow rates of consumption by stripedbass (alewife and blueback herring eachmake up less than 5 percent of stripedbass diet by weight) (Smith and Link,2010), while studies that sampledstriped bass in rivers and estuaries during the spring spawning runs foundmuch higher rates of consumption (greater than 60 percent of striped bassdiet by weight in some months and sizeclasses)

(Walter and Austin, 2003;Rudershausen et al., 2005). Translating these snapshots of diet composition intoestimates of total removals requiresadditional data on both annual percapita consumption rates and estimates of annual abundance for predatorspecies.The diets of other predators, including other fish (e.g., bluefish, spinydogfish),

along with marine mammals(e.g., seals) and birds (e.g., double-crested cormorant),

have not beenquantified nearly as extensively, makingit more difficult to assess theimportance of river herring in thefreshwater and marine food webs. As aresult, some models predict a significant negative effect from predation (Hartman, 2003; Heimbuch, 2008), while otherstudies did not find an effect(Tuomikoski et al., 2008; Dalton et al.,2009).In addition to predators native to theAtlantic coast, river herring arevulnerable to invasive species such asthe blue catfish (Ictalurusfurcatus) andthe flathead catfish (Pylodictis olivaris).

These catfish are large, opportunistic predators native to the Mississippi Riverdrainage that were introduced intorivers on the Atlantic coast. They havebeen observed to consume a wide rangeof species, including

alosines, andecological modeling on flathead catfishsuggests they may have a large impacton their prey species (Pine, 2003;Schloesser et al., 2011). In August 2011,ASMFC approved a resolution callingfor efforts to reduce the population sizeand ecological impacts of invasivespecies and named blue and flatheadcatfish specifically, as species ofconcern, due to their increasing abundance and potential impacts onnative anadromous species.

Non-native species are a particular concern becauseof the lack of native predators, parasites, and competitors to keep theirpopulations in check.Predation and multispecies models,such as the MS-VPA (NEFSC, 2006),have tremendous data needs, and moreresearch needs to be conducted beforethey can be applied to river herring.However, given the potential magnitude of predatory interactions, it is an area ofresearch worth pursuing (ASMFC,2012).Two papers have become available since the ASMFC (2012) stockassessment that discuss striped basspredation on river herring inMassachusetts and Connecticut estuaries and rivers, showing temporaland spatial patterns in predation (Daviset al., 2012; Ferry and Mather, 2012).Davis et al. (2012) estimated thatapproximately 400,000 blueback herringare consumed annually by striped bassin the Connecticut River springmigration.

In this study, striped basswere found in the rivers during thespring spawning migrations of bluebackherring and had generally left thesystem by mid-June (Davis et al., 2012).Many blueback herring in theConnecticut River are thought to beconsumed prior to ascending the riveron their spawning migration, and are,therefore, being removed from thesystem before spawning.

Alternatively, Ferry and Mather (2012) discuss theresults of a similar study conducted inMassachusetts watersheds withdrastically different findings for stripedbass predation.

Striped bass werecollected and stomach contentsanalyzed during three seasons from Maythrough October (Ferry and Mather,2012). The stomach contents of stripedbass from the survey were examinedand less than 5 percent of the clupeidcategory (from 12 categories identified to summarize prey) consisted ofanadromous alosines (Ferry and Mather,2012). Overall, the Ferry and Mather(2012) study observed few anadromous alosines in the striped bass stomachcontents during the study period. Thesetwo recent studies echo similarcontradictory findings from previousstudies showing a wide variation inpredation by striped bass with spatialand temporal effects;

however, theyexhibit no consistent trends along thecoast.Summary and Evaluation for Factor CWhile data are limited, the bestavailable information indicates thatriver herring are not likely affected to alarge degree by diseases caused byviruses,

bacteria, protozoans, metazoans, or microalgae.

Much of the Federal Register/Vol.

78, No. 155/Monday, August 12, 2013/Notices 48963information on diseases in alewife orblueback herring comes from studies onlandlocked species; therefore, even ifstudies indicated that landlocked alewife and blueback herring werehighly susceptible to diseases andsuffered high mortality rates, it is notknown whether anadromous riverherring would be affected in the sameway. While it may be possible thatdisease threats to river herring couldincrease in prevalence or magnitude under various climate change scenarios, there are currently no data available tosupport this supposition.

We haveincluded disease as a threat in thequalitative threats assessment described in detail below.Alewife and blueback herring areconsidered to be an important foragefish for many marine and anadromous predators, and therefore, may beaffected by predation, especially if somepopulations of predators (e.g., stripedbass, spiny dogfish) continue toincrease.

There may also be effects frompredation by invasive species such asthe blue and flathead catfish.

Somepredation and multispecies models haveestimated an effect of predation on riverherring, while others have not. Ingeneral, the effect of predation on thepersistence of river herring is not fullyunderstood;

however, predation may beaffecting river herring populations andconsequently, it is included as a threatin the qualitative threats assessment described below.D. Inadequacy of Existing Regulatory Mechanisms As wide-ranging anadromous species,alewife and blueback herring are subjectto numerous Federal (U.S. andCanadian),

state and provincial, Tribal,and inter-jurisdictional laws,regulations, and agency activities.

Theseregulatory mechanisms are described indetail in the following section.International The Canadian DFO manages alewifeand blueback herring fisheries thatoccur in the rivers of the CanadianMaritimes under the Fisheries Act(R.S.C.,

1985, c. F-14). The MaritimeProvinces Fishery Regulations includesrequirements when fishing for orcatching and retaining river herring inrecreational and commercial fisheries (DFO, 2006; http://laws-lois.justice.gc.ca).

Commercial and recreational riverherring fisheries in the CanadianMaritimes are regulated by license,fishing gear, season and/or othermeasures (DFO, 2001). Since 1993, DFOhas issued few new licenses for riverherring (DFO, 2001). River herring areharvested by various gear types (e.g.,gillnet, dip nets, trap) and theregulations depend upon the river andassociated location (DFO, 2001). Theprimary management measures areweekly closed periods and limiting thenumber of licenses to existing levels inall areas (DFO, 2001). Logbooks areissued to commercial fishermen in someareas as a condition of the license, andpilot programs are being considered inother areas (DFO, 2001). Themanagement objective is to maintainharvest near long-term mean levelswhen no specific biological andfisheries information is available (DFO,2001).DFO (2001) stated that additional management measures may be requiredif increased effort occurs in response tostock conditions or favorable markets.There has been concern as fisheryexploitation rates have been abovereference levels and fewer licenses arefished than have been issued (DFO,2001). In 2001, DFO reported that insome rivers river herring were beingharvested at or above reference levels(e.g., Miramichi),

while in other riversriver herring were harvested at or belowthe reference point (e.g., St. John Riverat Mactaquac Dam). DFO (2001) believesprecautionary management involving noincrease or decrease in exploitation isimportant for Maritime river herringfisheries, given that biological andharvest data are not widely available.

Additionally, DFO (2001) added thatriver-specific management plans basedon stock assessments should beprioritized over general management initiatives.

Eastern New Brunswick is currently the only area in the Canadian Maritimes with a river herring integrated fisherymanagement plan (DFO, 2006). TheDFO uses Integrated Fisheries Management Plans (IFMPs) to guide theconservation and sustainable use ofmarine resources (DFO, 2010). An IFMPmanages a fishery in a given region bycombining the best available science onthe species with industry data oncapacity and methods for harvesting (DFO, 2010). The 6-year management plan (2007-2012) for river herring forEastern New Brunswick is implemented in conjunction with annual updates tospecific fishery management measures(e.g., seasons).

For example, it notes amanagement problem of gear congestion in some rivers and an approach toestablish a carrying capacity of the riverand find a solution to the gear limit byworking with fishermen (DFO, 2006). Atthis time, an updated Eastern NewBrunswick IFMP is not available.

FederalASMFC and Enabling Legislation Authorized under the terms of theAtlantic States Marine Fisheries

Compact, as amended (Pub. L.81-721),the purpose of the ASMFC is to promotethe better utilization of the fisheries (marine, shell, and anadromous) of theAtlantic seaboard "by the development of a joint program for the promotion andprotection of such fisheries, and by theprevention of the physical waste of thefisheries from any cause."Given management authority in 1993under the Atlantic Coastal Fisheries Cooperative Management Act (16 U.S.C.5101-5108),

the ASMFC may issueinterstate FMPs that must beadministered by state agencies.

If theASMFC believes that a state is not incompliance with a coastal FMP, it mustnotify the Secretaries of Commerce andInterior.

If the Secretaries find the statenot in compliance with the management plan, the Secretaries must declare amoratorium on the fishery in question.

Atlantic Coastal Fisheries Cooperative Management ActWe manage river herring stocks underthe authority of section 803(b) of theAtlantic Coastal Fisheries Cooperative Management Act (Atlantic Coastal Act)16 U.S.C. 5101 et seq., which states, inthe absence of an approved andimplemented FMP under the Magnuson-Stevens Act (MSA, 16 U.S.C. 1801 etseq.) and, after consultation with theappropriate Fishery Management Council(s),

the Secretary of Commercemay implement regulations to governfishing in the Exclusive Economic Zone(EEZ), i.e., from 3 to 200 nautical mi(nm) offshore.

The regulations must be:(1) Compatible with the effective implementation of an Interstate FisheryManagement Plan for American Shadand River Herring (ISFMP) developed by the ASMFC; and (2) consistent withthe national standards set forth insection 301 of the MSA.The ASMFC adopted Amendment 2 tothe ISFMP in 2009. Amendment 2establishes the foundation for riverherring management.

It was developed to address concerns that many Atlanticcoast populations of river herring werein decline or are at depressed but stablelevels, and that the ability to accurately assess the status of river herring stocksis complicated by a lack of fisheryindependent data.Amendment 2 requires states to closetheir waters to recreational andcommercial river herring harvest, unlessthey have an approved sustainable management plan in place. To beapproved, a state's plan must clearly 48964Federal Register/Vol.

78, No. 155/Monday, August 12, 2013/Notices meet the Amendment's standard of asustainable fishery defined as "acommercial and/or recreational fisherythat will not diminish the potential future stock reproduction andrecruitment."

The plans must meet thedefinition of sustainability bydeveloping and maintaining sustainability targets.

States without anapproved plan were required to closetheir respective river herring fisheries asof January 1, 2012, until such a plan issubmitted and approved by theASMFC's Shad and River HerringManagement Board. Proposals to re-open closed fisheries may be submitted annually as part of a state's annualcompliance report. Currently, the statesof ME, NH, RI, NY, NC, and SC haveapproved river herring management plans (see "State section of Factor D" formore information).

In addition to the state sustainability plan mandate, Amendment 2 makesrecommendations to states for theconservation, restoration, and protection of critical river herring habitat.

TheAmendment also requires states toimplement fisheries-dependent andindependent monitoring

programs, toprovide critical data for use in futureriver herring stock assessments.

While these measures addressproblems to the river herringpopulations in coastal areas, incidental catch in small mesh fisheries, such asthose for sea herring, occurs outsidestate jurisdiction and remains asubstantial source of fishing mortality according to the ASMFC. Consequently, the ASMFC has requested that the NewEngland and Mid-Atlantic FisheryManagement Councils (NEFMC andMAFMC) increase efforts to monitorriver herring incidental catch in small-mesh fisheries (See section on "NEFMCand MAFMC recommendations forfuture river herring bycatch reduction efforts").

Magnuson-Stevens FisheryConservation and Management Act(MSA)The Magnuson-Stevens FisheryConservation and Management Act(MSA) is the primary law governing marine fisheries management in Federalwaters. The MSA was first enacted in1976 and amended in 1996 and 2006.Most notably, the MSA aided in thedevelopment of the domestic fishingindustry by phasing out foreign fishing.To manage the fisheries and promoteconservation, the MSA created eightregional fishery management councils.

A 1996 amendment focused onrebuilding overfished fisheries, protecting Essential Fish Habitat (EFH),and reducing bycatch.

A 2006amendment mandated the use ofAnnual Catch Limits (ACL) andAccountability Measures (AM) to endoverfishing, provided for widespread market-based fishery management through limited access privilege

programs, and called for increased international cooperation.

The MSA requires that Federal FMPscontain conservation and management measures that are consistent with theten National Standards.

NationalStandard

1. 9 states that conservation andmanagement measures shall, to theextent practicable, (A) minimize bycatchand (B) to the extent bycatch cannot beavoided, minimize the mortality of suchbycatch.

The MSA defines bycatch asfish that are harvested in a fishery, butwhich are not sold or kept for personaluse. This includes economic discardsand regulatory discards.

River herring isencountered both as bycatch andincidental catch in Federal fisheries.

While there is no directed fishery forriver herring in Federal waters, riverherring co-occur with other species thathave directed fisheries (Atlantic

mackerel, Atlantic

herring, whiting,squid and butterfish) and are eitherdiscarded or retained in those fisheries.

Essential Fish Habitat Under the MSAUnder the MSA, there is arequirement to describe and identifyEFH in each Federal FMP. EFH isdefined as ". ..those waters andsubstrate necessary to fish for spawning,

breeding, feeding, or growth tomaturity."

The rules promulgated by theNMFS in 1997 and 2002 further clarifyEFH with the following definitions:

(1)Waters-aquatic areas and theirassociated

physical, chemical, andbiological properties that are used byfish and may include aquatic areashistorically used by fish whereappropriate; (2) substrate-sediment, hard bottom, structures underlying thewaters, and associated biological communities; (3) necessary-the habitatrequired to support a sustainable fisheryand the managed species' contribution to a healthy ecosystem; and (4)spawning,

breeding, feeding, or growthto maturity-stages representing aspecies' full life cycle.EFH has not been designated foralewife or blueback

herring, though EFHhas been designated for numerous otherspecies in the Northwest Atlantic.

Measures to improve habitats andreduce impacts resulting from thoseEFH designations may directly orindirectly benefit river herring.Conservation measures implemented inresponse to the designation of Atlanticsalmon EFH and Atlantic herring EFHlikely provide the most conservation benefit to river herring over any otherEFH designation.

Habitat features usedfor spawning,

breeding, feeding, growthand maturity by these two speciesencompasses many of the habitatfeatures selected by river herring tocarry out their life history.

Thegeographic range in which river herringmay benefit from the designation ofAtlantic salmon EFH extends fromConnecticut to the Maine/Canada border. The geographic range in whichriver herring may benefit from thedesignation of Atlantic herring EFHdesignation extends from the Maine/Canada border to Cape Hatteras.

The Atlantic salmon EFH includesmost freshwater, estuary and bayhabitats historically accessible toAtlantic salmon from Connecticut to theMaine/Canada border (NEFMC, 2006).Many of the estuary, bay and freshwater habitats within the current andhistorical range of Atlantic salmonincorporate habitats used by riverherring for spawning, migration andjuvenile rearing.

Among Atlanticherring EFHs are the pelagic waters inthe Gulf of Maine, Georges Bank,Southern New England, and middleAtlantic south to Cape Hatteras out tothe offshore U.S. boundary of the EEZ(see NEFMC 1998). These areasincorporate nearly all of the U.S. marineareas most frequently used by riverherring for growth and maturity.

Subsequently, in areas where EFHdesignations for Atlantic salmon andAtlantic herring overlap with freshwater and marine habitats used by riverherring, conservation benefits affordedthrough the designation of EFH for thesespecies may provide similar benefits toriver herring.Federal Power Act (FPA) (16 U.S.C.791-828) and Amendments The FPA, as amended, provides forprotecting, mitigating damages to, andenhancing fish and wildlife resources (including anadromous fish) impactedby hydroelectric facilities regulated bythe Federal Energy and Regulatory Commission (FERC). Applicants mustconsult with state and Federal resourceagencies who review proposedhydroelectric projects and makerecommendations to FERC concerning fish and wildlife and their habitat, e.g.,including spawning

habitat, wetlands, instream flows (timing, quality,quantity),

reservoir establishment andregulation, project construction andoperation, fish entrainment andmortality, and recreational access.Section 10(j) of the FPA provides thatlicenses issued by FERC containconditions to protect, mitigate damagesto, and enhance fish and wildlife based Federal Register/Vol.

78, No. 155/Monday, August 12, 2013/Notices 48965on recommendations received from stateand Federal agencies during thelicensing process.

With regard to fishpassage, Section 18 requires a FERClicensee to construct,

maintain, andoperate fishways prescribed by theSecretary of the Interior or the Secretary of Commerce.

Under the FPA, othersmay review proposed projects and maketimely recommendations to FERC torepresent additional interests.

Interested parties may intervene in the FERCproceeding for any project to receivepertinent documentation and to appealan adverse decision by FERC.While the construction ofhydroelectric dams contributed to somehistorical losses of river herringspawning

habitat, only a few new damshave been constructed in the range ofthese species in the last 50 years. Insome areas, successful fish passage hasbeen created; thus, restoring access tomany habitats once blocked.

Thus, riverherring may often benefit from FPAfishway requirements whenprescriptions are made to addressanadromous fish passage and during there-licensing of existing hydroelectric dams when anadromous species areconsidered.

Anadromous Fish Conservation Act (16U.S.C. 757a-757f) as AmendedThis law authorizes the Secretaries ofInterior and Commerce to enter into costsharing with states and other non-Federal interests for the conservation, development, and enhancement of thenation's anadromous fish.Investigations, engineering, biological

surveys, and research, as well as theconstruction, maintenance, andoperations of hatcheries, are authorized.

This Act was last authorized in 2002,which provided 5 million dollars for thefiscal years 2005 and 2006 (Pub. L. 107-372). There was an attempt toreauthorize the Act in 2012; however,this action has not yet been authorized.

Fish and Wildlife Coordination Act(FWCA) (16 U.S.C. 661-666)The FWCA is the primary lawproviding for consideration of fish andwildlife habitat values in conjunction with Federal water development activities.

Under this law, theSecretaries of Interior and Commercemay investigate and advise on theeffects of Federal water development projects on fish and wildlife habitat.Such reports and recommendations, which require concurrence of the statefish and wildlife agency(ies)

involved, must accompany the construction agency's.

request for congressional authorization, although the construction agency is not bound by therecommendations.

The FWCA applies to water-related activities proposed by non-Federal entities for which a Federal permit orlicense is required.

The most significant permits or licenses required are Section404 and discharge permits under theClean Water Act and Section 10 permitsunder the Rivers and Harbors Act. TheUSFWS and NMFS may review theproposed permit action and makerecommendations to the permitting agencies to avoid or mitigate anypotential adverse effects on fish andwildlife habitat.

Theserecommendations must be given fullconsideration by the permitting agency,but are not binding.Federal Water Pollution Control Act,and amendments (FWPCA) (33 U.S.C.1251-1376)

Also called the "Clean Water Act,"the FWPCA mandates Federalprotection of water quality.

The law alsoprovides for assessment of injury,destruction, or loss of natural resources caused by discharge of pollutants.

Of major significance is Section 404 ofthe FWPCA, which prohibits thedischarge of dredged or fill material intonavigable waters without a permit.Navigable waters are defined under theFWPCA to include all waters of theUnited States, including the territorial seas and wetlands adjacent to suchwaters. The permit program isadministered by the Army Corps ofEngineers (ACOE). The Environmental Protection Agency (EPA) may approvedelegation of Section 404 permitauthority for certain waters (notincluding traditional navigable waters)to a state agency; however, the EPAretains the authority to prohibit or denya proposed discharge under Section 404of the FWPCA.The FWPCA (Section 401) alsoauthorizes programs to remove or limitthe entry of various types of pollutants into the nation's waters. A point sourcepermit system was established by theEPA and is now being administered atthe state level in most states. Thissystem, referred to as the NationalPollutant Discharge Elimination System(NPDES),

sets specific limits ondischarge of various types of pollutants from point source outfalls.

A non-point source control program focusesprimarily on the reduction ofagricultural siltation and chemicalpollution resulting from rain runoff intothe nation's streams.

This effortcurrently relies on the use of landmanagement practices to reduce surfacerunoff through programs administered primarily by the Department ofAgriculture.

Like the Fish and WildlifeCoordination and River and HarborsActs, Sections 401 and 404 of theFWPCA have played a role in reducingdischarges of pollutants, restricting thetiming and location of dredge and filloperations, and affecting other changesthat have improved river herring habitatin many rivers and estuaries over thelast several decades.

Examples includereductions in sewage discharges into theHudson River (A. Kahnle, New YorkState DEC, Pers. comm. 1998) andnutrient reduction strategies implemented in the Chesapeake Bay (R.St. Pierre, USFWS, Pers. comm. 1998).Rivers and Harbors Act of 1899Section 10 of the Rivers and HarborsAct requires a permit from the ACOE toplace structures in navigable waters ofthe United States or modify a navigable stream by excavation or filling activities.

National Environmental Policy Act of1969 (NEPA) (42 U.S.C. 4321-4347)

The NEPA requires an environmental review process of all Federal actions.This includes preparation of anenvironmental impact statement formajor Federal actions that may affect thequality of the human environment.

Lessrigorous environmental assessments arereviewed for most other actions, whilesome actions are categorically excludedfrom formal review. These reviewsprovide an opportunity for the agencyand the public to comment on projectsthat may impact fish and wildlifehabitat.Coastal Zone Management Act (16U.S.C. 1451-1464) and Estuarine AreasActCongress passed policy on values ofestuaries and coastal areas through theseActs. Comprehensive planningprograms, to be carried out at the statelevel, were established to enhance,protect, and utilize coastal resources.

Federal activities must comply with theindividual state programs.

Habitat maybe protected by planning and regulating development that could cause damageto sensitive coastal habitats.

Federal Land Management and OtherProtective Designations Protection and good stewardship oflands and waters managed by Federalagencies, such as the Departments ofDefense, Energy and Interior (National Parks and National Wildlife

Refuges, aswell as state-protected park, wildlifeand other natural areas), contributes tothe health of nearby aquatic systemsthat support important river herring 48966Federal Register

/ Vol. 78, No. 155 / Monday, August 12, 2013 / Noticesspawning and nursery habitats.

Relevantexamples include the Great Bay, RachelCarson's and ACE Basin NationalEstuarine Research

Reserves, Department of Defense properties in theChesapeake Bay, and many NationalWildlife Refuges.Marine Protection, Research andSanctuaries Act of 1972 (MPRSA),

TitlesI and III and the Shore Protection Act of1988 (SPA)The MPRSA protects fish habitatthrough establishment and maintenance of marine sanctuaries.

The MPRSA andthe SPA regulate ocean transportation and dumping of dredge materials, sewage sludge, and other materials.

Criteria that the ACOE uses for issuingpermits include considering the effectsdumping has on the marineenvironment, ecological systems andfisheries resources.

Atlantic Salmon ESA Listing andCritical Habitat Designation In 2009, the Gulf of Maine (GOM) DPSof Atlantic salmon was listed asendangered under the Endangered Species Act (74 FR 29344). The GOMDPS includes all anadromous Atlanticsalmon whose freshwater range occursin the watersheds from theAndroscoggin River northward alongthe Maine coast to the Dennys River.Concurrently in 2009, critical habitatwas designated for the Atlantic salmonGOM DPS pursuant to section 4(b)(2) ofthe ESA (74 FR 29300; August 10, 2009).The critical habitat designation includes45 specific areas occupied by Atlanticsalmon at the time of listing, andincludes approximately 12,160 miles(19,600 kin) of perennial river, stream,and estuary habitat and 308 squaremiles (495 sq kin) of lake habitat withinthe range of the GOM DPS in the Stateof Maine.Measures to improve habitats andreduce impacts to Atlantic salmon as aresult of the ESA listing may directly orindirectly benefit river herring.

Atlanticsalmon are anadromous and spend aportion of their life in freshwater andthe remaining portion in the marineenvironment.

River herring occupy a lotof the same habitats as listed Atlanticsalmon for spawning,

breeding, feeding,growth and maturity.

Therefore, protection measures such as improvedfish passage or reduced discharge permits may benefit river herring.The critical habitat designation provides additional protections beyondclassifying a species as endangered bypreserving the physical and biological features essential for the conservation ofthe species in designated waters inMaine. One of the biological featuresidentified in the critical habitatdesignation for Atlantic salmon wasfreshwater and estuary migration siteswith abundant, diverse native fishcommunities to serve as a protective buffer against predation.

Co-evolved diadromous fish species such asalewives and blueback herring areincluded in this native fish community.

Because the ESA also requires that anyFederal agency that funds, authorizes, orcarries out an action ensure that theaction does not adversely modify ordestroy designated critical

habitat, theimpacts to alewife and blueback herringpopulations must be considered duringconsultation with NMFS to ensure thatAtlantic salmon critical habitat is notadversely affected by a Federal action.Atlantic Sturgeon ESA ListingIn 2012, five distinct population segments of Atlantic sturgeon werelisted under the ESA (77 FR 5914; 77 FR5880). The Chesapeake Bay, New YorkBight, Carolina, and South AtlanticDPSs of Atlantic sturgeon are listed asendangered, while the Gulf of MaineDPS is listed as threatened.

Measures to improve habitats andreduce impacts to Atlantic sturgeon maydirectly or indirectly benefit riverherring.

Atlantic sturgeon areanadromous; adults spawn in freshwater in the spring and early summer andmigrate into estuarine and marinewaters where they spend most of theirlives. As with Atlantic salmon, many ofthe habitats that Atlantic sturgeonoccupy are also habitats that riverherring use for spawning, migration andjuvenile rearing.

The geographic rangein which river herring may benefit fromAtlantic sturgeon ESA protections extends from the Maine/Canada borderto Florida.

Therefore, any protection measures within this range such asimproved fish passage or a reduction ofwater withdrawals may also provide abenefit to river herring.State Regulations A historical review of stateregulations was compiled and published in Volume I of the stock assessment.

The following section on stateregulations includes currentrequirements only and is cited fromVolume I of the assessment as compiledby Dr. Gary Nelson and Kate Taylor(ASMFC, 2012). Otherwise, updates areprovided by Kate Taylor, supplemental information from state river herringplans or state regulations.

MaineIn Maine, the Department of MarineResources (DMR), along withmunicipalities granted the rights toharvest river herring resources, cooperatively manage municipal fisheries.

Each town must submit anannual harvesting plan to DMR forapproval that includes a 3-day per weekescapement period or biological equivalent to ensure conservation of theresource.

In some instances, anescapement number is calculated andthe harvester passes a specific numberupstream to meet escapement goals.River herring runs not controlled by amunicipality and not approved assustainable by the ASMFC River Herringand American Shad Management Board,as required under Amendment 2, areclosed. Each run and harvest location isunique, either in seasonality, fishcomposition, or harvesting limitations.

Some runs have specific management plans that require continuous escapement and are more restrictive than the 3-day closed period. Othershave closed periods shorter than the 3-day requirement, but require anescapement number, irrespective of thenumber harvested during the season.Maine increased the weekly fishingclosure from a 24-hour closure in the1960s to a 48-hour closure beginning in1988. The closed period increased to 72hours beginning in 1995 to protectspawning fish. Most towns operate aweir at one location on each stream andprohibit fishing at any other location onthe stream. The state landings programcompiles in-river landings of riverherring from mandatory reportsprovided by the municipality undereach municipal harvest plan or they loseexclusive fishing rights. The statepermitted 22 municipalities to fish forriver herring in 2011. The river specificmanagement plans require theremaining municipalities to close theirruns for conservation and not harvest.There are several reasons for these state/municipal imposed restrictions on thefishery.

Many municipalities voluntarily restrict harvest to increase the numbersof fish that return in subsequent years.Some of these runs are large but havethe potential to become even larger. Thecommercial fishery does not exploit theestimated 1.5 to 2.0 million river herringthat return to the East Machias Riverannually.

These regulations have beenapproved through a sustainable fisheries management plan, as required underASMFC Amendment 2 to the Shad andRiver Herring FMP (Taylor, Pers.Comm., 2013).Recreational fishermen are allowed tofish for river herring year-round.

Thelimit is 25 fish per day and gear isrestricted to dip net and hook-and-line.

Recreational fishermen may not fish inwaters, or in waters upstream, of a Federal Register/Vol.

78, No. 155/Monday, August 12, 2013/Notices 48967municipality that owns fishing rights.Recreational fishermen are not requiredto report their catch. The MRFSS andMRIP programs do sample some of thesefishermen based on results queried fromthe database.

Recreational fishing forriver herring in Maine is limited andlandings are low. These regulations have been approved through asustainable fisheries management plan,as required under ASMFC Amendment 2 to the Shad and River Herring FMP(Taylor, Pers. Comm., 2013).New Hampshire The current general regulations are:(1) No person shall take river herring,alewives and blueback

herring, from thewaters of the state, by any method,between sunrise Wednesday and sunriseThursday of any week; (2) any trap orweir used during a specified timeperiod, shall be constructed so as toallow total escapement of all riverherring; and (3) any river herring takenby any method during the specified timeperiod shall be immediately releasedback into the waters from which it wastaken. Specific river regulations are:Taylor River-from the railroad bridgeto the head of tide dam in Hamptonshall be closed to the taking of riverherring by netting of any method; andSquamscott River-during April, Mayand June, the taking of river herring inthe Squamscott River and its tributaries from the Rt. 108 Bridge to the Great Damin Exeter is open to the taking of riverherring by netting of any method onlyon Saturdays and Mondays, the dailylimit shall be one tote per person ("tote"means a fish box or container measuring 31.5 in (80.01cm) x 18 in (45.72 cm) x11.5 in (29.21cm))

and the tote shallhave the harvester's coastal harvestpermit number plainly visible on theoutside of the tote. These regulations have been approved through asustainable fisheries management plan,as required under ASMFC Amendment 2 to the Shad and River Herring FMP.Massachusetts As of January 1, 2012, commercial and recreational harvest of river herringwas prohibited in Massachusetts, asrequired by ASMFC Amendment 2 tothe Shad and River Herring FMP(Taylor, Pers. Comm., 2013). Theexception is for federally permitted vessels which are allowed to land up to5 percent of total bait fish per trip(Taylor, Pers. Comm., 2013).Rhode IslandThe Rhode Island Division of Fishand Wildlife (RIDFW) will implement a5 percent bycatch allowance for Federalvessels fishing in the Atlantic herringfishery in Federal waters. RIDFW willalso implement a mandatory permitting process that will require vessels wantingto fish in the Rhode Island watersAtlantic herring fishery to, amongstother requirements, integrate in to theUniversity of Massachusetts Dartmouth, School for Marine Science andTechnology, river herring bycatchmonitoring program to ensuremonitoring of the fishery and minimizebycatch.

As of Jan 1, 2013, there is aprohibition to land, catch, take, orattempt to catch or take river herringwhich is a continuation of measures thatRIDFW has had in place since 2006when a moratorium was originally established (Taylor, Pers. comm., 2013).Connecticut Since April 2002, there has been aprohibition on the commercial orrecreational taking of migratory alewives and blueback herring from allmarine waters and most inland waters.As of January 1, 2012, commercial andrecreational harvest of river herring wasprohibited in Connecticut, as requiredby ASMFC Amendment 2 to the Shadand River Herring FMP (Taylor, Pers.Comm., 2013).New YorkCurrent regulations allow for arestricted river herring commercial andrecreational fishery in the Hudson Riverand tributaries, while all other statewaters prohibit river herring fisheries.

These regulations have been approvedthrough a sustainable fisheries management plan, as required underASMFC Amendment 2 to the Shad andRiver Herring FMP.New Jersey/Delaware As of January 1, 2012, commercial harvest of river herring was prohibited in New Jersey and Delaware, as requiredby ASMFC Amendment 2 to the Shadand River Herring FMP. Additionally, only commercial vessels fishingexclusively in Federal waters whileoperating with a valid Federal permitfor Atlantic mackerel and/or Atlanticherring may possess river herring up toa maximum of five percent by weight ofall species possessed (Taylor, Pers.Comm.).MarylandAs of January 1, 2012, commercial harvest of river herring was prohibited in Maryland, as required by ASMFCAmendment 2 to the Shad and RiverHerring FMP. However, an exception isprovided for anyone in possession ofriver herring as bait, as long as a receiptindicating where the herring waspurchased is in hand (Taylor, Pers.comm). This will allow bait shops tosell, and fishermen to possess, riverherring for bait that was harvested froma state whose fishery remains open, asan ASMFC approved sustainable fishery(Taylor, Pers. Comm).Potomac River Fisheries Commission (PRFC)/District of ColumbiaThe PRFC regulates only themainstem of the river, while thetributaries on either side are underMaryland and Virginia jurisdiction.

TheDistrict of Columbia's Department of theEnvironment (DDOE) has authority forthe Potomac River to the Virginia shoreand other waters within District ofColumbia.

Today, the river herringharvest in the Potomac is almostexclusively taken by pound nets. In1964, licenses were required tocommercially harvest fish in thePotomac River. After Maryland andVirginia established limited entryfisheries in the 1990s, the PRFCresponded to industry's request and, in1995, capped the Potomac River poundnet fishery at 100 licenses.

As of January1, 2010, harvest of river herring wasprohibited in the Potomac River, with aminimal bycatch provision of 50 lb (22kg) per licensee per day for pound nets.These regulations have been approvedthrough a sustainable fisheries management plan, as required underASMFC Amendment 2 to the Shad andRiver Herring FMP.VirginiaVirginia's Department of Game andInland Fisheries (VDGIF) is responsible for the management of fishery resources in the state's inland waters. As ofJanuary 1, 2008, possession of alewivesand blueback herring was prohibited onrivers draining into North Carolina (4VAC 15-320-25).

The Virginia MarineResources Commission (VMRC) isresponsible for management of fisheryresources within the state's marinewaters. As of January 1, 2012,commercial and recreational harvest ofriver herring was prohibited in allwaters of Virginia, as required byASMFC Amendment 2 to the Shad andRiver Herring FMP. Additionally, it isunlawful for any person to possess riverherring aboard a vessel on Virginia tidalwaters, or to land any river herring inVirginia (4 VAC 20-1260-30).

North CarolinaA no harvest provision for riverherring, commercial and recreational, within North Carolina was approved in2007. A limited research set aside of7,500 lb (3.4 mt) was established, and toimplement this harvest, a Discretionary Herring Fishing Permit (DHFP) was 48968Federal Register/Vol.

78, No. 155/Monday, August 12, 2013 / Noticescreated.

Individuals interested inparticipating had to meet the following requirements:

(1) Obtain a DHFP, (2)harvest only from the Joint FishingWaters of Chowan River during theharvest period, (3) must hold a validNorth Carolina Standard Commercial Fishing License (SCFL) or a RetiredSCFL, and (4) participate in statistical information and data collection programs.

Sale of harvested riverherring had to be to a licensed andpermitted River Herring Dealer. Eachpermit holder was allocated 125-250 lb(56-113 kg) for the 4-day season duringEaster weekend.

These regulations wereapproved through a sustainable fisheries management plan, as required underASMFC Amendment 2 to the Shad andRiver Herring FMP. The North CarolinaWildlife Resources Commission (NCWRC) has authority over the InlandWaters of the state. Since July 1, 2006,harvest of river herring, greater than 6inches (15.24 cm) has been prohibited in the inland waters of North Carolina's coastal systems.South CarolinaIn South Carolina, the South CarolinaDivision of Natural Resources (SCDNR)manages commercial herring fisheries using a combination of seasons, gearrestrictions, and catch limits. Today, thecommercial fishery for blueback herringhas a 10-bushel daily limit (500 lb (226kg)) per boat in the Cooper and SanteeRivers and the Santee-Cooper Rediversion Canal and a 250-lb-per-boat (113 kg) limit in the Santee-Cooper lakes. Seasons generally span thespawning season. All licensedfishermen have been required to reporttheir daily catch and effort to theSCDNR since 1998.The recreational fishery has a 1-bushel (49 lb (22.7 kg)) fish aggregate daily creel for blueback herring in allrivers; however, very few recreational anglers target blueback herring.

Theseregulations have been approved througha sustainable fisheries management plan, as required under ASMFCAmendment 2 to the Shad and RiverHerring FMP.GeorgiaThe take of blueback herring is illegalin freshwater in Georgia.

As of January1, 2012, harvest of river herring wasprohibited in Georgia, as required byASMFC Amendment 2 to the Shad andRiver Herring FMP.FloridaThe St. Johns River, Florida, harborsthe southernmost spawning run ofblueback herring.

There is currently noactive management of blueback herringin Florida.

As of January 1, 2012,harvest of river herring was prohibited, as required by ASMFC Amendment 2 tothe Shad and River Herring FMP.Tribal and First Nation Fisheries We have identified thirteen federally recognized East Coast tribes from Maineto South Carolina that have tribal rightsto sustenance and ceremonial fishing,and which may harvest river herring forsustenance and ceremonial purposesand/or engage in other river herringconservation and management activities.

The Mashpee Wampanoag tribe is the only East Coast tribe thatvoluntarily reported harvest numbers tothe State of MA that were incorporated into the ASMFC Management Plan assubsistence harvest.

The reportedharvest for 2006 and 2008 rangedbetween 1,200 and 3,500 fish per year,with removals coming from severalrivers. Aside from the harvest reportedby ASMFC for the Mashpee Wampanoag tribe, information as to what tribes mayharvest river herring for sustenance and/or ceremonial purposes is not available.

Letters have been sent to all 13potentially affected tribes to solicit anyinput they may have on theconservation status of the species and/or health of particular riverinepopulations, tribal conservation andmanagement activities for river herring,biological data for either species, andcomments and/or concerns regarding the status review process and potential implications for tribal trust resources and activities.

To date, we have notreceived any information from anytribes.Summary and Evaluation for Factor DAs described in Factor A, there aremultiple threats to habitat that haveaffected and may continue to affect riverherring including dams/culverts,

dredging, water quality, waterwithdrawals and discharge.

However,many of these threats are beingaddressed to some degree throughexisting Federal legislation such as theFederal Water Pollution Control Act,also known as the Clean Water Act, theCoastal Zone Management Act, theRivers and Harbors Act, the FPA,Marine Protection, Research andSanctuaries Act of 1972, the ShoreProtection Act of 1988, EFHdesignations for other species and ESAlistings for Atlantic salmon and Atlanticsturgeon.

Commercial harvest of alewife andblueback herring is occurring in Canadawith regulations,

closures, and quotas ineffect. In the United States, commercial harvest of alewife and blueback herringis also currently occurring in a fewstates with regulations that have beenapproved through a sustainable fisheries management plan, as required underASMFC Amendment 2 to the Shad andRiver Herring FMP. All other states hadpreviously established moratoria or, asof January 1, 2012, harvest of riverherring was prohibited, as required byASMFC Amendment 2 to the Shad andRiver Herring FMP. However, riverherring are incidentally caught inseveral commercial fisheries, but theextent to which this is occurring has notbeen fully quantified.

The New Englandand Mid-Atlantic Fishery Management Councils have adopted measures for theAtlantic herring and mackerel fisheries intended to decrease incidental catchand bycatch of alewife and bluebackherring.

In the United States, thirteenfederally recognized East Coast tribesfrom Maine to South Carolina havetribal rights to sustenance andceremonial

fishing, and may harvestriver herring for sustenance andceremonial purposes and/or engage inother river herring conservation andmanagement activities.

We have furtherevaluated the existing international,

Federal, and state management measures in the qualitative threatsassessment section below.E. Other Natural or Manmade FactorsAffecting the Continued Existence of theSpeciesCompetition Intra- and inter-specific competition were considered as potential naturalthreats to alewife and blueback herring.The earlier spawning time of alewifemay lead to differences in prey selection from blueback

herring, given that theybecome more omnivorous withincreasing size (Klauda et al., 1991a).This could lead to differences in preyselection given that juvenile alewifewould achieve a greater age and sizeearlier than blueback herring.

JuvenileAmerican shad are reported to focus ondifferent prey than blueback herring(Klauda et al., 1991b). However, Smithand Link (2010) found few differences between American shad and bluebackherring diets across geographic areasand size categories; therefore, competition between these two speciesmay be occurring.

Cannibalism has beenobserved (rarely) in landlocked systemswith alewife.

Additionally, evidence ofhybridization exists between alewifeand blueback

herring, but theimplications of this are unknown.Competition for habitat or resources hasnot been documented with alewife/blueback herring hybrids, as there islittle documentation of hybridization inpublished literature, but given the Federal Register/Vol.

78, No. 155/Monday, August 12, 2013/Notices 48969unknowns about their life history, it ispossible that competition between non-hybrids and hybrids could be occurring.

Artificial Propagation and StockingGenetics data have shown thatstocking alewife and blueback herringwithin and out of basin in Maine hashad an impact on the genetic groupings within Maine (Bentzen, 2012,unpublished data); however, the extentto which this poses a threat to riverherring locally or coast-wide isunknown.

Stocking river herringdirectly impacts a specific river/watershed system for river herring inthat it can result in passing fish abovebarriers into suitable spawning andrearing habitat, expanding populations into other watersheds, and introducing fish to newly accessible spawninghabitat.The alewife restoration program inMerrymeeting Bay, Maine, focuses onstocking lakes and ponds in theSebasticook River watershed and SevenMile Stream drainage.

The highestnumber of stocked fish was 2,211,658 in2009 in the Sebasticook River and93,775 in 2008 in the Kennebec River.The annual stocking goal of therestoration projects range from 120,000to 500,000 fish, with most fish stockedin the Androscoggin and Sebasticook watersheds.

The Union River fishery inEllsworth, Maine, is sustained throughthe stocking of adult alewives above thehydropower dam at the head-of-tide.

Fish passage is not currently required atthis dam, but fish are transported around the dam to spawning habitat intwo lakes. The annual adult stockingrate (from 2011 forward) is 150,000 fish.Adult river herring are trapped at acommercial harvest sites below the damand trucked to waters upstream of thedam. The highest number of stockedfish in the Union River was 1,238,790 in1986. In the Penobscot River watershed, over 48,000 adult fish were stocked intolakes in 2012, using fish collected fromthe Kennebec (39,650) and Union Rivers(8,998).

The New Hampshire Fish andGame stocks river herring into theNashua River, the Pine Island Pond, andthe Winnisquam Lake using fish fromvarious rivers which have included theConnecticut,

Cocheco, Lamprey,Kennebec, and Androscoggin Rivers.MA Division of Marine Fisheries (DMF)conducts a trap and transport stockingprogram for alewife and bluebackherring.

Prior to the moratorium in thestate, the program transported between30,000 and 50,000 fish per year into 10-15 different systems.

Since themoratorium, effort has been reduced toprotect donor populations andapproximately 20,000 fish per year havebeen deposited into five to ten systems.Many of the recent efforts have beenwithin system, moving fish upstreampast multiple obstructions to theheadwater spawning habitat.

RhodeIsland's Department of Environmental Management (DEM) has been stockingthe Blackstone River with adultbroodstock which was acquired fromexisting Rhode Island river herring runsand other sources out of state. In April2012, over 2,000 river herring pre-spawned adults were stocked into theBlackstone River. A small number ofalewife (200-400 fish) were stocked inthe Bronx River, NY, in 2006 and 2007from Brides Brook in East Lyme, CT.Furthermore, an experimental stockingprogram exists in Virginia wherehatchery broodstock are marked andstocked into the Kimages Creek, atributary to the James River. A total of319,856 marked river herring fry werestocked in this creek in 2011.The Edenton National Fish Hatchery(NFH) in North Carolina and theHarrison Lake NFH in Virginia havepropagated blueback herring forrestoration purposes.

Edenton NFH iscurrently rearing blueback herring forstocking in Indian Creek and Bennett's Creek in the Chowan River watershed inVirginia.

This is a pilot project to see ifhatchery contribution makes asignificant improvement in runs ofreturning adults (S. Jackson, USFWS,Pers. comm., 2012). Artificial propagation through the Edenton NFHfor the pilot program in the ChowanRiver watershed is intended forrestoration

purposes, and it is notthought that negative impacts toanadromous blueback herringpopulations will be associated withthese efforts.Landlocked Alewife and BluebackHerringAs noted above, alewives andblueback herring maintain two lifehistory variants; anadromous andlandlocked.

It is believed that theydiverged relatively recently (300 to5,000 years ago) and are now discretefrom each other. Landlocked alewifepopulations occur in many freshwater lakes and ponds from Canada to NorthCarolina as well as the Great Lakes(Rothschild, 1966; Boaze & Lackey,1974). Landlocked blueback herringoccur mostly in the southeastern UnitedStates and the Hudson River drainage.

At this time, there is no substantive information that would suggest thatlandlocked populations can or wouldrevert back to an anadromous lifehistory if they had the opportunity to doso (Gephard and Jordaan, Pers. comm.,2012). The discrete life history andmorphological differences between thetwo life history variants providesubstantial evidence that uponbecoming landlocked, landlocked herring populations become largelyindependent and separate fromanadromous populations.

Landlocked populations and anadromous populations occupy largely separateecological niches, especially in respectto their contribution to freshwater, estuary and marine food-webs (Palkovacs and Post, 2008). Thus, theexistence of landlocked life forms doesnot appear to pose a significant threat tothe anadromous forms.Interbreeding Among Alewife andBlueback Herring (Hybridization)

Recent genetic studies indicate thathybridization may be occurring in someinstances among alewife and bluebackherring where populations overlap(discussed in the River Herring StockStructure Working Group Report,NMFS, 2012a). Though interbreeding among closely related species isuncommon, it does occasionally occur(Levin, 2002). Most often, different reproductive strategies, home ranges,and habitat differences of closely relatedspecies either prevent interbreeding, orkeep interbreeding at very low levels. Incircumstances where interbreeding doesoccur, natural selection often keepshybrids in check because hybrids areless fit in terms of survival or theirability to breed successfully (Levin,2002). Other times, intermediate environmental conditions can providean environment where hybrids canthrive, and when hybrids breed with themember of the parent species, this canlead to "mongrelization" of one or bothparent species; a process referred to asintrogressive hybridization (Arnold,1997). Introgressive hybridization canalso occur as a result of introductions ofclosely related species, or man-made ornatural disturbances that createenvironments more suitable for thehybrid offspring than for the parents(e.g., the introduction of mallards hasled to the decline of the American blackduck through hybridization andintrogression)

(Anderson, 1949; Rhymer,2008).Though evidence has come forwardthat indicates that some hybridization may be occurring between alewife andblueback

herring, there is not enoughevidence to conclude whether or nothybridization poses a threat to one orboth species of river herring.

Mostimportantly, there is not enoughevidence to show whether hybridssurvive to maturity and, if so, whetherthey are capable of breeding with each 48970Federal Register/Vol.

78, No. 155/Monday, August 12, 2013/Notices other or breeding with either of theparent species.Summary and Evaluation of Factor EThe potential for inter- and intra-specific competition has beeninvestigated with respect to alewife andblueback herring.

Differences have beenobserved in the diet activity patternsand in spawning times of anadromous

alosids, and this may reduce inter- andintra- specific competition.

However, itis possible that competition isoccurring, as similarities in prey choicehave been identified.

Stocking is a toolthat managers have used for hundreds ofyears with many different species offish. This tool has been used as a meansof supporting restoration (e.g., passingfish above barriers into suitablespawning and rearing habitat,expanding populations into otherwatersheds, and introducing fish tonewly accessible spawning habitat).

Inaddition, stocking has been used tointroduce species to a watershed forrecreational purposes.

Stocking of riverherring has occurred for many years inMaine watersheds, but is less commonthroughout the rest of the range of bothspecies.

Stocking in the United Stateshas consisted primarily of trap andtruck operations that move fish fromone river system to another or over animpassible dam. Artificial propagation of river herring is not occurring to asignificant extent, though bluebackherring are being reared on a small scalefor experimental stocking in NorthCarolina.

We have considered natural ormanmade factors that may affect riverherring, including competition, artificial propagation and stocking, landlocked river herring, and hybrids.

Severalpotential natural or manmade threats toriver herring were identified, and wehave considered the effects of thesepotential threats further in thequalitative threats assessment described below.Threats Evaluation for Alewife andBlueback HerringDuring the course of the StatusReview for river herring, 22 potential threats to alewife and blueback herringwere identified that relate to one ormore of the five ESA section 4(a)(1)factors identified above. The SRTconducted a qualitative threatsassessment (QTA) to help evaluate thesignificance of the threats to bothspecies of river herring now and into theforeseeable future. NMFS has usedqualitative analyses to estimateextinction risk in previous statusreviews on the West Coast (e.g., Pacificsalmon, Pacific herring, Pacific hake,rockfish, and eulachon) and East Coast(e.g., Atlantic

sturgeon, cusk, Atlanticwolffish),

and the River Herring SRTdeveloped a qualitative ranking systemthat was adapted from these types ofqualitative analyses.

The results fromthe threats assessment have beenorganized and described according tothe above mentioned section 4(a)(1)factors.

They were used in combination with the results of the extinction riskmodeling to make a determination as towhether listing is warranted.

When ranking each threat, Teammembers considered how variousdemographic variables (e.g., abundance, population size, productivity, spatialstructure and genetic diversity) may beaffected by a particular threat. WhileFactor D, "inadequacy of existingregulatory mechanisms,"

is a different type of factor, the impacts on thespecies resulting from unregulated orinadequately regulated threats should beevaluated in the same way as the otherfour factors.QTA MethodsAll nine SRT members conducted anindependent, qualitative ranking of theseverity of each of the 22 identified threats to alewives and bluebackherring.

NERO staff developed factsheets for the SRT that contained essential information about theparticular threats under each of the fiveESA section 4(a)(1) factors, attempts toameliorate these threats, and how thethreats are or may be affecting bothspecies.

These fact sheets were reviewedby various experts within NMFS toensure that they contained all of the bestavailable information for each of thefactors.Team members ranked the threatsseparately for both species at arangewide scale and at the individual stock complex level. Each Teammember was allotted five likelihood points to rank each threat. Teammembers ranked the severity of eachthreat through the allocation of thesefive likelihood points across five ranksranging from "low" to "high." EachTeam member could allocate all fivelikelihood points to one rank ordistribute the likelihood points acrossseveral ranks to account for anyuncertainty.

Each individual Teammember distributed the likelihood points as he/she deemed appropriate with the condition that all fivelikelihood points had to be used foreach threat. Team members also had theoption of ranking the threat as "0" toindicate that in their opinion there wereinsufficient data to assign a rank, or "N/A" if in their opinion the threat was notrelevant to the species either throughout its range or for individual stockcomplexes.

When a Team member choseeither N/A (Not Applicable) or 0(Unknown) for a threat, all 5 likelihood points had to be assigned to that rankonly. Qualitative descriptions of ranksfor the threats listed for alewife andblueback herring (Table 1, 2) are:* N/A-Not Applicable.

• 0-Unknown.

• 1 Low-It is likely that this threatis not significantly affecting the speciesnow and into the foreseeable future, andthat this threat is limited in geographic scope or is localized within the species/stock complex' range.* 2 Moderately Low-Threat fallsbetween rankings 1 and 3.* 3 Moderate-It is likely that thisthreat has some effect on the speciesnow and into the foreseeable future, andit is widespread throughout the species/stock complex' range.* 4 Moderately High-Threat fallsbetween rankings 3 and 5.e 5 High-It is likely that this threatis significantly affecting the species nowand into the foreseeable future, and it iswidespread in geographic scope andpervasive throughout the species/stock complex' range.The SRT identified and ranked 22threats to both species both rangewide and for the individual stock complexes.

Threats included dams and barriers,

dredging, water quality and waterwithdrawals, climate change/variability, harvest (both directed and incidental),

disease, predation, management internationally, federally, and at thestate level, competition, artificial propagation and stocking,

hybrids, andfrom landlocked populations.

QTA ResultsThe SRT unequivocally identified dams and barriers as the most important threat to alewife and blueback herringpopulations both rangewide and acrossall stock complexes (the qualitative ranking for dams and barriers wasbetween moderately high and high).Incidental catch, climate change,dredging, water quality, waterwithdrawal/outfall, predation, andexisting regulation were among themore important threats after dams forboth species, and for all stockcomplexes (qualitative rankings forthese threats ranged betweenmoderately low and moderate).

Waterquality, water withdrawal/outfall, predation, climate change and climatevariability were generally seen as greaterthreats to both species in the southernportion of their ranges than in thenorthern portion of their ranges. Inaddition, the Team identified commercial harvest as being notably Federal Register

/ Vol. 78, No. 155 / Monday, August 12, 2013 / Notices48971more important in Canada than in theUnited States. The results of the threatsanalysis for alewives are presented inTables 1-5 and Figure 3. The results ofthe threats analysis for blueback herringare presented in Tables 6-10 and Figure4.QTA Conclusion The distribution of rankings acrossthreat levels provides a way to evaluatecertainty in the threat level for each ofthe threats identified.

The amount ofcertainty for a threat is a reflection ofthe amount of evidence that links aparticular threat to the continued survival of each species.

For threatswith more data, there tended to be morecertainty surrounding the threat level,whereas threats with fewer data tendedto have more uncertainty.

The sameholds true for datasets that were limitedover space and/or time.The results of the threats assessment rangewide and for all stock complexes reveal strong agreement and lowuncertainty among the reviewers thatdams and barriers are the greatest threatto both alewives and blueback herring.There was also strong agreement thattribal fisheries, scientific monitoring, and educational harvest currently poselittle threat to the species.

For thethreats of state, Federal andinternational management,

dredging, climate change, climate variability, predation, and incidental catch, therewas more uncertainty.

Among alewife and blueback stockcomplexes, Canada, the Mid-Atlantic, and South Atlantic diverged the mostfrom the other stock complexes withrespect to certainty of threats.

In Canadathere was more certainty surrounding the threats of climate change andclimate variability for both species, andless certainty surrounding the threat ofdirected commercial harvest andincidental catch for alewives comparedto the certainty surrounding thesethreats for the other stock complexes.

Inthe mid-Atlantic for alewives andSouth-Atlantic for bluebacks, there wasmore uncertainty surrounding climatevariability and climate changecompared to the certainty surrounding these threats for the other stockcomplexes.

Based on the Team member rankings, dams and other barriers present thegreatest and most persistent threatrangewide to both blueback herring andalewife (Tables 12-13). Dams andculverts block access to historical migratory corridors and spawninglocations, in some instances, even whenfish passage facilities are present.Centuries of blocked and reduced accessto spawning and rearing habitat haveresulted in decreased overall production potential of watersheds along theAtlantic coast for alewives and bluebackherring (Hall et al., 2012). This reducedproduction potential has likely been oneof the main drivers in the decreased abundance of both species.

The recentASMFC Stock Assessment (2012)attempted to quantify biomass estimates for both alewife and blueback herringbut was unable to develop an acceptable model to complete a biomass estimate.

Therefore, it is difficult to accurately quantify the declines from historical biomass to present-day

biomass, thoughsignificant declines have been noted.Studies from Maine show that damshave reduced accessible habitat to afraction of historical levels, 5 percent foralewives and 20 percent for bluebackherring (Hall et al., 2011).Rangewide, for alewife and bluebackherring, no other threats rose to the levelof dams, but several other stressors ranked near the moderate threat level.The Team ranked incidental catch,water quality, and predation as threatslikely to have some effect on the speciesnow and into the foreseeable future thatare widespread throughout the species'range. Incidental catch is primarily fromfisheries that use small-mesh mobilegear, such as bottom and midwatertrawls. Sources of water qualityproblems vary from river to river andare therefore unique to each of the stockcomplexes.

And finally, predation bystriped bass, seals, double-crested cormorants (and other fish-eating avianspecies, e.g., northern gannets) andother predators is known to exist, butdata are lacking on the overallmagnitude.

Overall, the degree ofcertainty associated with these mid-level threats is much lower, primarily due to lack of information on how thesestressors are affecting both species.The SRT's qualitative rankings andanalysis of threats for alewife rangewide and for each stock complex:BILLING CODE 3510-22-P 48972 Federal Register/Vol.

78, No. 155/Monday, August 12, 2013/Notices High5CoLow432100 Canada*1 Northern New England* Southern New England* Mid-Atlantic

• Overall RangeFigure 3. Median qualitative ranking of threats to alewives range-wide and for each stockcomplex.

Federal Register/Vol.

78, No. 155/Monday, August 12, 2013/Notices 48973Table 1. Qualitative ranking of threats for the alewife rangewide.

Status Review Teammembers ranked threats by distributing 5 likelihood points among 5 ranks: 1- low, 2-medium/low, 3- medium, 4-medium/high, 5-high. The mean represents the overall Teamaverage rank, mode represents the rank which received the most likelihood points, andrange represents the range of ranks that were assigned likelihood points for each threat.N=number of Team members who ranked the threat between 1 and 5; likelihood pointsfor threats that Team members ranked as either unknown or not applicable are notincluded.

Threats Mean SD Mode Range NDams and Other Barriers 4.3 0.7 5 3-5 9Water Quality (chemical) 2.8 1.0 3 1-5 9Incidental Catch 2.7 0.9 3 1-5 9Predation 2.6 1.1 3 1-5 9Water Withdrawal/Outfall (physical andtemp.) 2.4 0.8 2 1-5 9Dredging 2.4 1.0 2 1-4 9Climate change 2.4 0.9 3 1-4 8Climate variability 2.3 1.1 2 1-5 9Federal Management 2.3 1.1 2 1-5 9International Management 2.3 1.1 2 1-5 9State Management 2.2 1.2 1 1-5 9Directed Commercial Harvest 1.8 0.8 2 1-3 9Competition 1.6 0.7 1 1-4 9Artificial Propagation and Stocking 1.5 0.7 1 1-3 9Hybrids 1.5 0.7 1 1-3 2Recreational Harvest 1.4 0.6 1 1-3 9Tribal/First Nation Fisheries Management 1.3 0.5 1 1-3 7Disease 1.3 0.4 1 1-2 8Landlocked Populations 1.2 0.5 1 1-3 8Tribal/First Nation Fisheries Utilization 1.2 0.4 1 1-2 8Scientific Monitoring 1.0 0.2 1 1-2 9Educational Harvest 1.0 0.1 1 1-2 9 48974Federal Register/Vol.

78, No. 155/Monday, August 12, 2013/Notices Table 2. Qualitative ranking of threats for the Canadian stock complex of alewife.

StatusReview Team members ranked threats by distributing 5 likelihood points among 5 ranks:1- low, 2-medium/low, 3- medium, 4-medium/high, 5-high. The mean represents theoverall Team average rank, mode represents the rank which received the most likelihood points, and range represents the range of ranks that were assigned likelihood points foreach threat. N=number of Team members who ranked the threat between 1 and 5;likelihood points for threats that Team members ranked as either unknown or notapplicable are not included.

Threats Mean SD Mode Range NDams and Other Barriers 4.0 0.9 5 2-5 8State Management 2.4 0.9 2 1-4 6Incidental Catch 2.4 1.2 1 1-5 6Federal Management 2.4 0.9 2 1-4 6Water Withdrawal/Outfall (physical andtemp.) 2.3 0.7 2 1-3 6Directed Commercial Harvest 2.2 0.9 2 1-4 8International Management 2.2 0.9 2 1-4 8Water Quality (chemical) 2.1 0.7 2 1-3 7Predation 2.1 1.0 2 1-5 8Dredging 2.0 0.7 2 1-4 6Climate variability 1.9 0.9 2 1-5 8Climate change 1.6 0.7 1 1-4 8Hybrids 1.5 0.7 1 1-3 2Competition 1.4 0.5 1 1-3 9Disease 1.3 0.5 1 1-2 7Artificial Propagation and Stocking 1.3 0.5 1 1-3 7Tribal/First Nation Fisheries Management 1.2 0.4 1 1-2 5Recreational Harvest 1.2 0.4 1 1-2 6Tribal/First Nation Fisheries Utilization 1.2 0.4 1 1-2 6Landlocked Populations 1.1 0.3 1 1-2 7Scientific Monitoring 1.0 0.2 1 1-2 6Educational Harvest 1.0 0.0 1 1 6 Federal Register/Vol.

78, No. 155/Monday, August 12, 2013/Notices 48975Table 3. Qualitative ranking of threats for the Northern New England stock complex ofalewife.

Status Review Team members ranked threats by distributing 5 likelihood pointsamong 5 ranks: 1- low, 2-medium/low, 3- medium, 4-medium/high, 5-high. The meanrepresents the overall Team average rank, mode represents the rank which received themost likelihood points, and range represents the range of ranks that were assignedlikelihood points for each threat. N=number of Team members who ranked the threatbetween I and 5; likelihood points for threats that Team members ranked as eitherunknown or not applicable are not included.

Threats Mean SD Mode Range NDams and Other Barriers 4.3 0.7 5 3-5 9Incidental Catch 2.9 0.8 3 1-5 7Water Withdrawal/Outfall (physical andtemp.) 2.5 0.9 3 1-5 8Dredging 2.4 0.9 2,3 1-4 8State Management 2.4 1.1 2 1-5 9Predation 2.4 1.2 2 1-5 9Federal Management 2.4 1.1 2 1-5 9International Management 2.2 0.9 2 1-4 9Water Quality (chemical) 2.1 1.0 1 1-5 9Climate variability 2.0 1.0 2 1-5 9Directed Commercial Harvest 1.9 0.9 1 1-4 9Climate change 1.8 0.8 1 1-4 8Artificial Propagation and Stocking 1.6 0.7 1 1-3 9Hybrids 1.5 0.7 1 1-3 2Competition 1.5 0.6 1 1-3 9Recreational Harvest 1.3 0.5 1 1-3 9Disease 1.3 0.5 1 1-2 8Landlocked Populations 1.2 0.4 1 1-2 8Tribal/First Nation Fisheries Management 1.2 0.4 1 1-2 7Tribal/First Nation Fisheries Utilization 1.1 0.3 1 1-2 8Scientific Monitoring 1.0 0.1 1 1-2 9Educational Harvest 1.0 0.0 1 1 9 48976Federal Register/Vol.

78, No. 155/Monday, August 12, 2013/Notices Table 4. Qualitative ranking of threats for the Southern New England stock complex ofalewife.

Status Review Team members ranked threats by distributing 5 likelihood pointsamong 5 ranks: 1- low, 2-medium/low, 3- medium, 4-medium/high, 5-high. The meanrepresents the overall Team average rank, mode represents the rank which received themost likelihood points, and range represents the range of ranks that were assignedlikelihood points for each threat. N=number of Team members who ranked the threatbetween I and 5; likelihood points for threats that Team members ranked as eitherunknown or not applicable are not included.

Threats Mean SD Mode Range NDams and Other Barriers 4.2 0,7 4 3-5 9Incidental Catch 2.9 0.8 3 1-5 7Water Withdrawal/Outfall (physical andtemp.) 2.7 0.8 3 1-5 8Water Quality (chemical) 2.5 0.9 3 1-5 9Predation 2.5 1.1 2 1-5 9Dredging 2.5 0.9 3 1-4 8Federal Management 2.2 1.1 2 1-5 9Climate variability 2.2 1.0 2 1-5 9State Management 2.2 1.1 2 1-5 9Climate change 2.2 1.0 1,3 1-4 8International Management 2.0 0.8 2 1-4 9Directed Commercial Harvest 1.7 0.8 1 1-3 9Hybrids 1.5 0.7 1 1-3 2Artificial Propagation and Stocking 1.5 0.6 1 1-3 9Competition 1.4 0.6 1 1-3 9Disease 1.3 0.5 1 1-2 8Recreational Harvest 1.3 0.5 1 1-3 9Landlocked Populations 1.2 0.4 1 1-2 8Tribal/First Nation Fisheries Management 1.2 0.4 1 1-2 7Tribal/First Nation Fisheries Utilization 1.2 0.4 1 1-2 8Scientific Monitoring 1.0 0.1 1 1-2 9Educational Harvest 1.0 0.0 1 1 9 Federal Register/Vol.

78, No. 155/Monday, August 12, 2013/Notices 48977Table 5. Qualitative ranking of threats for the Mid-Atlantic stock complex of alewife.Status Review Team members ranked threats by distributing 5 likelihood points among 5ranks: 1- low, 2-medium/low, 3- medium, 4-medium/high, 5-high. The mean represents the overall Team average rank, mode represents the rank which received the mostlikelihood points, and range represents the range of ranks that were assigned likelihood points for each threat. N=number of Team members who ranked the threat between Iand 5; likelihood points for threats that Team members ranked as either unknown or notapplicable are not included.

Threats Mean SD Mode Range NDams and Other Barriers 3.8 1.0 4 3-5 9Incidental Catch 2.9 0.8 3 1-5 7Water Quality (chemical) 2.9 0.9 3 1-5 9Water Withdrawal/Outfall (physical andtemp.) 2.8 0.8 3 1-5 8Climate change 2.7 1.2 3 1-5 8Climate variability 2.6 1.2 2 1-5 9Predation 2.5 1.1 2 1-5 9Dredging 2.5 0.9 3 1-4 8Federal Management 2.3 1.1 2 1-5 9State Management 2.2 1.1 2 1-5 9International Management 1.8 0.8 1 1-4 9Directed Commercial Harvest 1.7 0.8 1 1-3 9Hybrids 1.5 0.7 1 1-3 2Artificial Propagation and Stocking 1.5 0.7 1 1-3 9Competition 1.4 0.6 1 1-3 9Disease 1.4 0.5 1 1-3 8Recreational Harvest 1.3 0.5 1 1-3 9Landlocked Populations 1.2 0.4 1 1-2 8Tribal/First Nation Fisheries Management 1.2 0.4 1 1-3 7Tribal/First Nation Fisheries Utilization 1 .1 0.3 1 1-2 7Scientific Monitoring 1.0 0.1 1 1-2 9Educational Harvest 1.0 0.0 1 1 9 48978 sqalttverninso Federal Register/Vol.

78, No. 155/Monday, August 12, 2013/Notices The SRT's qualitative rankings ofthreats for blueback herring rangewide and for each stock complex:Figure 4. Median qualitative ranking of threats to blueback herring rangewide and foreach stock complex.

Federal Register/Vol.

78, No. 155/Monday, August 12, 2013/Notices 48979Table 6. Qualitative ranking of threats for blueback herring rangewide.

Status ReviewTeam members ranked threats by distributing 5 likelihood points among 5 ranks: I- low,2-medium/low, 3- medium, 4-medium/high, 5-high. The mean represents the overallTeam average rank, mode represents the rank which received the most likelihood points,and range represents the range of ranks that were assigned likelihood points for eachthreat. N=number of Team members who ranked the threat between 1 and 5; likelihood points for threats that Team members ranked as either unknown or not applicable are notincluded.

Threats Mean SD Mode Range NDams and Other Barriers 4.2 0.8 4,5 3-5 9Water Quality (chemical) 2.8 1.0 3 1-5 9Incidental Catch 2.7 0.9 3 1-5 9Climate change 2.7 1.2 3,4 1-5 8Predation 2.6 1.1 3 1-5 9Climate variability 2.4 1.2 1,2,3 1-5 9Water Withdrawal/Outfall (physical andtemp.) 2.4 0.8 2 1-5 9Dredging 2.4 1.0 2 1-4 9Hybrids 2.4 1.0 3 1-4 2Federal Management 2.3 1.1 2 1-5 9International Management 2.3 1.1 2 1-5 8State Management 2.2 1.1 1 1-5 9Directed Commercial Harvest 1.8 0.8 1 1-3 9Competition 1.5 0.6 1 1-3 9Artificial Propagation and Stocking 1.5 0.7 1 1-3 9Tribal/First Nation Fisheries Management 1.3 0.5 1 1-3 7Recreational Harvest 1.3 0.5 1 1-3 9Disease 1.3 0.5 1 1-2 8Landlocked Populations 1.2 0.5 .1 1-3 7Tribal/First Nation Fisheries Utilization 1.2 0.4 1 1-2 8Scientific Monitoring 1.0 0.2 1 1-2 9Educational Harvest 1.0 0.1 1 1-2 9 48980Federal Register/Vol.

78, No. 155/Monday, August 12, 2013/Notices Table 7. Qualitative rankings of threats for the Canadian stock complex of bluebackherring.

Status Review Team members ranked threats by distributing 5 likelihood pointsamong 5 ranks: 1- low, 2-medium/low, 3- medium, 4-medium/high, 5-high. The meanrepresents the overall Team average rank, mode represents the rank which received themost likelihood points, and range represents the range of ranks that were assignedlikelihood points for each threat. N=number of Team members who ranked the threatbetween I and 5; likelihood points for threats that Team members ranked as eitherunknown or not applicable are not included.

Threats Mean SD Mode Range NDams and Other Barriers 3.9 0.9 4 2-5 8Incidental Catch 2.4 1.2 1,3 1-5 6State Management 2.4 0.9 2 1-4 6Hybrids 2.4 1.0 3 1-4 2Water Withdrawal/Outfall (physical andtemp.) 2.4 0.6 2 1-3 6Federal Management 2.4 0.9 2 1-4 6Directed Commercial Harvest 2.4 0.8 3 1-4 8Water Quality (chemical) 2.2 0.7 2 1-3 7Climate variability 2.2 1.2 1 1-5 8Predation 2.1 1.0 2 1-4 8International Management 2.1 0.9 2 1-4 8Dredging 2.0 0.7 2 1-3 6Climate change 2.0 1.0 1 1-4 8Competition 1.5 0.6 1 1-4 9Recreational Harvest 1.3 0.5 1 1-2 6Disease 1.3 0.5 1 1-2 7Artificial Propagation and Stocking 1.3 0.5 1 1-3 7Tribal/First Nation Fisheries Utilization 1.2 0.4 1 1-2 6Tribal/First Nation Fisheries Regulation 1.2 0.4 1 1-3 5Landlocked Populations 1.A 0.3 1 1-2 6Scientific Monitoring 1.0 0.2 1 1-2 6Educational Harvest 1.0 0.0 1 1 6 Federal Register/Vol.

78, No. 155/Monday, August 12, 2013/Notices 48981Table 8. Qualitative ranking of threats for the Northern New England stock complex ofblueback herring.

Status Review Team members ranked threats by distributing 5likelihood points among 5 ranks: 1- low, 2-medium/low, 3- medium, 4-medium/high, 5-high. The mean represents the overall Team average rank, mode represents the rankwhich received the most likelihood points, and range represents the range of ranks thatwere assigned likelihood points for each threat. N=number of Team members whoranked the threat between 1 and 5; likelihood points for threats that Team membersranked as either unknown or not applicable are not included.

Threats Mean SD Mode Range NDams and Other Barriers 4.3 0.7 5 3-5 9Incidental Catch 2.8 0.9 3 1-5 7Dredging 2.6 1.0 3 1-4 8Water Withdrawal/Outfall (physical andtemp.) 2.5 0.9 3 1-5 8State Management 2.4 1.1 2 1-5 9Hybrids 2.4 1.0 3 1-4 2Water Quality (chemical) 2.4 1.1 3 1-5 9Predation 2.4 1.2 2 1-5 9Federal Management 2.4 1.1 2 1-5 9Climate variability 2.2 1.2 2 1-4 9Climate change 2.1 1.0 2 1-4 8International Management 2.0 0.9 2 1-4 9Directed Commercial Harvest 1.9 0.9 1 1-3 9Artificial Propagation and Stocking 1.6 0.7 1 1-3 9Competition 1.5 0.7 1 1-3 9Recreational Harvest 1.3 0.5 1 1-3 9Disease 1.3 0.5 1 1-2 8Landlocked Populations 1.2 0.4 1 1-2 7Tribal/First Nation Fisheries Management 1.1 0.4 1 1-2 7Tribal/First Nation Fisheries Utilization 1.1 0.3 1 1-2 8Scientific Monitoring 1.0 0.1 1 1-2 9Educational Harvest 1.0 0.0 1 1 9 48982Federal Register/Vol.

78, No. 155/Monday, August 12, 2013/Notices Table 9. Qualitative ranking of threats for the Southern New England stock complex ofblueback herring.

Status Review Team members ranked threats by distributing 5likelihood points among 5 ranks: 1- low, 2-medium/low, 3- medium, 4-medium/high, 5-high. The mean represents the overall Team average rank, mode represents the rankwhich received the most likelihood points, and range represents the range of ranks thatwere assigned likelihood points for each threat. N=number of Team members whoranked the threat between I and 5; likelihood points for threats that Team membersranked as either unknown or not applicable are not included.

Threats Mean SD Mode Range NDams and Other Barriers 4.3 0.7 4,5 3-5 9Incidental Catch 2.8 0.9 3 1-5 7Water Withdrawal/Outfall (physical andtemp.) 2.6 0.8 2,3 1-5 8Dredging 2.6 1.0 3 1-4 8Water Quality (chemical) 2.6 1.0 3 1-5 9Predation 2.4 1.1 2 1-5 9Hybrids 2.4 1.0 3 1-4 2Climate change 2.3 1.0 2 1-4 8Climate variability 2.3 1.1 2 1-5 9Federal Management 2.2 1.1 2 1-5 9State Management 1.9 1.1 1 1-5 9International Management 1.9 0.8 2 1-4 9Directed Commercial Harvest 1.6 0.8 1 1-3 9Artificial Propagation and Stocking 1.6 0.7 1 1-3 9Competition 1.5 0.7 1 1-4 9Disease 1.3 0.5 1 1-2 8Recreational Harvest 1.2 0.4 1 1-2 9Landlocked Populations 1.2 0.4 1 1-2 7Tribal/First Nation Fisheries Management 1.1 0.4 1 1-2 7Tribal/First Nation Fisheries Utilization 1.1 0.3 1 1-2 8Scientific Monitoring 1.0 0.1 1 1-2 9Educational Harvest 1.0 0.0 1 1 9 Federal Register/Vol.

78, No. 155/Monday, August 12, 2013/Notices 48983Table 10. Qualitative ranking of threats for the Mid-Atlantic stock complex of bluebackherring.

Status Review Team members ranked threats by distributing 5 likelihood pointsamong 5 ranks: 1- low, 2-medium/low, 3- medium, 4-medium/high, 5-high. The meanrepresents the overall Team average rank, mode represents the rank which received themost likelihood points, and range represents the range of ranks that were assignedlikelihood points for each threat. N=number of Team members who ranked the threatbetween 1 and 5; likelihood points for threats that Team members ranked as eitherunknown or not applicable are not included.

Threats Mean SD Mode Range NDams and Other Barriers 3.9 1.0 4 3-5 9Water Quality (chemical) 3.0 0.9 3 1-5 9Incidental Catch 2.8 0.9 3 1-5 7Water Withdrawal/Outfall (physical andtemp.) 2.7 0.8 3 1-5 8Climate change 2.7 1.2 3 1-5 8Dredging 2.6 1.0 3 1-4 8Climate variability 2.6 1.2 2,3 1-5 9Predation 2.4 1.1 2 1-5 9Hybrids 2.4 1.0 3 1-4 2Federal Management 2.3 1.1 2 1-5 9State Management 2.2 1.1 2 1-5 9Directed Commercial Harvest 1.9 0.9 1 1-4 9International Management 1.7 0.8 1 1-4 9Competition 1.5 0.7 1 1-3 9Artificial Propagation and Stocking 1.5 0.7 1 1-3 9Disease 1.4 0.5 1 1-3 8Recreational Harvest 1.3 0.5 1 1-3 9Landlocked Populations 1.2 0.4 1 1-2 7Tribal/First Nation Fisheries Management 1.1 0.4 1 1-2 7Tribal/First Nation Fisheries Utilization 1.1 0.3 1 1-2 7Scientific Monitoring 1.0 0.1 1 1-2 9Educational Harvest 1.0 0.0 1 1 9 48984Federal Register/Vol.

78, No. 155/Monday, August 12, 2013/Notices Table 11. Qualitative ranking of threats for the Southern Atlantic stock complex ofblueback herring.

Status Review Team members ranked threats by distributing 5likelihood points among 5 ranks: 1- low, 2-medium/low, 3- medium, 4-medium/high, 5-high. The mean represents the overall Team average rank, mode represents the rankwhich received the most likelihood points, and range represents the range of ranks thatwere assigned likelihood points for each threat. N=number of Team members whoranked the threat between 1 and 5; likelihood points for threats that Team membersranked as either unknown or not applicable are not included.

ThreatsDams and Other BarriersWater Quality (chemical)

Climate changeClimate variability Water Withdrawal/Outfall (physical andtemp.)DredgingIncidental CatchPredation Federal Management State Management HybridsDirected Commercial HarvestInternational Management Competition DiseaseArtificial Propagation and StockingRecreational HarvestLandlocked Populations Tribal/First Nation Fisheries Management Tribal/First Nation Fisheries Utilization Scientific Monitoring Educational HarvestMean3.83.03.02.8SD1.10.91.31.42.8 0.82.72.62.62.32.21.91.81.71.61.51.51.31.21.11.11.01.01.01.01.21.11.10.70.80.80.70.60.70.50.40.30.30.10.0Mode4342,4333322211111111111Range3-51-51-51-51-5N898981-41-51-51-51-51-31-31-41-31-31-31-31-21-21-21-21 Federal Register/Vol.

78, No. 155/Monday, August 12, 2013/Notices 48985Table 12. Summary table of threat ranking for alewife rangewide.

Threat Threat Level Section 4 FactorDams and Other Barriers Medium High AWater Quality (chemical)

Medium AIncidental Catch Medium BPredation Medium CDredging Medium Low AWater Withdrawal/Outfall (physical andtemp.) Medium Low AClimate change Medium Low AClimate variability Medium Low ADirected Commercial Harvest Medium Low BInternational Management Medium Low DFederal Management Medium Low DState Management Medium Low DCompetition Medium Low EArtificial Propagation and Stocking Medium Low ERecreational Harvest Low BTribal/First Nation Fisheries Management Low BScientific Monitoring Low BEducational Harvest Low BDisease Low CTribal/First Nation Fisheries Utilization Low DHybrids Low ELandlocked Populations Low E 48986Federal Register/Vol.

78, No. 155/Monday, August 12, 2013/Notices Table 13. Summary table of threat ranking for blueback herrin rangewide.

Threat Threat Level Section 4 FactorDams and Other Barriers Medium High AClimate change Medium AWater Quality (chemical)

Medium AIncidental Catch Medium BPredation Medium CWater Withdrawal/Outfall (physical andtemp.) Medium Low ADredging Medium Low AClimate variability Medium Low ADirected Commercial Harvest Medium Low BInternational Management Medium Low DFederal Management Medium Low DState Management Medium Low DCompetition Medium Low EHybrids Medium Low ERecreational Harvest Low BTribal/First Nation Fisheries Management Low BScientific Monitoring Low BEducational Harvest Low BDisease Low CTribal/First Nation Fisheries Utilization Low DArtificial Propagation and Stocking Low ELandlocked Populations Low EBILLING CODE 3510-22-C Extinction Risk AnalysisIn order to assess the risk ofextinction for alewife and bluebackherring, trends in the relativeabundance of alewife and bluebackherring were assessed for each speciesrangewide, as well as for each species-specific stock complex.

As notedpreviously, for alewife, the stockcomplexes include Canada, NorthernNew England, Southern New Englandand the mid-Atlantic.

For bluebackherring, the stock complexes areCanada, Northern New England,Southern New England, mid-Atlantic and Southern.

Criteria Established by SRT forEvaluating RiskPrior to conducting the trend analysismodeling, the SRT established criteriathat would be used to evaluate the riskto both species as well as to theindividual stock complexes.

At theSRT's request, the NEFSC conducted modeling to develop trends in relativeabundance by estimating the population growth rate for both species bothrangewide and for each individual stockcomplex.

The SRT established two tiersthat could be used separately or incombination to interpret the results ofthe modeling in order to assess risk toalewife and blueback herring rangewide and for the individual stock complexes.

We concur that these tiers areappropriate.

Tier A relates to what isknown about the geographic distribution, habitat connectivity andgenetic diversity of each species, andTier B relates to the risk thresholds established for the trend analysis thatwas conducted by the NEFSC. Thesetiers are subject to change in the futureas more information becomes available.

For example, Tier A is based onpreliminary genetic data addressing possible stock complexes, which couldchange in the future. Data related toboth tiers were assessed to determine ifsufficient information was available tomake a conclusion under one or both ofthe tiers. The SRT decided that, becauseof significant uncertainties associated with the available data and a significant number of data deficiencies for bothspecies, it was not necessary to haveinformation under both tiers in order tomake a risk determination, and weconcur with this decision.

The goal of Tier A was to maintainthree contiguous stock complexes thatare stable or increasing as this: (1)Satisfies the need to maintain bothgeographic closeness and geographic distance for a properly functioning metapopulation (see McElhany et al.,2000); (2) ensures that the recovered population does not include isolatedgenetic groups that could lead to geneticdivergence (McDowall, 2003, Quinn,1984); (3) provides some assurance thatthe species persists across a relatively wide geographic area supporting diverseenvironmental conditions and diversehabitat types; and (4) ensures that theentire population does not share thesame risk from localized environmental catastrophe (McElhany et al., 2000).Tier B information was used todirectly interpret the results of thetrends in relative abundance modeling Federal Register/Vol.

78, No. 155/Monday, August 12, 2013/Notices 48987conducted by the NEFSC. As described below, relative abundance of bothalewife and blueback herring was usedto estimate growth rate (along with the95 percent confidence intervals for thegrowth rates) for each species rangewide and for each stock complex.

Tier Bestablished risk criteria depending onthe outcomes of the population growthrate modeling.

As indicated in theforeseeable future section above, a 12- to18-year timeframe (e.g., 2024-2030) foreach species was determined to beappropriate.

After subsequent discussions, the SRT decided that theprojections into the foreseeable futurewould not provide meaningful information for the extinction riskanalysis.

As noted previously, the trendanalysis provides a steady population growth rate. If the population growthrate is positive and everything elseremains the same into the foreseeable future (e.g., natural and anthropogenic mortality rates do not change),

theabundance into the foreseeable futurewill continue to increase.

If thepopulation growth rate is negative, thenthe abundance into the foreseeable future will continue to decline.Currently, there is insufficient information available to modify any ofthe factors that may change the growthrates into the foreseeable future, andthus, performing these projections willnot provide meaningful information forthe extinction risk of either of thesespecies.The baseline for the overall riskassessment assumes that there hasalready been a significant decline inabundance in both species due to areduction in carrying capacity andoverfishing as indicated in variouspublications (Limburg and Waldman,2009; Hall et al., 2012), as well as otherthreats.

The estimated population growth rates reflect the impacts from thevarious threats to which the species arecurrently exposed.

The SRTrecommended that NEFSC use data from1976 through the present to minimizethe overfishing influence from distantwater fleets that occurred in earlieryears but has since been curtailed byfisheries management measures.

TheSRT recommended that the NEFSC alsorun a trajectory using a plus/minus 10-percent growth rate to test modelsensitivity with respect to changes inthe model variables.

This approach hasbeen used in analyses for other species(e.g., Atlantic

croaker, Atlantic cod) andcan serve as a means of showingsensitivities in the model to potential variables (e.g., population growth ratechanges, climate change) (Hare andAble, 2007; Hare, NMFS Pers. comm.,2012). Following completion of themodel results, we determined that theplus/minus 10-percent change inpopulation growth rate would notprovide additional information thatwould change the conclusions as towhether the populations aresignificantly increasing, stable ordecreasing.

Without the projections ofthe population growth rate into theforeseeable future, the plus/minus 10-percent would merely provide anadditional set of bounds around thepopulation growth rate estimate, and,therefore, we determined that runningthe model with the plus/minus 10-percent was not necessary.

The population growth rates derivedfrom the analysis help identify whetherstability exists within the population.

Mace et al. (2002) and Demaster et al.(2004) recognized that highly fecund,short generation time species like riverherring may be able to withstand a 95to 99 percent decline in biomass.

Bothalewives and blueback herring mayalready be at or less than two percent ofthe historical baseline (e.g., Limburgand Waldman, 2009), though theseestimates are based on commercial landings data, which are dependent upon management and are not a reliableestimate of biomass.

However,recognizing historical declines for bothspecies, the modeled population growthrates were used to gauge whether thesestock complexes are stable, significantly increasing or decreasing.

Relativeabundance of a stock is considered to besignificantly increasing or decreasing ifthe 95-percent confidence intervals ofthe population growth rate do notinclude zero. In contrast, if the 95-percent confidence intervals do containzero, then the population is considered to be stable, as the increasing ordecreasing trend in abundance is notstatistically significant.

The SRT determined and we agreethat a stable or significantly increasing trajectory suggests that these speciesmay be within the margins of being self-sustainable and thus, if all of the growthrates for the coast-wide distribution andthe stock complexes are stable orsignificantly increasing, the species is atlow risk of extinction (the riskcategories were defined by adapting thecategories described above for theQTA-Low risk-it is likely that thethreats to the species' continued existence are not significant now and/orinto the foreseeable future; Moderately Low-risk falls between low andmoderate rankings; Moderate-it islikely that the threats are having someeffect on the species continued existence now and/or into theforeseeable future; Moderately High-the risk falls between moderate andhigh; High-it is likely that the threatsare significantly affecting the species'continued existence now and/or into theforeseeable future).

If the coast widepopulation growth rate is stable orsignificantly increasing and one stockcomplex is significantly decreasing butall others are stable or significantly increasing, the species is at a moderate-low risk. A significantly decreasing population growth rate for several stockcomplexes would be an indicator thatthe current abundance may not besustainable relative to currentmanagement measures and, therefore, may warrant further protections.

Thus,if the population growth rates for two ofthe stock complexes are significantly decreasing but the coast-wide index issignificantly increasing, the species is atmoderate-high risk. If the growth ratesfor three or more of the stock complexes are significantly decreasing and/or thecoast-wide index is significantly decreasing, the species is at high risk.Risk Scenarios

• Low risko Coast wide trajectory-Stable tosignificantly increasing o Stock complex trajectories-All stable to significantly increasing

" Moderate-Low risko Coast wide trajectory-Stable tosignificantly increasing o Stock complex trajectories-One significantly decreasing,all othersstable to significantly increasing

• Moderate-High risko Coast wide trajectory-Stable tosignificantly increasing o Stock complex trajectories-Two ormore significantly decreasing

• High risko Coast wide trajectory-Significantly decreasing o Stock complex trajectories-Three or more significantly decreasing Trend Analysis ModelingThe sections below includesummaries/excerpts from the NEFSCReport to the SRT, "Analysis of Trendsin Alewife and Blueback HerringRelative Abundance,"

June 17, 2013, 42pp. (NEFSC, 2013). For detailedinformation on the modeling conducted, please see the complete report whichcan be found at http://www.nero.noaa.gov/prot resiCandidateSpeciesProgram/

RiverHerringSOC.htm or see FORFURTHER INFORMATION CONTACT sectionabove for contacts.

48988Federal Register/Vol.

78, No. 155/Monday, August 12, 2013/Notices Data Used in the Trend AnalysisModelingRangewide DataRelative abundance indices frommultiple fishery-independent surveytime series were considered as possibledata inputs for the rangewide analysis.

These time series included the NEFSCspring, fall, and winter bottom trawlsurveys as well as the NEFSC shrimpsurvey. For alewife, two additional timeseries were available:

Canada's DFOsummer research vessel (RV) survey ofthe Scotian Shelf and Bay of Fundy(1970-present),

and DFO's GeorgesBank RV survey (1987-present, conducted during February and March).For the NEFSC spring and fall bottomtrawl surveys, inshore strata from 8 to27 m depth and offshore strata from 27to 366 m depth have been mostconsistently sampled by the RVAlbatross IV and RV Delaware II sincethe fall of 1975 and spring of 1976. Priorto these time periods, either only aportion of the survey area was sampledor a different vessel and gear were usedto sample the inshore strata (Azarovitz, 1981). Accordingly, seasonal alewifeand blueback herring relativeabundance indices were derived fromthese trawl surveys using both inshoreand offshore strata for 1976-2012 in thespring and 1975-2011 in the fall.Additional relative abundance indiceswere derived using only offshore stratafor 1968-2012 in the spring and 1967-2011 in the fall (from 1963-1967 the fallsurvey did not extend south of HudsonCanyon).

These time series weredeveloped following the samemethodology used in the ASMFC riverherring stock assessment (ASMFC,2012).Through 2008, standard bottom trawltows were conducted for 30 minutes at6.5 km/hour with the RV Albatross IVas the primary survey research vessel(Despres-Patanjo et al., 1988). However,vessel, door and net changes did occurduring this time, resulting in the needfor conversion factors to adjust surveycatches for some species.

Conversion factors were not available for net anddoor changes, but a vessel conversion factor for alewife was available toaccount for years where the RVDelaware II was used. A vesselconversion factor of 0.58 was applied toalewife weight-per-tow indices from theRV Delaware IL. Alewife number-per-tow indices did not require a conversion factor (Byrne and Forrester, 1991).In 2009, the survey changed primaryresearch vessels from the RV Albatross IV to the RV Henry B. Bigelow.

Due tothe deeper draft of the RV Henry B.Bigelow, the two shallowest series ofinshore strata (8-18 m depth) are nolonger sampled.

Concurrent with thechange in fishing vessel, substantial changes to the characteristics of thesampling protocol and trawl gear weremade, including tow speed, net typeand tow duration (NEFSC, 2007).Calibration experiments, comprising paired standardized tows of the twofishing vessels, were conducted tomeasure the relative catchability between the two vessel-gear combinations and develop calibration factors to convert Bigelow surveycatches to RV Albatross equivalents (Miller et al., 2010). In the modeling, theNEFSC developed species-specific calibration coefficients which wereestimated for both catch numbers andweights using the method of Miller et al.(2010) (Table 14). The calibration factorswere combined across seasons due tolow within-season sample sizes from the2008 calibration studies (fewer than 30tows with positive catches by one orboth vessels).

Table 14. Coefficients and associated standard errors used to convert RV Bigelowcatches of alewife and blueback herring to RV Albatross IV equivalents for the 2009-2011 NEFSC bottom trawl surveys.Number BiomassSpecies Coefficient SE Coefficient SEAlewife 1.05 0.16 0.72 0.11Blueback herring 0.87 0.17 1.59 0.45Bottom trawl catches of river herringtend to be higher during the daytimedue to diel migration patterns (Loesch etal., 1982; Stone and Jessop, 1992).Accordingly, only daytime tows wereused to compute relative abundance andbiomass indices.

In addition, thecalibration factors used to convert RVBigelow catches to RV Albatross equivalents were estimated using onlycatches from daytime tows. Daytimetows, defined as those tows betweensunrise and sunset, were identified foreach survey station based on samplingdate, location, and solar zenith angleusing the method of Jacobson et al.(2011). Although there is a clear generalrelationship between solar zenith andtime of day, tows carried out at the sametime but at different geographic locations may have substantially different irradiance levels that couldinfluence survey catchability (NEFSC,2011). Preliminary analyses (LisaHendrickson, NMFS, 2012-unpublished data) confirmed that riverherring catches were generally greaterduring daylight hours compared tonighttime hours.In addition to the NEFSC spring andfall trawl surveys, the NEFSC winterand shrimp surveys were considered forinclusion in the analysis.

For the wintersurvey (February),

the sampling areaextended from Cape Hatteras, NC,through the southern flank of GeorgesBank, but did not include the remaining portion of Georges Bank or the Gulf ofMaine. With the arrival of the RVBigelow in late 2007, the NEFSC wintersurvey was merged with the NEFSCspring survey and discontinued.

Alewife and blueback herring indices ofrelative abundance were developed forthe winter survey from 1992-2007 usingdaytime tows from all sampled inshoreand offshore strata. The shrimp surveyis conducted during the summer (July/August) in the western Gulf of Maineduring daylight hours. Relativeabundance indices were derived foralewife and blueback herring from1983-2011 using all strata that wereconsistently sampled across the surveytime series in the NEFSC winter andshrimp surveys.Stratified mean indices of relativeabundance of alewife from Canada'ssummer RV survey and Georges BankRV survey were provided by Heath Federal Register/Vol.

78, No. 155/Monday, August 12, 2013/Notices 48989Stone of Canada's DFO. In thesesurveys, alewife is the predominant species captured;

however, someblueback herring are likely included inthe alewife indices because catches arenot always separated by river herringspecies (Heath Stone, DFO Pers. comm.,2012). Furthermore, some Georges Bankstrata were not sampled in all years ofthe survey due to inclement weatherand vessel mechanical problems (Stoneand Gross, 2012).Due to the restricted spatial coverageof the winter, shrimp and CanadianGeorges Bank surveys, these surveyswere not used in the final rangewide analyses.

Accordingly, relativeabundance (number-per-tow) from theNEFSC spring and fall surveys was usedin the rangewide models for bluebackherring, and number-per-tow from theNEFSC spring survey, NEFSC fallsurvey, and the Canadian summersurvey were used in the rangewide models for alewife.Data from 1976 through the presentwere incorporated into the trendanalysis.

This time series permitted theinclusion of the spring and fall surveys'inshore strata. In addition, with thistime series, the required assumption that the population growth rate willremain the same was reasonable.

Priorto 1976, fishing intensity was muchgreater due to the presence of distantwater fleets on the East Coast of theUnited States.Years with zero catches were treatedas missing data. For alewife, there wereno years with zero catches in the spring,fall and Scotian shelf surveys.

Zerocatches of blueback herring occurred inthe fall survey in 1988, 1990, 1992 and1998.Stock-Specific DataStock-specific time series of alewifeand blueback herring relativeabundance were obtained from theASMFC and Canada's DFO. Available time series varied among stocks andincluded run counts, as well as young-of-year (YOY), juvenile and adultsurveys that occurred solely within thebays or sounds of the stock of interest(for alewife see Table 15 in the NEFSC's"Analysis of Trends in Alewife andBlueback Herring Relative Abundance,"

and for blueback

herring, see Table 16).All available datasets were included inthe stock-specific

analyses, with theexception of run counts from the St.Croix and Union Rivers. These datasetswere excluded due to the artificial impacts of management activities on runsizes. The closure of the Woodland Damand Great Falls fishways in the St. CroixRiver prevented the upstream passage ofalewives to spawning habitat.

Incontrast, fluctuations in Union Riverrun counts were likely impacted bylifting and stocking activities used tomaintain a fishery above the Ellsworth Dam. In the southern Gulf of St.Lawrence trawl survey, all river herringwere considered to be alewife becausesurvey catches were not separated byriver herring species (Luc Savoie DFO,Pers. comm., 2012). No blueback herringabundance indices were available forthe Canadian stock. Select strata werenot used to estimate stock-specific indices from the NEFSC trawl surveysbecause mixing occurs on thecontinental shelf. Accordingly, anyNEFSC trawl survey indices, evenestimated using only particular strata,would likely include individuals frommore than one stock.Each available dataset in the stock-specific analyses represented aparticular age or stage (spawners, young-of-year, etc.) of fish.Consequently, each time series wastransformed using a running sum over 4years. The selection of 4 years for therunning sum was based on thegeneration time of river herring.

For age-and stage-specific data, a running sumtransformation is recommended toobtain a time series that more closelyapproximates the total population (Holmes, 2001). In order to compute therunning sums for each dataset, missingdata were imputed by computing themeans of immediately adjacent years.For both species 4 years were imputedfor the Monument River, and 1 year wasimputed for the DC seine survey. Foralewife, 1 year was also imputed for theMattapoisett River, Nemasket River, andthe southern Gulf of St. Lawrence trawlsurvey. For blueback

herring, 1 year wasalso imputed for the Long Island Sound(LIS) trawl survey and Santee-Cooper catch-per-unit-effort (CPUE).If possible data from 1976 through thepresent were incorporated into eachstock-specific model, with the firstrunning sum incorporating data from1976 through 1979. However, for somestocks, observation time series beganafter 1976. In these cases, the firstmodeled year coincided with the firstrunning sum of the earliest survey.MARRS Model Description Multivariate Autoregressive State-Space models (MARSS) were developed using the MARSS package in R (Holmeset al., 2012a). This package fits linearMARSS models to time series data usinga maximum likelihood framework basedon the Kalman smoother and anExpectation Maximization algorithm (Holmes et al., 2012b).Each MARSS model is comprised ofa process model and an observation model (Holmes and Ward, 2010; Holmeset al., 2012b). The model is described indetail in the NEFSC (2013) final reportto the SRT (posted on the Northeast Regional Office's Web site-http://

www.nero.noao.gov/prot res!Can didateSpeciesProgram!

RiverHerringSOC.htm).

Population projections and model analysis.

For each stock complex, the estimated population growth rate and associated 95 percent confidence intervals wereused to classify whether the stock'srelative abundance was stable,significantly increasing or decreasing.

As noted previously, relative abundance of a stock was considered to besignificantly increasing or decreasing ifthe 95 percent confidence intervals ofthe population growth rate did notinclude zero. In contrast, if the 95percent confidence intervals includedzero, the population was considered tobe stable because the increasing ordecreasing trend in abundance was notsignificant.

Model ResultsRangewide AnalysesFor the rangewide

analysis, as shownin Table 15 below, the preferred modelrun for alewife indicates that the 95-percent confidence intervals spanningthe estimated population growth rate donot include 0 and are statistically significantly increasing.

For bluebackherring rangewide,

however, the 95-percent confidence intervals do include0, and thus, it is not possible to statethat the trend rangewide for this speciesis increasing.

We, therefore, concludebased on our criteria described abovethat blueback herring rangewide arestable.

48990Federal Register/Vol.

78, No. 155/Monday, August 12, 2013/Notices Table 15. Population growth rate maximum likelihood estimates (ML.Est),

associated standard errors (Std.Err) and lower and upper 95 percent confidence intervals (low.CI,up.CI) for each rangewide model run. The preferred model run (lowest AIC) for eachspecies is highlighted in grey.Species Run ML.Est Std.Frr Iow.Cl u.ClIIndependent with equal variances 0.034 0.006 0.022 0.046Independent with unequal variances

,; 0.0321Ofý 0 , 2, '. P0..4Y)Alewife Unconstrained 0.030 0.005 0.020 0.041Unequal variances with one covariance term 0.035 0.013 0.009 0.062Equal variance and covariance 0.034 0.005 0.023 0.045Independent with equal variances

`0. .0X 0.,1iBlueback herring Independent with unequal variances 0.022 0.036 -0.047 0.093Unconstrained 0.026 0.045 -0.063 0.112Equal variance and covariance 0.040 0.052 -0.064 0.144Stock-Specific AnalysesAs shown in Table 16 below, the 95-percent confidence intervals spanningthe estimated population growth rate forthe Canadian stock complex do notinclude 0 and are statistically significantly increasing.

For the otherthree stock complexes,

however, theconfidence intervals do include 0, andthus, the Northern New England,Southern New England and mid-Atlantic alewife stock complexes arestable.As Canada does not separate alewifeand blueback herring in their surveys(e.g., they indicate that all fish arealewife),

we were unable to obtain datafrom Canada specifically for bluebackherring.

For three of the remaining fourstock complexes, the 95-percent confidence intervals spanning theestimated population growth rate doinclude 0 and thus, the trend for thesestock complexes is stable. For the mid-Atlantic stock complex, the population growth rate and both 95-percent confidence intervals are all statistically significantly decreasing.

Thus, weconclude that this stock complex issignificantly decreasing.

BILLING CODE 3510-22-P Table 16. Population growth rate maximum likelihood estimates (ML.Est),

associated standard errors (Std.Err) and lower and upper95-percent confidence intervals (low.CI, up.CI) for each stock-specific model run. The preferred model run (lowest AIC) for eachstock is highlighted in grey.SpeciesStockRunAlewifeMid AtlanticIndependent with equal variances Independent with unequal variances Unconstrained Unequal variances with one covariance termSouthern New EnglandNorthern New EnglandEqual variance and covariance Independent with equal variances Independent with unequal variances Equal variance and covariance Independent with equal variances Unconstrained Equal variance and covariance MLEst StdErr low.Cl up.CI0.004 0.034 -0.061 0.073::p-O021:~.3 W P 9~ 0-0.013 0.029 -0.071 0.044-0.021 0.035 -0.088 0.054-0.004 0.046 -0.092 0.0880.008 0.032 -0.052 0.0720.005 0.032 -0.057 0.0690.038 0.036 -0.034 0.1080.036 0.041 -0.048 0.114Cn00CD0z0DCJICanadaIndependent with equal variances Blueback herring Southern Independent with equal variances

-0.004 0.047 -0.091 0.091Independent with unequal variances 0: O ',02 04LUnconstrained 0.024 0.042 -0.058 0.103Equal variance and covariance

-0.001 0.046 -0.091 0.092Mid Atlantic Independent with equal variances

-0.070 0.008 -0.085 -0.055Independent with unequal variances 05Equal variance and covariance

-0.072 0.013 -0.097 -0.046Southem New England Independent with equal variances

.,,Northern New England Independent with equal variances

-4 48992Federal Register/Vol.

78, No. 155/Monday, August 12, 2013/Notices BILLING CODE 3510-22-C Model Assumptions and Limitations The available data for each analysisvaried considerably among species andstocks. Some stocks such as SouthernNew England blueback herring had onlyone available data set; however, otherstocks such as Southern New Englandalewife and mid-Atlantic bluebackherring had eight or more available timeseries. Within each analysis, all inputtime series must be weighted equally,regardless of the variability in thedataset.

Furthermore, only the annualpoint estimates of relative abundance are inputs to the model; associated standard errors for the time series arenot inputted.

However, some observation timeseries may be more representative of thestock of interest than other time series.For example, for Northern New Englandalewife, available datasets included runcounts from five rivers and Maine'sjuvenile alosine seine survey. Each timeseries of run counts represents thespawning population in one particular river, whereas the juvenile seine surveysamples six Maine rivers including Merrymeeting Bay (ASMFC, 2012).Accordingly, it is possible that thejuvenile seine survey provides a betterrepresentation of Northern New Englandalewife than the run counts from anyparticular river because the seine surveysamples multiple populations.

Likewise, for Southern New England alewife,available datasets included the LongIsland Sound (LIS) trawl survey, NewYork juvenile seine survey, and runcounts from six rivers. The LIS trawlsurvey samples Long Island Sound fromNew London to Greenwich Connecticut with stations in both Connecticut andNew York state waters, including themouths of several rivers including theThames, Connecticut, Housatonic, Eastand Quinnipiac (CTDEP, 2011; ASMFC,2012). The NY juvenile seine surveysamples the Hudson River estuary(ASMFC, 2012), and run counts arespecific to particular rivers. As aconsequence, the LIS trawl survey maybe more representative of the SouthernNew England alewife stock because itsamples not only a greater proportion ofthe stock, but also samples LIS wheremixing of river-specific populations likely occurs.Several sources of uncertainty aredescribed in detail in the modelingreport. It is important to understand anddocument these sources of uncertainty.

However, even with severalassumptions and these sources ofuncertainty, we are confident that themodel results are useful in determining the population growth rates both coast-wide and for the individual stockcomplexes, and thus, for providing information to be used in assessing therisk to these species and stockcomplexes.

Extinction Risk Conclusion In performing our analysis of the riskof extinction to the species, weconsidered the current status and trendsand the threats as they are impacting thespecies at this time. Currently, neitherspecies is experiencing high rates ofdecline coast-wide as evidenced by therangewide trends (significantly increasing for alewife and stable forblueback herring).

Thus, using theextinction risk tiers identified by theSRT, we have concluded the following:

Alewife-* Tier A: There is sufficient information available to conclude thatthere are at least three contiguous populations that are stable tosignificantly increasing.

• Tier B: The species is at "Low risk"as the coast-wide trajectory issignificantly increasing and all of thestock complexes are stable orsignificantly increasing.

Blueback herring-s Tier A: There is insufficient information available to make aconclusion under Tier A as we wereunable to obtain data from Canada todetermine the population growth ratefor rivers in Canada. Thus, we were onlyable to obtain information for four of thefive stock complexes identified for thespecies.* Tier B: The species is at "Moderate-low risk "as the coast-wide trajectory isstable and three of the four stockcomplexes are stable. The estimated population growth rate of the mid-Atlantic stock complex is significantly decreasing based on the available information.

However, the relativeabundance of the species throughout itsrange (as demonstrated through thecoast-wide population growth rate) isstable, and thus, the SRT concluded thatthe mid-Atlantic stock complex does notconstitute a significant portion of thespecies range. We concur with thisconclusion.

In other words, the dataindicate that the mid-Atlantic stockcomplex does not contribute so much tothe species that, without it, the entirespecies would be in danger ofextinction.

Many conservation efforts areunderway that may lessen the impact ofsome of these threats into theforeseeable future. One of the significant threats identified for both species isbycatch in Federal fisheries, such as theAtlantic herring and mackerel fisheries.

The New England and Mid AtlanticFishery Management Councils haverecommended management measuresunder the MSA that are expected todecrease the risk from this particular threat. Under both the Atlantic HerringFishery Management Plan and theMackerel/Squid/Butterfish FisheryManagement Plan, the Councils haverecommended a suite of reporting, vessel operation, river herring catch capprovisions, and observer provisions thatwould improve information on theamount and extent of river herring catchin the Atlantic herring and mackerelfisheries.

NMFS has partially approvedthe measures as recommended by theNew England Council and will beimplementing the measures inSeptember or October 2013. Anotherthreat that has been identified for bothspecies is loss of habitat or loss of accessto spawning habitats.

We have beenworking to restore access to spawninghabitats for river herring and otherdiadromous fish species through habitatrestoration projects.

While severalthreats may lessen in the future, giventhe extensive decline from historical levels, neither species is thought to becapable of withstanding continued highrates of decline.Research NeedsAs noted above, there is insufficient information available on river herring inmany areas. Research needs wererecently identified in the ASMFC RiverHerring Stock Assessment Report(ASMFC, 2012); NMFS Stock Structure, Climate Change and Extinction RiskWorkshop/Working Group Reports(NMFSa, 2012; NMFSb, 2012; NMFSc,2012) and associated peer reviews; andNew England and Mid-Atlantic FisheryManagement Council documents (NEFMC, 2012; MAFMC, 2012). Wehave identified below some of the mostcritical and immediate research needs toconserve river herring taking therecently identified needs intoconsideration, as well as information from this determination.

However, theseare subject to refinement as acoordinated and prioritized coast-wide approach to continue to fill in data gapsand conserve river herring and theirhabitat is developed (see "ListingDetermination" below).9 Gather additional information onlife history for all stages and habitatareas using consistent andcomprehensive coast-wide protocols (i.e., within and between the UnitedStates and Canada).

This includesinformation on movements such asstraying rates and migrations at sea.Improve methods to develop biological benchmarks used in assessment modeling.

Federal Register/Vol.

78, No. 155/Monday, August 12, 2013/Notices 489939 Continue genetic analyses to assessgenetic diversity, determine population stock structure along the coast (U.S. andCanada) and determination of riverorigin of incidental catch in non-targeted ocean fisheries.

Also, obtaininformation on hybridization andunderstand the effects of stocking ongenetic diversity.

e Further assess human impacts onriver herring (e.g., quantifying bycatchthrough expanded observer and portsampling coverage to quantify fishingimpact in the ocean environment andimprove reporting of commercial andrecreational harvest by waterbody andgear, ocean acidification) a Continue developing models topredict the potential impacts of climatechange on river herring.

This includes, as needed to support these efforts,environmental tolerances andthresholds (e.g., temperature) for all lifestages in various habitats.

• Develop and implement monitoring protocols and analyses to determine river herring population responses andtargets for rivers undergoing restoration (e.g., dam removals,

fishways, supplemental stocking).

Also, estimatespawning habitat by watershed (withand without dams).* Assess the frequency andoccurrence of hybridization betweenalewife and blueback herring andpossible conditions that contribute to itsoccurrence (e.g., occurs naturally or inresponse to climate change, dams, orother anthropogenic factors).

a Continue investigating predatorprey relationships.

Listing Determination The ESA defines an endangered species as any species in danger ofextinction throughout all or a significant portion of its range, and a threatened species as any species likely to becomean endangered species within theforeseeable future throughout all or asignificant portion of its range. Section4(b)(1) of the ESA requires that thelisting determination be based solely onthe best scientific and commercial dataavailable, after conducting a review ofthe status of the species and after takinginto account those efforts, if any, thatare being made to protect such species.We have considered the available information on the abundance of alewifeand blueback

herring, and whether anyone or a combination of the five ESAfactors significantly affect the long-term persistence of these species now or intothe foreseeable future. We havereviewed the information receivedfollowing the positive 90-day finding onthe petition, the reports from the stockstructure, extinction risk analysis, andclimate change workshops/working groups, the population growth ratesfrom the trends in relative abundance estimates and qualitative threatsassessment, the Center for Independent Experts peer reviewers'

comments, otherqualified peer reviewer submissions, and consulted with scientists, fishermen, fishery resource

managers, and Native American Tribes familiarwith river herring and related researchareas, and all other information encompassing the best available information on river herring.

Based onthe best available information, the SRTconcluded that alewife are at a low riskof extinction from the threats identified in the QTA (e.g., dams and otherbarriers to migration, incidental catch,climate change, dredging, water quality,water withdrawal/outfall, predation, and existing regulation),

and bluebackherring are at a moderate-low risk ofextinction from similar threatsidentified and discussed in the QTAdiscussion above. We concur with thisconclusion, and we have determined that as a result of the extinction riskanalysis for both species, these twospecies are not in danger of extinction or likely to become so in the foreseeable future. Therefore, listing alewife andblueback herring as either endangered or threatened throughout all of theirranges is not warranted at this time.Significant Portion of the RangeEvaluation Under the ESA and our implementing regulations, a species warrants listing ifit is threatened or endangered throughout all or a significant portion ofits range. In our analysis for this listingdetermination, we initially evaluated the status of and threats to the alewifeand blueback herring throughout theentire range of both species.

As statedpreviously, we have concluded thatthere was not sufficient evidence tosuggest that the genetically distinctstock complexes of alewife or bluebackconstitute DPSs. We also then assessedthe status of each of the individual stockcomplexes in order to determine whether either species is threatened orendangered in a significant portion of itsrange.As noted above in the QTA section,the SRT determined that the threats toboth species are similar and the threatsto each of the individual stockcomplexes are similar with some slightvariation based on geography.

Waterquality, water withdrawal/outfall, predation, climate change and climatevariability were generally seen as greaterthreats to both species in the southernportion of their ranges than in thenorthern portion of their ranges. In lightof the potential differences in themagnitude of the threats to specificareas or populations, we next evaluated whether alewife or blueback herringmight be threatened or endangered inany significant portion of its range. Inaccordance with our draft policy on"significant portion of its range," ourfirst step in this evaluation was toreview the entire supporting record forthis listing determination to "identify any portions of the range[s]

of thespecies that warrant furtherconsideration" (76 FR 77002; December9, 2011). Therefore, we evaluated whether there is substantial information suggesting that the hypothetical loss ofany of the individual stock complexes for either species (e.g., portions of thespecies' ranges) would reasonably beexpected to increase the demographic risks to the point that the species wouldthen be in danger of extinction, (i.e.,whether any of the stock complexes within either species' range should beconsidered "significant").

As noted inthe extinction risk analysis

section, allof the alewife stock complexes as wellas the coastwide trend are either stableor increasing.

For blueback

herring, 3 ofthe stock complexes and the coastwide trend are all stable, but the mid-Atlantic stock complex is decreasing.

The SRTdetermined that the mid-Atlantic stockcomplex is not significant to the species,given that even though it is decreasing, the overall coastwide trend is stable.Thus, the loss of this stock complexwould not place the entire species atrisk of extinction.

We concur with thisconclusion.

Because the portion of theblueback herring stock complex residingin the mid-Atlantic is not so significant that its hypothetical loss would renderthe species endangered, we concludethat the mid-Atlantic stock complexdoes not constitute a significant portionof the blueback herring's range.Consequently, we need not address thequestion of whether the portion of thespecies occupying this portion of therange of blueback herring is threatened or endangered.

Conclusion Our review of the information pertaining to the five ESA section 4(a)(1)factors does not support the assertion that there are threats acting on eitheralewife or blueback herring or theirhabitat that have rendered either speciesto be in danger of extinction or likely tobecome so in the foreseeable future,throughout all or a significant portion ofits range. Therefore, listing alewife orblueback herring as threatened orendangered under the ESA is notwarranted at this time.

48994Federal Register/Vol.

78, No. 155/Monday, August 12, 2013/Notices While neither species is currently endangered or threatened, both speciesare at low abundance compared tohistorical levels, and monitoring bothspecies is warranted.

We agree with theSRT that there are significant datadeficiencies for both species, and thereis uncertainty associated with available data. There are many ongoingrestoration and conservation efforts andnew management measures that arebeing initiated/considered that areexpected to benefit the species;however, it is not possible at this timeto quantify the positive benefit fromthese efforts.

Given the uncertainties and data deficiencies for both species,we commit to revisiting both species in3 to 5 years. We have determined thatthis is an appropriate timeframe forconsidering this information in thefuture as a 3- to 5-year timeframe equates to approximately one generation time for each species, and it is therefore unlikely that a detrimental impact toeither species could occur within thisperiod. Additionally, it allows for timeto complete ongoing scientific studies(e.g., genetic analyses, ocean migration

patterns, climate change impacts) andfor the results to be fully considered.

Also, it allows for the assessment of datato determine whether the preliminary reports of increased river counts inmany areas along the coast in the last 2years represent sustained trends. Duringthis 3- to 5-year period, we intend tocoordinate with ASMFC on a strategy todevelop a long-term and dynamicconservation plan (e.g., priorityactivities and areas) for river herringconsidering the full range of bothspecies and with the goal of addressing many of the high priority data gaps forriver herring.

We welcome input andinvolvement from the public. Anyinformation that could help this effortshould be sent to us (see ADDRESSES section above).References CitedA complete list of all references citedin this rulemaking can be found on ourWeb site at http://www.nero.noaa.govl prot res/CandidateSpeciesProgram/

RiverHerringSOC.htm and is available upon request from the NMFS office inGloucester, MA (see ADDRESSES).

Authority:

The authority for this action isthe Endangered Species Act of 1973, asamended (16 U.S.C. 1531 et seq.).Dated: August 6, 2013.Alan D. Risenhoover,

Director, Office of Sustainable Fisheries, performing the functions and duties of theDeputy Assistant Administrator forRegulatory Programs National MarineFisheries Service.[FR Doc. 2013-19380 Filed 8-9-13; 8:45 am]BILLING CODE 3510-22-P NOTE: This is a REVISED version of the plan, originally posted to the DEC website inA ugust 2011. Changes were made as a result of public comment received by Sept 22, 2011.SNew York StateaDepartment of Environmental Conservation Sustainable Fishing Plan for New York River Herring StocksKathryn A. Hattala, Andrew W. KahnleBureau of Marine Resources, Hudson River Fisheries UnitandRobert D. AdamsHudson River Estuary ProgramSeptember 2011Submitted for reviewto theAtlantic State Marine Fisheries Commission REVISED VERSION:

September 2011, based on public comment received.

EXECUTIVE SUMMARYAmendment 2 to the Atlantic States Marine Fisheries Commission Shad and river HerringInterstate Fishery Management Plan requires member states to demonstrate that fisheries forriver herring (alewife and blueback herring) within their state waters are sustainable.

Asustainable fishery is defined as one that will not diminish potential future reproduction andrecruitment of herring stocks. If states cannot demonstrate sustainability to the Atlantic StatesMarine Fisheries Commission (ASMFC),

they must close their herring fisheries.

New York State proposes to maintain a restricted river herring (alewife and blueback herring)fishery in the Hudson River and tributaries and to close river herring fisheries elsewhere in theState. This proposal conforms to Goal 1 of the New York State Hudson River Estuary ActionAgenda.Stock StatusBlueback herring and alewife are known to occur and spawn in New York State in the HudsonRiver and tributaries, the Bronx River, and several streams on Long Island. The Hudson River istidal to the first dam at Troy, NY (rkm 245). Data on stock status are available for the HudsonRiver and tributaries.

Few data are available on river herring in streams in Bronx County,southern Westchester County, or on Long Island. River herring are absent in the New Yorkportion of the Delaware River.Hudson River: Commercial and recreational fisheries exploit the spawning populations of riverherring in the Hudson River and tributaries.

Fixed and drifted gill, cast and scap/lift nets are usedin the main stem Hudson, while scap/lift and cast nets are used in the tributaries.

Recreational fishers often use commercial net gears because permit fees remain at 1911 levels. Anglers alsoare allowed take of river herring with variety of small nets and hook and line. In the last tenyears, about 250 fishers annually purchased commercial gill net permits and approximately 240purchased commercial scap net permits.

However only 84 gill net and 93 scap/lift fishersreported using the gear licensed.

Fishers using commercial gears are required to report landingsannually.

Most river herring taken in the Hudson and tributaries are used as bait in therecreational striped bass fishery.

Anglers and subsistence fishers take a few river herring fromLong Island streams.Data on commercial harvest of river herring are available since the early 1900s. Landings peakedin the early 1900s and in the 1930s and then declined through the 1980s. Landings increased again through 2003, but have since declined.

Reported commercial harvest has remained below50,000 river herring per year since the early 1990s. A series of creel surveys and estimates since2001 indicated substantial and increasing harvest of river herring by recreational anglers fromthe Hudson River and tributaries.

We estimated that approximately 240,000 river herring wereharvested by recreational anglers in 2007. The extent of the loss of river herring through bycatchin ocean commercial fisheries remains largely unknown but is expected to be significant.

2 REVISED VERSION:

September 2011, based on public comment received.

Fishery dependent data on river herring status since 2000 are available from commercial reportsand from on-board monitoring.

Catch per unit effort (CPUE) in fixed (anchored) gill nets fishedin the main stem river has increased.

Conversely, CPUE in scap nets fished in tributaries initially

declined, but then varied without trend. Mean length of river herring observed in the commercial harvest has declined slightly since 2000. We feel that the CPUE in fixed gear below the BearMountain Bridge provides the best annual measure of abundance because it intercepts riverherring migrating past the gear to upriver spawning locations..

Fishery independent data on size and age composition of river herring spawning in the HudsonRiver Estuary are available from 1936 and intermittently since the late 1970s. Sample size hasbeen small in most years. The largest fish were collected in the 1930s. Size of both bluebackherring and alewife has declined over the last 30 years. Age data were obtained from scales in1936 and the late 1980s. Since then, ages were estimated from age length keys developed byMaine, Massachusetts, and Maryland.

Observed and estimated age at length of Hudson Riverfish varied substantially among methods and thus age can only be used for trends within method.Annual mean age since the late 1980s has remained stable in blueback herring and femalealewife, but declined in male alewife.

Because of the uncertainty with estimated ages, weestimated annual mortality with length-based methods.

Estimates varied substantially depending on assumed model inputs and therefore actual total mortality on the stocks remains unknown.However, we should emphasize that mortality on stocks must have been high in the last 30 yearsto have so consistently reduced mean size and presumably mean age. Within method, estimates of total mortality generally increased for both species since 1980. This increase was mostpronounced in alewife.Young of year production has been measured annually by beach seine since 1980. CPUE ofalewife remained low through the late 1990s and has since increased erratically.

CPUE of youngof year blueback herring has varied with a very slight downward trend since 1980.Streams on Long Island, Bronx and south shore of Westchester County: Limited data have beencollected for some of the river herring populations in these areas. The data are not adequate tocharacterize stock condition.

Delaware River in New York: No records exist to document the presence of river herring in thisportion of the river.Proposed Fishery for the Hudson RiverGiven the inconsistent measures of stock status described above, we do not feel that the datawarrant a complete closure of the Hudson River fishery at this time. New York State proposes afive year restricted fishery in the main-stem Hudson River, a partial closure of the fishery intributaries, and annual stock monitoring.

We set a sustainability target for juvenile indices.

Wewill monitor, but not set targets for mean length from fishery independent spawning stocksampling and CPUE in the commercial fixed gill net fishery in the lower river below the BearMountain Bridge. We will also monitor age structure, frequency of repeat spawning, and totalmortality from fishery independent sampling if we can resolve problems with age determination 3

REVISED VERSION:

September 2011, based on public comment received.

and mortality estimation.

A summary of existing and proposed restrictions is provided.

Proposed restrictions to therecreational fishery include:

a ten fish per day creel limit for individual anglers with a boat limitof 50, and a 10 fish creel limit per day for paying customers with a boat limit of 50 for chartervessels, no fishing within 825 ft (250m) of any man made or natural barrier in the main river andtributaries, no use of nets in tributaries, and the continuation of various small nets in the mainriver. Proposed restrictions to the commercial fishery and use of commercial gears include:

acommercial verification requirement; a net ban in the upper 28 km of the main-stem estuary,shad spawning flats, or tributaries; gill net mesh and size restrictions; a ban on fixed gears ornight fishing above the Bear Mountain Bridge; seine and scap/lift net size restrictions; extension of existing 36 hour4.166667e-4 days <br />0.01 hours <br />5.952381e-5 weeks <br />1.3698e-5 months <br /> lift period to all commercial net gears; increased net fees to account forinflation since 1911 when fees were set or the preferred option of creation of a new HudsonRiver Commercial Fish Permit; extension of the current Marine and Coastal District Charter/Party boat license to the tidal Hudson and tributaries at a cost of $250.00 annually; and monthlymandatory reporting of catch and harvest.We should note that Draft Addendum 3 to Amendment 6 of the ASMFC Interstate Management Plan for striped bass stipulates that states should reduce fishing mortality on spawning stocks by50%. If this draft is approved by the ASMFC Striped Bass Management Board, we may have torestrict effort in the recreational striped bass fishery.

Restrictions may include a reduction in useof bait such as river herring.

Any reduction in effort will likely reduce demand for river herringand thus reduce losses in the Hudson stocks.Proposed Moratorium for streams on Long Island, Bronx County, the southern shore ofWestchester County, and the Delaware River and its tributaries north of Port Jervis NY. Due tothe inability to determine stock condition for these areas, the ASMFC Amendment 2 requiresthat a moratorium on river herring fishing be implemented.

This SFP does not directly address ocean bycatch but focuses on fisheries in New York Statewaters. New York is working with the National Marine Fisheries

Service, the New EnglandFishery Management Council and the Mid-Atlantic Fishery Management Council to deal withthis issue. Both councils are in the process of amending the Atlantic Herring and the AtlanticMackerel, Squid and Butterfish Plans to reduce bycatch of river herring.4 REVISED VERSION:

September 2011, based on public comment received.

1 CONTENTS2 INTRODUCTION

...........................................................................................................

63 MANAGEM ENT UNITS ...................................................................................................

63.1 Description of the M anagement Unit Habitat .............................................................

73.1.1 Hudson River and tributaries

................................................................................

73.1.2 Long Island and W estchester County .................................................................

83.1.3 Delaware River .....................................................................................................

93.2 Habitat Loss and Alteration

..........................................................................................

93.3 Habitat W ater Quality ................................................................................................

104 STOCK STATUS .................................................................................................................

104.1 Fisheries Dependent Data ..........................................................................................

114.1.1 Commercial Fishery ............................................................................................

114.1.2 Recreational Fishery ...........................................................................................

154.2 Fishery Independent Surveys ....................................................................................

164.2.1 Spawning Stock Surveys -Hudson River ..........................................................

164.2.2 Hudson River Spawning Stock -Characteristics

.................................................

174.2.3 Spawning Stock Surveys -Long Island ............................................................

194.2.4 Volunteer and Other river herring monitoring

...................................................

194.2.5 Young-of-the-Year Abundance

..........................................................................

204.2.6 Conclusion

.........................................................................................................

215 PROPOSED FISHERY CLOSURES

................................................................................

215.1 Long Island, Bronx County and W estchester County ..............................................

215.2 Delaware River ..........................................................................................................

216 PROPOSED SUSTAINABLE FISHERY ........................................................................

226.1 Hudson River and Tributaries

....................................................................................

226.1.1 Proposed Restrictions

-Recreational Fishery ...................................................

236.1.2 Proposed Restrictions

-Commercial Fishery ...................................................

257 PROPOSED M EASURES OF SUSTAINABILITY

........................................................

277.1 T argets ............................................................................................................................

2 77.2 Sustainability M easures ..............................................................................................

288 REFERENCES

.....................................................................................................................

295 REVISED VERSION:

September 2011, based on public comment received.

2 INTRODUCTION Amendment 2 to the Atlantic States Marine Fisheries Commission Shad and River HerringInterstate Fishery Management Plan was adopted in 2009. It requires member states todemonstrate that fisheries for river herring (alewife and blueback herring) within state waters aresustainable.

A sustainable fishery is defined as one that will not diminish potential futurereproduction and recruitment of herring stocks. If states cannot demonstrate sustainability toASMFC, they must close their herring fisheries.

The following proposes a plan for a sustainable fishery for river herring in waters of New YorkState. The goal of this plan is to ensure that river herring resources in New York provide a sourceof forage for New York's fish and wildlife and provide opportunities for recreational andcommercial fishing now and in the future.The fisheries that existed back in colonial days in the Hudson Valley of New York undoubtedly included river herring among the many species harvested.

River herring, comprised of bothalewife (Alosa pseudoharengus),

and blueback herring (Alosa aestivalis) were among the fishmentioned by early explorers and colonists

-the French Jesuits, Dutch and English.Archaeological digs along the Hudson in Native American middens indicates that the fisheryresources in the river provided an important food source to Native Americans.

Written records for river herring harvest in New York begin in the early 1900. Landings peakedin the early 1900s and in the 1930s and then declined through the 1980s. Landings increased again through 2003, but have since declined.

Factors in addition to fishing have affected thestocks: habitat destruction (filling of shallow water spawning habitat) and water quality problemsassociated with pollution that caused oxygen blocks in major portions of the river (Albany andNew York City). Water quality has improved over the last 30 years.New York State does not augment wild river herring stocks with hatchery progeny.

The NewYork City Parks Department initiated an experimental restoration program in which alewife werecaptured in a Long Island Sound tributary in Connecticut and released in the Bronx River abovethe first barrier.

Limited returns to the river suggest that some reproduction has occurred fromthese stockings.

A variety of non-governmental organizations along with state and federalagencies are working on development of fish passage for alewife in Long Island streams3 MANAGEMENT UNITSThe management unit for river herring stocks in New York State comprises three sub-units.

Allunits extend throughout the stock's range on the Atlantic coast.* The largest consists of the Hudson River Estuary from the Verrazano Narrows at NewYork City to the Federal Dam at Troy including numerous tributary streams (Figure 1)." The second is made up of all Long Island streams that flow into waters surrounding LongIsland and streams on the New York mainland (Bronx and Westchester Counties) that6 REVISED VERSION:

September 2011, based on public comment received.

flow into the East River and/or Long Island Sound (Figure 2).* The third subunit consists of the non-tidal Delaware River and tributaries upriver of PortJervis, NY.Range of the New York river herring along the Atlantic coast is from the Bay of Fundy, Canadaand Gulf of Maine south to waters off Virginia (NAI 2008).A listing of most Hudson River tributaries, and streams on Long Island, and the Bronx andsouthern Westchester Counties are in Appendix Table A.3.1 Description of the Management Unit Habitat3.1.1 Hudson River and tributaries Habitat Description The Hudson River Estuary is tidal its entire length of 246 km from the Battery (tip of Manhattan Island) in New York City to the Federal Dam at Troy (Figure 1). The estuary is fresh waterabove Newburgh (km 90).The estuarine portion of the Hudson River is considered a "drowned" river valley in that thevalley slopes steeply into the river. Many of the tributaries below the Troy Dam are tidal for ashort distance (usually about a kilometer) ending at a natural or man-made

barrier, often built ona natural barrier.

There are approximately 67 primary and secondary, both named and unnamed,tributaries to the tidal portion of the Hudson River Estuary (Figure 1). Schmidt and Cooper(1996) catalogued 62 of these tributaries for the presence or absence of barriers to migratory fish.They found that only one had no barrier for migratory fish, 31 were blocked (either partially orcompletely) by natural barriers, and the remaining 30 had artificial

barriers, dams or culverts, that reduced or eliminated access for fish. We estimated stream length of all these tributaries tobe about 97 km that is accessible to river herring below the first impassable man-made or naturalbarrier.The Mohawk River is the largest tributary to the Hudson River. It enters the Hudson 2 km northof the Troy Dam. Cohoes Falls, a large scenic waterfall of 20 m is the first natural barrier on theMohawk just upriver of the confluence with the Hudson. Access into the Mohawk system wascreated through the Waterford Flight -a series of five locks and dams, built as part of the ErieCanal to circumvent the falls. The canal lock and dam system was built in 1825, to connect theHudson to central New York and Lakes Ontario and Erie. The Canal parallels and/or is part ofthe Mohawk River for the river's entire length to Rome, a distance of 183 km. A series ofpermanent and seasonal pools make up the canal where it intertwines with the Mohawk River.Permanent pools created from hydro-power dams are found in the Waterford section.

Temporary pools are created each year in early spring by removable dams (series of gates) that increasewater levels to 14 feet (4.3 m) while the canal is in operation (May through November).

Duringthe winter months, the river is returned to its natural state of riffles and pools.7 REVISED VERSION:

September 2011, based on public comment received.

Habitat UseHudson River alewife, blueback herring and American shad are spring spawners.

Alewives arethe first of the herring to enter the estuary, arriving as early as mid-March with continued spawning through early May. Blueback herring prefer slightly warmer temperatures and arrivelater, usually in April.Adults of both species spawn in Hudson River tributaries and in the shallow waters of the mainstem Hudson. Alewife prefer to spawn over gravel, sand and stone in back water and eddieswhereas bluebacks tend to spawn in fast moving water over a hard bottom. Herring spawn in thetidal freshwater Hudson from Kingston (km 144) to Troy (km 256) (Figure 1) and its tributaries for approximately six to ten weeks, dependent on water temperature (Smith 1985, Hattala et al.2011). Once spawning ends, most mature fish quickly return to ocean waters. The nursery areaincludes the spawning reach and extends south to Newburgh Bay (km 90), encompassing thefreshwater portion of the Estuary.Some blueback herring of the Hudson River migrate above the Federal Dam at Troy. A fewcontinue upriver in the non-tidal Hudson as far as Lock 4 on the Champlain Canal (NAI 2007).However, most fish turn west into the Mohawk River. This larger portion migrates as far inlandas Rome (439 km inland),

via the Erie Canal and the Mohawk River. The canal system opens inNew York on or about May st. Since most alewives are already spawning by then, they do notmove into the system (J. Hasse, NYSDEC retired, personal communication).

Blueback herring began colonizing the Mohawk River in the 1970s. By 1982, they had migratedinto Oneida Lake in the Great Lakes drainage.

The number of herring using the Mohawkincreased through the 1990s, but since 2000 herring have rarely occurred in the upper end of theRiver. Blueback herring were historically unable to access the Mohawk River until the locks ofthe Erie Canal provided upstream passage into the system. Now that they are established,

however, they have become important forage for local sport fish populations.

3.1.2 Long Island and Westchester CountyThe herring runs in streams on Long Island are comprised almost exclusively of alewife (B.Young, NYSDEC retired, personal communication).

Most streams are relatively short runs tosaltwater from either head ponds (created by dammed streams) or deeper kettle-hole lakes. Eithercan be fed by a combination of groundwater, run-off or area springs.

Spawning occurs in Aprilthrough May in the tidal freshwater below most of the barriers.

Natural passage for spawningadults into the head ponds or kettle lakes is present in very few streams.There have been limited efforts to understand river herring runs on Long Island since 1995.Several known runs of alewives on Long Island occur in East Hampton, Southampton, Riverhead and Brookhaven.

With the advent of a more aggressive restoration effort in Riverhead on thePeconic River other runs have come to light. Since 2006, an annual volunteer alewife spawning8 REVISED VERSION:

September 2011, based on public comment received.

run survey has been conducted.

This volunteer effort basically documents the presence orabsence of alewives in Long Island Coastal Streams.

In 2010 a volunteer investigation wasinitiated to quantify the Peconic River alewife run. Size and sex data have been collected for2010 and 2011. A crude estimate of the runs size was also made in 2010, this effort wasimproved during 2011 with the placement of a video camera for recording alewife passagethrough the fish passage.

These efforts have been undertaken to understand the Long IslandCoastal streams and to improve the runs that exist there.We have no record of river herring in any of the streams in southern Westchester County. In theBronx River (Bronx County) alewives were introduced to this river in 2006 and 2008 and someadult fish returned in 2010. Monitoring of this run is in its early stages.3.1.3 Delaware RiverNo records exist to document the presence of river herring in the New York portion of theDelaware River.3.2 Habitat Loss and Alteration Hudson River: Much spawning and nursery habitat in the upper half of the tidal Hudson was lostdue to dredge and fill operations to maintain the river's shipping channel to Albany. Most of thisloss occurred between the end of the 19th century (NYS Department of State 1990) and the firsthalf of the 20th century.

Preliminary estimates are that approximately 57% of the shallow waterhabitat (1,821 hectares or 4,500 acres) north of Hudson (km 190) was lost to filling (Miller andLadd 2004). Work is in progress to map the entire bottom of the Hudson River. Data from thisproject will be used to characterize and quantify existing spawning and nursery habitat.

Whilemost of the dredge and fill loss affected American shad, it is suspected that herring were alsoaffected as they spawn along the shallow water beaches in the river.Very little, or no, habitat has been lost due to dam construction.

The first major dam wasconstructed in 1826 at Rkm 256 at Troy. Prior to the dam, the first natural barrier occurred atGlens Falls, 32 km above the Troy Dam. The construction of the dam is not known to havereduced spawning or nursery habitat.The introduction of zebra mussels in the Hudson in 1991, and their subsequent explosive growthin the river, quickly caused pervasive changes in the phytoplankton (80% drop) and micro- andmacro- zooplankton (76% and 50% drop respectively) communities (Caraco et al. 1997). Waterclarity improved dramatically (up by 45%) and shallow water zoobenthos increased by 10%.Given these massive changes, (Strayer et al. 2004) explored potential effects of zebra musselimpact on young-of-year (YOY) fish species.

Most telling was a decrease in observed growthrate and abundance of YOY fishes, including both alewife and blueback herring.

It is not yetclear how this constraint affects annual survival and subsequent recruitment.

Long Island: Most all streams on Long Island have been impacted by human use as the9 REVISED VERSION:

September 2011, based on public comment received.

population expanded.

Many streams were blocked off with dams to create head ponds, initially used to contain water for power or irrigation purposes.

The dams remain; only a few withpassage facilities.

Many streams were also impacted by the construction of highways, withinstallations of culverts or other water diversions which impact immigrating fish.Recent efforts at restoration look to provide fish passage over or around these barriers, or evenremoval of small obstructions.

Permanent fish passage was recently installed on the CarmansRiver in the South Shore Estuary near Shirley, NY. This project was the result of advocacy andcooperation by environmental groups and local, state and federal agencies.

Additional protections for the River are assured due to legislation enacted in 2011, and community awareness is building.

An earlier cooperative effort resulted in the installation of a rock ramppassage in the Peconic River within the Peconic Bays Estuary.

Local citizens monitor the springalewife run in this river. As awareness of these successful efforts spreads, interest in replicating that success on other systems grows.3.3 Habitat Water QualityThe Hudson has a very long history of abuse by pollution.

New York City Department ofEnvironmental Protection recognized pollution, primarily sewage, as a growing problem as earlyas 1909. By the 1930s over a billion gallons a day of untreated sewage were dumped into NewYork Harbor. (NYCDEP http:/!/home2.nvc.gov/htmni/dep/htm l/news/hwqs.shtml

)New York City was not the only source of sewage. Most major towns and cities along theHudson added their share. It was so prevalent that the Hudson was often referred to as an opensewer. Biological demand created by the sewage created oxygen blocks that occurred seasonally (generally mid to late summer) in some sections of the river. One of the best known blocksoccurred near Albany in the northern section of the tidal estuary in the 1960s through the 1970s.This block often developed in late spring and remained through the summer months. Itessentially cut off the upper 40 km of the Hudson for use as spawning and nursery habitat.

Asecond oxygen block occurred in the lower river in the vicinity of New York City in latesummer. This block could potentially have affected emigrating age zero river herring.

Thissummer oxygen-restricted area occurred for decades until 1989 when a major improvement in asewage treatment plant came on line in upper Manhattan.

It took decades, but water quality ingeneral has greatly improved in both areas since the implementation of the Clean Water Act inthe 1970s and subsequent reduced sewage loading to the river.4 STOCK STATUSFollowing is a description of all available data for the Hudson's river herring stocks, plus a briefdiscussion of their usefulness as stock indicators.

Sampling data are summarized in Tables 1 and2. Sampling was in support of Goal 1 of the Hudson River Estuary Action Agenda and has beenpartially funded by the Hudson River Estuary Program.10 REVISED VERSION:

September 2011, based on public comment received.

4.1 Fisheries Dependent Data4.1.1 Commercial FisheryCommercial fisheries for river herring in New York State waters occur in the Hudson RiverEstuary and in marine waters around Long Island. Current commercial fishing restrictions forNew York waters are listed in Appendix Table B.The present commercial fishery in the Hudson River and tributaries exploits the spawningmigration of both alewife and blueback herring.

The primary use of commercially caughtherring is for bait in the recreational striped bass fishery.

The herring fishery occurs from Marchinto early June annually, although some fishers report catching herring as late as July.Ocean bycatchRiver herring occur as bycatch in many commercial fisheries which are in the known migratory range of the Hudson stock from North Carolina up to the Gulf of Maine. Fishery bycatch ismostly un-documented but has the potential to harvest Hudson stock and many other stocksalong the coast. In some years, estimated bycatch of river herring in the Atlantic herring fisheryequaled or exceed the total of all coastal in-river landings (Cieri et al. 2008). More recentanalyses by the National Marine Fisheries Service's Northeast Fisheries Science Center (2011)indicated that total annual incidental catch of river herring in all fishing fleets sampled by theNortheast Fisheries Observer Program during 1989-20 10 ranged from 108 to 1867 mt. It is notknown how much of current ocean river herring bycatch consists of Hudson River fish.This SFP does not directly address ocean bycatch but focuses on fisheries in New York Statewaters. New York is working with the National Marine Fisheries

Service, the New EnglandFishery Management Council (www.nefinc.org) and the Mid-Atlantic Fishery Management Council (w-,w.mafinc.org) to deal with this issue. Both councils are in the process of amendingthe Atlantic Herring (Amendment

5) and the Atlantic

Mackerel, Squid and Butterfish (Amendment

14) Plans to reduce bycatch of river herring.Gear Use in the Hudson River and Tributaries The fixed gill net fishery occurs in the mainstem river from km 40 to km 75 (Piermont to BearMountain Bridge, Figure 1). In this stretch, the river is fairly wide (up to 5.5 km) with wide,deepwater

(- six to eight m) shoals bordering the channel.

Fishers use particular locations withinthis section away from the main shipping channel.

Over the past ten years, an average of 22active fishers participated in this lower river fixed gill net fishery annually.

Nets are 3.7 to 183 m(12 to 600 ft) long. Above the Bear Mountain Bridge gill net fishers use both drift (-58%) andfixed gill nets (-42%). These gears are used up to km 225 (Castleton) where the river is muchnarrower (1.6 to 2 km wide). Approximately 60 fishers participate in this mid river gill netfishery.

Nets range in size from 7.6 to 183 m (25 to 600 ft).11 REVISED VERSION:

September 2011, based on public comment received.

The other major gear used in the river herring fishery is scap nets (also known as lift and/or dipnets). The scap/lift net fishery occurs from km 70 to km 130 (Peekskill to New Baltimore),

primarily in the major river herring spawning tributaries.

Scap/lift nets range in size from 0.2 to121.9 m2 (0.5 to 400ft2). On average, about 96 fishers participate annually.

Marine permits are required of fishers to use seines or scap nets greater than 36 ft2, dip or scoopnets exceeding 14 in. in diameter, and all gill nets. Marine permit holders are required to reporteffort and harvest annually to the Department.

Many marine permit holders are recreational anglers taking river herring for personal use as bait or food. It should be noted that over the lastten years, an average of over 260 gill net and 260 scap nets permits were sold annually.

According to the required annual reports,

however, only 36% of the permitees actively catchfish.In addition to Marine permits, New York has a bait license that allows the take and sale of baitfish (river herring included) using seines and cast nets. As no reporting is required for thislicense, harvest of river herring using this license is unknown.Commercial Landings and License Reporting Recorded landings of river herring in New York State began in the early 1900s. Anecdotal reports indicate that herring only played a small part in the historic commercial fishing industryin the Hudson River. Total New York commercial landings for river herring include all herringcaught in all gears and for both marine and inland waters. Several different time series of data arereported including several state sources, National Marine Fisheries Service (NMFS), and morecurrently Atlantic Coastal Cooperative Statistics Program (ACCSP).

NMFS data do not specifyriver or ocean source(s) and landings are often reported as either alewife or blueback

herring, butnot both in a given year. It is unlikely that only one species was caught. From 1995 to thepresent, the Department has summarized landings and fishing effort information from mandatory state catch reports required for Hudson River marine permits.

Full compliance for this reporting started in 2000. All Hudson River data are sent to NMFS and ACCSP for incorporation into thenational databases.

Because of the discrepancies among the data series and the lack of information to assign thelandings to a specific water body source, only the highest value from all sources is used to avoiddouble counting.

Several peaks occur in the river herring landings for New York (Figure 3). Thefirst peak occurred in the early 1900s followed by a lull (with some gaps) until the period priorto, during, and after World War II when landing peaked a second time. By the 1950s landingswere in a serious decline.

A few unusual peaks occurred in the NMFS data series. In 1966, 1.9million kg were landed (omitted on Figure 3), followed by a series of years of low landings withanother peak in 1982. Landings were low, with some data gaps during the rest of the 1980sthrough 1994.Hudson River landings12 REVISED VERSION:

September 2011, based on public comment received.

Since 1995, landings have been separated between the Hudson and other water (marine).

Harvestin the river was relatively low in 1995, but grew in response to the need for bait for theexpanding striped bass recreational fishery.

In-river landings peaked in 2003 and have slowlydeclined since then (Figure 4). The reason for the decline is unknown.

The striped bass fisheryand the need for bait have not diminished.

It is possible that recreational fishers have shiftedharvest to non commercial gears which do not have a mandatory reporting requirement.

Thelandings from these "personal use" gears are unknown.

Reporting rate from fishers usingcommercial gear is unknown.The primary outlet for harvest taken by Hudson River marine permits is for the in-river baitindustry.

Since 2000, most commercially caught river herring have been taken by scap/lift nets(10 year mean of 48% of the catch) (Figure 5). The remaining 52% was split between drift andfixed gill nets.Commercial DiscardsFrom 1996 to 2010, river herring were not reported as discards on any mandatory reportstargeting herring in the Hudson River or tributaries.

Our commercial fisheries monitoring data,however, (See program description below) suggests otherwise.

Since 1995, we have observed a0.12% rate of discard in the anchored gill net fishery.

Reasons for discards are unspecified.

Discard rates are unknown for ocean fisheries.

Hudson River Commercial Catch Rates -Mandatory ReportsRelative abundance of river herring is tracked through catch per unit effort (CPUE) statistics offish taken from the targeted river herring commercial fishery in the Estuary.

All commercial fishers annually fill out mandatory reports.

Data reported include catch, discards, gear, effort,and fishing location for each trip. Data within week is summarized as total catch divided by totaleffort (square yards of net x hours fished),

separately by gear type (fixed gill nets, drift gill nets,and scap nets). Annual means are summarized in two ways. Above the Bear Mountain Bridgeand within the spawning reach, annual CPUE is calculated as total catch/total effort. Below theBear Mountain Bridge (km 75) and thus below the spawning reach, annual CPUE is calculated asan annual sum of weekly CPUE. Here, nets capture fish moving through to reach upriverspawning locations and run size is determined by number (density) of spawners each week aswell as duration (number of weeks) of the run. The sum of weekly CPUE mimics area under thecurve calculations where sampling occurs in succeeding time periods.

The downside of usingreported CPUE to monitor relative abundance is that results can be influenced by inter-annual,

location, and inter-gear differences in reporting rate.We use the CPUE of the fixed gear fishery below the Bear Mountain Bridge for estimating relative abundance because effort expended by the fishery below this bridge is much greater(-70% of fixed gill net effort) than in the river above this point (remaining 30%). Moreover, fixed gear below the bridge (rkm 40 to 75) is always fished in relatively the same location eachyear, is passive in nature, and intercepts fish that pass by. Annual CPUE for the lower river fixedgill net remained relatively flat until 2006 and has since increased (Figure 6).13 REVISED VERSION:

September 2011, based on public comment received.

We do not consider the CPUE of gears fished above the Bear Mountain Bridge and within thespawning reach as reliable an annual abundance indicator as that from fixed gill nets below thebridge. Upriver gears catch fish that are either staging (getting ready to spawn) or moving intoareas to spawn and gears are generally not employed until fish are present.

The gears includedrift gill nets, scap nets and some fixed gill nets (Figure 5). Drift gill net CPUE is also morevariable as it can be actively fished -set directly into a school of fish. Drifted gill net CPUEvaried widely without trend through the time period. Scap net CPUE declined slightly from2000 through 2003, and has since remained relatively stable (Figure 6). Fixed gill nets fishedwithin the spawning reach show the same recent increasing trend as lower in the river, but effortexpended is much less than below Bear Mountain Bridge.Hudson River Commercial Catch Rates -Monitoring ProgramUp until the mid-1990s, the Department's commercial fishery monitoring program was directedat the American shad gill net fishery, a culturally historic and economically important fishery.We expanded monitoring to the river herring fishery in 1996, but were limited by available manpower and the ability to connect with the fishers.

Monitoring focused on the lower riverfixed gill net fishery since we considered it to be a better measure of annual abundance trends(see section above).Data were obtained by observers onboard fishing vessels.

Technicians recorded data on numbersof fish caught, gear type and size, fishing time and location.

Scale samples, lengths and weightsare taken from a subsample of the fisher's catch. CPUE was calculated by the method used forsummarizing mandatory report data (above).Since 1996, 66 trips targeting river herring (lower river: 53; mid and upper river: 13) have beenmonitored.

These trips were sporadic and sample size is low, from one to 11 trips per year.Because of these few samples, the resulting CPUE is considered unreliable for tracking relativeabundance.

However, active monitoring provided the only data on catch composition of thecommercial harvest and we consider these data to be useful.Commercial Catch Monitoring-Size and Age Structure Commercial fixed gill net fishers use I 3/4 to 2 3/4 inch stretch mesh sizes to target herring.

Catchcomposition include fish caught in all meshes. For trend analysis of size change, we subset thedata to include only fish caught in similar size mesh each year; these include gill nets of 2 1/22 and2 3/4 inch mesh.Catch composition varied annually most likely due to the low number of monitored trips eachyear, and the timing of when the trips occurred.

Annual sample size was relatively low, rangingfrom 40 to 185 fish from 2001 to 2007 (Table 3). Alewives were observed more often thanblueback herring.

The species difference may be the result of when the samples occurred (earlyor late in the run). The sex ratio of alewife in the observed catch was nearly equal (- 50:50) in allyears; more blueback herring females were caught than males (60:30 ratio). From 2001 to 2010,14 REVISED VERSION:

September 2011, based on public comment received.

a slight decline was observed in mean total length (mm) for both alewife and blueback herring(Figure 7).Age data for samples collected during the commercial monitoring program are yet to be analyzed(see discussion in Age section under FI programs below).4.1.2 Recreational FisheryHudson River and tributaries:

The recreational river herring fishery exists throughout the main-stem Hudson River, and its tributaries including those in the tidal section and above the TroyDam (Mohawk River). Herring are sought from shore and boat by angling (jigging) and multiplenet gears (see Appendix B). Boat fishers utilize all allowable gears while shore fisherspredominantly use scap/lift nets, or angling (jigging).

Some recreational herring fishers use theircatch as food (smoking/pickling).

However, the recreational herring fishery is driven primarily by the need for bait in the striped bass fishery.The magnitude of the recreational fishery for river herring is unknown for most years. NYSDECcontracted with Normandeau Associates, Inc. to conduct creel surveys on the Hudson River in2001 and 2005 (NAI 2003 and 2007). Estimated catch of river herring in 2001 was 34,777 fishwith a 35.2% retention rate. When the 2001 data were analyzed, NAI found that the total catchand harvest of herring was underestimated due to the angler interview methods.

In the 2001survey, herring caught by fishers targeting striped bass were only considered incidental catch,and not always included in herring total catch and harvest data. Fishers were actually targeting herring and striped bass simultaneously.

Corrections were made to the interview process for the2005 survey and estimated catch increased substantially to 152,117 herring with an increased retention rate of 75.1% (Table 4). Although some fish were reported as released, we considerthese mortalities due to the herring's fragile nature. We also adjusted the 2001 catch using the2005 survey data. The adjusted catch rose to 93,157 fish.We also evaluated river herring use by striped bass anglers using data obtained from ourCooperative Angler Program (CAP). The CAP was designed to gather data from recreational striped bass anglers through voluntary trip reports.

Volunteer anglers log information for eachstriped bass fishing trip including fishing time, location, bait use, and fish caught, including length, and weight, and bycatch.

In 2006 through 2010, volunteer anglers were asked to providespecific information about herring bait use. The annual proportion of angler days where herringwas used for bait ranged from 71% to 93 % with a mean of 77%. The proportion of herring usedby anglers that were caught rather than purchased increased through the time period (Table 4).Herring caught per trip varied from 1.6 to 4.8 and with the highest values in the last two years.Herring purchased per trip ranged from 0.63 to 1.5 with the lowest value in 2009. We calculated the total number of herring caught or purchased by striped bass anglers in 2007 as the estimated number of striped bass trips from a statewide creel survey (90,742)

• average proportion ofangler days using herring in the CAP in 2007 (0.77)

• number of herring caught or purchased pertrip in the CAP (1.8 and 1.7). The result was 125,502 caught and 115,816 bought for a total of241,318 herring used.15 REVISED VERSION:

September 2011, based on public comment received.

The number of river herring taken from the Hudson River and tributaries for personal use as foodby anglers is unknown.Long Island. Alewives can be caught in many of the small streams on Long Island, though onlythe Peconic River sees more than occasional effort. No creel data are available but anecdotal information (B. Young, NYSDEC retired, personal communication) suggests that harvest isrising in the more easily accessible streams.

Herring taken are used for personal consumption aswell as for bait.The town of Southampton, on Long Island's East End, has local ordinances in place to preventfishing (dipping) during the alewife spawning runs.Bronx and Westchester Counties:

We do not know if any fishery occurs in the streams in Bronxand Westchester Counties that empty into the East River and Long Island Sound.4.2 Fishery Independent Surveys4.2.1 Spawning Stock Surveys -Hudson RiverSeveral surveys have sampled the alewife and blueback herring spawning stocks of the HudsonRiver and tributaries.

The spawning stocks are made up of the fish which have escaped fromcoastal and in-river commercial and recreational fisheries.

The earliest data is from a biological survey of the Hudson in 1936 by the then New York StateConservation Department (Greeley 1937). The sample size was small (25 fish) but indicates thefish were relatively large compared to recent data. More recent data on river herring come fromseveral Department surveys.

The longest dataset (1975-2000) is from an annual survey ofchemical contaminants in fish that targeted multiple species within the Hudson River estuary.Fish were collected by electro-fishing and river herring sample size varied among years. In mostyears, length data were recorded for a sub sample of herring.

The Department also conducted atwo-year electro-fishing survey in 1989 and 1990, to examine the population characteristics ofblueback herring in the Hudson and the Mohawk River, the Hudson's largest tributary.

Datawere obtained on length, age, and sex.Limited data on river herring stock characteristics have also been collected during annualmonitoring of American shad and striped bass spawning stocks. Sampling occurs in the main-stem Hudson River between km 145 and 232 from late April through early June. Fish arecollected by haul seines and electro-fishing.

The 10.2 cm stretch mesh in the haul seines wasspecifically designed to catch shad and striped bass and avoid river herring, but some large (>280mm) herring were occasionally retained in these gears. Herring were an incidental catch ofthe electro-fishing.

Data were collected on length, age, and sex of river herring caught in bothgears.16 REVISED VERSION:

September 2011, based on public comment received.

In 1987, the Department began to target adult river herring during the spring spawning stocksurvey. From 1987 to 1990, two small mesh (9.5 mm) beach seines (30.5 and 61m) wereoccasionally used with some success.

In 1998, we specifically designed a small haul seine (91 m)with an appropriate mesh size (5.1 cm) to target herring.

It was designed to capture all sizes ofherring present with the least amount of size, and age, bias. We have used this gear since 1999.Sampling occurs during the shad and bass survey within the area described above, using thesame field crew.We only use data from the least size-biased gears to describe characteristics of the herringspawning stock: electro-fishing, the beach seine (61m) and the herring haul seine (91 m). Assample size varied among years, all data were combined to characterize size and weightcomposition of the spawning population.

Mean total length and weight data are summarized foradults only (>=170mm TL).4.2.2 Hudson River Spawning Stock -Characteristics Mean Size and GrowthMean size of fish has been calculated for all years that samples were obtained (Figure 8). Samplesize is relatively small, however, in most years presented (n<34 fish). Adequate samples (n>34),following the method described by Lynch and Kim (2010) to characterize length (depicted withan X over the graph's data point) were collected in the late 1980s, early 1990s, then occasionally since 2001 for both species.

Lengths have declined since the early 1980s. Since 2000, mean sizeof female alewife has been stable, but declined slightly in males (Figure 8). Mean size ofblueback herring has declined for both sexes from 1989 to the present.AgeThe Department samples from the 1989-1990 were primarily blueback herring.

The agingmethod used was that of Cating (1954), developed for American shad. More recent scale samplesfrom Department surveys remain un-aged and therefore we have limited age or repeat spawndata directly from scales of Hudson River fish. In attempting to age Hudson River herring scales,we relied on techniques used by other state agencies.

As an alternative, and for a very generalpicture of potential age structure, we estimated annual age structure using length at age keysfrom datasets provided by Maine, Massachusetts, and Maryland for alewife and Massachusetts and Maryland for blueback herring.

We found that three state agencies differ enough in theirtechnique to produce variation in the results.Blueback herring:

Age estimates using length-age keys differed from ages assigned by theDepartment for the 1989- 1990 samples and from each other for most years (Figure 9). Ingeneral, keys from MD and MA were mostly in agreement for male blueback herring in mostyears, but MA aged females slightly older (Figure 10). Ages from two through eight werepresent in the spawning stock. Most fish were ages three, four, and five. Mean age remained17 REVISED VERSION:

September 2011, based on public comment received.

relatively stable among years within method (Figure 11).Alewife:

Age estimates using length-age keys from the three states differed from each other foralewife (Figure 12). In general, the ME key resulted in the youngest ages, followed by older agesfrom MA, then MD. Ages from two through eight or nine were present in the spawning stock.Peak age varied with key used and by sex; most fish were ages three or four for males and fouror five for females.

Mean age was youngest for the ME key, older for MA, and oldest for MDage key (Figure 13). Mean age for males was greater in 2001 and 2003, then dropped andremained relatively stable for 2005 through 2010. Mean age for females was slightly lower in2008 and 2009 but by 2010 returned to the same level as estimated for 2001 and 2003.Maximum age that the Hudson River herring stock can attain is unknown.

Jessop (B. JessopDFO retired, personal communication) reported a maximum age of 12 for both alewife andblueback herring for the St. John's River in New Brunswick.

Given current uncertainty about aging methods and age of Hudson River river herring, wesuggest that available estimates should only be used for a general discussion of age structure andfor trends within estimate method. We do not feel that age estimates should be used to monitorchanges in stock status or to set sustainable fishing targets until aging methods can be verified.

This issue is currently being discussed in the ongoing ASMFC River Herring stock assessment where resolution to the differences in ageing methods is being sought..Mortality Estimates The variation in annual age structure translated into comparable variation in estimates of totalmortality when various age-based estimation methods were used. This difficulty in estimating ages precluded the use of age-based mortality estimators.

As an alternative, we explored use ofthe Beverton-Holt length-based method (Gedamke and Hoenig 2006) using growth parameters for length calculated from the 1936 length at age data (see section above). Since the definition oflength at full recruitment (Lc) given by Nelson et al. (2010) seemed arbitrary, we estimated totalmortality using the Nelson et al. (2010) and two additional Lc values. Results from the lengthbased method were also influenced by Loo. The Beverton-Holt method also relies on severalpopulation assumptions including continuous recruitment to the stock that the population is inequilibrium.

Neither of these assumptions are true for Hudson herring stocks.Total mortality estimates for alewife of both sexes varied tremendously within and among yearsdepending on assumed model inputs (Figure 14). Estimates increased until 2006, after which adecline occurred to 2010. An even greater variation occurred for blueback herring (Figure 15)with a series of very high peaks followed by low values. Given this demonstrated sensitivity tomodel inputs, we suggest that total mortality of Hudson River river herring stocks remainsunknown.

However, we should emphasize that mortality on stocks must have been high in thelast 30 years to have so consistently reduced mean size and presumably mean age. We do notfeel that estimates of total mortality should be used to monitor stock change during the proposedexperimental fishery unless uncertainty in estimation methodology can be resolved.

Currentuncertainty precludes use of total mortality to set sustainability targets.18 REVISED VERSION:

September 2011, based on public comment received.

4.2.3 Spawning Stock Surveys -Long IslandYoung (2011) sampled alewife in the Peconic River 32 times throughout the spawning season in2010. Sampling occurred by dip net just below the second barrier to migration at the lower endof a tributary stream. A rock ramp fish passage facility was completed at the first barrier near theend of February 2010. The author collected data on total length and sex and estimated thenumber of fish present based on fish that could be seen below the barrier.

Peak spawningoccurred during the last three weeks of April. The minimum estimate of run size was 25,000 fishand was the total of the minimal visual estimates made during each sample event. Males rangedfrom 243- 300 mm with a mean length of 263 mm. Females ranged from 243-313 mm with amean of 273 mm.4.2.4 Volunteer and Other river herring monitoring The Department's Hudson River Fisheries Unit (HRFU), Hudson River Estuary Program and theEnvironmental Defense's South Shore Estuary Reserve Diadromous Fish Workgroup (SSER)have begun to incorporate citizen volunteers into the collection of data on temporal variation ofand physical characteristics associated with spawning of river herring in tributaries.

These datawere not provided by the fishery dependent and independent sample programs discussed above.The volunteer programs also bring public awareness to environmentally important issues.Long Island StreamsThe SSER began a volunteer survey of alewife spawning runs on the south shore of Long Islandin 2006. The survey is designed to identify alewife spawning in support of diadromous fishrestoration projects.

The survey also evaluates current fish passage projects (i.e. Carmans Riverfish ladder),

and sets a baseline of known spawning runs. Data were available for surveys in2006 -2008. Monitoring occurred on six to nine targeted streams annually, with volunteer participation ranging from 24 to 68 individuals.

Monitoring takes place from March throughMay. Alewife were seen as early as March 5 (2006) and as late as May 31 (2008). Data indicated that alewife use multiple streams in low numbers.

It is not clear whether each stream supports aspawning population since total sightings were very low. The Carmans and Swan Rivers showedthe most alewife activity and likely support yearly spawning migrations.

The first permanent fishladder on Long Island was installed in 2008 on the Carmans River. Information gathered duringthis study will aid in future construction of additional fish passage (Kritzer et al. 2007a, 2007band Hughes and O'Reilly 2008).In addition to the SSER, other interested individuals have also monitored Long Island runs (seeAppendix Table A). Anecdotal data provides valuable information on tracking existing in-stream conditions, whether streams hold active or suspected runs, interaction with human land uses andsuggestions for improvement (L. Penney, Town of East Hampton, personal communication).

Arock ramp was constructed around the first barrier to migration on the Peconic River in early19 REVISED VERSION:

September 2011, based on public comment received.

2010 (B. Young, retired, NYS Dept of Environmental Conservation, personal communication).

The Peconic River Fish Restoration Commission set up an automated video counting apparatus at the upriver end of this ramp. Data are still being analyzed.

The Department has conducted a similar river herring volunteer monitoring program annuallysince 2008 for tributaries of the Hudson River Estuary (Dufour et al. 2009, NYSDEC 2010,Hattala et al. 2011). We designed this project to gather presence-absence and temporalinformation about river herring spawning runs from the lower, middle and upper tributaries ofthe Estuary.

Between nine and 11 tributaries were monitored annually by 70 to 213 volunteers in2008, 2009, and 2010. Herring were seen as early as 31 March and as late as 1 June. Riverherring were observed in all but one of the tributaries.

However, several tributaries with knownstrong historical runs had very few sightings.

Water temperature seemed to be the mostimportant factor determining when herring began to run up a given tributary.

Sightings of herringwere most common at water temperature above 50 F. Tributaries in the middle part of the estuarywarmed the fastest each spring and generally had the earliest runs.4.2.5 Young-of-the-Year Abundance Since 1980, the Department has obtained an annual measure of relative abundance of young-of-the-year (YOY) alewife and blueback herring in the Hudson River Estuary.

Although theprogram was designed to sample YOY American shad, it also provides data on the two riverherring species.

Blueback herring appear more commonly than alewife.

In the first four years ofthe program, sampling occurred river-wide (rkm 0-252), bi-weekly from August throughOctober, beginning after the peak in YOY abundance occurred.

The sampling program wasaltered in 1984 to concentrate in the freshwater middle and upper portions of the Estuary (km 88-225), the major nursery area for young herring.

Timing of samples was changed to begin in lateJune or early July and continue biweekly through late October each year. Gear is a 30.5 m by 3.1m beach seine of 6.4 mm stretch mesh. Collections are made during the day at approximately 28standard sites in preferred YOY herring habitat.

Catch per unit effort is expressed as annualgeometric and arithmetic means of number of fish per seine haul for annual weeks 26 through 42(July through October).

This period encompasses the major peak of use in the middle and upperestuary.From 1980 to 1998, the Department's geometric mean YOY annual index for alewife was low,with only one year (1991) over one fish per haul. Since 1998, the index has increased erratically (Figure 16).From 1980 through 1994, the Department's geometric mean YOY annual index for bluebackherring averaged about 24 fish per haul, with only one year (1981) dropping below 10 fish perhaul (Figure 16). After 1994, the mean dropped to around 17 fish per haul, and then began thesame high-low pattern observed for alewife.The underlying reason for the wide inter-annual variation in YOY river herring indices is notclear. The same erratic trend that occurred since 1998 has also occurred in American shad20 REVISED VERSION:

September 2011, based on public comment received.

(Hattala and Kahnle 2007). The increased inter-annual variation in relative abundance indices ofall three Alosines may indicate a change in overall stability in the system.4.2.6 Conclusion Over the last 30 years, the Hudson River stocks of alewife and blueback herring have showninconsistent signs in stock status trends. Calculated CPUE for commercial gill net gears hasincreased in recent years, while CPUE in scap nets fished in tributaries initially

declined, but hasremained relatively stable since 2003. Apparent mortality increased on mature fish and asmortality rose, mean total length and weight declined.

Similar trends occur in the both the fisherydependent and independent data. Recruitment has become extremely variable since the mid-1990s for both species.

Some decline is occurring for YOY blueback herring while, counter-intuitively, there has been an increasing trend for YOY alewife.

Anecdotal evidence from anglersand commercial fishermen suggest a decline in abundance in tributaries yet a dramatic increaseof herring in the main-stem river in the last few years.The upsurge in river herring used as bait for striped bass has placed herring in a tenuous position.

With this continuing demand, declining size, and increasing mortality, careful management isneeded despite variable but stable recruitment.

5 PROPOSED FISHERY CLOSURES5.1 Long Island, Bronx County and Westchester CountyLimited data that have been collected for Long Island river herring populations are not adequateto characterize stock condition or to choose a measure of sustainability.

Moreover, there are nolong-term monitoring programs in place that could be used to monitor future changes in stockcondition.

In 2010, the Peconic River Fish Restoration Commission installed a rock ramp toprovide fish passage at the first dam on the Peconic River system. In the spring of 2011, a fishcounting apparatus was installed upriver of this ramp. In addition, the Commission initiated biological fish sampling of species, sex, length and scales. If these operations continue in thefuture and if these provide information that could be used to set and monitor a sustainability target, we will consider a fishery for this river. Little data have been collected for river herringpopulations in the Bronx and Westchester Counties.

For the above reasons, New York State will close all fisheries for river herring in Long Islandstreams and in the Bronx and Westchester County streams that empty into the East River andLong Island Sound.5.2 Delaware River21 REVISED VERSION:

September 2011, based on public comment received.

We have no data that suggest river herring occur in New York waters of the Delaware River.New York State proposes to close fishing for river herring in New York waters of the DelawareRiver to prevent future harvest should the Delaware stock rebound and expand upriver.

Thisclosure conforms to similar closures planned for the Delaware River and Bay by Pennsylvania, New Jersey, and Delaware.

6 PROPOSED SUSTAINABLE FISHERY6.1 Hudson River and Tributaries Given the mixed picture of stock status provided by available data on Hudson River herring,New York State proposes a restricted fishery in the main-stem Hudson River coupled with apartial closure of the fishery in all tributaries.

We do not feel that the data warrant a completeclosure of all fisheries.

We propose that the restricted fishery would continue for five yearsconcurrent with annual stock monitoring.

We propose a five-year period because the full effectof our proposed restrictions will not become apparent until all age classes in the population havebeen exposed to the change. Most of the fish in the Hudson River herring spawning stocks areestimated to be three through seven years old and these ages predominate in the fishery.Sustainability targets would be set juvenile indices.

We would monitor, but not yet set targets formean length from fishery independent spawning stock sampling and CPUE in the commercial fixed gill net fisheries in the lower river below Bear Mountain Bridge. We will also monitor agestructure, frequency of repeat spawning, and total mortality (Z) if we can resolve uncertainties about aging methods and mortality estimate methodology.

Stock status would be evaluated during and after the five year period and a determination made whether to continue or changerestrictions.

Moreover, we do not know how much of the apparent high mortality is caused bybycatch in ocean fisheries and thus outside current scope of restrictions proposed in this plan.Recreational harvest of river herring is much greater than reported harvest from commercial gears. Data from a creel survey in 2005 estimated approximately 152,000 herring were taken inthe recreational fishery (NAI 2007) while some 31,000 herring were reported from commercial gears (Table 2). For this reason, we feel that restrictions to the recreational fishery will likelyhave a greater impact on take of herring than commercial restrictions.

We should note that Draft Addendum 3 to Amendment 6 of the ASMFC Interstate Management Plan for striped bass stipulates that states should reduce fishing mortality on spawning stocks by50%. If this draft is approved by the ASMFC Striped Bass Management Board, we may have torestrict effort in the recreational striped bass fishery.

Restrictions may include a reduction in useof bait such as river herring.

Any reduction in effort will likely reduce demand for river herringand thus reduce losses in the Hudson stocks.A summary of the following fishery restrictions are contained in Tables 5 and 6. Theserestrictions were based on public comments received from public information meetings held in22 REVISED VERSION:

September 2011, based on public comment received.

the Hudson valley in 2010 in addition to the need to reduce harvest.

Public suggestions forrestrictions are listed in Appendix C.6.1.1 Proposed Restrictions

-Recreational FisheryRecreational fishing seasonCurrently none; proposed season is March 15 to June 15.Recreational Creel LimitCurrently there are no restrictions on daily take of river herring in the Hudson and its tributaries.

To reduce harvest and waste, we propose to implement a restrictive recreational creel limit oftenriver herring per day, or a total maximum boat limit of 50 per day for a group of boat anglers,whichever is less. A Charter boat captain (see Commercial Fishery Restrictions) will beresponsible for a possession limit of 10 river herring per paying customer or a total maximumboat limit of 50 herring per day, whichever is less. Charter boat captains are required, atminimum, to hold a US Coast Guard "six pack" license, i,e. a maximum number of sixpassengers can be on board. However, most vessels fishing the Hudson relatively small (20 to 30ft) with an average of four fares maximum.Most of the river herring harvest is driven by striped bass fishermen catching herring for bait.Anecdotal reports and comments at public meetings suggest that many anglers take many moreherring than they need for a day's fishing.

The proposed creel limit will prevent such overharvest and avoid waste. We obtained an idea of potential harvest reduction from the proposed creelsurvey from data in the Cooperative Angler Program described in Section 2.1.3. Data wereavailable on herring harvest during 502 trips. Since trip level reports often included more thanone angler, we divided the reported herring catch by the number of anglers for an estimate ofcatch per angler trip. These data indicated that 56 percent of the catch per angler trips caught sixor more herring suggesting that a five fish limit could reduce harvest by 56 percent.To track harvest, New York will implement the on line creel survey/ diary program coordinated by ACCSP. It is scheduled to go live by Jan. 1, 2012. New York will increase public outreach toencourage angler use of this program.

We will also continue the Cooperative Angler Program forcomparison and for individuals not savvy with on-line tools.Prohibit Harvest by Nets in Tributaries Recreational anglers generally use hook and line jigging) in the main-stem river and are allowedto use personal use gears (without a license) of scap/lift nets (36 sq ft or less), small dip nets, andcast nets. They are not required to report this catch and the number of herring taken by thesegears is unknown.

Anecdotal reports and observations suggest tributaries are popular locations for recreational harvest by these net gears, especially in the middle section of the estuary (Figure1).23 REVISED VERSION:

September 2011, based on public comment received.

Information from the volunteer angler program along with anecdotal data on recreational harvestsuggests that abundance of river herring, mostly alewife, has declined in some spawningtributaries.

This may be due to the increased vulnerability to harvest as herring often concentrate in these tributaries in large schools to spawn. Tributaries with an impassable barrier close to themouth confine fish to even smaller areas. For these reasons, we feel it prudent to closerecreational harvest by nets from tributaries until measures of stock condition improve.

We didnot feel that it was feasible or desirable to enforce a closure on angling for river herring intributaries.

In the main-stem Hudson, personal use nets will be allowed to continue but with a reduced sizefor scap/ lift nets (16 sq ft instead of 36 sq ft); seine, cast, and dip nets sizes will remain the same(Table 5).Closed areasAlthough personal-use net fishing by recreational anglers will not be allowed in tributaries, angling will continue.

However, to further relieve fishing pressure in areas of fish concentration, in addition to the net ban, no fishing will be allowed within the River Herring Conservation Area(RHCA) defined as stream length within 250 m (825 feet) of any type of barrier, natural or man-made. This is similar to a fishing ban within 50 rods of fishways instituted in New York in 1895.Many of the Hudson's tributaries have natural (rapids) or man-made barriers a short distance infrom the main river. River herring concentrate in great numbers below these barriers makingthem very vulnerable to any fishing.

This closed area will allow them to spawn in thisundisturbed stretch.

The RHCA closure will effectively end all fishing in the eight smallesttributaries, or 14% of the tributaries in the estuary.Above the Troy Dam, an area closure is already in effect for the "Waterford Flight",

Lock 2 toGuard Gate 2, a series of dams and locks at the entrance to the Mohawk River. Within theMohawk, a RHCA will be in effect below any of the remaining locks and dams up to Lock 21 inRome.Escapement periodNone are proposed.

Licensing and reporting In 2011, New York State implemented a recreational marine fishing registration.

All anglersfishing for anadromous fish must register prior to fishing for migratory fish of the sea. For theHudson this includes river herring and striped bass. The recreational and commercial fisheries for American shad were closed in the Hudson River in 2010.24 REVISED VERSION:

September 2011, based on public comment received.

By Jan 1 2012 New York, in cooperation with ACCSP, will start up an online angler survey. TheDepartment will increase public outreach to strongly encourage fishers to use this new tool to aidin understanding recreational catch and harvest.6.1.2 Proposed Restrictions

-Commercial FisheryLicense Required:

Currently, fishers using commercial, non-personal use size gears to take and /or sell fish must bein possession of a Marine Permit for that gear. Marine permits have an annual reporting requirement, but no requirements for proof that harvest was for commercial purposes.

Recreational fishermen commonly purchase marine permits and use commercial gears becauseof the low cost. We propose to strengthen the commercial aspects of these gears by requiring proof that harvest was sold as a requirement for license renewal.The overlap with gears licensed under the NY bait license will be minimized by requiring aMarine Permit to take river herring.

Cast nets will be included under the Marine Permit licensing system.Closed areaWe propose to continue the current closures as listed in Table 6 and implement a new closure:Prohibit Harvest by Nets in Tributaries:

Closing the tributaries to harvest by nets willlikely reduce overall harvest, but the actual size of this reduction is not known. We do not knowthe size of recreational net harvest from tributaries.

We can infer current commercial harvestfrom tributaries by the number of fish taken in scap nets since most river herring taken intributaries are taken by this gear and most scap nets are fished in tributaries.

Mean annualreported harvest by commercial scap nets in the last five years was about 15,000 river herring or48% of the total reported commercial harvest.

The mean number of commercial fishing tripsusing scap nets during this time period was 611 trips which were about 59% of all reported tripsin the estuary and tributaries.

Elimination of commercial net harvest from these waters willeliminate commercial fishing in 175 miles, or approximately 65% of linear spawning streams inthe Estuary and above the Troy Dam.Gear Restrictions All current gear restrictions will remain in place (Table 6). Other changes include:Gill nets: Currently both anchor and drift gill nets are used in the mid and upper estuaryabove the Bear Mountain Bridge (> rkm75). Both gears catch herring, but losses can be higher inanchored nets because they are often not tended as frequently as drifted nets. This is especially the case with recreational fishermen who are often not experienced in use of gill nets. Wepropose to ban use of fixed gill nets in the Hudson River above Bear Mountain Bridge; drift gill25 REVISED VERSION:

September 2011, based on public comment received.

nets are required to be tended by owners as they are fished. We don't know what reduction inharvest would result, but some will occur and the change will certainly reduce waste of fish.Scap /Lift nets: Currently there are no limits on size of scap nets to be used. Mandatory reports indicate that the largest nets in use are 400 sq ft (20 by 20 ft). The proposed maximumnet size is 10 ft by 10 ft.Fyke and Trap nets: Although currently legal for the take of river herring, no commercial harvest is reported from these gears. We propose that their use not be allowed for harvest of riverherring.Commercial Net Permit and FeesCommercial gears in the main-stem Hudson and tributaries are licensed under a NYSDECBureau of Marine Resources Marine Permit. Access to obtain a Marine Permit remains open,with no prior requirements.

These commercial gears are often used by recreational fishermen because current permit fees are very low. Most fees were set in 1911 by the then New YorkForest, Fish and Game Commission and no fee increases have occurred through the present time.Commercial gears such as gill nets can take high numbers of herring and are not considered to berecreational gear in New York. For the purposes of harvest in ocean waters (Marine and CoastalDistrict),

gill nets are considered commercial gear and their use for recreational purposes is notpermitted.

We propose regulations to increase fees to account for inflation, to emphasize that nets arecommercial gears, and to discourage casual use by recreational anglers.

Current fee structure canbe found in New York Code of Rules and Regulation-Part 35 (seehittp:iwxN.vwv.dec.ny.gov/regs/4019.html

). We considered two alternatives.

1. Increased gear and fishing vessel fees.a. In 1911, fees were $5.00 per each trap, seine or gill net, and $1.00 per scap net.These fees would translate to $115.00 per gill net or seine and $25.00 per scap netin today's (2011) dollars.b. Gill nets and seines can also be licensed by the linear foot of net rather than as atype of net. We propose that the current $ 0.05 per foot be increased to $1.00 perfoot. Data from the mandatory reports indicates that the most recent (2010)licensed gill net lengths ranged from 10 ft ($10 fee) to 600 ft ($600 fee). Seineshave no maximum length restriction in place; current use is 50 ft ($50 fee) to 100ft ($100 fee).c. Another way to differentiate between recreational and commercial fishermen is toreinstitute the 1911 fishing vessel registration for the Hudson River, which is stillactive for other waters of NY. The 1911 fee of $15.00 for the smallest motorized vessel translates to $350.00 per vessel in today's dollars.2. A single commercial gear permit.26 REVISED VERSION:

September 2011, based on public comment received.

This approach simplifies the above combination of gear fees and is our preferred alternative.

We would create a Hudson River Commercial Fish Gear Permit (HRCFGP):

forindividuals who want to harvest river herring or Atlantic menhaden; fee of $150. Thiswould be instead of individual gear licenses.

a. Qualifications needed: proof of previous sale to a licensed retail bait shop; if abusiness (retail bait shop), proof of business incorporation (LLC)b. If applicant holds a valid New York food fish or crab permit(s);

cost of HRCFGPto be offset by valid permit fee(s)c. To include all restrictions as listed in Table 6.d. Gears to be used include anchored (fixed) and drifted gill nets, scap/lift nets,seines and cast nets (see Table 6 for size limitations)

Gear restrictions outlined above will still apply to any alternative chosen.Closed Fishing DaysA 36-hour escapement period per week, from 6 AM prevailing time on Friday to 6 PMprevailing time on Saturday, is in effect for commercial gill nets from March 15 to June 15. Wepropose to expand this closure to include all commercial nets.Reporting Current mandatory reports of daily catch and effort data are submitted annually.

We willcontinue to require these reports, but decrease the time of report submission to monthly.Charter Boat LicenseIn order to distinguish Charter Boat operators from recreational

anglers, we propose to use theexisting Marine & Coastal District Party & Charter Boat License (CPBL), as it exists for NY'sMarine District.

CPBL holders will follow all regulation as established for the Marine Districtwith two exceptions:

creel and size limit for striped bass will comply with limits set for theHudson River above the G. Washington Bridge and the creel limit for a charter boat will be 20river herring per day. Hudson valley charters can take up to three to six individuals per trip.7 PROPOSED MEASURES OF SUSTAINABILITY 7.1 TargetsJuvenile IndicesWe propose to set a sustainability target for juvenile indices using data from the time period of27 REVISED VERSION:

September 2011, based on public comment received.

1983 through 2010 for both species.

We will use a more conservative definition of juvenilerecruitment failure than described in section 3.1.1.2 of Amendment 2 to the ASMFC Interstate Fisheries Management Plan for Shad and River herring (ASMFC 2009). Amendment 2'sdefinition is that recruitment failure occurs when three consecutive juvenile index values arelower than 90 % of all the values obtained in the base period. We will use a 75% cut off level.The 75% level for alewife is 0.35 (instead of 0.19) and 11.14 (instead of 2.86) for bluebackherring (Figure 16).The fishery will close system-wide if recruitment

failure, defined as three consecutive yearsbelow the recruitment failure limit, occurs in either species and will remain closed until we seethree consecutive years of recruitment greater than the target values.7.2 Sustainability MeasuresThere are several measures of stock condition of Hudson River herring that can be used tomonitor relative change among years. However, these measures have limitations (described below) that currently preclude their use as targets.

These include mean length in fisheryindependent

samples, catch per unit effort (CPUE) in the reported commercial harvest and agestructure.

We propose to monitor these measures during the fishery and use them in concert withthe sustainability target to evaluate consequences of a continued fishery.Mean LengthMean total length reflects age structure of the populations and thus some combination ofrecruitment and level of total mortality.

Mean total lengths of both river herring species in theHudson River system has declined over the last 20 years and the means are now the lowest of thetime series. Since this has been a persistent change in the face of stable recruitment, we suggestthat the reduction in length has been caused by excessive mortality of adults within the river andduring their ocean residency (bycatch).

The bycatch fishery is a large unknown and not solelycontrolled by New York State to effect a change. Current annual reproduction now relies on afew returning year classes making the populations vulnerable to impacts of poor environmental conditions during the spawning and nursery seasons.

We propose to monitor mean total lengthsduring the proposed fishery.Catch per Unit Effort in Report Commercial We suggest that CPUE values of the reported harvest reflect general trends in abundance.

However, annual values can be influenced by changes in reporting rate and thus we do not feelthat CPUE should be used as a target. Rather, we will follow changes within gear types andfisheries for general trends.Age structure and Total mortality We will monitor age structure, frequency of repeat spawning, and total mortality (Z) if we can28 REVISED VERSION:

September 2011, based on public comment received.

resolve uncertainties about aging methods and estimate methodology discussed in Status Section4.2.

2.8 REFERENCES

ASMFC. (Atlantic States Marine Fisheries Commission).

2009. Amendment 2 to the Interstate fisherymanagement plan for shad and river herring.

Washington, D.C. USA.Caraco, N.F., J.J. Cole, P.A. Raymond, D. L Strayer, M.L. Pace, S.E.G. Findlay and D.T. Fischer.

1997,Zebra mussel invasion in a large turbid river: phytoplankton response to increased grazing.Ecology 78:588-602.

Cating, J. P. 1954. Determining age of Atlantic shad for their scales. U.S. Fish and Wildlife ServiceBulletin 54: 187-199.Cieri, M., G. Nelson, and M. Armstrong.

2008. Estimate of river herring bycatch in the directed Atlanticherring fishery.

Report prepared for the Atlantic States Marine Fisheries Commission.

Dufour, M. R. Adams, K. Hattala, and L. Abuza. 2009. 2008 volunteer river herring monitoring program.NYS Department of Environmental Conservation, New Paltz, NY.Gedamke, T, and J. M. Hoenig. 2006. Estimating mortality from mean length data in nonequilibrium Situations, with Application to the Assessment of Goosefish.

Transactions of the AmericanFisheries Society 135:476-487.

Greeley J.R. 1937. Fishes of the area with annotated list IN A biological survey of the lower Hudsonwatershed.

Suppplement to the twenty-sixth annual report, 1936, State of New York Conservation Department.

J.B. Lyons Company Albany NY, USA.Hattala, K., M. Dufour, R. Adams, K. McShane, J. Kinderd, and R. Lowenthal.

2011. Volunteer riverherring monitoring program 2010 report. NT State Department of Environmental Conservation, New Paltz, NY.Hattala, K. and A. Kahnle. 2007. Status of the Hudson River, New York, American shad stock. INASMFC Stock assessment Report No. 07-01 (supplement) of the Atlantic States Marine Fisheries Commission.

American shad stock assessment report for peer review, Volume II. Washington, D.C.,USA.

Hughes, A, and C, O'Reilly.

Monitoring Alewife Runs in the South Shore Estuary Reserve Report on the2008 Volunteer Survey. July 2008. <http://www.estuary.cog.ny.us/council-priorities/living-resources/alewife_survey/Alewife%202008.pdf>.

[accessed September 2008].Kahnle, A., D. Stang, K. Hattala, and W. Mason. 1988. Haul seine study of American shad and stripedbass spawning stocks in the Hudson River Estuary.

Summary report for 1982-1986.

New YorkState Dept. of Environmental Conservation, New Paltz, NY, USA.29 REVISED VERSION:

September 2011, based on public comment received.

Kritzer, J, A. Hughes and C, O'Reilly.

2007a. Monitoring Alewife Runs in the South Shore EstuaryReserve Report on the 2006 Volunteer Survey 06. <http://www.estuary.cog.ny.us/council-priorities/living-resources/alewife-survey/2006ý/%20Aewife%20Survey%20Report.pdf>.

[accessed September 2008].Kritzer, J, A. Hughes and C, O'Reilly.

2007b. Monitoring Alewife Runs in the South Shore EstuaryReserve Report on the 2007 Volunteer Survey. <http://www.estuary.cog.ny.us/council-priorities/living-resources/alewife-survey/2007%20Alewife%2OSurvey%20Final.pdf>.

[accessed September 2008].Lynch, R. and B. Kim. 2010. Sample size, the margin of error and the coefficient of variation.

http:i/interstat.statiournials.net/YEARI2OI 0/articles/1001 004.pdfMiller, D., and J. Ladd. 2004. Channel morphology in the Hudson River Estuary:

past changes andopportunity for restoration.

In Currents-newsletter of the Hudson River Environmental Society,Vol. XXXIV, No. 1. Available:

http://www2.marist.edu/-en04/CUR34-1.pdf.

Normandeau Assoociates Inc. 2003. Assessment of Hudson River Recreational Fisheries.

Final Reportprepared for the New York State Dept. of Environmental Conservation, Albany NY, USA.Normandeau Assoociates Inc. 2007. Assessment of Spring 2005 Hudson River Recreational Fisheries.

Final Report prepared for the New York State Dept. of Environmental Conservation, Albany NY,USA.Normandeau Associates, Inc. 2008. Spawning stock characteristics of alewife (Alosapseudoharengus) and blueback herring (Alosa aestivalis) in the Hudson River Estuary and tributaries, including theMohawk River. Final Report prepared for the New York State Dept. of Environmental Conservation, Albany NY, USA.Nelson, G. A, P. Brady, J. Sheppard and M.P. Armstrong.

2010. An assessment of river herring stocks inMassachusetts.

Massachusetts Division of Marine Fisheries Technical Report.NEFSC (Northeast Fisheries Science Center).

2011. Part I -preliminary analyses for Amendment 14 tothe Atlantic

Mackerel, Squid, and Butterfish Fishery Management Plan. Prepared for the 10 May2011 FMAT meeting, Woods Hole, MA.NYCDEP http://home2.nyc.eov/hltnl/dep/html/news/hwqs.shtmii NYSDEC (New York State Department of Environmental Conservation).

2010. Volunteer river herringmonitoring program 2009. NY State Department of Environmental Conservation, New Paltz, NY.Schmidt, R. and S. Cooper. 1996. A catalog of barriers to upstream movement of migratory fishes inHudson River tributaries.

Final report to the Hudson River Foundation from Hudsonia, Annandale NY, USA..Smith, C.L. 1985. Inland fishes of New York State. New York Department of Environmental Conservation, Albany, NY, USA.Strayer, D.L., K.A. Hattala, and A.W. Kahnle. 2004. Effects of an invasive bivalve (Dreissena 30 REVISED VERSION:

September 2011, based on public comment received.

polymorpha) on fish in the Hudson River estuary.

Can. J. Fish. Aquat. Sci. 61:924-941.

Young, B. 2011. Report on Peconic River alewife run -2010. Peconic River Fish Restoration Commission, Ridge, NY, youngb53@optimum.net.

31 REVISED VERSION:

September 2011, based on public comment received.

Upper Hudson (non-fdo4 MohawkAver/Erie Cana.. ........ ..... ........ ........ ........... .... ........... .. 4 _ y 1-----------

-_FedffmlDw anT ýPesten R-ilAlany (w 232)Upperj, Es-vs ayCuadeton 90 Bddge

........

..........

... ..nWin....

............

ig (Catskill reekRoeliffAnseanKill ESopUS CreekKbtgston cn 146)Mid Evrnwy,3 Rouadozt CreeBlack CreekYougJufepsie (lwn 122)Wappingers CaeekQcssaick Creek ' kill &eekNewburgh Bay (km 95)Moodna CreekWest Point (kmt 83)-.arMo 41tag n EjdAnnswtlle CreekHaverstraw Bay (kmi 55) Crcton RwerTappanZee (kn 45)Tappan Zee BidgeG. Washington BAdgeBattery (In 0 New York CityVerrazano Narrows.Lawr++.

EsheatyFigure 1 Hudson River Estuary with major spawning tributaries for river herring.

(see Appendix Table A for complete list)32 REVISED VERSION:

September 2011, based on public comment received.

33 REVISED VERSION:

September 2011, based on public comment received.

ýCTHudson RiverNY6PLon g Island SoundMontauk,tonNew YorkAtlantic OceanFigure 2 Long Island, Bronx and Westchester

Counties, New York, with some river herring (primarily alewife) spawning streams identified (See Appendix Table A for list)34 REVISED VERSION:

September 2011, based on public comment received.

Commercial landings of River herring in NY25020015010050-4~ ~ M4 -4 M4 4 -M -4r M4 -4 r 4 .4 .4 .. 4 4Figure 3 Commercial landings of river herring from all waters of New York State.rOn4540353025201510501995 1998 2001 2004 2007 2010Figure 4 Commercial landings of river herring in the Hudson River and NY Oceanwaters.CL0.600.500.400.300.200.100.00.. .................

..........

... ..................

I ...........

..... ...... .. ------------- -------", Fixed bBMB.. -Drift-M-- Fixed aBMB* Scap2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010Figure 5 Percent commercial catch by gear of river herring in the Hudson River (a/b BMB=above and below BearMountain Bridge).35 REVISED VERSION:

September 2011, based on public comment received.

IIIL.001.801.601.40I.OO0.200.00Lower River Catch-Per-Unit-Effort 3.00 Mid & Upper River Catch-per-Unit-Effort

weighted mean2.500400.002000 2001 2002 2003 2004 2005 2006 2007 2008 2000 20102000 2001 2202 2003 2004 2005 2006 2007 200R 20 2010Figure 6 Catch per Unit Effort (number of fish per hours fished) by area of the river and gear. Lower estuary = below Bear Mountain Bridge[rkm 75]; Mid & Upper estuary = above the Bear Mountain Bridge.EE00CaJ3002902802702602SO240230220210200.. .............

00 ...............

ý,,s Alewife-M

-V-Alewlfe-F

---

Blueback herring-M

-- Blueback herring-F

....... ...........

............

! ...............

--T .... ..... .... ... "2001 2002 2003 2004 2005 2006 2007 2008 2009 2010Figure 7 Mean total length of river herring collected from commercial fishery monitoring trips in the Hudson River Estuary36 REVISED VERSION:

September 2011, based on public comment received.

340.0Alewife +Male224-.4-0I-.320.0300.0280.0260.0240.0zzU.u'AFermale 200.U1936 1976 1979 198Z 1985 1988 1991 1994 1997 2000 2003 2006 2009340.00320.00300.00280.00260.00: 240.00220.00200.00Blueback herring4-* MaleFemaleS x.1936 1976 1979 1982 1985 1988 1991 1994 1997 2000 2003 2006 2009Figure 8 Mean total length of river herring in the Hudson River Estuary.

Symbols with an "X"indicate adequate sample size (N>34) to characterize the stock.37 REVISED VERSION:

September 2011, based on public comment received.

Blutback herring male200ISO "100z5050 I I-Blueback herring female4-0-HRmale89 IMamale89 i-s-Mdmale89 200150E 100z5002 3 4 5 6 7 8 9Age2 3 4 5 6 7 8 9Age200 Blueback herring male150 --*-HRmale90

-~MAmale90 100 MdmaIe9O5002 3 4 5 6 7 8 9AgeBlueback herring female150100EDz 500-*-HKRfemale90

--o<Maferna1e90

--",MDfemale9O 2 3 4 5 6 7 8 9AgeFigure 7 Hudson (HR) age structure and estimated age structure of Hudson River blueback herring based on length-at-age keysfrom Massachusetts (MA) and Maryland (MD) blueback herring.38I REVISED VERSION:

September 2011, based on public comment received.

Blueback herring 1992 1260.050.40.0:10 020.010.0O,0Z ki -*-M-Ma02 3 4 S 6 7 8 9 10AgeBlueback herring 2009Blueback herring 2010z80.070.0 160.050.040.030.020.010.0 -f0.0 M-Ma-M -M-Md-F-Ma--# F-Md35.030.025.0IU.Uz= 15.0z10.05.00.0--,-M-Ma2 3 4 5 6 7 8 9 10Age2 3 4 5 6 7 8 9 10AgeFigure10 Estimated age structure of Hudson River blueback herring based on length-at-age keys fromMassachusetts (MA) and Maryland (MD).Blueback h erring mean ageUbe6.005.004.003.002.001.000.00I.........

i ..........

-TIn-..... M da-1-M.0 Md-rn.Ma-fN- Md-f-- w NY-ni7, NO-2.00 -EMdF1.00 U NY-rn1989 1990 1991 19922009 2010Figure 11 Mean age of Hudson River blueback herring based on length-at-age keys from Massachusetts (MA) and Maryland (MD).39I REVISED VERSION:

September 2011, based on public comment received.

Alewife Male 2001Alewife Female 20012502O0150z100so-4-MA90807060z 40302010#MA--*-ME0 i-s 5 I A ----2 3 4 5 6 7 a 9Alewife Male 20030 4- 40-2 3 4 5 6 7 a 9Alewife Female 2003200 0-MA-MD1505o--1 ME100908070602 403020100-4-MA0 4 -- .------2 3 4 5 6 7 8 9Alewife Male 20092 3 4 5 6 7 8 9300250200E 150z21O050 MAMDAgeAlewife Female2009 7060 -,-MA50MD40100 -2 3 4 5 6 7 S 9Ag.Alewife Female 20102 3 4 5 6 7 8 9AgeAlewife Male 2010150 - MA100 MD4,Ez502 3 4 5 6 7 8 9Age605040E 30z20100-=,- MAMD*~ME2 3 4 5 6 7 8 9IAgeFigure 12. Estimated age structure of Hudson River alewife based on length-at-age keys from Maine(ME), Massachusetts (MA) and Maryland (MD).40 REVISED VERSION:

September 2011, based on public comment received.

4.80 Alewife -male4.60EMA A MD U ME4.404.204.003.803.603.403.203.001990 2001 2002 2003 2004 2005 2006 2007 2008 2009 20105.00 Alewife -female4.80N MA EMD <ME4.604.40w 4.204.003.803.603.403.203.001990 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010Figure 13. Mean age of Hudson River alewife, ages estimated from age-length keys from Maine (ME),Massachusetts (MA) and Maryland (MD).41 REVISED VERSION:

September 2011, based on public comment received.

Male alewife8.0 -7.06.065.00E 4.03.0.02.00J-'4-Lc=250

..... -Lc=240"- -Lc=230AI-.Uat %0.0 1 1 1 ý I -I I1980 1982 1984 1986 1988 1990 1992 1994 1996 19982000 2002 2004 2006 2008 2010Female alewife4.0-"* Lc= 2503.5-W- Lc=240EU 3.0 -r-Lc=230

" 2.50E 2.01.01890.51980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010Figure 14. Length-based mortality estimates for Hudson River alewife.

Lc =minimum length of fishcaught in the sample gear.42 REVISED VERSION:

September 2011, based on public comment received.

16.0 Male Blueback herring14.0 Lc=250=

EM~* 10.0V08, 6.06.0S4.02.0 N-0.091981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009Female Blueback herring14.0 1E0E*1012.010.0-4"* Lc=250Lc=240-,r- Lc=2308.06.04.02.00.01981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009Figure 15 Length-based mortality estimates for Hudson River blueback herring.

Lc =minimum length offish caught in the sample gear.43 REVISED VERSION:

September 2011, based on public comment received.

7.00 Alewife6.006- -GeometricMean w CI-5 25th percentile 1983-2010 5.00--10th percentile 1983-2010 4.00IMI3.0002.00 #i !1.00 1 1 10.00 1 98 19 91980 1983 1986 1989 1992 1995 1998 2001 2004 2007 201070.0060.0050.003 40.00E30.0020.0010.000.00Blueback herring.t A- p GeometrilcMean w C125th percentile 1983-2010 S10th percentile 1983-2010 I ------------

1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 2010Figure 16. Annual young-of-the-year indices (with 95% CI) for alewife and blueback herring collected inthe Hudson River Estuary.44 REVISED VERSION:

September 2011, based on public comment received.

Table 1. Summary of available fishery-dependent river herring data in Hudson River and Marine District of New York.Data Type Time eriod/ Details Description Usefulness as indexFishey .Dependent.-

Comtercial Harvest Historic data: Provide catch and effort data Gives historic perspective

-1904-1994:

NMFS Not separated by area ( river v marine) Provides trend data for state as a whole, but does not-1994-present:

Hudson (see below)- River data reporting rate unknown separate river(s) from ocean until 1994.NYSDEC; Marine waters- VTR/dealer report since 2002-1994- present:

transfer of historic NMFSdata to ACCSP, data available inconfidential and non-confidential formMarine monitoring River herring most likely occur as bycatch No port sampling in NY for 'herring' in variety of fisheries Hudson River Began in 1995 through the present Data from 2000 to present good Emigration area CPUEMandatory reports Enforcement of reports in 2000 Reporting rate unknown -Fixed GN below BMB:Catch and effort statistics Data separated by gear used: o Good indicator of abundance Licenses are open access with low fees, Fixed gill net below Bear Mountain Bridge (BMB); o increasing trendmany recreational fishers purchase and use assive gear below spawning area; consistent manner of Spawning area CPUEommercial gears to obtain bait ishing; weekly sum of CPUE approximating "area under o Drift GN -variableurve" method o Scap -FlatIn spawning area above BMB o Fixed GN- increasing Drift gill (main-stem HR only) -active gearFixed gill (main-stem HR only) -less effort than belowMB-Scap/lift net (main-stem HR and tributaries)

Hudson R. Fishery Began in 1999 through the present -Number of annual trips are low; co-occurs

& conflicts Characterize catchMonitoring Onboard monitoring with FI samplingCatch and effort statistics Catch samples lowI" Catch subsample NEED improved sample size to be usefulFishery Dependent

-Recreational Harvest (primarily Creel surveys:

2001: provides point estimate of effort for striped bass, Combination of effort for striped bass and pointsought as bait for 2001, river-wide, all year ancillary river herring (RH) data estimate of RH harvest; combine with below CAPstriped bass; some 2005, spring only -2005 provides point estimate of RH harvest & effort for data to estimate magnitude of recreational harvest forharvest for personal 2007, state-wide angler survey; effort for striped bass 2005 to the present.consumption) striped bassCooperative Angler Data 2006-present Diary program for striped bass anglers; includes data for Good RH use per trip- used above with rec. harvestProgram RH catch or purchase, use by trip to estimate total recreational harvest45 REVISED VERSION:

September 2011, based on public comment received.

Table 2. Summary of available fishery-independent river herring data in Hudson River, New York.Data type Time peiod/Agency Description Usefulness as indexIndependent-H dson RiverSpawning stock 1936: Biological Survey Historic data, low sample size of 25 fish, species, Indication of size change to presentsex, length & age1975-1985:

NYSDEC contaminant Sample size low and extremely variable by year Indication of size change to presentsampling1989-1990 NYSDEC Hudson-Mohawk Focused study, large sample size (1,100 fish): Primarily blueback herringRiver. species, sex, length & age1999-2001 Normandeau Assoc. Inc. (NAI) Contract to assess gears for spawning stock survey Primary gear used was size selective gill nets;Developed own age key; not clear how compares to precludes use for length analyses; needmethod of other Atlantic coast states adjustment for ages2001 to present:

NYSDEC spawning stock Focused spawning stock survey; >300 fish Sample design precludes use for catch-per-unit-survey collected most years; species, sex, length & scales effort data(ageing not complete)

Overview of all above Problems Ok to useSpotty adequate sample size in most years (>34 Good sample size for data 1989-99, 2001,-03,-05, per species, sex) to provide trend for length and -08 to presentweightAgeing technique varies greatly from 1936, 1980s, Used ME, MA & MD age-length keys to estimateNAI; techniques appear different from other Judson ages;Atlantic coast states Results:

a slight non-consistent bias of age-Mortality estimates from age structure (above) lifference, possibly attributed to ageing technique unusable as index k/or growth differences (MD fish grow faster thanAA)Suggest use trend in mean ageMortality estimates from age structure (above)rusable as indexBeverton-Holt length based too dependent onInputs (length at recruitment and age)Volunteer River herring surveys 2006 to present; documents presence/absence of -Not yet useful as index; provide a mechanism toriver herring in Hudson tributaries and in some Long mprove future sampling for adult runsIsland streamsYoung-of-year Indices 1980 to present:

annual yoy sampling July-Oct sampling within nursery area Both species index variablestandardized since 1984; Geometric mean number per haul Alewife increasing Catchability may be affected by habitat change Blueback slight decreasing trend46 REVISED VERSION:

September 2011, based on public comment received.

I ý Selected conservative target of 25th percentile 47I REVISED VERSION:

September 2011, based on public comment received.

Table 3. Commercial river herring fishery monitoring data for the Hudson River Estuary.On-board Observations on Comnrwrcial TripsNofYear tripsAlewifeNumber SexratioM F U M FBlueback herringNumber Sex ratio Number Sexratio PercentM F U M F M F U M F Total Alewife BluebackUnidentified "river herring"7 7.f1996 11997 51998 11999 42000 62001 72002 82003 22004 112005 12006 32007 62008 12009 32010 143Fr F5 25 178 0.17 0.83114737 719 18 0.51 0.49p. 7192 178 851 0.52 0.48431717 7124 168 8 0.42 0.58428114187 179 4 0.51 0.498 480 42 2 0.66 0.343487 73 32 480 0.09 0.9119 41 1225 0.32 0.685 6 0.45 0.5543 0%/07 7208 100%F 7r114 100%/07 *421 17%552 7%/o7 71221 1000/07 71328 3%171 100%0.39 0.61 1904 16%456 94%* 7247 00/0335 4%0.50 0.50 44 0%/o468 79%2 5233 53%100%/00°/000/000/093%0°/097%00/01%0(/0100%/016%0°/%21%47%500 796 297282465326844373361700.38 0.626 0.32 0.6848 REVISED VERSION:

September 2011, based on public comment received.

Table 4. Estimated recreational use and take of river herring by Hudson River anglers.Herring Use*% of all CAP Trips N-SB N Total Trips using Estimated using herring as Trips N bought caught / RH Estimated herring as HerringYear bait using RH / trip trip use/trip SB trips** bait** Use2001 53,988 39,500 93,157**2005 89% 2.36 72,568 64,500 152,117**

Cooperative Angler Program Data2006 93% 263 1.47 2.57 4.042007 70% 331 1.66 1.80 3.46 90,742 69,700 241,318***

2008 71% 445 0.86 1.64 2.502009 77% 492 0.63 3.80 4.432010 74% 527 0.67 4.80 5.48*Data from NYSDEC -HRFU Cooperative Angler Program (unpublished data)**Creel survey data: NAI 2003, NAI 2007; 2001 estimated use modified using 2005 RH use per trip* 2001 trips using herring as bait***Estimate calculated from overall average RH/trip (CAP) and Estimated SB trips from NYSDEC statewide angler survey49 REVISED VERSION:

September 2011, based on public comment received.

Table 5. Current and proposed recreational fishery regulations for a river herring fishery in the HudsonRiver.Regulation Current 2010 Recreational Proposed change- newSeason All year March 15 to June 15Creel/ catch limits None 10 per day per angler or a maximum boat(any size, any number) limit of 50 per day for a group of boatanglers (whichever is lower)Charter boats: (see commercial fishingtable)Closed areas None below Troy Dam -the River Herring conservation Area: No-Closure from Guard gate 2 to fishing within 825fi (250m) of a man-madeLock 2 on the Mohawk River or natural barrier-Closure from Guard gate 2 to Lock 2 onMohawk RiverGear restrictions

-Angling All tributaries, including the Mohawk River-Scap/lift net: 36 sq ft or above Troy: Angling only, no netssmaller Main river below Troy Dam: Angling or theDip net: 14" round or 13"x 13" use of nets to obtain bait for personal usesquare only as follows:Seine: 36 sq ft or smaller Scap/lift net 16 sqfi or less-Cast net; 1Oft diameter Dip net: 14" round or 13"x 13" squareSeine 36 sq ft or smallerCast net 10 ft diameterEscapement (no fishing days) None NoneLicense Marine Registry Marine RegistryReporting None New York angler diary on ACCSP website50 REVISED VERSION:

September 2011, based on public comment received.

Table 6. Current and proposed commercial fishery regulations for a river herring fishery in the HudsonRiver.Regulation Current 2010 Commercial Proposed change -newSeason Mar 15 -Jun 15 Mar 15 -Jun 15Creel/ catch limits None Charter boats: IO fish per day per payingcustomer or a maximum boat limit of 50fish per day, (whichever is lower) *Closed areas No gill nets above 190-Castleton Bridge No gill nets above 190 -Castleton BridgeNo nets on Kingston Flats -No nets on Kingston Flats-No nets in tributaries Gear restrictions Allowed gears Allowed gears for river herring-Gill net -Gill net0 600 ft or less 0 600 ft or less0 3.5 in stretch mesh or smaller o 3.5 in stretch mesh or smallerD No fishing at night in HR 0 No fishing at night in HR aboveabove Bear Mt Bridge Bear Mt BridgeSeine >36 sq ft 0 No fixed gill nets above the BearNo seine >100 ft allowed above 190 Mt Bridgebridge -Seine; no seine >100 ft allowed aboveScap/lift net no size 190-Castleton BridgeFyke or trap net Scap/lift net lOft by lOft maximumCast net not exceeding ten ft diameter Cast net not exceeding ten ft diameterEscapement (no 36 hr lift (applies only to gill nets 36 hr liftfishing days) allowed in the main river) Applicable to all net gearsMarine Permit Marine Permit Marine permit only license to take-Fees implemented in 1911 anadromous river herring, the only netGill net $0.05/foot gears allowed include drift and fixed gillScap net <10 sq ft $1.00 net, scap/lift net,seine and cast netScap net> 1Osq ft $2.00 Fees updated to include any of theSeine $0.05/foot following:

Trap nets $3 to $10 Ia. Gill or seine net -$115; scap net $25Fyke net $1 to $2 1 b.Gill or seine $1 per footBait license I c.Fishing vessel $350-Cast net $102. Create Hudson River commercial fishpermit; includes use of gillnets, scap/lift nets, seines and cast nets with all otherrestrictions as listed in this table;qualifications needed (see Sec 6.1.2, page26)Charter*

Boat None for Hudson above the Tappan Require existing Maine &Coastal DistrictLicense Zee Bridge Party boat/ Charter license for tidalHudson and its tributaries-

$250.00Reporting Mandatory daily catch& effort; one Mandatory daily catch& effort; reportsI annual report due monthly51 REVISED VERSION:

September 2011, based on public comment received.

Appendix A. River herring streams of New York including tributaries of the Hudson River Estuary, andthe Mohawk River; streams in the Bronx and Westchester Counties and on Long Island. (This list maynot be complete).

Hudson RiverRiver Mile County Prirary Tributary 18 Westchester Saw Mill24 Rockland Sparkill Creek25 Westchester Wicker's Creek28 Westchester Pocantico River33 Westchester Sing-Sing 34 Westchester Croton River38 Westchester Furnace Brook38 Rockland Minisceongo 39 Rockland Cedar Pond Brook43 Westchester Dickey Brook44 Westchester Annsville Creek44 Westchester Annsville Creek44 Westchester Annsville Creek46 Orange Popolopen Creek52 Putnam Phillipse Brook52 Putnam Indian Brook53 Putnam Foundry Brook55 Putnam Breakneck Brook57 Orange Moodna Creek58 Dutchess Malzingah Brook (Gordon's Brook)59 Dutchess Fishkill Creek67 Dutchess Hunters Brook67 Dutchess Wappingers Creek69 Ulster Lattintown Creek69 Ulster South Lattintown 75 Dutchess Falkill76 Ulster Twaalfskill 78 Dutchess Maritje Kill81 Dutchess Crum Elbow84 Dutchess Indian Kill84 Ulster Black Creek87 Dutchess Fallsburg Creek87 Dutchess Lands man Kill91 Ulster Roundout98 Columbia South Bay Creek98 Dutchess Saw Kill100 Dutchess Stony Creek101 Ulster Esopus Creek105 Columbia Cheviot Creek110 Columbia RoeliffJansen Kill112 Greene Catskill Creek118 Greene Murderers Creek121 Columbia Stockport Creek121 Columbia Stockport Creek126 Greene Co~sackie 128 Columbia Mill Creek131 Albany Hannacroix 132 Albany Coeymans135 Renssalaer Schodack136 Renssalaer Vlockie Kill137 Albany Vloman Kill137 Renssalaer Papscanee 142 Albany Nomans Kill144 Renssalaer Mill Creek149.5 Renssalaer Wynants Kill150 Renssalaer Poesten KillAbove Troy Dam Mohawk RiverSecondary Tribl Secondary Trib2 M to barrierFt to barrier100 3281,620 5,315240 787950 3,117450 1,4762,860 9,384820 2,6902,100 6,8904,500 14,7652,610 8,563Peekskill Hollow Sprout Brook 1,140 3,740Peekskill Hollow 2,310 7,5793,000 9,843840 2,7561,160 3,8061,240 4,068880 2,887160 5254,740 15,552100 328980 3,215180 591Hunters Brook 3,380 11,090S. Lattintown 550 1,8051,100 3,609100 328Highland Brook 400 1,312190 623270 8861,200 3,9371,670 5,4792,000 6,5622,100 6,8903,820 12,533890 2,920970 3,1832,290 7,5131,850 6,070380 1,2479,320 30,579Kaaterskill Creek 4,940 16,208930 3,051Claverack Creek 1,250 4,101Claverack Creek Kinderhook Cree 1,780 5,840Sickles Creek (dry) 1,270 4,1671,870 6,1351,650 5,414300 984Muitzes Kill 10,900 35,7631,880 6,1681,130 3,708Moordener Kill 1,550 5,0862,970 9,745210 689430 1,411310 1,017183,000 600,42352 REVISED VERSION:

September 2011, based on public comment received.

Appendix Table A continued.

County StreamBronx Bronx RiverHutchinson RiverWestchester Beaver Swamp BrookBlind BrookByram RiverMamaroneck RiverNew Rochelle CreekOtter CreekLong IslandShore Stream&.or Pond with outlet Tributary Alewife Present?South Beaverdam Creek UnknownSouth Browns River UnknownSouth Carlls River Confirmed South Carmans River Confirmed South Connetquot River Westbrook, Rattlesnake Creek UnknownSouth Massapequa Creek Confirmed South Mud Creek UnknownSouth Patchogue River UnknownSouth Penataquit Creek UnknownSouth Swan River UnknownSouth Champlin Creek UnknownSouth Forge River UnknownSouth Pipes Creek UnknownNorth Beaver Brook UnknownNorth Cold Spring Brook UnknownNorth Fresh Pond/Baiting Hollow Confirmed North Mill River, Oyster Bay UnknownNorth Nissequogue River Confirmed North Setauket Mill pond UnknownNorth Stony Hollow Run, Ctrpt. UnknownNorth Sunken Meadow Creek Confirmed North Wading River UnknownEast Fnd Alewife Brook Confirmed East End Alewife Creek/Big Fresh Pond Confirmed East End Big Reed Pond Confirmed East End Fly Pond Restoration stocking effortEast End Gardiner Bay Creeks UnknownEast End Georgica Pond UnknownEast End Halsey's-Neck Pond UnknownEast End Hog Creek UnknownEast End Hook Pond UnknownEast End Ligonee Brook Confirmed East End Mill Pond -MecoxBay Ext. UnknownEast End Peconic River Confirmed East End Sagaponack Pond -Jeremy's Hole UnknownEast End Scoy Pond Restoration stocking effortEast End Silver Lake/Moore's Drain UnknownEast End Stepping Stones Pond Unknown53 REVISED VERSION:

September 2011, based on public comment received.

Appendix Table B. Summary of current (2010) fishery regulations for alewife and bluebackherring in New York State.Fishery / AreaCommercial Harvest:Inland watersHudson River Estuary:

G. Washington Bridge north to Troy Dam (River kilometer 19-245)-Season: 15 March through 15 June-36 hour Escapement period (Friday 6 am to Saturday 6pm, prevailing time)-Net size restriction:

limit of 600 ft, mesh size restriction:

mesh <3.5 inch stretch mesh-Net deployment restrictions (distance between fishing gear > 1500 ft)-Area restrictions (drifted gears allowed in certain portions of the river)Long Island: No restrictions, except for some towns which have restricted fishing within theirtownshipMarine Waters: Hudson River -G. Washington Bridge south; and waters including NY Harbor andaround Long Island-No limits or season.Delaware River: NY portion, north of Port Jervis-No commercial fishery exists in this portion; no rules prohibit itBaitfish harvest:

Take of bait fish (including alewife and blueback herring) are allowed with Bait Licensein the Inland water of New York State. Allowed gears are seines (all Inland waters) and cast nets in theHudson River only.Recreational Harvest:-No daily limit-No season-Harvest can be by hook and line, and some net gears: dip nets (14inches round), scoop nets (13x 13 inches square),

cast net (maximum of 10 feet in diameter) and seine and scap / lift nets 36square feet or less. Anglers must be registered with the New York Recreational Marine Registry.

54 REVISED VERSION:

September 2011, based on public comment received.

Appendix C. Current regulations for river herring fisheries in the Hudson River watershed, and publicsuggestions for change summarized from meetings held in April, 2010. Published in the NYSDECwebsite:

http://www.dec.nvy.gzov/animals/57672.littn Reguation Crrn 201.0 Comnmercial Public suggestions for changeSeason Mar 15 -Jun 15Creel/ catch limits None -Possession limit of 24 fish forcharter boats*-Have a 100 fish daily limit-Have some kind of quotaClosed areas -No gill nets above 190 Bridge -Add: Close tributaries to nets-No nets on Kingston FlatsGear restrictions

-Gill net -Gill neto 600 ft or less o Shorten length to 100 oro 3.5 in stretch mesh or 200 ftsmaller o Add mesh size restriction o No fishing at night in HR o Limit net sizeabove Bear Mt Bridge -Allow no nets-Seine >36 sq ft-No seine >100 ft above 190bridgeEscapement (no fishing 36 hr lift (no gill nets allowed in -36 to 72 hr closuredays) the main river) -Stay away from the weekend-does not apply to scap nets in higher demand for bait)tributaries License Marine Permit -*require a charter boat licenseVaries by gear $1 to $30 -Raise the price of a permit-Increase fee to $75 to $200Include cast nets as commercial Marine Permit (currently need abait license)Make a lottery for obtaining marine permitReporting Mandatory daily catch& effort55I REVISED VERSION:

September 2011, based on public comment received.

Season All year -Be more restrictive

-Choose a season to protect alewife-Choose closure (season) based onwater temperature Creel/ catch limits None -5 to 10 a day(any size, any number) -Allow a special limit for Charterboats: 24/day-Need to know difference betweencreel and possession limit?-Make a slot size &/or size limitClosed areas None -Close all the tributaries to fishing-Close the Mohawk to herring fishing-Have rotating tributary closures(changes every 3 years)-Close parts of tributaries Gear restrictions Angling -No nets, angling only36 sq ft scap or smaller -No nets in tributaries 14" round or 13"x13" dip net -No nets or smaller gear36 sq ft seineMaximum 10 ft diam. Castnet*Escapement (no fishing days) None Close fishing 3 or 4 days a weekAllow herring harvest either on odd oreven days of the weekClose the run during peak of spawningTime closures (hours during the day ornight)Opposed to day closuresMake no-fishing days enough to protectspawningHave sliding closures during the week,i.e. "lure" daysLicense Marine License $10Reporting None -Have a call-in number for harvestlike a HIP #) to get betterinformation

-Create a website for anglers to inputwhat they catchOther issues (other than a fishery) that are creating problems for river herring-Chlorine discharge problems-Ocean harvest is the problem-not the river fishery-Increased silt (covers eggs)Long Island streams:

The lack of data means that no fishery will be allowed under the "sustainable" definition in the ASMFC Amendment

2. Information on habitat and passage issues will be gathered.

56 In a springtime ritual, adults and children went to their localstreams and caught great quantities of the small fish. Prized as oneof the best-tasting fried fish, smelt were brought home for dinner,sold locally, and shipped to distant markets.

Many animals-seals, striped bass, codfish, great blue herons, and others-feasted onrainbow smelt during the springtime bonanza.

Although small insize, this fish played a big role in the ecosystem and economy.Now rainbow smelt are declining, even in streams that oncehosted abundant runs each spring. The diminishing numbershave become evident in the Gulf of Maine. Recognizing the plightof the rainbow smelt, the U.S. government listed it in 2004 as afederal Species of Concern.T-he state governments of Maine, Massachusetts, and NewHampshire are working together to understand the rainbowsmelt's status and threats, and to plan a regional conservation effort for the species.

Scientific research by the three-state collab-orative focuses on the status of the smelt population and thecondition of spawning areas in streams, which may be a key factorin the rainbow smelt's decline.Ice-fishing shacks (above) are evidence ofNew England's long tradition of fishing forrainbow smelt. Scientists (below) from threestates are studying causes of the smelt'srecent decline, including loss of suitablestream habitats (bottom) for spawning.

a0I Rainbow Smeltat a Glance-- 0*1,-Native to coastal waters ofnortheastern United States andCanadian Maritimes.

Eats shrimp, marine worms,amphipods, euphausiids, mysids,and smaller fish.Eaten by porpoises, seals,salmon, trout, bluefish, stripedbass, Atlantic cod, and birds.Slender fish averaging 6 to 8inches long.Can live up to 6 years, but moretypically lives 3 or 4 years.Lives in estuaries,

harbors, andoffshore waters during summer,fall, and winter.Migrates into rivers and streamsto spawn beginning in latewinter (Massachusetts) to latespring (eastern Maine). *:Red dots indicate streamswhere rainbow smelt are,~known to spawn.O4I~A New England Tradition Historically, people in New England valued rainbow smelt as aneasy-to-catch, abundant source of fresh protein after the long winter.The commercial fishery for rainbow smelt is one of the oldest inNew England, and for many years it was among the most valuable.

More recently, the catch along the Gulf of Maine coast has dwindled, although parts of eastern Maine still have strong commercial fisher-ies. Recreational fishing for rainbow smelt continues to be a popularpastime in Massachusetts, New Hampshire, and Maine.Fish in PerilRainbow smelt were so plentiful a hundred years ago that farmerscaught them by the barrelful and had enough to eat, use as bait, andeven spread on their fields as fertilizer.

In many places now, it wouldbe difficult to fill a single barrel with rainbow smelt. The species haslargely disappeared from the southern part of its geographic range,and its numbers along the coast of the Gulf of Maine have droppeddramatically.

In general, rainbow smelt are least abundant in Massa-chusetts and increase slightly toward eastern Maine. Reliable data onpopulation size are not available, but Maine fishery data show thatrainbow smelt landings have dropped tremendously since the 1800s.While a decrease in fishing effort may contribute to the drop in land-ings, the overall trend is clear: rainbow smelt are in trouble.0oIE4-1~0 I x 4 D 10 ODAt present,rainbow smeltlive only northof Long Island Sound(green area).

• ~; r2U..~~

3/4~tMQStr< 4942t" 2"~ t 2< -~4 t t,~ C Q7t~47*.~,pr -~'(tLS ~;7 >2cc~j' i~&rt1/4~ I.3 A7ittUr-ir,~ ~'t,.7*I<'1 *72 N-VK~* 29~2.7 .. v.**A1~3/4-siF"i.wr4I"A-1ii1'TI'*6" What Can You Do?Individual citizens and towns can take important steps to help the rainbow smelt recover.Local efforts are essential and can make a big difference in the survival of the species.1. Use less fertilizer on your property.

Water carries fertilizer into streams, where the nutrients promote growth of algae on smelt eggs.2. Fix dams and culverts blocking smelt from spawning areas.Many dams and culverts prevent rainbow smelt and other fishfrom swimming upstream and downstream.

In collaboration with owners and government

agencies, dams can be removed,culverts reconfigured, and culverts replaced with bridges.3. Plant shrubs and trees along stream banks andrefrain from clearing existing vegetation.

Vegetated buffers help to filter out pollutants,

sediment, and excess nutrients before they enter the waterway.

Shrubsand trees also shade streams, keeping the water cool for fish.4. Maintain natural stream channels and substrate; restorethose altered with concrete walls or other structures.

Faster-flowing water in altered streams can lead to scouring orcrowding of smelt eggs. Low water velocity and unnatural substrates can reduce egg attachment and incubation success.5. Use less road salt and sand near streams.When salt and sand are washed into streams,they can kill smelt eggs.-6. Clean storm drains annually.

Debris and infrequent maintenance can clog storm drains,forcing water to flow over ground. The water carries sedimentinto streams, which smothers smelt eggs.7. Get to know your smelt runs.Find out where smelt spawn in your town and insist thatlocal officials protect these valuable habitats.

I A RegionalConservation PlanforAnadromous Rainbow Smelt inthe U.S. Gulf of Maine0,---ByClaire L. Enterline Maine Department of Marine Resources 172 State House Station,

Augusta, ME 04333Bradford C. ChaseMassachusetts Division of Marine Fisheries 1312 Purchase Street, 3rd Floor, New Bedford, MA 02740Jessica M. CarloniNew Hampshire Fish and Game Department 225 Main Street, Durham, NH 03824Katherine E. MillsUniversity of Maine,Gulf of Maine Research Institute 350 Commercial Street, Portland, ME 04101IM" FFSTHE UNIVERtSITY OFUMAINE ACKNOWLEDGEMENTS This project would not have been possible without the leadership ofJohn W Sowles and Seth L. Barker (Maine Department of MarineResources)

And the many tireless hours of field, laboratory, and statistical work,and administration by:Massachusetts Division of Marine Fisheries:

Matthew H. Ayer, Scott P Elzey, Christopher Wood, Katie L'Heureux, Kim Trull, Carolyn Woodhead, John Boardman, Mike Bednarski, Steve Correia, and Stephanie Cunningham New Hampshire Fish and Game Department:

Douglas Grout, Cheri Patterson, Simon Beirne, Joshua Borgeson, Eric Bruestle, Joshua T. Carloni, Jessica Devoid, Michael Dionne,Robert Eckert, Rebecca Heuss, Elizabeth Morrissey, Conor O'Donnell, Lon Robinson, Kevin Sullivan, Christopher Warner, Kristi Wellenberger, and Rene6 ZobelMaine Department of Marine Resources:

Linda Mercer, Ernie Atkinson, Tim Bennett, Denise Blanchette, Colby Bruchs, Amy Hamilton Vailea, Joseph Gattozzi, Jon Lewis,Marcy Lucas, Celeste Mosher, Anne Simpson, Peter Thayer,Chris Uraneck, Thomas Watson, and the Maine Marine PatrolUniversity of Maine Sea Grant Extension:

Christopher BartlettDowneast Salmon Federation:

Dwayne ShawSubmitted as part of:A Multi-State Collaborative to Develop & Implement aConservation Program for Three Anadromous Finfish Species ofConcern in the Gulf of MaineNOAA Species of Concern Grant Program Award #NA06NMF4720249A Cover illustration by Victor Young©2012 CONTENTSIntroduction

............................................................................................................................................................

31 -Species Status .....................................................................................................................................................

61.1 -Basic Biology ...................................................................................................................................................

6Life History ................................................................................................................................................

6Habitat Use ................................................................................................................................................

6Genetic Stock Structure in Gulf of M aine ...............................................................................................

81.2 -Historical Smelt Fisheries

...............................................................................................................................

12M id-Atlantic

.............................................................................................................................................

12New Jersey ................................................................................................................................................

12New York ..................................................................................................................................................

13Connecticut

..............................................................................................................................................

14Rhode Island .............................................................................................................................................

14M assachusetts

...........................................................................................................................................

15H istorica l F ish eries ...............................................................................................................................

15R ecen t Tren ds .......................................................................................................................................

15New Hampshire

........................................................................................................................................

16Historical Fisheries

...............................................................................................................................

16R ecen t Tren ds .......................................................................................................................................

16M ai n e .......................................................................................................................................................

17Historical Fisheries

...............................................................................................................................

17R ecen t Tren ds .......................................................................................................................................

17Canadian Provinces

...................................................................................................................................

18Historical Fisheries

...............................................................................................................................

18R ecen t Tr ends .......................................................................................................................................

18Summary ..................................................................................................................................................

191.3 -Population Status in the Gulf of M aine ..........................................................................................................

20Previous Smelt Population Studies ............................................................................................................

20Current Fisheries Dependent M onitoring

............................................................................................

21New Hampshire Creel Survey ................................................................................................................

21M aine Creel Survey ..............................................................................................................................

22Current Fisheries Independent M onitoring

..........................................................................................

22State Inshore Trawl Surveys ...................................................................................................................

22M aine and New Hampshire Juvenile Abundance Surveys ...................................................................

23New Hampshire Egg Deposition M onitoring

....................................................................................

23M aine Spawning Stream Use M onitoring

........................................................................................

24Regional Fyke Net Sampling

.................................................................................................................

24Establishing Gulf of M aine Spawning Site Indices ....................................................................

242008-2011 Results .........................................................................................................................

25Study Area Summary .....................................................................................................................

29Conclusions About Regional Fyke Net Sampling

.......................................................................

302 -Threats to Rainbow Smelt Populations in the Gulf of M aine ......................................................................

422.1 -Threats to Spawning Habitat Conditions and Spawning Success ..............................................................

42Spawning Site Characteristics

....................................................................................................................

42Obstructions

.............................................................................................................................................

43D a m s ..................................................................................................................................................

4 3R oa d crossings

......................................................................................................................................

4 4 Contents continued Channelization and Flow Disruptions

.................................................................................................

44Discharge and Velocity

..........................................................................................................................

44Substrate and Channel Stability

............................................................................................................

45W atershed characteristics

..........................................................................................................................

452.2 -Threats to Embryonic Development and Survival

......................................................................................

48W ater Chemistry

.......................................................................................................................................

49Water Temperature

...............................................................................................................................

49Specific Conductivity

............................................................................................................................

49Dissolved Oxygen .................................................................................................................................

50p H ......................................................................................................................................................

5 0T u rb id ity .............................................................................................................................................

5 0D a ta A n a lysis .......................................................................................................................................

5 1Nutrient Concentrations

...........................................................................................................................

51Tota l N itrogen .....................................................................................................................................

5 1Total Phosphorus

..................................................................................................................................

52T N /T P R atio .......................................................................................................................................

52Data Analysis

.......................................................................................................................................

52Periphyton

................................................................................................................................................

52Heavy M etal Concentrations

....................................................................................................................

53W atershed characteristics

..........................................................................................................................

54C o n clu sion s ..............................................................................................................................................

4 42.3 -Threats to Smelt in M arine Coastal W aters .................................................................................................

59Fish Health ...............................................................................................................................................

59Fishing M ortality

......................................................................................................................................

60Overflshing in historical fisheries

.............................................................................................

... 60Incidental catch in small mesh fisheries

............................................................................................

61Predator-prey relationships

........................................................................................................................

62P rey A va ila bility ...................................................................................................................................

6 2Predator Population Shifts ....................................................................................................................

62Community shifts .................................................................................................................................

63Climate-driven environmental change .......................................................................................................

633 -Conservation Strategies

...................................................................................................................................

653.1 -Regional Conservation Strategies

...................................................................................................................

653.2 -State M anagement Recommendations

...........................................................................................................

69M assachusetts

...........................................................................................................................................

69Smelt Stocking Efforts ...........................................................................................................................

70Habitat Restoration

..............................................................................................................................

70Recommendations

.................................................................................................................................

70New Hampshire

........................................................................................................................................

71Population monitoring

.........................................................................................................................

71Habitat Restoration

..............................................................................................................................

72Recommendations

.................................................................................................................................

73M ain e .......................................................................................................................................................

7 3Continue monitoring smelt populations at multiple life stages .............................................................

73Improving connectivity and access to spawning grounds .....................................................................

74Assessing causes for local decline .............................................................................................................

75M arked larval stocking at monitored sites .........................................................................................

75Recommendations

.................................................................................................................................

76Literature Cited ........................................................................................................................................

77A p p en d ix ..................................................................................................................................................

8 52

• ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN INTRODUCTION The rainbow smelt (Osmerus mordax) is a small anadromous fish that over-winters in estuaries and bays prior to spawning each spring in coastal streamsand rivers. Smelt have supported culturally important commercial and recre-ational fisheries throughout New England since at least the 1800s. However,in recent years, concerns have risen about the population status of rainbowsmelt. The species has disappeared from the southern end of its geographic range, which once extended to the Chesapeake Bay and now may extendonly as far south as Buzzards Bay, Massachusetts.

High numbers of rainbowsmelt that once supported commercial fisheries in New England have declinedprecipitously since the late 1800s to mid-1900s.

While recreational fisheries for rainbow smelt continue, declining catches have also been noted by anglers,particularly since the 1980s.Based on these observations of range contraction and abundance

declines, the National Oceanic and Atmospheric Administration (NOAA) listed rainbowsmelt as a federal Species of Concern in 2004; New Hampshire also lists sea-run rainbow smelt as a Species of Special Concern.

Although rainbow smeltpopulation declines have been widely documented, the causes are not wellunderstood.

In listing the species, factors identified as potential contributors included structural impediments to their spawning migration (such as damsand blocked culverts) and chronic degradation of spawning habitat due tostormwater inputs that include toxic contaminants, nutrients, and sediment.

High numbers ofrainbow smelt thatonce supported commercial fisheries in New England havedeclined precipitously since the late 1800sto mid-1900s.

Following the designation of rainbow smelt as a species of concern, theMaine Department of Marine Resources received a 6-year grant from NOAAsOffice of Protected Resources to work in collaboration with the Massachusetts Division of Marine Fisheries and New Hampshire Fish and Game Department to document the status of and develop conservation strategies for rainbowsmelt (NA06NMF4720249).

This conservation plan represents a summary ofkey elements of the project, which focused on several objectives:

1) Documenting range contraction and range-wide population declinesbased on historical data and accounts2) Evaluating the status of rainbow smelt populations in the Gulf ofMaine region3) Developing a population index to track the strength of spawning runs4) Assessing a range of potential threats to rainbow smelt populations

5) Proposing management actions to help conserve rainbow smeltthroughout the Gulf of Maine region.This study has significantly advanced our understanding of the biology,status, and threats to rainbow smelt in the Gulf of Maine. A major contribu-tion was the development of standardized procedures for indexing the abun-dance of spawning rainbow smelt. Four years of fyke net sampling of spawningANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN

• 3 runs throughout the Gulf of Maine region have provided important baselineinformation about the status of the species.

Observations of truncated agestructures within the spawning run, high male to female ratios in some rivers,and lower survival rates and a higher portion of age-1 spawners than historically observed all indicate that Gulf of Maine rainbow smelt populations are cur-rently stressed.

Further evidence of the decline can be derived from a survey of historically active spawning sites throughout the state of Maine, using a study from the1970s (Flagg 1974) as a valuable baseline for comparison.

The recent surveyfound that 13% of the historically active spawning streams no longer supportrainbow smelt spawning, and most of the streams that remain active now sup-port smaller runs than they did historically.

The substantial decline in strongspawning runs merits concern and attention.

Many threats to rainbow smelt spawning habitat were identified as part ofthis study. Obstructions such as dams and improperly designed culverts mayphysically impede smelt migration to appropriate spawning sites. Further,extremely high or low flows can impede swimming ability or impair the cuessmelt rely on to undertake this migration.

Once on the spawning grounds,water quality conditions may affect the hatching and survival of smelt eggs. Inmany rivers studied as part of this project, pH, turbidity, nutrient levels, anddissolved contaminants warranted concern for water quality.

Field observations also showed an association between nitrogen levels and periphyton growth atspawning

grounds, and laboratory experiments demonstrated that high periph-yton growth significantly impaired the survival of smelt embryos.Many of these threats-particularly flow patterns and water quality-are not driven by factors within the spawning rivers themselves, but rather by ac-tivities in the surrounding watersheds.

Across a suite of water quality and heavymetal parameters, we found that high levels of development in the watershed were associated with poorer conditions for rainbow smelt, while high propor-tions of forest in the watershed supported high quality stream conditions.

In conjunction, watershed development was negatively associated with thestrength of smelt spawning runs, while forested watersheds supported strongerruns in their receiving streams.Our goal in assessing threats to rainbow smelt was to identify conditions that appear to negatively and positively affect smelt throughout their life cycleso that management actions can effectively target these factors.

Based on ourassessment of critical

threats, management recommendations to protect andrestore rainbow smelt populations include:" Maintain the federal Species of Concern designation for rainbow smelt" Continue monitoring population trends and biological characteristics in the extant range, and expand efforts towards estimating rainbowsmelt population size" Restore historical or degraded spawning habitat" Maintain and, where necessary, improve fishery monitoring to ensurethat fishing effort is compatible with sustainability of local and regionalrainbow smelt populations

• Expand research initiatives to anticipate direct and indirect effects of4

• ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN dimate change and variability on rainbow smelt" Invest in research to further study environmental requirements, stress-ors, and drivers in order to effectively manage recovery" Stock marked larvae to re-establish rainbow smelt runs at restoredsites, as needed and as appropriate given considerations of geneticdiversity and donor population viability This Conservation Plan provides:

a description of the life history ofanadromous rainbow smelt; an account of the historical fishing pressure on thespecies; a summary of the current population status and monitoring efforts;explanation of the threats to the species at different life stages, including themarine phase; and conservation and management strategies for the region andfor each state in the Gulf of Maine. Our intent is that this information willprovide important baseline information regarding the status of smelt popula-tions at the present time and that it will offer coastal and fishery managersguidance on appropriate actions and priorities to protect and restore rainbowsmelt moving forward.ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN e 5 1 -SPECIES STATUSAnadromous smelt serveas an important preyspecies for commercially and culturally valuablespecies, such as Atlanticcod, Atlantic salmon,trout, Atlantic gray seals,striped bass.Rainbow smelt (Osmerus mordax) are small anadromous fish that live innearshore coastal waters and spawn in the spring in coastal rivers immediately above the head of tide in freshwater (Buckley 1989, Kendall 1926, Murawski etal. 1980). Landlocked populations of smelt also naturally occur in lakes in theNortheast U. S. and Canada and have been introduced to many freshwater sys-tems, including the Great Lakes. Anadromous smelt serve as an important preyspecies for commercially and culturally valuable

species, such as Atlantic cod,Atlantic salmon, trout, Atlantic gray seals, striped bass (Clayton et al. 1978,O'Gorman et al. 1987, Kircheis and Stanley 1981, Kirn 1986, Stewart et al.1981). Historically, the range of rainbow smelt extended from Chesapeake Bayto Labrador (Buckley 1989, Kendall 1926), but over the last century, the rangehas contracted and smelt are now only found east of Long Island Sound.1.1 -BASIC BIOLOGYLife HistorySmelt are small-bodied and short-lived, seldom exceeding 25 cm in lengthor five years of age in the Gulf of Maine region (Murawski and Cole 1978,Lawton et al. 1990). By age two, smelt are fully mature and recruited to localrecreational fisheries and spawning runs. Life history appears to be influenced by latitude; few age-I smelt become mature and participate in Canadian smeltruns, however in Massachusetts, New Hampshire, and southern Maine, age-I individuals are present in the spawning runs (Collette and Klein-MacPhee 2002). Studies in Massachusetts found that the majority of age-1 spawners weremale (Murawski and Cole 1978, Lawton et al. 1990). Our current spawn-ing surveys have found that runs in the Gulf of Maine are dominated by age-2smelt, with few older smelt in Massachusetts, New Hampshire, and southernMaine; however the older ages are better represented in midcoast and easternMaine. Fecundity estimates of approximately 33,000 eggs for age-2 smelt and70,000 eggs for age-3 smelt were reported by Clayton (1976).Habitat UseAnnual movements and habitat use by adult rainbow smelt have been large-ly assumed based on discrete sampling or patterns in recreational and commer-cial fishing.

Mark and recapture studies have focused on distinct phases of thelife cycle, such as movements between spawning areas (Murawski et al. 1980),composition of late and early populations of spawning adults (McKenzie 1964)and winter movements within a river system (Flagg 1983). Larger annual andregional migrations have been synthesized from anecdotal reports by anglersand commercial fishermen as well as from beach seine and spawning surveys.6

• ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLANI Rainbow smelt overwinter in estuaries and bays and then spawn in earlyspring in pool and riffle areas above the head-of-tide in coastal streams andrivers. The spawning habitat characteristics are discussed in detail in sections2.1 -Threats to Spawning Habitat Conditions, and 2.2 -Threats to Embry-onic Development and Survival.

Because males have a longer physiological spawning period, they may return to spawning grounds multiple times withinthe same year (Marcotte and Tremblay 1948). Mark and recapture studies haveobserved the same male at different spawning sites within a given year, suggest-ing that males are able to spawn multiple times (Murawski et al. 1980, Rupp1968). Murawski et al. (1980) hypothesized that spawning in different streamsmay be facilitated by passive tidal transport, however this has not been directlyobserved.

Females, on the other hand, rarely ascend to the spawning groundsmore than once in a season, based on recent mark-recapture surveys (C.Enterline, unpublished data). Because female smelt are broadcast

spawners, their spawning is expected to occur in a single event as most or all of their eggsare deposited in a single event.Spawning females deposit demersal (sinking) adhesive eggs that attachto the substrate and hatch in 7-21 days, depending on temperature.

Uponhatching, larvae are immediately transported downstream into the tidal zone,at which point the larvae begin feeding on zooplankton.

Larval dispersion is mostly passive in response to river flow and coastal circulation

patterns, but there is also an active (swimming) component (Bradbury et al. 2006b).Although horizontal movements of smelt larvae appear passive, they activelymigrate vertically in response to tidal flow in order to maintain their positionin zooplankton rich water and minimize downstream movement (Laprise andDodson 1989, Dauvin and Dodson 1990, Sirois and Dodson 2000). Thisactive swimming behavior is overwhelmed by passive transport in localcirculation patterns.

The importance of local circulation on larvae dispersion isdiscussed more in the genetic stock structure section below.Juvenile smelt remain in the estuary, bay, or sheltered coastal area throughthe summer, and sometimes through the early fall (NHF&G and ME DMRJuvenile Abundance

Surveys, 1979-2011, analysis for current study). In GreatBay, NH, juvenile smelt are most abundant in August, while in the Kennebecand Merrymeeting Bay estuary complex in Maine, abundance is more evenlydistributed between August, September, and October (Figure 1.1.1). In Maine,catches of juvenile smelt occur from July to October, while in New Hampshire, catches range from June to November.

Habitat use in marine waters is largely unknown but can be inferredthrough interviews with coastal fishermen and state trawl surveys.

Smelt maymigrate in search of optimum water temperatures, moving offshore during thesummer months to greater depths with cooler water (Buckley 1989). Basedon low catches by fishermen in freshwater and larger catches in brackish andsaltwater in May, the presumed end of the spawning run, it has been assumedthat adults return to estuaries and coastal waters immediately after spawning(Bigelow and Schroeder 1953). However, recent findings indicate that rainbowsmelt may remain within estuaries and bays contiguous to their spawning sitesfor up to two months after spawning (C. Enterline, unpublished data).Recent trawl surveys have found small schools of smelt as far from the coastas 60 km and in depths up to 77 m (data from the Maine-New Hampshire andANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN

• 7 Massachusetts Trawl Surveys).

Spring trawl surveys find smelt further from thecoast and in deeper water (spring avg. depth = 29.7 m) than during fall trawlsurveys (fall avg. depth = 19.9 m) (Figures 1.1.2 and 1.1.3; t-test comparing depth, p = 0.0338 < 0.05), however the average spring catch is smaller com-pared to the fall (spring average catch 2001-2012

= 31, fall average catch 2000-2011 = 129, Wilcoxon non-parametric test of means, p < 0.0001 < 0.05), likelybecause adult smelt are within coastal streams and rivers as part of the spawningevent during the spring period. The smelt that are caught further offshore inthe spring are smaller, with lengths associated with age- I fish; these are likelyyoung fish that are not recruited to the spawning run.As offshore water temperatures drop in the fall, smelt likely move towardsthe coast, eventually migrating into the upper estuaries where they overwin-ter (Buckley 1989; Clayton 1976; McKenzie 1964). Anecdotal reports fromrecreational hook-and-line ice-fishermen describe smelt moving in tidal riverswith the nighttime flood tide and out with the ebb tide, and some moving asfar up as the head of tide each night. These foraging movements are the basisfor robust recreational fisheries in the fall and winter at many locations in theGulf of Maine.Genetic Stock Structure in the Gulf of MaineUnderstanding the genetic structure of a species and the driving factorsbehind that structure is central to well-designed species management.

Aspecies may be comprised of one or more genetic stocks, separated by different spawning areas or physical barriers.

Managing a species at too large a scale (i.e.,assuming there is only one stock when there are multiple) may lead to the lossof genetic structure and the benefits of local adaptation.

Managing at too smalla scale (i.e., assuming stocks are isolated within individual rivers when in factthere is some mixing),

neglects the important role of gene flow and results inloss of genetic variation (Kovach et al., in press).From 2006-2010, we collected genetic samples at 18 spawning site indexstations spanning the Gulf of Maine to understand if unique genetic stocksexisted and the extent of gene flow between spawning populations.

All informa-tion presented in this conservation plan was reported by the University of NewHampshire and in detail by Kovach et al. (in press). The three most genetically divergent populations were found in Cobscook Bay, Maine, Massachusetts Bay,and Buzzards Bay, Massachussetts.

Penobscot and Casco bays in Maine alsoshowed some differentiation.

Gene flow was high between rivers from downeastcoastal Maine, the Kennebec River, ME, and Great Bay, NH to northernMassachusetts; all were dominated by the same genetic signal. Midcoast Mainealso seemed to be part of this large stock, but also showed distinct signals fromPenobscot Bay and Casco Bay (Figure 1.1.4). These groupings can assistmanagement decisions on stocking

efforts, with the goals of maintaining distinct stocks where possible, while still preserving gene flow to maintain andreplenish genetic diversity.

Although the study did not find evidence of genetic bottlenecking, geneticvariation was significantly reduced in the two most distinct regions:

BuzzardsBay (Weweantic River), and Cobscook Bay (East Bay Brook) (Kovach et al.,in press). The reduced diversity in the Weweantic River is consistent with its8 ° ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN location at the southern extent of the species range, where populations can havereduced gene flow and lower spawning population sizes (Schwartz et al. 2003).The reduced variation in Cobscook Bay is more likely due to isolation bycirculation patterns.

The reduced diversity and distinctive nature of these smeltruns warrant further population monitoring and possibly updated protection measures.

The divergence patterns observed may be explained partly by coastal circu-lation patterns (Kovach et al., in press). Because the movement of smelt larvaeis largely passive during the early development (Bradbury et al. 2006b), theirdispersal is determined first by river flow and secondly by marine circulation.

The Gulf of Maine Coastal Current (GMCC) has a counter-clockwise pat-tern, which is strongest in the summer months when smelt larvae are presentin coastal waters. The GMCC consists of two distinct portions.

The EasternMaine Coastal Current (EMCC) flows from the Bay of Fundy southwest alongthe coast and, in the area of Penobscot Bay, often splits southward and offshore.

The remaining portion of the EMCC combines with outflow from Penobscot Bay and continues southwestward towards coastal New Hampshire and Mas-sachusetts, creating the Western Maine Coastal Current (WMCC; Pettigrew et al., 1998, 2005). Backflow eddies are associated with large rivers (like thePenobscot) and to a lesser extent with Casco Bay, and as a result, larvae may bemaintained within the nearshore area. Continuing further southwest along thecoast, Massachusetts Bay maintains high larval retention as the strength of theWMCC pattern has largely diminished by this point (Incze et al. 2010).Figure 1.1.1. Mean smelt catch600 by month In the Maine and NewHampshire Juvenile Abundance 500" Surveys 1979-2011 for all surveysites combined.

Error bars repre-400 sent one standard error from themean.300-200"100 -jd 0"500400-300"200-1000 M I5 7 8 9 10 11MonthANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN

• 9 Figure 11.2. Smeltcatches In the fall statenearshore trawl surveysfor Massachusetts, NewHampshire, and Maine2000-2011.

Figure 1.1.3. Smeltcatches In the springstate nearshore trawlsurveys for Massachu-setts (2000-2011),

New Hampshire, andMaine (2000-2012).

10

• ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN Figure 1.i4. Genetic differentia-tion of smelt stocks In the Gulfof Maine from Kovach et aL,("in press').

Divergence may beexplained by circulation

patterns, where the Gulf of Maine CoastalCurrent carries larvae fromdowneast coastal Maine to NewHampshire and northern Mas-sachusetts, while other localized circulation patterns maintainthe distinctiveness of Penobscot Bay, Casco Bay, Massachusetts Bay, and Buzzards Bay. The colorboxes display the 6 genetic signals-boxes with the same colorsIndicate the same signal. Lengthof boxes represents number ofsamples taken from the region.Ca88oEMCCWMCC-Massachusetts BayBuzzards BayJ110IKilometers ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN

• 11I By the late 1800s, withthe advancement of railtransport, smelt were animportant export productshipped on ice from theCanadian Maritimes andMaine to the Boston andNew York markets.1.2 -HISTORICAL SMELT FISHERIES Smelt fishing is a longstanding tradition in many coastal communities ofNew England and the Canadian Maritimes.

During winter and early spring,smelt schools enter estuaries and embayments and aggregate in preparation forthe spring spawning run. During this period of migration, commercial, andrecreational fisheries target smelt through the ice and from shore. Some shorefisheries also occur in fall, mainly with hook and line, during foraging move-ments that precede the spawning migration.

Fishing methods for smelt vary bystate; including weirs, hook and line, seines, dip nets, bag nets, and gill nets.This section will describe the historical range of rainbow smelt and thefisheries that targeted them. We focus on the Gulf of Maine, but provide somebackground on populations throughout the range. We rely heavily on the classicwork "The Smelts" by Kendall (1926) and the thorough recent literature reviewfound in Fried and Schultz's (2006) investigation in Connecticut.

The earliest record of smelt harvest in the U. S. was likely by Captain JohnSmith in 1622; Smith noted the smelts were so plentiful that the Native Ameri-cans would harvest the fish by simply scooping them up in baskets (in Kendall1926). There is little additional information about early New England smeltharvests until the mid-1800s, although extensive subsistence and local com-mercial harvest occurred before this time, based on occasional references andtown records.

Early uses of smelt included livestock feed and fertilizer to enrichfarm fields. The abundance of smelt in the mid-1800s can be pictured fromthe account of French settlers along the Buctouche River in New Brunswick harvesting 50 to 60 barrels (36 gallons/barrel) annually to serve as fertilizer foreach homestead (Perley 1849 in Kendall 1926). About this time, food marketsdeveloped for smelt as human populations grew in coastal cities. By the late1800s, with the advancement of rail transport, smelt were an important exportproduct shipped on ice from the Canadian Maritimes and Maine to the Bostonand New York markets (Kendall 1926).Mid-Atlantic Smelt are considered a cold water fish, with a historical center of abundance north of Cape Cod but southerly populations ranging south to the Mid-Atlan-tic. Early references of smelt range include Virginia, Maryland and Delaware(Goode 1884, Kendall 1926, Bigelow and Schroeder 1953), but we found noinformation on smelt populations or harvests for these states. Later references on smelt range list New Jersey as the southern limit (Scott and Scott 1988,Collette and Klein-MacPhee 2002). Overall, references south of Delaware Bayare not well documented.

The presence of smelt in states south of New Jerseymay have been sparse, an indication of occupancy at the edge of the species'range, or alternatively the fisheries may have faded before the onset of recordedcommercial harvest data in the early 20th century.New JerseyIn 1833, smelt were observed to be plentiful in New Jersey with "wagon-loads" of smelt harvested in Newark Bay, yet by 1849, smelt were reported asdeclining (New York Times 1881 in Fried and Schultz 2006). The Delaware12 -ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLANI River had been listed as a southern smelt run, including an early observation ina tributary, the Schulykill River, of cast net fishing for smelt during late winter(Norris 1862). Spring runs of smelt, also called frost fish, were reported in theDelaware, Hackensack, Passaic and Raritan rivers during the late 1860s. Bythis time, only the Raritan River supported a lucrative commercial

fishery, withannual catches nearing 10,000 lbs (NJCF 1872). The New Jersey Commis-sioners of Fisheries (NJCF) 1872 report also suggested that industrial waterpollution in the rivers was severely impacting all anadromous fisheries.

Thelast regular commercial catch in New Jersey was reported in 1921 (Fried andSchultz 2006).Smelt were considered endangered in New Jersey by 1877 and the statelaunched an effort in the 1880s to study the reproductive biology of smelt andto stock smelt fry hatched from eggs collected in viable smelt runs to depletedsmelt runs (NJCF 1886).No evidence of stocking success has been located and by 1941 smelt wereconsidered extirpated from New Jersey (Camp 1941 in Fried and Schultz2006). The New Jersey Fish and Game Department has conducted trawlsurveys throughout their coastal waters since the early 1980s, and no smelthave been detected during this time.New YorkHistorical references indicate that tributaries near the Hudson River andLong Island once supported prominent recreational and commercial fisheries but that overfishing and poor water quality likely caused declines be-fore the end of the 19th century (Kendall 1926). The smelt trade at the FultonMarket in New York City was reported to average 1,352,000 lbs annually inthe 1870s (Scott 1875 in Kendall 1926). By 1887, the smelt fishery was nolonger considered commercially viable (New York Times 1881, Mather 1887,Mather 1889; in Fried and Schultz 2006). State fishery agencies in New Yorkbecame concerned about the declining status of smelt in the late 1800s andembarked on extensive stocking efforts that included placing 127 million eggsin Long Island streams during 1896-1898 (Kendall 1926). The stocking effortsfaded when smelt eggs became scarce in the early 20th century (Kendall 1926).Commercial catches declined and became sporadic in the 20th century.

Rou-tine commercial harvests exceeding 1,000 lbs annually were last reported in the1950s (Fried and Schultz 2006).Since the 1970s, annual surveys in New York have detected rainbow smelt,but catches have become increasingly infrequent and have been rare since the1990s. The Hudson River Estuary Monitoring Program has conducted ichthy-oplankton and juvenile fish surveys throughout the estuary since 1973, and thedata show a dramatic decrease in smelt abundance since the mid-1990s, withonly trace numbers detected today (ASA A&C 2010). Fish sampling effortsconducted by New York State Department of Environmental Conservation (NY DEC) have produced similar results, with very few adults detected sincethe 1980s. Today, smelt are considered extirpated or at extremely low numbersin the Hudson River system (C. Hoffman, NY DEC, pers. comm. Sept. 2010).ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN 9 13 Connecticut A synopsis of early fisheries records shows that smelt runs were present inmost tidal rivers in coastal Connecticut, and economically important commer-cial fisheries targeted the seasonal occurrence of smelt (Visel and Savoy 1989,Fried and Schultz 2006). Smelt were targeted primarily with haul seines andgill nets in the Housatonic, Connecticut and Pawcatuck rivers (Visel and Savoy1989). Hook and line angling was also common in the 19th century at numer-ous locations; smelt were described as an important export fish to New YorkCity markets.

Smelt landings were reported as peaking in Connecticut in the1880s at 27,000 lbs and steadily declining with minor and intermittent land-ings since the 1930s (Fried and Schultz 2006). There was a modest increase inlandings in the 1960s when several thousand pounds were reported annually.

The last years with significant smelt runs in Horseneck Brook of Greenwich, were 1965 and 1966 (Visel and Savoy 1989).By the 1980s, smelt were recognized as nearly absent from Connecticut's coastal rivers. Similar to regions south of New England, concern centeredon the role of point and non-point pollution sources (Visel and Savoy 1989).The decline of smelt in Connecticut prompted dedicated efforts to documenttheir presence in the 2000s. The smelt fishery was formally closed to harvest in2005, and smelt were listed as a state endangered species in 2008. Fried andSchultz (2006) carried out intensive surveys in five estuaries along the centraland eastern Connecticut coast. They documented no evidence of smelt spawn-ing but did catch 9 adults while seining in the upper Mystic River during 2004.State beach seine surveys infrequently encounter smelt, however there havebeen recent observations of a few adult smelt in 2007 (T. Wildman, CT DEPInland Fisheries

Division, pers. comm. Nov. 2010). The State of Connecticut iscurrently considering listing smelt as extirpated from the state.Rhode IslandSmelt landings first appear in Rhode Island records in 1880 with landingsof 95,000 lbs, which remains the peak annual harvest for this state (Fried andSchultz 2006). Since that point, landings records steadily declined with minimallandings reported after 1932. Landings rebounded slightly during 1965-1970 when several thousand pounds were reported annually.

Since this time, minimalcommercial landings have been reported (Fried and Schultz 2006). In response todeclining populations, the Rhode Island Division of Fish and Wildlife (RIDFW)began a smelt stocking and monitoring program in 1971 (RIDFW 1971). Overthe next seven years, approximately 44 million smelt eggs were transferred frompopulations in Massachusetts and New Hampshire to four rivers in RhodeIsland. Extensive monitoring was conducted at the four recipient rivers, andno evidence was found of successful recruitment following stocking (RIDFW1978). The monitoring only found evidence of a viable smelt run in thePawcatuck River where low densities of smelt eggs were observed in 1974. Thestocking effort was considered unsuccessful and discontinued in 1977 (RIDFW1978). In the last decade smelt were briefly listed as endangered in Rhode Island,then delisted and considered extirpated with a chance of a trace populations present.

Adult smelt have been captured on rare occasions during coastal pondand bay surveys since the 1990s (A. Libby, RI DFW, pers. comm. Oct 2011).14

• ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN Massachusetts Historical Fisheries Early accounts indicate that smelt populations in Massachusetts supported culturally important sustenance fisheries that evolved into small-scale commer-cial and recreational fisheries as coastal populations grew. The smelt fisheries prior to 1874 targeted fall and winter feeding aggregations with baited hooksand used dip nets and seine nets during the spring spawning runs (Kendall1926). The local importance of these fisheries and the potential abundance ofthe populations is reflected in accounts that describe over nine million smelttaken from the Charles River at Watertown in 1853 (Storer 1858), and over2,300 fishermen at Hough's Neck in Quincy in one day targeting smelt (Kend-all 1926). Overfishing concerns were raised in the 1860s that were attributed towith the use of nets during the spawning run. This concern led the Massachu-setts State Legislature to prohibit net fishing for smelt during the spawning runin 1868 (Kendall 1926).In 1874, a law prohibited the taking of smelt by any method other thanhook and line in all state waters with a few exempted rivers -most of theseexemptions were revoked by the end of the century.

Kendall (1926) relatesaccounts of rebounding smelt fisheries in the 1870s and praise for the net ban.Catch records are sporadic and largely town or county specific during the latterhalf of the 19th century.

However, there was a general declining trend in thisperiod, and by the 191 Os and 1920s there was growing concern about smeltfisheries in Massachusetts and the influence of industrial pollution.

A quote theMassachusetts Commissioners on Fisheries and Game in 1917 expressed theconcern of the period, "The smelt fishery of Massachusetts is in a depletedcondition, and strenuous and radical measures will be required to save this spe-cies from extinction" (MCFG 1917).Smelt fisheries are poorly documented in Massachusetts after Kendall's 1926 report. The annual reports of the state fisheries agency depict contrasting trends along a gradient.

In southern Massachusetts, there was a sharp declinein commercial importance and the disappearance of smelt in some locations.

However, north of Cape Cod and in the greater Boston area, an active andpopular fall and winter sportfishery persisted through the 1970s. Fried andSchultz (2006) summarized federal commercial catch records that show threetime-series peaks in Massachusetts harvest:

1880 (82,034 lbs), 1919 (39,000lbs), and 1938 (25,000 lbs). The early landings data were based on the available town and county records and are expected to be incomplete (Kendall 1926).It is likely that no records adequately describe the true extent of smelt harvestat any time in Massachusetts's history.

The view provided by the combinedhistorical and anecdotal accounts suggests that smelt supported important sea-sonal fisheries that attracted large numbers of anglers and that smelt occurrence and abundance greatly exceeded the species' present status.A quote the Massachu-setts Commissioners onFisheries and Game in1917 expressed theconcern of the period,"The smelt fishery ofMassachusetts is in adepleted condition, andstrenuous and radicalmeasures will be requiredto save this species fromextinction."

Recent TrendsStriking changes appear to have occurred in smelt detection and abundance in Massachusetts since Kendall's report (1926). Contemporary studies beganwith river-specific work in the Jones and Parker rivers in the 1970s (Lawton etal. 1990, Murawski and Cole 1978, and Clayton 1976). These studies werethe first to report biological characteristics of the spawning runs and timing ofANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN

• 15 movements in Massachusetts.

Concerns over declines in smelt abundance grewafter these studies, as sportfisheries' catches declined sharply in the late 1980s.The MA DMF responded to concerns from the sportfishing community with asurvey of all smelt spawning habitats on the Gulf of Maine coast within Massa-chusetts during the 1990s (Chase 2006) and the initiation of fyke net monitor-ing in 2004 to develop population indices.Specific mention of Buzzards Bay is warranted because it is presently thesouthern limit of the documented spawning range. Buzzards Bay lies directlysouth of Cape Cod, which separates the Virginian marine ecoregion to thesouth from the Gulf of Maine/Bay of Fundy ecoregion to the north (Spalding et al. 2007). No historical records have been found of spawning runs on CapeCod, a likely result of its glacial formation and flat gradient.

Goode (1884)reported smelt harvest in coastal weir fisheries in Buzzards Bay in 1880. Morerecently, an anadromous fish survey from 1967 reported 10 rivers in BuzzardsBay with active smelt spawning runs (Reback and DiCarlo 1972). An estuarine survey of the Westport River in Buzzards Bay in 1966-1967 found smelt inseine and trawl surveys and reported a known spawning run and associated fish-ery in the river (Fiske et al. 1968). Smelt runs in the region have since quietlyfaded to low levels of detection.

Fisheries monitoring during the last 10 yearshas documented the presence of smelt in only three Buzzards Bay rivers; with alone viable spawning run in the Weweantic River.New Hampshire Historical Fisheries Significant smelt fisheries of commercial and cultural importance haveoccurred in the Great Bay estuary of New Hampshire since the 18th centuryor earlier.

Hook and line fishing has mainly occurred in winter through iceon tidal waters. Additionally, bow nets were traditionally fished under the ice,and weirs were deployed during spring spawning runs (Warfel et al. 1943).Historical fisheries in New Hampshire are poorly described relative to Maineand Massachusetts.

Kendall (1926) provides very little information on coastalNew Hampshire smelt runs, focusing more on landlocked populations.

Hedoes provide annual smelt harvest estimates for coastal fisheries as follows:

1888-36,000 lbs, 1908 -2,600 lbs, and 1924 -3,835 lbs. The reported peak ofcommercial catch in New Hampshire was between 1940-1945, with anestimated 150,000 lbs harvested per year (Figure 1.2.1; Fried and Schultz2006). It is expected that the historical records substantially underreported actual harvest from the Great Bay fisheries.

Recent TrendsThe state of New Hampshire has monitored smelt fisheries in Great Baysince the 1970s, when concerns were voiced from fishery participants aboutdeclining catches.

To this end, an angler creel survey was started in 1978 anda smelt egg deposition survey began in 1979. A project was also launched atthat time to improve commercial harvest data by mandating bow net and weirnet fishermen to record their catches in log books. In 1981, a statewide smeltfishery management plan was written by the New Hampshire Fish and GameDepartment (NHF&G) to maintain sea-run smelt populations and supportcommercial and recreational fisheries (NHF&G 1981).16

• ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN Data collected by the NHF&G indicate declining population trends inrecent decades.

The angler creel survey data depict a reduction in CPUE andtotal catch during the 2000s (Sullivan 2010). The smelt egg survey shows eggdensities in the 2000s that are an order of magnitude lower than the 1980s(Sullivan 2007); the survey was discontinued in 2008 due to concerns overmethodology and very low presence of smelt eggs. The commercial harvest re-cords in New Hampshire have also recorded declines since 1987 (Figure 1.2.1).Commercial dip net and bow net permits remain active, but the fisheries havedeclined to low levels of catch and effort (J. Carloni, NHF&G, pers. comm.,2011). Despite the apparent decreasing trends, recreational fishing for smelt inGreat Bay still remains a popular winter fishery that attracts higher catch andeffort than fisheries to the south in Massachusetts.

MaineHistorical Fisheries Commercial and sustenance smelt fisheries were important to Maine'scolonial inhabitants as early as the 18th century, but are poorly documented.

Kendall (1926) provides detailed accounts of valuable commercial hook andline and net fisheries from the 1880s to 1920s. The opening of export marketsto New York and Boston after the mid-i 800s, coupled with growing use ofseine and bag nets, led to increases in harvest and the development of a signifi-cant commercial fishery.

Goode (1884) provides the first reported commercial smelt harvest records for Maine, with landings exceeding a million pounds inthe 1880s. In 1894 the smelt fishery was reported to support 1,100 fishermen with shore fishery landings that were the fourth most valuable behind lobster,clams, and sea herring (Whitten 1894). Statewide records are absent beforethis time, however subsequent catch data show a steep decline after the 1890s(Squires et al. 1976; Figure 1.2.1). The last year the Maine catch exceeded amillion pounds was in 1903. As early as 1920, a report by the Maine Commis-sion of Sea and Shore Fisheries described the depleted status of smelt runs andthe negative impacts of targeting spring spawning aggregations for commercial harvest (MECSSF 1920). An early management response to this decline wasperforming egg transfers from both landlocked and sea-run smelt populations to depleted runs (Kendall 1926); these were largely undocumented.

While thecommercial fishery continued to decline in the 20th century, the recreational fishery that targeted smelt both through the ice and during spawning runsincreased in catch and effort starting in the 1940s. The rental ice shack fishery,in particular, grew in economic importance as out-of-state anglers were attract-ed to Maine's coastal rivers.Recent TrendsRecognizing the traditional importance of the smelt fishery and continued population

declines, the Maine Department of Marine Resources (ME DMR)developed a Smelt Management Plan in 1976 (Squires et al. 1976). The planoutlined present conditions and made recommendations to improve fisheries and spawning habitat.

It also attributed the dramatic decline observed in themid 20th century to increased industrial pollution in Maine's rivers after WorldWar II (Figure 1.2.1). The ME DMR also launched studies at this time torecord the presence and distribution of smelt in coastal Maine and investigate ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN -17 Anadromous smeltpopulations in Canadahave long supported valuable commercial fisheries that greatlyexceed the collective harvest from the UnitedStates.causes of the historic decline (Flagg 1974). Flagg's (1974) work on Maine'ssea-run smelt documented catches at camp fisheries on the Kennebec Riverand Merrymeeting Bay, and catalogued spawning runs on 134 coastal streams.As part of the present study, the ME DMR has reinstituted creel surveys andspawning habitat investigations so that current catch records can be comparedto the 1970s monitoring.

Maine continues to have important recreational fisheries featuring winter ice fishing on tidal rivers and spring dipnet fishing atspawning runs, although annual harvest is at historic lows. A modest commer-cial harvest continues in downeast Maine, largely centered on the Pleasant Riverin Columbia Falls, where gill and bag nets are allowed to fish in late winter.Canadian Provinces Historical Fisheries Anadromous smelt populations in Canada have long supported valuablecommercial fisheries that greatly exceed the collective harvest from the UnitedStates. Among provinces, New Brunswick has had the largest fishery, whichhistorically targeted smelt for use as fertilizer and bait (Goode 1884). Growingexport markets were driven by the Canadian

harvests, which were, and continueto be, the largest commercial harvests in the species' range. Records are sparsebefore the 20th century, however Kendall (1926) cites accounts of fast develop-ing export markets to Boston and New York in the 1870s that created demandfor large harvests

-exceeding two million pounds by the 1880s. In 1901, theshipment records of one export company in New Brunswick approached eightmillion pounds. The highest aggregate landings reported for Canada was justover nine million pounds in 1914 (Kendall 1926). A report from the U. S.Bureau of Fisheries in 1920 noted that while the Maine smelt fishery had de-clined in the early 1900s, the New Brunswick fisheries had undergone "remark-able" growth to support the market demands in the U.S. (USDOC 1920). TheMiramichi River in New Brunswick was long a center of the province's smeltfishery.

Shipments of smelt to U. S. markets from the Miramichi River regionexceeded 4.3 million lbs for the winter fishery in 1924 (Kendall 1926), makingthe fishery one of the most valuable industries in the Province at that time.Recent TrendsNew Brunswick and Nova Scotia continue to support important commer-cial fisheries.

There is less evidence of population declines in these provinces than in the U. S. portion of the range. The capitalization of a Great Lakes fish-ery for smelt in the 1960s and 1970s resulted in high landings that suppressed prices and may have reduced effort in the New Brunswick fishery (McKenzie 1964, DFO 2011). In spite of depressed prices, the eastern New Brunswick smelt fishery remained stable between 1988 and1998, with total reportedlandings between 1.5 and 2.5 million lbs, a sum that may under represent actual landings (DFO 2011).The smelt fisheries of the St. Lawrence River have shown a decline com-parable to U. S. fisheries.

Reduced commercial and recreational fisheries andspawning habitat abandonment in the St. Lawrence River tributaries triggered survey and restoration efforts in the 2000s (Trencia et al. 2005). The fisheries remain culturally important today while operating at historically low harvest18 ° ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN levels with ongoing restoration efforts by Quebec's Ministry of NaturalResources (Verreault et al. 2012).SummaryDramatic changes have occurred in both Gulf of Maine smelt fisheries andthe distribution of smelt on the East Coast since the start of the 20th century.Culturally and economically important smelt fisheries have disappeared orfaded to historic lows. The trend is evident of wide-scale abandonment of thehistoric southern extent of the range, where commercial smelt fisheries wereviable before the 20th century.

Currently, the southern extent of the speciesrange is likely in the Buzzards Bay, Massachusetts region, with higher popula-tion levels observed in more northern rivers.Popular recreational fisheries remain in Maine and New Hampshire, although these fisheries also appear to be harvesting at historically low levels.The traditional Massachusetts ice shack fisheries have been reduced to verylow levels of participation and catch, and they are faced with warmer wintersthat bring insufficient ice to support shacks. The causes of this steep declinein smelt fisheries on the U. S. East Coast have not been defined, but have beendiscussed for over a century.

Industrial pollution at spawning rivers, structural

barriers, and overfishing have received the most attention as causal factors.Watershed alterations, natural predation and climate change are potential fac-tors that have been implicated more recently.

Figure 1.2.1. Commercial smeltCommercial harvest of rainbow smelt in ME and NH landings for Maine (1887-2009) and New Hampshire (1950-2009).

1400 Data sources:

U.S. Commissioners Report, U.S. Bureau of Fisheries, 1200 State of Maine landings data (as~1. summarized by Squlers et al.Zl 1976), and NMFS website.~.800 vI~ -MEC600-Hc 4000Z200I (,) 0) U) -I- C' 0) UZ) -t- C') 0) UZ) ,. C"- C') 0) U) ,-rrrrr rrrrr rrrrN NYearANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN

• 19I Studies conducted inthe late 1950's described several life history char-acteristics that we alsoobserved in the presentstudy, such as declining average length as thespawning run progresses and few smelt over theage of three.1.3 -POPULATION STATUS IN THE GULF OF MAINEConcerns have grown over the health of anadromous rainbow smelt popu-lations throughout much of their range. This concern has prompted interestin assessing smelt populations and developing restoration strategies.

Limitedinformation is available from both fisheries-dependent and independent sourceson the present status of populations in New England.

The Species of Concern(SOC) project reviewed existing smelt population data in New England toconsider the potential for developing indices of abundance, and initiated fieldprojects during 2008-2011 to establish new data series to provide information on the status of smelt runs.Previous Smelt Population StudiesThe earliest smelt population studies occurred in northern portions of theirrange, likely in response to the commercial importance of smelt fisheries inthese regions.

Kendall (1926) focused on smelt fisheries but did provide smeltlength data gathered from various sources during the 1850s to 1920s. Notmuch information can be gleaned from these sparse data, except to say the max-imum size of smelt from that time period of about 26-28 cm (total length) isquite similar to the maximum size found in the present study (27 cm). Warfelet al. (1943) reported smelt age data for Great Bay, NH; this study providedsome of the first age data for the area and perhaps the first documentation ofage-I smelt participating in the spawning run. Summary statistics for Warfel etal. (1943) and the following studies are presented in Table 1.3.1.McKenzie (1958 and 1964) followed the Great Bay study with a detailedstudy of the life history of smelt and their fisheries in the Miramichi River ofNew Brunswick during 1949-1953.

McKenzie (1964) demonstrated severallife history characteristics that have been confirmed in the present study, suchas: declining average length of smelt as the run progresses, a more balanced sexratio in the winter fishery than during the spawning run and few smelt olderthan age-4.The age composition in the Miramichi River during 1949-1953 had consistently higher representation of age-3 (22-49% annually) and age-4 (2-8% annually) than seen in the present study and had older fish presenteach year, although at low proportions (age-5 and age-6 at <0.5% and <0.1%,respectively).

Murawski and Cole (1978) calculated an annual survival rate (S)of 0.35 for the overall proportions in McKenzie's age composition data, a valuefound to be the highest among reported survival data for anadromous rainbowsmelt (Chase et al. 2012).The ME DMR devoted considerable time to the assessment of smelt fisher-ies in the 1970s and 1980s (Squiers et al. 1976, Flagg 1983). The majority ofthe effort was fishery-dependent assessments of the winter smelt fishery.

Thesize composition data from these winter fishery studies may not be directlycomparable to spawning run size composition.

However, summary data onsampling proportion by age and mean length at age are included in Table 1.3.1because the data document the size composition of smelt populations at thetime and the relatively larger contribution of older smelt in the catch.Murawski and Cole (1978) provided size, age and mortality data from theParker River, Massachusetts spawning run and winter fishery during 1974-1975. This study sampled both the winter sport fishery catch and spring20

• ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN spawning run with a fyke net, providing a valuable comparison to the ParkerRiver data in the present study. Five age classes were represented in the fykecatches, with a majority at age-2. Murawski and Cole (1978) also providedone of the few estimates of smelt population mortality and survival rates. Theyreported mean values of the annual survival rate (S) of 0.28 and the instanta-neous total mortality (Z) of 1.27 for both sexes using three analysis methodsfor the spawning runs. They considered the estimated overall annual mortality rate of 72% of the adult population to be high and that increases in fishingpressure could limit reproductive success in the Parker River.Lawton et al. (1990) investigated biological aspects of the Jones River(MA) smelt spawning run during 1979-1981.

The study used a lift net at theupstream limit of smelt spawning habitat to collect mature smelt. All biologi-cal data collected by the lift net may not be directly comparable to the presentstudy, wherein a fyke net was deployed downstream of the lowermost spawninghabitat.

However, the study did produce an age/length key based on length-stratified age subsamples that should be representative of the spawning rundemographics and comparable to the fyke net age/length data. Five age classeswere found in the Jones River with an age-2 majority for most years and veryfew age-5 smelt. For the three spawning seasons sampled, age-2 and age-3 smeltcomprised 83-99% of the spawning smelt. Lawton et al. (1990) also estimated the Jones River spawning population by extrapolating smelt egg densities to to-tal spawning habitat area. The spawning stock abundance model calculated thespawning run of 1981 to exceed four million adult smelt. They also reportedevidence of a strong 1978 year class with relative contributions of this cohortevident in the subsequent three spawning runs.The smelt runs of the St. Lawrence River have supported culturally and ec-onomically important fisheries in Qu6bec for decades.

Declining smelt fisher-ies landings attracted the interest of the Qu~bec Ministry of Natural Resources to conduct biological monitoring in the 1990s. Pouliot (2002) reported on sizeand age sampling of the spawning run in a St. Lawrence River tributary, theFouquette River, during 1991-1996.

A standardized dipnet sampling methodwas used at night at the spawning habitat.

The results provide the first detailedpopulation demographics and mortality estimates for smelt in the St. LawrenceRiver watershed.

The Fouquette River smelt runs during the 1990s contained four or five cohorts in most years. Estimates of the annual rate of total mortal-ity were 74% for females and 73% for males.Current Fisheries Dependent Monitoring New Hampshire Creel SurveyNHF&G has conducted winter creel surveys since 1978. The surveyoccurs from ice in to ice out, generally between the months of Decemberand March. Four locations are sampled:

the Lamprey, Oyster/Bellamy andSquamscott rivers as well as Great Bay. From 1983-1986 no survey was con-ducted due to lack of funding, and in 2002 and 2006 fishing, and subsequently

surveys, were not possible due to lack of ice cover.Biologists interview all anglers (or a sub-sample when large groups of an-glers are present) for catch and effort information during a two hour survey pe-riod per day, visiting locations on a rotating basis. The information collected isANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN

• 21 expanded to provide estimates of catch, effort and catch per unit effort (CPUE)by month and location.

Biological information from the smelt catch, including length, sex and scales for ageing, are taken from 150 fish weekly.The average CPUE for 1987-2011 is 4.48 fish/hour over the entire sampleperiod. High CPUEs have not been observed in the last ten year period (2000-2011, max CPUE 5.6), compared to the previous twenty year period (1980-1989 max CPUE = 10.3; 1990-1999 max CPUE = 10.6; Figure 1.3.1). Inmost recent years, the CPUE has been below the series average (4.48) until2011 when it increased to 5 fish/angler hour. There has not been a peak inCPUE over 6 fish/angler hour since 1995. The CPUE shows large inter-annual variability, and seems to follow a 5-10 cyclical pattern (Figure 1.3.1).Maine Creel SurveyAdopting sampling methods currently used by NHF&G (Sullivan 2009)and methods used in a 1979-1982 study conducted by the ME DMR (Flagg1983), ME DMR again began conducting creel surveys in 2009 in the Ken-nebec River and Merrymeeting Bay area. As part of this survey, ME DMR staffvisited participating camps two or three times per week on a rotating basisto collect biological information about the recreational catch. Staff collected biological information from a subset of each angler's catch (up to 100 fish perangler),

including length, sex, scale samples for ageing and fin clip samples forgenetic sampling.

The number of anglers, fishing hours, and the number offishing lines used was also recorded.

CPUE was calculated as the total number of smelt caught per line-hour of fishing, as opposed to NHF&G calculation of CPUE as smelt caught perangler hour -ME DMR currently calculates CPUE using line-hour to remainconsistent with surveys conducted by ME DMR 1979-1982.

The recentsurvey found a slightly lower CPUE (0.48), compared to the 1979-1982 studyCPUE (0.64), however inter-annual variability was significantly larger than thecomparison between the two study periods (Figure 1.3.2, Flagg 1983). Whileannual fluctuations in CPUE occurred in both surveys, the recent survey hadthe lowest CPUE recorded (0.17) during the two time series.Catch Card boxes were also posted at each camp for fishermen to voluntari-ly report information about their total smelt catch and any bycatch; responses varied widely between sites and between years. There were 122 responses in2009, 6 in 2010, and 37 in 2011 for all camps combined.

It is our hope thatwith continued interaction with anglers and camp owners that the numberof responses will increase.

Despite the low number of responses in 2010, theCatch Cards still reflected the catch patterns found in creel survey data.Current Fisheries Independent Monitoring State Inshore Trawl SurveysThe three state fisheries agencies perform inshore small-mesh trawl sur-veys twice a year, in the spring (MA DMF in May, NH/ME in late May, earlyJune) and fall (MA DMF in September, NH/ME in October, early November).

The MA DMF has been performing surveys since 1978, while the ME DMRbegan sampling the New Hampshire and Maine waters in fall 2000. Thesesurveys provide information about marine habitat use and migration patterns22 o ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN of rainbow smelt, as discussed in section 1.1 -Basic Biology.

However, thissurvey is designed to monitor groundfish abundance, and has limited applica-tion for pelagic species like rainbow smelt. The data are helpful in determining the presence or absence of smelt in certain regions and depths, and can give apicture of inter-annual age cohort strength from size data, but are not powerfulin showing trends in rainbow smelt abundance.

However, trends in catches inboth state surveys seem to have a 5-10 year cyclical pattern similar to the creelsurveys and juvenile abundance surveys (Figure 1.3.3), although the causalfactor behind these cycles is unknown.Maine and New Hampshire Juvenile Abundance SurveysIn 1979, ME DMR established the Juvenile Alosine Survey for the Kenne-bec/Androscoggin estuary to monitor the abundance of juvenile alosines at 14permanent sampling sites, sampled June through November.

Four sites are onthe upper Kennebec River, three on the Androscoggin River, four on Merry-meeting Bay, one each on the Cathance, Abadagasset, and Eastern rivers. Thesesites are in the tidal freshwater portion of the estuary.

Since 1994, ME DMRadded six additional sites in the lower salinity-stratified portion of the Ken-nebec River. The seine is made of 6.35 mm stretch mesh nylon, measures 17 mlong and 1.8 m deep with a 1.8 m x 1.8 m bag at its center. The net samples anarea of approximately 220 m2.Of all the river sections, the lower Kennebec catches considerably morejuvenile smelt than all upstream sections; the average catch over the time periodfor the lower Kennebec was 92 smelt/haul/year, while all others were under 10smelt/haul/year, and catches are sporadic.

Though the highest average annualcatch occurred in 2005 (316 smelt/haul) in the lower Kennebec, juvenile smeltabundance in this river segment has been low since 2007, with three of the fourlowest average annual catches occurring in the past four years. Trends in abun-dance also seem to follow a 5-10 cyclical pattern similar to the other surveys(Figure 1.3.4).The NHF&G has conducted an annual Juvenile Abundance Survey since1997. It is designed as a fixed station survey, as opposed to a stratified randomsurvey, because strong tidal currents, rocky shorelines, and various anthropo-genic structures limit the amount of suitable beach seining locations.

A total of15 fixed locations are sampled monthly from June to November.

The stationsare located within the Great Bay and Hampton-Seabrook estuaries.

Seine haulsare conducted by boat using a 30.5 m long by 1.8 m high bag seine with 6.4mm mesh deployed 10 -15 m from the shore. Over the sampling period, thePiscataqua River has seen the highest CPUE (177 smelt/haul/year),

howeverthe highest annual average catch occurred in Great Bay in 2001 (225 smeltper haul). The lowest average catch over the entire sampling period was in theHampton Beach/Seabrook area (11 smelt/haul/year).

While these abundance data also seem to follow a cyclical

pattern, there has been a decline in the juve-nile rainbow smelt being captured in recent years -excluding the first year ofsampling, the four lowest average annual catches have occurred within the past6 years (Figure 1.3.5).New Hampshire Egg Deposition Monitoring New Hampshire Fish and Game Department conducted egg deposition sampling from 1978-2007 using methodologies described by Rupp (1965). AANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN

• 23 ring of known area (20.3 cm2) was tossed on natural substrate, and the numberof eggs within the ring was counted.

Egg counts were conducted weekly, frommid-March to mid-April, in the Oyster, Bellamy,

Lamprey, Squamscott andWinnicut rivers. The mean number of eggs per square centimeter was used asan index of spawning stock abundance.

Validation of the index was attempted by regressing the index with catch per unit effort (CPUE) of the creel surveybut showed very poor correlation.

The egg deposition sampling was discon-tinued in 2008 because comparisons between this dataset and other indices ofsmelt abundance (creel and juvenile surveys) did not correlate well, while trendsin the other surveys did correlate well with each other.Maine Spawning Stream Use Monitoring In 2005 and 2007-2009, biologists with the ME DMR worked with theMaine Marine Patrol to document coastal rivers and streams currently beingused by rainbow smelt for spawning.

The survey collected information aboutthe spawning habitat (substrate, possible obstructions),

and the strength of therun as characterized by the density of egg mats or number of spawning adultspresent.

We compared the current use and strength of runs to information collected by ME DMR in the early 1970s (Flagg 1974) and to information compiled in 1984 by the U. S. Fish and Wildlife Service (USFWS 2012).Of the 279 streams surveyed

, the majority either supported smaller runsthan they did historically or no longer support spawning, while only a smallpercentage (19%) seem to currently support strong runs (Table 1.3.2, Figure1.3.6). Spawning decline was concentrated in southern Maine, lower CascoBay, the Kennebec River, and the east side of Frenchman's Bay. Spawning runsremain strong in northern Casco Bay, the Medomak, St. Georges, and Penob-scot Rivers, and around Pleasant Bay and Cobscook Bay.Regional Fyke Net SamplingEarlier research on anadromous smelt populations in the Gulf of Maine hasprimarily consisted of short-term efforts that monitor smelt size and agestructure during spawning runs. These efforts have not produced long-term population indices of abundance for smelt, and presently, no indices exist inNew England.

The smelt SOC project targeted the spring spawning runs asa source of information on population status. The objective was to producefishery-independent indices of abundance, with the understanding that onlymature smelt participate in the spawning runs. The approach was to recordbiological data from spawning runs; to conduct analyses on size and age com-position, catch-per-unit-effort, and mortality; and to make comparisons aspossible among rivers and to previous studies.Establishing Gulf of Maine Spawning Site IndicesMethods.

As part of this project, fyke net stations were selected at coastalrivers in Maine, New Hampshire, and Massachusetts for monitoring during2008-2011 (Figure 1.3.7, Table 2.1.1). The stations were chosen for suitability to maintain a fyke net in a known smelt run and to represent a range of runsizes and watershed conditions.

The fyke net was set at mid-channel in theintertidal zone below the downstream limit of smelt egg deposition.

The fykenet opening faced downstream, and nets were hauled after overnight sets. Thisapproach was adopted to intercept the spawning movements of smelt that occur24

• ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN at night during the flood tide. Fyke net catches were assumed to be representa-tive of the size and sex composition of the spawning run. With each haul, smeltwere counted, sexed, measured (total length) and released.

Scales were sampledweekly at some stations for ageing.After pilot deployments in 2007-2008 to identify suitable

stations, eightfyke net stations were monitored in Massachusetts, three stations in NewHampshire and six in Maine (Figure 1.3.7). The sampling period inMassachusetts targeted 11 weeks from the first week of March to the thirdweek of May to cover the known smelt spawning period. The sampling dura-tion in New Hampshire and Maine varied due to a later ice-out and spawningseason that occurs later with increasing latitude.

2008-2011 ResultsSmelt were captured at most fyke stations during the spring spawning runsof 2008-2011.

The annual catches ranged from few individual smelt in somerivers to several thousand in the larger smelt runs. The following sections andgraphics describe major findings in the fyke net catch data that portray popula-tion trends across the species' distribution on the Gulf of Maine coast.Seasonality.

Because smelt migrate from marine to freshwater habitats tospawn during the spring freshet, they are affected by a range of environmental factors most related to temperature and precipitation.

Understanding how anunpredictable environment can influence the timing, location and strength of asmelt run is valuable for managing smelt populations.

Accordingly, characteris-tics of the onset, peak, and overall duration of a smelt run can provide measuresof population health. In this study, the onset and ending of the spawning runwere based on the average date of first and last capture, respectively.

Spawn-ing run peak was determined based on the average date of maximum catch. Inseveral cases, the onset and the ending of the spawning run were inconclusive and had to be estimated using best professional judgment.

Run duration wasdetermined based on the average yearly duration of the run from 2008-2011.

Because smelt migratefrom marine to fresh-water habitats to spawnduring the spring freshet,they are affected by arange of environmental factors most related totemperature andprecipitation, Inter-system variability was noted in the timing of the spawning run(Figure 1.3.8). Within most systems in Massachusetts and New Hampshire, thespawning run had begun by mid-March.

Within several Maine systems, how-ever, the spawning run was delayed and did not start until late-April.

Similarpatterns were observed in the peak and ending, with Massachusetts and NewHampshire systems having earlier peaks and earlier ending dates than those inMaine. Differences in run timing among states are presumably attributable toregional differences in climate, with cooler, more northerly systems displaying adelayed spawning run.Run duration also varied with location.

The longest run durations wereobserved for the Fore and Jones rivers, Massachusetts, and Tannery Brook,Maine. In these systems, average run duration appeared to exceed 70 days.Conversely, the shortest runs were observed to occur in the North, Weweantic, and Saugus rivers, Massachusetts, where average run duration did not exceed40 days. The causes for the differences in run duration are unknown, par-ticularly because previous studies have demonstrated shorter run durations innorthern latitudes, with runs in individual tributaries often lasting less thantwo weeks in New Brunswick (McKenzie 1964) and Quebec (Pouliot 2002).In the case of the U. S. Gulf of Maine surveys, population abundance and yearANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN

• 25 class strength may be influential, however the causal factors are not currently understood.

Catch Per Unit Effort (CPUE). The number of fish captured per a givenamount of sampling, known as catch per unit effort (CPUE), is a measureused by fishery scientists to assess the relative abundance of a fish population, under the assumption that higher catches for a given amount of sampling effort(e.g., time, gear, habitat area, samplers) represents greater abundance.

For thefyke net survey, the number of smelt caught per haul was used as a measure ofCPUE. Yearly measures of CPUE were based on the geometric mean of weeklyaverage CPUE.The results of this study demonstrated that CPUE varied widely amongrivers and years. For the entire region, the two highest overall CPUE werefound in Maine (Deer Meadow Brook = 58.07, Schoppee Brook = 37.83),while the two lowest were found in Massachusetts (Westport River = 1.01,North River = 1.37). There was an overall trend of higher CPUE in Mainecompared to New Hampshire and Massachusetts

-out of the 17 index sites,four out of the top five highest CPUE were found in Maine (Table 1.3.3).Considering abundance by state, in Massachusetts, the Fore River had thehighest overall CPUE (20.42),

while the Westport River had the lowest (1.01).In New Hampshire, the highest overall value was found at the Oyster River(5.62), while the lowest was at the Winnicut River (1.64). In Maine, thehighest was found at Deer Meadow Brook (58.07),

and the lowest at LongCreek (11.39, Table 1.3.3).Yearly CPUE peaked in five of eight Massachusetts rivers in 2008, suggest-ing that in these systems, the largest smelt runs were observed at the beginning of the study (Table 1.3.3, Figure A. 1.1). In New Hampshire, the highest annualCPUE for all rivers was seen in 2011 (Table 1.3.3, Figure A. 1.2). In southernand midcoast Maine (Long Creek, Mast Landing, and Deer Meadow Brook),the highest annual CPUE was seen in 2008 or 2009, while in eastern Maine(Tannery,

Schoppee, and East Bay brooks),

the highest annual geometric meanvalues were seen in 2010 (Table 1.3.3, Figure A. 1.3). It should be noted thatwhen CPUE is calculated as simply the number of smelt per haul, the highestCPUE for East Bay Brook occurred in 2008 (Figure A. 1.3).At this time, high levels of variability in CPUE and the limited duration ofthe study prohibit a statistical analysis of trends in relative abundance.

How-ever, the CPUE data from 2008-2011 for some stations should be valuable as areference point for future comparisons.

Length and Age Composition.

Length and age information yieldsimportant insights into the health of a fish population.

As a general rule, thepresence of a variety of age classes is indicative of a healthy population.

Further,populations containing older and larger individuals, which have a relatively high reproductive potential, are considered healthier than those containing onlyyounger, smaller individuals.

Smelt are fast growing fish that mature at smallsize and become fully recruited to the spawning stock at age-2 in the study area.We measured total length of captured smelt to the nearest millimeter (mm).Smelt ages were determined from scale samples.The age class composition of the runs varied between sites, but displayed 26 -ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN geographical patterns.

We found that runs in the southern portion of the Gulfof Maine (represented by the Fore River, Massachusetts, and Mast Landing,Maine) displayed two dominant age modes: one comprised mainly of age-Ismelt and second mode comprised of mainly age-2 smelt (Figure 1.3.9 and1.3.10).

Age-1 smelt were common in Massachusetts and, in some years, werethe dominant age class; yet this age class was present at much lower frequencies in spawning runs in the northern range of the study area (Table 1.3.4, Figures1.3.9-1.3.14).

In the mid-portion of the region (represented by Deer Meadowand Tannery brooks, Maine), age- I fish were encountered infrequently

-theruns instead were dominated by age-2 fish, and the frequency of age-3 individ-uals was much higher than seen in more southern runs. Older ages (4-5) werealso seen in these runs at higher rates than at all other runs, and these were theonly sites to have age-6 fish represented in the runs (Table 1.3.4, Figures 1.3.11and 1.3.12).

In the northeastern portion of the Gulf of Maine (represented bySchoppee and East Bay brooks),

runs were composed primarily of age-2 fish,with few to no age-I fish observed.

Age-3 fish were observed, but at a lowerfrequency than the mid-portion of the region. The occurrence of older ages(4-5) was higher than the southern runs, but not as high as the mid-portion (Table 1.3.4, Figures 1.3.13 and 1.3.14).

The fact that fish at age-4 or olderwere unusual in Massachusetts, but relatively common in Maine samples,suggests higher levels of mortality in southerly systems.Length at age also varied between sites, but again showed a geographic pattern.

Because large sample sizes of age-2 males were present in each run, itis informative to compare the average lengths between sites using this category.

The largest length at age was observed in the southern portion of the region(Fore River avg. age-2 male = 184 mm, Mast Landing = 178 mm, Table 1.3.4),indicating a faster growth rate at lower latitudes.

Though the Oyster Rivergeographically lies between these sites, age-2 males were comparatively smallerthan the other southern sites (162 mm). This smaller age-at-length comparedto surrounding sites may be evidence of a stressed population in the OysterRiver, although further evidence would be needed to substantiate this idea.Comparisons between previous studies show that length-at-age is observed todecline moving northward (Table 1.3.1). We observed a similar trend, how-ever the smallest length-at-age was observed in the mid-portion of the region(Deer Meadow Brook avg. age-2 male = 157 mm, Tannery Brook = 142 mm,Table 1.3.4). Sites at the most northeastern portion of the Gulf of Maine hadlarger age-2 males than in this mid-portion, but smaller than the southern sites(Schoppee Brook = 163 mm, East Bay Brook = 166 mm, Table 1.3.4). Thispattern in age-at-length, as well as the pattern in run compositions discussed above, is coincident with the genetic stock structure of rainbow smelt reportedby Kovach et al. (in press) and discussed in section 1.1 -Basic Biology, whichfound that the fish from Tannery Brook had a genetically differentiated signalthat was also seen in fish from Deer Meadow Brook, but not in any other sites.Because it was not possible to develop age-at-length keys for all sites dueto low sample numbers at some sites, median length (calculated from all fishat a site) and length distributions are useful in understanding region-wide trends. Median lengths were lowest for males in the Massachusetts sites, andfor females in the New Hampshire sites, and were generally higher for Mainesites (Table 1.3.5, Figure 1.3.15).

The driving factor behind these patternsANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN -27 seems to be the age composition of each of these runs rather than the length atage -runs in the southern portion of region are composed of a large proportion of age-1 fish, while runs in the mid- and northeastern portion have a higherproportion of age 3+ fish (Table 1.3.4). While not all fish were aged, modescorresponding to specific ages can help in affirming this idea. Length frequency figures for all sites with enough samples to produce relevant figures are includedin the Appendix (Figures A. 1.4 -A. 1.14).Sex Ratio. Although spawning runs of most anadromous fishes are malebiased, those displaying a substantially higher proportion of males may beindicative of a stressed population.

Because the limiting factor for population growth is often the abundance of females, populations dominated by malesmay be less robust than those containing a higher proportion of females.

In thisstudy, sex ratio was determined based on the ratio of the aggregate 2008-2011 catch of males to the catch of females.The results of the fyke net survey demonstrated that each system contained a smelt population that was male biased. Overall, this survey observed anaverage sex ratio of 4:1. Of the systems sampled, the most heavily male biasedwere the Parker River, MA, and the Squamscott and Oyster rivers, NH, whichall displayed a male to female ratio of greater than 8:1. The lowest male tofemale ratios were found in three systems in Maine: Tannery Brook, SchoppeeBrook, and the East Bay River. In each of these systems, the sex ratio was lessthan 2: 1. We acknowledge that these sex ratios are biased themselves due tothe behavior of male smelt spending more time on the spawning grounds thanfemales (Murawski et al. 1980).Mortality.

Limited work has been done on population metrics for anadro-mous rainbow smelt throughout their range, but a few studies have calculated population mortality and survival rates based on age structure (Murawski andCole 1978, Pouliot 2002). Survival and mortality analyses have potential biasesthat may limit their accuracy.

Few age cohorts are available for the assessment:

the age-I cohort is excluded from mortality estimates because they are partially recruited to the spawning run, and age-4 smelt are presently uncommon.

Sec-ondly, the sampling method cannot distinguish the occurrence of repeat spawn-ing movements of individual smelt; this behavior could bias measurements ofmortality and survival.

Under the assumption that these biases were consistent among studies, we calculated mortality and survival estimates for sites that hadsufficient age data for 2008-2011 and compared them to previous studies.Within the study area, survival rates (S) and instantaneous total mortal-ity (Z) were calculated using the Chapman and Robson equation (Chapmanand Robson 1960) at five stations in Maine and one each in Massachusetts andNew Hampshire.

However, the presence of some small sample sizes, few yearsof observations and the above discussed biases limit the reporting of these datato a relative comparison across the region and to past studies.

Tannery Brook,Maine, had the highest average survival for 2008-2011 at S = 0.33, followed byS = 0.26 for 2009-2011 at Deer Meadow Brook in Maine. For sites that had atleast three years of data, the Fore River, Massachusetts, had the lowest averagesurvival at S = 0.17. The range of these spawning population survival estimates places the higher values in the present study among the highest reported byprevious studies in the U.S. (Murawski and Cole 1978, Lawton et al. 1990) and28 ° ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN Canadian Provinces (McKenzie 1964, Pouliot 2002), and the sites at the lowerrange are the lowest survival values reported for anadromous rainbow smelt.Study Area SummaryMassachusetts.

Of the eight fyke net stations monitored in Massachusetts, six caught enough smelt to allow summary comments on run demographics, but only the Fore River had a sufficient sample size to generate age composition data each year. The age and length data in Massachusetts suggest the pres-ence of a truncated age distribution, a sign of stressed populations due to highmortality and potentially poor recruitment.

Male smelt in Massachusetts havesimilar median lengths compared to male smelt in New Hampshire and Maine.However, female smelt in Massachusetts had higher median length than theother states; a statistic driven by larger age-2 to age-4 females.

Massachusetts stations are dominated by length modes that indicate age-I and age-2 smelt,with very low presence of smelt older than age-4.The proportion of age-1smelt in Parker River and Jones River spawning runs markedly exceeds thatfound in previous studies.

Changes in the contribution of age-1 smelt to thespawning run between previous studies and the present study, and the higherproportion of these small smelt in Massachusetts compared to New Hampshire and Maine raises interesting questions on the significance of these apparentdifferences.

Smelt at the southern stations may experience faster growth intheir first year and are reaching a body size that supports maturity sooner thannorthern runs.New Hampshire.

The presence of mature smelt was documented in fykecatches in the Bellamy, Salmon Falls, Lamprey, Squamscott, Winnicut andOyster rivers during 2008, and the standardized fyke net sampling protocolwas followed in the Squamscott and Winnicut rivers from 2008-2011, andin the Oyster River from 2010-2011.

Sufficient age samples were collected at the Oyster and Squamscott rivers in 2011 to prepare length frequency andage-length graphs. Two length modes are apparent in both rivers composed ofage-I and age-2 smelt. However, more overlap is seen in these modes than isfound in Massachusetts smelt age-length data. Few smelt reached age-4 in NewHampshire rivers. For each available age key, age-4 comprised less than 2% ofthe annual age sample. Growth rates appear to be slower within New Hamp-shire runs, as age-3 smelt occur at smaller lengths than seen in Massachusetts and no age-2 smelt larger than 19 cm have been sampled.Maine. All six Maine fyke net stations produced sample sizes large enoughto summarize information on smelt run status. Median smelt length for theMaine stations was slightly larger than at the other states because these runshad a lower proportion of age-1 smelt, but higher proportion of age 3+ smelt;however, average length at age was smaller, indicating a slower growth ratecompared to sites further south. The Maine smelt runs also averaged higherCPUE rates and showed more balanced age distributions and sex ratios thanseen in southern runs. These patterns were most evident in catch data from theeasternmost Maine stations.

All these observations indicate relatively healthier smelt runs in Maine than in Massachusetts and New Hampshire.

The agecomposition of smelt in Maine's spawning runs contributes to less separation between length modes and an extended age-2+ mode. These features couldreflect interesting potential differences in growth rates, maturation, and survivalThe age and length datain Massachusetts suggestthe presence of a trun-cated age distribution, a sign of stressedpopulations due to highmortality and potentially poor recruitment.

ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN

• 29 in Maine than at the southern runs.Conclusions About Regional Fyke Net SamplingA common goal in fisheries management is to base decisions on a long-Table 1.3.1. Mean length at age term stock assessment that generates defensible biological benchmarks on theand proportion at age of anad- health of the fish stock. The present study does not achieve this goal, but itromous rainbow smelt sampled starts the process of providing information on spawning run CPUE, temporalduring spawn/ng runs In earlier characteristics, and size and age composition of rainbow smelt in three states.studies in the study area andCanadian Maritime Provinces.

The sampling period from 2008-2011 is too brief for conclusions onAft length data were converted topopulation trends. However, such baseline information is vital for all fish stocktotal length.assessments.

The task of assessing the status of rainbow smelt in the Gulf ofMean Length at AgeLocation Region Citation Year Sex N Age-1 Age-Z Age-3 Age-4 Age-S Age-6Miramichi River NB McKenzie (1958) 1949-53 M NA 157 178 196 211Miramichi River NB F NA 162 186 212 238Fouquette River Quebec Pouliot (2002) 1991-96 M NA 133 166 198 215 227Fouquette River Quebec F NA 135 173 213 237 245Great Bay NH Warfel et al. (1943) 1942 both 287 86 145 171 220 264Penobscott River ME Squiers et al (1976) 1974-75 both 260 165 196 226 264Kennebec River ME Flagg (1984) 1980-82 M 1012 174 202 221 229Kennebec River ME F 680 180 215 239 249Parker River MA Murawski 1974-75 M 2097 141 188 208 236 242Parker River MA and Cole (1978) F 584 140 197 219 245 249Jones River MA Lawton et al. (1990) 1979-81 M 31394 132 184 208 221 242Jones River MA F 5009 130 190 222 244 254Proportion

(%) at AgeLocation Region Citation Year Sex N Age-i Age-2 Age-3 Age-4 Age-5 Age-6Miramichi River NB McKenzie (1964) 1949-53 both NA 66,2 29.3 4.1 0.4Great Bay NH Warfel et al. (1943) 1942 both 287 3.5 65.9 29.6 1.0 0Kennebec River ME Flagg (1984) 1979-82 both 1700 59.9 33.0 5.5 0.5Penobscott River ME Squiers et al (1976) 1974 both 133 42.1 39.1 17.3 1.5Penobscott River ME 1975 both 127 17.3 67.7 14.2 0.8Parker River MA Murawski 1974 M 343 38.0 42.5 15.9 3.2 0.4Parker River MA and Cole (1978) 1974 F 50 15.7 50.5 20.8 10.8 2.2Parker River MA 1975 M 113 9.9 81.2 7.9 0.8 0.2Parker River MA 1975 F 40 3.9 76.6 16.4 2.4 0.7Jones River MA Lawton et al. (1990) 1979 M 364 15.0 64.6 19.7 0.7 <0.1Jones River MA 1979 F 235 15.1 66.7 16.7 1.0 0.5Jones River MA 1980 M 428 0.2 88.4 11.1 0.3 0Jones River MA 1980 F 353 0 86.0 12.8 1.2 0Jones River MA 1981 M 250 2.9 55.7 37.9 3.5 0Jones River MA 1981 F 160 0.4 36.0 48.7 14 0.9Notes1. Lawton et al. (1990) and Murawski and Cole (1978) age composition is based on age key proportions assigned to total length sample.2. Murawski and Cole (1978) mean length combines 1975 winter creel survey with 1974 and 1975 spawning run data.3. McKenzie (195S and 1964) length data are converted to TL from SL Age-6 smelt were caught in most years at low frequency

(<0.1%).4. Pouliet (2002) fork length data were converted to total length using the conversion, TL = (FL-0.5584)/0.9142, from Chase et al. (20061.5. Flagg (1983) and Squiers et al (1976) size and age data are both from winter smelt ice fishery, but included for comparative value.30 , ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN Maine is further complicated by the case of having distinct stock structure forsome rivers, instead of a coast-wide stock complex.

Finally, the assessment ofanadromous fish is confounded by their migration between marine andfreshwater

habitats, where different factors influence their growth and survival.

Despite these challenges, the fyke net data from the present study show a gradi-ent of conditions with signs of stressed populations in southern Gulf of Maineand less evidence of stress moving north along the Maine coast, as evidenced byyounger age distributions, smaller age-at-length, and lower CPUE rates.Status N PercentNot historically listed, and currently do not support spawning 42 15%Historical runs that do not currently support spawning 35 13%Currently support smaller runs than historically 95 34%Currently support strong runs 53 19%Historical runs that were not visited, current status Is unknown 54 19%Annual CPUE OverallRim 5tQ 2008 2009 2010 2011Weweantlc R. MA 2.81 1.27 1.47 1.57 1.78Westport R. MA 1.00 1.00 1.00 1.02 1.01Jones R. MA 9.13 5.58 7.56 5.13 6.85Fore R. MA 33.55 10.41 22.00 15.70 20.42Saugus R. MA 6.30 1.19 1.07 2.49 2.76North R. MA 1.39 1.12 1.08 1.90 1.37Crane R. MA 3.03 1.97 2.12 3.39 2.63Parker R. MA 7.63 2.56 1.66 2.47 3.58Squamscott R. NH 3.45 1.44 1.08 6.26 3.06Winnicut R. NH 1.60 1.34 1.36 2.25 1.64Oyster R. NH --5.45 5.79 5.62Long Cr. ME -18.69 5.56 9.93 11.39Mast Landing ME 52.00 29.84 8.81 13.80 26.11Deer Meadow Bk. ME 11.11 100.82 24.86 95.46 58.07Tannery Bk. ME 15.28 28.26 41.87 14.03 24.86Schoppee Bk. ME 38.42 37.25 37.83East Bay R. ME 15.48 4.42 21.66 11.86 13.35Table 1.3.2. Current state of smeltspawning runs in Maine with re-spect to their historical status.Table 1.3.3. Catch per unit effort(CPUE) of rainbow smelt at fykenet spawning survey Index sites,by annual CPUE and overall CPUEfor the entire sampling period,2008-2011.

ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN o 31 Proportion

(%) at AgeLocation Region YearSex Length N Age N Age-1 Age-2 Age-3 Age-4 Age-S Age-6East Bay Brook ME 2008East Bay Brook ME 2009East Bay Brook ME 2010East Bay Brook ME 2011bothbothbothboth89923613871211636826126892.2 6.70.8 62.3 33.92.0 80.7 13.772.0 26.73.03.61.25.41.10.1Schoppee Brook ME 2010Schoppee Brook ME 2011both 2034 281 0.9 90.2 3.5both 1831 245 2.2 90.7 7.1Tannery Brook ME 2008Tannery Brook ME 2009Tannery Brook ME 2010Tannery Brook ME 2011Deer Meadow ME 2008Deer Meadow ME 2009Deer Meadow ME 2010Deer Meadow ME 2011Mast Landing ME 2008Mast Landing ME 2009Mast Landing ME 2010Mast Landing ME 2011bothbothbothbothbothbothbothbothbothbothbothboth2001177818929081792016136619461620110635518337472344172851353201089012826827560.0 34.23.9 78.6 7.92.5 49.6 45.46.9 36.6 48.05.0 77.1 17.90 90.2 5.72.8 26.0 64.71.5 83.6 6.915.2 58.6 24.20.6 85.6 13.975.5 8.7 13.844.5 53.5 0.85.86.41.48.53.45.06.72.02.91.71.20.7<0.10.40.60.10.43.21.00.71.50.90.30.1Oyster RiverOyster RiverFore RiverFore RiverFore RiverFore RiverNH 2010 both 421 185 65.8 29.0 4.5NH 2011 both 401 231 11.2 75.1 13.5MA 2008MA 2009MA 2010MA 2011bothbothbothboth19588461441124138066049348651.9 41.4 6.215.5 52.5 31.489.6 7.9 2.448.3 48.7 2.60.1<0.1<0.1Mean Length at AgeLocation Region Year Sex N Age-1 Age-2 Age-3 Age-4 Age-5 Age-6East Bay Brook ME 2008-11 M 322 145 166 197 215 241East Bay Brook ME 2008-11 F 338 155 173 212 238 241Schoppee Brook ME 2010-11 M 225 146 163 195 204Schoppee Brook ME 2010-11 F 299 152 169 206 234Tannery Brook ME 2008-11 M 339 135 142 166 183 190Tannery Brook ME 2008-11 F 322 137 146 178 198 211 215Deer Meadow ME 2008-11 M 397 138 157 185 209 220 226Deer Meadow ME 2008-11 F 250 125 160 194 222 208Mast Landing ME 2008-11 M 447 132 178 192 211Mast Landing ME 2008-11 F 312 137 190 209 232 256Oyster River NH 2008-11 M 344 117 162 179 209Oyster River NH 2008-11 F 60 114 167 180Fore River MA 2008-11 M 1113 141 184 202 215Fore River MA 2008-11 F 507 142 194 217 249 251 266Table L3.4. Mean length at ageand proportion at age of anadro-mous rainbow smelt sampled atfyke net stations for 2008-2011 for the present study. Age keyswere applied to length samplesfor proportion at age.32 9 ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN MALEState River Code I Years N Mean SE Median Min MaxMA Weweantie WWMA Jones JRMA Fore FRMA Saugus SGMA North NRMA Crane CNMA Parker PRNH Squamscott SQNH Oyster OYME Long Creek LCME Mast Landing MLME Deer Meadow DMME Tannery Brook TBME Schoppee SBME East Bay EB444444422444424188 1511249 1564396 166401 16279 150262 1611217 167340 154344 1491191 1693099 1634367 1664214 1522303 1642368 1722.010.930.431.302.181.440.981.851.740.410A00.330.270.240.311451431571531491561561591561681691631521631691041061091131181218686881101058310412513623825424124021722125522722S228227241223222250ITotalME Schoppee SB 2 2303 164 0.24 125 222ME East Bay EB 4 2368 172 0.31 136 250Total 26018I Total26018FEMALEState River Code Sex Ratio N Mean SE Median Min MaxMA Weweantic WW 3.4 55 149 4.29 139 107 225MA Jones JR 2.5 492 160 1.69 144 100 258MA Fore FR 4.0 1090 168 1.06 154 111 270MA Saugus SG 7.7 52 172 5.01 157 129 248MA North NR 3.4 23 154 4.71 153 113 214MA Crane CN 2.8 94 169 3.31 162 114 257MA Parker PR 9.5 128 194 3.18 204 112 272NH Squamscott SQ 3.7 93 135 3.86 118 86 239NH Oyster OY 5.7 60 151 4.80 166 88 224ME Long Creek LC 3.3 360 178 0.99 176 118 251ME Mast Landing ML 2.7 1136 177 0.86 180 93 263ME Deer Meadow DM 3.6 1209 165 0.71 159 83 258ME Tannery Brook TB 1.8 2366 157 0.46 154 108 236ME Schoppee SB 1.5 1564 174 0.53 170 129 256ME East Bay EB 1.7 1389 183 0.59 176 122 263Table 1.3.5. Rainbowsmelt length data fromcatches at fyke netstations, 2008-2011.

A few stations wereexcluded because of lowsample sizes or poten-tially biased samplesfrom few hauls. Smeltof unknown sex wereexcluded from this table.Sex ratio Is the ratio ofmales to females.Total10111ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN e 33 New Hamps108CL-6*11.C1-E 4WELU2shire Creel SurveyIFigure 1.3.1. New Hampshire Fishand Game Creel Survey catch perunit effort (CPUE) calculated asnumber of fish caught per hour offishing 1978-2011 Figure 1.3.2. Catch per unit effort(CPUE) as smelt caught per line-hour of fishing observed duringthe rainbow smelt winter creelsurvey In Maine during 1979-1982 and 2009-2011.

Figure 1.3.3. Inshore Trawl Surveyaverage annual smelt catches (innumbers of fish) from MA DMFstate survey (1978-2011) andME DMR/NHF&G combined statesurvey (2000-2012).

1Ii1ill111111U0.90.80.70.60.50.40.30.20.10Maine Creel Survey200 W. .........1801601401120180 .MEfte400402034 a ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLANI 350300250-200~150S100500W-----------800 17006o00S5003 400<~ 300S200100New Hampshire Juwenile Abundance Surey__ILFigure 1.3.4. Average annualcatch of rainbow smelt YOY In MEDMR Juvenile Abundance SurveyIn the lower Kennebec River.Other sites are excluded due tolow catches.Figure 1.3.5. Average annualcatch of rainbow smelt YOY inNHF&G Juvenile Abundance Sur-vey. The 11 locations within thePiscataqua River and Little/Great Bay were grouped Into two cohortsto show annual trends. The Hamp-ton/Seabrook area was excludeddue to low catches.01 Little and Great BaysN Piscataqua RiwrANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN -35I Figure 1.3.6. Current status ofsmelt spawning runs In Maine andhistorical sites where the currentstatus remains unknown.36

• ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLANI Dee Meadow BrookOys erRiPaFwvrkolM iJanelan River-A Rb~erSchW BwookFigure 1.3.7. Fyke net monitoring stations In Massachusetts, NewHampshire, and Maine2008-2011.

110I,,, I O MFigure 1.3.8. Smelt runs progressin a bell-curve shape over the sea-son, where the beginning of therun sees few smelt and the num-ber steadily Increases to a peak Inthe run (red portion of the bars inthe figure),

after which point therun steadily declines (blue por-tion of the bars). These patternsare shown here, along with theaverage beginning and end dateof each run 2008-201:.

Stationsare arranged from south to northstarting at the x-axis origin.ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN a 37 S00400'o02008o0te-I 8Age,2 A4g-3 m Alel" 5 4 5e-S+Fle0300SOD200I00Age N .380530/ N I1 ilS~2 9 10 13 12 13 14 15 16 17 18 19 2n 21 22 23 24 IS 26 27 28TOWl Legt ml2 14g-I 56I42 l5483 maVe-4 mAge-5+1/F N 1441l 9 10 11 12 13 14 15 16 l 7 35 19 20 21 2 23 24 2S 26 27 28Tota Length (m)4003000200100-2009No.--1F1 --I I I I048-1 A W40-2 IN4o-3 mAge4 IN48-5.8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28Total Length (-I8l:I 2011AgF N: 458L/F N. 8460440-3 m504- 5A4e.3 5440-4 5440ta-S.

44e N -486L/F N = 12412o 00]1000oH] rv.IIll g8 9 00 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28Total le ngth (iM)Figure 1.3.9. Age composition of Fore River, MA, fyke net catchIn 2008-2011.

Both genderswere combined with number ofage samples reported as "Age N"and length frequency sample sizereported as "i/F N ýSon.,40-I m44*2 m403 53404800 2008NOIL/F N. 1705SOD4001 2009200]8 9 10 12 12 13 140 9 9 10 11 12 13 14 15 16II I.-O4-4 5 11e-2 5411 -3 5Ape-4IS 16 17 18 19 20 22 22 23 24 25 26 27 28Tota Leng4t h (1)Total17 18 19 20 01 22 23 24 25 26 27 28length (s)25004A40. 5443-2 ii4e-3 mA504 44m.-S 5"2010AS, N -2680440-1 54402 w440-3 a40. 53401.A-2011201) Ap8e100So,L/F N -3S7No.NO0200Ri R -0..-I ami.A40N 275L/F N .183308 5 10 11 12 13 14 IS 16 17 18 15 20 2121 23 24 25 26 27 2STotal Length1 Bot e )8 9 10 11 12 13 14 16 16 17 18 19 20 21Total Length I-m)22 11 24 IS 26 Z7 22(Figure 1.3.10. Age composition ofMast Landing, ME, ilike net catchIn 2008-2011.

Both genderswere combined with number ofage samples reported as "Age N"and length frequency sample sizereported as "L/F N"ý38 to ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN 2008714so250Ae-I S*Ap2 SIAV3 NAge4Age N: 175400200OWL0- *60o2 3AP-3 MAI- 3440-50A1e N 135L/F N 2016No1ilium08 9 10 11 12 13 14 IS 16 17 10 19 20 21 22 23 24 25 26 27 28Total Length (mn)8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28Totl. Length 1Cm)200GM0o0g4-1 MAg12 3mae-3 mIAe-4 3041-52010A0 N" 320 200L/FN5 -13672008 9 10 01 12 13 14 15 16 17 IS 1 20 221 22 23 24 2S 26 27 20Total Legt (-)DA02-4 3A0e-2 34A4-3 3a4 4 44S201110 21 12 53 14 15 16 17 18 19 20 22 22 23 24 25 2S 27 28Tot." 140 I-0)Figure 1.3.11. Age composition of Deer Meadow Brook, ME, fykenet catch In 2008-2011.

Bothgenders were combined withnumber of age samples reportedas "Age N" and length frequency sample size reported as "L/F N".Figure 1.3.12. Age composition ofTannery Brook, ME, fyke net catchin 2008-2011.

Both genderswere combined with number ofage samples reported as 'Age N"and length frequency sample sizereported as "L/F N".ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN a 39I 2010 2011AgeN 281 0e0N = 245L/F N 2036 L/FN = 1831No. No,400 400200 200-IO I ...8 9 20 11 12 13 14 15 16 17 i8 19 20 21 22 23 24 25 26 27 28 8 9 10 a1 12 13 14 15 16 17 18 19 20 22 22 23 24 25 26 27 28Total Length (nmj Total Length (-)Figure 1.3.13. Age composition of Schoppee Brook, ME, fyke netcatch In 2010-2011.

Both gen-ders were combined with numberof age samples reported as "AgeN" and length frequency samplesize reported as "L/F N'.500082000A.I iB4.-Z R88.8 UA,04 *E0.-S000e-2 m008-2 §Age-3 30A0-42008200940eN.63V, -.92.40eN068/FN9 237200.1,.'..-9 20 21 12 13 14 15 26 17 IS 19 20 21 22 23 24 25 26 27 28Tow Length io)8 9 10 11 12 13 1415 16 17 18 19 20 22 22 D 24 25 26 27 28Total Length (rm)4002010AV N 2611 9N. 1308500400300-200-100-0No,No.C0e8- 11A9.2 384.3 IA0-4 3Age-52011Age N 2681SF N

• 12116 9 10 11 12 1 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28Tool Length c-)8 9 10 11 12 13 14 IS 16 17 19 19TOWa Length. (M)j20 21 22 23 24 25 26 27 28Figure 1.3.14. Age composition of East Bay Brook, ME, fyke netcatch In 2008-2011.

Both gen-ders were combined with numberof age samples reported as "AgeN" and length frequency samplesize reported as "L/F N.40

• ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN MALE250200-100-50I I 10 1 1

• 0 0 0

• iI I I I I ILC ML DM TB SB EBWW JR FR SG CN PRSQ OYRiverFEMALE250200-E 10150100--soI I I i1i jj iFigure 1.3.15. Median total lengthof smelt caught at 14 fyke netstations In the study area, 2008-2011 The top of the box plotsIs the 75th percentile and thebottom Is the 25th percentile.

Theline in the box Is the median andthe error bars mark the 10th and90th percentiles.

The stations arearranged on the x-axis from thesouthernmost MA station to thenorthernmost ME station.

Stationmedians for females and maleswere found to be significantly different with Kruskal-Wallis test,KW = 1324.94, df -13, p <0.001;and KW -2000.77, df -13, p<0.001, respectively.

WWV JR FR SG CN PRSQ OYRiverLC ML DM TB SB EBANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN

• 41 Dams, overfishing, andpollution have typically been considered themost important factorsaffecting diadromous fish, including rainbowsmelt.2 -THREATS TO RAINBOW SMELT POPULATIONS IN THE GULF OF MAINERainbow smelt encounter a variety of potential threats during their fresh-water and marine life stages. Dams, overfishing, and pollution have typically been considered the most important factors affecting diadromous fish, includ-ing rainbow smelt (Saunders et al. 2006, Limburg and Waldman 2009). Whilethese factors may have played major roles in the declines of rainbow smelt,other factors may also be responsible for recent declines.

Changes in trophicinteractions, community shifts, watershed land use, and climate-driven environmental conditions may all need to be considered when evaluating factors that affect rainbow smelt populations.

2.1 -THREATS TO SPAWNING HABITAT CONDITIONS AND SPAWNINGSUCCESSSpawning Site Characteristics Across their distribution range, smelt spawning runs are variable in regardto habitat use, spawning substrate, spawning period, and water temperature range (Bigelow and Schroeder 1953, Hurlbert 1974, Kendall 1926, Pettigrew 1997, Rupp 1959). Investigations of Massachusetts smelt runs have found thatspawning begins between late February and mid-March when water tempera-tures reach 4-6 'C and concludes in May (Chase 1990, 2006; Chase and Childs2001; Crestin 1973; Lawton et al. 1990). In New Hampshire, spring runsbegin in early to mid-March when the water temperatures reach 3-6 'C andconclude in May (NHF&G, current study). In Maine, the timing of the runvaries geographically, beginning in late March in waters west of the KennebecRiver, in mid-April in waters between the Kennebec River and the Penobscot River, in late April to early May in the Penobscot River and advancing to mid-May in most waters in downeast Maine. Water temperature at the beginning of runs varies from 1.5-9 'C, and most runs in Maine last four to five weeks(ME DMR, current study). There is also some evidence that rainbow smeltmay spawn in the main stem of large rivers in Maine earlier than runs begin insmaller streams close to these rivers. In rivers such as the Kennebec, Penobscot, Union, and Pleasant, spawning may occur under the ice or directly following ice-out in mid-March to early April (ME DMR, current study).The best documentation of the physical characteristics of smelt spawninghabitats in the Gulf of Maine is provided by a detailed assessment ofMassachusetts rivers that was conducted between 1988 and 1995 (Chase 2006).This study identified both stream attributes and water chemistry conditions that were suitable for smelt spawning.

Chase (2006) documented and mappedsmelt spawning habitat at 45 locations in 30 rivers on the Gulf of Maine coastof Massachusetts.

Rainbow smelt egg deposition was documented to take placeover stream sections ranging from 16 m to 1,111 m in length, with an average42 ° ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN of 261 m. In most cases, the downstream limit of egg deposition occurred nearthe interface of salt and fresh water, while the upstream limits were typically delimited by physical impediments that prevented further passage.

Whenpassage allowed, smelt would continue spawning in freshwater riffles beyondtidal influence.

The average patch size of substrate where smelt eggs wereobserved was 2,336 M2, with a range of 16 m2 to 13,989 m2.Smelt were found to spawn in shallow riffles where water velocity increased in stream channels.

Within the streams where smelt eggs were found, channelwidth averaged 6.8 m. Depth transects conducted in 16 of these streams foundthat the average depth of spawning riffles was 0.28 m, and the range of averagedepths was 0.1 -0.5 m under baseflow conditions.

However, smelt eggs werefound in depths up to 1.5 m in three surveyed rivers. The average water veloc-ity at the riffle transects was 0.39 m/s, with a range of 0.1 to 0.9 m/s. Thesemeasurements and observations of associated egg deposition led Chase (2006)to hypothesize that 0.5 -0.8 m/s was an optimal range for adult attraction andegg survival.

Observations in smelt spawning rivers in Massachusetts led Chase (2006)to conclude that the ideal channel configuration for spawning habitat may be-gin with a deep channel estuary where the salt wedge rises to meet a moderategradient riffle at the tidal interface and follows into the freshwater zone withample vegetative buffer and canopy and an extended pool-riffle complex thatspreads out egg deposition and provides resting pools. However, this scenariowas not common in Massachusetts spawning rivers, and likely is not in manyother rivers and streams in the Gulf of Maine. Many of the spawning streamsand rivers were altered by: (1) a range of passage obstructions (undersized cul-verts, dams, etc.) that limited or completely blocked the smelts' ability to reachtheir spawning

grounds, (2) channelization and flow alterations that changedwater velocity and substrate conditions, and (3) removal of riparian vegeta-tion, leading to increased amounts of polluted runoff flowing directly into thestream, as well as reduced canopy cover leading to increased water temperature.

These three categories represent major threats to spawning habitat and to smeltspawning

success, and they are described further in the following sections.

Inmany cases, these threats are present simultaneously in more developed water-sheds, compounding the threats to successful smelt spawning.

Obstructions DamnsIndustrial development depended on rivers for power, and over 500 damsremain on rivers in Maine, New Hampshire, and Massachusetts that may havea large impact on diadromous species (Martin and Aspe 2011). Dams blockaccess to spawning habitats for many anadromous

species, but their effecton rainbow smelt is particularly acute. The small body size of rainbow smeltmakes them unable to jump to heights necessary to migrate through fishladders, which pass other diadromous fish over dams. In Maine, at least 13 outof 275 (5%) historical and current spawning sites are either reduced in areaor the spawning habitat is blocked by coastal dams (Abbott, USFWS, pers.comm., 2012). In New Hampshire, although smelt spawning occurs in mostof the coastal rivers, head-of-tide dams exist on all of these rivers (with theANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN -43 exception of the Winnicut River), reducing habitat and forcing smelt to spawnwithin areas subject to tidal influence.

Although the exact number has not beendocumented, the same situation exists in Massachusetts, where head-of-tide dams limit spawning habitat.Road crossings The majority of smelt spawning streams in the Gulf of Maine are smallcoastal streams that are not dammed. More frequently, barriers are road-stream crossings.

Undersized, improperly installed, or poorly maintained culverts atroad-stream crossings can severely impair smelt migration.

This can occur whenculverts have become perched, where the downstream side stream height is wellbelow the culvert height, or when culverts are undersized to such an extentthat they create velocity barriers or reduce freshwater flow to levels that impedeenvironmental cues for smelt. Reducing stream habitat fragmentation is criticalfor increasing access to smelt spawning habitat.

In Maine, there is an ongo-ing effort to ground-survey all stream barriers.

At the time of this report, 35%of the state has been surveyed.

Of the 88 smelt historical or current spawningsites falling within this surveyed

portion, 34 (39%) sites have potential barri-ers to passage.

Extending the scope to the entire state, 127 historical or cur-rent spawning sites out of a total of 275 are crossed by roads at least once, andmultiple times in many cases. While some of these crossings may have adequatepassage, it is estimated that two-thirds of these crossings are undersized andmay present passage problems for smelt (A. Abbott, USFWS, pers. comm..,2012). The frequency of the problem is magnified in Massachusetts where onlyI of 45 mapped smelt spawning habitats were unaltered by road crossings orimpediments (Chase 2006).Channelization and Flow Disruptions Discharge and VelocityIn Massachusetts, New Hampshire, and Maine, most smelt runs occur insmall coastal rivers or streams with low seasonal baseflows where spring streamdischarge is sufficiently high to attract adults and support egg incubation.

Inthe Northeast United States, early spring flows are typically enhanced by snowmelt and precipitation, but discharge may decline progressively later in theseason. In a survey of 45 spawning rivers in Massachusetts, aside from theMerrimack River, only nine had average spring discharges over I m3/s (35 cubicfeet per second (cfs)), and only four exceeded a spring average of 10 m3/s (353cfs) (Chase 2006).During the current study, when USGS gauge stations were present, we re-corded river discharge weekly at our smelt spawning sampling sites. None of thesurvey stations in Maine were located on rivers with gauge stations; however,measurements were available for two New Hampshire sites and four Massa-chusetts sites (Table 2.1. 1). Over a two year period (2008-2009),

we found anaverage discharge of 1.83 m3/s (65 cfs) across all sites, with most values (75%)under 1.99 m3/s (70 cfs) (Table 2.1.2). Discharge varied significantly betweenthe sites, and was directly correlated to watershed size (Spearman's rank correla-tion = 0.78).Although high discharge is not a threat to smelt spawning, if it results insharp increases in velocity it impairs smelts' ability to reach their44 a ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN spawning grounds.

In watersheds with large amounts of impervious surface andnot managed for stormwater, infiltration of runoff is reduced and the smootherimpervious surfaces allow water to run off the surface and into streams faster.The combined result is a rapid increase in both volume and velocity (Cooper1996, Klein 1979). Substantial variability in velocity may be found within acoastal stream depending on specific location (e.g. pool versus riffle),

and tim-ing (precipitation events and tidal stage will affect daily velocities).

However, aspart of the current study we found that velocities at all spawning index sites fellwithin a fairly narrow range (0.32 m/s -0.58 m/s) when measurements weretaken within riffles when no tidal influence was present (Table 2.1.2). Velocityexceeded 0.79 m/s only 10% of the time, and generally the catch per unit effortof spawning adult smelt was lower during those high velocity events.Conversely, low discharge may also threaten successful spawning.

Suffi-cient freshwater flows are necessary for other anadromous species to cue theirmigrations and enable them to successfully locate their spawning site (Yako etal. 2002). Low discharge associated with urbanization may also lead to insuf-ficient water mixing, resulting in higher water temperatures, lower dissolved oxygen, increased sedimentation, and increased concentrations of pollutants and contaminants (Klein 1979). Reductions in baseflow can be caused by waterwithdrawals and impounding as well as increases in impervious surface (Klein1979, Simmons and Reynolds 1982). In many cases, withdrawals duringthe spring months may be expected to remove a small proportion of available spring flows. However, concerns are growing in urban areas where humanpopulation growth has increased water demands.

Furthermore, a gradual butmeasured loss in snow pack over the last century has led to a reduction ofspring baseflow in coastal streams, a situation that could compound concernsover water withdrawals.

Substrate and Channel Stability Natural stream and river channels that are vegetated and dynamic canabsorb the impacts of flooding by accommodating changes in discharge andwater levels. However, in urbanized areas with extensive impervious surfaceor where streams have been channelized by fixed walls, the runoff from largerain events flows directly into streams, leading to increases in the frequency and severity of flooding.

In turn, these events can cause channel erosion andalteration of the stream bed (Klein 1979). The timing of flood events can causepositive responses to smelt spawning substrata by scouring sediment andperiphyton before spawning occurs or negative responses by scouring awaylarge egg sets (Chase 2006). Booth and Reinelt (1993) report that pool andriffle habitat may be altered and channel stability may be degraded whenimpervious surface exceeds 10-15% of the watershed area These impacts canbe mitigated by restoring riparian buffers along stream and river banks.Watershed characteristics Watershed activities can have a substantial influence on many of the condi-tions identified above as potentially affecting rainbow smelt spawning habitat.Land cover in a watershed affects habitat conditions and biological communi-ties in receiving waters in a variety of ways (Burcher et al. 2007, Allan 2004).Urbanization and agricultural activities can contribute to erratic flow levels,Warmer water temperatures, channel alterations, sedimentation, chemical andANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN

• 45 Our analysis found thatweak spawning runs ex-isted in rivers surrounded by urbanized watersheds, while rivers drainingforested watersheds sup-ported stronger smeltspawning populations.

bacterial pollution, and nutrient loading (Wang et al. 200 la, Allan 2004).In addition, barriers to spawning passage are more likely to exist due to roadnetworks in more urbanized watersheds than in less developed areas. Thesewatershed-associated factors can all influence the suitability of streams forrainbow smelt spawning.

Associations between watershed characteristics and spawning site use havebeen observed for other anadromous species.

Limburg and Schmidt (1990)noted that spawning activity of anadromous fishes (mostly alewife) in tributar-ies to the Hudson estuary was inversely related to the proportion of urban landuse in the surrounding watershed.

In the Pacific Northwest, Pess et al. (2002)found that median densities of spawning coho salmon were 1.5-3.5 timeshigher in forest-dominated areas than in urban or agricultural areas. Theseexamples indicate that there may be linkages between spawning success andwatershed characteristics.

While the causal factors have not been identified, urbanization may influence in-stream habitat by increasing water velocities as-sociated with flood events, changing substrate, removing canopy cover and thusincreasing water temperature, and other habitat changes.In this study, we evaluated correlations between rainbow smelt catch perunit effort at the spawning index sites and land use in the adjacent watersheds at two spatial scales: (1) the full drainage basin and (2) the 210 meter bufferimmediately adjacent to the stream. Watersheds within which rainbow smeltspawning runs were sampled represented a wide variety of conditions (Table2.1.1). A principal components and cluster analysis suggests that the smeltspawning watersheds can be classified into three distinct types: (1) urbanized, (2) forested, and (3) wetlands/agricultural (Figure 2.1.1). Correlations be-tween the aggregate mean CPUE of spawning rainbow smelt over 2008-2011 (standardized based on net coverage of the stream width) indicate that weakspawning runs exist in rivers surrounded by urbanized watersheds, while riversdraining forested watersheds support strong smelt spawning populations.

Interestingly, the negative association between development and CPUE wassubstantially stronger at the scale of the full drainage basin than when only theriparian buffer zone was considered (Table 2.1.3). This appears to be becausemany rivers within urbanized watersheds have extensive riparian wetlands intheir buffer zones. The presence of these wetlands at the 21 0-m scale weakensthe influence of urbanization on smelt spawning.

Other land cover types andthe number of downstream crossings, at either the scale of the watershed orriparian buffer zone, were not significantly correlated to the strength of rainbowsmelt spawning populations.

46 e ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN Fyke Net Location Hydrologic Information Watershed Intormation Average AverageChannel Discharge Velocity DrainageRiver Latitude Longitude Town State Width (m) (mIs) Watershed (HUC 10) Area (kmn) Land Cover (11/20)Westport River 41.6209 -71.0598 Westport MA 11.3 Buzzards Bay 26.5 Forest I Agriculture Weweantic River 41.7662 -70.7461 Wareham MA 35,7 Buzzards Bay 148.2 Forest I Agriculture Jones River 41.9960 -70,7233 Kingston MA 27.3 1.92 0.492 South Coastal Basin 69.3 Forest I WetlandFore River 42.2225 -70.9732 Braintree MA 13.7 1.92 0.623 Boston Harbor 74.7 Development I ForestSaugus River 42.4680 .71.0077 Saugus MA 55.4 Boston Harbor 55.8 Development I ForestNorth River 42.5221 -70.9116 Salem MA 9.1 0.49 0.454 North Coastal Basin 12.6 Development I ForestCrane River 42.5966 -70.9364 Danvers MA 8.2 0.17 0.497 North Coastal Basin 14.0 Development I ForestParker River 42.7505 .70.9282 Newbury MA 54.8 0.516 Plum Island Sound 66.0 Forest I WetlandSquamscott River 42.9824 -70.9461 Exeter NH 101.0 5.65 0.384 Exeter River 276.9 Forest I WetlandWinnicut River 43.0389 -70.8455 Greenland NH 36.6 1.05 0.3 Great Bay 45.5 Forest I WetlandOyster River 43.1310 .70.1310 Durham NH 32.0 Great Bay 59.9 Forest I Development Long Creek 43.6332 3133 S. Portland ME 24.3 0.64 Fore River 17.5 Development I ForestMast Landing 43.8587 -70.0842 Freeport ME 15.2 0.468 Casco Bay Basin 20.7 Forest / WetlandDeer Meadow Brook 44.0369 -69.5874 Newcastle ME 24.0 0.489 Sheepscot River 27.6 ForestlWetland Tannery Brook 44.5706 .68.7888 Bucksport ME 07.7 0.402 Penobscot River and Bay 13.2 Forest I Agriculture Schoppee Brook 44.6627 -67.5533 Jonesboro ME 16.0 0.583 Roques Bluffs Frontal Drainages 9.3 Forest / WetlandEast Bay Brook 44.9547 -67.1041 Perry ME 21.9 0.217 Cobscook Bay 3.0 Forest I WetlandDischarge (ms Velocity (m/s)Minimum Value 0.04 0.050Lower Quantile (25%) 0.35 0.323Mean 1.83 0.478Upper Quantile (75%) 1.99 0.579Maximum Value 12.81 :1.483Table 2.1.1. Rainbow smeltspawning habitat station loca-tions for water quality monitoring.

Drainage areas are GIS calcula-tions set from the location of fykenet placement.

Table 2.1.2. Discharge and velocitymeasurements from spawningsurvey Index sites. Discharge measurements taken fromUSGS gauge stations upstreamof spawning sites and velocitymeasurements taken by statebiologists at the spawning sites(discharge n = 6, velocity n m 13)in active riffle areas.Table 2.1.3. Spearman's rankcorrelation between rainbow smeltspawning CPUE and land coverat two spatial scales. Correlation coefficients In bold type indicatesignificance at the p = 0.5 levelCorrelation with smelt spawning CPUELand Cover Watershed Level Stream Buffer Zone (210m)% developed

-0.62 -0.48% developed open space (parks, golf courses)

-0.47 -0.32% forest 0.60 0.60% wetland -0.29 -0.28% agriculture

-0.06 0number of downstream crossings

-0.46 -0.46ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN

• 47 Figure 2.1.1. Cluster analysis(Ward's method) of study water- _ __,sheds based on dominant landuses (as indicated by the propor-tion of developed, developed open, forest, agriculture, andwetland areas) and watershed characteristics (i.e., population

density, stream crossings, and 0proportion of Impervious sur- 0 * ,_ ,face). Station codes: NR -North E -, -_ _River, LC = Long Creek, CR=Crane River, FR = Fore River, SR= Saugus River, WE = Weweantic River, WN = Winnicut River, SQ= Squamscott River, JR = Jones 0River, PR = Parker River, EB X 0 W x W. ix a_Z __ U~ LL. W U) (L U4 ~ lEast Bay Brook, OY = Oyster River,TB = Tannery Brook, SB = Schop- Developed Agriculture Forestedl pee Brook, DM = Deer Meadow WetlandBrook, ML = Mast Landing.2.2 -THREATS TO EMBRYONIC DEVELOPMENT AND SURVIVALSmelt deposit demersal (sinking),

adhesive eggs at fast-flowing riffles,where they attach to the substrate or aquatic vegetation.

The duration of eggincubation is related to water temperature (McKenzie 1964), and in the Gulfof Maine, eggs hatch 7-21 days after fertilization (Chase et al. 2008, McKenzie1964). The success of this reproductive strategy depends on access from ma-rine waters, low predation, and suitable water and habitat quality for successful recruitment.

In many watersheds, the tidal interface is the physical locationfavored for the development of commerce and community centers.

This changein landscape can lead to hydrologic alterations, particularly in urban areas,leaving streams vulnerable to point and non-point source pollutants; nutrientenrichment; and reduced streamflow, shading and riparian buffer.Changes in spawning habitat may be a major factor in the decline of smeltpopulations.

However, up to this point, the degree to which water qualityimpairment may be impacting smelt populations in the Gulf of Maine has notbeen described.

With this concern in mind, we developed monitoring pro-grams to assess baseline water and habitat conditions at smelt spawning habitatindex sites spanning the entire Gulf of Maine and explored possible impactson spawning success resulting from changing habitat conditions.

This informa-tion is applied to support recommendations for conserving and restoring smeltpopulations and habitats.

Four indicators were measured to assess water quality at smelt spawningindex sites: basic water chemistry, nutrient concentrations, periphyton growthand heavy metal concentrations.

The sampling was guided by a Quality Assur-ance Program Plan (QAPP) for monitoring water and habitat quality at smeltspawning habitats in coastal rivers on the Gulf of Maine coast (Chase 2010).The QAPP integrates smelt life history with existing state and federal waterquality criteria, with the objective of developing a standardized process toclassify the suitability of smelt spawning habitat.

Beyond characterizing smelt48 -ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN

habitat, it is our hope these data will contribute to water quality and habitatrestoration efforts at coastal rivers in New England.Summary statistics were generated for water quality data by site and thencompared to thresholds assembled from existing water quality criteria (Table2.2.1). The U.S. Environmental Protection Agency (EPA) developed criteriafor turbidity, total nitrogen (TN) and total phosphorus (TP) based on the25th percentile of the distribution of observed values in an ecoregion (US EPA2000). The 25th percentile is the value of a given parameter where 25% of allobservations are below and 75% are above. The 25th percentile was adoptedby EPA as the threshold between degraded conditions and minimally impactedlocations.

Additionally, the Massachusetts Department of Environmental Protection (MassDEP) established Surface Water Quality Standards (SWQS)for temperature, pH and dissolved oxygen (DO) as part of their Clean WaterAct waterbody assessment process (MassDEP 2007). These thresholds wereselected to protect designated categories of aquatic life, including fish habitat.Stations were classified as Suitable (minimally impacted) or Impaired for eachparameter.

Water quality data were also evaluated to explore the potential ofestablishing new thresholds specifically derived from smelt spawning habitatmeasurements.

Water Chemistry Basic water chemistry parameters were measured during smelt spawn-ing runs at 19 index station stations:

the 16 fyke survey sites and 3 additional spawning sites of interest in Massachusetts (Figure 1.3.7 and Table 2.1. 1)following the QAPP protocol.

Yellow Springs Incorporated (YSI) water chem-istry sondes were used to measure water temperature

(*C), DO (mg/L and %saturation),

specific conductivity (mS/cm),

pH and turbidity (NTU, Neph-elometric Turbidity Units) in freshwater at the spawning grounds.

At moststations, discrete water chemistry measurements were recorded three times perweek. The seasonality of water chemistry monitoring was not synchronized forall stations due to the later onset of the spawning season at the northern end ofthe study area. For this reason, detailed comparisons of some parameters, suchas temperature, should be made cautiously.

Water Temperature Water temperature has an important influence on smelt metabolism, theonset of smelt spawning and the duration of egg incubation.

Median watertemperatures during the spawning period were fairly consistent across the studyarea, with a range of 8.8 -12.9 'C (Table 2.2.2, Figure A.2.1). No measure-ments exceeded the water temperature criterion of 28.3 0C adopted from Mass-DEP SWQS to protect aquatic life. The relatively high temperature threshold has little relevance for smelt that spawn in the cool water of the spring freshet;however, the temperature data have value for documenting baseline conditions and may have future application for monitoring reference values, such asstation medians or 75th percentiles.

Specific Conductivity Specific conductivity is proportional to the concentration of major ions insolution corrected to the international standard of 25 °C. High conductance inANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN a 49 freshwater can indicate high watershed contributions of natural alkaline com-pounds or ionic contributions from pollution sources.

For this reason, conduc-tivity has been discussed as a potential proxy for pollution

sources, urbaniza-tion, and eutrophication.

Median specific conductivity during the spawningperiod ranged from 0.031 -0.997 uS/cm (Table 2.2.2, Figure A.2.2). The fourhighest medians occurred at urban sites near the Boston metropolitan area.Dissolved OxygenAdequate dissolved oxygen (DO) concentrations are necessary for embry-onic survival and normal development.

The QAPP provides a DO criterion of > 6.0 mg/L to protect aquatic life. Median DO concentrations during thespawning period ranged from 9.5 -12.5 mg/L (Table 2.2.2, Figure A.2.3), andmedian DO saturation levels ranged from 91.0 -107.8% (Table 2.2.2, FigureA.2.4). All individual DO measurements were well above the DO threshold.

Similar to water temperature, the DO threshold may have limited relevance because of the high concentrations of DO found in turbulent riffles during thespring freshet.

The distribution of DO saturation data does show increasing supersaturation in urban Massachusetts and a declining DO saturation mov-ing north in the study area. Supersaturation of oxygen can indicate eutrophic conditions, where due to the photosynthetic cycle of the algal communities, supersaturation is observed during the daylight hours, but anoxic conditions arepresent during darkness (Carlton and Wetzel 1987).pHIncreased acidification of water bodies in New England is a widely recog-nized threat to fish populations, as low water pH can increase the impact of alu-minum toxicity and disrupt fish respiration.

Geffen (1990) conducted laborato-ry experiments to examine the influence of pH on smelt embryo survival; trialsfound that survival was most influenced by the duration of low pH exposureand embryo developmental stage. For example, high mortality occurred to earlystage smelt eggs (4-6 days post-fertilization) at 5.5 pH when exposure rangedfrom 6-11 days. Fuda et al. (2007) conducted similar experiments and foundsurvival was not affected until pH was < 5.0. The QAPP adopted the water pHcriterion of > 6.5 to _< 8.3 from MassDEP (2007) to protect aquatic life. Moststations had pH measurements in a range that was not a concern for rainbowsmelt. Median pH during the spawning period ranged from 5.92 -7.67 (Table2.2.2, Figure A.2.5). Of the 19 rivers sampled, seven were classified as Impaired(> 10% of individual measurements below pH 6.5). Among the stations classi-fied as Impaired, only four had routine measurements below 6.0 pH: the threesouthernmost Massachusetts stations and Schoppee Brook in Maine.Turbidity Turbidity in water is the result of suspended inorganic and organic matter;it can be caused by natural fluctuations in sediment transport or by changesin productivity.

The QAPP adopted the turbidity criterion of : 1.7 (NTU)from the EPA Northeast Coastal Zone ecoregion (US EPA 2000). Most rivershad median turbidity values >1.7 NTU, and all were classified as Impaired forhaving at least 10% of measurements

> 1.7 NTU (Table 2.2.2, Figure A.2.6).Several stations in New Hampshire and southern Maine had median valueswell above the threshold.

However, this elevated turbidity may result from thenatural suspension of sediments, either due to soil type or the naturally high50

• ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN turbidity in the spring associated with snow melt and higher runoff. Adoptingthe study's 25th percentile of 1.9 NTU would still result in all stations beingclassified as Impaired.

The turbidity data will be further evaluated to determine if a more appropriate turbidity threshold can be established by removing pre-cipitation effects through an analysis of baseflow data.Data AnalysisMedian values of water temperature, DO, specific conductivity, pH andturbidity were compared among sampling stations (Kruskal-Wallis, p < 0.001),and a multiple comparison test was used to determine which stations weresignificantly different from others (Siegal and Castellan, 1988; R code, krus-kalmc; p = 0.05; Figures A.2.1 -A.2.6). Significant differences were foundfor all parameters; trends between parameters were common among rivers andregions.

Conductivity was especially variable among sites and may be relatedto watershed characteristics; in the most urban sites (Crane and North rivers,Massachusetts) conductivity was significantly higher than most other sites,whereas at the forested sites (Deer Meadow and East Bay Brooks, Maine),conductivity was significantly lower than most other sites, The relation of thesevariables to spawning smelt populations is discussed in the Watershed Charac-teristics Section.Nutrient Concentrations Nitrogen and phosphorus are vital nutrients for plants but can cause exces-sive growth and degrade the health of aquatic life at high concentrations.

Theinfluence of nutrient pollution on water and habitat quality in rivers and lakesis a growing concern in the United States (Mitchell et al. 2003). The healthor trophic state of aquatic habitat is influenced most by light, carbon sources,nutrients, hydrology and food web structure (Dodds 2007). Among theseinfluences in developed watersheds, nutrient enrichment is most dependent on human activity and may be most amenable to remediation efforts.

Totalnitrogen and total phosphorus were recorded weekly at index stations in thefreshwater portion of the streams on the spawning grounds from 2008-2011.

Field sampling procedures are documented in the QAPP (Chase 2010), and thelaboratory analysis followed EPA-approved Quality Assurance

/Quality Control(QA/QC) protocols.

Nutrient concentrations for smelt spawning habitat were classified us-ing EPA recommended thresholds for freshwater streams and rivers that weredeveloped from the distribution of available water quality data (US EPA 2000).These EPA thresholds for Suitable habitat for the study area are 0.57 mg/L fortotal nitrogen (TN) and 23.75 ug/L for total phosphorus (TP). The EPA alsorecommends that states develop their own nutrient water quality criteria forprotecting specific designated uses of aquatic habitat under Clean Water Actassessment and remediation processes (US EPA 2000). In this light, the TNand TP data recorded for this study were compared to the EPA nutrient criteriaand the data distributions were evaluated for potential smelt habitat-specific thresholds (Table 2.2.3)Total NitrogenMeasurements of TN at 20 stations during 2008-2011 showed a trend ofANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN

• 51 higher concentrations in urban areas (Table 2.2.3, Figure A.2.7). The rangeof median concentrations for all stations was 0.216 -1.395 mg/L. Only fivestations were classified as Suitable for TN (_< 10% of measurements below0.57 mg/L; EPA 2000), with four of these stations at the northeastern end ofthe study area. All others were classified as Impaired.

The TN 25th percentile generated from the study sites was 0.340 mg/L, which was 40% lower than theEPA ecoregion threshold.

Total Phosphorus Measurements of TP displayed a more stable trend across the study area(Table 2.2.3, Figure A.2.8). The range of median concentrations for all stationswas 12.18 ug/L to 36.72 ug/L. Only 4 stations were classified as Suitable for TP(< 10% of measurements below 23.75 ug/L; EPA 2000). All others were classi-fied as Impaired.

The TP 25th percentile generated from the study stations was17.56 ug/L; 26% lower than the EPA ecoregion threshold.

TN/TP RatioWhile total concentrations of nitrogen and phosphorus are important forplant production, the balance or ratio of TN to TP can also influence growthand species composition.

Most TN:TP ratios were in a range expected forfreshwater systems in New England (15:1-30:1).

Higher ratios indicating highnitrogen and possible phosphorus limitation were found at the most urbanized

stations, and low ratios most influenced by high phosphorus were only found ata few stations where watershed development was low.Data AnalysisComparisons of median TN, TP and TN:TP ratios among sampling sta-tions found significant differences for all three parameters (Kruskal-Wallis, p <0.001). A multiple comparison test was used to determine which stations weresignificantly different from others (Siegal and Castellan, 1988; R code, krus-kalmc; p = 0.05). The box plots in Figures A.2.7 -A.2.8 represent a graphicdisplay of the multiple comparisons.

The high TN concentrations at CraneRiver and North River (> 1.0 mg/L) in Massachusetts were significantly dif-ferent from all stations except the Saugus River. The four stations with medianTN < 0.3 mg/L were significantly lower than most the remaining

stations, allbut one found in urban areas of Massachusetts and New Hampshire.

Periphyton Periphyton is the complex of benthic algae, detritus and other microorgan-isms that attaches to the river bed and is an important indicator of primary pro-duction and environmental disturbances in aquatic habitats.

Periphyton growthresponds to nutrient enrichment and can reach excessive or nuisance growth ineutrophied systems (Biggs 1996). Eutrophication has been identified as a majorconcern for smelt spawning habitat due to the potential impact of excessive pe-riphyton growth on smelt embryo survival at spawning riffles in Massachusetts (Chase 2006). These concerns have also been raised for smelt runs in tributar-ies to the St. Lawrence River in less urban regions of Qu~bec (Lapierre et al.1999). Periphyton monitoring was conducted to provide a biological responsevariable for nutrient concentrations that may be directly related to successful embryonic survival.

Laboratory experiments studying the effect of periphyton 52 , ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN growth on smelt embryo survival complimented the field monitoring, The labresults demonstrated that embryo survival was significantly lower on substrata with high periphyton growth/concentrations than on clean surfaces (Wyatt etal. 2010).Field monitoring measured the growth of periphyton on spawning groundsubstrate at the index sites during the spawning period to determine howgrowth may differ between sites. Ceramic tiles were deployed to collect pe-riphyton during the 2008-2009 spawning period at riffle habitat where smeltdeposit eggs. Periphyton growth on the tiles was collected biweekly to quan-tify daily growth and describe algal species composition.

Ash-free dry weight(AFDW, g/m2/day) was calculated as a measure of periphyton biomass.

Averageperiphyton growth ranged from 0.006 to 0:120 g/m2/day at 12 smelt spawninghabitat stations (Table 2.2.3). The range of periphyton growth included verylow growth at the easternmost Maine stations to high growth at urban centersin Massachusetts.

No algal biomass thresholds are available specifically for smelt spawninghabitat.

In the absence of published thresholds, the 25th percentile of 0.0143g/m2/day was calculated from the AFDW medians observed during this studyand compared to all values. All river stations exceeded this threshold and wereclassified as Impaired for periphyton, except for Deer Meadow Brook, Chan-dler River and East Bay Brook, Maine. The periphyton data suffer from highvariability and low sample sizes at some sites. However, there appears to bepotential value in using the 50th percentile (0.0533 g/m2/day) as a threshold for moderately impacted rivers. At the stations with medians above the 50thpercentile (Figure 2.2.1), the periphyton could be characterized as excessive growth that could impede egg incubation and appears to be associated withhigher TN and urbanization.

However, more work is needed to understand the range of periphyton growth at different spawning

streams, how this var-ies annually in response to environmental conditions, and the point at whichperiphyton growth impairs embryo survival.

Heavy Metal Concentrations Heavy metals such as cadmium,

chromium, copper, iron, lead, manganese,

mercury, silver and zinc can be absorbed by both fish embryos and larvae andlead to developmental abnormalities and reduced survival (Finn 2007, Jezierska et al. 2009, Wegwu and Akaninwor 2006). Short-term, high-intensity con-tamination mostly occurs in the spring months during snowmelt

periods, whenmild water acidification that is associated with snow melt leads to free metalions being leached from sediments (Jezierska et al. 2009). Long term exposureto lower concentrations of heavy metals may be of equal concern.

The toxic ef-fects of aluminum on salmonid embryos are seen when pH is below 6.5; at thislevel, pH can inhibit the swelling of the egg shell, reducing the amount of spacefor the embryo to develop and move, and leading to stunted growth or physicalabnormalities (Finn 2007). Cadmium, lead and copper at low levels can exac-erbate these effects at any pH (Jezierska et al. 2009). Above critical thresholds,

mercury, lead, cadmium,

chromium, iron, and zinc have all been shown toreduce the number of embryos successfully hatching (Wegwu and Akaninwor 2006), as well as to disturb skeletal growth, impair hemoglobin (red blood cell)ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN

• 53 formation, cause osmoregulatory

failure, and limit overall growth because theorganism's energy is spent ridding the body of the toxic contaminants (Finn2007; Jezierska et al. 2009).We sampled heavy metal concentrations and other minerals (calcium andmagnesium) at all index sites during baseflow conditions over the course of thespawning period in 2010 and 2011 to describe the range of concentrations towhich smelt embryos are chronically exposed.

Although not part of this study,corollary laboratory experiments should be performed to ascertain which metalsand what concentrations reduce survival and impair normal development insmelt embryos and larvae.Of the heavy metals, silver, cadmium, and mercury concentrations werebelow detection levels for all sites during all sampling periods (detection levels0.002 mg/l, 0.5 ug/l, 0.5 ug/l, respectively).

Chromium was detected only onceduring the sampling period, in the Oyster River, New Hampshire (0.003 mg/l;detection level 0.002 mg/I). Although these metals were not detected, or de-tected only once, it should not be assumed that they are not present.

They mayin fact be present either at concentrations below the detection levels or duringrunoff or precipitation events neither of which our sampling captured.

All othermetal concentrations were detected at most sites, and the range of values fol-lowed a log distribution.

As log distributions are typical of metal concentrations in many regions, the values we measured likely represent much of the range ofmetal concentrations present in the region during the smelt spawning season(Table 2.2.4).A principal components analysis (PCA) was performed using the 2010-2011 average concentrations (log transformed to produce normal distributions) to determine which metal and mineral concentrations trended together, andwhich seemed to vary on their own. From this analysis, we find that lead (Pb;abbreviations refer to labels in associated figure, and are not the full elemental symbols with ionic sign), copper (Cu), and zinc (Zn) are highly related andtrend opposite from aluminum (Al). This pattern indicates that when highvalues of lead, copper, and zinc were present, aluminum values were low, andvice versa. Being drivers of water hardness, calcium (Ca) and magnesium (Mg)were highly related to hardness and alkalinity, but notably nickel (Ni) was alsohighly related to these variables (Figure 2.2.2).The relationship between metal concentrations and watershed characteris-tics is explored in the following section.Watershed characteristics As suggested throughout the preceding

sections, watershed land use canaffect water quality in receiving streams and rivers in a variety of ways. Thedevelopment of wetlands, agricultural fields, or forested areas replaces poroussoils with impervious

surfaces, which increases the velocity of water flowing offthe land and the supply of suspended sediments, nutrients, and contaminants to adjacent streams (Brenner and Mondok 1995, Corbett et al. 1997, Strayer etal. 2003, US EPA 2004). In addition, agricultural areas contribute nutrients-both nitrogen and phosphorus-to receiving streams.

In aquatic ecosystems, these nutrients can promote algal blooms, deplete oxygen, and degrade fishhabitat (Carpenter et al. 1998, Howarth et al. 2000).54

• ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN Understanding how water quality, nutrient levels, and heavy metal concen-trations are related to watershed land use is important for developing manage-ment strategies to minimize impacts to rainbow smelt eggs and larvae.Correlations between watershed land use and water quality parameters, nutrient levels, periphyton growth, and heavy metal concentrations were evalu-ated using Spearman's rank correlation statistic.

Results are presented in Table2.2.5 at the scale of the full drainage basin and riparian buffer zone. Severalkey patterns emerge from these correlation results that are relevant to rain-bow smelt conservation.

First, patterns are very similar at full watershed andriparian buffer scales, indicating that land use in the broader watershed exertsa similar influence on water quality as land use immediately adjacent to thereceiving stream. Second, the percent of development and forest in the water-shed show the strongest associations with water quality, with the direction ofinfluence occurring in opposition to one another.

For example, higher percent-ages of developed areas are associated with higher stream dissolved (available) nitrogen and heavy metals concentrations; conversely, highly forested water-sheds are associated with lower concentrations of nitrogen and metals (Craw-ford and Lenat 1994). Because periphyton growth is dependent on available nutrients (like dissolved nitrogen),

and because heavy metals can negatively affect embryo development and survival, this pattern suggests that protecting forested areas is important for maintaining water quality conditions that arebeneficial to rainbow smelt.Conclusions When compared to the established EPA thresholds, the water quality datacollected during 2008-2011 show widespread impairment due to elevatedTN, TP, and turbidity and more localized impairment from acidification andexcessive periphyton growth. More work is needed to evaluate existing criteriaand to establish new thresholds that are specific to smelt spawning habitat.For example, the turbidity criterion is likely too low to be relevant for streamriffles during spring; conversely, the water temperature and DO criteria may betoo high, as smelt embryos require a lower temperature than the current EPAthreshold.

The highest median values for TN, conductivity and periphyton were associated with urban sites. Most sites with few identified impairment were at the northern end of the study area.These results provide a range of water quality conditions that affectsuccessful embryonic survival.

From high impairment in urban settings to suit-able water quality in rural settings, these sites are examples of both conditions requiring remediation and demonstrating restoration targets.

We encourage resource managers to use these baseline conditions to consider potential reme-diation measures (e.g., riparian

buffers, stormwater improvements, point sourcereductions) to improve impairments and to plan for protecting locations withsuitable conditions for supporting smelt spawning success.Understanding how waterquality, nutrient levels,and heavy metal con-centrations are relatedto watershed land use isimportant for developing management strategies to minimize impacts torainbow smelt eggsand larvae.ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN

• 55 Table 2.2.1. Water chemistry criteria related to smelt spawn-Ing habitat.

The water chemistry parameters were adopted toprotect Aquatic Life at Class BInland Waters (MassDEP 2007),and US EPA reference conditions (25th percentile) for the Northeast Coastal Zone sub-Ecoreglon (USEPA 2000). Potential criteria arepresented based on 25th and50th percentiles from 2008-2011 project data. Blank cells Indicateeither that no criterion exists orthe derived percentile has limitedrelevance for smelt habitat.Existing Water Ouallty CriteriaSuitable Minimally Minimally Moderately Impacted Impacted Impacted25th Percentile 25th Percentile 50th Percentile Parameters (MassDEP 2007) (US EPA 2000) (2008-2011 data) (2008-2011 data)Temperature

(°C) 5 28.3Sp. Conductivity (mS/cm) 5 0.131pH > 6.5 to 5 8.3DO (mg/L) a 6.0Turbidity (NTU) S 1.7 5 1.9 5 2.1TN (mg/L) 5 0.570 5 0.340 5 0.452TP (ug/L) 5 23.75 5 17.56 5 20.43Perlphyton Biomass (g/m2/d) 5 0.0143 5 0.0533_ _mTI_ Cond. DO % SN NTUMWia Rie Cods e in Exceed Median ECW Mjn Exceed Meia Exceed Median Excee Medfaia ExceedMA Weatpoft W~p 95S5 0% 0.130 96.1 0.6 0 .92 1.4 -3MA Woantlic WW 11.05 0% 0.092 95.9 10.55 0% 6.23 W% 2.2 40MA Jones JR 9.71 0% 0.200 100.0 11.74 0% 639 M% 2.8 OWMA Fore FR 10,26 0% 0.558 105.1 12D6 0% 7.09 2% 2.1 A16MA Saugus SG 8-89 0% 0.663 102.3 11.98 0% 7.28 0% 2.9 ofMA North NR 9,57 0% 0.962 105.0 12.45 0% 7,24 0% 2.0 74%MA Crone CN 9.22 0% 0.997 99.1 11.89 0% 7.18 1% 3A 90MA Essex ER 9,83 0% 0200 1Q5.2 12,32 0% 6,71 28 13 293MA parker PR 9.11 0% 0.2.2 105.1 11.88 0% 7.02 1% 1,8NH Squamsoott So 11t89 0% 0.152 100.4 10.93 0% 6.93 2% 1.8 57%NH WiMIcut WR 11.50 0% 0.315 97.6 1121 0% 7.43 0% 4.3 0%NH Oyster OY 10,48 0% 0.195 101.0 11.31 0% 7.38 0% 4.4 ME Long Creek LC 10.36 0% 0.526 97.0 11.07 0% 7.25 0% 6.9 1OO%ME Mast Landing ML 8.79 0% 0.134 98.1 11,52 0% 7.11 8% 8.8ME DeerMeadow DM 10.99 0% 0,031 9810 11.14 0% 6.84 1.. 24ME Tannery Brook TB 12O58 0% 0.157 98.1 10.43 0% 7.67 4% 1,8ME Sdchppe SB 9.48 0% 0.089 92.5 10.2 0% 6.27 2.1 %ME Chandler River CR 12,86 0% 92.8 9.47 0% 6.72 21 2.0 10,ME EtBa Es 9.80 0% 0.044 95.1 10.78 0% 7.31 7% 2.02Mth Periet 9.51 0.131 96.6 10.85 6.72 1.950th Percentile 9.83 0.197 98.1 11.21 7.09 2.1Table 2.2.2. Basic water chemistry measured at 19 smelt fyke netindex stations in the U. S. Gulf ofMaine and Buzzards Bay, Mas-sachusetts.

Median values werecalculated from all available datafrom 2008-2011.

The percent-age of samples at each stationthat exceed the QAPP (Chase2010) thresholds are presented In shaded cells, indicating anImpaired classification for the pa-rameter.

No water quality criteriaare available for conductivity orDO saturation.

56

• ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN TP TNN:P AFDW (alelday)

Sbft Rhr Code N Median Exceed N Median Exceed N Medan N MedinMA Westport WP 25 19.20 ýo 25 0... W 25 332 0MA Weweenbc WW 26 37.80 4% 23 0.283 '11 23 7.8 0MA Jones JR 48 16.70 13* 47 0.659 9 47 34.1 8 0.0169MA Fore FR 47 21.10 M% 48 0.530 l% 47 23.1 8 0.0154MA Saugus SG 10 26.95 70% 11 0.917 1*% 10 35.1 0MA North NR 47 21.06 % 49 1,395 10% 47 88.0 6 0.0828MA Csrne CN 48 21.89 A 48 1.285 10ft 48 58.9 8 0.1198MA Essex ER 11 12.80 9% 11 0.411 9% 11 31,3 0MA MIN MR 45 21.80 3 46 0.844 45 28.0 8 0.0685MA Parker PR 11 17.80 0% 11 023 411 31.0 0NH Squamscott SO 37 17.44 9 37 OA20 % 37 22.7 9 0.0598NHi Winnlcut WR 37 20.10 36 0.516 M 36 25.3 9 0.08867NH Oyster OY 15 22.70 15 0.387 20 15 18.3 0ME Long Creek LC 30 20375 29 0.425 ... 29 23.8 4 0.0625ME Mat Landing ML 37 18.81 22% 37 0.258 0% 37 11.9 0ME DeerMeadow DM 37 17.0 .35 0.253 0% 35 16.5 4 0.0068ME Tannery Brook TB 32 23.84 0 32 0&332 0% 32 13.9ME Scloppe.

SB 18 27.00 81% 18 0.479 11% 18 15.5 0ME Chandler River CR 10 14.95 0% 9 0.342 111% 9 24.8 4 0.0111ME East Say Es 34 11,15 6% 33 0.216 0% 33 17.7 4 0.006525th Pevetl 17.56 0.340 17.4 0.01435lih Percentle 20.43 0.452 24.1 0M0533Table 2.2.3. Nutrient and periphy-ton measurements for all Indexstations In the U. S. Gulf of Maineand Buzzards Bay, Massachu-setts. The percentage of samplesat each station that exceed theQAPP (Chase 2010) thresholds are presented in shaded cells,Indicating an Impaired classifica-tion for the parameter.

No criteriaare available for the N:P ratio orperlphyton.

AnalyteUnit 2010Detection Limit2011Detection Limit2010-2011 2010-2011 2010-2011 Mean Low HighValue Value ValueAluminum mg/LArsenic ug/LCadmium ug/LCalcium mg/LChromium mg/LAlkalinity mg/LCopper mg./LIron mg/LLead ug/LMagnesium mg/LNickel mg/LSilver mg/LZinc mg/LTotal Hardness mg/LMercury ug/L0.0050.50.50.050.00210.00050,050.50.050.00050.0020.0020.350.50.010.50.50.050.00210.00050.050.50.050.00050.00050.0020.33Not Sampledin 20110.13471.30BDL13.780.00329.140.00130.621.054.270.0016BDL0.00654.6BDL0.00590.51BDL0.550.0033.260.00050.160.380.270.0005BDL0.0022.5BDL1.00004.00BDL52.000.003100.000.00772.703.1039.000.0050BDL0.021430.0BDLTable 2.2.4. Analytes measured inwater samples taken at baseflowat smelt spawning Index sites2010-2011.

Detection limits andmean, low, and high concentra-tions are shown for each analyte.BDL -below detection limit.ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN -57I Full watershed Stream buffer%dev %devopen

%forest %wetiand

%a I %dev %devooen

%forest %wetland

%aa4water qualityconductivity DO conc.pHturbidity TPTNAFOWalkalinity hardnessMetalsAdAsCaCuFeMiNIPbZn0.S50.670.360.320.260.870.620.830.83-0.s30.S40.830.580.260,860.890.640.70.90.650.390.470.340.770.490.770.78-0.440A50.750.450.430.840.890.630.74-0.83-0.38-0.25-0.22-0.46-0.81-0.57-0.66-0.70.46-0.44-0.68-0.37-0.32-0.74-0.74-0.72-0.87-0.16 -0.12-0.18 -0.19-0.42 0.01-0.14 -0.04-0.21 0.040.1 -0.16-0.1 0.23-0.23 -0.14-0.24 -0.110.2 0.220.13 0.22-0.3 -0.16-0.41 00.22 0,42-0.05 -0.06-0.21 -0.04-0.45 -0.81-0.29 -0.350.94 0.920.51 0.560.42 0.430,28 0.510.36 0.310.85 0.74069 0.550.8 0.760.88 0.88-0.39 -0.280.61 0.570.86 0.820.42 0.350.19 0.410.89 0.92O.8i 0.830.81 0.590.74 0.71-0.79-0.34-0.3-0.12-0.48-0.74-0.58-0.66-0.680.56-0.67-0.41-0.3-0.67-0.69-0.64-0.82-0.01 -0.24-0.05 -0.2-0.33 -0.14-0.21 -0.18-0.11 -0.020.24 -0.190.04 0.02.0.05 -0.25-0.16 -0.330.02 0.130.15 -0.04-0.19 -0.36-0.25 -0.090,24 0,34-0.01 -0.26-0.08 -0.2-0.36 -0.8-0.25 -0.44Table 2.2.5. Spearman's rankcorrelation between water qualitymetrics and land cover at twospatial scales (e.g., full water-shed and riparian buffer zone).Correlation coefficients In boldtype indicate significance at thep-O.05 level.Figure 2.2.1. Annual medianperiphyton growth (ash-free dryweight, g/M2/day) displayed bysample station with 50th per-centile of station median valuesmarked by green line. Refer toTable 2.2.2 for river codes.0.350,30 92008 320090,250.20AFDW0.150.00 -JR FR NR CM MR SQ WR IC DM TB CR EBRiver58 e ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN oMgAlk Ca Figure 2.2.2. Principal compo-..nents analysis (PCA) performed 0 on 2010-2011 average metaland mineral concentrations (log-2C transformed).

The first compo-nent Is driven most by hardnessZn (a variable which represents thetotal mineral concentration ofwater, driven by calcium and mag-neslum),

magnesium, calcium,alkalinity, and nickel. The secondcomponent Is driven most In thepositive direction by aluminumand arsenic and less so by iron,Component 1 (56.6 %) and In the negative direction byzinc, copper, and lead.2.3 -THREATS TO SMELT IN MARINE COASTAL WATERSSmelt spend at least half the year in marine coastal waters during thesummer and fall months. As adults and juveniles they are a schooling fish thatattract a wide range of predators.

While monitoring this life phase can be moredifficult than monitoring discrete spawning runs, it is no less important whenconsidering the species decline.

During this period, smelt are susceptible toenvironmental influences on survival, shifts in natural mortality and to capturein small mesh fisheries targeting other species.

These topics are discussed below,using the best available information to discuss how each issue may be affecting smelt populations;

however, to fully understand the implications, each requiresfurther study.Fish HealthImproving understanding of fish health status as well as the abundance, geographic distribution, and vectors of areas of study necessary to support thedevelopment and implementation of conservation strategies designed to protectand restore rainbow smelt populations.

Pathogens can adversely affect bothjuveniles and adults in both general and acute ways, including organ failure,energy loss, interruption of hormonal pathways and reproductive weakness (D.ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN -59

Bouchard, University of Maine, pers. comm., 2011).We characterized pathogen presence endemic to smelt at fourteen spawn-ing index sites spanning the Gulf of Maine over a two-year period, 2009-2010 (Bouchard 2010). Sampling did not detect bacterial pathogens of regulatory concern but did detect endemic parasites that are well documented for similaranadromous species.

Parasitological results were typical of wild fish populations, with various trematodes (e.g., black grub), cestodes, nematodes and protozoaobserved at all sites. A microsporidian parasite detected in various tissues ofmany individuals in this study was not identified as to species, but is consis-tent with (Glugea hertwigi),

which was confirmed at one site: the Fore River,Massachusetts.

This parasite has been documented extensively in freshwater smelt can be detrimental to successful spawning because this parasite infests thegonads of smelt (Jimenez et al. 1982, Nsembukya-Katuramu et al. 1981). Theobservation of large numbers of (Philometra spp.)-like nematodes in the gonadsof the majority of female fish in the study is also consistent with reports of thisparasite as an opportunistic pathogen of spawning female fish in other species(Moravec and de Buron 2009).Virology results revealed a viral agent from adults from Casco Bay, Maine;however, it is difficult to place any significance to this agent at the present timebecause the virus is not similar to currently catalogued agents (IPNV, IHNV,ISAV, and VHSV have been ruled out by PCR techniques).

More analysis onthis agent is needed to fully understand the physiological effects it may be hav-ing. Fish from a majority of the sites spanning the entire Gulf of Maine regionshowed evidence of erythrocytic

disease, or degradation of red blood cells,leading to anemic effects (Bouchard 2010). This last point may be of specificconcern and warrants further investigation to understand the extent of diseaseand causal factors.Fish from a majority ofthe sites spanning theentire Gulf of Maineregion showed evidenceof erythrocytic

disease, ordegradation of red bloodcells, leading to anemiceffects.Fishing Mortality Overfishing in historicalfisheries While historical fisheries for rainbow smelt landed thousands (and inMaine millions) of pounds annually in the 1800s, because the relative size ofthe entire population was unknown, it is not possible to quantify the effect ofthese targeted fisheries on smelt populations.

As populations declined in the 20th century, and as regulations limitedfishing gear and take in response to this decline, targeted fishing effort has alsobeen reduced.

Today, few targeted commercial fisheries exist: a dip and bownet fishery is open to permitted individuals in Great Bay, New Hampshire; anda gill and bag net fishery are allowed during a regulated time period to per-mitted individuals on five rivers in downeast Maine. Large-scale recreational hook-and-line ice fisheries also exist in Great Bay, New Hampshire, and onmany rivers and embayments in Maine (most notably the Kennebec River andMerrymeeting Bay area). While these fisheries are not thought to contribute high mortality for the smelt populations they target, the current extraction ratesare unknown.

Studies by the ME DMR in the late 1970s estimated that the ice-fishery on the Kennebec River extracted less than 5% of the total smelt popula-tion in the river (Flagg 1983). In Maine there is also a large recreational dip netfishery that targets adult smelt on the spawning grounds during the spring runs.60

• ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN While there is a limit of 2 quarts of smelt per person per day in this springfishery, the contribution to mortality is unknown.Incidental catch in small mesh fisheries Five small mesh fisheries operate in the Gulf of Maine, all capable of en-countering rainbow smelt. Because smelt is not a regulated species for federally permitted fisheries, incidental catch (bycatch) is not required to be reported, although it is in some cases. Thus, it is difficult to determine the total amountof smelt bycatch;

however, the relative impact on the species can be assessedbased on reports from the Northeast Fisheries Science Center Observer (NEF-SC) Program, which monitors catch from a representative sample of each fleet(NEFCS 2012). The following analyses represent all Gulf of Maine states.The Northern shrimp fishery operates in nearshore coastal waters duringthe winter and early spring months. Since 1992, the fishery has been requiredto install a finfish excluder device in their nets, the Nordmore grate. Prior to1992, total bycatch in this fishery comprised almost two-thirds of the catch(Howell and Langan 1992). Subsequent surveys have found that the grate isextremely effective in limiting bycatch; Eayrs et al. (2009) observed reductions to 4-8% of the total catch over a two-year period.Using NEFSC observer

records, the effect of the Nordmore grate on reduc-ing smelt bycatch can specifically be seen. In the period directly preceding therequirement of the excluder device (1989-1992),

there were 197 observed tripson vessels targeting Northern shrimp, and smelt were caught on 38 (19%) ofthese trips. A total of 201 lbs of smelt were caught during these trips combined, for an average of 5.3 lbs per trip. The highest was 46 lbs of smelt bycatch,although 87% of these trips caught less than 10 lbs. In the period directlyfollowing the excluder panel requirement (1993-2006),

the amount of smeltbycatch on observed trips decreased, although not significantly (Wilcoxon ranked sum test: p = 0.129 > 0.05). During this period, smelt were observed on74 (24%) out of 303 observed trips. A total of 289 lbs of smelt bycatch werecaught during these trips, with an average weight per trip of 3.1 lbs. The high-est smelt catch was 31 lbs, and 92% of these trips had less than 10 lbs. Recentdata (2007-2011) show that smelt bycatch has decreased significantly from thelast two time periods (Wilcoxon ranked sum test: p < 0.0001 < 0.05). Duringthis most recent period, smelt bycatch was observed on only 22 162 (14%)observed trips, all of which saw less than 10 lbs. The average smelt bycatch forthis recent period was 0.5 lbs, with a maximum catch of 2 lbs.Vessel Trip Reports (VTRs) were implemented in 1996, at which point itbecame mandatory for vessels to report all catch. From the VTR reports, smeltwere only reported in the shrimp fishery post-2006, but reported annuallysince then. From 2006-2011, smelt were reported in 35 trips out of 14,339trips (0.2%). Of the trips that did report smelt, the average catch was 5.3 lbs,the highest 100 lbs (one occurrence),

and 94% of trips reported less than 10lbs. Further work is needed to estimate the total amount of smelt taken in theshrimp fishery using both observer and VTR data.The mackerel, whiting (silver hake), Atlantic

herring, and loligo squidfisheries are all also capable of encountering smelt as bycatch.

These fisher-ies operate on multiple scales with various gear types, including pound (trap)nets at fixed locations close to shore, offshore

trawling, and bag netting.

SmeltANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN

• 61 bycatch has been reported on VTRs in the Atlantic herring and whiting fisher-ies, however too few reports have been given from the mackerel fishery to drawany inferences, and no smelt bycatch has been reported from the loligo squidfishery.In the Atlantic herring fishery, some smelt bycatch was reported in eachyear 1996-2011, although was reported on fewer than five reports in 1997,2002, and 2008-2011.

For the total period, smelt were reported in 135 tripsout of 5463 total Atlantic herring trips (2.4%). The average reported catchwas 5.1 lbs, the highest was 100 lbs (one occurrence),

and 84% of these tripsreported less than 10 lbs.In the whiting (silver hake) fishery, smelt bycatch was reported for 71 tripsout of a total of 20,204 trips (0.3%) for 1996-2011.

In seven of these years,fewer than 5 VTRs reported smelt (1999, 2004, 2005, 2008-2011).

The aver-age reported catch was 6.4 lbs, the highest was 42 lbs, and 73% of these tripsreported less than 1 Olbs.If these data are representative of smelt bycatch in these fisheries, it is likelythat they are not having a large effect on smelt populations at this time.However, because we do not have a population estimate for smelt, it is notpossible to ascertain the mortality rate due to bycatch in these fisheries.

Further,the effect of small-mesh fisheries in the past cannot be determined.

To fullyunderstand the effect of small-mesh fisheries on smelt populations, more workis necessary to ensure that the observer and VTR programs are accurately capturing the extent of smelt bycatch.Predator-prey relationships Prey Availability Rainbow smelt are voracious feeders on amhipods, euphausiids, mysids,shrimps, marine worms, and any available small fishes (e.g., silverside, mummi-chog, herring)

(Scott and Scott 1988). We do not know of existing broad-scale data to evaluate changes in the prey of rainbow smelt over time, however, theprey base was likely affected by changes in primary production and zooplank-ton community composition during the'1990s (Greene et al. 2012), and suchvariability should be expected as a result of oceanographic and climate variabil-ity. In addition, the balance between small prey species and larger fishes mayshift as a result of ocean acidification (Wootton et al. 2008), which will likelyaffect calcifying organisms such as zooplankton and shrimp.Predator Population ShiftsPredators of rainbow smelt include a variety of aquatic birds (e.g.,mergansers, cormorants, gulls, terns), fish (e.g., Atlantic cod, Atlantic salmon,striped bass, bluefish),

and seals (Collette and Klein-MacPhee 2002). Whilethe abundance of some of these predators has declined since the 1990s, othershave increased.

For example, striped bass populations have increased dramati-cally over the past 20 years, although the recovery has not been seen consis-tently along the coast. Maine striped bass populations have actually declinedor remained at low levels compared to other regions (ASMFC 2011). Stripedbass predation has been shown to have a significant impact on blueback herringpopulations in Connecticut River, and has been attributed as one of the factors62 , ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN limiting blueback herring restoration in this river (Davis et al. 2009). Similarly, populations of grey seals in the Gulf of Maine have increased dramatically overthe past few decades (NEFSC 2010). Like striped bass, grey seals are capableof ingesting large amounts of forage fish, and are found feeding in nearshore coastal waters in late spring when smelt are present in large schools.

Althoughnot as closely documented, cormorant populations have also sharply increased in recent years and are known to prey heavily on smelt. Striped bass, cormo-rants, and grey seals have received protections as managed species that haveincreased their populations sharply in short periods of time. Although theseare natural predators that smelt have coexisted with while adapting to Gulf ofMaine environments, it is possible that the impact of increasing predation ondeclining smelt populations results in proportionally higher natural mortality than in the past.Recent shifts in predator range may also increase the exposure of smeltto predators.

Friedland et al. (2012) suggested that the survival post-smolt Atlantic salmon may be affected by increasing predator abundance in the Gulfof Maine; increasing predator abundance that is due not necessarily to increas-ing population size, but to northward shifts in range due to recent changesin climatic and oceanic conditions.

Because many of these species prey on awide range of forage fish, this increasing predator abundance may affect smeltpopulations as well, although more research would be necessary to assess thisrelationship.

Community shiftsDramatic declines of diadromous fish populations have been observedacross North America (Limburg and Waldman 2009; Hall et al. 2012).Saunders et al. (2006) proposed that coherent declines within a co-evolved diadromous community could negatively affect individual species.

WhileSaunders et al. (2006) focused on benefits that may have been lost for Atlanticsalmon through community-level shifts, several of these could also affect rain-bow smelt. In particular, the decline of species such as alewives, blueback her-ring, and American shad-which are present in rivers and estuaries as juveniles during the same time as rainbow smelt-could have resulted in the loss of aprey buffer for rainbow smelt juveniles, making them more vulnerable topredation.

Climate-driven environmental changeIt is anticipated that climate change will influence temperature and pre-cipitation patterns in New England, and some of these effects may already beevident in recent environmental trends. Surface water temperature has beenmonitored monthly nearly continuously since 1905 (ME DMR 2011). Thistemperature series shows periods of warming during the 1940s-1950s and againfrom the 1990s to mid-2000s, with the warmest water on record observed in2006 (Figure 2.3.1). Because smelt are a cold water species, their geographic distribution shift northward may be influenced by the trend in warmer waters.In addition to warmer coastal waters, freshwater conditions have changedin recent years as well. During the 1980s and 1990s, the Northeast experi-enced an increase in heavy precipitation events, and warmer temperatures havereduced ice cover and prompted earlier spring flows (Hodgkins et al. 2003,ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN

• 63 Frumhoff et al. 2007). On New England streams that are substantially affectedby snowmelt, the winter/spring center of volume dates and peak flow datesadvanced by 1-2 weeks between 1970 and 2000 (Hodgkins et al. 2003). Watertemperature and flow changes may affect spawning migration timing (Juaneset al. 2004, Ellis and Vokoun 2009), development rates, and early life stagesurvival in rainbow smelt. More research is needed to understand how climate-related environmental changes influence smelt abundance and distribution changes and to anticipate future implications for rainbow smelt.With concern to species communities and shifts that are due to climatechange, evidence suggests that the balance between small prey species and largerfishes may shift as a result of ocean acidification (Wootton et al. 2008). As theamount of atmospheric carbon increases, the amount of dissolved carbon inoceanic water also increases, in turn decreasing the pH of seawater.

At lowerpH values, the development and survival of calcifying marine organisms likecoralline algae and phytoplankton are inhibited.

Because these organisms arethe base of the marine food chain and the direct diet of many of smelts' preyspecies, a decline in these organisms may also negatively affect smelts' prey base.This hypothesis has been examined on the Pacific coast, but with no conclu-sive results, and has only begun to be considered in the Gulf of Maine. Moreresearch is needed to fully understand the effect of climate change on speciescomposition changes in this region.131210* 10iE7Figure 2.3.1. Mean annual 6surface water temperature at U) to to U) to L U) Ul) U) U') to0 +_- (" i~ t U) to 1'- 03 0) 0Boothbay Harbor, Maine, from O 0) 0) 0) 0) 0) 0) 0) 0) 0) 01905-2010.

64

• ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN 3 -CONSERVATION STRATEGIES We recommend that rainbow smelt remain federally listed as a Species ofConcern.

Populations have disappeared from their southern range in a shortperiod of time and are also declining in their present distribution in the Gulf ofMaine. The species should continue to be monitored, and factors contributing to its decline should continue to be assessed.

3.1 -REGIONAL CONSERVATION STRATEGIES Recommendation 1: Continue monitoring programsEach state within the present distribution of rainbow smelt in the Gulf ofMaine currently monitors populations through inshore trawl, juvenile abun-dance, fyke net, and/or creel surveys.In states at the extreme southern limit of the range where spawningpopulations have not been documented within the past ten years, inshore trawlsurveys are likely the most effective way to monitor the remnant populations.

In the Gulf of Maine states, trawl surveys provide the only source of data onthe marine life phase of smelt. It is necessary that these surveys continue todocument smelt presence and quantify abundance, and it is recommended thatbiological information is collected from a sub-sample of catches.The regionally standardized fyke net survey developed for this study shouldbe continued in the Gulf of Maine. A standardized survey is necessary toprovide long-term data that can track inter-annual variability across distinctspawning stocks. This information is critical for detecting whether populations are declining or showing signs of stress, as may be characterized by truncated age distributions, decreases in length at age, and decreases in CPUE over time.The juvenile abundance surveys should also be continued in New Hampshire and Maine as the only surveys targeting this life stage. Further, creel surveysshould be maintained at recreational fishing sites to provide a measure of theimpact of the fishery as well as information about changes in population sizeand biological characteristics over time.Because some pathological concerns were found as part of this project (seesection 2.3 -Threats to Smelt in Marine Coastal Waters),

Gulf of Maine statesshould periodically monitor rainbow smelt from multiple spawning stocksfor pathology, including parasite occurrence, viral agents, and systemic physi-ological problems.

Further, states should cooperate with Canadian provinces tocompare parasite and disease prevalence in the entirety of the species' range.We recommend thatrainbow smelt remainfederally listed as aSpecies of Concern andthat current population monitoring effortscontinue in the Gulf ofMaine.ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN

• 65 Restoring in-stream habitat (e. g. substrate, water volume andvelocity, pool and riffleareas), riparian buffer,improving and preserv-ing watershed functions, and restoring access areimportant management strategies to improvelocal smelt populations.

Recommendation 2: Restore historical or degraded spawning habitatSpawning habitat degradation and obstructions to access have been identi-fied as two important factors that have reduced successful spawning.

Restoring in-stream habitat (e. g. substrate, water volume and velocity, pool and riffleareas), riparian buffer, improving and preserving watershed functions, andrestoring access are important management strategies to improve local smeltpopulations.

Where possible, head-of-tide dams should be removed.

Eggs deposited be-low dams are subject to periods of salinity during high tide and may be exposedto air at low tide if freshwater flows coming over the dam are low. Perchedculverts and small water control barriers can also have this effect. When theseobstructions are removed, smelt are able to ascend into freshwater, where waterchemistry is more stable over time and water level is relatively constant.

Whileundersized culverts (less than 1.2x bank-full width) may not completely blockaccess, they can limit the number of smelt that reach the spawning grounds bycreating velocity barriers.

Restoration projects to improve road-stream cross-ings should design replacement culverts that target minimum water depth of 6inches with average velocities in the culvert of 0.5 m/s or less, and flood veloci-ties below 1.5 m/s (see section 2.1 -Threats to Spawning Habitat Conditions and Adult Spawning).

Additionally, water quality at the spawning grounds must support healthyembryonic development and survival.

We found that diminished rainbow smeltspawning runs existed in rivers surrounded by urbanized watersheds, whilerivers draining forested watersheds supported strong smelt spawning popula-tions. Comparing watershed conditions to water quality, higher concentrations of nutrients and toxic contaminants were associated with developed areas,while highly forested watersheds were associated with lower concentrations of nutrients and metals. This pattern suggests that protecting forested areasis important for maintaining water quality conditions that are beneficial torainbow smelt. Furthermore, regional efforts to purchase conservation landsshould consider parcels in watersheds that support smelt spawning habitats.

When development does occur in watersheds with smelt spawning

habitat, theamount of impervious surface should be minimized, and stormwater mitigation techniques should be implemented to curtail the impacts on water quality (e. g.riparian

buffers, vegetated stormwater retention pools, underground filtration

systems, etc.).Recommendation 3: Smelt Fishery Management ActionsThe results of the present study documented evidence of high population mortality (truncated age distribution) and poor recruitment (low abundance) in smelt populations in the southern portion of the study area. The time seriesof population data collected among the fishery dependent and independent surveys is too brief to determine the causes of these stressors on smelt popula-tions. However, overfishing was consistently identified as a significant concernin the latter half of the 19th century and the early 20th century in the southernportion of smelt's distribution.

The sustainabiliry of current smelt fisheries, both recreational and commer-cial, will require management strategies to quantify natural mortality and fish-66 -ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN ing mortality.

We recommend that each state in the study area review currentsmelt fishery regulations and identify locations where present management maynot be sufficient to protect distinct populations that display evidence of stress.We recommend that states estimate fishing mortality from all targeted smeltfisheries and review bag limits on both commercial and recreational fisheries that target smelt.Recommendation 4: Expand research to estimate population sizeand assess the potential impacts of ecosystem and climate changesThe surveys carried out as part of this project did not enable us to developa population estimate for rainbow smelt. However, the standardized fyke netsurvey established by the study should be continued with additional researchin order to assess smelt population status in the region, understand the im-pact of targeted fishing and incidental

bycatch, and to understand the relativecontributions of each spawning stock to the regional population.

This may beaccomplished through a large-scale mark and recapture effort that targets eachgenetic stock (Kovach et al., in press; section 1.1 -Basic Biology).

Taggingstudies carried out as part of this project to understand habitat use and within-season repeat spawning behavior documented few inter-annual returns (lessthan 1%), although approximately 200 smelt per year were tagged (assumed tobe less than 10% of the entire run based on estimated fyke net catch efficien-cies). Future tagging studies should tag a representatively larger sample of thespawning population to effectively monitor inter-annual repeat spawning andestimate population size. Additionally, improved and validated age structure data are needed to support future estimates of population size. Efforts shouldbe made to maintain sufficient age structure sample sizes in each state.Further research is needed to understand how changes in prey availability and predator abundance affect smelt populations.

Other studies have foundconnections between increasing predator populations and depressed forage fishpopulations (see section 2.3 -Threats to Smelt in Marine Coastal Waters).Because these studies looked at predators that also feed on anadromous smelt,the impact on smelt populations should also be examined.

Species that are important prey of rainbow smelt may be particularly af-fected by changes in the chemistry of marine waters. Increases in the amount ofcarbon in the atmosphere are associated with increases in the amount of carbonin salt water, which leads to a reduction in oceanic pH that may negatively im-pact small prey species, such as calcareous plankton (Wooton et al. 2008). Thisrelationship needs to be better quantified to understand the effect of a smallerprey base on smelt populations.

Conversely, predator populations that haveshifted in their range in response to climate conditions may be preying uponforage fish populations more than in previous times (Friedland et al. 2012).Further studies are necessary to understand how rainbow smelt will be affectedby changes to their prey and predators as a consequence of climate change.Climate change may also impact smelt populations by changing the extentof available spawning areas. Smelt spawn directly above the head of tide, andthe upstream extent of the freshwater spawning area is typically either a natu-ral barrier or road crossing.

Thus, a rise in sea level that extends the tidal limitto these barriers may greatly reduce the number of spawning sites or the areaANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN

• 67 Expanded research tounderstand reasons forsystemic health issuesand reduced survival isneeded to effectively guide management actions.within sites that is suitable for spawning.

Conversely, a rise in sea level couldincrease habitat by raising tidewater above natural barriers allowing access tonew reaches.

Future research should model the potential effects for various sealevel rise projections.

Expanded research to understand reasons for systemic health issues andreduced survival is needed to effectively guide management actions.

While it ishelpful to understand overall relationships such as watershed composition andsmelt population responses, it is only a starting point. For example, researchinto dose responses to specific water quality constituents at all life stages wouldenable managers to develop smelt specific water quality criteria.

These criteriamay then be used to guide water treatment goals around which non-point orpoint source controls can be designed.

This would be especially important inthose already developed watersheds that are impractical to restore to forest.Controlled studies in both laboratory and field settings are critical to improveour understanding of cause and effect, not just correlations, and to developmeasureable relationships.

Lastly, post-restoration monitoring is necessary toevaluate the success of any prescribed restoration technique.

Recommendation 5: Implement stocking of marked larvae, withcontinued monitoring and genetic considerations Rainbow smelt are currently extirpated or have severely declined in manycoastal rivers and streams that once supported robust spawning populations.

Historical fishing pressure at the spawning grounds and degraded habitat andwater quality may be causal factors.

When improvements are made to waterquality and habitat in these streams, restoration practices, such as stocking, maybe appropriate to re-establish rainbow smelt runs at these sites.Successful stocking efforts must include marking and subsequent recapture of hatchery stocked smelt to quantify effectiveness of restoration efforts.

Utiliz-ing recent advances in smelt culture techniques, Ayer et al. (2012) developed methods for marking otoliths in larval rainbow smelt with oxytetracycline (OTC) for monitoring returns.

Using these methods, the Massachusetts Divi-sion of Marine Fisheries began a pilot program to stock OTC-marked smeltlarvae in the Crane River, MA, after water quality suitability was confirmed andpassage improvements were made to upstream spawning habitat (Chase et al2008). Over 10 million marked smelt larvae have been stocked into the CraneRiver since 2007, and spawning adult smelt with OTC-marked otoliths havebeen recaptured, providing a positive response for the project to continue stock-ing and monitoring.

New restoration sites for rainbow smelt are being examined in both Mas-sachusetts and Maine. In many situations, the protection and enhancement of existing habitat and water quality at both donor smelt runs and potential stocking sites will be preferential to initiating a stocking effort. Before anystocking begins, these sites will be sampled for baseline population data, and asite suitability assessment will be conducted, which will include water qualitymonitoring, streambed characterization, and flow measurements.

Further, thegenetic information presented in this plan (section 1.1 -Basic Biology) must beused in determining the appropriate parent stock. Managing at too fine a scalecan lead to reduced allelic diversity and ignores the natural occurrence of gene68

• ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN flow, while managing at too large a scale can reduce genetic diversity and ignorelocal adaptations.

Another important consideration is the status of donor popu-lations to support stocking efforts.

Careful planning should be made to removea minimal proportion of a donor smelt run's productivity for stocking.

Finally,long-term post-stocking monitoring should be performed to demonstrate stocking success.3.2 -STATE MANAGEMENT RECOMMENDATIONS Massachusetts Massachusetts has a long history of implementing management measuresto ensure sustainable smelt fisheries.

Concern over the capability of net fisher-ies during smelt spawning runs to negatively impact the long-term viability ofsmelt runs was documented in the 1860s (Kendall 1926). In 1874, the Massa-chusetts state legislature banned harvest using nets during the spawning periodand limited harvest to hook and line for most coastal rivers in Massachusetts.

By the start of the 20th century, nearly all smelt runs had this protection, andlocal smelt fisheries continued mainly as sportfisheries with little change untilrecent decades.The only location in Massachusetts that presently allows net fishing forsmelt during the spawning run is the Weweantic River in Wareham.

This fish-ery is conducted under authority of M.G.L 67 of 1931 that gives the Town ofWareham the responsibility to manage a smelt fishery from March 1 to March31. This recreational fishery continues today with a 36 smelt/day bag limit foreach permitted fisherman and limits the net size to 5 square feet. This locationwas monitored as a smelt fyke net station during the present study. The smeltcatch at the Weweantic River station had low CPUE for Massachusetts riversand a size composition dominated by the age-I mode. MA DMF intends toinitiate cooperative efforts with the Town of Wareham to ensure this uniquesouthern smelt run can be sustained.

Following the net bans of the 19th and early 20th centuries, no smelt lawsor regulations were made in Massachusetts until 1941 when three provisions were added to M.G.L. Chapter 130 that focused specifically on smelt fisher-ies. Section 34 of Chapter 130 standardized the spawning run ban for harvestduring March 15 to June 16. Section 35 standardized the method of harvest tohook and line only in Massachusetts.

Section 36 gave the Division of Ma-rine Fisheries authority to close smelt spawning river beds to entry during thespawning season. Following these three laws, no changes to smelt regulations were made until 2009 when a daily bag limit of 50 smelt per angler was adopt-ed. Unlike Maine and New Hampshire that drafted smelt management plansin the 1970s and 1980s, no such plan has been prepared in Massachusetts.

Declining recreational smelt catches in the 1980s prompted a review of thestatus of smelt fisheries and spawning runs by the MA DMF A survey of allcoastal drainages on the Gulf of Maine coast of Massachusetts was conducted from 1988-1995, during which 45 smelt spawning locations were documented and mapped in 30 coastal rivers (Chase 2006). The report for this survey in-cluded specific habitat and water quality recommendations for each smelt run.Following the survey, effort was directed toward acquiring smelt population ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN -69 data. A grant was received from NOAA's Office of Protected Resources to de-velop fyke net indices at six smelt runs during 2004-2005 (Chase et al. 2006).This approach and the six fyke net stations were adopted for the present study.These contemporary

efforts, when compared to the historical records and fish-ery accounts from the 1960s and 1970s, present evidence of a sharp decline inMassachusetts smelt populations in the past 2-3 decades.

Locations that oncesupported popular winter ice fisheries for smelt no longer have fisheries, andsome known spawning runs have had no recent evidence of spawning activity.

Smelt Stocking EffortsThe transfer of smelt eggs from larger donor smelt runs to smaller runs orrivers with no smelt spawning was a common practice late in the 19th centuryin Massachusetts, followed by a large dedicated effort during 1910 to 1920(Kendall 1926). The ease with which smelt eggs could be collected and theappearance of large numbers of excess eggs in some settings contributed to thezeal behind decades of stocking.

Unfortunately, documentation of responses to stocking is essentially absent, other than brief narratives in annual agency re-ports. Short-term increases in smelt spawning run size appear to have occurredin some systems, especially for coastal to inland lake transfers.

However, noevidence can be found of long-term benefits of coastal to coastal river transfers.

Smelt egg transfers continued periodically through the 1980s with strong sport-fishing constituency support.

Recent requests to stock smelt eggs led to a MADMF evaluation that attempted to quantify the number of eggs transferred, egg survival and returning adult smelt (Chase et al. 2008). Returning spawningadults were documented in a pilot river with no smelt run during the first yearof possible

returns, but low egg survival and expected low recruitment conclud-ed with MA DMF discouraging the use of smelt egg transfers and prioritizing

passage, water quality, and habitat quality improvements over stocking as meth-ods for restoring smelt populations.

MA DMF presently does not support theuse of egg transfers but is conducting a pilot study on the stocking of oxytetra-cycline marked larvae as a potential substitute for egg stocking in specific caseswhere population enhancement can be coupled with habitat improvements andmonitoring.

Habitat Restoration The survey of smelt spawning habitat provided recommendations forspecific habitat improvement projects (Chase 2006), four of which have sincebeen conducted.

Each of these projects has focused on improving spawningsubstrate.

Two of these projects were able to take advantage of planned culvertreplacements to add substrate improvements as part of the scope of work, whilethe other projects specifically targeted grant and mitigation funds to augmentspawning substrate.

The experience gained from these projects will assist futureefforts in the region.Recommendations

1) Apply the information gained from the present study and recent smelthabitat improvement projects to identify potential restoration sites and designsmelt spawning habitat improvements that meet the life history requirements of smelt. Projects that can remove barriers and extend habitat connectivity forsmelt and other diadromous fish should be prioritized 70 e ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN

2) Continue monitoring smelt fyke net stations from the present studythat have been identified as having promise to support long-term indices ofabundance (i.e., Weweantic River, Jones River, Fore River and Parker River).Improve and maintain data collection at fyke net stations to support futuredevelopment of biological population benchmarks

3) Develop water quality criteria that relate to designated uses within theMassachusetts Wetlands Protection Act in order to protect the specific habitatsof anadromous fish, including smelt spawning habitat4) Conduct a smelt habitat survey of the Buzzards Bay region ofMassachusetts that was not mapped during the previous Gulf of Maine surveyin Massachusetts

5) Develop a state smelt conservation plan similar those completed forMaine (1976) and New Hampshire (1981)New Hampshire The recreational smelt fishery in New Hampshire has been monitored andregulated for decades, and current fishing pressure is not believed to pose a ma-jor threat to the smelt population in the state. Ensuring that fishing pressure iscompatible with a sustainable smelt population requires continuing monitoring efforts that are already underway, including creel surveys, spring spawning runsurveys, and biological sampling during the ice fishery and young-of-the-year seine surveys.

Current monitoring of the fishery does not capture recreational fishing for smelt that occurs in the fall prior to the onset of ice. There is alsoa limited hook and line commercial fishery for smelt in New Hampshire withlocal markets that is not well recorded.

Developing surveys that obtain datafrom these portions of the fishery would be helpful for appropriately charac-terizing fishing related mortality.

Currently, the daily limit for recreational smelt fishing is 10 liquid quarts, which is approximately equivalent to half of a5 gallon bucket. Given that smelt is a species of concern, this limit would bere-evaluated if in the future fishing pressure is believed to pose a major threatto the population.

Neighboring states of Maine and Massachusetts, which havelarger smelt runs, have a daily limit of 2 quarts and 50 fish, respectively.

Population monitoring The most current statewide fisheries management plan for rainbow smeltwas written in 1981, but it predominately focuses on lake smelt populations.

The objectives for smelt management were to maintain or increase the popula-tion of smelt and to provide for commercial and recreational fisheries.

Man-agement measures implemented following development of the plan includedclosure of the fishery to net or weir fishermen from March 1 to December 15,a 10 quart daily possession limit, and implementation of a smelt egg transferprogram that occurred intermittently until 1991.To evaluate the effectiveness of the management measures and detecttrends in smelt abundance, an annual creel survey of the recreational ice fisherywas implemented, and a smelt egg deposition index was developed.

Data havebeen collected for the smelt egg index from 1979-2006.

The intent of theindex was to provide a fisheries independent relative measurement of spawningstock abundance.

Validation of the index was attempted in 1993 by regressing ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN

• 71 it with catch per unit effort of the winter fishery, but results showed very poorcorrelation between the two. The Department also compared data from thecreel survey with the abundance of young of the year (YOY) rainbow smelt col-lected via a seine survey that was initiated in 1997. This comparison resulted ina much stronger correlation with age-2 smelt CPUE from the creel survey. TheDepartment discontinued egg deposition surveys in 2006 as a result of poordata correlation with other surveys, but will continue to monitor rainbow smeltthrough juvenile abundance

surveys, creel surveys, as well as spawning surveysat the fyke net index stations that were implemented for this project.Habitat Restoration Improving water quality in the Great Bay Estuary is expected to benefitsmelt using New Hampshire waters. An increase in the concentration of dis-solved nutrients and substantial increases in nutrient loading have been de-tected in the estuary in recent years. These observations prompted the NewHampshire Department of Environmental Services (NH DES) and the U. S.Environmental Protection Agency to develop nutrient criteria for the estuary.Applying these criteria will result in water quality being classified as impaired inthe entire estuary, including all of its tributaries.

These noted nutrient increases have the potential to spur periphyton growth, which may reduce the viability and hatching of smelt eggs, as discussed in section 2.2 -Threats to Embryonic Survival and Development.

The current nutrient criteria assessment is motivat-ing local action to reduce nutrient

loading, which should result in improvedwater quality and reduced periphyton during the smelt spawning season.Habitat assessment and restoration are key conservation strategies thatwill be pursued in New Hampshire to enhance spawning conditions for smelt.While main stem spawning habitats are well known in the major tributaries toGreat Bay, a comprehensive assessment of other potential spawning locations in smaller tributaries would be beneficial.

Habitat improvement projects thatwould benefit smelt include mitigating siltation and removing head-of-tide dams to increase the amount of freshwater area available for spawning.

Cur-rently most spawning in New Hampshire occurs in intertidal areas. Intertidal bars have developed in some tributaries following recent flood events; smelteggs are deposited on these rocky bars and are then exposed to air at low tide.Grading of these bars to minimize their intertidal exposure would reduce eggmortality.

In addition, head-of-tide dams currently block smelt migration on most ofthe major tributary rivers to Great Bay. One of these obstructions has recentlybeen removed; the dam in place for 55 years on the Winnicut River in Green-land, NH, was recently demolished, restoring spawning habitat for smelt.Following the dam's construction in 1957, there was a steady decline of a oncewell-known large smelt run. Other head-of-tide dams in the Great Bay Estuaryare under consideration for removal.

The potential benefits to smelt will be akey factor in deliberations about the future options for these dams.Finally, siltation in some rivers has reduced smelt spawning habitat.

Damremoval should increase stream flows and help remove accumulated sediments, and actions to reduce nutrient inputs will also reduce sediment inputs to theGreat Bay Estuary and its tributaries.

These actions should improve smeltspawning habitat conditions in the tributaries.

72 e ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN Recommendations:

1) Continue monitoring efforts in place including:

winter creel survey,juvenile abundance seine survey, spring spawning run fyke net sampling2) Improve water quality and support NH DES in developing nutrientcriteria for Great Bay Estuary3) Identify habitat restoration projects to enhance smelt spawning conditions.

4) Continue to support dam removal projects to connect smelt to historical spawning habitats5) Conduct a smelt spawning habitat assessment of coastal areas in NewHampshire.

MaineThrough this project, we have found that while rainbow smelt populations are contracting rapidly in range, there are still strong populations in Maine.However, our surveys have also shown that smelt populations in the state arenot as strong as previous Department studies have found. Comparing the num-ber and strength of spawning runs currently to that of the late 1970's, we havefound that many runs have declined, while others are extirpated (see section1.3 -Population Status).

Data collected during our fyke net survey and creelsurveys has also shown that length at age has declined compared to historical records in upper Casco Bay and Kennebec River populations.

Because smeltcontinue to support an economically important and sizable recreational fisheryin Maine, as well as a locally economically important commercial fishery inWashington County, it is imperative to pursue management measures that willsustain and restore this species.Continue monitoring smelt populations at multiple life stagesThe state surveys that are currently in place target four important lifehistory stages for rainbow smelt. The annual fyke net survey, which began in2008, monitors the adult spawning runs at six index sites spanning the Mainecoast. From this survey, we collect information about the inter-annual variabil-ity of the spawning stock, the strength of age classes, and mortality rates. Thegenetic information combined with movement and habitat studies show thatwhile adult smelt may not home to the same stream each year, they do showfidelity to larger bay and estuary areas. Thus, by monitoring adult smelt duringthe spawning season, we can observe changes in a specific stock over time. Theother surveys do not have this ability.

While the inshore trawl survey can trackrelative population abundance over time, it likely catches mixed genetic stocksand annual CPUEs may be skewed by stock variability.

The creel survey that targeted the Kennebec River and Merrymeeting Bay beginning in 2009 was expanded with the help of the Downeast SalmonFederation in 2010 to survey anglers on the Pleasant and Narraguagus rivers.Flagg (1984) estimated an extraction rate of less than 5% on the KennebecRiver in the late 1970s. However, the population during that time period waslikely larger than at present (see section 1.3 -Population Status in the Gulf ofMaine); the fishery may have a more significant effect when population levelsare low. Given the cultural and economic value of these fisheries, the creelANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN -73 Local smelt runs may beaffected by a combina-tion of factors, including habitat degradation, access problems, andcurrent fishing practices.

survey should be expanded to target aggregations of fishing camps in otherlocations (e.g., Great Salt Bay on the Damariscotta River), and efforts should bemade to repeat the mark-recapture survey performed by Flagg (1984) to deter-mine a current extraction rate.The juvenile abundance survey is extremely important in understanding the reproductive success and early life stage survival in the Kennebec River andMerrymeeting Bay. Because we also monitor adult populations in this riversystem through creel surveys, it may be possible to compare data from thetwo surveys to quantitatively link adult winter catches to late summer juvenileabundance as NHF&G has been able to do. Additionally, by further under-standing how juvenile abundance varies between river segments, we may be ableto identify important juvenile habitat.Improving connectivity and access to spawning groundsIn many locations where smelt runs have historically declined or disap-peared on the Maine coast, the decline is due to the inability of smelt to reachthe spawning grounds.

Road crossings on small coastal streams are oftenprovided by undersized or hanging culverts or by small historic water controldams that no longer have purpose.

Undersized culverts present problems whenvelocities increase during rain events because the water is constricted to a widthsmaller than the natural streambed.

Because smelt are not strong swimmers, high water velocities can impede their ability to swim through the culvert,and thus to reach their spawning grounds.

Hanging culverts (those where thedownstream water level is lower than the culvert height) and dams that aredownstream of the spawning grounds completely block access. Unlike otheranadromous fishes (e.g., alewife and salmon) that can ascend fish ladders orjump vertical obstructions, smelt are unable to pass vertical obstructions oversix inches.State agencies in Maine, including ME DMR, are currently working tocatalogue such obstructions and prioritize which should be removed or rede-signed to allow for anadromous fish passage.

As part of this effort, a web-based tool will be publicly available so that municipalities and land trust organizations can identify road crossings in their area where improvements could re-establish smelt habitat access. In many cases, removing these barriers can have immedi-ate effects in opening smelt spawning passage into a stream when strong runsexist nearby. If this is not the case, stock enhancement may be considered inthe absence of other habitat degradation.

The ME DMR will continue to workwith other state agencies, municipalities, and non-governmental organizations to identify barriers to historical smelt habitat and restore access.Assessing causes for local declineSome smelt populations in Maine have declined or become extirpated, while others remain strong. In some cases, local declines can be attributed tohistorical overfishing;

however, habitat degradation, access problems, and cur-rent fishing practices may also be impacting smelt populations in the state.Effective stormwater management techniques can reduce the impact of de-velopment on water quality in urbanized watersheds in the state. As an exam-ple, the Maine Department of Environmental Protection has worked with theSouth Portland Water District and businesses within the Long Creek watershed 74 -ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN to build stormwater retention areas that reduce the amount of nutrients andcontaminants flowing directly into the stream. While the stream quality stillshows the effects of development, impairment is reduced and the stream is ableto support a limited smelt spawning run. Because this regional smelt projecthas found that development within a watershed can impact water quality tothe point where smelt embryonic health and survival are impaired, watershed management efforts that reduce runoff into receiving streams are recommended in urbanized or developing watersheds.

Current fishing regulations regarding anadromous rainbow smelt limit takeby season and location.

Recreational fishing is allowed July 1 through March14; there is no catch limit, but the gear is restricted to hook and line or dipnet. During the spawning season (March 15 through June 30), take is limitedto two quarts per person per day, and it is predominantly a dip net fishery.While the state Marine Patrol does actively enforce this regulation regarding gear and catch limitations, the number of violations that go without reprimand is unknown.

Further, it is currently unknown what impact the recreational fishery may have on smelt populations.

With the creel survey of the ice fisherybeginning again in 2009, the ME DMR now has the opportunity to assess theextraction rate of the winter fishery and determine if a limit on take is neces-sary. However, at this point there is no survey of the spring dip net fishery; theeffect of fishing mortality during the spawning season and the subsequent lossof possible embryos is unknown.

Future work should include an effort to quan-tify fishing mortality due to both the recreational winter and spring fishery.

Inlocations where there is evidence of stressed smelt runs, management actionshould be considered to limit mortality during spawning runs.Commercial fishing for smelt is allowed in only six tidal rivers in the state,all in Washington County: the East Machias,

Pleasant, and Narraguagus riversfrom January 1 through April 10, without any limit on quantity; and the Indi-an, Harrington, and Chandler rivers with no limit on quantity or time period.Anyone fishing commercially for smelt must possess a Pelagic License from theME DMR. With possession of this license, the fisherman is required to submitlandings data to the ME DMR. The ME DMR is working with DowneastSalmon Federation to survey the biological composition of the catches todetermine if the fishery may be impacting life history or age structure.

Thiscollaboration is necessary to monitor the fishery, and should continue in the fu-ture. If over time there is evidence of smelt population decline in this region orevidence that the commercial fishery may be contributing to a high mortality, management actions should address the fishing effort possibly by limiting takeor further gear restrictions.

Marked larval stocking at monitored sitesAs part of this project, the ME DMR revisited historical spawning runsto document their current status and found that many sites no longer supportspawning or support only limited runs (see section 1.3 -Population Statusin the Gulf of Maine). When the decline at these sites can be attributed tohistorical fishing pressure that no longer exists or to habitat degradation or pas-sage constraints that have been addressed, larval stocking may be an option toreintroduce smelt.Adapting methods by Ayer et al. (2012), the ME DMR began a projectANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN a 75 With continued population monitoring and threat assessment incollaboration withfisheries

managers, university scientists, recreational andcommercial fishermen, and interested

citizens, rainbow smelt popula-tions could be maintained or possibly expanded.

to restore rainbow smelt populations to North Haven, Maine, an island in thecenter of Penobscot Bay that supported robust smelt populations up until the1950s. After visits by ME DMR to identify the most appropriate stream for theproject, the North Haven Community School completed pre-monitoring andfound no water quality impairments that would affect smelt embryo survival.

In spring 2012, the ME DMR and school worked together to mark larvae withoxytetracycline (OTC) for release at the stream. The school and ME DMR willcontinue to monitor adult returns in subsequent years to determine the suc-cess of the project.

Following this model, the ME DMR hopes to continue tore-establish smelt populations at sites where restoration projects have improvedhabitat quality or connectivity.

However, habitat restoration must always pre-cede any stocking efforts.Recommendations With continued population monitoring and threat assessment in collabora-tion with fisheries

managers, university scientists, recreational and commercial fishermen, and interested

citizens, the rainbow smelt populations in MaineI could be maintained or possibly expanded.

To this end, the ME DMR hasbegun to implement restoration

efforts, including a stocking project in NorthHaven and assessment of culvert replacements that would provide access tohistorical habitat.

Future work in the state of Maine to protect this species ofconcern should include:1) Continuing monitoring of smelt populations through fyke net sampling, creel surveys, the inshore trawl survey, and the juvenile abundance survey2) Developing a mark-recapture study to estimate the current extraction rateof recreational ice fishing on the Kennebec River and Merrymeeting Bayand other rivers and embayments that support recreational ice fishing3) Restoring stream connectivity and access to historical spawning groundswith monitoring to assess pre- and post-construction conditions and smeltpopulations

4) Assessing threats to smelt habitat and evaluating connections betweendegraded habitat and local smelt population decline5) Stocking rainbow smelt larvae marked with oxytetracycline intohistorical smelt spawning streams that maintain good habitat, whilemaintaining the genetic structure as identified by this project and annuallymonitoring stocking success.76 ° ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN LITERATURE CITEDAllan, J.D. 2004. Landscapes and riverscapes:

the influence of land use on streamecosystems.

Annual Review of Ecology, Evolution, and Systematics 35:257-284.

ASA A&C (ASA Analysis and Communication).

2010. 2008 year class report for theHudson River Estuary Monitoring Program.

Prepared for Dynegy Roseton L.L.C.by ASA Analysis and Communication, Washingtonville, NYASMFC (Atlantic States Marine Fisheries Commission).

2011. Striped Bass StockAssessment Update. 132 pp. www.asmfc.org Ayer, M.A., S.P. Elzey, and B.C. Chase. 2012. Laboratory marking of anadromous rainbow smelt embryos and larvae and the implications for restoration.

Pages60-62 in The Fourth North American Workshop on Rainbow Smelt: ExtendedAbstract Proceedings.

Massachusetts Division of Marine Fisheries, Technical Report TR-5 1.Bigelow, H.B., and W.C. Schroeder.

1953. Fishes of the Gulf of Maine. U.S. Fish andWildlife Service Fishery Bulletin 53:1-577.

Biggs, B.J.E 1996. Patterns in benthic algae of streams.

Pages 21-56 in Stevenson, R.J., M.L. Bothwell, and R.L. Lowe, editors.

Algal Ecology:

Freshwater BenthicEcosystems.

Academic Press, San Diego, California.

Booth, D., and L. Reinelt.

1993. Consequences of urbanization on aquatic systems:measured

effects, degradation thresholds, and corrective strategies.

Pages 545-550in Proceedings Watershed

'93 A National Conference on Watershed Management.

Alexandria, Virginia.

Bouchard, D. 2010. August 2010 Executive Summary for Atlantic Sturgeon andRainbow Smelt Pathology Analysis.

Maine Department of Marine Resources Cooperative Agreement

1. CT-I 3A-20090126*4283.

Bradbury, I.R., K. Gardiner, P.V.R. Snelgrove, S.E. Campana, P. Bentzen, and L.Guan, (2006b).

Larval transport, vertical distribution, and localized recruitment in anadromous rainbow smelt (Osmerus mordax).

Canadian Journal of Fisheries and Aquatic Sciences 63:2822-2836.

Brenner, EJ., and J.J. Mondok. 1995. Nonpoint-source pollution potential in anagricultural watershed in northwestern Pennsylvania.

Water Resources Bulletin31(6):1 101-1112.

Buckley, J.L. 1989. Species Profiles:

Life Histories and Environmental Requirements of Coastal Fishes and Invertebrates (North Atlantic);

Rainbow Smelt. U.S. Fishand Wildlife Service Biological Report 82(11.106).

U.S. Army Corps of Engineers, TR EL-82-4.

I1pp.Burcher, C.L., H.M. Valett, and E.E Benfield.

2007. The land-cover cascade:relationships coupling land and water. Ecology 88(l):228-242.

Carpenter, S.R., N.E Caraco, D.L. Correll, R.W. Howarth, A.N. Sharpley, and V.H.Smith. 1998. Nonpoint pollution of surface waters with phosphorus andnitrogen.

Ecological Applications 8(3):559-568.

Chase, B.C. 1990. Summary report for the North Coastal Basin. Massachusetts Division of Marine Fisheries, Massachusetts Bay Smelt Spawning HabitatProgram, STAP Doc.90-102.ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN

• 77 Chase, B.C. 2006. Rainbow smelt (Osmerus mordax) spawning habitat on the Gulfof Maine Coast of Massachusetts.

Massachusetts Division of Marine Fisheries, Technical Report TR-30.Chase, B.C. 2010. Quality Assurance Program Plan (QAPP) for Water QualityMeasurements Conducted for Diadromous Fish Habitat Monitoring, Version 1.0,2008-2012.

Massachusetts Division of Marine Fisheries, Technical Report TR-42.Chase, B.C., M.H. Ayer, K.J. MacGowan, and K. Taylor. 2006. Population Indicesof Rainbow Smelt Spawning Runs in Massachusetts.

Massachusetts Division ofMarine Fisheries, Completion Report for NMFS Project No: O-2003-NER4, Gloucester, MA.Chase, B.C., and A.R. Childs. 2001. Rainbow smelt (Osmerus mordax) spawninghabitat in the Weymouth-Fore River. Massachusetts Division of Marine Fisheries, Technical Report TR-5.Chase, B.C., C. Enterline, K. Mills, M.H. Ayer, S.P Elzey, J. Fischer, S.J. Correia, andC.H. Wood. 2012. Rainbow smelt (Osmerus mordax) spawning population moni-toring on the Gulf of Maine coast of New England.

Pages 13-15 in The FourthNorth American Workshop on Rainbow Smelt: Extended AbstractProceedings.

Massachusetts Division of Marine Fisheries, Technical Report TR-51.Chase, B.C., J.H. Plouff, and M. Gabriel.

2008. An evaluation of the use of egg trans-fers and habitat restoration to establish an anadromous rainbow smelt (Osmerusmordax) spawning population.

Massachusetts Division of Marine Fisheries, Techni-cal Report TR-33.Chapman, D.G., and D.S. Robson. 1960. The analysis of a catch curve. Biometrics 16:354368.

Clayton, G.R. 1976. Reproduction, first year growth, and distribution of anadro-mous rainbow smelt, Osmerus mordax, in the Parker River and Plum Island Soundestuary, Massachusetts.

Master's Thesis. University of Massachusetts, Amherst,Massachusetts.

Clayton, G.R., C.E Cole, S.A. Murawski, and J.D. Parrish.

1978. Common marinefishes of coastal Massachusetts.

University of Massachusetts, Amherst,Massachusetts.

Collete, B.B., and G. Klein-MacPhee.

2002. Bigelow and Schroeder's Fishes of theGulf of Maine. Smithsonian Institution Press, Washington, D.C.Cooper, C. 1996. Hydrologic effects of urbanization on Puget Sound lowland streams.Master's Thesis. University of Washington, Seattle.Corbett, C.W, M. Wahl, D.E. Porter, D. Edwards, and C. Moise. 1997. Nonpointsource runoff modeling:

a comparison of a forested watershed and an urbanwatershed on the South Carolina coast. Journal of Experimental Marine Biologyand Ecology 213:133-149.

Crawford, J.K., and D.R. Lenat. 1994. Effects of Land Use on the Water Quality andBiota of Three Streams in the Piedmont Province of North Carolina.

Hydrobiolo-gia. 294:185-199.

Crestin, D.S. 1973. Some aspects of the biology of adults and early life stages of therainbow smelt, Osmerus mordax (Mitchill),

from the Weweantic River Estuary,Wareham-Marion, Massachusetts, 1968. Master's Thesis. Boston University, Boston, Massachusetts.

Dauvin, J.C., and J.J. Dodson. 1990. Relationship between feeding incidence andvertical and 644 longitudinal distribution of rainbow smelt larvae (Osmerusmordax) in a turbid well-mixed estuary.

Marine Ecology Progress Series 60:1-12.Davis, J.P., E.T. Schultz, and J. Vokoun. 2009. Assessment of river herring and striped78 e ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN bass in the Connecticut River: abundance, population structure, and predator/prey interactions.

Ecology & Evolutionary Biology Articles Paper 26.http://digitalcommons.uconn.edu/eebarticles/26 DFO (Department of Fisheries and Oceans, Canada).

2011. Integrated Smelt FisheryManagement Plan, Eastern New Brunswick Area, Gulf Region, 2007-2011.

DFOIFMP, New Brunswick, Canada. http://www.glf.dfo-mpo.gc.ca/e0008079 Dodds, W.K. 2007. Trophic state, eutrophication and nutrient criteria in streams.TRENDS in Ecology and Evolution Vol.22 No. 12.Eayrs, S., N. Stevens, D. Salerno, and M. Raymond.

2009. A contemporary assessment of the bycatch of regulated species and the Nordmore Grate in thenorthern shrimp fishery.

Northeast Consortium Final Report PZ09020.http://www.northeastconsortium.org/pdfs/awards 2007/Eayrs%2007/Eayrs%20 07%20Final%2OReport.pdf Ellis, D., and J.C. Vokoun. 2009. Earlier spring warming of coastal streams andimplications for alewife migration timing. North American Journal of Fisheries Management 29:1584-1589.

Finn, R.N. 2007. Review. The physiology and toxicology of salmonid eggs and larvaein relation to water quality criteria.

Aquatic Toxicology 81:337-354.

Fiske, J.D., J.R. Curley, and R.P. Lawton. 1968. A study of the marine resources ofthe Westport River. Massachusetts Division of Marine Fisheries, Monograph SeriesNo. 7.Flagg, L.N. 1974. Distribution, food habits, and management of striped bass andsmelt in Maine coastal waters. Completion report AFS-4, Maine Department ofMarine Resources.

Flagg, L.N. 1983. Evaluation of anadromous fish resources.

Final Report AFS-2 l-R,Maine Department of Marine Resources.

Fried, H.A., and E.T. Schultz.

2006. Anadromous rainbow smelt and tomcod inConnecticut:

Assessment of populations, conservation status, and need forrestoration plan. Master's thesis. University of Connecticut, Storrs.Friedland, K.D., J.P. Manning, J. Link, J.R. Gilbert, A.T. Gilbert, and A.EO'Connell, Jr. 2012. Variation in wind and piscivorous predator fields affecting the survival of Atlantic salmon, Salmo salar, in the Gulf of Maine. Fisheries Man-agement and Ecology 19:20-25.

Frumhoff, P.C., J.J. McCarthy, J.M. Melillo, S.C. Moser, and D.J. Wuebbles.

2007.Confronting Climate Change in the U.S. Northeast:

Science, Impacts, andSolutions.

Synthesis report of the Northeast Climate Impacts Assessment (NE-CIA). Cambridge, MA: Union of Concerned Scientists (UCS). http://www.north-eastclimateimpacts.org/pdf/confronting-climate-change-in-the-u-s-northeast.pdf Fuda, K.M., B.M. Smith, M.P. Lesser, B.J. Legare, H.R. Breig, R.B. Stack, and D.L.Berlinsky.

2007. The effects of environmental factors on rainbow smelt Osnierusmordax embryos and larvae. Journal of Fish Biology 71:539-549.

Geffen, A.J. 1990. Response of rainbow smelt, Osmerus mordax (Mitchill),

eggs to lowpH. Journal of Fish Biology 37:865-871.

Goode, G.B. 1884. The Fisheries and Fishery Industries of the United States. U.S.Commission of Fish and Fisheries, Washington D.C.Greene C.H., B.C. Monger, L.P. McGarry, M.D. Connelly, N.R. Schnepf, A.J. Persh-ing, L.M Belkin, P.S. Fratantoni, D.G. Mountain, R.S. Pickart, R. Ji, J.J. Bisagni,C. Chen., S.M.A. Hakkinen, D.B. Haidvogel, J. Wang, E. Head, P. Smith, andA. Conversi.

2012. Recent Arctic Climate Change and its Remote Forcing ofNorthwest Atlantic Shelf Ecosystems.

Oceanography 25:2.ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN

• 79 Hall, C.J., A. Jordaan, and M.G. Frisk. 2012. Centuries of anadromous foragefish loss: consequences for ecosystem connectivity and productivity.

BioScience 62(8):723-731.

Hodgkins, G.A., R.W. Dudley, and T.G. Huntington.

2003. Changes in the timingof high river flows in New England over the 20th century.

Journal of Hydrology 278:244-252.

Howarth, R., D. Anderson, J. Cloern, C. Elfring, C. Hopkinson, B. Lapointe, T.Malone, N. Marcus, K. McGlathery, A. Sharpley, and D. Walker. 2000. Nutrientpollution of coastal rivers, bays, and seas. Issues in Ecology 7:1-15.Howell, W.H., and R. Langan. 1992. Discarding of commercial groundfish species inthe Gulf of Maine shrimp fishery.

North American Journal of Fisheries Manage-ment 12:568-580.

Hurlbert, P.J. 1974. Factors affecting spawning site selection and hatching success inanadromous rainbow smelt (Osmerus mordax),

Mitchell.

Master's Thesis. University of Maine, Orono.Incze, L., H. Xue, N. Wolff, D. Xu, C. Wilson, R. Steneck, R. Wahle, P. Lawton, N.Pettigrew, and Y Chen. 2010. Connectivity of lobster (Homarus americanus) populations in the coastal Gulf of Maine: part II. Coupled biophysical dynamics.

Fisheries Oceanography 19:(1) 1-20.Jezierska, B., K. Lugowska, and M. Witeska.

2009. The effects of heavy metals onembryonic development of fish (a review).

Fish Physiology and Biochemistry 35:625-640.

Jimenez, D., J.E. Pelczarski, and H.R. lwanowicz.

1982. Incidence of piscine erythro-cytic necrosis (PEN) and Glugea hertwigi (Weissenberg) in rainbow smelt(Osmerus mordax) in selected Massachusetts streams.

Estuaries 5(2):145-149.

Juanes, E, S. Gephard, and K.E Beland. 2004. Long-term changes in migration timing of adult Atlantic salmon (Salmo salar) at the southern edge of the speciesdistribution.

Canadian Journal of Fisheries and Aquatic Sciences 61:2392-2400.

Kendall, WC. 1926. The smelts. Bulletin of the Bureau of Fisheries, Vol. XLII, Doc.No. 1015.Kircheis, EW, and J.G. Stanley.

1981. Theory and practice of forage-fish management in New England.

Transactions of the American Fisheries Society 110:729-737.

Kirn, R.A. 1986. Assessment of rainbow smelt (Osmerus mordax) as forage in LakeChamplain.

Master's thesis. University of Vermont, Burlington.

Klein, R.D. 1979. Urbanization and stream quality impairment.

Water Resources Bulletin 15(4):948-962.

Kovach, A.I., T.S. Breton, C.L. Enterline, and D.L. Berlinsky.

In press. Identifying the spatial scale of population structure in anadromous rainbow smelt (Osmerusmordax).

Fisheries Research.

Lapierre, L., C. Cormier, G. Trencia, and G. Verreault.

1999. Periphyton of four St.Lawrence estuary tributaries used for rainbow smelt spawning.

Pages 33-38 in FirstNorth American Workshop on Rainbow Smelt, Quebec.Laprise, R., and J.J. Dodson. 1989. Ontogeny and importance of tidal vertical migra-tions in the retention of larval smelt Osmerus mordax in a well-mixed estuary.Marine Ecology Progress Series 55:101-111.

Lawton, R., P. Brady, C. Sheehan, S. Correia, and M. Borgatti.

1990. Final report onspawning sea-run rainbow smelt (Osmerus mordax) in the Jones River and impactassessment of Pilgrim Station on the population, 1979-1981.

Pilgrim NuclearPower Station Marine Environmental Monitoring

Program, Report Series No. 4,72 pp.80 ° ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN

Limburg, K.E., and R.E. Schmidt.

1990. Patterns of fish spawning in Hudson Rivertributaries:

response to an urban gradient?

Ecology 71(4):1238-1245.

Limburg, K.E., and J.R. Waldman.

2009. Dramatic declines in North Atlanticdiadromous fishes. BioScience 59:955-965.

Marcotte, A., and J.L. Tremblay.

1948. Notes sur la biologie de l'6perlan (Osmerusmordax Mitchell) de la Province de Qu~bec. Sta. Biol. Saint-Laurent, Que.Contrib.

18:107 pp.Martin, E.H., and C.D. Apse. 2011. Northeast Aquatic Connectivity:

An Assessment of Dams on Northeastern Rivers. The Nature Conservancy, Eastern Freshwater Program.

http://cw-environment.usace.army.mil/restore/fishpassage/pdfs/NEA-quaticConnectivityReport.pdf MassDEP (Massachusetts Department of Environmental Protection).

2007.Massachusetts Surface Water Quality Standards.

Massachusetts Department ofEnvironmental Protection, Division of Water Pollution

Control, Technical ServicesBranch, Westborough, MA (Revision of 314 CMR 4.00, January 2007).MCFG (Massachusetts Commissioners on Fisheries and Game). 1917. Fifty-first an-nual report, for the year 1916. Boston, MvA.McKenzie, R.A. 1958. Age and growth of smelt, Osmerus mordax (Mitchill),

of theMiramichi River, New Brunswick.

Journal of the Fisheries Research Board ofCanada 15(6):1313-1327.

McKenzie, R.A. 1964. Smelt life history and fishery of the Miramichi River, NewBrunswick.

Journal of the Fisheries Research Board of Canada, Bulletin 144.MECSSF (Maine Commission of Sea and Shore Fisheries).

1920. SecondBiennial Report of the Commission of Sea and Shore Fisheries of the State ofMaine. Auburn, Maine.ME DMR (Maine Department of Marine Resources).

2011. Boothbay HarborEnvironmental Data. http://www.maine.gov/dmr/rm/environmentaldata.html

Mitchell, D.F, M. Liebman, L. Ramseyer, and B. Card. 2003. Riffles and Reservoirs

-nutrient criteria and downstream effects.

Proceedings of the Water Environment Federation, National TMDL Science and Policy, 22:572-592.

Moravec, F, and I. De Buron. 2009. New data on three gonad-infecting species ofPhilometra (Nematoda, Philometridae) from estuarine fishes in South Carolina, USA. Acta Parasitologica 54(3):244-252.

Murawski, S.A., and C.F Cole. 1978. Population dynamics of anadromous rainbowsmelt Osmerus mordax, in a Massachusetts river system. Transactions of theAmerican Fisheries Society 107:535-542.

Murawski, S.A., G.R. Clayton, R.J. Reed, and C.H. Cole. 1980. Movements ofspawning rainbow smelt, Osmerus mordax, in a Massachusetts Estuary.

Estuaries 3(4):308-314.

NEFSC (Northeast Fisheries Science Center).

2010. Gray Seal (Halichoerus grypusgrypus):

Western North Atlantic Stock in U.S. Atlantic and Gulf of Mexico MarineMammal Stock Assessments.

NOAA Technical Memorandum NMFS-NE-219.

NEFSC (Northeast Fisheries Science Center).

2012. Northeast Fisheries ObserverProgram.

http://www.nefsc.noaa.gov/fsb/program.html NHF&G (New Hampshire Fish and Game Department).

1981. Fisheries manage-ment plan for rainbow smelt (Osmerus mordax).

New Hampshire Fish and GameDepartment,

Concord, NH.NJCF (New Jersey Commissioner of Fisheries).

1872. Second Annual Report of theCommissioners of Fisheries of the State of New Jersey. Trenton, New Jersey.ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN

• 81 NJCF (New Jersey Commissioner of Fisheries).

1886. Report of the New Jersey FishCommissioners to the One Hundred and Tenth Legislature of the State of NewJersey. Vol. III, Camden, New Jersey.NOAA. 2012. http://www.nefsc.noaa.gov/fsb/program.html Norris, T. 1862. Remarks on a species of Osmerus taken in the Schuykill, belowFairmont Dam. Proceedings, Academy of Natural Sciences of Philadelphia, 1861(1862), pp. 58-59, Philadelphia, PA.Nsembukya-Katuramu, S., E.K. Balon, and R. Mahon. 1981. A comparison ofspawning, harvested, and die-off rainbow smelt, Osmerus mordax, in eastern LakeErie. Journal of Great Lakes Research 7(2):144-154.

O'Gorman, R., R. Bergstedt, and T.H. Eckert. 1987. Prey fish dynamics and salmo-nine predator growth in Lake Ontario 1974-1984.

Canadian Journal of Fisheries and Aquatic Sciences 44(Supplement 2):390-403.

Pess, G.R., D.R. Montgomery, E.A. Steel, R.E. Bilby, B.E. Feist, and H.M.Greenberg.

2002. Landscape characteristics, land use, and coho salmon(Oncorhynchus kisutch) abundance, Snohomish River, Washington, USA.Canadian Journal of Fisheries and Aquatic Sciences 59:613-623.

Pettigrew, P. 1997. Suivi de la reproductio de L'eperlan arc-en-ciel de la rive sud der'Estuaire du Saint-Laurent en 1994, 1995, et 1996. Ministere de l'Env. et de laFaune, Direction regionale du Bas-Saint-Laurent, Riviere-du-Loup,

Fevrier, 52 pp.Pettigrew, N.R., D.W. Townsend, H. Xue, J.P. Wallinsa, P.J. Brickley, and R.D.Hedand. 1998. Observations of the Eastern Maine Coastal Current and itsoffshore extensions in 1994. Journal of Geophysical Research 103:30623-30640.

Pettigrew, N.R., J.H. Churchill, C.D. Janzen, L.J. Mangum, R.P. Signell, A.C.Thomas, D.W. Townsend, J.P. Wallinga, and H. Xue. 2005. The kinematic andhydrographic structure of the Gulf of Maine Coastal Current.

Deep Sea Research1152:2369-2391.

Pouliot, G. 2002. Dynamique de la population d'eperlans arc-en-ciel (Osmerus mordax)du sud de l'estuaire du Saint-Laurent par l'analyse des cohortes de reproducteurs frequentant la riviere Fouquette entre 1994 et 2001. Societe de la faune et desparcs du Quebec, Direction de l'amenagement de la faune de la region du Bas-Saint-Laurent.

47 p.Reback, K.E., and J.S. DiCarlo.

1972. Completion Report on the anadromous fishproject.

Massachusetts Division of Marine Fisheries, Publication No. 6496:113 p.RIDFW (Rhode Island Division of Fish and Wildlife).

1971. Restoration and Estab-lishment of Sea Run Fisheries.

Rhode Island Division of Fish and Wildlife, FederalAid in Fish Restoration, Project F-26-R-6, Job Completion Report.RIDFW (Rhode Island Division of Fish and Wildlife).

1978. Restoration and Estab-lishment of Sea Run Fisheries.

Rhode Island Division of Fish and Wildlife, FederalAid in Fish Restoration, Project F-26-R-13, Job Completion Report.Rupp, R.S. 1959. Variation in the life history of the American smelt in inland watersof Maine. Transactions of the American Fisheries Society 88:241-252.

Rupp, R.S. 1965. Shore-spawning and survival of eggs of the American smelt, Osmerusmordax (Mitchill),

in Maine. Transactions of American Fisheries Society 90 (1):42-48.Rupp, R.S. 1968. Life history and ecology of the smelt (Osmerus mordax) in the inlandwaters of Maine. Maine Department of Inland Fish Game, Federal Aid in FishRestoration, Project F- 10-R, Final Report.82

• ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN

Saunders, R., M.A. Hachey, and C.W Fay. 2006. Maine's diadromous fishcommunity:

past, present, and implications for Atlantic salmon recovery.

Fisheries 31(11):537-547.

Schwartz, M.K., L.S. Mills, Y. Ortega, L.E Ruggiero, and EW. Allendorf.

2003.Landscape location affects genetic variation of Canada lynx. Molecular Ecology12:1807-1816.

Scott, W, and M. Scott. 1988. Atlantic Fishes of Canada. Canadian Bulletin ofFisheries and Aquatic Sciences 219:731 pp.Siegel, S. and N.J. Castellan.

1988. Nonparametric Statistics for the Behavioural

Sciences, 2nd edition.

McGraw-Hill, New York.Simmons, D. and R. Reynolds.

1982. Effects of urbanization on baseflow of selectedSouth-Shore

streams, Long Island, NY Water Resources Bulletin 18(5):797-805.

Sirois, P., and J.J. Dodson. 2000. Critical periods and growth-dependent survival oflarvae of an estuarine fish, the rainbow smelt Osmerus mordax. Marine EcologyProgress Series 203:233-245.

Spalding, M.D., H. Fox, G. Allen, N. Davidson, Z. Ferdafia, M. Finlayson, B.Halpern, M. Jorge, A. Lombana, S. Lourie, K. Martin, E. Mcmanus, J. Molnar,C. Recchia, and J. Robertson.

2007. Marine Ecoregions of the World: A Biore-gionalization of Coastal and Shelf Areas. BioScience 57(7):573-583.

Squires, T.S., L. Flagg, and L. Austin. 1976. Smelt Management Plan. Maine Depart-ment of Marine Resources, Completion Report Project #AFSC-13/FWAC-2.

Stewart, D.J., J.E Kitchell, and L.B. Crowder.

1981. Forage Fishes and TheirSalmonid Predators in Lake Michigan.

Transactions of the American Fisheries Society 110:751-763.

Storer, H.R. 1858. Fishes of Massachusetts.

Memoirs of the American Academy ofSciences United States Government Printing Office.Strayer, D.L., R.E. Beighley, L.C. Thompson, S. Brooks, C. Nilsson, G. Pinay, andR.J. Naiman. 2003. Effects of land cover on stream ecosystems:

roles of empirical models and scaling issues. Ecosystems 6:407-423.

Sullivan, K. 2007. Monitoring of the rainbow smelt resource and winter ice fishery.New Hampshire Fish and Game Department, Federal Aid in Fish Restoration, Project F-61-R, Progress Report, Durham.Sullivan, K. 2009. Rainbow smelt creel survey design and calculation methodologies.

New Hampshire Fish and Game Department, Federal Aid in Fish Restoration, Project F-61-R, Progress Report Appendix 1, Durham.Sullivan, K. 2010. Monitoring of the rainbow smelt resource and winter ice fishery.New Hampshire Fish and Game Department, Federal Aid in Fish Restoration, Project F-61-R, Progress Report, Durham.Trencia, G., G. Verreault, M. Legault, and V. Tremblay.

2005. L' eperlan arc-en-ciel (Osmerus mordax) anadrome du sud de l'estuaire du Saint-Laurent, une population en voie de designation comme espece vulnerable.

Le Naturaliste

Canadien, Vol.129 No. 1.US DOC (U.S. Department of Commerce).

1920. Progress in Biological Inquiries.

Report of the Division of Scientific Inquiry for the Fiscal Year of 1920. Depart-ment of Commerce, Bureau of Fisheries, Doc. No. 896, Washington, D.C.US EPA (U.S. Environmental Protection Agency).

2000. Nutrient Criteria Technical Guidance Manual. Rivers and Streams.

Office of Water, USEPA, Document EPA822-B-00-OO2Washington, D.C.ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN a 83 US EPA (U.S. Environmental Protection Agency).

2004. National water qualityinventory.

USEPA, Report to Congress, 2004 reporting cycle, Washington, DC.USFWS (U.S. Fish & Wildlife Service).

2012. http://www.fws.gov/GOMCP/identify.

html#diafish Verreault, G., G. Trencia, and M. Legault.

2012. The rainbow smelt restoration planfor the St. Lawrence estuary:

where are we eight years later? Pages 62-64 in FourthNorth American Workshop on Rainbow Smelt: Extended Abstract Proceedings, Massachusetts Division of Marine Fisheries, Technical Report TR-5 1.Visel, T., and T. Savoy. 1989. The historical decline of the rainbow smelt "Osmerusmordax"in Connecticut's coastal waters. Proposal to the University of Connecticut Center for Environmental Health, University of Connecticut Cooperative Exten-sion Service.

Reprinted in 2005.Wang, L., J. Lyons, P. Kanehl, and R. Bannerman.

2001 a. Impacts of urbanization onstream habitat and fish across multiple spatial scales. Environmental Management 28(2):255-266.

Warfel, H.E., T.P. Frost, and WH. Jones. 1943. The smelt, Osmerus mordax, in GreatBay, New Hampshire.

Transactions of the American Fisheries Society 72:257-262.

Wegwu, M., and J. Akaninwor.

2006. Assessment of heavy-metal profile of the NewCalabar River and its impact on juvenile Clarias gariepinus.

Chemistry

&Biodiversity 3:79-87.Wootton, J.T., C.A. Pfister, and J.D. Forester.

2008. Dynamic Patterns andEcological Impacts of Declining Ocean pH in a High-Resolution Multi-Year Dataset.

Proceedings of the National Academy of Sciences of the United States ofAmerica 105(48):18848-18853.

Whitten, D.R. 1894. Commissioner of Sea and Shore Fisheries of the State of Maine,1894, Augusta, ME, 38 pp.Wyatt, L.H., A.L. Baker, and D.L. Berlinsky.

2010. Effects of sedimentation andperiphyton communities on embryonic Rainbow Smelt, Osmerus mordax. AquaticSciences 72:361-369.

Yako, L.A., M.E. Mather, and E Juanes. 2002. Mechanisms for migration of anadro-mous herring:

an ecological basis for effective conservation.

Ecological Applications 12(2):521-534.

84 -ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN APPENDIX25 :Weweantic River20 0- 2008-dr-20100 -"-20113,1 3/21 4/10 4/30 5/20 6/9150Jones River100

• 2008-,-20100 4 20113/1 321 4/10 4/30 5S20 6/9300 Fore River41 50 16-2009100so -d-210 ---- 20113J1 3/21 4!10 4(30 5/20 6,95oo Parker River400 o- 20oo300200 -2009100 -2010 Figure A.1.1. Catch-per-unit-effort 0 ..... -X- 2011 (number of smelt per haul) atV'1 3121 4/10 4/30 5i20 6/9 selected Massachusetts fyke netstations, 2008-2011.

ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN

• 85 200100so0Squamscott River*ý 200$0-5-2009-0-r-2010 hi4-2011.

31131 3/21 4,10 4130 S"20 6/93:M106i430Winnicut River*+ 200$-1-2009--2010042011121 4/10 4130 5/20 6/91006040200/Oyster River-","2010--0-2011Figure A.1.2. Catch-per-unit-effort (number of smelt per haul) atNew Hampshire fyke net stations, 2008-2011.

3;214;10 4,30 5,,20 6i986 0 ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN 600500400 430020010003/1AAA'.3Mast Landing4I+ 2008-.-2009---2010-)- 20113V214/10 4/30 5/20 6,9700MX4.(0150010005003/"15001000500Deer Meadow Brook2008-,-2009-,-,,-2010

-- 20113/21 4/10 4/30 5/20 619Tannery Brook4,* 2003--2009-,"-20100- 20113!13/21 41104/30 5/20 6i9400300200100 i3/1East Bay River*- 2008-2009S2010N&AW 20113121 4/10 4/30 5/20 6/9Figure A.1.3. Catch-per-unit-effort (number of smelt per haul) atselected Maine fyke net stations, 2008-2011.

ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN

• 87 soI 2009 -2.01,1094 0 U 12L 1 ID 2S20"noo"Figure A.3-4. Length frequency of rainbow smelt caught In theWeweantic River, MA, fyke net,2008-2011.

go2006904020 Ji1i. nulL..2009.W096 IN -013NVION 374210)N4,40iii upijilMi

-0a 9 20 U1 a ll o6 1617 0 tol a a U 0 2121 2 2 002 27 2TONI1 L'"lv1 ao)01 10 1IO U U " " Ii111 1 I 1 0 1 22 1 4a 20 26 2.7Towj 1440 1iM)W0Ia f IN -1714amok1 (N. fil2011.~fqlO4 01N291am*41 IN-10910.a~lI111.9a* 6 0 I U 12 II14 20 16 7 1* I 1021 22 2 24 2526 27 300 9 10 II 12 14 10 1I 17 10 09 1 21 22 21 2 20 2 27Figure A.1.5. Length frequency ofrainbow smelt caught In the JonesRiver, MA, fyke net, 2008-2011.

No.2006_Jill...Iii f.m98 041.114271P am*.J Im41000)2009,Fo.1-- (11.14711 394.6 IN -6"0.u.IjjW40I mI0i _. .. -. AM64. .11 9 W10 12 U It 14 16 27 A1 6. Z) 11 UZ 23 24 JS 26 21, aTotjid LeU, ton)*9 I 1* 12 U!It 14 A2 17V 3 1 2 20 U1 24 25 A4 27 2low 'Ifoh tw2010X00Nd.100amUFO-*. IN-am0hu0.

WF0odw (0.2271,31043 IN. 1010IJllJ-No.we0-Um.A L ~ -U --9a 1r U Ua 1 14 15 U 17 nI 20 21 22 2214 a 26 27 R.NTOW R OALth CO A N*

• 1011 U 14 s1 18 a0 39 1 02 22 50 72Figure A.1.6 Length frequency ofrainbow smeit caught In the ForeRiver, MA, fyke net 2008-2011.

88 a ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN 40 2010 2*011 ll40 zn"Fel (,$40 4220 209 0 WU 2U 14 IS U17 IS It20 Z12 23 Z4 2 524 7 n1 ol2* dI11729012 34557ZTMW La~t AM 04111 10"IIFigure A.1. 7. Length frequency of rainbow smelt caught at theOyster River, NH, fyke net, 2010-2011.soNa.to9 S IOU U U m3 14 is 10 712111920 21 2223 24 252 all7 240S30No.22I R 30 U 12 13 14 IS l6 17 IN 192D 221 22 X3 M4 2S 26 22 .9Figure A.1.8. Length frequency ofrainbow smelt caught at the Lam-prey River, NH, fyke net, 2008.Figure A.1.9. Length frequency of rainbow smelt caught at theSquamscott River, NH, fyke net2011ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN

• 89I 40.2002009M0UFWm. tmN SZ7)am0w N4 132).Ivm*2 jfl-mample 04-5.221No.1gW1010,10.I1.--iM-.A1 9 10 U 12 U 14 15 16 17 IN 2 W 20 2 22 23 24 IS 21 Z? 28%WLS2 A04ao)la 9 2 U2 11 22 4 is 26 1? 0 is 20 21 U2 32 14 2S 20 27 IsTA LWOO laiR20220No.1105020104W0202182fi...*

tN -76DIMS 0~211l 3W2W 3we..0,1ilL0 tý.402. (N.""muI0e M mo MA"U A UaS 20 11 1 2 U 1.4 IS It 17 26 19 20 21 nl 2l l4 25 26 27 28TOWi L-0 w9 9 20 94 U It 104 IS1 17I IS 1 21 20 72M8[- fmFigure A.1.10. Length frequency of rainbow smelt caught at MastLanding, ME, fyke net, 2008-2011.

2100,200=SOD0~J~Sa# fN~7 allE~il 79~ISO2009'Fe..* (N.14A25,ai"24444No.24'I9 IN 241 U Ing 12 4 IS 16 12 19 If 29 21 22 23 24 25 26 27 n6Tow Lii(e)2010 Wffnaie IN-1774*Mast (N-2l21)* 9 20 I 12 13 14 IS 16 1? A1 V 29 0 21 Z2 23 24 2!5 26 27 25TOWb"w d~1 9 10 22 Ulf0 141 22 ly U. 2)119301 22 22 7224 26 2? 7a,fem.0I (N-3347UM*IO IN262273Wx9.200SODI SO It 1 2 23 14 U5 15 17 26 26 A0 1 27 71 24 S0 2i"? ATOW 1040 (CM)Figure A.1.11. Length frequency of rainbow smelt caught at DeerMeadow Brook, ME fyke net2008-2011.

90

• ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLANI zoOnW1FOi3 N.434W,JJill'.1.03010Ii..3010* (N

• 11Sgl4. ----*9 U) I 1.2 U 14 15 16 11 18 19 20 21 V U 25 4 25 26, 27 ISVAW LaN"I1-3 0 11 12 U 14 15 16 V7 IS 1.9 Z0 U1 U22 25 24 I 26 Z 20Taw3 £0340 lIwo2010ýWsr4*! p4'?772)sm40*Am.. j".3121WI.lIi].iJi.IL~.

0ý9 to0U12 U 14I"Ewo*0L03 -1 9 10 11 12 U 14 1 16 10 19 It N0 U1 a 23 14 as 16 22 28UWUW40"W, Figure A.1.12. Length frequency ofrainbow smelt caught at TanneryBrook, ME, fyke net, 2008-2011.

2W0NI.(?.Laa

• 1110W11.113222254W222 8f Fftam~ aowFigure A.1.13. Length frequency ofrainbow smelt caught at SchoppeeBrook, ME, tyke net, 2010-2011.

20"5,m311 INAZ1IZ0091.50 1DOOM. (W, S1O)9M.no111d..IA. 1.0Na 9 2.0 1.1 U I HA 15 A 1.7 A* " 20 1 U0 Z) 24 25 20 IF 2850.no. a".6 (N-691)a I,0 9 to2 U U2 13 14 is 1 1.7 U to3a 2011 22 53 24 25 x6 2? 25-- I8 3 20 11 U 13 N4 15 16 I1 18 U1 20 2a 22 is 24 27 a 21 24lo efat ".640OW2011JFOXVU 00-0#1I613440 M-22113Io.S$ 8 11 12 U3 14 is 16 1?1* Uigu 20 r1 AS .1 24 20 20 IV 2oBok M, f")20Figure A.1.14. Length frequency ofrainbow smelt caught at East BayBrook, ME, tyk net 2008&2011.

ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN

• 91I Figure A.2.1. Water temperature data distributions for 19 smeltsampling stations In study area.The top of the box plots is the75th percentile and the bottomIs the 25th percentile.

The linewithin the box Is the median andthe error bars represent the 10thand 90th percentiles.

The stationsare arranged on the x-axis fromthe southernmost MA station tothe northernmost ME station.Station medians were found tobe significantly different withKruskai-Wallis test (KW -93.21,df -18, p < 0.001).2520 10 -C.EITalil ggI i t50WP WWN JR FR SG NR CN ER PR SQ WR OY LC ML DM TB SB CR EBRiverFigure A.2.2. Specific conductiv-ity data distributions for 18 smeltsampling stations in study area.The top of the box plots Is the75th percentile and the bottomis the 25th percentile.

The linewithin the box Is the median andthe error bars represent the lothand 90th percentiles.

The stationsare arranged on the x-axis fromthe southernmost MA station tothe northernmost ME station.Station medians were found tobe significantly different withKruskaI-Wallls test (KW -1374.4,df -17, p < 0.001).U09C-o)1.4-1.2 -1.0-0.80.6-0.4-0.2-0.0TWP WW JR FR SG NR CN ER PR SQ WR OY LC ML DM TB SB EBRiver92 9 ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN

-JE0E01614-12 8-6-4-T T T TT T T T T T T T TqHTT TFigure A.2.3. Dissolved oxygen(mg/L) data distributions for 19smelt sampling stations In studyarea. The top of the box plots isthe 75th percentile and the bot-tom Is the 25th percentile.

Theline within the box Is the medianand the error bars represent the10th and 90th percentiles.

Thestations are arranged on the x-axisfrom the southernmost MA stationto the northernmost ME station.Station medians were found tobe significantly different withKruskaI-Wallis test (KW -439.51,df -18, p < 0.001). The green linemarks the MassDEP DO criterion (6.0 mg/L) for protecting AquaticLife.WP WW JR FR SG NR CN ER PR SQ WR OY LC ML DM TB SB CR EBRiver14U0a12010080t --T T T TFigure A.2.4. Dissolved oxygen(% saturation) data distributions for 19 smelt sampling stationsIn study area. The top of the boxplots is the 75th percentile andthe bottom Is the 25th percentile.

The line within the box is the me-dian and the error bars represent the 10th and 90th percentiles.

Thestations are arranged on the x-axisfrom the southernmost MA stationto the northernmost ME station.Station medians were found tobe significantly different withKruskal-Wallis test (KW -439.51,df -18, p < 0.001).tjUWP WI JR FR SG NR CN ER PR SQ WR OY LC ML DM TB SB CR EBRiverANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN

• 93I Figure A.2.5. Water pH datadistributions for 19 smelt sam-pling stations in study area. Thetop of the box plots is the 75thpercentile and the bottom Is the25th percentile.

The line withinthe box Is the median and theerror bars represent the 10th and90th percentiles.

The stations arearranged on the x-axis from thesouthernmost MA station to thenorthernmost ME station.

Stationmedians were found to be sig-nflcantly different with Kruskal-Wallis test (KW -1041.3, df -18,p < 0.001). The green lines markthe lower MassDEP pH criterion (a6.5 and s 8.3) for protecting Aquatic Life.Figure A.2.6. Turbidity (NTU)data distributions for 19 smeltsampling stations in study area.The top of the box plots is the75th percentile and the bottomis the 25th percentile.

The linewithin the box is the median andthe error bars represent the 10thand 90th percentiles.

The stationsare arranged on the x-axis fromthe southernmost MA station tothe northernmost ME station.Station medians were found tobe significantly different withKruskai-Waills test (KW -660.8,df -18, p < 0.001). The green linemarks the EPA turbidity criterion for minimally Impacted waterquality (:5 1. 7 NTU).9.08.5'8.0-7.5.7.0-6.5-6.0-5.5-5.0-4,540T T7T-PVWV JR FR SG NR CN ER PR SQWROY LO ML DM TB SB CR EBRiver242220181614z 121086420WP WV JR FR SG NR CN ER PR SQ WR OY LC ML DM TB SB CR EBRiver94

• ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN

-J76)EzI--2.52.01.51.00.5-0.0T TFigure A.2. 7. Total nitrogen (TN)data distributions for 20 smeltsampling stations in study area.The top of the box plots is the 75thpercentile and the bottom Is the25th percentile.

The line withinthe box Is the median and theerror bars represent the 10th and90th percentiles.

The stations arearranged on the x-axis from thesouthernmost MA station to thenorthernmost ME station.

Stationmedians were found to be sig-nificantly different with Kruskal-Wallis test (KW -408.4, df -19,p < 0.001). The green line marksthe EPA total nitrogen criterion forminimally Impacted water quality(< 0.57 mg/L).WPWWW JR FR SG NR CN ER MR PR SQ WR OY LC ML DM TB SB CR EBRiver6050,40,30-a.t tFigure A.2.8. Total phosphorus (TP) data distributions for 20smelt sampling stations In studyarea. The top of the box plots Isthe 75th percentile and the bot-tom Is the 25th percentile.

Theline within the box Is the medianand the error bars represent the10th and 90th percentiles.

Thestations are arranged on the x-axisfrom the southernmost MA stationto the northernmost ME station.Station medians were found tobe significantly different withKruskal-Wallis test (KW -174.7,df -19, p < 0.001). The green linemarks the EPA total phosphorus criterion for minimally Impactedwater quality (< 23.75 ugIL).20110.0-WPWWAJR FR SG NR CN ER MR PR SQ WR OY LC ML DM TB SB CR EBRiverANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN e 95 I96 o ANADROMOUS RAINBOW SMELT REGIONAL CONSERVATION PLAN For a copy of this report online, please visitwww.restorerainbowsmelt.

comand click on the "Learn More" tabFor a printed copy, please contact your state marine agency:Massachusetts Division of Marine Fisheries Website:

http://www.

mass.gov/dfrele/dmfl Boston Offices:

(617) 626-1520Gloucester Regional Office: (978) 282-0308New Bedford Regional Office: (508) 990-2860New Hampshire Fish and Game Department Website:

http://www.wildlife.state.nh.usl Durham Marine Fisheries Division:

(603) 868-1095Maine Department of Marine Resources Website:

http ://www.maine.gov/dmr/index.htm Sea Run Fisheries Division:

(207) 287-9972Bureau of Marine Sciences:

(207) 633-9500MAR12003.INDD/NHFG 2012

New York State Department of Environmental Conservation Office of General Counsel, 14tb Floor625 Broadway, Albany, New York 12233-1500 Fax: (518) 402-9018 or (518) 402-9019Website:

www.dec.ny.gov Joe MartensActing Commissioner January 28, 2011VIA ELECTRONIC MAILAND HAND DELIVERYHon. Maria E. VillaHon. Daniel P. O'Connell Administrative Law JudgesNew York State Department ofEnvironmental Conservation Office of Hearings and Mediation Services625 Broadway, 1st FloorAlbany, New York 12233-1550 Re: Enterzy Nuclear Indian Point. Units 2 and 3CWA Section 401 WQC Application Proceeding NRC--Atomic Safety and Licensing Board's Dec. 3,2010 FSEIS

Dear ALJs Villa and O'Connell:

This letter constitutes Department staff's filing in compliance with the Ruling onProposed Issues for Adjudication and Petitions for Party Status dated December 13, 2010,issued in the Entergy Indian Point §401 WQC proceeding

("Issues Ruling"),

and with item 3 ofthe Scheduling Order attached to the Issues Ruling. Specifically, page 9 of the Issues Rulingand item "3" of the Scheduling Order directed Department staff to:" advise the ALJs and the parties as to whether the Nuclear* Regulatory Commission, Atomic Safety and Licensing Board'sDecember 3, .2010 Final Supplemental Environmental ImpactStatement

('FSEIS')

is sufficient for Department Staff to -makethe findings required by Section 617.11 of 6 NYCRR.'It is Department staff's position after due deliberation that, in conjunction with or asotherwise supplemented by the Final Environmental Impact Statement by the Department concerning the Applications to Renew SPDES Permits for Three Hudson River Power Plantsaccepted June 25, 2003, along with the Department's records of proceedings (administrative hearing records) for both the Entergy Indian Point SPDES permit (DEC No.: 2-5522-00011/00004) and Entergy Indian 'Point §401 WQC application (DEC Nos.: 3-5522-00011/0030 and 3-5522-00105/00031),

as well as with the NRC's record of proceeding (hearing file and record) for Entergy's license renewal for Indian Point Units 2 and 3 (DocketDepartment staff notes that the December 3,2010 FSEIS was prepared by staff of the NRC, not by.theAtomic Safety and Licensing Board, and that such FSEIS is not yet actually "final."

Nos. 50-247-LR and 50-286-LR; ASLBP No. 07-858-03-LR-BDO1),

including but not limitedto any contentions,

-attachments,

reports, declarations, comments; and administrative hearingsrelating to or arising from the publication by the NRC Staff on. December 3, 2010, of theFSEIS for the renewal of the Indian Point nuclear operating

licenses, the NRC Staffs FSEIS(insofar as it may be further supplemented or amended by future proceedings noted herein)would be sufficient

.for the purpose of making findings as required by 6 NYCRR §617.11.Department staff notes that, consistent with the provisions of 6 NYCRR §617.15(c),

a finaldecision by a Federal agency is not controlling on any agency decision on the proposed action,but may be considered by the agency. In addition, consistent with the provisions of 6 NYCRR§617.11 (e), Department staff further notes that, because the Indian Point nuclear facilities arelocated in the coastal area (as defined in 6 NYCRR §6i7.2[f]),

the agency' cannot make a finaldetermination on the proposed action until there has been a written finding that the action isconsistent with applicable policies set forth in 19 NYCRR §600.5.Thank you for your courtesies and attention to this matt6r.Very truly yours,Mark D. SanzaAssistant CounselVia U.S. Mail and E-Mail:Elise N. Zoli, Esq.John C. Englander, Esq.Goodwin Procter, LLPExchange PlaceBoston, Massachusetts 02109Rebecca Troutman, Esq.Riverkeeper, Inc.20 Secor RoadOssining, New York 10562Melissa-Jean Rotini, Esq.Assistant County AttorneyCounty of Westchester Room 600, 148 Martine AvenueWhite Plains, New York 10601Richard L. Brodsky, Esq.2121 Saw Mill River RoadWhite Plains, New York 10607ezoli@goodwinprocter.com jenglander@goodwinprocter.com rfitzgerald@goodwinprocter.com rtroutman@riverkeeper.org mjrl @westchestergov.corn richardbrodsky@gmail.com Daniel Riesel, Esq. driesel@sprlaw.com Sive, Paget & Riesel, P.C.460 Park Avenue,.

10't FloorNew York, New York 10022Steven Blow, Esq. stevenblow@dps.state.ny.us Assistant General CounselNew York State Department of Public ServiceAgency Building ThreeEmpire State PlazaAlbany, New York 12233-1350 Sam M. Laniado, Esq. sml@readlaniado.com David B. Johnson, Esq. dbj@readlaniado.com Read and Laniado, LLP25 Eagle StreetAlbany, New York 12207-1901 Michael J. Delaney, Esq. mdelaney@dep.nyc.gov

Director, Energy Regulatory AffairsNew York City Department of Environmental Protection 59 17 Junction Boulevard, 109t FloorFlushing, New York 11373-5108 Robert J. Glasser, Esq.Robert J. Glasser, P.C.284 South AvenuePoughkeepsie, New York 12601Bob.glasser~robertjg~asserpc.com Via E-Mail Only:Ned Sullivan, President Hayley Mauskapf, Esq.Paul Schwartzberg Scenic Hudson, Inc.Karl S. Coplan, Esq.Daniel E. Estrin, Esq.Pace Environmental Litigation Clinic, Inc.Deborah Brancato, Esq.Phillip H. MusegaasRiverkeeper, Inc.nsullivan@scenichudson.org hMauskapf@scenichudson.org Schwartzberg@scenichudson.org kcoplan@law.pace.edu destrin@law.pace.edu dbrancato@riverkeeper.org phillip@,riverkeeper.org

, IGeoffrey H. Pettus, Esq.Natural Resources Defense CouncilFrank V. Bifera, Esq.Hiscock & Barclay, LLPKelli M. Dowell, Esq.Entergy Services, Inc.gfettus@nrdc.org fbifera@hblaw.com kdowell@entergy-com EDMSI390924vI.