NL-08-023, Entrainment and Impingement at IP2 and IP3: a Biological Impact Assessment

Entrainment and Impingement at IP2 and IP3: a Biological Impact Assessment

ML080390059
Person / Time
Site:	Indian Point
Issue date:	01/31/2008
From:	Barnthouse L, Heimbuch D, Vanwinkle W, Young J AKRF, ASA Analysis & Communication, LWB Environmental Services, Van Winkle Environmental Consulting
To:	Office of Nuclear Reactor Regulation
References
NL-08-023, TAC MD5411, TAC MD5412
Download: ML080390059 (278)

Text

ENTRAINMENT AND IMPINGEMENT AT IP2 AND IP3:

A BIOLOGICAL IMPACT ASSESSMENT Lawrence W. Barnthouse LWB Environmental Services, Inc.

Douglas G. Heimbuch AKRF, Inc.

Webster Van Winkle Van Winkle Environmental Consulting Boise, Idaho John Young ASA Analysis & Communications, Inc.

January 2008

TABLE OF CONTENTS Executive Summ ary ....................................................................................................... 1 G lo ssary ......................................................................................................................... 3

1. Introduction ....................................................................................... ............. . . 5

2. Approach to Impact Assessment ...................................................................... 6 2.1 Definition of "adverse environmental impact".................. 8 2.1.1 Definition of adverse environmental impact in the context of fishery m anagem ent ............................................................... I............. 8 2.1.2 Definition of AEI in the context of ecological risk assessment ......... 10 2.1.3 Definition of adverse environmental impact in the context of entrainment and impingement... ............. ............................... 11 2.2 Why entrainment losses alone are insufficient to demonstrate A E I ............................................................................................... . . 12 2.3 Role of the conditional mortality rate (CMR) in impact assessm ent .................................................................................... . . 13 2.4 Role of long-term datasets in impact assessment ......................... 14 2.5 Indicators of adverse impacts potentially related to CWIS ....... 18

3. Evaluation of changes in abundance of fish populations with life stages susceptible to entrainment ............................................................................. 20 3.1 Species addressed ......................................................................... 22 3.2 Impact hypotheses and stressor metrics .... ......................... 23 3.2.1 CWIS .............................................. ............................................. . . 23 3.2.2 Fishing ......................................................................................... . . 24 3.2.3 Zebra mussels ............................................................................... 25 3.2.4 Predation by striped bass ............................................................... 26 3.2.5 Temperature ......... ;....................................................................... 26 3.3 Response m etrics ............................................................ ................... 27

3.3.1 Response metrics for striped bass, white perch, American shad, alewife, blueback herring, and bay anchovy ................................. 27 3.3.2 Response metrics for spottail shiner ............................................. 28 3.3.3 Response metrics for Atlantic tomcod .......................................... 28 3.4 Tests of impact hypotheses .......................................................... 29 3.4.1 Striped bass ................................................................................... 30 3.4.1.1 CW IS ............................................................................................. 31 3.4.1.2 Fishing ........................................................................................... 33 3.4.1.3 Zebra mussels ............................................................................. 34 3.4.1.4 Summary evaluation of hypotheses ............................................... 36 3.4.2 White perch .................................................................................... 36 3.4.2.1 CWIS ............................................................................................. 37 3.4.2.2 Zebra mussels ............................................................................... 39 3.4.2.3 Striped bass predation ................................ 40 3.4.2.4 Summary evaluation of hypotheses ...................... 42 3.4.3 American shad .............................................................................. 43 3.4.3.1 CW IS ............................................................................................. 44 3.4.3.2 Fishing ........................................................................................... 45 3.4.3.3 Zebra mussels ............................................................................... 47 3.4.3.4 Striped ba'ss predation.................................... 48 3.4.3.5 Summary evaluation of hypotheses ............................................... 50 3.4.4 Atlantic tomcod ............................................................................. 51 3.4.4.1 CW IS ............................................................................................. 52 3.4.4.2 Elevated summer temperatures .................................................... 54 34.4.3 Striped bass predation ................................................................... 55 3.4.4.4 Summary evaluation of hypotheses ............................................... 56 ii

3.4.5 Alewife and blueback herring ........................................... 57 3.4.5.1 CWJS ....................................................................... 57 3.4.5.2 Zebra mussels.............................................................. 59 3.4.5.3 Striped bass predation .............................................. I.......61 3.4.5.4 Summary evaluation of hypotheses..................................... 63 3.4.6 Bay anchovy...................I............................................. 64 3.4.6.1 CWJS....................................................................... 64 3.4.6.2 Striped bass predation .................................................... 66 3.4.6.3 Summary evaluation of hypotheses..................................... 67 3.4.7 Spottail Shiner............................................................. 67 3.5 Summary evaluation of trends analysis ............. :.................... 68

4. Evaluation of impacts of cooling-water withdrawals on spawning potential ........................................................... 69 4.1 History of the SSBPR model ............................................. 69 4.2 Explanation of the SSBPR concept...................................... 70 4.3 Application to Hudson River fish populations ......................... 72 4.3.1 Striped bass................................................................ 73 4.312 American shad ............................................................ 73

5. Community-Level Trends Analysis .. ...................................... 74 5.1 Methods .................................................................... 75 5.2 Results and Discussion ................................................... 78

6. Conclusions................................................... 78

7. References.............................................................................. 80 TablelI...........................................................................................1 Table 2........................................................ I..................................

Table 3..........................................................................................1II

T ab le 4 ......................................................................................................................... iv T ab le 5 .................................................................................................... ...................... V T ab le 6 ...........................................  ;............................................................................. v i T ab le 7 ........................................................................................................................ v ii T ab le 8 ....................................................................................................................... v iii T ab le 9 ......................................................................................................................... ix F igu re 1.......................................................................................................................... x F ig ure 2 ........................................................................................................................ xi F igure 3 ....................................................................................................................... x ii F igu re 4 ...................................................................................................................... x iii F ig ure 5 ...................................................................................................................... x iv F igu re 6 ....................................................................................................................... xv F igu re7 a ..................................................................................................................... xv i.

F igu re 7b .................................................................................................................... xv i F igu re 8a ................................................................................................................... x v ii Figure 8b ......................................................................................................... xvii Figure 9a .................................................................................................................. xviii F igu re 9b .................................................................................................................. xviii Figure 10 ................................................... xix F igu re 11 ...................................................................................................................... xx F igu re 12 a ................................................................................................................. x xi Figure 12b . xxi F igu re 13a ................................................................................................................. x x ii F igu re 13b ................................................................................................................. xx ii F igu re 14 .................................................................................................................. xx iii iv

Figure 15 .................................................................................................................. xxiv Figure 16a .................................................................................................................. xxv Figure 16b .................................................................................................................. xxv Figure 17 ........................................................................................................... xxvi Figure 18a ............................................................................................................... xxvii Figure 18b............................................................................................................. xxvii Figure 19a .............................................................................................................. xxviii Figure 19b ............................................................................................................. xxviii Figure 20 .................................................................................................................. xxix Figure 21 a .................................................................................................................. xxx Figure 2 1b ................................................................................................................... xxx Figure 22 .................................................................................................................. xxxi Figure 2,3 ................................................................................................................. xxxii Figure 24a .............................................................................................................. xxxiii Figure 24b ............................................................................................................. xxxiii Figure 25a .............................................................................................................. xxxiv Figure 25b ............................................................................................................. xxxiv Figure 25c ..............................................................................................  ;............... xxxiv Figure 26 ................................................................................................................... xxxv Figure 27 ...........................................................................  !.................................... xxxvi Figure 28a ............................................................................................................ xxxvii Figure 28b ............................................................................................................ xxxvii Figure 29a .......................................................................................................... xxxviii Figure 29b ............................................................................................................ xxxviii Figure 30a .............................................................................................................. xxxix v

Figure 30b ............................................................................................................. xxxix F ig u re 3 1a .................................................................................................................... xl Figure 3 1b .................................................................................................................... xl Figure 32a ................................................................................................................... xli Figure 32b ................................................................................................................... xli F igu re 3 3 .................................................................................................................... x lii Figure 34a ................................................................................................................. xliii Figure 34b ................................................................................................................. xliii Figure 34c ........................................................................................................ xliii Figure 35a ................................................................................................................. xliv Figure 35b ................................................................................................................. xliv Figure 35c ...................................................................... xliv Figure 36 ................................................................................................................... xlv Figure 37a ................................................................................................................. xlvi Figure 37b ................................................................................................................. xlvi Figure 38 .................................................................................................................. xlvii Figure 39a .............................................................................................................. xlviii Figure 39b ............................................................................................................... xlviii Figure 40 ................................................................................................................... xlix Fig u re 4 1a ...................................................................................................................... I F ig ure 4 1b ...................................................................................................................... I F ig u re 4 2 ....................................................................................................................... li F igu re 4 3 ...................................................................................................................... lii F ig u re 4 4 a ................................................................................................................... liii Figure 44b ................................................................................................................... liii vi

Executive Summary This report evaluates whether entrainment and impingement by the respective cooling water intake structures ("CWIS") at Indian Point Unit 2 ("IP2") and Indian Point Unit 3 ("IP3")

have caused an adverse environmental impact ("AEI"), using biologically-based definitions of AEI that are consistent with established definitions and standards of ecological risk assessment and fisheries management.

The approach involves three elements. First, we use the extensive Hudson River fisheries datasets to determine (1) whether changes in the status of species of interest identified by the New York State Department of Environmental Conservation ("NYSDEC") have occurred since IP2 and IP3 began commercial operation, (2) whether cooling-water withdrawals by IP2 and IP3 during this period could have been responsible for any such changes, or (3) whether alternative stressors including striped bass predation, zebra mussels, and harvesting are the more probable cause of perceived changes.

Second, we use a widely-accepted method for quantifying the impacts of harvesting on the sustainability of fish populations, termed the Spawning Stock Biomass per Recruit

("SSBPR") model, to determine whether entrainment and impingement at IP2 and IP3 could have adversely affected the sustainability of the Hudson River striped bass and American shad populations.

Third, we examine long-term trends in the abundance of all Hudson River fish species for which adequate trends data sets can be developed to determine whether species with high susceptibility to entrainment at IP2 and IP3 are more likely to have declined in abundance over the past 30 years than are species with low susceptibility to entrainment.

All three elements of the assessment support a conclusion that IP2 and IP3 have not caused an AEI. Evaluation of alternative hypotheses concerning the causes of changes in abundance of Hudson River fish populations found no evidence supporting the hypothesis that IP2 and IP3 contributed to these changes. Instead, the evaluation shows that overharvesting is the most likely cause of recent declines in abundance of American shad, with striped bass predation being a' potentially significant contributing factor. Increased predation by the rapidly growing Hudson River striped bass population is the most likely cause of recent declines in the abundance of Atlantic tomcod, river herring and bay anchovy. Striped bass predation probably 1

contributed to the decline in abundance of white perch, although other unknown causes were also involved.

Two additional lines of evidence support a conclusion that entrainment and impingement at IP2 and IP3 have not resulted in AEI. Application of the SSBPR model to stock assessment data for striped bass and American shad shows that mortality caused by entrainment at IP2 and IP3 is negligible, particularly compared to fishing mortality, and does not impair the ability of these populations to sustain themselves. Analysis of community-level trends data show that species with relatively high susceptibility to entrainment at IP2 and IP3 are no more likely to have declined in abundance since 1974 than are species with relatively low susceptibility to entrainment.

Considered together, the evidence evaluated in this report shows that the operation of IP2 and IP3 has not caused effects on early life stages of fish that reasonably would be considered "adverse" by fisheries scientists and/or managers. The operation of IP2 and IP3 has not destabilized or noticeably altered any important attribute of the resource.

2

Glossary Ichthyoplankton: Eggs and larvae of fish with limited swimming abilities that float in the water-column and are passively transported by currents Entrainment: The drawing of ichthyoplankton and other small aquatic organisms through a cooling water intake structure into the cooling system of a power plant Impingement: The trapping of fish and other aquatic organisms against intake screens by the force of the water being drawn through a cooling water intake structure Individual: A single organism Population: A group of plants, animals, or other organisms, all of the same species, that live together and reproduce Community: An assemblage of species populations that occur together in space and time Yolk-sac larvae (YSL): Fish larvae that have recently hatched and are still receiving nutrition from yolk deposited in the eggs before they were spawned Post yolk-sac larvae (PYSL): Fish larvae that have absorbed the yolk and obtain nutrition by feeding Young-of-the-year (YOY): Fish that have completed the transformation from the larval to the juvenile stage and have grown large enough to be captured by the gear used in the generators' Beach Seine Survey and Fall Shoals Survey Longitudinal River Survey (LRS): The Hudson River generators' annual riverwide ichthyoplankton survey 3

Beach Seine Survey (BSS): The Hudson River generators' annual survey of YOY and older fish abundance in the shorezone Fall Shoals Survey (FSS): The Hudson River generators' annual survey of YOY and older fish abundance in the shoal zone Early life stage: The collective term for the egg, YSL, PYSL, and early juvenile (juveniles too small to be captured by the gear used in the BSS and FSS) life stages Conditional mortality rate (CMR): A measure of the mortality imposed on a population by a stressor such as a cooling water intake structure Recruit: A fish that has grown large enough to be caught in gears used by agencies performing stock assessments for harvested fish species; as used in the spawning stock biomass per recruit model, a one-year-old fish Spawning stock biomass per recruit (SSBPR): The expected lifetime reproduction of a typical female recruit, measured in terms of the expected future egg production or biomass Density-dependence: A relationship between the abundance of a population and the growth rates or mortality rates of individuals belonging to that population Stressor: An anthropogenic or environmental factor that increases mortality or decreases growth of organisms belonging to a population exposed to that factor j

Stressor metric: A measure of the intensity Of a stressor Response metric: A measure of the response of an exposed population to one or more stressors 4

1. Introduction This report evaluates whether entrainment and impingement by the respective cooling water intake structures ("CWIS") at Indian Point Unit 2 ("IP2") and Indian Point Unit 3 ("IP3")

has caused an adverse environmental impact ("AEI"), as that term is employed in §316(b) of the Clean Water Act ("CWA") and 6 NYCRR §704.5 and reasonably may be interpreted by the scientific community.' Our evaluation of whether entrainment and impingement by the respective CWIS at IP2 and IP3 has caused AEI is based on biologically-based definitions of "adverse environmental impact" consistent with established definitions and standards of ecological risk assessment (USEPA 1998) and fisheries management (Restrepo et al. 1998, Quinn and Deriso 1999). Our approach involves three elements.

First, we use the extensive Hudson River fisheries datasets (prepared under the direction and oversight of the New York State Department of Environmental Conservation ("Department" or "NYSDEC")) to determine (1) whether changes in the status of species of interest identified by NYSDEC have occurred since IP2 and IP3 began commercial operation, (2) whether cooling-water withdrawals by IP2 and IP3 during this period could have been responsible for any such changes, or (3) whether alternative stressors including striped bass predation, zebra mussels, and harvesting are the more probable cause of perceived changes.

..Second, we use a widely-accepted method for quantifying the impacts of harvesting on the sustainability of fish populations, termed the Spawning Stock Biomass per Recruit

("SSBPR") model, to determine whether entrainment and impingement at IP2 and IP3 could have adversely affected the sustainability of the Hudson River striped bass and American shad populations.

Third, we examine long-term trends in the abundance of all Hudson River fish species for

'Which adequate trends data sets can be developed to determine whether species with high As applicable here, the CWIS for IP2 and IP3 extend from the point at which water is withdrawn from the Hudson River (the "River") up to, and including, the intake pumps. See, e.g., In Re Matter of Bowline, LLC, 2001 WL 1587359 (N.Y. Dept. Env. Conserv.) (Nov. 30, 2001), at *6-7 (relying on USEPA definition, now codified at 40 C.F.R § 125.93); 40 C.F.R. § 125.93. The CWIS at IP2 and IP3 are shown schematically in Figures IV-12 through IV-15 of the Draft Environmental Impact Statement for State Pollutant Discharge Elimination System Permits for Bowline Point, Indian Point 2 & 3, and Roseton Steam Electric Generating Stations, dated December 1999 (ihe "DEIS"), subsequently incorporated into the Final Environmental Impact Statement by the New York State Department of Environmental Conservation, accepted June 25, 2003 (the "FEIS"). See FEIS, p. 12. These intake structures generally commence with bar racks and debris barriers at the point of entry, include modified Ristroph traveling screens and fish return systems upstream of the point of entry, and terminate with the circulating water pumps.

5

susceptibility to entrainment at IP2 and IP3 are more likely to have declined in abundance over the past 30 years than are species with low susceptibility to entrainment.

Although the technical analyses documented in this report emphasize entrainment, the conclusions reached apply to the combined impacts of entrainment and impingement. There are two reasons for this. First, the trends data that are the primary focus of this assessment reflect the combined effects of entrainment and impingement. Second, entrainment is the focus of the Department, as the existing retrofits (i.e., Ristroph screens and fish returns) have resolved the Department's concerns regarding impingement (Draft SPDES Permit, Special Condition 27).

2. Approach to Impact Assessment Populations2 and communities3 are the proper focus for evaluating adverse impacts of cooling-water withdrawals on the Hudson River estuary. The fundamental reason for focusing on populations and communities is that, whereas all individual organisms have finite life spans, populations and communities can persist. Because populations and communities can persist in spite of the inevitable mortality of the individual organisms, populations and communities can be managed and restored. Most commonly, fisheries management agencies establish harvesting policies to manage populations of fish while allowing harvesting of individual fish to continue (Restrepo et al. 1998). The U.S. Environmental Protection Agency ("USEPA") develops biological assessment methods, based on measures of aquatic community composition, to help states, tribes, territories, and interstate commissions identify communities that are impaired and in need of restoration (USEPA 2002). Established principles of population and community ecology underly both fisheries management and biological assessment. These scientific disciplines also provide a sound foundation for assessing impacts of entrainment and impingement on the biological resources of the Hudson River.

Our evaluation is primarily based on an analysis of empirical data collected over the 30 years during which IP2 and IP3 have been operating, in a manner that appropriately accounts for other potential causes of changes in fish populations. This is because factors other than entrainment and impingement affect the abundance of fish populations, including short-term 2 A population is a group of plants, animals, or other organisms, all of the same species, that live together and reproduce (Gotelli 1995).

3 A community is an assemblage of species populations that occur together in space and time (Begon et al. 1996).

6

natural environmental fluctuations, long-term environmental change, introductions of exotic species, pollution, and over-harvesting (Pew Oceans Commission 2003). The preamble to USEPA's Phase II Rule, 69 Fed. Reg. 41588 (July 9, 2004), also acknowledges the potential influence of these factors on Hudson River fish populations. Where potentially adverse changes in Hudson River fish populations have occurred over the past 30 years, we attempt to determine whether those changes are reasonably attributable to entrainment and impingement, or whether they are more likely to have resulted from other factors.

This impact assessment focuses on eight of the ten species identified for quantitative assessment in NYSDEC's October 1, 1992 Scope of Work for the DEIS: (1) striped bass; (2) white perch; (3) American shad; (4) Atlantic tomcod; (5) alewife; (6) blueback herring; (7) bay anchovy; and (8) spottail shiner. All of these species have been included in §316(b) studies for Indian Point and other Hudson River power plants since the 1970s (TI 1980). Six of these species, striped bass, white perch, Atlantic tomcod, alewife, and bay anchovy, were listed by USEPA as Representative Important Species ("RIS") for the Hudson River (TI 1980). Although not officially listed as RIS, blueback herring was included in the list of species studied because of its abundance in impingement collections at Indian Point, and American shad was included because of its commercial importance (TI 1980).

NYSDEC finalized the Scope of Work for the DEIS following a public scoping meeting and the integration of comments received from the generators, state and federal agencies, and environmental organizations. Two of the species identified in the Scope of Work, blue crab and shortnose sturgeon, are not addressed in this report. These two species are not addressed here because there is broad consensus that the CWIS at IP2 and IP3 have no impact on these species.

See, e.g., DEIS, p. V-125, 126 (sturgeon); Technical Comments on the DEIS, Pisces Conservation, Ltd., June 2000 ("Pisces Comments"), p. 27 ("There seems no basis for suggesting that power plants are linked to [changes in Atlantic and shortnose sturgeon abundance]."); DEIS,

p. V-157 (based on preferred habitat, blue crab eggs and larvae not entrained at IP2 and IP3; very high impingement survival); Pisces Comments, p. 28-29 (numbers of blue crab within the estuary have risen dramatically since 1980).

7

2.1 Definition of "adverse environmentalimpact" Neither §316(b) of the CWA (including USEPA's Phase II Existing Facilities Rule), nor New York regulation provides a definition of the term "adverse environmental impact." See, e.g., 6 NYCRR §704.5. However, both regulations governing fisheries management in the United States and other USEPA guidance provide a foundation for a scientifically appropriate definition of this term.

2.1.1 Definition of adverse environmental impact in the context of fishery management In the context of fisheries management, mortality per se could not be considered an AEI, because the act of fishing necessarily causes mortality. To the contrary, fisheries management agencies, including NYSDEC, actively encourage the responsible harvesting of fish. For example, NYSDEC has issued a guide to saltwater fishing in the New York City area (http://www.dec.ny.gov/outdoor/8377.html) that discusses equipment, fish identification, and specific fishing locations in all five New York City boroughs.

Fishery policy in waters under the control of the U.S. federal government, including estuaries and rivers utilized by anadromous fish, is established in the Magnuson-Stevens Fishery Conservation and Management Act ("Magnuson-Stevens Act"). The amended Act states:

Fishery resources are finite but renewable. If placed under sound management before over-fishing has caused irreversible effects, the fisheries can be conserved and maintained so as to provide optimal yields on a continuing basis.

16 U.S.C. §1801(a)(5).

Federal guidelines implementing the Magnuson-Stevens Act state that "[c]onservation and management measures shall prevent over-fishing while achieving on a continuing basis, the optimum yield ("OY") from each managed fishery for the U.S. fishing industry." 70 Fed. Reg.

36240, 36250 (June 22, 2005). Thus, a fish population is viewed by managers as a renewable resource for which mortality in the form of harvesting is permissible, provided that this mortality does not threaten the long-term productivity of the population. Over-fishing that threatens the long-term sustainability of harvests is considered to be adverse. The National Oceanic and Atmospheric Administration ("NOAA") guidelines and other related technical guidance documents (e.g., Restrepo et al. 1998) provide specific procedures for determining whetherover-8

fishing is occurring. Fishery management councils are required to take action to reduce harvest levels if over-fishing is found to exist. 70 Fed. Reg. 36240, 36257 (June 22, 2005).

The Magnuson-Stevens Act is often cited as the "Sustainable Fisheries Act." The term "sustainable" is often used in. a wider environmental policy context to refer to an approach to economic development and resource utilization that meets the needs of the present without compromising the ability of future generations to meet their own needs (World Commission on Environment and Development 1987). Sustainable uses of resources preserve those resources for future use; non-sustainable uses degrade or destroy the resources so that they may be unavailable in the future (World Commission on Environment and Development 1987).

Applying the definition of sustainable use provided by the World Commission on Environment and Development, sustainable use in the context of a fish population refers to a resource-management approach that permits the population to persist indefinitely into the future, while continuing to perform its normal ecological function and support normal human use.

Ecological function is included as part of the definition of sustainable use of fish populations because fish have a role in the maintenance of healthy aquatic systems that can be compromised by over-fishing (Dayton et al. 2002). Predatory fish, such as striped bass, control the abundance of other fish species upon which they prey, and forage fish, such as bay anchovy, serve as both food for other fish species and as controls on the abundance of smaller organisms at the base of the marine food chain (Dayton et al. 2002). Over-fishing has led to a wide variety of direct and indirect changes in the structure and function of fish communities throughout the world (Dayton et al. 2002).

The sustainability of a population is a function of the abundance and other characteristics of the population (e.g., age and size structure) and also of the ability of members of the population to reproduce and replace themselves. Thus, with respect to the harvest-related mortality imposed on a fish population, an adverse impact consists of harvest-related reductions in abundance, changes in age/size structure, increases in mortality rates, or reduction in reproduction rates that threaten the capacity of the population to persist, perform its normal ecological function, and support normal human uses.

9

2.1.2 Definition of AEI in the context of ecological risk assessment USEPA's Guidelines for Ecological Risk Assessment (USEPA 1998) provide a general discussion of adverse ecological effects' of environmental stressors, including criteria for evaluating whether or not observed or predicted changes should be considered adverse. These guidelines were expressly issued to "set forth current scientific thinking and approaches for conducting and evaluating ecological risk assessments" (USEPA 1998, p. 8). This guidance discusses advei'se ecological effects of environmental stressors, including criteria for evaluating whether or not observed or predicted changes should be considered adverse. According to USEPA and the scientific community, adverse ecological effects are changes that "alter valued structural or functional attributes of the ecological entities under consideration" (USEPA 1998,

p. 106). USEPA (1998, p. 106) further states that the following criteria should be considered when determining whether an observed or predicted effect is adverse:

• Nature and intensity of effects;

• Spatial and temporal scale; and

• Potential for recovery.

"Nature and intensity of effects" refers to the types of effects that have occurred (or are predicted to occur), and the magnitude of the measured or predicted effects, the statistical significance of measured effects, and the ecological significance. of the effects. "Spatial and temporal scale" refers to the size and location of the area within which an effect occurs, and the duration of the period required for the effect to appear. "Potential for recovery" refers to the expected rate and extent of return of an affected population or community following elimination of the stressor responsible for an effect that has been determined to be ecologically significant.

USEPA's definition and criteria for determining ecological adversity are consistent both with accepted principles of fishery management and with the current scientific understanding of the potential effects of harvesting on fish populations and communities. As noted in the introduction to this Section, in the context of §316(b) and §704.5, the ecological entities of interest are the populations and communities potentially affected by entrainment at CWIS. A definition of AEI of CWIS consistent with the Guidelines for Ecological Risk Assessment (USEPA 1998) should be expressed in terms- of undesirable alterations in the structural or functional attributes of these populations and communities. An assessment whether adverse 10

impacts have occurred (or will occur) should address the three criteria provided in the Guidelines.

2.1.3 Definition of adverse environmental impact in the context of entrainment and impingement The definition of sustainable use in the Magnuson-Stevens Act and the definition of ecological adversity in USEPA's Guidelines for Ecological Risk Assessment provide a reasoned basis for a definition of AEI applicable to entrainment and impingement at CWIS. A sustainable approach to managing a fishery would ensure the long-term persistence and productivity of the population being managed. A non-sustainable approach, in contrast, would cause harvest-related reductions in abundance, changes in age/size structure, increases in mortality, or reductions in reproduction that could threaten the capacity of a population to persist, perform its normal ecological function, and support normal human uses. Since the ecological function of a population is understood by scientists to include interactions with other populations, non-sustainable use of a population can affect an entire community.

Abundance, age/size structure, mortality, and reproduction are examples of the "structural and functional attributes" discussed in the USEPA Guidelines. Hence, non-sustainable management of a fishery would be an example of an AEI according to USEPA's definition. Entrainment mortality differs from mortality caused by harvesting only in that the mortality is imposed on early life stages of fish or shellfish rather than on adults. Excessive levels of entrainment mortality could potentially affect most of the same structural and functional attributes affected by harvesting.

In sum, the term AEI, as it relates to entrainment and impingement, is reasonably and appropriately defined as follows:

An adverse environmental impact due to entrainment and impingement consists of adverse changes in important population or community characteristics sufficient to threaten the sustainability of susceptible populations or to cause significant or potentially irreversible changes in population or community structure and function.

Such a definition would be consistent with recognized principles of both natural resource management and ecological risk assessment, as discussed above.

11

2.2 Why entrainmentlosses alone are insufficient to demonstrateAEI Context is essential to understanding what the term AEI reasonably may mean with respect to fisheries biology. As a matter of science and logic, losses, even large numbers of early life stage individuals do not necessarily equate to AEI. This is because fish species inhabiting the Hudson River exhibit either "periodic" or "opportunistic" life history traits (Winemiller and Rose 1992). -From an ecological perspective, periodic fish species are characterized by high fecundity (i.e., they spawn a large number of eggs), large size, and long life spans during which a female fish may spawn many times (Winemiller and Rose 1992). Striped bass is an example of a periodic species (Winemiller and Rose 1992). Opportunistic species are characterized by small body size, short life spans, and the ability to disperse offspring widely throughout the environment (Winemiller and Rose 1992). Bay anchovy is an example 'of an opportunistic species. Periodic and opportunistic traits are advantageous to fish species that live in unstable or unpredictable environments, such as the Hudson River, which experiences significant within-year and between-year variation in environmental conditions (e.g., temperature, salinity, freshwater flow, etc.). In other words, the reproductive strategies of these fish in these unstable conditions, including the very large numbers of eggs produced, ensure that sufficient offspring will survive to sustain the populations, even in unstable environments characterized by the presence of multiple stressors.

Entrainment losses consist mainly of eggs and larvae. Only a small fraction of the entrained fish would survive to adulthood, even if IP2 and IP3 did not exist. For example, an 18-year-old Hudson River striped bass was found to contain more than 3 million eggs (Hoff et al.

1988). A 16-year-old female striped bass examined by Olsen and Rulifson (1992) was found to contain nearly 5 million eggs. Since striped bass can live for up to 30 years (Secor and Piccoli 1996), a single fish could potentially spawn tens of millions of eggs over her entire lifespan.

According to early life stage survival estimates developed by Secor and Houde (1995), more than 99.99% of young striped bass eggs die from natural causes within 60 days following spawning. Less than one striped bass egg in 100,000 is likely to survive to become a one-year-old fish, and less than one in a million is likely to survive to reach six years of age, the median age at which female striped bass become sexually mature (EPRI 2005).

12

Because nearly all of the eggs and larvae entrained at IP2 and IP3 would have died in any case, counts of total numbers entrained reveal nothing meaningful about the potential impact of IP2 and IP3 on fish populations. What matters is whether or not entrainment significantly reduces the number of fish that survive the early period of high natural mortality. As discussed in the next sections, this fact was recognized more than 30 years ago by the scientists who performed the first entrainment impact assessments for IP2 and IP3, in conjunction with other Hudson River generating stations.

2.3 Role of the conditionalmortality rate (CMR) in impact assessment The first assessments of the effects of cooling-water withdrawals on Hudson River fish populations, conducted on behalf of the Consolidated Edison Company of New York and various federal regulatory agencies were based on mathematical models that predicted the potential effects of entrainment losses on the abundance and other characteristics of fish populations, especially striped bass (Barnthouse et al. 1984). Many of these models were developed to support U.S. Atomic Energy Commission licensing proceedings for IP2 and IP3, and were incorporated in environmental impact statements prepared to support these proceedings (Barnthouse et al. 1984). At the time they were first developed, in the early and mid-1970s, modeling was undertaken because no actual fisheries data were available to test whether cooling-water withdrawals would have adverse impacts on important fish populations. When data from riverwide ichthyoplankton sampling became available in the late 1970s, scientists studying entrainment impacts developed an empirical model, termed the Empirical Transport Model

("ETM", Boreman et. al. 1981), and used it to estimate the impact of entrainment on the abundance of juvenile fish. The metric calculated using the ETM, which was termed the

conditional mortality rate" ("CMR"), provides an estimate of the fraction by which the abundance of young-of-the-year fish is reduced due to entrainment. A similar model, termed the Empirical Impingement Model ("EIM", Barnthouse and Van Winkle 1988), was used to estimate a CMR for impingement.

It was recognized at the time that the CMR could not be used to predict long-term impacts on populations, however, because neither the ETM, nor the EIM, accounts for the density-dependent processes that can partially offset mortality due to entrainment and 13

impingement (Barnthouse et al. 1984). CMRs could, however, be used to compare the relative potential effectiveness of alternative technologies intended to reduce entrainment and impingement mortality. As discussed by Englert et al. (1988), CMRs calculated using the ETM also were used to develop the cross-plant outage credits that were included in the Hudson River Settlement Agreement ("HRSA"). CMRs were also used in the DEIS to compare alternative entrainment mitigation approaches. In all of these applications, CMRs were used usefully as measures of mortality caused by entrainment and impinfgement, not as measures of the 'impacts of that mortality on the long-term abundance or sustainability of susceptible populations.

Because it does not account for density-dependent effects, the CMR is not a valid measure of long-term entrainment impacts. Depending on the strength of density-dependence in a given population, a particular CMR value corresponds to either a negligible or a substantial impact on the sustainability of a population. 4 CMRs can, however, be used as a measure of the annual rate of mortality imposed by entrainment and as inputs to assessment models that estimate the combined impacts of entrainment mortality and fishing mortality on the sustainability of populations (Goodyear 1977, 1993). For this assessment, CMRs are used for both of these purposes. They are not, however, used as measures of AEI, because CMRs are not appropriately used in that fashion and superior methods for assessing adverse impacts are available. As discussed in the following sections, analysis of long-term trends in the abundance of important Hudson River fish populations, available from .30 years of intensive data collection, is the best method available for assessing impacts of IP2 and IP3 on Hudson River fish populations. The trends analysis is supplemented by an analysis of the impacts of IP2 and IP3 on the sustainability of the Hudson River striped bass and American shad populations, using the SSBPR model.

2.4 Role of long-term datasets in impact assessment Today, nearly 30 years of data are available from both generator and agency-sponsored monitoring programs. Together, these overlapping datasets provide information concerning long-term trends in the abundance and distribution of eggs, larvae, and juveniles of all of the species addressed in this report. For some commercially harvested species, data on long-term 4 Although there can be substantial uncertainty concerning the strength of density-dependence in specific populations, there is strong theoretical and empirical evidence that the great majority of biological populations, including fish populations, are regulated in part by density-dependent mechanisms (Murdoch 1994, Turchin 1999, Rose et al. 2001, Brook and Bradshaw 2006).

14

trends in the abundance, age distribution, and mortality of adult fish are available. These datasets can be used both to assess trends in the status of important fish populations and to test alternative hypotheses concerning potential causes of adverse changes.

In this report, information concerning long-term trends on key population characteristics and on the intensities of potential stressors is used to test specific hypotheses concerning the expected impacts of cooling-water withdrawals, termed "risk hypotheses" in USEPA's Guidelines for Ecological Risk Assessment (USEPA 1998). These hypothesis tests are used to distinguish changes that could have been caused by cooling-water withdrawals from changes that are most likely related to other causes.

The following generator-sponsored long-term datasets are the primary datasets used in assessing the effects of the CWIS at IP2 and IP3:

Longitudinal River Ichthyoplankton Survey ("LRS"). This program samples eggs, larvae, and juvenile fish, weekly from April through July.

The region between the George Washington Bridge and the Federal Dam at Troy ( Figure 1) has been sampled with only minor changes in methodology since 1974. In 1988, the LRS was extended to sample the region between the Battery and the George Washington Bridge.

Beach Seine Survey ("BSS"). This program samples juvenile fish, also called "young-of-the-year" fish ("YOY") (i.e., fish spawned earlier in the year) on alternate weeks from June through October. Sampling is conducted from the George Washington Bridge to the Federal Dam. The BSS has been conducted annually with only minor changes in methodology since 1974.

Fall Shoals Survey ("FSS"). This program samples YOY and older fish in offshore habitats, on alternate weeks from the BSS. Approximately 200 samples are collected per week, from Manhattan to the Federal Dam. The FSS uses two different gears in order to sample as much of the Hudson River as possible: a 1-mi2 Tucker trawl and a 3-m beam trawl. This 15

program was also initiated in 1974, however, the beam trawl was not used until 1985. From 1974 through 1984 an epibenthic sled was used to sample near the river bottom. To ensure comparability between years, only the data collected from 1985 onward are used in this assessment.

Atlantic Tomcod Mark-Recapture Program. This program has been conducted in most years since 1974 to generate estimates of the number of tomcod in the winter spawning population. 5 Box traps and bottom trawls are used to collect fish for marking and recapture.

The above datasets were selected as the primary datasets for this assessment because they have been conducted continuously since the mid-1970s. They cover nearly all of the period of commercial operation of IP2 (1973 startup) and all of the period of commercial operation of IP3 (1976 startup). These four datasets provide the most comprehensive and consistent estimates of long-term trends in the abundance of multiple life stages of important Hudson River fish populations. More detailed descriptions of these datasets are provided in ASA (2007).

A variety of other programs, conducted by the generators, NYSDEC, and federal resource management agencies provide information that can be used to test the validity of the primary trends data. These programs include:

Striped Bass Mark-Recapture Program. This program was initiated in 1984, to estimate the contribution of the Hudson River striped bass hatchery (established as a condition of the HRSA) to the Hudson River population. The program targets 1-year-old and 2-year-old striped bass, and is conducted from November through March. Data from this program are used to estimate the numbers of striped bass greater than 150 mm in length overwintering in the lower estuary. Growth and survival rate estimates are also obtained from this program.

5 The program was not conducted in 1984 and 1986.

16

NYSDEC Beach Seine Survey. Since 1976, the NYSDEC Division of Marine Resources has conducted a beach seine survey in the lower Hudson River estuary. The program focuses on the Tappan Zee and Haverstraw Bay. It samples juvenile fish using a method similar, but not identical to, the generators' BSS.

Juvenile Alosid Survey. NYSDEC conducts a beach seine survey in the middle and upper regions of the estuary (above River Mile 55) to estimate the relative abundance of YOY American shad and other juvenile fishes.

This program was initiated in 1980 and continues to the present.

Western Long Island Survey. NYSDEC conducts a survey for subadult striped bass in the bays around western Long Island Sound. Sampling is conducted using a 200-ft. beach seine. The program was initiated in 1984 and is continuing, although it has been modified over time.

Spawning Stock Assessment. NYSDEC conducts a haul seine survey in the Hudson River to provide information on length, age and sex distribution, and mortality rates for adult American shad and striped bass.

The program was initiated in 1982 and continues to the present.

Commercial Fishery Monitoring. NYSDEC monitors the commercial gill net fishery for American shad. The objective of the program is to determine the relative abundance and age structure of the commercial catch of American shad.

As shown in Appendix A, indices derived from these datasets are strongly correlated with indices derived from the primary datasets. These correlations support the use of the primary datasets in this assessment.

In addition to the Hudson River monitoring programs, information on population status and trends for important fish species is also available from the National Marine Fisheries Service 17

("NMFS") and the Atlantic States Marine Fisheries Commission ("ASMFC"). Quantitative stock assessments, which include estimates of age structure, natural mortality, and fishing mortality, are available for striped bass (ASMFC 2005) and American shad (ASMFC 2007a).

These assessments provide additional information for determining whether these populations have been harmed by CWTS.

2.5 Indicatorsof adverse impactspotentially relatedto CWIS As discussed above, an adverse impact of CWIS would consist of entrainment and impingement-related adverse changes in important population or community characteristics sufficient to threaten the sustainability of relevant populations, or to cause significant or potentially irreversible changes in community structure and function. Characteristics that influence the sustainability of a fish population include the total size of the po'pulation, the relative abundances of different life stages or age groups, the sizes and reproductive rates of the individual fish, and the rates of mortality of fish at different life stages or ages. Measures of any of these population characteristics could, at least in principle, be used as indicators of adverse impact. Some of these measures are not suitable as indicators of adverse impacts potentially caused by CWIS, however, because they measure changes that cannot be reasonably attributed to cooling-water withdrawals. For example, a reduction in fecundity could be an indicator of a potential impact caused by a toxic chemical but, because impingement and entrainment do not affect fecundity, this characteristic is not an appropriate indicator of impacts caused by CWIS.

Similarly, some indicators of impact are not particularly useful in narrowing the potential causes of impacts. For example, a prolonged downward trend in the abundance of adult fish could be the result of any number of causes, including over-fishing or environmental factors.

CWIS may impose mortality on early life stages of fish (i.e., eggs, larvae, and YOY) in addition to the mortality that would have occurred naturally. Therefore, characteristics that are either directly or indirectly affected by increased mortality of these life stages are potentially useful as indicators of harm related to CWIS. Increased mortality imposed on a particular life stage would reduce the fraction of organisms in that stage that survive to the next stage.

Accordingly, this assessment focuses on whether CWIS have had a measurable influence on the survival of early life stages of fish in the Hudson River.

18

As discussed in Section 2.1 of this report, however, mortality of early life stages as a result of CWIS is insufficient, of itself, to establish that an adverse impact has occurred. It is necessary, in addition, to evaluate whether the magnitude, spatial extent, and duration of this mortality are large enough to constitute an adverse impact (USEPA 1998). Fisheries scientists have developed metrics, termed "biological reference points," for determining whether harvested fish populations are being harmed by over-fishing (Restrepo et al. 1998). These reference points, expressed in terms of either the total spawning stock biomass ("SSB") or the SSBPR, are viewed as indicators of the risk that over-fishing will lead to future declines in abundance and harvest.

The methods that fisheries scientists use to estimate effects of fishing mortality on SSB and SSBPR can also be used to estimate impacts of entrainment-related mortality on SSB and SSBPR (Goodyear 1993). Hence, the indicators used to determine whether fish populations are being adversely affected by fishing can also be used as indicators of whether these. same populations are being adversely affected by cooling-water withdrawals. Accordingly, for species for which published agency stock assessment reports provide relevant information, this assessment addresses whether the magnitude of entrainment mortality (as measured using the CMR) is sufficient to produce an ecologically significant reduction in SSB or SSBPR.

Information needed to estimate SSBPR is available for both striped bass and American shad. A coastwide SSB estimate is available for striped bass.

The following indicators have been selected for this assessment:

1. Long-term declines in the abundance of Y0Y fish belonging to species with life stages susceptible to impingement and entrainment, see, infra, Section 3;

2. Reductions in the spawning potential of female fish below the sustainable level as estimated using the SSBPR approach, see, infra, Section 4; and

3. Long-term trends in the abundance of species with high susceptibility to entrainment at IP2 and IP3 as compared to species with low susceptibility to entrainment at IP2 and IP3, see, infra, Section 5.

19

The analyses documented in Sections 3, 4, and 5 of this report evaluate whether any such declines or reductions in spawning potential have occurred and, if so, whether they may reasonably be attributed to the CWIS of IP2 and IP3.

3. Evaluation of changes'in abundance~of fish populations with life stages susceptible to entrainment In complex ecological systems, such as the Hudson River estuary, fish populations are influenced by many factors in addition to CWIS, including water quality impairment, introductions of non-native species, and overfishing (Pew Oceans Commission 2003). Many of these factors are discussed in the preamble to USEPA's Final Phase II Existing Facilities Rule.

69 Fed. Reg. 41575, 41588 (July 9, 2004). For this reason, investigations of the causes of changes in fish populations must consider multiple hypotheses, weighing the evidence for and against each hypothesis (Hilborn and Mangel 1997, Suter et al. 2007). This approach has been termed "ecological detection" by Hilborn and Mangel (1997) and "ecoepidemiology" by Suter et al. (2007).

Most environmental factors affecting Hudson River fish populations vary in intensity over time. Knowledge of these variations can be used to predict the change in each metric that should have occurred, if that stressor had been affecting a particular fish population. To test each hypothesis, this analysis utilizes rules for evaluating causal associations provided by Suter et al.

(2007, p. 50). These authors identified five criteria that should guide analyses of potential causes of adverse environmental effects:

1. Co-occurrence: An effect occurs where and when its cause occurs and does not occur in the absence of its cause.

2. Sufficiency: The intensity or frequency of a cause should be adequate to produce the observed magnitude of effect.

3. Temporality: A cause must precede its effect.

.4. Manipulation: Changing the cause must change its effect.

5. Coherence: The relationship between a cause and effect must be consistent with scientific knowledge and theory.

20

Evaluations of co-occurrence discussed in this sections rely on a commonly-used and relatively straightforward statistical method known as correlation analysis (Clarke and Kempson 1997). In simple terms, correlation is a measure of whether two different variables are related to one another and, if so, how strong that relationship is (Clarke and Kempson 1997). A positive correlation between two variables indicates that as the value of one variable increases, so does the other. For example, height and weight among people are positively correlated. Although some taller people weigh less than shorter people, on average the taller a person is, the more that person is likely to weigh. Conversely, a negative correlation indicates that, as the value of one variable increases, the other decreases (Clarke and Kempson 1997). For example, weight and fuel efficiency among automobiles are negatively correlated. Although some heavier cars get better gas mileage than some lighter cars, on average the heavier a car is, the lower its gas mileage will be.

The existence and strength of correlations between stressor metrics and response metrics provides evidence concerning the co-occurrence criterion. If, for example, entrainment mortality at IP2 and IP3 is reducing thesurvival of eggs and larvae of a particular fish species, then there should be a negative correlation between entrainment mortality and a measure of the fraction of eggs and larvae that survive to reach older life stages. This means that in years when mortality due to IP2 and IP3 is high, survival should be relatively low, and in years when mortality due to IP2 and IP3 is low, survival should be high. Data showing the presence of a negative correlation between early life stage survival and IP2 and IP3-related mortality would constitute evidence supporting this impact hypothesis; data showing the absence of a correlation would constitute evidence against this hypothesis.

Evaluations of sufficiency in this assessment rely on measures of the magnitude of the stressor, as compared to the magnitude required to cause the observed response. For example, the rate of fishing mortality imposed on the striped bass and American shad populations can be compared to overfishing thresholds established by the ASMFC.

Evaluations of temporality in this assessment rely on time trends of the various stressor and response metrics. For any stressor to be a potential cause of a decline in the survival or abundance of a fish population, the decline should be preceded by an increase in the intensity of the stressor. If the decline in survival or abundance precedes the increase in the stressor, then the stressor cannot have caused the decline.

21

Evaluations of manipulation in this assessment rely on observations of responses of populations to deliberate changes in the magnitudes of stressors, e.g., the harvesting restrictions imposed on the striped bass fishery in the 1980s.

Evaluations of coherence in this assessment rely on the consistency of the responses with all relevant scientific information.

Because the focus of the permit proceedings is on entrainment and impingement of age 0 fish, the analysis will focus primarily on age 0 response metrics. The steps in the analysis include:

1. Develop a conceptual model of each stressor, including (1) a description of the stressor itself, (2) the reasonably expected causal mechanisms through which fish populations would be affected, (3) the species that would likely be affected, (4) the life stages (e.g., juveniles)'that would likely be affected, (5) the life history characteristics (e.g., survival and growth) that would likely be affected, and (6) the type of measurable effects that would likely occur (increase or decrease);

2. Identify appropriate sets of "stressor metrics" and "response metrics" that can be used to test the potential influence of the various stressors;

3. Summarize the expected effect of the stressor on each response metric;

4. Apply the five evaluation criteria discussed above to the available data for each fish species; and

5. Summarize conclusions regarding (1) whether changes in the response metrics could have been caused by entrainment by CWIS at IP2 or IP3, or (2) whether other stressors are more likely to be responsible for these changes.

3.1 Species addressed The DEIS assessed entrainment and impingement impacts on striped bass (Morone saxatilis), white perch (Morone Americana), Atlantic tomcod (Microgadus tomcod), bay anchovy (Anchoa mitchilli), American shad (Alosa sapidissima), alewife (Alosa 22

pseudoharengus), blueback herring (Alosa aestivalis), and spottail shiner (Notropis hudsonius)

(DEIS, Sections 5 and 6). This report assesses entrainment and impingement impacts on these same species, focusing on the most economically important species (striped bass) and on the three species (white perch, American shad, and Atlantic tomcod) identified in the draft permit fact sheet as being of potential concern with respect to IP2 and IP3. Fact Sheet, Draft SPDES Permit, Attachment B, at 1 of 8. The datasets used in these analyses are documented in the 2005 Year Class Report (ASA 2007). The stressor and response metrics are documented in Appendix B.

3.2 Impact hypotheses and stressormetrics This section documents expected effects of CWIS and four other stressors that are widely regarded as potentially having affected Hudson River fish populations: fishing, invasion of the Hudson River by zebra mussels (Dresseinapolymorpha), temperature (Atlantic tomcod only) and predation by striped bass.

3.2.1 CWIS CWIS may cause mortality of fish due to entrainment and impingement. For most species, this mortality is largely limited to eggs, larvae, and YOY. Because most of the susceptible life stages are planktonic 6 and are widely dispersed throughout the estuary due to tidal and nontidal flows, cooling-water withdrawals would not be expected to alter the spatial distributions of the affected species. In addition, the CWIS would not be expectedto reduce the survival of fish that have grown through the most susceptible life stages, or to reduce fish growth rates at any life stage.

As discussed in Section 2.3, the CMR is a direct estimate of the rate of mortality caused by entrainment and impingement, independent from natural mortality. Similar measures are used by fisheries scientists to estimate the rate of mortality imposed on adult fish by fishing. The CMR can have values ranging between 0.0 and 1.0. The higher the value of the CMR, the greater the mortality imposed on early life stages of fish.

6 Planktonic organisms are small organisms such as fish larvae that have limited swimming capabilities and are passively transported up and downriver with tidal currents.

23

Expected effects of CWIS on the life stages potentially susceptible to entrainment and impingement (i.e., eggs, larvae, and YOY) are summarized in Figure 2. As shown in Figure 2, CWIS should affect the survival rates of the susceptible life stages, but should not affect the survival of stages that are not susceptible to entrainment or impingement. If entrainment or impingement were having a measurable impact on a fish population, then in years when the IP2 and IP3 CMR is high, the survival rates of susceptible life stages of that species should be lower than in years when the IP2 and IP3 CMR is low. As a consequence, long-term trends in IP2 and IP3 CMR values for that species should be negatively correlated with long-term trends in the survival rates of susceptible life stages.

Although entrainment would not affect the number of eggs spawned by females of susceptible species, it is still possible that entrainment could directly affect the abundance of early life stages. The reason for this is that the LRS is conducted during the period in which entrainment at IP2 and IP3 is occurring. Therefore, entrainment could affect the abundance estimates derived from LRS data. If entrainment at IP2 and IP3 is reducing early life stage abundance, then the IP2 and IP3 CMR values should also be negatively correlated with PYSL abundance estimates.

3.2.2 Fishing Fishing imposes mortality primarily on harvestable-sized 7 fish. 8 For managed Hudson River fish species (i.e., striped bass and American shad), harvesting is largely limited to age 1 and older fish'(ASMFC 1998, 2002). Fishing has predictable effects on the age distribution of adult fish and on the abundance (numbers and biomass) of the spawning stock (Dayton et al.

2002). Measures of age distribution and spawning stock abundance are used by fisheries managers as indicators of fishing (Restrepo et al. 1998). Fishing reduces the total reproductive output of a fish population (Goodyear 1993).

The most appropriate estimate of stress due to fishing is the annual rate of fishing mortality (F) imposed on the population. Estimates ofF for two of the species addressed in this analysis, striped bass and American shad, are available from the ASMFC.

7 Harvestable-size fish are fish that fall within the size range for which harvesting is permitted.

8 Fish outside the permitted range are frequently caught by trawls and other fishing gear. Although they are returned to the ocean, substantial mortality may still occur. This mortality is termed "bycatch" mortality.

24

Expected effects of fishing on age 0 life stages are summarized in Figure 3. Over-harvesting reduces the size of the adult population and necessarily the total number of eggs produced per year. The reduction in egg production would be expected to reduce the number of eggs surviving to become one-year-old fish. Fishing should not reduce the survival or growth rate of any age 0 life stage, however, because early life stages of fish are not susceptible to harvesting.

3.2.3 Zebra mussels Zebra mussels invaded the Hudson River in the early 1990s (Caraco et al. 1997). Zebra mussels form dense beds on the bottom of colonized water bodies. Because of their high filtering capacity, zebra mussels remove phytoplankton from the water column, thus reducing the food base that supports pelagic fish larvae, such as American shad, striped bass, and white perch (Strayer et al. 2004). Because less food is available to support fish species that feed in open water, the survival and growth of these species may decrease. The increased water clarity caused by zebra mussel filtration can result in improved growth of rooted vegetation. The survival and growth of species that inhabit vegetated areas may increase because of increased' habitat availability (Strayer et al. 2004). Zebra mussels are limited to fresh water, and are not found in substantial numbers below approximately river kilometer ("RKM") 100 in the Hudson River.

For this reason, zebra mussels could potentially alter the spatial distributions of some species, reducing their abundance above RKM 100 as compared to below RKM 100.

There is no readily available quantitative metric for zebra mussel abundance. Due to the discontinuous nature of the zebra mussel invasion (absent prior to 1992; highly abundant after 1992), however, the qualitative evaluation can use presence/absence to develop predicted effects, and the quantitative analysis can use a simple index to distinguish between these two periods (e.g., "0" for all years prior to 1993 and "I" for 1993 and later). Expected effects of zebra mussels on age 0 life stages are summarized in Figure 4. Zebra mussels would be expected to reduce the survival and growth rates of post yolk-sac larvae and YOY utilizing freshwater regions of the Hudson River. These changes in survival and growth could result in a shift in the relative abundance of YOY present in predominantly freshwater regions (Regions 6-12; Figure

1) as compared to marine and brackish regions (Regions 0-5; Figure 1). Specifically, if zebra 25

mussel activity reduces the growth and survival of pelagic fish species in freshwater regions as compared to marine and brackish regions, then during the post-invasion period a greater fraction of the populations of pelagic species, such as striped bass, white perch, alewife, and river herring, should be found in marine and brackish regions than during the pre-invasion period.

3.2.4. Predation by striped bass 9

Increased abundance of yearling and older striped bass, which are piscivorous (Gardinder and Hoff 1982, Walter et al. 2003), could lead to increased predation mortality.

Savoy and Crecco (2004) have attributed a recent decline in American shad and blueback herring populations in the Connecticut River to predation by large adult striped bass on spawning adults of these species.

Because the abundance of striped bass early life stages has been found to be strongly correlated with the relative abundance of adults (Pace et al. 1993; Barnthouse et al. 2003),

estimates of striped bass larval abundance from the LRS can be used as a surrogate for adult striped bass abundance.

Predation on adults would, like harvesting, reduce the number of spawning adults and, as a consequence, the number of eggs spawned. The reduction in egg production would be expected to reduce the number of eggs surviving to become one-year-old fish. Predation on YOY would directly reduce YOY abundance, over and above and reductions resulting from reduced egg production (Figure 5).

3.2.5 Temperature Changes in temperature can cause either increases or decreases in the growth and survival of affected species, depending on species-specific temperature tolerances. Long-term trends in Riverwide temperatures could potentially lead to long-term changes in the abundance of sensitive species, such as Atlantic tomcod (FEIS, pp. 65-66). Expected effects of elevated summer temperatures on age 0 temperature sensitive species are summarized in Figure 6.

Elevated summer temperatures would be expected to cause decreases in survival and growth of temperature-sensitive species during this period. Growth and survival of early life stages would 9 Piscivorous fish are fish that eat other fish.

26

not be depressed, however, because these life stages are present only during the winter and early spring, when temperatures would be well below adverse effects thresholds.

According to McLaren et al. (1988), the growth of juvenile Atlantic tomcod in the Hudson River ceases during the summer when river temperatures regularly exceed 25°C. The lethal temperature for juvenile Atlantic tomcod is 26.5°C (McLarenf et al. 1988). Temperature records available from the Poughkeepsie Water Works (PWW) were used to develop a degree-day index for evaluating the potential effects of elevated summer temperatures on Atlantic tomcod. A degree-day is defined as the number of degrees by which the temperature measured at the PWW on that day exceeds 24'. If, for example, the temperature measured at the PWW on a given date was 27°C, then the degree-day value for that date would be 3. If the temperature on a date is 240 or less, then the degree-day value for that date is recorded as 0. The degree-day index for a years is calculated by summing the degree-days for all days during that year.

3.3 Response metrics Because not all data sets are suitable for evaluating all species, the response metrics used in this assessment are not the same for all species.

3.3.1 Response metrics for striped bass, white perch, American shad, alewife, blueback herring, and bay anchovy For species other than spottail shiner and Atlantic tomcod, the LRS and BSS provide the most reliable data concerning survival, growth, and spatial distribution. Because the durations of egg and YSL life stages are comparatively short, such that individuals can hatch and develop through one or both of these stages between survey dates, most of the fish captures in the LRS are PYSL. The PYSL stage is typically much longer, so that PYSL are susceptible to sampling for at least one and possibly two or more survey dates. For these reasons, estimates of total larval abundance from the LRS are best interpreted as estimates of the abundance of PYSL.

Although the beach seine used in the BSS and the beam trawl used in the FSS do not capture larvae, they effectively sample YOY fish present in the sampled habitats (shore zone for the BSS and shoal zone for the FSS). The response variables that can be calculated from the generators' survey data are:

27

1. Abundance of PYSL, as measured in the LRS;

2. Survival from the PYSL to the YOY stage, as measured by the ratio of densities of larvae in the LRS dataset to densities of juveniles in the BSS or FSS,

3. Abundance of YOY, as measured in the BSS or FSS;

4. YOY growth, as measured by the average length of YOY fish from the BSS or FSS; and

5. Spatial distribution of PYSL and YOY relative to river regions with high zebra mussel densities, as measured by the per cent of the total population occurring downriver from RKM 100.

3.3.2 Response metrics for spottail shiner Because the LRS does not adequately sample areas of the Hudson River inhabited by spottail shiner, for this species, no estimates of egg and larval abundance are available. However, the BSS provides estimates of both YOY abundance and adult abundance (age 1 and 2 adults) for this species. For the purpose of trends analysis, adult abundance is used as a surrogate for egg production.

3.3.3 Response metrics for Atlantic tomcod Because a substantial fraction of Atlantic tomcod larvae and YOY occur downriver from the regions sampled by the generators' surveys, for Atlantic tomcod, the data provided by the Atlantic tomcod mark-recapture program should be more reliable than the LRS, BSS, or FSS data for estimating survival rates. The mark-recapture program provides annual estimates of age-1 abundance, spawning stock size, and total egg production that can be used to calculate the fraction of eggs produced during a given year that survive to become age-1 spawners the following year. The LRS data can be used to characterize both year-to-year variations in early life stage abundance and the distribution of Atlantic tomcod larvae and juveniles within the Hudson River.

28

For this species, the response variables include:

1. Abundance of PYSL and early juveniles, as estimated from the LRS;

2. Abundance of Age-i and Age-2 fish, as estimated from the mark-recapture program;

3. Total age 0 survival, as measured by the ratio of total egg production each year to age 1 abundance during the following year;

4. Juvenile growth, as measured from growth rates of juveniles from the FSS; and

5. Spatial distribution of PYSL and early juveniles, as measured by the

,fraction of the total PYSL/juvenile population found in river regions 1-5 (LRS dataset).

3.4 Tests of impact hypotheses The predicted impacts of the stressors on the response metrics are summarized below and in Tables 1 (striped bass, white perch, American shad, river herring, bay anchovy, and spottail shiner) and 2 (Atlantic tomcod):

CWIS: Entrainment at IP2 and IP3 would be expected to reduce survival from the PYSL to the YOY stage, and could also reduce the abundance of PYSL. Entrainment should have no effect on growth or spatial distribution.

Fishing: Fishing would be expected to reduce the abundance of eggs and early larvae because of reduced spawner abundance, but should not reduce the survival of any age 0 life stage.

Zebra mussels: Zebra mussel activity would be expected to decrease both PYSL survival and YOY growth, and also to shift the spatial distribution of juveniles toward the lower regions and away from the freshwater regions where zebra mussels are abundant.

29

Temperature: Since Atlantic tomcod are known to be sensitive to high summer water temperatures, increased summer temperatures would be expected to decrease the growth and survival rates of life stages of this species that are present in the Hudson during this season.

Striped bass predation: Predation by older striped bass would be expected to decrease juvenile abundance, if the juveniles are susceptible to predation, and early life stage abundance, if adults are susceptible to predation.

Appendix B documents the stressor and response metrics and statistical methods used in this analysis. The subsections below present the results of the analyses performed for each species, and evaluate the consistency of these results with the impact hypotheses.

3.4.1 Striped bass Figure 7a depicts long-term trends in the abundance of striped bass PYSL and YOY in the Hudson. Figure 7b depicts long-term trends in striped bass PYSL to YOY survival. The abundance of juvenile striped bass in the Hudson has shown no trend, even though the abundance of striped bass early life stages has greatly increased. The increase in abundance of striped bass larvae has occurred, concurrently with an increase in the abundance of the Hudson River spawning stock of striped bass (Barnthouse et al. 2003). The increase in spawning size has been attributed to coastwide restrictions on harvesting that were imposed to promote the recovery of the Chesapeake Bay striped bass stock (Young-Dubovsky et al. 1995). As first noted by Pace et al. (1993), and later confirmed by Barnthouse et al. (2003), there is nocorrelation between the abundance of striped bass PYSL and striped bass YOY (Figure 8a). There is a strong negative relationship between PYSL abundance and PYSL survival, however (Figure 8b).

This negative correlation has been interpreted by both Pace et al. (1993) and Barnthouse et al.

(2003) as evidence for density-dependent mortality of striped bass larvae. This density-dependent mortality is reflected in the long-term trend in PYSL to YOY survival (Figure 7b),

which has declined through time as the size of the spawning population has increased.

30

3.4.1.1 CWIS, Co-occurrence Appendix B (Tables B- 1I and B-12) summarizes the results of the correlation analysis for striped bass. If entrainment at IP2 and IP3 were reducing the survival or abundance of early life stages of striped bass, then there should be a negative correlation between the CMR and striped bass PYSL survival, PYSL abundance, or both. However, as shown in Figure 9, there is no correlation between the IP2 and IP3 CMR and either PYSL survival (Figure 9a) or PYSL abundance (Figure 9b) for striped bass. Hence, the CWIS hypothesis fails the co-occurrence criterion for striped bass.

Sufficiency There are no independent measures of sufficiency that can be applied to this hypothesis.

The objective of this report is to determine, using all available and relevant evidence whether the magnitude of entrainment and impingement at Indian Point have been sufficient to cause a reduction in the abundance of important Hudson River fish species. Hence, the sufficiency criterion is inapplicable to the CWIS hypothesis.

Temporality If entrainment at IP2 and IP3 were reducing the survival or abundance of early life stages of striped bass, then a decline in PYSL survival, or PYSL abundance should have occurred after the startup of commercial operations of IP2 (1974) and IP3 (1976). However, as shown in Figure 7, no such declines occurred. PYSL abundance was relatively stable until 1985, and then rapidly increased. Striped bass PYSL survival has declined over time (Figure 7b), but the decline did not begin until several years after the startup of iP2 and IP3. Hence, the CWIS hypothesis fails the temporality criterion for striped bass.

Manipulation No experimental manipulations of plant operations have been performed for the purpose of evaluating entrainment impacts on fish populations. However, outages, including refueling and maintenance outages mandated by the HRSA (Englert et al. 1988), have frequently occurred 31

during the months when entrainable striped bass are present in the River. The peak abundance of striped bass eggs and larvae typically occurs during May and June (Boreman and Klauda, 1988).

IP2 was offline during the entire months of May and June in 1976, 1989, 1991, 1997, 1998, and 2000. IP3 was offline during the entire months of May and June in 1975, 1982, 1993, and 1994.

If entrainment at Indian Point were reducing the survival of striped bass PYSL, then PYSL survival should have been higher in years when one unit was offline than in years when both units were operating. As shown in Figure 10a, the measured PYSL survival values are inconsistent with this expectation. Figure 10a shows the time series of annual PYSL survival indices from 1975 through 2002. The horizontal line in Figure 10a shows the median survival index value for this time period. The median is defined as the midpoint of the entire distribution of survival index values, meaning that one-half of the survival indices are above the median and one-half are below the median. If striped bass PYSL survival were higher in years of one-unit operation than in years of 2-unit operation, then significantly more survival index values for years of one-year operation should be higher than the median than lower than the median.

However, Figure 10a shows that the PYSL survival index was higher than the median for only 3 of the 11 years of one-unit operation. The PYSL index was lower than the median in 8'years of one-unit operation.

This result is confirmed by Figure 10b, which shows the relationship between the striped bass PYSL survival index and the May-June total water withdrawals by IP2 and IP3 for the years 1975-2002. There is no correlation between withdrawals by IP2 and IP3 and striped bass PYSL survival. Hence, the CWIS hypothesis fails the manipulation criterion for striped bass.

Coherence As noted above, the objective of this report is to determine, using all available and relevant evidence whether the magnitude of entrainment and impingement at Indian Point have been sufficient to cause a reduction in the abundance of important Hudson River fish species.

Including "coherence" as an explicit evaluation criterion for CWIS would be redundant. Hence, the coherence criterion is inapplicable to the CWIS hypothesis.

32

3.4.1.2 Fishing Co-occurrence Fishing indirectly affects the abundance of early life stages of fish by reducing the abundance of spawning adults (Goodyear 1993). If a population is being overfished, then reducing the rate of fishing should cause the spawning population, and therefore the number of eggs spawned, to increase. As discussed by Young-Dubovsky et al. (1994), a coastwide ban on harvesting of striped bass was imposed in 1986. Estimates of fishing mortality and adult population abundance developed by the ASMFC (2005) show that the coastwide adult population has increased greatly since 1986. As shown in Figure 7a, the abundance of striped bass PYSL began increasing in 1988 and increased steadily throughout the 1990s. This is the same period during which the adult striped bass population was expanding. Hence, the overfishing hypothesis satisfies the co-occurrence criterion.

Sufficiency Fishing mortality estimates for individual striped bass spawning stocks are not estimated by the ASMFC, because much of the fishing occurs along the Atlantic coast when fish from the individual spawning stocks are mixed (ASMFC 2003). Since the magnitude of fishing mortality imposed specifically on Hudson River striped bass has never been estimated, it is not possible to determine whether the fishing hypothesis satisfies the sufficiency criterion.

Temporality The ban on striped bass harvesting preceded the increase in abundance of striped bass PYSL in the Hudson River by approximately 2 years. Hence, the fishing hypothesis satisfies the temporality criterion.

Manipulation The 1986 ban on striped bass harvesting was described by Young-Dubovsky et al. (1996) as an "adaptive management experiment." In other words, fishing was deliberately reduced in order to observe the response of the striped bass population to reduced harvesting. The fact that the adult population of striped bass began to increase immediately following the ban was 33

interpreted by Young-Dubovsky et al. (1994) as strong evidence that overfishing was, if not the only cause, at least the primary cause of the depressed abundance of Atlantic striped bass prior to the ban. Because the response of the population to this management was consistent with the expectations from the fishing hypothesis, the fishing hypothesis satisfies the manipulation criterion.

Coherence Atlantic striped bass are managed as a single coastwide fishery because a large fraction of the harvest occurs when fish originating in Chesapeake Bay, the Delaware River, and the Hudson River are mixed and migrating along the Atlantic coast (ASMFC 2003, Waldman et al. 1990, Waldman and Fabrizio 1994). If reduced harvesting had been the cause of increases in the abundance of early life stages of striped bass in the Hudson River, then similar increases should have occurred in the Chesapeake Bay and the Delaware River as well. As shown in the ASMFC's 2003 stock assessment, the abundance of juvenile striped bass in both Chesapeake Bay and the Delaware River grew rapidly after the harvest ban. Hence, the overfishing hypothesis is consistent with the coherence criterion.

3.4.1.3 Zebra mussels Co-occurrence As documented in Appendix B (Table B-11), the zebra mussel index is negatively correlated with the striped bass PYSL survival index. This correlation is consistent with the zebra mussel hypothesis. Hence, the zebra mussel hypothesis satisfies the co-occurrence criterion.

Sufficiency The potential effects of zebra mussel activity on early life stages of fish are indirect, and related to reductions in prey abundance and changes in habitat quality. No experiments have been performed that could quantify the relationship between zebra mussel activity and fish growth or survival, and no mathematical models that could be used to quantify the indirect 34

effects of zebra mussel activity have been developed. Hence, whether or not the zebra mussel hypothesis satisfies the sufficiency criterion is unknown.

Temporality Zebra mussels first became abundant in the Hudson River in 1992 (Caraco et al. 1997).

However, as shown in Figure 7b, striped bass PYSL survival began declining in the 1980s and had already fallen to a very low level by 1990. Because the decline in striped bass PYSL survival preceded, rather than followed, the appearance of zebra mussels in the River, the zebra mussel hypothesis fails the temporality criterion.

Manipulation No deliberate manipulations of zebra mussel populations in the Hudson River have been performed, therefore, this criterion is inapplicable to the zebra mussel hypothesis.

Coherence Because the proposed mechanism through which zebra mussel activity could have affected striped bass in the Hudson River involves reducing food availability, the growth as well as the survival of striped bass PYSL and YOY should have been reduced. Although Strayer et al. (2004) found a negative relationship between the growth rate of YOY striped bass and the presence of zebra mussels, no significant correlation was found in the analyses performed to support this report (Appendix B, Table B-11). Zebra mussel activity should also have shifted the distribution of striped bass PYSL and YOY downriver, away from the freshwater zone in which zebra mussels are abundant. Strayer et al. (2004) found no downstream shift in the distribution of striped bass PYSL and YOY. In the analyses performed to support this report (Appendix B, Table B- 11), no downstream shift in the distribution of PYSL was found, and an upstream shift (i.e., a shift in the opposite direction from the shift predicted by the zebra mussel hypothesis) in the distribution of YOY was found. The negative effect of zebra mussel activity on striped bass YOY growth that was reported by Strayer et al. (2004) conflicts with the findings in Appendix B, moreover, neither Strayer et al. (2004) nor the present analysis (Appendix B) found the predicted relationship between zebra mussel activity and striped bass PYSL and juvenile distribution.

Hence, the zebra mussel hypothesis fails the coherence criterion for striped bass.

35

3.4.1.4 Summary evaluation of hypotheses Table 3 summarizes the consistency of the striped bass trends data with the CWIS, overfishing, and zebra mussel hypotheses. Two of the five evaluation criteria - sufficiency and coherence - are inapplicable to the CWIS hypothesis. However, this hypothesis fails all three of the remaining criteria. Hence, the CWIS hypothesis can be rejected as an explanation for long-term trends in the abundance of age 0 striped bass in the Hudson River. The zebra mussel hypothesis passes the co-occurrence criterion, but fails the temporality and coherence criteria.

Because striped bass PYSL survival declined several years prior to the invasion of the Hudson River by zebra mussels, and because predicted effects of zebra mussels on the growth and distribution of striped bass PYSL and YOY were not observed, the zebra mussel hypothesis also can be rejected as an explanation for long-term trends in the abundance of age 0 striped bass in the Hudson River.

The overfishing hypothesis, in contrast, passes four of the five evaluation criteria. The remaining criterion (sufficiency) is inapplicable to this hypothesis. The abundance of striped bass PYSL in the Hudson began increasing shortly following a reduction in striped bass harvesting. The reduction in harvest was specifically intended to promote striped bass reproduction, and was followed by simultaneous increases in striped bass reproductive success in all three of the major east coast spawning populations. It is reasonable to conclude, therefore, that elimination of overfishing is the most likely cause of trends in the abundance of early life stages of striped bass in the Hudson River.

3.4.2 White perch Figure 11 depicts long-term trends in the abundance of white perch YOY and PYSL in the Hudson. As shown in Figure 11, the abundance of juvenile white perch declined steadily throughout the 1980s, but has increased since 1990. Despite the recent increase, over the entire time series, there is a statistically significant decline in YOY abundance (Appendix B, Table B-13 and Figure B-4). There is no long-term trend in the annual abundance of PYSL (Figure 11),

however, which suggests that larval production is stable. There is no relationship between PYSL abundance and YOY abundance in white perch (Figure 12a). The survival rate of white perch 36

from the PYSL to the juvenile stage has declined (Appendix B, Table B-13). Moreover, there is a strong positive relationship between PYSL survival and YOY abundance (Figure 12b, Appendix B, Table B-14). Because YOY abundance in white perch is closely related to PYSL survival but not to PYSL abundance, we can conclude that the decline in YOY abundance was due to a decline in PYSL survival rather than to a decline in white perch reproduction.

3.4.2.1 CWIS Co-Occurrence Appendix B, Table B-13 and B-14*summarize the results of the correlation analysis for white perch. If entrainment at Indian Point had caused the observed decline in white perch PYSL survival, there should be a negative relationship between the entrainment CMR for white perch and white perch PYSL survival. This means that in years when the CMR was high, white perch PYSL survival should have been low, and in years when the CMR was low, white perch PYSL survival should have been high. However, as shown in Figure 13a, the opposite relationship exists. The IP2 and IP3 CMR is positively correlated with PYSL to juvenile survival, meaning that the CMR was high in years when PYSL survival was high and the CMR was low in years when PYSL survival was low.

There is a negative, relationship between the IP2 and IP3 CMR and white perch PYSL abundance (Figure 13b), but this correlation is significant only at the 10% level. Figure 14 plots time trends in both the CMR and in PYSL to juvenile survival for white perch. The two trend lines show similar patterns, with values decreasing from the mid-1970s to the mid-1980s, fluctuating until the mid-1990s, and then increasing. It is important to note that the recent increase in survival occurred during a period in which the capacity factors for IP2 and IP3 have been higher than in earlier years (Darla Gray, Entergy Corp., personal communication).

Although there is a weak negative relationship between the CMR for IP2 and IP3 and white perch PYSL abundance, the much stronger positive relationship between the CMR and PYSL to YOY survival must be accorded a higher weight. Because this positive correlation clearly conflicts with the CWIS hypothesis, the CWIS hypothesis fails the co-occurrence criterion for white perch.

Sufficiency 37

There are no independent measures of sufficiency that can be applied to this hypothesis.

The objective of this report is to. determine, using all available and relevant evidence whether the magnitude of entrainment and impingement at Indian Point have been sufficient to cause a reduction in the abundance of important Hudson River fish species. Hence, the sufficiency criterion is inapplicable to the CWIS hypothesis.

Temporality As shown in Figure 14, white perch PYSL survival began to decline in 1977, one year following the startup of commercial operation at IP3. Since the startup of 2-unit operation preceded the decline in white, perch PYSL survival, the CWIS hypothesis satisfies the temporality criterion.

Manipulation As discussed in Section 3.4.1.1, outages of IP2 or IP3 have frequently occurred during the entrainment season at Indian Point. The peak abundance of white perch eggs and larvae typically occurs during May and June (Klauda 1988). IP2 was offline during the entire months of May and June in 1976, 1989, 1991, 1997, 1998, and 2000. IP3 was offline during the entire months of May and June in 1975, 1982, 1993, and 1994. If entrainment at Indian Point were reducing the survival of white perch PYSL, then PYSL survival should have been higher in years when one unit was offline than in years when both units were operating. As shown in Figure 15a, the measured PYSL survival values are inconsistent with this expectation. Figure 15a shows the time series of annual PYSL survival indices from 1975 through 2002, which are the years for which cooling water flow data were available. The horizontal line in Figure 15 shows the median survival index value for this time period. The median is defined as the midpoint of the entire distribution of survival index values, meaning that one-half of the survival indices are above the median and one-half are below the median. If white perch PYSL survival were higher in years of one-unit operation than in years of 2-unit operation, then significantly more survival index values for years of one-year operation should /be' higher than the median than lower than the median. However, Figure 15a shows that the PYSL survival index was higher than the median for only 4 of the 11 years of one-unit operation. The PYSL index was equal to the 38

median in one year (1989) of one-unit operation, and lower than the median in 6 years of one-unit operation.

This result is confirmed by Figure 15b, which shows the relationship between the white perch PYSL survival index and the May-June total water withdrawals by IP2 and IP3 for the years 1975-2002. There is no correlation between withdrawals by IP2 and IP3 and white perch PYSL survival. Hence, the CWIS hypothesis fails the manipulation criterion for white perch.

Coherence As noted above, the objective of this report is to determine, using all available and relevant evidence whether the magnitude of entrainment and impingement at Indian Point have been sufficient to cause a reduction in the abundance of important Hudson River fish species.

Including "coherence" as an explicit evaluation criterion for CWIS would be redundant. Hence, the coherence criterion is inapplicable to the CWIS hypothesis.

3.4.2.2 Zebra mussels Co-Occurrence As shown in Appendix B, Table B-13, the zebra mussel index is negatively correlated with PYSL to YOY survival in white perch. Hence, the zebra mussel hypothesis satisfies the co-occurrence criterion.

Temporality As shown in Figure 14, however, the decline in white perch PYSL to YOY survival occurred primarily between 1974 and 1986, prior to the zebra mussel invasion. PYSL to YOY survival has actually been increasing since 1993, the first year in which zebra mussels were abundant enough to potentially affect fish populations (Strayer et al. 2004). Hence, the zebra mussel hypothesis fails the temporality criterion.

Sufficiency The potential effects of zebra mussel activity on early life stages of fish are indirect, and related to reductions in prey abundance and changes in habitat quality. No experiments have 39

been performed that could quantify the relationship between zebra mussel activity and fish growth or survival, and no mathematical models that could be used to quantify the indirect effects of zebra mussel activity have been developed. Hence, whether or not the zebra mussel hypothesis satisfies the sufficiency criterion is unknown.

Manipulation No deliberate manipulations of zebra mussel populations in the Hudson River have been performed, therefore, this criterion is inapplicable to the zebra mussel hypothesis.

Coherence Because the proposed mechanism through which zebra mussel activity could have affected white perch in the Hudson River involves reducing food availability, the growth as well as the survival of white perch PYSL should have been reduced. Although Strayer et al. (2004) reported a negative relationship between zebra mussel activity and white perch growth, the analysis performed to support this assessment (Appendix B, Table B-13) found no significant relationship between zebra mussels and white perch growth. Moreover, the percent of white perch juveniles downriver from RKM 100 is negatively, instead of positively, correlated with the zebra mussel index (Appendix B, Table B-13). This negative correlation implies that over this same period of years, the percentage of the population present downriver from RKM 100 has declined, rather than increasing as predicted by the zebra mussel hypothesis. This result is also consistent with the findings of Strayer et al. (2004). Hence, the zebra mussel hypothesis partially, but not fully, satisfies the coherence criterion.

3.4.2.3 Striped.basspredation Co-occurrence There is a weak negative correlation between the striped bass index and the white perch PYSL index (Appendix B, Table B-13). This relationship provides weak evidence supporting the hypothesis that striped bass are preying on adult white perch. There is much stronger negative correlation between the striped bass index and the YOY index (Figure 16a). This correlation is consistent with the hypothesis that striped bass are preying on juvenile white perch.

There is also a strong negative correlation between the striped bass index and white perch PYSL 40

to YOY survival, however, this relationship is difficult to interpret because striped bass would not be expected to prey on larval white perch. Overall, the striped bass hypothesis satisfies the co-occurrence criterion with respect to predation on YOY white perch.

Sufficiency Striped bass larger than 200 mm in length have been shown to feed on white perch (Gardinier and Hoff 1982, Dunning et al. 1997). Appendix C to this report documents an analysis of prey consumption by Hudson River striped bass. This analysis compares the change in striped bass prey consumption requirements (August through October) between earlier (1983-1990) and more recent (1991-2004) periods to changes in abundance of YOY fish in the Hudson River between these same two periods. The analysis shows that the increase in prey consumption from the earlier to the later period would be sufficient to explain the decline in YOY white perch abundance between these two periods if 1% of the age 1 and age 2 striped bass seasonal predatory demand was satisfied by YOY white perch, or if 0.3% of the age 1.through age 13 striped bass seasonal predatory demand was satisfied by YOY white perch. Hence, the striped bass predation hypothesis satisfies the sufficiency criterion for white perch.

Temporality A sustained decline in white perch YOY abundance began in 1989, at the same time the striped bass index began to increase (Figure 16b). However, the historic peak in YOY abundance occurred in 1980 (Figure 16b), and PYSL to YOY survival declined substantially between 1975 and 1985 (Figure 14). White perch PYSL to YOY survival and YOY abundance are strongly correlated (Figure 12b), implying that declining YOY abundance must have been at least in part caused by a decline in PYSL to YOY survival. The decline in PYSL to YOY survival that declined between 1975 and 1985 cannot be explained by striped bass predation.

Hence, the striped bass predation hypothesis only partially satisfies the temporality criterion.

Manipulation No deliberate manipulations of striped bass predation in the Hudson River have been performed, therefore, this criterion is inapplicable to the striped bass hypothesis.

41

Coherence If predation by striped bass had caused the decline in abundance of YOY white perch in the Hudson River, then the YOY abundance of other known striped bass prey species, including river herring, American shad, bay anchovy, and Atlantic tomcod should also have declined. As shown in other Sections of this report, YOY abundance for all of these species has declined since the late 1980s, when striped bass abundance began to increase. Moreover, other published studies have concluded that striped bass predation is reducing the abundance of some prey species. Savoy and Crecco (2004) attributed recent declines in the abundance of both blueback herring and American shad in the Connecticut River to striped bass predation. Hartman (2003) estimated that the coastwide annual prey consumption by striped bass between 1 and 10 years of age increased by more than a factor of 8 between 1982 and 1995, from 17,900 metric tons (mt) to 147,900 mt. Uphoff (2003) calculated even larger estimates of striped bass consumption, and attributed a 90% decline in the abundance of Atlantic menhaden in upper Chesapeake Bay from 1980 through 1999 to predation by striped bass.

Because parallel declines in other susceptible species have occurred, and because the other published studies have documented the influence of striped bass predation on susceptible prey species, the striped bass predation hypothesis satisfies the coherence criterion.

3.4.2.4 Summary evaluation of hypotheses Table 4 summarizes the consistency of the white perch trends data with the CWIS, zebra mussel, and striped bass predation hypotheses. Two of the five evaluation criteria - sufficiency and coherence - are inapplicable to the CWIS hypothesis. The CWIS hypothesis fails the co-occurrence and manipulation criteria. Although the CWIS hypothesis satisfies the temporality criterion because the observed decline in white perch PYSL survival followed the startup of IP2 and IP3, the inconsistency of this hypothesis with the co-occurrence and manipulation hypotheses means that the temporal correspondence between the beginning of the decline in survival and the startup of IP2 and IP3 is very likely a coincidence. Hence, the CWIS hypothesis can be rejected as an explanation for long-term trends in the abundance of age 0 white perch in the Hudson River.

42

The zebra mussel hypothesis passes the co-occurrence criterion and at least partially satisfies the coherence criterion. However, it fails the temporality criterion because the declines in white perch PYSL survival and YOY abundance began prior to the appearance of zebra mussels in the Hudson River. Although zebra mussel activity might have contributed to a decline in white perch PYSL to YOY survival and YOY abundance from 1993 to 2004, zebra mussels could not have been the primary explanation for long-term trends in white perch survival and abundance.

The striped bass predation hypothesis satisfies four of the five criteria. The fifth, manipulation, is inapplicable to this hypothesis. However, the strong relationship between white perch PYSL survival and YOY abundance over the entire period from 1974 to 2004 (Figure 12b) cannot be explained by the predation hypothesis, because striped bass abundance did not begin to increase until 1987. Hence, although striped bass predation likely contributed to the decline in white perch PYSL to YOY survival and YOY abundance, from 1987 onward, predation could not have been the primary cause of declines that took place between 1975 and 1985.

3.4.3 American shad Figure 17 depicts long-term trends in the abundance of American shad YOY and PYSL in the Hudson. The abundance of both life stages has declined significantly since the initiation of the generators' monitoring program, with declines in the abundance of both life stages beginning in the late 1980s. As shown in Figure 18, there is a strong positive correlation between PYSL abundance and YOY abundance in American shad (Figure 18a), and no relationship between PYSL survival and YOY abundance (Figure 18b). Because YOY abundance is correlated with PYSL abundance but not with PYSL survival, we can conclude that the decline in YOY abundance is a consequence of reduced reproduction rather than reduced PYSL survival.

Four hypothetical causes for these changes are evaluated below: 'the. Indian Point CWIS, overfishing, zebra mussels, and striped bass predation.

43

3.4.3.1 CWJS Co-Occurrence There is no correlation between PYSL survival and the entrainment CMR at IP2 and IP3 (Figure 19a). The IP2 and IP3 CMR is also uncorrelated with American shad PYSL abundance (Figure 19b). Hence, the CWIS hypothesis fails the co-occurrence criterion.

Sufficiency There are no independent measures of sufficiency that can be applied to this hypothesis.

The objective of this report is to determine, using all available and relevant evidence whether the magnitude of entrainment and impingement at Indian Point have been sufficient to cause a reduction in the abundance of important Hudson River fish species. Hence the sufficiency criterion is inapplicable to the CWIS hypothesis.

Temporality American shad PYSL abundance grew from the mid-1970s, when IP2 and IP3 began commercial operations, until 1986 (Figure 17). The highest values for both PYSL and YOY abundance occurred in 1986, 10 years after the startup of commercial operations at IP3 and 12 years after the startup of IP2 (Figure 17). Hence, the CWIS hypothesis fails the temporality criterion.

Manipulation As discussed in Section 3.4.1.1, outages of IP2 or IP3 have frequently occurred during the entrainment season at Indian Point. Although American shad eggs and larvae occur only at very low densities in the vicinity of Indian Point (DEIS, Figure V-68), the peak abundance of American shad eggs and larvae typically occurs during May and June (DEIS, Figure V-67). IP2 was offline during the entire months of May and June in 1976, 1989, 1991, 1997, 1998, and 2000. IP3 was offline during the entire months of May and June in 1975, 1982, 1993, and 1994.

If entrainment at Indian Point were reducing the survival of American shad PYSL, then PYSL survival should have been higher in years when one unit was offline than in years when both units were operating. As shown in Figure 20a, the measured PYSL survival values are 44

inconsistent with this expectation. Figure 20a shows the time series of annual PYSL survival indices from 1985 through 2002. The horizontal line in Figure 20a shows the median survival index value for this time period. The median is defined as the midpoint of the entire distribution of survival index values, meaning that one-half of the survival indices are above the median and one-half are below the median. If American shad PYSL survival were higher in years of one-unit operation than in years of 2-unit operation, then significantly more survival index values for years of one-year operation should be higher than the median than lower than the median.

However, Figure 20a shows that the PYSL survival index was higher than the median for 5 of the 8 years of one-unit operation. The PYSL index was lower than the median in 3 years of one-unit operation. This difference could easily have arisen by chance. Moreover, 3 of the 5 years with the highest survival rates (1996, 1999, and 2002) were years of 2-unit operation.

This result is confirmed by Figure 20b, which shows the relationship between the American shad PYSL survival index and the May-June total water withdrawals by IP2 and IP3 for the years 1975-2002. There is no correlation between withdrawals by IP2 and IP3 and American shad PYSL survival. Hence, the CWIS hypothesis fails the manipulation criterion for American shad.

Coherence The objective of this report is to determine, using all available and relevant evidence whether the magnitude of entrainment and impingement at Indian Point have been sufficient to cause a reduction in the abundance of important Hudson River fish species. Including "coherence" as an explicit evaluation criterion for CWIS would be redundant. IHence, the coherence criterion is inapplicable to the CWIS hypothesis.

3.4.3.2 Fishing Co-Occurrence If a population is being overfished to the point at which spawner abundance is reduced, then the number of eggs and larvae produced by those spawners should decline. Historically, American shad supported very large unregulated commercial fisheries along the east coast of both the United States and Canada (ASMFC 1999). These harvests have declined dramatically 45

in recent years. In its most recent stock assessment for American shad (ASMFC 2007), the ASMFC found that the abundance of adult American shad in the Hudson River peaked in 1985 and 1986 and has since declined. This decline in adult abundance occurred during the same period in which the abundance of American shad PYSL and YOY in the Hudson River declined (Figure 17). Hence, the fishing hypothesis satisfies the co-occurrence criterion.

Sufficiency There is conflicting information concerning whether the magnitude of fishing mortality imposed on Hudson River American shad has been sufficient to cause the declines in spawner abundance. According to the ASMFC (2007), many American shad stocks have declined in abundance in recent decades. Although the declines appear to be related to an increase in the mortality of adult shad, the contribution of fishing to the increase in mortality is unclear and probably differs between spawning populations. According to Hattala and Kahnle (2007), the Hudson River population of American shad is probably being overfished, however, other sources of mortality cannot be excluded as contributing causes. Although there is still substantial uncertainty concerning causes of decline in American shad population, this assessment accepts Hattala and Kahnle's (2007) results and concludes that the overfishing hypothesis satisfies the sufficiency criterion.

Temporality The decline in American shad spawner abundance coincided with the decline in abundance of PYSL and YOY (Figure 17). Hence, the overfishing hypothesis satisfies the temporality criterion.

Manipulation Amendment 1 to the Interstate Fisheries Management Plan for Shad and River Herring, (ASMFC 1999) directed all states to phase outthe coastal fishery for American shad over a five year period beginning in 2000. The phase-out should reduce fishing mortality on American shad. If the coastal fishery had been contributing to decreased abundance of Connecticut River American shad, then the abundance of this population should increase as a result of this action.

Data on fishing mortality and population abundance from the post-closure period are not yet 46

available, so it is not yet possible to evaluate whether the overfishing hypothesis satisfies the manipulation criterion.

Coherence As noted above, there is still substantial uncertainty concerning the impact of fishing on the Hudson River American shad population. However, available data are consistent with a conclusion that fishing is at least a significant contributor to the recent decline in abundance of Hudson River American shad (Hattala and Kahnle 2007). Hence, the overfishing hypothesis satisfies the coherence criterion.

3.4.3.3 Zebra mussels Co-occurrence As shown in Appendix B, Table B-15, the American shad PYSL survival index is positively correlated with the zebra mussel index, rather than negatively correlated as predicted by the zebra mussel hypothesis. As can easily be seen from;Figure 17, American shad PYSL to YOY survival has increased since the zebra mussel invasion. Hence, the zebra mussel hypothesis fails the co-occurrence criterion for American shad.

Sufficiency The potential effects of zebra mussel activity on early life -stages of fish are indirect, and related to reductions in prey abundance and changes in habitat quality. No experiments have been performed that could quantify the relationship between zebra mussel activity and fish growth or survival, and no mathematical models that could be used to quantify the indirect effects of zebra mussel activity have been developed. Hence, whether or not the zebra mussel hypothesis satisfies the sufficiency criterion is unknown.

Temporality The decline in abundance of American shad PYSL and YOY began in the late 1980s (Figure 17), several years prior to the zebra mussel invasion. Hence, the zebra mussel hypothesis fails the temporality criterion.

47

Manipulation No deliberate manipulations of zebra mussel populations in the Hudson River have been performed, therefore, this criterion is inapplicable to the zebra mussel hypothesis.

Coherence Because the proposed mechanism through which zebra mussel activity could have affected American shad in the Hudson River involves reducing food availability, the growth as well as the survival of American shad PYSL and YOY should have been reduced. Although Strayer et al. (2004) found a decline in growth rate of American shad PYSL and YOY following the zebra mussel invasion, this relationship was not significant even at the 20% level (Strayer et al. 2004, Fig. 7). No relationship between American shad YOY growth and zebra mussel activity was found in the analysis performed to support this assessment (Appendix B, Table B-15). Zebra mussel activity should also have shifted the distribution of American shad PYSL and YOY downriver, away from the freshwater zone in which zebra mussels are abundant. Strayer et al. (2004) found a net downriver shift in the distribution of American shad YOY, but a net upriver shift in the distribution of PYSL. In the analysis performed to support this assessment (Appendix B, Table B- 15), no significant shifts in the distribution of either life stage was found.

The observed changes in growth and distribution predicted by the zebra mussel hypothesis were not observed. Hence, the zebra mussel hypothesis fails the coherence criterion for American shad.

3.4.3.4 Striped basspredation Co-occurrence American shad PYSL abundance, which reflects spawner abundance and reproduction, is negatively correlated with the striped bass. index (Figure 21a), although this relationship is significant only at the 10% level. This correlation provides weak support for the hypothesis that striped bass are preying on adult American shad. There is a negative relationship between the striped bass index and the American shad YOY index, (Figure 21 b), however, this relationship is not statistically significant. Hence, the striped bass predation hypothesis appears to marginally satisfy the co-occurrence criterion for predation.

48

Sufficiency Striped bass larger than 200 mm in length have been shown to feed on alosids such as American shad (Gardinier and Hoff 1982, Dunning et al. 1997). However, the prey consumption analysis documented in Appendix C to this report did not address predation on YOY American shad. Hence, with respect to YOY American shad, whether or not striped bass predation satisfies the sufficiency criterion is unknown. Kahnle and Hattala (2007) have argued that the great majority of adult striped bass in the Hudson are feeding on river herring rather than shad, and the striped bass predation is insufficient to significantly affect the abundance of adult Hudson River American shad. This assessment accepts the conclusions of Kahnle and Hattala (2007) that striped bass predation on adult Hudson River American shad is probably low.

Temporality As can be seen from Figure 22, the increase in striped bass spawner abundance that began in the late 1980s closely coincides with the decline in American shad PYSL abundance. As shown in Figure 17, American shad YOY abundance has. declined over this same period. Hence, the striped bass predation hypothesis satisfies the temporality criterion with respect to predation on both adults and YOY.

Manipulation No deliberate manipulations of striped bass predation in the Hudson River have been performed, therefore, this criterion is inapplicable to the striped bass hypothesis.

Coherence If predation by striped bass had caused the decline in abundance of American shad PYSL and YOY in the Hudson River, then the PYSL and YOY abundance of other known striped bass prey species, including white perch, river herring, bay anchovy, and Atlantic tomcod should also have declined. As discussed in other Sections of this report, no declines in white perch or bay anchovy PYSL abundance have occurred. However, PYSL abundance for river herring and Atlantic tomcod declined over the same period in which PYSL abundance for American shad declined. YOY abundance for all of the above species has declined since the late 1980s, when 49

striped bass abundance began to increase. Moreover, other published studies have concluded that striped bass predation is reducing the abundance of some prey species. Savoy and Crecco (2004) attributed recent declines in the abundance of both blueback herring and American shad in the Connecticut' River to striped bass predation on spawning adults, however, Kahnle and Hattala (2007) concluded that predation of striped bass on adult American shad in the Hudson River is relatively low. On the other hand, Hattala and Kahnle (2007) acknowledged that predation by striped bass on young American shad could be substantial and could be contributing to a decline in recruitment of young shad to the adult population.

Hartman (2003) estimated that the coastwide annual prey consumption by striped bass between 1 and 10 years of age increased by more than a factor of 8 between 1982 and 1995, from 17,900 mt to 147,900 mt. Uphoff (2003) calculated even larger estimates of striped bass consumption, and attributed a 90% decline in the abundance of Atlantic menhaden in upper Chesapeake Bay from 1980 through 1999 to predation by striped bass.

Because parallel declines in YOY abundance of other susceptible species have occurred, and because the other published studies have documented the influence of striped bass predation on susceptible prey species, the striped bass predation hypothesis satisfies the coherence criterion with respect to predation on YOY American shad, but not with respect to predation on adults.

3.4.3.5 Summary evaluation of hypotheses Table 5 summarizes the consistency of the American shad data with the CWIS, overfishing, zebra mussel, and striped bass predation hypotheses. Two of the five evaluation criteria - sufficiency and coherence - are inapplicable to the CWIS hypothesis. The CWIS hypothesis fails the co-occurrence, temporality, and manipulation criteria. Hence, the CWIS hypothesis can be rejected as an explanation for long-term trends in the abundance of age 0 American shad in the Hudson River.

The overfishing hypothesis satisfies the co-occurrence, sufficiency, temporality, and coherence criteria for American shad. The manipulation criterion is inapplicable at present, although applicable data may become available once the response of the population to the phase-out of the ocean intercept fishery has been observed.

50

The zebra mussel hypothesis fails the co-occurrence, temporality, and coherence criteria for American shad. Whether the sufficiency criterion is satisfied is unknown, and the manipulation criterion is inapplicable. Hence, the zebra mussel hypothesis can be rejected as an explanation for long-term trends in the abundance of age 0 American shad in the Hudson River.

The striped bass predation hypothesis satisfies two and possibly three of the five criteria.

Because no estimates of potential striped bass predation on YOY American shad have been developed, whether this hypothesis satisfies the sufficiency criterion is unknown. The manipulation criterion, is inapplicable to this hypothesis. The simultaneous declines in abundance of susceptible life stages of other prey species in the Hudson River and the published studies documenting impacts of striped bass predation on prey species support for the predation hypothesis. However, substantial uncertainty remains concerning the fraction of the American shad YOY population that might be consumed.

It appears reasonable to conclude that the recent decline in abundance of Hudson River American shad is most likely a result of overfishing, but striped bass predation may be a contributing cause.

3.4.4. Atlantic tomcod Figure 23 depicts long-term trends in the abundance of Atlantic tomcod as measured by the LRS and the Atlantic Tomcod mark-recapture program. The LRS index reflects the abundance of late PYSL and early juvenile fish. The mark-recapture index reflects the combined abundance of age 1 and older (predominantly age 2) fish. The abundance of Atlantic tomcod has declined since the initiation of the generators' monitoring programs, with the abundance of age 1 and older fish abundance showing an abrupt decline beginning in 1990. The trend in abundance in the LRS time series- is less clear, but the LRS index also has declined since 1990. Using Atlantic tomcod survival rates derived from annual mark-recapture surveys, for each year, the total egg to age 1 survival rate is estimated by comparing the total egg production during that year to the number of age 1 fish estimated to be present in the Hudson River during the following year. As shown in Figure 24, there is no relationship between egg deposition and resulting age 1 abundance (Figure 24a). There is a positive relationship between egg to age 1 survival and age 1 51

abundance (Figure 24b). Hence, the decline in Atlantic tomcod -abundance is related to a decrease in survival rather than a decrease in egg production.

Atlantic tomcod are uncommon in freshwater reaches of the Hudson River, therefore, they should not be susceptible to the effects of zebra mussel activity. This potential stressor is not evaluated as a cause of changes in the abundance of this species. Three hypothetical causes for these changes are evaluated below: the Indian Point CWIS, elevated summer temperatures, and striped bass predation.

3.4.4.1 CWIS Co-occurrence As shown in Figure 25a, there is no correlation between the IP2 and IP3 CMR and egg-to-age 1 survival. There is a negative correlation between the IP2 and IP3 CMR and the Atlantic tomcod LRS index (Figure 25b), but this correlation is significant only at the 10% level (Appendix B, Table B-17). There is no correlationbetween the IP2 and IP3 CMR and the mark-recapture index (Figure 25c). Because the IP2 and IP3 CMR are negatively correlated with only one of the three response metrics, and only at the 10% level, the CWIS hypothesis only weakly satisfies the co-occurrence criterion.

Sufficiency There are no independent measures of sufficiency that can be applied to this hypothesis.

The objective of this report is to determine, using all available and relevant evidence whether the magnitude of entrainment and impingement at Indian Point have been sufficient to cause, a reduction in the abundance of important Hudson River fish species. Hence, the sufficiency criterion is inapplicable to the CWIS hypothesis.

Temporality As shown in Figure 23, the decline in abundance of Atlantic tomcod in the mark-recapture survey did not begin until the mid-1980s and the decline in the LRS survey did not begin until 1990. Hence, the CWIS hypothesis fails the temporality criterion.

52

Manipulation Although American tomcod spawn in December and January, entrainable larvae and juveniles are still abundant in the lower estuary during May and June (DEIS, Figure 5-56). IP2 was offline during the entire months of May and June in 1976, 1989, 1991, 1997, 1998, and 2000. IP3 was offline during the entire months of May and June in 1975, 1982, 1993, and 1994.

If entrainment at Indian Point were reducing the survival of Age 0 Atlantic tomcod, then egg to age 1 survival should have been higher in years when one unit was offline than in years when both units were operating. As shown in Figure 26a, the measured PYSL survival values are inconsistent with this expectation. Figure 26a shows the time series of egg to age 1 indices from 1976 through 2001. The horizontal line in Figure 26a shows the median survival index value for this time period. The median is defined as the midpoint of the entire distribution of survival index values, meaning that one-half of the survival indices are above the median and one-half are below the median. If Atlantic tomcod survival were higher in years of one-unit operation than in years of 2-unit operation, then significantly more survival index values for years of one-year operation should be higher than the median than lower than the median. However, Figure 26a shows that the PYSL survival index was higher than the median for 3 of the 7 years of one-unit operation. The PYSL index was lower than the median in 4 years of one-unit operation.

This result is confirmed by Figure 26b, which shows the relationship between the Atlantic tomcod egg to age 1 survival index and the May-June total water withdrawals by IP2 and IP3 for the years 1975-2002. There is no correlation between withdrawals by IP2 and IP3 and Atlantic tomcod egg to age 1 survival. Hence, the CWIS hypothesis fails the manipulation criterion for Atlantic tomcod.

Coherence The objective of this report is to determine, using all available and relevant evidence whether the magnitude of entrainment and impingement at Indian Point have been sufficient to cause a reduction in the abundance of important Hudson River fish species. Including "coherence" as an explicit evaluation criterion for CWIS would be redundant. Hence, the coherence criterion is inapplicable to the CWIS hypothesis.

53

3.4.4.2 Elevated summer temperatures Co-occurrence As shown in Appendix B, Table B-17, eggto age 1 survival is negatively correlated with the PWW degree-day index. Egg to age 1 survival is not, however, correlated with the August cooling water flows at IP2 and IP3, which is an index of the thermal loading to the River from IP2 and IP3. Hence, the temperature hypothesis satisfies the co-occurrence criterion, although.

there is no evidence that IP2 and IP3 contribute to a temperature effect.

Sufficiency As discussed by McLaren et al. (1988), summer temperatures in the Hudson River frequently exceed optimal levels for juvenile Atlantic tomcod, and occasionally can exceed the lethal tolerance temperature (26.5°C) for this species (McLaren et al. 1988). Although the temperature of the Hudson River is highly variable between locations, depth strata, and years, it can be concluded that the temperature hypothesis satisfies the sufficiency criterion.

Temporality Figure 27 compares long-term trends in PWW degree-day index to long-term trends in the abundance of age 1 and age 2 Atlantic tomcod, for the period 1987-2001. For each year, the degree-day index is paired with the mark-recapture estimates generated during the following winter (e.g., the 1987 temperature value is paired with the mark-recapture value for the winter of 1987-1988). As shown in Figure 27, a decline in Atlantic tomcod occurred from 1990-2001.

However, elevated temperatures that could have explained this decline did not occur. There is no long-term trend in the PWW degree-day index, and three of the four lowest values of the index have occurred since 1990. Hence, the temperature hypothesis fails the temporality criterion.

Manipulation No deliberate manipulations of Hudson River water temperatures have been performed, therefore, this criterion is inapplicable to temperature hypothesis.

54

Coherence If elevated temperatures were adversely affecting Atlantic tomcod in the Hudson River, then other temperature-sensitive species should also be declining. As noted in the FEIS (pp 66-67), the abundance of rainbow smelt in the Hudson River has also been declining. In addition, the temperature hypothesis is consistent with laboratory data on thermal tolerances in Atlantic tomcod and with the geographic distribution of this species. As noted by McLaren et al. (1988),

the Hudson River is the southern-most reproducing Atlantic tomcod population. Hence, the temperature hypothesis satisfies the coherence criterion.

3.4.4.3 Striped bass predation Co-occurrence Both the Atlantic tomcod mark-recapture index and the LRS index are negatively correlated with the striped bass index (Figure 28). Hence, the striped bass predation hypothesis satisfies the co-occurrence criterion.

Sufficiency Striped bass larger than 200 mm in length have been shown to feed on Atlantic tomcod (Gardinier and Hoff 1982, Dunning et al. 1997). Appendix C to this report documents an analysis of prey consumption by Hudson River striped bass. This analysis compares the change in striped bass prey consumption requirements (August through October) between earlier (1983-1990) and more recent (1991-2004) periods to changes in abundance of YOY fish in the Hudson River between these same two periods. The analysis shows that the increase in prey consumption from the earlier to the later period would be sufficient to explain the decline in YOY Atlantic tomcod abundance between these two periods if 1.4% of the age 1 and age 2 striped bass seasonal predatory demand was satisfied by YOY Atlantic tomcod, or if 0.4% of the age 1 through age 13 striped bass seasonal predatory demand was satisfied by YOY Atlantic tomcod. Hence, the striped bass predation hypothesis satisfies the sufficiency criterion.

55

Temporality The increase in striped bass abundance coincides in time with the declines in both Atlantic tomcod abundance metrics (Figure 29). Hence, the striped bass predation hypothesis satisfies the temporality criterion.

Manipulation No deliberate manipulations of striped bass predation in the Hudson River have been performed, therefore, this criterion is inapplicable to the striped bass hypothesis.

Coherence If predation by striped bass had caused the decline in abundance of Atlantic tomcod in the Hudson River, then the YOY abundance of other known striped bass prey species, including white perch, river herring, American shad, and bay anchovy, should also have declined. As shown in other Sections of this report, YOY abundance for all of these species has declined since the late 1980s, when striped bass abundance began to increase. Moreover, other published studies have concluded that striped bass predation is reducing the abundance of some prey species. Savoy and Crecco (2004) attributed recent declines in the abundance of both blueback herring and American shad in the Connecticut River to striped bass predation. Hartman (2003) estimated that the coastwide annual prey consumption by striped bass between 1 and 10 years of age increased by more than a factor of 8 between 1982 and 1995, from 17,900 mt to 147,900 mt.

Uphoff (2003) calculated even larger estimates of striped bass consumption, and attributed a 90% decline in the abundance of Atlantic menhaden in upper Chesapeake Bay from 1980 through 1999 to predation by striped bass.

Because parallel declines in other susceptible species have occurred, and because the other published studies have documented the influence of striped bass predation on susceptible prey species, the striped bass predation hypothesis satisfies the coherence criterion.

3.4.4.4 Summary evaluation of hypotheses Table 6 summarizes the consistency of the Atlantic tomcod data with the CWIS, temperature, and striped bass predation hypotheses. Two of the five evaluation criteria -

56

sufficiency and coherence - are inapplicable to the CWIS hypothesis. The CWIS hypothesis weakly satisfies the co-occurrence criterion, but fails the temporality, and manipulation criteria.

The CWIS hypothesis can be rejected as an explanation for long-term trends in the abundance of age 0 Atlantic tomcod in the Hudson River.

The temperature hypothesis satisfies the co-occurrence, sufficiency, and coherence criteria, but fails the temporality criterion. The manipulation criterion is inapplicable to this hypothesis. Hence, the temperature hypothesis cannot be rejected. However, failure to satisfy the temporality criterion indicates that factors other than temperature were responsible for the decline in abundance of Atlantic tomcod that occurred after 1990.

The striped bass predation hypothesis satisfies all of the applicable criteria. The correlations between striped bass abundance and Atlantic tomcod abundance, the temporal correspondence between the timing of the striped bass increase and the Atlantic tomcod decline, the estimates of striped bass prey consumption, the simultaneous declines in abundance of susceptible life stages of other prey species in the Hudson River, and the published studies documenting impacts of striped bass predation on prey species all provide relatively strong support for the predation hypothesis.

3.4.5 Alewife and blueback herring Figure 30 depicts long-term trends in the abundance of alewife and blueback herring PYSL and YOY in the Hudson. These two species must be considered together for purposes of evaluating impacts of CWIS, because their larvae are indistinguishable. PYSL abundance for both species combined (Figure 30a) was stable until 1985, and has since declined. With respect to YOY abundance, these two species have tended to vary together (Figure 30b). YOY abundance in both species declined abruptly in the mid-1980s and has fluctuated without apparent trend since that time, but without returning to previous abundance levels.

3.4.5.1 CWIS Co-occurrence IP2 and IP3 entrainment CMR is uncorrelated with river herring PYSL survival (Figure 31a), river herring PYSL abundance .(Figure 31b), alewife YOY abundance (Figure 32a), and 57

blueback herring YOY abundance (Figure 32b). Hence, the CWIS hypothesis fails the co-occurrence criterion.

Sufficiency There are no independent measures of sufficiency that can be applied to this hypothesis.

The objective of this report is to determine, using all available and relevant evidence whether the magnitude of entrainment and impingement at Indian Point have been sufficient to cause a reduction in the abundance of important Hudson River fish species. Hence the sufficiency criterion is inapplicable to the CWIS hypothesis.

Temporality As shown in Figures, 30a and 30b, alewife and blueback herring PYSL and YOY abundance did not decline until the mid-1980s, nearly a decade after the startup of commercial operations at IP2 and IP3. Hence, the CWIS hypothesis fails the temporality criterion.

Manipulation The peak abundance of river herring eggs and larvae typically occurs during May and June (DEIS, Figures V-71 and V-74). IP2 was offline during the entire months of May and June in 1976, 1989, 1991, 1997, 1998, and 2000. IP3 was offline during the entire months of May and June in 1975, 1982, 1993, and 1994. If entrainment at Indian Point were reducing the survival of river herring PYSL, then PYSL survival should have been higher in years when one unit was offline than in years when both units were operating. As shown in Figure 33a, the measured PYSL survival values are inconsistent with this expectation. Figure 33a shows the time series of annual PYSL survival indices from 1974 through 2002. The horizontal line in Figure 33a shows the median survival index value for this time period. The median is defined as the midpoint of the entire distribution of survival index values, meaning that one-half of the survival indices are above the median and one-half are below the median. If river herring PYSL survival were higher in years of one-unit operation than in years of 2-unit operation, then significantly more survival index values for years of one-year operation should be higher than the median than lower than the median. However, Figure 33a shows that the PYSL survival index was higher 58

than the median for 4 of the 11 years of one-unit operation. The PYSL was index lower than the median in 7 years of one-unit operation.

This result is confirmed by Figure 33b, which shows the relationship between the river herring PYSL survival index and the May-June total water withdrawals by IP2 and IP3 for the years 1975-2002. There is no correlation between withdrawals by IP2 and IP3 and river herring PYSL survival. Hence, the CWIS hypothesis fails the manipulation criterion for alewife and blueback herring.

Coherence The objective of this report is to determine, using all available and relevant evidence whether the magnitude of entrainment and impingement at Indian Point have been sufficient to cause a reduction in the abundance of important Hudson River fish species. Including "coherence" as an explicit evaluation criterion for CWIS would be redundant. Hence, the coherence criterion is inapplicable to the CWIS hypothesis.

3.4.5.2 Zebra mussels Co-occurrence As shown in Appendix B, Tables B-19 and B-21, there is no correlation between the zebra mussel index and any abundance index for either alewife or blueback herring. Hence, the zebra mussel hypothesis fails the co-occurrence criterion for both species.

Sufficiency The potential effects of zebra mussel activity on early life stages of fish are indirect, and related to reductions in prey abundance and changes in habitat quality. No experiments have been performed that could quantify the relationship between zebra mussel activity and fish growth or survival, and no mathematical models that could be used to quantify the indirect effects of zebra mussel activity have been developed. Hence, whether or not the zebra mussel hypothesis satisfies the sufficiency criterion is unknown.

59

Temporality The decline in abundance of alewife and blueback herring PYSL and YOY occurred during the mid7 1980s, more than 5 years prior to the invasion of the river by zebra mussels (Figure 30). Hence, the zebra mussel hypothesis fails the temporality criterion.

Manipulation No deliberate manipulations of zebra mussel populations in the Hudson River have been performed, therefore, this criterion is inapplicable to the zebra mussel hypothesis.

Coherence Because the proposed mechanism through which zebra mussel activity could have affected river herring in the Hudson River involves reducing food availability, the growth as well as the survival of river herring PYSL and YOY should have been reduced. Strayer et al. (2004) found a decline in the growth rate of YOY alewife following the zebra mussel invasion using both the utility beach seine index and the NYSDEC beach seine index. Only the decline in the growth rate calculated from the NYSDEC index was statistically significant, and only at the 20%

level. No relationship between alewife or blueback herring growth and zebra mussel activity was found in the analysis performed to support this assessment (Appendix B, Tables B-19 and B-21). Zebra mussel <activity should also have shifted the distribution of river herring PYSL and YOY downriver, away from the freshwater zone in which zebra mussels are abundant. Strayer et al. (2004) found net downriver shifts in the distribution of alewife and blueback herring YOY, but a net upriver shift in the distribution of PYSL. None of these shifts was statistically significant, even at the 20% level. In the analysis performed to support this assessment (Appendix B, Tables B-19 and B-21), no significant shift in the distribution of blueback herring was found, but an upstream shift in the distribution of alewife YOY was found. Only one of the predicted effects of the zebra mussel invasion on river herring was observed, in only one out of three analyses, and at a significance level (20%) not usually accepted in scientific studies.

Hence, the zebra mussel hypothesis fails the coherence criterion for alewife and blueback herring.

60

3.4.5.3 Striped bass predation Co-occurrence The river herring PYSL abundance index, which reflects spawner abundance and reproduction, is negatively correlated with the striped bass index (Figure 34a). The alewife YOY index, and the blueback herring YOY index are also negatively correlated with the striped bass index, although only at the 10% significance level (Appendix B, Tables B-19 and B-21).

(Figures 34b and 34c). Hence, the striped bass predation hypothesis satisfies the co-occurrence criterion for predation, on both adults and YOY.

Sufficiency Striped bass larger than 200 mm in length have been shown to feed on alosids, including alewife and blueback herring (Gardinier and Hoff 1982, Dunning et al. 1997). According to Savoy and Crecco (2004) and Davis et al. (2007), adult striped bass in the Connecticut River feed heavily on spawning blueback herring. Recently, Kahnle and Hattala (2007) reported that river herring were the most common prey item in the stomachs of adult striped bass captured in the Hudson River. Appendix C to this report documents an analysis of prey consumption by Hudson River striped bass. This analysis compares the change in striped bass prey consumption requirements (August through October) between earlier (1983-1990) and more recent (1991-2004) periods to changes in abundance of YOY fish in the Hudson River between these same two periods. The analysis shows that the increase in prey consumption from the earlier to the later period would be sufficient to explain the decline in YOY river herring abundance between these two periods if 3% of the age 1 and age 2 striped bass seasonal predatory demand was satisfied by YOY river herring, or if 0.9% of the age 1 through age 13 striped bass seasonal predatory demand was satisfied by YOY river herring. Hence, the striped bass predation hypothesis satisfies the sufficiency criterion with respect to predation on YOY river herring. No quantitative estimates of consumption of adult river herring by striped bass are available.

61

Temporality The decline in river herring abundance coincides in time with the increase in the striped bass index (Figure 35). Hence, the trends analysis supports the hypothesis that predation by striped bass has contributed to the decline in alewife and blueback herring abundance. Alewife and blueback herring do not return to the Hudson as spawning adults until an age of at least four years (ASMFC 1998). Hence, if only juvenile river herring were susceptible to predation by striped bass, a four-year time lag would be expected between the increase in striped bass abundance and the decline in PYSL abundance. The fact that no such time lag is apparent over the substantial time series available (Figure 35a), is consistent with the hypothesis that spawning adults are also susceptible to predation. Hence, the predation hypothesis satisfies the temporality criterion for both predation on adults and predation on YOY.

Manipulation No deliberate manipulations of striped bass predation in the Hudson River have been performed, therefore, this criterion is inapplicable to the striped bass hypothesis.

Coherence If predation by striped bass had caused the decline in abundance of river herring in the Hudson River, then the YOY abundance of other known striped bass prey species, including white perch, American shad, Atlantic tomcod, and bay anchovy, should also have declined. As shown in other Sections of this report, YOY abundance for all of these species has declined since the late 1980s, when striped bass abundance began to increase. Moreover, other published studies have concluded that striped bass predation is reducing the abundance of some prey species. Savoy and Crecco (2004) attributed recent declines in the abundance of both blueback herring and American shad in the Connecticut River to striped bass predation. This conclusion is supported by a recent study of the diet composition of striped bass present in the Connecticut River during the spring shad and river herring spawning run (Davis et al. 2007). These' authors found that striped bass between 600 and 800 mm in length feed predominantly on adult river herring. These results are consistent with the results published by Kahnle and Hattala (2007),

who found that river herring were the most abundant of the identifiable prey items in the stomachs of adult striped bass captured in the Hudson River. Hartman (2003) estimated that the 62

coastwide annual prey consumption by striped bass between 1 and 10 years of age increased by more than a factor of 8 between 1982 and 1995, from 17,900 metric tons (mt) to 147,900 mt.

Uphoff (2003) calculated even larger estimates of striped bass consumption, and attributed a 90% decline in the abundance of Atlantic menhaden in upper Chesapeake Bay from 1980 through 1999 to predation by striped bass.

Because parallel declines in other susceptible species have occurred, because predation by striped bass on adult river herring has been demonstrated, and because the other published studies have documented the influence of striped bass predation on susceptible prey species, the striped bass predation hypothesis satisfies the coherence criterion.

3.4.5.4 Summary evaluation of hypotheses Table 7 summarizes the consistency of the alewife and blueback herring data with the CWIS, temperature, and striped bass predation hypotheses. Two of the five evaluation criteria -

sufficiency and coherence - are inapplicable to the CWIS hypothesis. The CWIS hypothesis fails the co-occurrence, temporality, and manipulation criteria. Hence, the CWIS hypothesis can be rejected as an explanation for long-term trends in the abundance of age 0 river herring in the Hudson River.

The zebra mussel hypothesis fails the co-occurrence, temporality, and coherence criteria for river herring. Whether the sufficiency criterion is satisfied is unknown, and the manipulation criterion is inapplicable. Hence, the zebra mussel hypothesis can be rejected as an explanation for long-term trends in the abundance of age 0 river herring in the Hudson River.

The striped bass predation hypothesis satisfies all of the applicable criteria. The correlations between striped bass abundance and river herring abundance, the temporal correspondence between the timing of the striped bass increase and the river herring decline, the estimates of striped bass prey consumption, the simultaneous declines in abundance of susceptible life stages of other prey species in the Hudson River, and the published studies documenting predation by striped bass on spawning adult river herring, and studies documenting impacts of striped bass predation on prey species all provide relatively strong support for the predation hypothesis.

63

3.4.6. Bay anchovy

. Bay anchovy is a marine species and, because zebra mussels occur only in the freshwater' zone of the Hudson River, bay anchovy should not be susceptible to the effects of zebra mussel activity. This potential stressor is not evaluated as a cause of changes in the abundance of this species. Two hypothetical causes for these changes are evaluated below: the Indian Point CWIS and striped bass predation.

Figure 36 depicts long-term trends in the abundance of bay anchovy YOY and PYSL in the Hudson. The abundance of juvenile bay anchovy, as measured by the FSS, has declined since 1985. There has been no trend in abundance of PYSL.

3.4.6.1 CWIS Co-occurrence As shown in Figure 37, the PYSL to YOY survival rate (Figure 37a) and the PYSL index (Figure 37b) are both uncorrelated with the IP2 and IP3 CMR. Hence, the CWIS hypothesis fails the co-occurrence criterion.

Sufficiency There are no independent measures of sufficiency that can be applied to this hypothesis.

The objective of this report is to determine, using all available and relevant evidence whether the magnitude of entrainment and impingement at Indian Point have been sufficient to cause a reduction in the abundance of important Hudson River fish species. Hence, the sufficiency criterion is inapplicable to the CWIS hypothesis.

Temporality There has been no decline in bay anchovy PYSL abundance, and bay anchovy YOY abundance did not decline until the late 1980s, more than 10 years following the startup of IP2 and IP3. Hence, the CWIS hypothesis fails the temporality criterion.

64

Manipulation The peak abundance of bay anchovy eggs and larvae typically occurs during June and July (DEIS, Figures V-78). IP2 was offline during the entire months of June and July in 1976, 1998, and 2000. IP3 was offline during the entire months of June and July in 1975, 1982, 1987, 1993, 1994, and 1997. If entrainment at Indian Point were reducing the survival of bay anchovy PYSL, then PYSL survival should have been higher in years when one unit was offline than in years when both units were operating. As shown in Figure 38a, the measured PYSL survival values are inconsistent with this expectation. Figure 38a shows the time series of annual PYSL survival indices from 1985 through 2002. The horizontal line in Figure 38a shows the median survival index value for this time period. The median is defined as the midpoint of the entire distribution of survival index values, meaning that one-half of the survival indices are above the median and one-half are below the median. If bay anchovy PYSL survival were higher in years of one-unit operation than in years of 2-unit operation, then significantly more survival index values for years of one-year operation should be higher than the median than lower than the median. However, Figure 38a shows that the PYSL survival index was higher than the median for 4 of the 7 years of one-unit operation and lower than the median for the other 3 years. This difference could easily have arisen by chance.

This result is confirmed by Figure 38b, which shows the relationship between the bay anchovy PYSL survival index and the June-July total water withdrawals by IP2 and IP3 for the years 1975-2002. There is no correlation between withdrawals by IP2 and IP3 and bay anchovy PYSL survival. Hence, the CWIS hypothesis fails the manipulation criterion for bay anchovy.

Coherence The objective of this report is to determine, using all available and relevant evidence, whether the magnitude of entrainment and impingement at Indian Point have been sufficient to cause a reduction in the abundance of important Hudson River fish species. Including "coherence" as an explicit evaluation criterion for CWIS would be redundant. Hence, the coherence criterion is inapplicable to the CWIS hypothesis.

65

3.4.6.2 Striped bass predation Co-occurrence Bay anchovy juvenile abundance is negatively correlated with the striped bass index (Figure 39a). Hence, the striped bass hypothesis satisfies the co-occurrence criterion.

Sufficiency Striped bass larger than 200 mm in length have been shown to feed on clupeids such as bay anchovy (Gardinier and Hoff 1982, Dunning et al. 1997). However, the prey consumption analysis documented in Appendix C to this report did not address predation on bay anchovy.

Hence, whether the striped bass predation hypothesis satisfies the sufficiency criterion for bay anchovy is unknown.

Temporality The increase in striped bass abundance coincides in time with the decline in bay anchovy juvenile abundance (Figure 39b). Hence, the striped bass hypothesis satisfies the temporality criterion for bay anchovy.

Manipulation No deliberate manipulations of striped bass predation in the Hudson River have been performed, therefore, this criterion is inapplicable to the striped bass hypothesis.

Coherence If predation by striped bass had caused the decline in abundance of bay anchovy YOY in the Hudson River, then the YOY abundance of other known striped bass prey species, including white perch, American shad, river herring, and Atlantic tomcod should also have declined. As discussed in other Sections of this report, YOY abundance for all of the above species has declined since the late 1980s, when striped bass abundance began to increase. Moreover, other published studies have concluded that striped bass predation is reducing the abundance of some prey species.

66

Hartman (2003) estimated that the coastwide annual prey consumption by striped bass between 1 and 10 years of age increased by more than a factor of 8 between 1982 and 1995, from 17,900 mt to 147,900 mt. Uphoff (2003) calculated even larger estimates of striped bass consumption, and attributed a 90% decline in the abundance of Atlantic menhaden in upper Chesapeake Bay from 1980 through 1999 to predation by striped bass.

Because parallel declines in other susceptible species have occurred, and because the other published studies have documented the influence of striped bass predation on susceptible prey species, the striped bass predation hypothesis satisfies the coherence criterion with respect to predation on YOY bay anchovy.

3.4.6.3 Summary evaluation of hypotheses Table 8 summarizes the consistency of the bay anchovy data with the CWIS and striped bass predation hypotheses. Two of the five evaluation criteria - sufficiency and coherence - are inapplicable to the CWIS hypothesis. The CWJS hypothesis fails the co-occurrence, temporality, and manipulation criteria. Hence, the CWIS hypothesis can be rejected as an explanation for long-term trends in the abundance of age 0 bay anchovy in the Hudson River.

.The striped bass hypothesis satisfies three of the five criteria. The manipulation criterion is inapplicable to this hypothesis, and whether this hypothesis satisfies the sufficiency criterion is unknown. The simultaneous declines in abundance of susceptible life stages of other prey species in the Hudson River and the published studies documenting impacts of striped bass predation on prey species all provide relatively strong support for the predation hypothesis.

However, substantial uncertainty remains concerning the fraction of the bay anchovy YOY population that might be consumed.

3.4.7. Spottail shiner Figure 40 depicts long-term trends in the abundance of spottail shiners and YOY in the Hudson River. The abundance of shiners has significantly declined, while the abundance of YOY has significantly increased. The increase in abundance of YOY spottail shiner is inconsistent with all of the hypotheses evaluated in this report. Hence, there is no need to perform a formal evaluation using the criteria from Suter et al. (2007).

67

As shown in Figure 41, there is no correlation between the IP2 and IP3 CMR and either spottail shiner response metric. This result is not unexpected because, as discussed in the DEIS (Figure V-107), spottail shiner is a freshwater species that is uncommon in the vicinity of Indian Point. The causes of recent changes in the abundance of this species cannot be identified using the data available for this report; however, the CWIS hypothesis can be rejected.

3.5 Summary evaluation of trends analysis The results of the trends analysis are inconsistent with the hypothesis that entrainment at IP2 and IP3 is reducing the survival or abundance of any of the eight Hudson River fish species considered in this, assessment. Overfishing is the most likely cause of the recent decline in abundance of American shad, with striped bass predation being a potentially important contributing factor. For other species, the striped bass predation hypothesis is the most strongly supported hypothesis. This hypothesis satisfies the co-occurrence, sufficiency, temporality, and coherence criteria for many of the species evaluated. With respect to the co-occurrence criterion, the striped bass index is negatively correlated with abundance indices for white perch, American shad, Atlantic tomcod, river herring, and bay anchovy. With respect to sufficiency, the analyses documented.in Appendix C show that the increase in prey consumption by Hudson River striped bass in recent years is sufficient to account for observed declines in the YOY abundance of white perch, Atlantic tomcod, and river herring. With respect to. temporality, the increase in striped bass abundance that occurred following the imposition of harvest restrictions in the mid-1980s coincides in time with the declines in abundance of one or more life stages of all of these species.

With respect to coherence, striped bass predation has been implicated in declines of susceptible species in other mid-Atlantic northeastern estuaries (Hartman 2003, Uphoff 2003, Savoy and Crecco 2004) and striped bass have been shown to prey on all of the species listed above (Gardinier and Hoff 1982, Dunning et al.' 1997, Savoy and Crecco 2004, Kahnle and Hattala 2007).

The available evidence is sufficient to reject Indian Point CWIS as having a measurable effect on any of the species evaluated. Within the limits of the data available for this assessment, it can reasonably be concluded that striped bass predation is a far more likely cause of declines in 68

the abundance of YOY white perch, American shad, Atlantic tomcod, river herring, and bay anchovy than are any of the other potential causes evaluated.

4. Evaluation of impacts of cooling-water withdrawals on spawning potential Fisheries scientists have developed a variety of quantitative methods for determining whether the sustainability of a fish population is being harmed by excessive harvesting. From the perspective of population dynamics, entrainment and impingement have been characterized (somewhat over simplistically) as a type of "fishing," imposed on early life stages rather than on adult fish (Goodyear 1977). For this reason, these methods may be used to determine whether entrainment or impingement by IP2 and IP3's respective CWIS could have adversely affected Hudson River fish populations that support managed fisheries. The method to be used, the SSBPR model, has a long history of application both in power-plant impact assessment studies and in fisheries management (Goodyear 1993).

4.1 History of the SSBPR model One of the critical questions in fisheries management is how much spawning stock (essentially, the number of adult fish) must be protected from harvesting to allow a population to replace itself and persist through time (i.e., a sustainable population) (Mace and Sissenwine 1993). The so-called spawning stock biomass per recruit or SSBPR model is the most widely used approach for answering this important question for fish populations subjected to commercial and recreational fishing (Sissenwine and Shepherd 1987, Gabriel et al. 1989, Goodyear 1993, Mace and Sissenwine 1993, Rosenberg et al. 1994). Further, since it was originally developed by Goodyear (1977) as a method for assessing whether entrainment and impingement of striped bass at Hudson River power plants could, in combination with fishing mortality, threaten the ability of the population to sustain itself, its application to entrainment and impingement is well-supported.

The SSBPR model uses information on age-specific mortality and fecundity (i.e., the number of eggs produced by a female fish of a given age) to calculate the expected lifetime reproduction of a one-year-old female fish (a "recruit," in fisheries terminology). Expected lifetime reproduction is a function both of the average fecundity of female fish at each age and 69

the probability that the female will survive to reproduce at that age (Goodyear 1977). Mortality due to fishing, CWIS, or other causes reduces expected lifetime reproduction either by reducing the probability of survival (in the case of fishing), reducing the probability that spawned eggs will'survive to become one-year-old recruits (in the case of CWIS), or reducing the fecundity of female fish (e.g., through adverse environmental conditions, such as toxic chemicals). For the population to persist, each one-year-old female fish must produce at least one female egg that survives to become a one-year-old female recruit (Mace and Sissenwine 1993, Goodyear 1993).

An average female has the potential to produce far more eggs than are required to replace her (Mace and Sissenwine 1993). For example, a female striped bass can spawn 3 million or more eggs in a single year (Hoff et al. 1988; Olsen and Rulifson 1992) and can live for up to 30 years (Secor and Piccoli 1996). For the population to maintain itself at a stable level, only one of the female eggs produced by each fish over her lifetime must survive to adulthood. This massive surplus of eggs ensures that the population will be able to persist in spite of natural and potentially extreme fluctuations in environmental conditions. This massive surplus of eggs also ensures that even substantial harvesting by commercial and recreational fishermen will not adversely affect the population.

4.2 Explanation of the SSBPR concept The use of SSBPR in fisheries management derives from recognition that the lifetime reproductive capacity of a typical recruit provides a useful measure of the replacement capability of a population (Goodyear 1977, 1993, Sissenwine and Shepherd 1987, Mace and Sissenwine 1993, Rosenberg et al. 1994). At low levels of fishing mortality, the lifetime reproductive capacity of a typical female recruit is far larger than is necessary to sustain the population. As fishing mortality increases, the expected life span of each fish decreases, resulting in a reduction in lifetime reproductive capacity. If fishing mortaiity exceeds a critical threshold, the number of eggs produced by a female over her lifetime will fall below the replacement level. Once egg production falls below this level, recruitment (the number of fish entering the population each year) will begin to decline, and will continue to decline unless fishing is reduced to a level that once again allows lifetime egg production to meet or exceed the replacement level (Sissenwine and Shepherd 1987, Mace and Sissenwine 1993).

70

In a review of over-fishing definitions used in the management of marine fish stocks, Rosenberg et al. (1994),found that most of these definitions were based on the SSBPR model, and used the SSBPR model to evaluate over-fishing definitions used to manage the marine fish stocks. NOAA guidelines (Restrepo et al. 1998) for implementing National Standard 1 of the Magnuson-Stevens Act identify the SSBPR model as one of the methods that can be used to establishing over-fishing reference points that comply with the Act.

SSBPR is estimated as:

SSBPR = limif where 1i = probability of survival from age 1 to age i mi = fraction of the population of age i which are mature females; and f = average fecundity of a female fish at age i (average number of eggs/female of age i).

The probability of survival to age i is estimated by combining age-specific rates of natural mortality, fishing mortality, and entrainment/impingement mortality:

a=i-I li=He-(M,+F+ P-)

where Ma = age-specific instantaneous natural mortality rate at age a; Fa = instantaneous fishing mortality rate at age a; and Pa = instantaneous power-plant mortality rate at age a.

The impact of fishing and power-plant mortality on expected lifetime egg production is expressed as the ratio of SSBPR including both sources of mortality to SSBPR without these sources of mortality. This ratio is often termed the "spawning potential ratio" ("SPR"):

71

SPR = SSBPRfished SSBPRUnfshed Rates of fishing mortality that would produce a given SPR value are used by fisheries management agencies to establish acceptable limits on fishing mortality. Historically, the two reference points most commonly used by fisheries managers are F. 35 and F. 20. F. 35 is the fishing mortality rate that will lead to an SPR value of 0.35. F. 35 has often been used as a default goal for achieving maximum sustained yield ("MSY"), i.e., the maximum amount of adult fish (in pounds or kilograms) that can be removed from the population each year by fishermen without affecting the sustainability of the population. Values of F greater than F. 35 would lead to harvests greater than could be sustained over time. F.20 is the fishing mortality rate that will lead to SPR value of 0.2, a default value indicating over-fishing. If F consistently exceeds F. 20, then significant declines in the adult population may occur. Although some fish stocks may be able to maintain recruitment at F.20 , other stocks are more sensitive to fishing and cannot sustain exploitation at this level (Mace and Sissenwine 1993, Rosenberg et al. 1994).

4.3 Application to Hudson Riverfish populations Quantitative stock assessments and biological reference points are available for two of the species addressed in this report: striped bass (ASMFC 2005) and American shad (ASMFC 2007). As long as mortality caused by entrainment and impingement is limited to fish that are younger than one year old (which is true for both striped bass and American shad), the CMR calculated using the generators' empirical entrainment and impingement models provides a direct measure of the reduction in SSBPR caused by IP2 and IP3 (Goodyear 1977). The likelihood that entrainment and impingement at IP2 and IP3 have adversely affected the sustainability of these two species is evaluated in two ways. First, estimates of reduction in SSBPR due IP2 and IP3 are compared to reductions caused by fishing mortality. Second, estimates of combined reductions in SSBPR due to both IP2 and IP3 and fishing are compared to the biological reference points that are currently used to manage these species.

72

4.3.1 Striped bass As shown in Figure 42, the striped bass CMR for the 30 years for which data are available corresponds to an SPR of 0.92. In other words, IP2 and IP3 reduce the spawning potential of the Hudson River striped bass population to 92% of the value for an unfished population. Fishing for striped bass at the current target rate established by the ASMFC (F=0.30)10 corresponds to an SPR of 0.13. This means that fishing for striped bass, under the current management approach, has reduced the reproductive potential of a typical 1-year-old female striped bass to only 13% of the value that would be expected in an unfished striped bass population. The threshold fishing rate for striped bass is currently set at F=0.41 (ASMFC 2003).

This value corresponds to an SPR of 0.096. If the rate of fishing were to rise above F=0.41, the ASMFC would be required to declare the population to be over-fished and would take action to reduce harvesting.

As shown in Figure 42, even when effects of fishing are combined with effects of IP2 and IP3, the combined SPR is still above the threshold. Hence, either alone or in combination with fishing, entrainment and impingement at IP2 and IP3 have not jeopardized the sustainability of the Hudson River striped bass population as defined by ASMFC regulations. Further, as is clear from Figure 42, the impacts of fishing on the sustainability of the Hudson River striped bass population dwarf any impact of IP2 and IP3. Eliminating entrainment and impingement of striped bass at IP2 and IP3 would not have a measurable influence on the sustainability of the population.

4.3.2 American shad The ASMFC (ASMFC 2007a, 2007b) recently used the SSBPR model to assess impacts of increased mortality on the sustainability of Atlantic coastal American shad populations, including the Hudson River American shad population. Because the relative contributions of fishing mortality and natural mortality to the increase are uncertain, the ASMFC expressed the maximum sustainable rate of mortality in terms of total mortality (Z) rather than fishing mortality. The ASMFC selected Z.30, the total mortality rate at which SSBPR would fall to 30%

10 For assessment purposes, Atlantic striped bass are treated as a single mixed population, and the same fishing mortality rate is assumed to be applicable to all of the individual spawning populations that contribute to the mixed coastal fishery.

73

of an assumed baseline value, as an excess mortality threshold analogous to F 30 . Using alternative assumptions concerning the operation of the American shad fishery, the ASMFC developed a range of estimates of Z 30 of Z=0.54 to Z=0.73 for the Hudson River American shad population.

Empirical estimates of total annual mortality in Hudson River American shad are available for the years 1984-2004 (ASMFC 2007a). Total mortality has exceeded Z 3o in most years during this period. Hattala and Kahnle (2007) have contended that the excessive mortality imposed on Hudson River American shad is due primarily to overfishing. However, regardless of the actual cause, it is clear' that entrainment at Indian Point is a negligible contributor to American shad mortality. Figure 43 compares reductions in spawning potential of American shad due to IP2 and IP3 to reductions due to other causes, including fishing. The calculations were performed using the Hudson-specific life history parameters from Tables 1.1.5.1-b (age-invariant natural mortality) and 1.1.5.2-b of ASMFC (2007a) and the revised Type 1 fishery model from ASMFC (2007b).

As shown in Figure 43, entrainment at IP2 and IP3 would reduce the spawning potential of Hudson River American shad by only 1% compared to the baseline value. According to the ASMFC (2007a), the current rate of total mortality on age 1 and older American shad (Z=0.87) corresponds to an SPR of 0.23, well below the threshold level. Because it was derived from an analysis of long-term trends in abundance and age structure of the Hudson River shad population, the total mortality rate estimate already includes the effects of entrainment at IP2 and IP3. If this contribution (as estimated using the CMR) is removed, the decrease in total mortality and increase in SPR level are negligibly small (Figure 43). Eliminating entrainment at IP2 and IP3 would result in less than a 1% increase in spawning potential, leaving the SPR still substantially below the threshold defined by the ASMFC.

5. Community-Level Trends Analysis Cooling-water withdrawals impose some incremental additional mortality on species susceptible to entrainment. If entrainment at IP2 and IP3 were having an adverse impact on the Hudson River fish community, then species with high susceptibility to entrainment would be more likely to have declined in abundance over the past 30 years than would species with low susceptibility. Among those species that declined in abundance, the magnitude of the decline 74

should have been greater for species with high susceptibility than for species with low susceptibility. Among species that increased in abundance, the magnitude of the increase should have been lower for species with high susceptibility than for species with low susceptibility.

This hypothesis can be tested using data available from the generators' riverwide survey programs, using data for all Hudson River fish species for which an adequate trends dataset could be developed. The method used to perform the test is analysis of correlations between indices of entrainment susceptibility, as calculated using distributional data obtained from the LRS, and indices of trends in age 0 abundance, obtained from the BSS and FSS.

Evaluating the correlation between entrainment susceptibility and change in YOY abundance requires selecting those species for which data are available for both variables.

Entrainment susceptibility at IP2 and IP3 can be estimated by evaluating the distribution of entrainable life stages in the region from which IP2 and IP3 withdraws water in comparison to all the regions sampled. The generators' LRS program is designed to collect such data. The expected effect of continued annual entrainment losses of early life stages of a species, if losses are severe enough to reduce population size, is a decrease in YOY abundance. YOY is the best stage to look for the effect of entrainment losses because entrainment occurs prior to the YOY stage, and because most susceptible species are still in the river during the YOY stage and thus their abundance is measurable. The generators' BSS and FSS sampling programs are designed to monitor YOY abundance.

5.1 Methods The evaluation involves three steps: (1) calculate a species-specific numeric index of entrainment susceptibility based on data from the LRS; (2) calculate a species-specific numeric index of change in YOY abundance based on data from the BSS and FSS; and (3) determine whether entrainment susceptibility is related to change in age 0 abundance.

Susceptibility to entrainment at IP2 and IP3 was evaluated using an index of standing crop estimated from the generators' LRS for the 31-year period 1974-2004 (Appendix D).

Indian Point is located in Region 4 (Figure 1), but because of tidal and nontidal flows, can withdraw water originating in the two adjoining regions as well. Therefore, relative abundance of a species in Regions 3-5 (Figure 1), as'compared to the riverwide abundance of that species, was 75

used to define a susceptibility index termed EntSus. For each sampled year (and each seasonal period when possible), EntSus is estimated for each species as the ratio of standing crop in Regions 3-5 to standing crop in all sampled regions. For those species occurring in more than one of the three seasonal periods, annual EntSus values are calculated as an average across periods, p, weighted by abundance for each period:

EntSus = *p SCi.EntSusip YpSCip where EntSusi - fraction of species in the Hudson River estuary in the IP2 and IP3 region in year i; SCip = sum of abundance of the species within seasonal period p in year i; and EntSusip = value of EntSus for seasonal period p in year i.

Annual EntSus values for each species for each of 31 years (1974-2004) in which the yolk-sac or post yolk-sac stages appeared in the Hudson River are provided in Appendix D.

The BSS and FSS programs were selected as the best potential indicators of long-term relative abundance of fish in the estuary. These programs have sampled the estuary using similar gear and methodology since the early 1970s, although there have been variations in the regions sampled and in time of initiation and end of the sampling across the years. To maintain consistent sampling effort and maximize comparability of results, data are restricted to Regions 1-12, and weeks 31-42, approximately August through October.

As documented in Appendix D, abundance data by species are categorized into two salinity zones, three habitats, and two time periods. The two salinity zones are brackish (Regions 1-6; river miles 12-61) and freshwater (Regions 7-12; river miles62-152). The three habitats sampled by these surveys are (a) shorezone (bottom area in water 10 ft or less in depth), (b) benthic (volume of water between river bottom and 3 ft above the bottom), and (c) water column (water volume not included in either the shorezone or benthic habitats). Time series of abundance data are divided into two equal periods: Period 1, covering the years 1974 through 1989, and Period 2, covering the years 1990-2005.

76

Because freshwater and marine species typically have strong salinity preferences, data from the non-preferred salinity zones (brackish zone for freshwater guild; freshwater zone for marine guild) were excluded when calculating overall relative change in abundance from Period 1 to Period 2 for species in these two guilds. So that species with greatly differing abundances could be compared in the same scale, the between-period changes were expressed as a relative change index (i.e., abundance in Period 2 divided by abundance in Period 1). Details concerning these calculations are provided in Appendix D.

The quantity and quality of abundance and distribution data Vary greatly among species.

The inclusion of species collected only rarely, or only in a small number of years, would weaken the analysis. Selection criteria are needed to eliminate species caught'too infrequently to provide meaningful estimates of EntSus or meaningful abundance trends. However, any single choice of selection criteria can be questioned. For this reason, a sensitivity analysis was performed to evaluate influence of selection criteria on the outcome of the hypothesis test. The sensitivity analysis was performed by defining two cases, or sets of species, termed "Case A" and "Case B."

Species included in both cases were selected based on the annual numbers of organisms collected in the LRS and BSS/FSS surveys. Species were included in the Case A analysis if (1) an average of at least 100 larvae per year of occurrence was collected in LRS samples during 1974-2005 and (2) at least 100 YOY were collected in BSS or FSS samples in at least one salinity zone-habitat combination in at least one of the two time periods. Species were included in the Case B analysis if (1) an average of at least 1000 larvae per year of occurrence was collected in LRS samples 1974-2005 and (2) at least 1000 YOY were collected in BSS or FSS samples in at least one salinity zone-habitat combination in at least one of the two time periods. The species included in Case B are a subset of the species included in Case A. The selection criteria and the species included in each case are more fully documented in Appendix D.

Three correlation metrics (Pearson, Spearman, and Kendall) were used to evaluate the association between entrainment susceptibility and YOY abundance change. There is no simple mathematical relation between any two of these three methods, and when the true correlation coefficient is not zero, it is likely that each coefficient is sensitive to different types of departures from independence (Sokal and Rohlf, 1995).

77

5.2 Results and Discussion Table 9 shows the correlation coefficients and probability values, for both Case A and Case B, for all three correlation indices. None of the correlations are statistically significant.

Figure 44 provide plots of mean entrainment susceptibility vs. the normalized index of relative change in YOY abundance from Period 1 to Period 2 for both Case A and Case B.

These figures illustrate the same two patterns. First, more species decreased in abundance than increased. For the 21 species in Case A, 71% decreased and 19% increased (Figure 44a). For the 11 species in Case B, 73% decreased and 17% increased (Figure 44b).

Second, the regression of relative abundance change on EntSus is not statistically significant for any case, even at the 20% level. This means that relative change from the earlier to the later period was the same for species with high susceptibility to entrainment (high EntSus) as for species with low susceptibility to entrainment. This result is inconsistent with the hypothesis that the susceptibilities of species to entrainment at Indian Point influenced their rates of increase or decrease over the period 1974-2005. Although the number of taxa (19) included in this analysis is small compared to the total number of species present in the Hudson, these taxa represent approximately 94% (Case A) and 88% (Case B) of all age 0 fish captured in the BSS/FSS programs from 1974-2005.

The guild to which each of the 21 species in Case A belongs is indicated in Figure 44a.

Although each guild is represented by only four to six species, at least one species in each guild increased in abundance. This pattern further reinforces the conclusion that the long-term trends in abundance of the fish species inhabiting the Hudson River estuary are similar across all guilds and are unrelated to entrainment at IP2 and IP3.

6. Conclusions The FEIS and the Draft Permit for IP2 and IP3 stated that three fish species (Atlantic tomcod, American shad, and white perch) have declined in abundance in recent years, and attributed these declines to cooling-water withdrawals at IP2 and IP3. Analyses performed to test alternative hypotheses concerning the causes of these declines show that cooling water withdrawals by IP2 and IP3 did not cause these declines. Overharvesting is the most likely cause of recent declines in the abundance of American shad, with striped bass predation being a 78

potentially significant contributing factor. Striped bass predation is the most likely cause of the decline in abundance of Atlantic tomcod (as well as river herring and bay anchovy). Striped bass predation probably contributed to the decline in abundance of YOY white perch, although other unknown causes were also involved. The striped bass hypothesis is supported not only by analysis of species abundance trends, but also by four recently-published studies of striped bass predation (Hartman 2003, Uphoff 2003, Savoy and Crecco 2004, Kahnle and Hattala 2007) and by an analysis of the increase in prey consumption needed to support the recent growth of the Hudson River striped bass population (Appendix C).

Two additional lines of evidence support a conclusion that entrainment and impingement at IP2 and IP3 have not resulted in AEI. Application of the SSBPR model to stock assessment data for striped bass and American shad (Section 4) shows that mortality caused by entrainment at IP2 and IP3 is negligible, particularly compared to fishing mortality, and does not impair the ability of these populations to sustain themselves. Analysis of community-level trends data (Section 5) shows that species with relatively high susceptibility to entrainment at IP2 and IP3 are no more likely to have declined in abundance since 1974 than are species with relatively low, susceptibility to entrainment.

Considered together, the evidence evaluated in this report shows that the operation of IP2 and IP3 has not caused effects on early life stages of fish that reasonably would be considered "adverse" by fisheries scientists and/or managers. The effects of mortality at IP2 and IP3 on the survival and abundance of susceptible populations cannot be detected, even after 30 years of intensive monitoring. Those changes that have occurred are more likely attributable to predation by the Hudson River's rapidly growing striped bass population.

For all of the above reasons, from the perspective of a science-based definition of AEI, the available data demonstrate that entrainment and impingement associated with cooling-water withdrawals by IP2 and IP3 have not had an adverse impact on Hudson River fish populations and communities.

79

7. References ASA. 2007. 2005 Year Class Report for the Hudson River Estuary monitoring program. ASA Analysis and Communication, Washingtonville, NY.

Atlantic States Marine Fisheries Commission (ASMFC). 1999. Amendment 1 to the Interstate Fishery Management Plan for shad and river herring. Fishery Management Report no. 35.

ASMFC, Washington, D.C.

ASMFC 2003. Amendment 6 to the Interstate Fishery Management Plan for Atlantic striped bass. Fishery Management Report No. 41. ASMFC, Washington, D.C.

ASMFC 2004. 2004 review of the Interstate Fishery Management Plan for shad and river herring. ASMFC, Washington, D.C.

ASMFC 2005. 2005 stock assessment report for Atlantic striped bass. AMSFC, Washington, D.C.

ASMFC 2007a. American shad stock assessment for peer review. Stock Assessment Report No.

07-01 (supplement), ASMFC, Washington, D.C.

ASMFC 2007b. Terms of reference & advisory report to the American shad stock assessment peer review. Stock Assessment Report No. 07-01. ASMFC, Washington, D.C.

Barnthouse, L.W., J. Boreman, S.W. Christensen, C.P. Goodyear, W. Van Winkle, and D.S.

Vaughan. 1984. Population biology in the courtroom: the Hudson River controversy. Bioscience 34:14-19.

Barnthouse, L.W., D. Glaser, and J. Young. 2003. Effects of historic PCB exposures on the reproductive success of the Hudson River striped bass population. Environmental Science and Technology 37:223-228.

Begon, M., J. L. Harper, and C. R. Townsend. 1996. Ecology: Individuals,populations, and communities. 3P Edition. Blackwell Science, Oxford, UK.

Bigelow, H. B., and W. C. Schroeder. 1953. Fishes of the Gulf of Maine. U.S. Wildlife Service Fishery Bulletin 74:1-576.

Boreman, J., C. P. Goodyear, and S. W. Christensen. 1981. An empirical methodology for estimating entrainment losses at power plants sited on estuaries. Transactionsof the American FisheriesSociety 110:255-262.

Boreman, J. and R. J. Klauda. 1988. Distribution of early life stages of striped bass in the Hudson River estuary, 1974-1979. American FisheriesSociety Monograph 4:53-58.

80

Brook, B. W., and C. J. A. Bradshaw. 2006. Strength of evidence for density dependence in abundance time series of 1198 species. Ecology 87:1445-145 1.

Brosnan, T. M., A. Stoddard, and L. J. Hettling. 2006. Hudson River sewage inputs and impacts:

past and present. pp. 335-348 in Levinton, J. S., and J. R. Waldman (eds.) The Hudson River Estuary. Cambridge University Press, New York, NY. 471 p.

Caraco, N.F., Cole, J.J., Raymond, P.A., Strayer, D.L., Pace, M.L., Findlay, S.E.G., and Fischer, D.T. 1997. Zebra mussel invasion in a large, turbid river: phytoplankton response to increased-grazing. Ecology 78: 588-602.

Clarke, G. M., and R. E. Kempson. 1997. Introduction to the Design andAnalysis of Experiments. John Wiley & Sons, New York.

Cole, J. J., and N. F. Caraco. 2006. Primary production and its regulation in the tidal-freshwater Hudson River. pp. 107-120 in Levinton, J. S., and J. R. Waldman (eds.) The Hudson River Estuary. Cambridge University Press, New York, NY. 471 p.

Davis, J., E. Schultz, and J. Vokoun. 2007. Assessment of river herring and striped bass in the Connecticut River: Abundance, population structure, and predator/prey interactions. 2006 Progress Report. Submitted to the Connecticut Department of Environmental Protection.

Dayton, P. K., S. Thrush, and F. C. Coleman. 2002. Ecological effects of fishing in marine ecosystems of the United States. Prepared for the Pew Oceans Commission.

Dunning, D.J., J.R. Waldman, Q.E. Ross, and M.T. Mattson. 1997. Use of Atlantic tomcod and other prey by striped bass in the lower Hudson River estuary during winter. Transactionsof the American FisheriesSociety 126:857-861.

Englert, T. L., J. Boreman, and H. Y. Chen. 1988. Plant flow reductions and outages as mitigative measures. American FisheriesSociety Monograph 4:274-279.

Findlay, S., C. Wigand, and W. C. Nieder. 2006. Submersed macrophyte distribution and function in the tidal freshwater Hudson River. Pp. 230-241 in Levinton, J. S., and J. R.

Waldman (eds.) The Hudson River Estuary. Cambridge University Press, New York, NY. 471 p.

Gabriel, W. L., M. P. Sissenwine and W. J. Overholtz. 1989. Analysis of spawning stock biomass per recruit: an example for Georges Bank Haddock. North American Journalof FisheriesManagement 9:383-391.

Gardinier, M. and T.B. Hoff. 1982. Diet of striped bass in the Hudson River estuary. New York Fish and Game Journal. 29:152-165.

Goodyear, C. P. 1977. Assessing the impact of power plant mortality on the compensatory reserve of fish population pp. 186-195. In: Proceedings of the Conference on Assessing the 81

Effects of Power Plant.Induced Mortality on Fish Populations. Pergamon Press, N.Y. W. Van Winkle, [ed].

Goodyear, C.P. 1993. Spawning stock biomass per recruit in fisheries management: foundation and current use p. 67-8 1. In: S. J. Smith, J. J. Hunt and D. Rivard [ed.] Risk evaluation and biological reference points for fisheries management. CanadianSpecial Publicationin Fisheriesand Aquatic Sciences. 120.

Gotelli NJ. 1995. A primer of ecology. Sunderland MA, USA: Sinauer Associates.

Hartman, K.J. 2003. Population-level consumption by Atlantic coastal striped bass and the influence of population recovery upon prey communities. FisheriesManagement and Ecology 10:281-288.

Hilborn, R., and M. Mangel. 1977. The EcologicalDetective. Princeton University Press, Princeton, NJ.

Hattala, K. A., and A. W. Kahnle. 2007. Status of the Hudson River, New York, American shad stock. Pp 209-301 in American shad stock assessment for peer review, Assessment Report No.

07-01 (supplement), ASMFC, Washington, D.C.

Hoff, T. B., J. B. McLaren, and J. C. Cooper. 1988. Stock characteristics of Hudson River striped bass. American FisheriesSociety Monograph 4:59-68.

Howarth, R. W., R. Marino, D. P. Swaney, and E. W. Boyer. 2006. Wastewater and watershed influences on primary productivity and oxygen dynamics in the lower Hudson River estuary. pp.

121-139 in Levinton, J. S., and J. R. Waldman (eds.) The Hudson River.Estuary. Cambridge University Press, New York, NY: 471 p.

Kahnle, A. W., and K. A. Hattala. 2007. Striped bass predation on adult American shad:

Occurrence and observed effects on American shad abundance in Atlantic coastal rivers and estuaries. Pp. 182-194 in: American shad stock assessment for peer review, Assessment Report No. 07-01 (supplement), ASMFC, Washington, D.C.

Levinton, J. S., and J. R. Waldman (eds.) The Hudson River Estuary. Cambridge University Press, New York, NY. 471 p.

Luo, J. and J. A. Musick. 1991. Reproductive biology of the bay anchovy in Chesapeake Bay.

Transactions of the American FisheriesSociety 120:701-710.

Mace, P. M. and M. P. Sissenwine. 1993. How much spawning pre recruit is enough? p. 101-118. In: S. J. Smith, J.,J. Hunt and D. Rivard [ed.] Risk evaluation and biological reference points for fisheries. management. CanadianSpecial Publicationin FisheriesandAquatic Sciences 120.

82

McLaren, J. B., T. H. Peck, W. P. Dey, and M. Gardinier. 1988. Biology of Atlantic tomcod in the Hudson River Estuary. American FisheriesSociety Monograph 4:102-112.

Murdoch, W. W. 1994. Population regulation in theory and practice. Ecology 75: 271-287.

Olsen, E.J., and R.A. Rulifson. 1992. Maturation and fecundity of Roanoke River-Albermarle Sound striped bass. Transactionsof the American FisheriesSociety 121:524-537.

Pace, . L., and D. J. Lonsdale. 2006. Ecology of the Hudson River zooplankton community. Pp.

217-229 in Levinton, J. S., and J. R. Waldman (eds.) The Hudson River Estuary. Cambridge University Press, New York, NY. 471 p.

Pace, M. L., S. B. Baines, H. Cyr, H, and J. A. Downing. 1993. Relationships among early life stages of Morone americanaand Morone saxatilis from long-term monitoring of the Hudson River Estuary. CanadianJournalof FisheriesandAquatic Sciences. 50:1976-1985.

Pew Oceans Commission. 2003. America's living oceans: Charting a course for change.

Quinn, T. J., II, and R. B. Deriso. 1999. QuantitativeFish Dynamics. Oxford University Press, New York, NY. 542 p.

Rosenberg, A., P. Mace, G. Thompson, G. Darcy, W. Clark, J. Collie, W. Gabriel, A. MacCall, R. Methot, J. Powers, V. Restrepo, T. Wainwright, L. Botsford, J. Hoenig, and K. Stokes. 1994.

Scientific review of definitions of over-fishing in U.S. fishery management plans. NOAA Technical Memorandum NMFS/F-SPO-17. National Marine Fisheries Service, Silver Spring, MD.

Sokal, R. R., and F. J. Rohlf. 1995. Biometry-The Principlesand Practiceof Statistics in BiologicalResearch. 3 rd Edition. W. H. Freeman and Company, San Francisco, CA.

Restrepo, V. R., G. G. Thompson, P. M. Mace, W* L. Gabriel, L. L. Low, A. D. MacCall, R. D.

Mehot, J. E. Powers, B. L. Taylor, P. R. Wade, and J. F. Witzig. 1998. Technical guidance on the use of precautionary approaches to implementing the Magnuson-Stevens Fishery Conservation and Management Act. NOAA Technical Memorandum NMFS-F/SP-3 1. National Oceanic and Atmospheric Administration, Washington, D.C.

Rose, K. A., J. H. Cowan, Jr., K. 0. Winemiller, R. A. Myers, and R. Hilbom. 2001.

Compensatory density-dependence in fish populations: importance, controversy, understanding and prognosis. Fish andFisheries 2:293-327.

Savoy, T. F., and V. A. Crecco. 2004. Factors affecting the recent decline of blueback herring and. American shad in the Connecticut River. American FisheriesSociety Monograph 9:361-378.

Secor, D.H. and P.M. Piccoli. 1996. Age- and sex-dependent migrations of striped bass in the Hudson River as determined by microanalysis of otoliths. Estuaries 19:778-793.

83

Secor, D. H., J. R. Rooker, E. Zlokovitz, and V. S. Zdanowicz. 2001. Identification of riverine, estuarine, and coastal contingents of Hudson River striped bass based on otolith elemental fingerprints. Marine Ecology ProgressSeries 211:245-253.

Sissenwine, M. P. and J. G. Shepherd. 1987. An alternative perspective on recruitment over-fishing and biological reference points. CanadianJournalof FisheriesandAquatic Sciences 44:913-918.

Strayer, D. L., K. A. Hattala, and A. w. Kahnle. 2004. Effects of an invasive bivalve (Dreissena polymorpha) on fish in the Hudson River estuary. CanadianJournalof FisheriesandAquatic Sciences 61:924-941.

Suter, G. W. II, S. M. Cormier, and S. B. Norton. 2007. Ecological epidemiology and causal analysis. Ch. 4 in G. W. Suter II (ed.) EcologicalRisk Assessment, 2 nd Edition. Taylor &

Francis, Boca Raton, FL.

Texas Instruments (TI) 1980. 1978 year class report for the multiplant impact study: Hudson River estuary. Prepared for Consolidated Edison Co. of New York, Inc., 4 Irving Place, New York, NY 10003.

Turchin, P. 1999. Population regulation: a synthetic view. Oikos 84:153-159.

Uphoff, J.H. Jr. 2003. Predator-prey analysis of striped bass and Atlantic menhaden in upper Chesapeake Bay. FisheriesManagement andEcology 10:313-322.

U.S. Environmental Protection Agency. 1998. Guidelines for ecological risk assessment.

EPA/630/R-95/002F. U.S. Environmental Protection Agency, Washington, D.C.

U.S. Environmental Protection Agency. 2002. Summary of biological assessment programs and biocriteria for states, tribes, territories, and interstate commissions: streams and wadeable rivers.

EPA-822-R-02-048. U.S. Environmental Protection Agency, Washington, D.C.

Waldman, J. R., D. J. Dunning, Q. E. Ross, and M. T. Mattson. 1990. Range dynamics of Hudson River striped bass along the Atlantic coast. Transactionsof the American Fisheries Society 119:910-919.

Waldman, J. R., and M. C. Fabrizio. 1994. Problems of stock definition in estimating relative contributions of Atlantic striped bass to the coastal fishery. Transactionsof the American FisheriesSociety 123:766-778.

Walter, J.F. III., A.S. Overton, K.H. Ferry, and M.E. Mather. 2003. Atlantic coast feeding habits of striped bass: a synthesis supporting a coast-wide understanding of trophic biology.

FisheriesManagement and Ecology 10:349-360.

84

/

Winemiller, K. 0., and K. A. Rose. 1992. Patterns of life-history diversification in North American fishes: implications for population regulation. CanadianJournalof Fisheriesand Aquatic Sciences 49:2196-2218.

World Commission on Environment and Development. 1987. Our Common Future. Report to the United Nations General Assembly, August 4, 1987.

Young-Dubovsky, C., G. R. Shepherd, D. R. Smith, and J. Field. 1996. Striped Bass Research Study: Final Report. Jointly published by the U.S. Fish and Wildlife Service and the National Oceanic and the National Marine Fisheries Service, Washington, D.C.

85

Table 1. Expected effects of stressors on Hudson River fish populations (except Atlantic tomcod): age 0 growth, age 0 survival, and age 0 spatial distribution, and adult age structure.

Response metric CWIS Fishing Zebra mussels Predation by striped bass PYSL Abundance

• T PYSL---Juv survival 40 Juvenile abundance Juvenile growth Spatial distribution i

Table 2. Expected effects of stressors on Hudson River fish Atlantic tomcod population: Age 0 survival, age 1 survival, juvenile growth, and spatial distribution.

Response metric CWIS Temperature Striped bass predation PYSL/early juvenile abundance Egg to age 1 survival Age 1 &2 abundance 40 Age 1 to age 2 survival 4 Juvenile growth Spatial distribution 9

ii

Table 3. Consistency of hypotheses with evaluation criteria: striped bass.

CWIS Fishing Zebra Mussels Co-occurrence + +

Sufficiency N/A unknown unknown Temporality + -

Manipulation + N/A Coherence N/A +

Summary evaluation CWIS and zebra mussel hypotheses rejected Most likely cause: fishing iii

Table 4. Consistency of hypotheses with evaluation criteria: white perch.

CWIS Zebra mussels Striped bass predation Co-occurrence + +

Sufficiency N/A unknown +

Temporality + + (?)

Manipulation N/A N/A Coherence N/A +(?) +

Summary evaluation CWIS hypothesis rejected.

Zebra mussels and striped bass predation may have contributed declines occurring in later years, but other unknown causes were I responsible for declines occurring between 1975 and 1985.

I, iv

Table 5. Consistency of hypotheses with evaluation criteria: American shad.

CWIS Overfishing Zebra Striped bass mussels predation Co-occurrence + -+ (?)

Sufficiency N/A + unknown unknown Temporality -+ +

Manipulation -_N/A N/A N/A Coherence N/A + +

Summary CWIS and zebra mussel hypotheses rejected evaluation Most likely cause: fishing, with striped bass predation a potential contributing factor V

Table 6. Consistency of hypotheses with evaluation criteria: Atlantic tomcod.

CWIS Temperature Striped bass predation Co-occurrence + +

Sufficiency N/A + +

Temporality +

Manipulation N/A N/A Coherence N/A + +

Summary evaluation CWIS hypothesis rejected Temperature a significant influence, but cannot explain post-1990 decline Most likely cause of decline: striped bass predation vi

Table 7. Consistency of hypotheses with evaluation criteria: River herring.

CWIS Zebra mussels Striped bass predation Co-occurrence +

Sufficiency N/A N/A +

Temporality +

Manipulation N/A N/A Coherence N/A +

Summary evaluation CWIS and zebra mussel hypotheses rejected Most likely cause: striped bass predation vii

Table 8. Consistency of hypotheses with evaluation criteria: bay anchovy.

CWIS Striped bass predation Co-occurrence +

Sufficiency N/A Unknown Temporality +

Manipulation N/A Coherence N/A +

Summary evaluation - CWIS hypothesis rejected Striped bass predation most likely cause of change viii

Table 9. Pearson, Spearman, and Kendall correlation coefficients for the association between Loglo(R) and mean EntSus. A value ofp represents the probability of a sample correlation coefficient larger than the observed sample correlation coefficient, if the true correlation coefficient is zero.

Case N Pearson Spearman Kendall 19 r 0.225 0.182 0.129 A

p 0.355 0.457 0.442 12 0.157 -0.042 -0.046 B

_ p 0.625 0.897 0.837 ix

Figure 1. Hudson River map, with sample regions

) REGION RIVER MILES STUDY AREA,,

ALBANY (125-152)

CATSKILL (107-124)

SAUG ERTIES (94-106)I -

KINGSTON; (86)-93)

HYDE PARK (77-85)

Pc)UGHKEEPSIE (62-76) - 'ANSKAMMER!

'D:

CORNWALL (5r-81)

WEST POINT ~

.. ..T._:-

(47-55) ('394*)~ ~.*..:~~

INDIAN POINIT 13-4) POI IDIN OIT CROTON- HAVERSTRAW B L(34-38)

TAPPAN ZEE (24-33)

YONKERS (12-23) J BATTERY (0-11) , A / '

A*lantic Ocean x

Figure 2. Impacts of CWIS on Age 0 life stages, partitioned between abundance of each life stage and survival between life stages.

Survival I Abundance I Survival I Survival I Abundance I Abundance I xi

Figure 3. Impacts of fishing on Age 0 life stages.

I Decline in spawning Abundance 4 Abundance 4 Abundance 4 Abundance I xii

Figure 4. Impacts of zebra mussel activity on Age 0 life stages.

Abundance I Survival Growth Survival I Growth I Yb u Abundance I xiii

on Age 0 life stages.

Figure 5. Impact of striped bass predation I Decline in spawning Abundance I Abundance Survival Abundance Abundance I

xiv

Figure 6. Impact of elevated summer temperatures on Age 0 Atlantic tomcod.

Abundance I Survival Growth Survival Abundance xv

Figure 7a. Long-term trends in the abundance of striped bass PYSL and YOY.

7000 3500 un 0

- 6000 -PYSL 3000 D x 5000 2500 V -<

0

.j 4000 2000 -<

C/.

3000 1500 m 2000 1000 0 C,)

0.Lw 1000 -,500 0 0 Figure 7b. Long-term trend in striped bass PYSL to YOY survival.

0.03 x

- 0.025 Ca

• 0.02 "j 0.015 U) 0.01

¢ 0.005 0,

A , b b IZ 0 . b.* 6 Z , t (, C, Q, 01, t.

xvi

Figure 8a. Relationship between striped bass PYSL abundance and striped bass YOY abundance.

C'-

C:

M 3500 0

.-. 30000, x

CD

,C' 2500 a)

C.

ct- 2000 00

.a 1500 100 Cn

>- 1000 ,

al)

-.o 500

"- 0 0~0 0 1000 2000 3000 4000 5000 6000 7000 Striped bass PYSL abundance index (millions)

Figure 8b. Relationship between striped bass PYSL abundance and PYSL survival.

0.03 -

x a) o

-S 0.025

= 0.02 U)

-J U)

>- 0.015 0n Ut)

CO 0.01 La-0.005 U)0 o 1 0

0 1000 2000 3000 4000 5000 6000 7000 Striped Bass PSL Index (millions)

Figure 9a. Relationships between IP2 and IP3 CMR for striped bass and striped bass PYSL survival index.

0.03 X

0.025 -

0.02 -

U)

-j C/) 0.015 -

CL C')

U)

(U 0.01 -

CL (0 0.005 -

0 0 *  ! o.4c, 0 5 10 15 20 Striped Bass CMR (%)

Figure 9b. Relationship between IP2 and IP3 CMR for striped bass and striped bass PYSL abundance index.

7000 U)

C:

0 6000 X 5000 a)

C:

-j 4000 U)

U) 3000 (n

2000 0

-a L.2

• 0 U) 1000 0 0

4~A4 0404 0 0 1 0 5 10 15 20 Stripedýpss CMR (%)

Figure 10. (a) Striped bass PYSL to YOY survival during years in which 1 unit (blue) and 2 units (red) at Indian Point were operating during May and June, the peak months during which entrainable life stages of striped bass are present in the Hudson River. The horizontal line shows the median survival index value for the time series. (b) Relationship between total May-June withdrawals by IP2 and IP3 and striped bass PYSL survival.

(a) 0.03 0.025-0.02 0.0151

• 0.01 0 0.005 0 i .. .. . . . . . . . . . 1 11 . - . . V 4 21 A<O AA A N ýb ýD A 5:b ýb :b A ýb Nl:b Ncb Ncb 0i

• N41 NCP e e e NWO1e 1ýp (b)

X 0.03 aI) 0.025 -

C')

0.02 -

C')

0.0,15 -

_0 0+

M, 0.01 -

0 0#

0.005 -

0 0 50000 100000 150000 Combined May-June Flows for IP2 and IP3, 1975-2002 xix

Figure 11. Long-term trends in the abundance of white perch PYSL and YOY in the Hudson River.

-. 2500 3500 0 3000 CD E 2000 "-" ~--YOY 'CPYSL" x 2500 0 "V

CD -<'

" 1500 2000 0 ClCL 1500 CD

- 1000 x CL 1000 0 (D 500 500 CL 0- -0 xx

Figure 12a. Relationship between white perch PYSL abundance and YOY abundance.

En 3500

_0 C:

C-

-" 3000 0

X9 a 2500 V

_0 S 2000 (V

C 1500 -

.0

>- 1000o° "2 500 + +

a) 0 0-0500 '1000 1500 2000 2500 White perch PYSL abundance index (millions) a)

U,00 Figure 12b. Relationship between white perch PYSL survival and YOY abundance.

3500 X

03000 5 0C 3 2500

_0

.o 2000 -

U-13.

0a)

.- 1500 c 1000 +

-C00 2500 "

0 0.001 0.002 0.003 0.004 0.005 0.006 White Perch PYSL survival index xxi

Figure 13a. Relationship between the IP2 and IP3 CMR for white perch and the white perch PYSL survival index.

X 0.006 0.005 cn

-j 0.004

• 0 0.003 0.002

• 0 00*

0.001

• e. *

• 0.00 0

0 5 10 15 20 Indian Point CMR (%)

Figure 13b. Relationship between the IP2 and IP3 CMR for white perch and the white perch PYSL abundance index.

2500 X

(,

"0 C 2000

_0 C

-0 1500 1

• Cn 1000 I*-ý" Y.

0 0 4 0 ** 0

• 500 0

0 5 10 15 20 Indian Point CMR (%)

xxii

Figure 14. Long-term trends in IP2 and IP3 CMR for white perch and white perch PYSL survival.

0.006 20 18 0.005 - PYSL survial 16 Indian Point CMR 14 0.004 12 0.003 10 8

0.002 6 4

0.001 2

o Z 0 xxiii

Figure 15. (a) White perch PYSL to YOY survival during years in which 1 unit (blue) and 2 units s(red) at Indian Point were operating during May and June, the peak months during which entrainable life stages of white perch are present in the Hudson River. The horizontal line shows the median survival index value for the time series. (b) Relationship between total May-June withdrawals by IP2 and IP3 and white perch PYSL survival.

(a)

X V

a) 0.006 "0

0.005 0

0 0.004 0

0.003 0 8

CD) 0.002 8 8 0

a) a)

8 4, 4, 0 0.001

8. *

• 8 A

1975 1978 1981 1984 1987 1990 1993 1996 1999 2002 X 0.006 (b)

,-o 0.005 0) 0 0.004

-C 0.003 CL 3J 0_

0.002 C.,

0.001 ~ 8 8 * *8

• 8

• 8 0

0 50000 100000 150000 Combined May-June Flows for IP2 and IP3 xxiv

Figure 16a. Relationship between white perch YOY abundance and the striped bass predation index.

U) 2.00E+03 -

0.

U) 0

-C 1.50E+03 -

x C

0 1.OOE+03

-C U,) 0 CL.

U1) 0 4$'

5.OOE+02 - A 0.OOE+00 I I 0 1000 2000 3000 4000 5000 6000 7000 Striped Bass Abundance Index (millions)

Figure 16b. Long-term trends in white perch YOY abundance and the striped bass predation index.

7000 3500 6000 3000 5000 2500 4000 2000 3000 1500 2000 1000 1000 500 0 0

,A", 21(6ok § 1o 0ý1,ý0 ,"t 1 xxv

Figure 17. Long-term trends in abundance of American shad PYSL and YOY abundance in the Hudson River.

160 3500

-PYSL Abundance 140 - YOYAbundance 3000 120 2500 100 2000 80 1500

.60 1000 40 20 500 0 0

,pOo ,o c c, ,lb(:(b N~bNbNDJ N~ NIO& q Jq Nq Nq J N q q 16 p i xxvi

Figure 18a. Relationship between American shad PYSL abundance and YOY abundance in the Hudson River.

"-" 3,500 CU

-3 3,000 0¢-C 0

j 2,500 C-CU "o 2,000

.0.

>- 1,500 0>..

-a 1,000 CU c.-

- 500 CU a)0 E 0 0 50 100 150 200 American shad PYSL abundance (millions)

Figure 18b. Relationship between American shad PYSL survival and YOY abundance in the Hudson River.

x 3,500

• - 3,000 o

C:

CU

_0 t.- 2,500 C

CU 2,000 0U-1,500 oCU

>" 1500 1,000 CU E

0 0 0.02 0.04 0.06 0.08 0.1 American shad PYSL survival index XXVII

Figure 19a. Relationship between the 1P2 and TP3 CMR for American shad and American shad PYSL survival.

0.1 x

a) 0 0.09 C 0 E 0.08

_0 0.07

-J 0.06 0.05 CU)

Ca 0

• 4 0.04 0.03 0

-) 0.02 *~ 4

• 4 ,

0.01 E

0 0 1 2 3 4 Indian Point CMR (%)

Figure 19b. Relationship between the IP2 and IP3 CMR for American shad and American shad PYSL abundance.

I--, 160 C

0 140 x 120 a) 100

-j 0

o. 80

_0 60 CU 40 E

°E 20

, 0 0

0 2 4 6 8 xxviii Indian Point CMVR(%

Figure 20. (a) American shad PYSL to YOY survival during years in which 1 unit (blue) and 2 units s(red) at Indian Point were operating during May and June, the peak months during which entrainable life stages of American shad are present in the Hudson River. The horizontal line.

shows the median survival index value for the time series. (b) Relationship between total May-June withdrawals by IP2 and IP3 and American shad PYSL survival.

(a) 0.1 0.09 0.08 0.07 0.06 0.05 e 0 4 0.04 CU) 0.03 S: 0.02 0 .. 9 WO 0*0*

Z.

0.01 0 I9 I i i I I 1 I9 I I I I 1 I 2 1 985 1987 1989 1991 1993 1995 1997 1999 2001 (b) 0.1 0.09

'C 0,

0.08 0.07 0.06 0

0.05

_0 0.04 Cl) 0.03 Cu 0.02 CU 0.01 Eu 0

0 50000 100000 150000 Combined May-June Flows for IP2 and IP3 xxix

Figure 21 a. Relationship between American shad PYSL abundance and the striped bass predation index.

175 C:

0 150

• E a)

_0 C:

125 -

-j C', 100 -

_0 M 75 -

a)

Ec 50 -

S ~~0- -

25 -

0 4 0 I I I I I I 0 1000 2000 3000 4000 5000 6000 SB Predation Index (millions)

Figure 21b. Relationship between American shad YOY abundance and the striped bass predation index.

0 0

CU Cn C:

0 Ec 8 0

0 0 1000 2000 3000 4000 5000 6000 SB Predation Index (millions) xxx

Figure 22. Long-term trends in American shad PYSL abundance and in the striped bass predation index.

. 7000 160 C

0 6000 - SB Predation

-Amercan shad PYSL 140 B cD X 120 a 5000 120 (

100 4000 oT 80 -<

CU CD

- 3000 I-60 CC 2000 CCD S40 10 1000 20 --

0 0-*

U 0

(I) xxxi

Figure 23. Long-term trends in the abundance of Atlantic tomcod in the Hudson River.

200 - 30 180 LRS

- Mark-Recapture 25 160 140 20 1220 100 15 80 60 60 10 40 5 20 0 -91 0 xxxii

Figure 24a. Relationship between Atlantic tomcod egg deposition and resulting age 1 abundance.

12 10 C:

0 8 E

6 4

a)

W, 2

0 0 20 40 60 80 100 120 Egg deposition (Billions)

Figure 24b. Relationship between Atlantic tomcod egg to age 1 survival and age 1 abundance.

12 -

c 10 CO 8

'0 o.,

0 E

0 4 2

0 0 0.2 0.4 0.6 0.8 1 Atlantic tomcod egg to age 1 survival index xxxiii

Figure 25a. Relationship between IP2 and IP3 CMR and Atlantic tomcod egg to age 1 survival.

0.9 0.8 -

X a) 0.7 -

C 2) 0.6 -

76 0"

0m 0.5 -

0 0.4 -

0.3 -

• 0

0) 0 0) 0.2 -

e 0.1 -

0 N0 v

0 5 10 15 20 25 30 Indian Point Entrainment CMR (%)

Figure 25b. Relationship between IP2 and IP3 CMR and Atlantic tomcod LRS index.

200 180 160 x 140 (D

120

._j 100 "0

0 80 0

E

¢-

60 40 20 0

0 10 20 30 40 Indian Point Entrainment CMR (%)

Figure 25c. Relationship between IP2 and IP3 CMR and Atlantic tomcod mark-recapture index 30 25 X

a) 20

_0 15 8

E

.2 10-C.2 C

5 0

-~ :::: *0* 0 0 10 20 30 40 Indian Point Entrainment CMR (%)

xxxiv

Figure 26. (a) Atlantic tomcod age 0 survival during years in which 1 unit (blue) and 2 units (red) at Indian Point were operating during May and June, the peak months during which entrainable life stages of Atlantic tomcod are present in the Hudson River. The horizontal line shows the median survival index value for the time series. (b) Relationship between combined IP2 and IP3 May-June withdrawals and Atlantic tomcod egg to age 1 survival.

(a) x 0.9 a)

_0 0.8 -

oa C) 0.7 -

0.6 -

CU 0.5 U0 0

0.4 0

E 0.3 0

• 0 c.

t" 0.2 U) 0*

0.1 A. 4 0 I I I I I I I I I I I I I I I I T I I I I I I I I I 00~~C Z01 NCJN~

C NNj N: NNoN<fl OPNc F N1 W (b) x 0.9 C6 0.8 0.7 CO, 0.6 0.5 0

0) 0.4 I-0.3 E

0 0

E 0.2j 0.1 e 0 ee 0 0 50000 100000 150000 Combined May-June Flows for IP2 and IP3 (MGal) xxxv

Figure 27. Comparison of long-term trends in the PWW degree-day index to long-term trends in the abundance of age 1 and age 2 Atlantic tomcod.

160 30

- PWW Degree-Days 140 -- Age 1-2 tomcod (millions) 25 120-o 120 100 80 15 60 10 40

_5 20 0 -0 NCP Ncb Zý Ncb CP "0 NCb

(

xxxvi

Figure 28a. Relationship between the striped bass predation index and the Atlantic tormcod LRS index.

250 0

E 200 x

CD CO150 ,

-j

_0 oo 100 E

o Cu*" 50

• 0-0 1000 2000 3000 4000 5000 6000 SB Predation Index (millions)

Figure 28b. Relationship between the striped bass predation index and the Atlantic tomcod mark-recapture index.

Cj C

0 30 x

25

! 20 0

0 E 15 0

C 10 5, *8

• 0 1000 2000 3000 4000 5000 6000 SB Predation Index (millions) xxxvii

Figure 29a. Long-term trends in the Atlantic tomcod LRS index and the striped bass predation index.

7000 200 C

o 180 6000 E 160 0 a 5000 140 0

~3

~~0 4000 120 0 C: 100 C 3000 80

.CL

< 2000 60 CD 40 m 1000 0 "20 a) 0 - 0 C',

Figure 29b. Long-term trends in the Atlantic tomcod mark-recapture index and the striped bass predation index.

- 7000 6 C

0 o 6000 5 5 0" a 5000 0 c3 0

4000 o3 C 3000

.2

< 2000 U)

Ca3 1000 1 0 0 xxxviii

Figure 30a. Long-term trend in abundance of river herring PYSL in the Hudson River.

4 3.5 3 - River herring PYSL

.2.5 2

1.5

.1 0.5 0

Figure 3(b. Long-term trends in abundance of alewife and blueback herring YOY in the Hudson River.

1200 35 1000 30 25 800 20 600 15 400 10 200 5 0 0 t K- Q0

'A 00 2ý 0 rý,

0 CO' Oj tK Q0 Cb Z(Zý

'A 'A

,0j Cb Cb 03 " NF' rlý rp rp N0 NONP N(P NCP NCP Nq, 03 NCbOO

Figure 3 Ia. Relationship between the IP2 and IP3 CMR and river herring PYSL survival.

0.04 x

CD 0.035 0.03 CD 0.025

-J W 0.02 0L 0.015 CD 0.01 -I 0.005

.4 8*

8 0 ~ 8 0 1 2 3 4 5 6 Indian Point CMR (%)

Figure 3 lb. Relationship between the IP2 and IP3 CMR and river herring PYSL abundance.

4 3.5 CO 3 e x

2.5t a;)

2-

• 8 1.5 -

17 e e

• 8 0.5.-

8 0

0 1 2 3 4 5 6 Indian Point CMR (%)

xl

Figure 32a. Relationship between the IP2 and IP3 CMR and alewife YOY abundance.

1200 Cn 1000 -

C:

0 800 -

x 600 -

0 400 -

(D 200 - b*

0 -I 0 1 2 3 4 5 6 Indian Point CMR (%)

Figure32b. Relationship between the IP2 and IP3 CMR and blueback herring YOY abundance.

U)

C 35 0

30 -

x (D

25 -

0I- 20 -

0)

C-e- 15 -

-o 10 -

U) * ,)

5 0

0 1 2 3 4 5 6 Indian Point CMR (%)

xli

Figure 33. (a) River herring (alewife and blueback herring) PYSL to YOY survival during years in which 1 unit (blue) and 2 units (red) at Indian Point were operating during May and June, the peak months during which entrainable life stages of river herring are present in the Hudson River. The horizontal line shows the median survival index value for the time series. (b)

Relationship between IP2 and IP3 May-June water withdrawals and river herring PYSL survival.

(a)

X

'0 CD 0.04 0.035 0 0.03

0) 0.025 C',

09 0.02 0.015

._j CD 0.01 ai) 0.005 00 . ."

1 1980. .19 "1 1 I 1I 1992 I I I1 2 01 1974 1977 1980 1983 1986 1989 1992 1995 1998 2001 (b) x "iO

- 0.04 0.035 0.03 0.025 0.02 0r)

C-0.015 t) 0- 0.01 0.005 0

0 50000 100000 150000 Combined May-June Flows for IP2 and 1P3 xlii

Figure 34a. Relationship between the striped bass predation index and river herring PYSL abundance.

4 0

a)

-J Li) ~,0. 0 2 0 0

0l

• 0 a) 1 0

I

• e

5) 0$ *0, 0 0

0 0 1000 2000 3000 4000 5000 6000 SB Predation Index (millions)

Figure 34b. Relationship between the striped bass predation index and alewife YOY abundance.

1200 1000 CU Li) 0 800

= 600

-- 0 0 400

-.... 0 U 200 0.... ........

0 0 0 1000 2000 3000 4000 5000 6000 SB Predation Index (millions)

Figure 34c. Relationship between the striped bass predation index and blueback herring YOY abundance.

35 35 0

E 30 25 0 20 E 15 0 10 . ~

0 00 C) 5 0 0 0

• 00 Y.. 0 0 00 ~ .... 0 (0 1000 2000 3000 -4000, 5000 6000 SB Predation Index (millions) xliii-

Figure 35a. Long-term trends in river herring PYSL abundance and in the striped bass predtation index.

.2 7000 ...........

E - SB PrndatUn 3.5 <

S6000 -- ierhng PYSL

- 3  :

3 CD

- 5000 3.

4000

~2.5 .5 C2 U 3000 ,* 1.5 M 2000 V1 1000 0.5 =

0 0 Figure 35b. Long-term trends in alewife YOY abundance and in the striped bass predation index.

7000 1200 6000 -- 1000

- SB Predation 5000 -- Alewfe YOY 800 4000 600 3000 2000 1000 200 0 T0 Figure 35c. Long-term trends in blueback herring YOY abundance and in the striped bass predation index.

7000 35 6000 -SB Predation 30

-Blueback herring YOY 5000 25 4000 . 20 3000 15 2000 10 1000 5 0 0 xliv

Figure 36. Long-term trends in abundance of bay anchovy PYSL and YOY.

6000 500 450 5000 400 350 4000 300 3000 250 200 2000 150 100 1000 50 0Pr IT---vf l Y- pYIhI qn 1 F I 1 )6 1 -I 0 Kb ~Rý Kb(bbZý Zb ,b~

xlv

Figure 37a. Relationship between the IP2 and IP3 CMR and bay anchovy PYSL to YOY survival.

200 x

CD 180 160 140 120 100 0~

80 C) 60 CO 40 20 0

0 5 10 15 20 25 Indian Point CMR (%)

Figure 37b. Relationship between the IP2 and IP3 CMR and bay anchovy PYSL abundance.

6000 e

a) 5000 CO 4000 3000 CO 2000 1000 0 _ *I AL&~A~

ý -- £*~~?

I I 2 0 5 10 15 20 25 Indian Point CMR (%)

xlvi

Figure 38. (a) Bay anchovy PYSL to YOY survival during years in which 1 unit (blue) and 2 units (red) at Indian Point were operating during May and June, the peak months during which entrainable life stages of river herring are present in the Hudson River. The horizontal line shows the median survival index value for the time series. (b) Relationship between total IP2 and IP3 June-July withdrawals and bay anchovy PYSL survival.

(a) 1000 x 100 CU 0 10 o

¢0 0

>1 1 '~

0 0

CU 0.1 CO 0.01 I I I I N

0 N~b~ 'A (b) 1000 X

0. 100 -

V 10 -

C',

>1

• 0 0 0 1-0 *

• 4 cc nO 0.1 0.01 I 0 50000 100000 150000 200000 Combined May-June Flows for IP2 and IP3 xlvii

Figure 39a. Relationship between bay anchovy YOY abundance and the striped bass predation index.

600 C/) 500 -

0

.E 400 -

x 300 -

0 200 -

4 0r-

'Z' CU) 100 -

co a,

0 0 1000 2000 3000 4000 5000 6000 7000 Striped Bass Predation Index (millions)

Figure 39b. Long-term trends in bay anchovy YOY abundance and the striped bass predation index.

7000 500 450 6000 400 5000 350 4000 300 250 3000 200 2000 150 100 1000 50 0 0

'A, o, o*'-' Rrv o$- 0, ,' o[* b " "i"o- ' 01-c(-:ý, I 3"

& ,z

Figure 40. Long-term trends in the abundance of spottail shiner eggs and YOY.

1.8 2.5 1.6 -- Eggs (millions)

-YOY (millions) 1.4 2 1.2 1.5 1

0.8 1

0.6 0.4 0.5 0.2 0

xlix

Figure 41 a. Relationship between the IP2 and IP3 CMR and spottail shiner egg to YOY survival.

xC) 140 -

120 -

_0 100 -

0 80 -

0) 60 -

C,)

40 -

_C 20 -

-f 0

0 2 4 6 8 10 Indian Point CMR (%)

Figure 41b. Relationship between the IP2 and IP3 CMR and spottail shiner YOY abundance.

2.5 2

U) 1.5 CU

• 0 1 a-

• e U) 0.5 -

e V.

0 -I.

0 2 4 6 8 10 Indian Poin4ICMR (%)

Figure 42. Relative influence of 1P2 and 1P3 vs. fishing on the spawning potential of Hudson River striped bass.

1.00 0.90 Spawning Potential Ratio (SPR) 0.80 - SPR at Threshold F 0

0.70 0.60 0.50 CL 0) 0.40

.C C:

0.30 (U

0.20 0.10 0.00 -H-Unfished population Indian Point Target F Target F + IP li

Figure 43. Comparative effects of Indian Point and fishing on Hudson River American shad SPR using data and modeling method from 2007 American shad stock assessment (ASMFC 2007a).

1.00 -r Spawning Potential Ratio (SPR) 09004- -SPR at Threshold Z 0 0.80 4-0.70 +

0.60 +

0 0.50 +

0)

C: 0.40 -I

-E 0.30 4-CU) 0.20 -

0.10 --

0.00 n pu I I Unfished population Indian Point Current Z Current Z - IP lii

Figure 44a. Relationship between relative change in YOY abundance from Period 1 to Period 2 and entrainment susceptibility for the 21 fish species included in Case A. Zero on the logarithmic Y axis corresponds to no change in abundance from Period 1 to Period 2.

CASE A 2

• Anadromous reEstuarine A Freshwater

• Marine A A 0o S, A 0 U

-2 0 0.1 0.2 0.3 0.4 0.5 Mean Entsus Figure 44b. Relationship between relative change in YOY abundance from Period 1 to Period 2 and entrainment susceptibility for the 11 fish species included in Case B. Zero on the logarithmic Y axis corresponds to no change in abundance from Period 1 to Period 2.

CASE B 2 O 0* Anadromous m Estuarine A Freshwater

• Marine

~0 *

-2 0 0.1 0.2 0.3 0.4 0.5 Mean Entsus liii

APPENDIX A Prepared by:

AKRF, Inc.

TABLE OF CONTENTS I. INTRODUCTION ................................................................................................................... 1 II. COMPARISON OF HUDSON RIVER GENERATORS' DATA ................... 1 A. M ethods ............................................................................................................................... 1 B. Results ................................................................................................................................. 2 Il. COMPARISON OF STRIPED BASS DATA WITH INDEPENDENT STUDIES ...... 2 A. Methods ............................................................................................................................... 3 B. Results ................................................................................................................................. 3 IV. COMPARISON OF ATLANTIC TOMCOD DATASET WITH INDEPENDENT STUDIES ............................................................................................................................ 3 A. M ethods ............................................................................................................................... 4 B. Results ................................................................................................................................. 4 V. LITERA TU RE CITED ................................................................................. 5 VI. TABLES ................................................................................................................................. 6 VII. FIGURES .............................................................................................................................. 20 i

I. INTRODUCTION Indices of relative abundance, derived from Hudson River Generator's Longitudinal River Ichthyoplankton Survey ("LRS"), Beach Seine Survey ("BSS"), and Fall Shoals Survey

("FSS") data, are used to analyze trends in abundance and to test the impact hypothesis for eight different species of finfish found in the Hudson River. These analyses are presented in Appendix B.

To confirm that the selection of relative abundance indices in Appendix B is valid, this document presents an examination of relationships that exist among LRS, BSS and FSS data. It also examines relationships that exist among LRS, BSS and FSS data and data from the Atlantic States Marine Fisheries Commission ("ASMFC"), as well as relationships that exist with the coast-wide striped bass abundance derived from its stock assessment (ASMFC 2005), the New York State Department of Environmental Conservation ("NYSDEC"), and the Hudson River Generators' mark-recapture studies of Atlantic tomcod ("ATMR") and striped bass. Correlation among these surveys validates the use of the LRS, BSS and FSS in Appendix B and demonstrates the robustness of the trends analysis and test of impact.

The strength of the correlation analysis can be evaluated using a power analysis. The power of a particular statistical test refers to the probability that the null hypothesis has been correctly rejected. In the case of a correlation analysis, the null hypothesis is defined as no significant correlation between surveys. The alternative hypothesis is defined as the presence of significant correlation between surveys. The power of a correlation analysis for different sample sizes is shown in Figure 1.

II. COMPARISON OF HUDSON RIVER GENERATORS' DATA A correlation analysis was used to validate the use of the BSS and FSS surveys. The analysis demonstrates that the abundance index derived from the BSS follow the abundance index derived from the FSS.

A. Methods Two datasets were compared in this analysis. Species-specific young-of-year indices based on the BSS were compared with species-specific FSS indices. See Appendix B for details on the development of these indices. The BSS and FSS indices are presented in Tables A-1 and A-2. The FSS indices were subset to the time period 1985 through 2004 to ensure that gear were comparable to the gear used in the BSS.

A Pearson correlation analysis was conducted, comparing the indices on a species-specific basis. A weighting factor based on the inverse of the variance was used, as described in the formula below:

1-WF (SEsS )2 + (SEFss )2 where:

I

WF = weighting factor for Pearson Correlation Analysis SEBss = standard error of BSS abundance estimate SEFss = standard error of FSS abundance estimate This analysis was conducted for white perch, striped bass, spottail shiner, bay anchovy, American shad, alewife, blueback herring, and Atlantic tomcod.

B. Results The correlation analysis shows that that seven of the eight species of fishconsidered in this analysis are significantly and positively correlated (Table A-3). The correlation coefficients among the seven species range from 0.5 to 0.80. According to Figure A-i, the sample size of 20 in the present correlation analysis results in the power for the test ranging from about 60% to about 100%. Spottail shiner is the only species that does not show a significant correlation between the two indices. The lack of correlation is most likely attributable to large variation in the FSS data within individual years (Table A-2). The coefficient of variation for spottail shiner catch rates range between 0.17 and 1 in the FSS. Based on the overall results of the analysis, it can be concluded that species and life stages that share both habitats and are sampled by the two surveys exhibit the same interannual variation. This variation is reflected in the indices of the two surveys.

Ill. COMPARISON OF STRIPED BASS DATA WITH INDEPENDENT STUDIES This analysis examines the relationship between the BSS striped bass data with independent studies conducted by the NYSDEC, the ASMFC and the Hudson River Generators.

Striped bass is sampled in a beach seine survey conducted by the NYSDEC. This survey is conducted in the Tappan Zee and Croton-Haverstraw region of the Hudson River. This is an area where a large proportion of the young-of-year ("YOY") striped bass found in the Hudson River are located in late summer and fall. The BSS and the NYSDEC beach seine survey overlap in this area, but the BSS samples a much larger area of the Hudson River, ranging from near the mouth of the river to Troy Dam. The two surveys have run concurrently since 1982.

The size and the method of setting the beach seines vary between the two surveys. A correlation analysis was conducted to validate the use of the BSS in Appendix B.

The results from the NYSDEC beach seine survey are also used in the stock assessment of striped bass performed by the ASMFC (2005). An additional 61 age-specific and age-aggregated fishery-independent and fishery-dependent indices were used in the striped bass stock assessment (ASMFC 2005). A correlation analysis between the BSS and the coast-wide striped bass population abundance was conducted to show whether the Hudson River striped bass contribute significantly to the abundance of the coast-wide population.

Finally, the Hudson River Generators conducted a mark-recapture study of striped bass from 1984 through 1993. A correlation analysis was conducted to demonstrate the validity of the BSS when compared to these mark-recapture data.

2

A. Methods Input data for this analysis included the ASMFC 2005 striped bass stock assessment -

both total stock estimates as well as indices of abundance for different spawning regions, BSS YOY data, and striped bass mark-recapture data presented in the Draft Environmental Impact Statement ("DEIS") (Central Hudson Gas & Electric Corp. et al. 1999).

A linear regression was used to determine the fraction of the overall striped bass stock that could be attributed to the three major spawning stock regions: the Hudson River, the Delaware Estuary, and the Chesapeake Bay. The total estimated population of age-I striped bass, as reported in the 2005 stock assessment (Table A-4), was compared with the indices of abundance for New York, New Jersey, Maryland, and Virginia (Table A-5) ("Model 1"). The index of New York abundance used by ASMFC was based on NYSDEC sampling data. A second linear regression was developed using BSS YOY data (Table A-i) to represent the New York component of the stock ("Model 2").

A correlation analysis using a Pearson model was used to compare the NYSDEC index, the BSS index, mark-recapture data collected by the Hudson River Generators (Table A-6), the estimate of the New York portion of the striped bass stock based on NYSDEC data (Table A-7),

and the estimate of the New York portion of the striped bass stock based on BSS data (Table A-7).

B. Results The correlation analysis between the BSS and the NYSDEC beach seine survey results in a significant positive correlation (Table A-8). This demonstrates that the two independent surveys of young-of-year striped bass in the Hudson River produce similar annual results. BSS and the coast-wide population abundance of striped bass are also significantly positively correlated. This positive correlation is not surprising, as the NYSDEC beach seine survey is one of many input parameters used in the coast-wide stock assessment of striped bass (ASMFC 2005). It has already been established that the NYSDEC beach seine survey and the BSS are positively correlated (See Section 11.13). However, the results show that the many other input parameters in the striped bass stock assessment do not mask this relationship and confirm that striped bass associated with the Hudson River contribute significantly to the population dynamics of the coast-wide striped bass population. Another independent survey, a mark-recapture study, shows a significant linear relationship with the BSS. In summary, the BSS correlates significantly and positively with other existing independent surveys of striped bass YOY and older. This shows the robustness of the BSS in predicting young-of-year striped bass abundance.

IV. COMPARISON OF ATLANTIC TOMCOD DATASET WITH INDEPENDENT STUDIES The ATMR study in the Hudson River has been conducted for 22 years, starting in 1974 (Normandeau Associates, Inc. 2006). Abundance indices of 1 and 2 year old Atlantic tomcod are calculated, using data from the ATMR program (Table A-9). Yearly egg production estimates are also provided in Normandeau (2006).

3

Atlantic tomcod data from the BSS, FSS, and the LRS were compared with data from the mark-recapture study conducted by the Hudson River Generators to validate the results of the ATMR program by determining if correlations among the datasets exist.

A. Methods There were multiple inputs used to conduct further examinations of the Atlantic tomcod data used in earlier analyses. These data included the Atlantic tomcod index presented in Appendix A (based on mark-recapture surveys), BSS data, FSS data, and LRS data (Table A-10).

Two different statistical methods were used to examine the Atlantic tomcod data.

A correlation analysis, based on the Pearson model, was conducted comparing the mark-recapture data of age-i Atlantic tomcod with young-of-year BSS and FSS data.

A second correlation analysis, also based on the Pearson model, compared the estimated of eggs derived from the mark-recapture study with the post yolk-sac index based on LRS data.

B. Results The relative abundance of Atlantic tomcod based on the FSS is significantly and positively correlated with their abundance based on the BSS (Table A-i 1). The mark-recapture program for Atlantic tomcod also correlates positively and significantly to the FSS and the BSS.

The egg deposition is borderline positively correlated to the post yolk-sac larvae Atlantic tomcod estimated from the LRS (Table A-12).

4

V. LITERATURE CITED Atlantic States Marine Fisheries Commission 2005. 2005 Stock Assessment Report for Atlantic striped bass: Catch-at-age VPA & tag release/recovery based survival estimation. A report prepared by the striped bass technical committee for the Atlantic striped bass management board. Atlantic States Marine Fisheries Center.

Central Hudson Gas & Electric Corp. et al. 1999. Draft Environmental Impact Statement for State Pollutant Discharge Elimination System Permits for Bowline Point, Indian Point 2

& 3, and Roseton Steam Electric Generating Stations.

Normandeau Associates, Inc. 2006. Abundance and stock characteristics of the Atlantic tomcod spawning population in the Hudson River, winter 2003-2004. Prepared for Entergy Nuclear Operations, Inc.

5

VI. 'TABLES 6

Table 1. Abundance indices and associated standard errors, based on BSS.

WHITE PERCH STRIPED BASS SPOTTAIL SHINER BAY ANCHOVY AMERICAN SHAD ALEWIFE Year Young-of-Year Youn-f-Year Young-of-Year Young-of-Year Young-of-Year Young-of-Year Index SE Index. SE Index SE Index SE Index SE Index SE 1974 566,346 61,280 1,373,138 264,598 658,945 87,448 2,999,066 973,844 2,123,265 232,509 583,238 74,805 1975 2,342,937 440,999 1,367,496 242,374 1,286,297 193,361 5,159,511 1,666,189 1,998,286 161,394 572,550 107,585 1976 1,944,220 255,910 864,743 70,734 1,324,434 203,989 .5,234,482 2,595,405 2,354,807 125,450 352,263 96,375 1977 953,799 87,722 1,375,537 124,595 495,690 66,445 4,616,994 875,014 2,123,707 114,152 517,792 49,081.

1978 2,675,700 402,374 3,042,920 614,048 1,363,313 148,541 329,478 57,321 4,021,203 251,047 1,027,891 174,698 1979 2,921,393 285,862 794,022 91,389 956,236 97,330 1,860,753 686,496 1,934,405 107,064 340,271 59,099 1980 1,884,895 231,650 1,265,254 147,121 633,323 72,196 3,445,878 818,900 1,632,041 117,820 93,783 17,894 1981 1,862,222 160,903 1,827,767 152,481 1,865,058 216,442 4,505,689 1,862,587 2,558,539 149,238 477,348 84,403 1982 1,967,754 287,490 934,550 97,768 477,090 62,605 2,740,240 1,735,314 1,768,839 150,312 116,606 24,817 1983 1,803,266 399,823 1,642,536 191,103 1,070,822 104,909 364,403 243,354 2,452,068 183,820 214,922 42,154 1984 703,959 145,133 1,300,754 173,872 616,182 128,367 1,887,240 963,767 1,060,902 74,374 49,776 10,864 1985 757,003 82,536 238,259 21,226 543,246 66,532 ,621,718 203,675 1,263,843 153,248 119,509 22,024 1986 1,036,321 97,303 298,745 31,415 388,736 69,297 975,435 779,300 2,207,907 125,447 119,468 48,899 1987 1,169,236 121,876 2,976,381 314,807 470,267 74,827 830,978 229,609 1,482,041 125,017 80,611 13,768 1988 1,738,310 255,364 1,172,303 68,239 419,874 49,588 546,894 225,975 997,414 59,920 87,080 15,727 1989 1,105,280 278,101 1,238,434 116,464 623,204 95,526 2,840,186 987,471 2,455,819 135,247 43,711 12,956 1990 588,162 75,727 1,486,911 89,409 808,662 101,694 208,541 65,810 2,004,620 162,122 157,159 25,580 1991 580,165 76,201 1,125,126 64,076 855,292 110,557 935,366 246,296 1,499,227 120,544 335,535 63,111 1992 463,555 53,444 1,046,654 53,265- 726,888 124,009 1,629,973 1,184,246 1,886,715 101,469 40,507 9,371 1993 806,848 97,157 1,640,132 90,969 655,117 95,425 1,183,278 462,699 815,539 68,698 69,438 11,826 1994 315,662 39,618 1,136,106 63,179 1,624,997 289,784 2,255,731 478,603 1,963,731 124,116 148,030 30,079 1995 425,062 49,042 1,404,935 89,202 603,130 94,204 2,507,280 721,809 552,490 48,911 91,731 22,716 1996 44,925 10,283 299,997 30,506 174,026 39,053 720,000 151,968 1,743,007 125,007 47,371 14,912 1997 571,160 114,812 1,892,597 169,399 1,197,799 170,583 3,496,618 815,723 1,573,674 106,235 291,323 54,177 1998 270,835 51,992 1,384,364 85,327 273,165 53,055 2,675,549 670,172 319,702 47,834 40,865 30,194 1999 1,411,184 169,447 1,715,282 142,568 2,040,399 243,244 858,192 298,574 1,399,557 107,459 445,167 79,622 2000 304,950 52,787 580,006 52,449 303,081 52,956 769,133 427,827 941,909 105,935 76,445 37,606 2001 1,019,516 119,666 2,392,216 170,860 2,143,066 610,761 613,810 401,115 2,479,221 176,132 330,876 70,451 2002 699,145 80,612 1,145,686 60,295 1,132,479 146,862 3,826,181 1,061,795 721,680 72,203 60,954 13,491 2003 2,177,013 228,303 2,282,684 118,276 2,102,568 257,006 1,703,952 451,911 1,071,881 69,880 452,292. 87,223 2004 632,961 89,075 807,661 70,743 1,031,399 152,802 404,497 145,762 444,880 131,585 218,118 35,902 7

Table A-I. Abundance indices and associated standard errors, based on BSS (continued).

BLUEBACK HERRING ATLANTIC TOMCOD Year - Young-of-Year Young-of-Year Index SE Index SE 1974 3,647,758 502,857 18,536 4,046 1975 10,888,524 1,249,788 39,688 11,253 1976 21,621,271 3,075,761 41,196 12,039 1977 31,795,371 4,717,652 8,178 -2,802 1978 22,993,451 4,200,939 37,401 11,147 1979 8,221,314 1,461,758 58,632 18,283 1980 8,892,467 2,207,337 17,337 6,016 1981 32,066,440 9,586,015 3,698 1,141 1982 10,164,307 1,750,817 70,051 14,120 1983 16,326,879 2,278,723 11,419 3,218 1984 3,577,323 786,742 50,486 12,104 1985 3,323,511 664,762 34,760 6,246 1986 1,555,182 357,032 28,125 5,369 1987 6,188,101 773,111 35,074 8,600 1988 5,887,963 1,008,925 21,020 5,249 1989 3,230,116 497,839 12,946 3,825 1990 9,436,487 1,274,900 16,941 5,709 1991 3,530,392 596,059 4,417 1,849 1992 6,642,282 1,599,250 43,740 10,403 1993 4,234,168 531,496 2,144 913 1994 9,584,696 1,308,960 1,198 579 1995 3,202,735 892,613 0 0 1996 4,044,353 890,186 9,182 5,836 1997 12,075,530 2,541,612 5,053 1,572 1998 155,761 32,365 1,384 616 1999 5,691,570 776,702 0 0 2000 2,342,499 572,561 9,823 3,892 2001 5,268,663 704,402 1,520 752 2002 1,438,577 299,230 0 0 2003 10,203,281 1,459,824 0 0 2004 5,091,421 620,888 5,928 1,647 8

Table A-2. Abundance indices and associated standard errors, based on FSS.

WHITE PERCH STRIPED BASS SPOTTAIL SHINER BAY ANCHOVY AMERICAN SHAD ALEWIFE year Young-of-Year Young-of-Year Young-of-Year Young -f-Year Young-of-Year Youn -f-Year

___ Index SE Index SE Index SE Index SE Index . SE Index SE 1985 1,685,851 165,213 164,284 16,636 85,977 39,236 218,612,898 21,269,766 1,591,435 190,139 2,105,489 381,844 1986 1,759,522 207,644 651,049 49,859 49,745 11,399 132,925,173 13,133,411 3,104,605 640,844 595,155 115,129 1987 1,579,037 136,932 4,889,589 239,032 20,977 5,401 246,910,112 26,982,497 647,070 157,299 695,124 245,872 1988 3,777,521 297,018 9,569,544 497,548 83,429 20,121 422,678,791 38,213,532 997,871 144,252 624,702 142,344 1989 3,167,143 357,848 4,235,166 333,577 3,591 1,550 349,952,337 26,107,654 2,754,815 198,752 505,822 105,987

.1990 548,583 167,722 2,883,805 200,426 17,347 5,614 161,039,442 14,450,450 1,139,272 235,276 807,620 138,564 1991 443,688 67,292 1,138,102 87,685 131,938 34,430 190,474,265 11,540,891 680,209 72,781 685,242 104,724 1992 1,064,922 136,793 1,186,233 113,756 23,041 8,964 185,902,303 13,738,226 1,306,732 147,744 746,514 158,432 1993 415,097 100,885 2,779,357 178,004 70,379 17,018 249,913,241 19,475,645 464,702 48,446 530,240 83,846 1994 566,404 53,440 3,439,449 209,768 34,772 5,983 206,642,043 14,141,476 1,036,782 88,932 571,174 82,018 1995 1,514,550 230,289 2,878,188 173,061 110,530 3,570 439,617,793 28,732,239 471,444 75,896 308,139 49,342 1996 414,924 60,068 2,396,874 172,968 73,863 15,117 102,941,191 5,959,974 2,859,373 451,439 1,076,096 124,312 1997 539,792 86,123 2,439,137 273,488 6,312 2,846 1283,382,412 17,014,202 913,970 107,851 1,233,697 154,951 1998 357,696 35,390 580,977 65,746 2,367 2,367 189,541,611 9,166,785 232,260 56,459 112,261 28,629 1999 2,021,946 166,188 2,655,600 220,747 25,220 5,712 165,375,818 9,972,244 853,411 135,639 2,543,734 197,641 2000 433,794 60,439 1,634,254 228,331 12,010 1,496 57,208,944 3,577,181 878,405 100,807 913,399 108,152 2001 869,631 93,161 1,184,609 105,581 20,724 9,574- 109,701,139 8,052,515 1,006,787 162,014 2,253,572 652,056 2002 401,209 46,026 982,555 156,264 14,619 4,774 171,692,430 10,652,063 497,537 57,524 255,519 37,190 2003 2,181,001 165,766 4,787,259 432,818 938 841 148,898,706 111,753,477 351,278 47,131 1941,836 102,643 2004 543,243 1159,067 991,181 119,540 40,935 8,459 218,178,981 117,899,774 336,973 63,105 1249,944 43,269 9

Table A-2. Abundance indices and associated standard errors, based on FSS (continued).

BLUEBACK HERRING ATLANTIC TOMCOD Young-of-Year Young-of-Year year Index SE Index SE 1985 63,437,557 9,471,265 3,818,562 537,609 1986 15,577,561 2,395,825 6,935,212 588,195 1987 38,342,783 9,373,512 3,431,206 257,718 1988 61,946,416 6,136,684 3,731,674 370,666 1989 33,621,840 3,107,711 13,006,674 1,862,570 1990 63,121,526 6,836,956 1,377,747 247,070 1991 43,421,773 5,346,974 263,792 37,402 1992 46,987,241 6,744,931 3,846,993 297,928 1993 20,223,194 1,817,165 3,742,238 1,013,814 1994 17,568,127 1,521,183 604,300 55,493 1995 14,114,745 1,634,192 84,328 16,082 1996 67,981,601 8,013,906 3,543,737 380,726 1997 29,241,071 3,323,567 2,392,903 208,967 1998 927,634 153,551 507,900 73,503 1999 22,609,332 2,329,531 19,312 6,888 2000 11,400,882 1,150,959 2,262,871 196,166 2001 23,294,104 4,713,494 897,887 240,836 2002 10,219,873 969,053 80,565 17,597 2003 17,724,162 1,789,797 355,046 74,484 2004 6,347,406 606,675 2,100,531 318,419 10

Table A-3. Correlations between BSS and FSS data Number Inverse-Variance Significance Taxa of Weighted Level Years Correlation Factors White Perch 20 0.69 0.0007 Striped Bass 20 0.69 0.0008 Spottail Shiner 20 -0.09 0.6969 Bay Anchovy 20 0.55 0.0122 American Shad 20 0.76 <0.0001 Alewife 20 0.50 0.0235 Blueback Herring 20 0.73 0.0002 Atlantic Tomcod 20 0.80 <0.0001 11

Table A-4. Estimated age-I striped bass population.

Striped Bass Age-I Population Year (thousands) 1982 1,534 1983 3,181 1984 2,401 1985 3,579 1986 2,763 1987 3,944 1988 5,219 1989 5609 1990 8,419 1991 8,644 1992 8,706 1993 11,065 1994 16,562 1995 13,338 1996 12,932 1997 15,586 1998 10,625 1999 10,982 2000 8,261 2001 15,490 2002 18,024 2003 5,976 2004 22,275 2005 12,721 Source: ASMFC 2005 12

Table A-5. Indices of abundance for Atlantic striped bass adjusted to January 1st Year Young-of-Year Young-of-Year Young-of-Year Young-of-Year New York Index New Jersey Index Maryland Index Virginia Index 1982 8.86 0.59 1.56 1983 14.17 0.12 3.57 2.71 1984 16.25 0.03 0.61 3.4 1985) 15 0.29 1.64 4.47 1986 1.92 0.18 0.91 2.41 1987 2.92 0.28 1.34 4.74 1988 15.9 0.41 1.46 15.74 1989 33.46 0.35 0.73 7.64-1990 21.35 1.03 4.87 11.23 1991 19.08 1 1.03 7.34 1992 3.6 0.47 1.52 3.76 1993 11.43 1.19 2.34 7.35 1994 12.59 1.78 13.97 18.11 1995 17.64 0.96 6.4 10.48 1996 16.23 1.98 4.41 5.45 1997 8.93 1.7 17.61 23 1998 22.3 1.01 3.91 9.35 1999 13.39 1.31 5.5 13.25 2000 26.64 1.9 5.34 2.8 2001 3.16 1.77 7.42 16.18 2002 22.98 1.07 12.57 14.17 2003 12.32 0.51 2.2 3.98 2004 17.36 2.43 10.83 22.89 2005 8.81 1.13 4.85 12.7 Source: ASMFC 2005 13

Table A-6. Abundance estimate of Hudson River striped bass, based on mark-recapture data.

Age-2+

Year Abundance 1984 213 1985 104 1986 108 1987 611 1988 560 1989 339 1990 344 1991 502 1992 238 1993 201 Source: Central Hudson Gas & Electric Corp. et al. 1999 14

Table A-7. Estimate of NY striped bass stock, based on NYSDEC and BSS data.

Estimate of Hudson River Year age-i striped bass Based on NYSDEG Data Based on BSS data 1974 1,510,636 1975 ____________ 1,504,429 1976 951,333 1977 1,513,275 1978 3,347,621 1979 ___________ 873,531 1980 1,391,949 1981 560,788 2,010,789 1982 896,882 1,028,131 1983 1,028,534 1,807,010 1984 949,416 1,431,004 1985 121,525 262,117 1986 184,820 328,660 1987 1,006,381 3,274,419 1988 2,117,831 __ 1,289,691 1989 1,351,336 1,362,444 1990 1,207,657 1,635,802 1991 227,860 1,237,790 1992 723,455 1,151,460 1993 796,877 1,804,365 1994 1,116,513 1,249,869 1995 1,027,268 1,545,617 1996 565,219 330,037 1997 1,411,465 2,082,111 1998 847,512 1,522,986 1999 1,686,163 1,887,041 2000 200,010 638,085 2001 1,454,505 2,631,759 2002 779,787 1,260,408 2003 1,098,791 2,511,259 2004 557,624 888,536 15

Table A-8. Striped Bass correlation coefficients New York Index BSS Index Mark-recapture New York Stock age-2 Abundance (based on NYDEC data)

New York Index ________ 0.53 0.55 1.00 BSS Index 0.53 0.68 0.53 Mark-recapture age-2 0.55 0.68 0.55 New York Stock (based on NYDEC data)- 1.00 0.53 0.55 New York Index BSS Index Mark-recapture New York Stock age-2 Abundance (based on BSS data)

New York Index _ 0.53 0.55 0.53 BSS Index 0.53 0.68 1.00 Mark-recapture age-2 0.55 0.68 0.68 New York Stock (based on BSS data) 0.53 1.00 0.68 Note: Correlation coefficients significant at the 10% level are shown.

Correlation coefficients significant at the 5% level are shown in bold.

16

Table A-9. Atlantic tomcod mark-recapture data Year Proportion Age-I Proportion Age-2 Population Egg Deposition Population Age-I (billions) (millions) 1975 3.6 1976 0.98 0.02 22 9.7 1977 0.933 0.067 65 2.4 1978 0.965 0.035 21 5.9 1979 0.989 0.01 51 8.8 1980 0.97 0.03 57 1981 0.943 0.056 1982 0.968 0.032 10.5 1983 0.843 0.155 97 5.9 1984 0.887 0.113 75 1985 .. 2 1986 0.957 0.043 25 1987 2.9 1988 0.837 0.163 43 5.3 1989 0.9 0.1 41 4.9 1990 0.715 0.285 87 2.6 1991 0.81 0.19 52 0.3 1992 0.715 0.285 7 2.2 1993 0.849 0.151 30 0.5 1994 0.662 0.338 7 2.2 1995 0.907 0.093 31 1996 0.483 0.517" 2.6 1997 0.8 0.2 47 0.7 1998 0.535 0.465 23 0.4 1999 0.664 0.336 10 0.2 2000 0.799 0.201 3 2.3 2001 0.935 0.065 28 2002 0.827 0.173 2003 .0.95 0.05 1.6 2004 0.952 0.048 28 Source: Normandeau Associates, Inc. 2006 17

Table A- 1. Atlantic Tomcod abundance index and associated standard errors, based on LRS ATLANTIC TOMCOD Post Yolk-Sac Larvae year Index SE 1974 128,306,743 19,426,263 1975 67,024,707 19,768,962 1976 42,777,042 13,470,065 1977 164,621,663 70,515,234 1978 54,313,088 10,307,482 1979 18,127,435 3,099,375 1980 95,402,234 13,128,146 1981 74,140,778 13,052,007 1982 28,419,800 7,665,326_

1983 42,683,202 8,311,722 1984 147,133,069 25,916,525 1985 109,664,584 11,132,251 1986 53,404,268 4,770,519 1987 138,570,516 12,594,732 1988 78,376,300 10,680,903 1989 185,450,859 23,858,579 1990 107,915,374 25,158,013 1991 116,333,462 14,859,973 1992 32,021,214 4,889,565 1993 126,394,886 20,139,893 1994 85,456,373 22,227,930 1995 79,816,881 6,641,688 1996 51,571,386 5,696,759 1997 110,409,961 28,829,551 1998 53,594,909 8,409,591 1999 17,392,702 2,076,588 2000 11,120,807 1,442,773 2001 93,816,691 8,320,053_

2002 4;3 82,650 649,979 2003 38,715,789 3,683,762 2004 115,401,578 16,005,570 18

Table A- 11. Atlantic tomcod correlation coefficients Age-I Young-of-year: Young-of-year:

Mark-recapture data BSS data FSS data Age-i: mark-recapture data __0.77 0.65 Young-of-year: BSS data 0.77 0.45 Young-of-year: FSS data 0.65 0.45 Note: Only correlation coefficients significant at the 10% level are shown.

Correlation coefficients significant at the 5% level are shown in bold.

Table A-12. Atlantic tomcod correlation coefficients Eggs: Post yolk-sac:

Mark-recapture data LRS data Eggs: mark-recapture data 0.41 Post yolk-sac: LRS data 0.41 Note: Only correlation coefficients significant at the 10% level are shown.

Correlation coefficients significant at the 5% level are shown in bold.

19

VII. FIGURES 20

Figure A- 1.

Power for Tests of Pearson Correlation a=0.05 two-sided

~,

0.9 'A / /

.0.8 0.7 0.6 0 0.5 0.4 0.3 0.2 0.1 0--

-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1 Correlation (rho)

-- N=10 -- N=20 - "N=30 ... N=40 21

APPENDIX B Prepared by:

AKRF, Inc.

TABLE OF CONTENTS

1. I1NTJRODUJCT.ION? .......................................................1

11. IN ICEJ................................................................. 1 A. FISH POPULATION ABUNDANCE ...................................................................... 1 B3. STRESSORS OF FISH POPULATIONS................................................................... 3

1. Power PlantMortality ........................................................................ 3

2. Zebra Mussels..................... ............................................................ 4

37. Striped Bass Predation......................................................................... 4

4. Temperature.................................................................................... 4 C. FISH POPULATION RESPONSE METRICS.............................................................. 5

1. Survival Indices................................................................................ 5

2. Abundance Indices............................................................................. 6

3. Growth Indices................................................................................. 6

4. Spatial DistributionIndices ................................................................... 6 I11. CORRELATION ANALYSES................................................................ 7 IV. LITERATURE CITED......................................................................... 9 V7. TABLES................................................... I...................................... 12

I. INTRODUCTION This Appendix documents the methods and data used in: (1) analyses of trends in fish population abundance; and (2) correlation analyses to address impact hypotheses. The rationale for and the results from the analyses of trends and the correlation analyses are discussed in the report titled: "Entrainment and Impingement at IP2 and IP3: A Biological Impact Assessment."

The analyses of trends in fish population abundance and the correlation analyses were based on indices developed from data collected by the Hudson River Generators' Longitudinal River Ichthyoplankton Survey ("LRS"), Beach Seine Survey ("BSS"), Fall Shoals Survey

("FSS"), and Atlantic Tomcod Mark-Recapture ("ATMR") Program. Three types of indices were defined for these analyses:

• indices of fish population abundance;

• indices of stressors of fish populations; and

• indices of fish population response to stressors.

The remainder of this Appendix is organized in three Sections. The first Section documents the three types of indices; the second Section documents the trend analysis methods and results; and the third Section documents the correlation analysis methods and results.

II. INDICES A. Fish Population Abundance Annual indices of fish population abundance were computed as the average of the weekly standing crop estimates presented in the Year Class Report for the Multiplant Impact Study of the Hudson River Estuary for the years 1974 through 1979 and the Hudson River Estuary Monitoring Program for the years 1980 through 2004 (collectively, ("Year Class Report")

(Applied Science Associates, Inc. 2000, 2001; ASA Analysis & Communication, Inc. 2001, 2002, 2003, 2004a, 2004b, 2005, 2006; Batelle New England Marine Research Laboratory 1983; Consolidated Edison Company of New York, Inc. 1996, 1997a, 1997b; EA Engineering, Science, and Technology 1990, 1991, 1996; Lawler, Matusky & Skelly Engineers 1989, 1992, 1996; Martin Marietta Environmental Systems 1986; Normandeau Associates, Inc. 1985a, 1985b; Texas Instruments, Inc. 1977, 1978, 1979, 1980a, 1980b, 1981; Versar, Inc. 1987). A separate annual index value was computed for each species and life stage. Indices of abundance for age- 1 and age-2 Atlantic tomcod and abundance of Atlantic tomcod eggs were based on abundance estimates from the ATMR Program (Normandeau Associates, Inc. 2006).

Weekly standing crop estimates for post yolk-sac larvae ("PYSL") were based on data collected by the LRS. Weekly standing crop estimates for young-of-year' ("YOY") fish inhabiting the beach zone of the Hudson River were based on data collected by the BSS. Weekly standing crop estimates for YOY fish inhabiting the shoals, bottom, and channel of the Hudson River were based on data collected by the FSS. These standing crop estimates, with associated standard errors, were provided in electronic format by ASA Analysis & Communication, Inc.

Young-of-year fish are sometimes also referred to as juvenile fish.

I

("ASA"). Data collection methods for the LRS, BSS, and FSS, and methods for estimating weekly standing crops (and associated standard errors) are documented in the Year Class Reports. Annual estimates of the number of age-1 and age-2 Atlantic tomcod and the number of Atlantic tomcod eggs spawned were developed by the ATMR program, and were provided by Normandeau Associates, Inc. ("NAI"). Data collection methods for the ATMR program and methods for estimating Atlantic tomcod abundance are documented in annual ATMR Program Reports prepared by NAI for the Hudson River Generators. In addition, estimates of the,'

variance of the estimate of the total number of age-1 and age-2 Atlantic tomcod were computed, as described below.

A set of regions and weeks that were consistently sampled among years was identified for each sampling program. Annual abundance indices based on LRS data were computed for 1974 through 2004, based on data from regions 1 through 12, and weeks 18 through 26. Annual abundance indices based on BSS data were computed for 1974 through 2004, based on data from regions 1 through 12, and weeks 31 through 42. Annual abundance indices based on FSS data were computed for 1979 through 2004, based on data from regions 1 through 12, and weeks 31 through 42. Data from the ATMR program were included for all years (1974 through 2004) in which the number of recaptured Atlantic tomcod exceeded one fish.

BSS data were used to develop YOY abundance indices for alewife, blueback herring, spottail shiner, striped bass, and white perch. FSS data were used to develop YOY abundance indices for American shad and bay anchovy. LRS data were used to develop the PYSL indices for striped bass, white perch, river herring (which included alewife, blueback herring and unidentified clupeids - three taxonomic groups that could not reliably be identified to species as PYSL), American shad, and bay anchovy. The LRS did not adequately sample areas of the river inhabited by spottail shiner larvae. To address the abundance of early life stages of spottail shiner, an index of egg abundance was developed based on spawning age spottail shiner (i.e.,

yearling and older) sampled by the BSS. The index of yearling and older spottail shiner was used as a surrogate index for spottail shiner egg abundance.

For each species, sampling program (LRS, BSS, and FSS), and year, the annual index of abundance (Ay) was computed using the following formula:

Ay Wmax [ CW°,Y 2,:Wmn(5J WW where 12 SCW, = SCRW, y R=I Wmi,, = first week of the season, 2

Wmax = last week of the season, SCRWY = estimated standing crop in region R, week Wand yeary, ky - = 1 if all 12 standard regions were sampled in week W of year y, and gWy = 0 otherwise.

For Atlantic tomcod, approximately unbiased Peterson-type mark-recapture estimates of abundance were computed as (Seber 1982):

y = (C' + 1I(My+1)

(my +1) 1 and the variance of the estimated abundance was estimated as (Seber 1982):

v() (CY + IlMY + Cy- myXM, - my)

(my + 1)2(my + 2) where Cy = number of fish (marked and unmarked) caught subsequent to marking, My = number of fish marked, and my = number of marked fish recaptured.

The abundance indices are presented in Tables B-1 through B-3.

B. Stressors of Fish Populations Four potential stressors of fish populations in the Hudson River estuary were identified:

(1) power plant mortality due to entrainment at Indian Point; (2) effects of the zebra mussel invasion on the Hudson River biota; (3) predation by increased abundance of striped bass in the Hudson River estuary; and (4) elevated late summer and fall bottom temperatures. For each stressor, an index was developed that was intended to track the intensity of the stressor.

1. Power Plant Mortality The index of entrainment mortality at Indian Point was the conditional mortality rate ("CMR"). An annual CMR for entrainment can be interpreted as the fractional reduction in age-I abundance of a year class of fish due to the effects of entrainment, assuming the absence of density-dependent mortality. Estimates of CMRs for entrainment at Indian Point from 1974 3

t through 1997 were taken from the Draft Environmental Impact Statement ("DEIS") for State Pollution Discharge Elimination System Permits for Bowline Point, Indian Point 2 & 3, and Roseton Steam Electric Generating Stations (Central Hudson Gas & Electric Corp. et al. 1999).

CMR estimates for entrainment at Indian Point for 1998 through 2003 were computed for this analysis using the same methods documented in the DEIS. CMR estimates were computed separately for striped bass, white perch, American shad, bay anchovy, spottail shiner, Atlantic tomcod, and river herring.

The indices of entrainment mortality are listed in Table B-4.

2. Zebra Mussels The invasive zebra mussel (Dreissenapolymorpha) first appeared in the Hudson in 1991 and became a dominant species in the Hudson River by September 1992 (Strayer et al. 1996).

Strayer et al. (2004) reported that "(z)ebra mussels were quantitatively important only in freshwater parts of the Hudson, and their effects extend from the head of the estuary (rkm 248) down to approximately rkm 100 (Strayer et al. 1996; Caraco et al. 1997; Pace et al. 1998)."

Based on this characterization, the indicator variable for zebra mussel effects was set to zero (i.e., no effect) for the period 1974 through 1992, and was set to one (i.e., effect was present) for the years 1993 through 2004. Also, an index of the spatial distribution of fish within the Hudson River was defined (see Section II.C.4, below), based on the relative abundance of fish downriver ofrkm 100.

The index of zebra mussel effects is listed in Table B-5.

3. Striped Bass Predation The index of striped bass predation was intended to represent the predatory pressure of adult striped bass on the fish community of the Hudson River estuary. Post yolk-sac larvae abundance was used as a surrogate for adult abundance under the assumption that PYSL abundance represented reproductive potential which, in turn was roughly proportional to spawning abundance. Accordingly, the striped bass PYSL abundance index based on the LRS was used as the index of striped bass predation.

The index of striped bass predation is listed in Table B-6.

4. Temperature For all species except Atlantic tomcod, the index of water temperature was based on water temperature in the bottom stratum of the river and was computed in two steps. First, a riverwide average temperature for each week within a season was computed. The weekly average value was computed as the weighted average, where the weighting factor for each region (1 through 12) was the volume of the bottom stratum in the region. The second step was to average the weekly values over all weeks (in which all 12 standard regions were sampled) within the season.

4

For Atlantic tomcod, an alternative index of water temperature was computed: a degree-day index based on data recorded at the Poughkeepsie Water Works ("PWW"). The annual PWW degree-day index was computed as the sum (January through December) of daily temperatures above 24'C. Days with water temperatures below 24 0 C did not contribute to the annual sum. The temperature of 24 0 C was chosen because growth in age-0 Atlantic tomcod from the Hudson River slows when water temperatures exceeded 20'C and ceased when water temperatures exceeded 24"C (Chambers and Witting, 2005).

The indices of water temperature are listed in Table B-7.

C. Fish Population Response Metrics

1. Survival Indices Each survival index was defined as a ratio of abundance indices from two life stages: the denominator of the ratio was the earlier life stage and the numerator was a subsequent life stage.

Therefore, the ratio was proportional to the fraction of the earlier life stage that survived to the subsequent life stage. Because the methods and data used for the abundance indices (see Section II.A, above) are species-specific, the definitions of the survival indices are also species-specific.

" The survival index for striped bass from PYSL to YOY was defined as the ratio of the YOY abundance index (based on BSS data) to the PYSL abundance index (based on LRS data).

" The survival index for white perch from PYSL to YOY was defined as the ratio of the YOY abundance index (based on BSS data) to the PYSL abundance index (based on LRS data).

" The survival index for alewife from PYSL to YOY was defined as the ratio of the alewife YOY abundance index (based on BSS data) to the river herring YOY abundance index (based on LRS data).

" The survival index for American shad from PYSL to YOY was defined as the ratio of the YOY abundance index (based on FSS data) to the PYSL abundance index (based on LRS data).

" The survival index for bay anchovy from PYSL to YOY was defined as the ratio of the YOY abundance index (based on FSS data) to the PYSL abundance index (based on LRS data).

• The survival index for spottail shiner from eggs to YOY was defined as the ratio of the YOY abundance index (based on BSS data) to the egg abundance index (based on BSS data).

• The survival index for Atlantic tomcod from age-1 to age-2 was defined as the ratio of the age-2 abundance index (based on ATMR data) to the age-I abundance index (based on ATMR data).

5

The survival index for Atlantic tomcod from eggs to age-I was defined as the ratio of the egg abundance index (based on ATMR data) to the age-I abundance index (based on ATMR data).

The survival indices are listed in Table B-8.

2. Abundance Indices Because some stressors can act directly on the abundance of certain life stages, the abundance indices listed in Tables B-1 through B-3 were also used as response metrics.

3. Growth Indices The growth index was intended to represent the relative amount of growth in juvenile fish that occurred during a standard set of weeks (31 through 42) in the fall of each year. Annual growth indices (1979 through 2004) were computed from BSS and FSS data.

The growth index for each species and year was computed in three steps. First, the average fish length was calculated for each week and region. Then, a weighted average length was computed for each week, where the weight for each region was the YOY abundances in the region. The third step was to conduct a log-linear regression analysis of the weighted-average length (Lw ) against week number (W):

kWk Wnx xep(W-Wi.).

The slope estimate (/) from that regression analysis represented the average growth rate during the fall season, and was used as the index of growth for the species in that year.

The growth indices are listed in Table B-9.

4. Spatial Distribution Indices This index was intended to address the possible effects of zebra mussels on fish distribution patterns, and was defined as the portion of the total population that occurred downstream of rkm 100.

For American shad and bay anchovy, the spatial distribution indices for YOY were based on data from the FSS for weeks 31 through 42. For striped bass, white perch, blueback herring, alewife and spottail shiner, the spatial distribution indices for YOY were based on data from the BSS for weeks 31 through 42. The spatial distribution indices for PYSL were computed for striped bass, white perch, bay anchovy, American shad, river herring, and Atlantic tomcod based on data from the LRS from weeks 18 through 26. For Atlantic tomcod, which spawn in late winter/early spring, data from the LRS included juveniles in addition to PYSL. Annual spatial indices based on LRS data were computed for 1974 through 2004. Annual spatial indices based on BSS data were computed for 1974 through 2004. Annual spatial indices based on FSS data were computed for 1979 through 2004.

6

For each species, life stage, region (R), and year, the fraction of the riverwide abundance inhabiting areas within the region or downriver of the region (FR,Y) was estimated using the following formula:

FR,y - 1 g where SCry 1 ZSC,.,Wy xJWy Z*c~,y W =W.W.

SW=Wmi,,

The upper boundary of Region 6 is between rkm 99 and rkm 100. Therefore, the index of spatial distribution was defined as *.y The spatial distribution indices are listed in Table B-I 0.

Ill. CORRELATION ANALYSES A correlation analysis was conducted to identify significant correlations between (1) stressor indices and (2) indices of fish population response metrics. For each stressor, a set of relevant response variables was selected based on impact hypotheses and life history considerations. For example, zebra mussel effects were paired with the proportion of a population downriver of rkm 100, and temperature was paired with juvenile growth rate.

A correlation analysis was also conducted to identify significant correlations between (1) abundance indices and (2) indices of fish population response metrics. Relevant combinations of abundance and response metrics were selected based on impact hypotheses and life history considerations.

The correlation analyses were conducted using Spearman (rank) correlation coefficients to account for possible non-Normality of the indices. The correlation analyses were based on annual index values and were conducted separately for each species.

Results from the correlation analyses are summarized in Tables B-I 1 through B-26.

Correlation coefficients significant atthe 0.05 level are printed in black and correlation coefficients significant at the 0. 10 level are printed in gray. A blank cell in the table indicates that the correlation coefficient was not significant at a probability level of 0. 10 or lower. Cells 7

shaded gray indicate pairs of indices that were not considered relevant, based on impact hypothesis and/or life history considerations.

8

IV. LITERATURE CITED Applied Science Associates, Inc. 2000. 1996 Year class report for the Hudson River estuary monitoring program ("Year Class Report"). Prepared for Central Hudson Gas & Electric Corporation.

Applied Science Associates, Inc. 2001. 1997 Year Class Report. Prepared for Central Hudson Gas & Electric Corporation.

ASA Analysis & Communication, Inc. 2001. 1998 Year Class Report. Prepared for Central Hudson Gas & Electric Corporation.

ASA Analysis & Communication, Inc. 2002. 1999 Year Class Report. Prepared for Dynegy Roseton L.L.C., Entergy Nuclear Indian Point 2 L.L.C., Entergy Nuclear Indian Point 3 L.L.C., and Mirant Bowline L.L.C.

ASA Analysis & Communication, Inc. 2003. 2000 Year Class Report. Prepared for Dynegy Roseton L.L.C., Entergy Nuclear Indian Point 2 L.L.C., Entergy Nuclear Indian Point 3 L.L.C., and Mirant Bowline L.L.C.

ASA Analysis & Communication, Inc. 2004a. 2001 Year Class Report. Prepared for Dynegy Roseton L.L.C., Entergy Nuclear Indian Point 2 L.L.C., Entergy Nuclear Indian Point 3 L.L.C., and Mirant Bowline L.L.C.

ASA Analysis & Communication, Inc. 2004b. 2002 Year Class Report. Prepared for Dynegy Roseton L.L.C., Entergy Nuclear Indian Point 2 L.L.C., Entergy Nuclear Indian Point 3 L.L.C., and Mirant Bowline L.L.C.

ASA Analysis & Communication, Inc. 2005. 2003 Year Class Report. Prepared for Dynegy Roseton L.L.C., Entergy Nuclear Indian Point 2 L.L.C., Entergy Nuclear Indian Point 3 L.L.C., and Mirant Bowline L.L.C.

ASA Analysis & Communication, Inc. 2006. 2004 Year Class Report. Prepared for Dynegy Roseton L.L.C., Entergy Nuclear Indian Point 2 L.L.C., Entergy Nuclear Indian Point 3 L.L.C., and Mirant Bowline L.L.C.

Batelle New England Marine Research Laboratory. 1983. 1980 and 1981 Year Class Report.

Prepared for Consolidated Edison Company of New York, Inc.

Caraco, N.F., Cole, J.J., Raymond, P.A., Strayer, D.L., Pace, M.L., Findlay, S.E.G., and Fischer, D.T. 1997. Zebra mussel invasion in a large, turbid river: phytoplankton response to increased grazing. Ecology 78: 588-602.

Central Hudson Gas & Electric Corp. et al. 1999. Draft Environmental Impact Statement for State Pollutant Discharge Elimination System Permits for Bowline Point, Indian Point 2

& 3, and Roseton Steam Electric Generating Stations.

Chambers, C and D Witting. 2005. Identifying environmental constraints on growth and survival of an anadromous gadid, Atlantic tomcod, at southern extreme of its geographic range.

Ecological Society of America. 2005 Annual Meeting.

Consolidated Edison Company of New York, Inc. 1996. 1992 Year Class Report. New York, New York.

9

Consolidated Edison Company of New York, Inc. 1997a. 1993 Year Class Report. New York, New York.

Consolidated Edison Company of New York, Inc. 1997b. 1994 Year Class Report. New York, New York.

EA Engineering, Science, and Technology. 1990. 1988 Year Class Report. Prepared for Consolidated Edison Company of New York, Inc.

EA Engineering, Science, and Technology. 1991. 1989 Year Class Report. Prepared for Consolidated Edison Company of New York, Inc.

EA Engineering, Science, and Technology. 1996. 1995 Year Class Report. Prepared for Consolidated Edison Company of New York, Inc.

Lawler, Matusky & Skelly Engineers. 1989. 1986 and 1987 Year Class Report. Prepared for Consolidated Edison Company of New York, Inc.

Lawler, Matusky & Skelly Engineers. 1992. 1990 Year Class Report. Prepared for Consolidated Edison Company of New York, Inc.

Lawler, Matusky & Skelly Engineers. 1996. 1991 Year Class Report. Prepared for Consolidated Edison Company of New York, Inc.

Martin Marietta Environmental Systems. 1986. 1984 Year Class Report. Prepared for Consolidated Edison Company of New York, Inc.

Normandeau Associates, Inc. 1985a. 1982 Year Class Report. Prepared for Consolidated Edison Company of New York, Inc.

Normandeau Associates, Inc. 1985b. 1983 Year Class Report. Prepared for Consolidated Edison Company of New York, Inc.

Normandeau Associates, Inc. 2006. Abundance and stock characteristics of the Atlantic tomcod spawning population in the Hudson River, winter 2003-2004. Prepared for Entergy Nuclear Operations, Inc.

Pace, M.L., Findlay, S.E.G., and Fischer, D.T. 1998. Effects of an invasive bivalve on the zooplankton community of the Hudson River. Freshwat. Biol. 39: 103-116.

Seber, G.A.F. 1982. The estimation of animal abundance and related parameters, 2nd edition.

Charles Griffin and Company Ltd., London.

Strayer, D.L., K.A. Hattala, and A.W. Kahnle. 2004. Effects of an invasive bivalve (Dreissena polymorpha) on fish in the Hudson River estuary. Can. J. Fish. Aquat. Sci. Vol. 61, 2004 Strayer, D.L., Powell, J., Ambrose, P., Smith, L.C., Pace, M.L., and Fischer, D.T. 1996. Arrival, spread, and early dynamics of a zebra mussel (Dreissenapolymorpha)population in the Hudson River estuary. Can. J. Fish. Aquat. Sci. 53: 1143-1149.

Texas Instruments, Inc. 1977. 1974 Year Class Report for the multiplant impact study of the Hudson River Estuary ("Year Class Report"). Prepared for Consolidated Edison Company of New York, Inc.

Texas Instruments, Inc. 1978. 1975 Year Class Report. Prepared for Consolidated Edison Company of.New York, Inc.

10

Texas Instruments, Inc. 1979. 1976 Year Class Report. Prepared for Consolidated Edison Company of New York, Inc.

Texas Instruments, Inc. 1980a. 1977 Year Class Report. Prepared for Consolidated Edison Company of New York, Inc.

Texas Instruments, Inc. 1980b. 1978 Year Class Report. Prepared for Consolidated Edison Company of New York, Inc.

Texas Instruments, Inc. 1981. 1979 Year Class Report. Prepared for Consolidated Edison Company of New York, Inc.

Versar, Inc. 1987. 1985 Year Class Report. Prepared for Consolidated Edison Company of New York, Inc.

11

V. TABLES 12

Table B-i:' Abundance Indices and Associated Standard Errors ("SE"), Based on Long River Survey Data.

White Perch fStriped Bass Bay Anchovy [ American Shad Year Post Yolk-Sac Larvae Post Yolk-Sac Larvae Post Yolk-Sac Larvae Post Yolk-Sac Larvae Class Index SE Index SE Index SE Index SE 1974 139,139,531 9,461,494 116,793,360 14,525,520 9,111,556 2,155,940 32,149,174 5,436,351 1975 418,776,213 14,897,579 167,352,740 11,297,813 167,900,084 21,837,003 38,104,249 3,668,122 1976 571,765,805 26,442,918 55,463,017 3,014,531 341,602,306 88,340,964 30,532,518 4,411,773 1977 628,980,330 32,916,730 147,319,974 9,345,100 108,551,600 47,407,559 31,792,930 6,593,648 1978 852,286,248 54,375,932 113,088,409 9,188,267 13,499,413 2,574,305 14,808,830 1,725,494 1979 889,355,233 27,210,046 111,789,357 10,177,101 31,217,251 4,193,924 76,008,019 8,374,974 1980 731,972,701 29,071,443 193,067,215 15,374,877 282,472,131 47,526,524 62,624,636 6,850,621 1981 878,432,947 57,291,346 565,580,988 29,382,161 386,003,879 40,370,163 107,959,543 9,223,464 1982 1,533,952,669 63,678,126 214,574,357 17,311,853 7,721,685 1,434,887 105,866,404 11,668,608 1983 689,913,421 28,117,162 134,838,042 8,271,457 45,952,457 8,165,287 108,436,433 21,821,939 1984 659,480,715 40,337,372 200,167,635 28,656,262 39,045,805 11,944,143 46,171,178 7,590,296 1985 1,421,323,747 59,947,138 93,874,968 7,700,762 349,889,115 30,127,176 84,264,727 11,412,620 1986 2,052,461,814 98,317,198 171,163,020 8,998,325 118,354,834 10,883,362 152,128,084 17,215,544 1987 1,012,538,712 32,052,565 405,324,057 16,848,690 189,564,190 11,607,205 27,892,890 3,374,299 1988 754,305,782 42,580,552 351,072,816 35,669,346 152,035,433 30,786,324 78,027,604 11,883,534 1989 925,022,100 102,183,412 1,071,325,339 99,670,379 14,134,359 3,081,790 86,573,611 8,951,649 1990 768,296,570 79,095,729 1,295,596,696 153,298,294 890,027 256,957 108,278,134 14,347,189 1991 907,921,874 61,907,978 1,896,058,025 203,606,883 5,602,678,703 551,771,800 43,259,681 5,089,006 1992 1,211,029,021 53,752,949 1,436,836,717 103,392,955 77,338,304 10,339,754 99,755,719 15,257,291 1993 1,231,794,687 50,130,673 2,008,989,233 181,226,826 573 ,83 9,976 50,894,605 33,386,515 6,848,737 1994 1,043,697,036 46,808,643 2,009,527,814 204,188,984 583,968,501 47,054,442 37,913,769 3,901,481 1995 623,420,693 29,028,682 939,209,970 99,781,400 839,521,735 64,631,235 24,920,433 3,668,256 1996 1,505,193,548 83,865,093 3,629,518,187 365,724,596 405,338,653 43,811,932 31,112,517 3,986,134 1997 307,236,756 17,277,642 1,252,166,315 211,669,199 1;009,992,702 213,235,143 19,546,174 4,202,344 1998 575,146,100 35,729,754 1,413,117,919 122,712,647 18,860,574 3,243,002 10,840,582 1,389,788 1999 673,636,250 39,842,187 3,468,043,472 358,992,219 287,637,139 29,957,432 19,920,980 4,244,449 2000 1,180,789,474 133,501,704 5,803,754,734 715,393,543 1,355,732 345,802 10,158,022 1,432,512 2001 734,730,398 61,307,779 5,258,385,169 340,997,297 51,298,063 22,554,315 48,974,089 9,013,780 2002 566,273,447 39,302,719 587,019,561 40,128,197 173,651,942 21,508,231 11,487,215 2,321,455 2003 692,003,842 45,947,390 1,853,946,447 202,927,363 6,523,373 2,802,470 11,636,329 1,626,253 2004 721,129,750 39,776,443 1,646,077,551 106,676,037 717,812,470 71,311,509 13,196,538 1,966. 124 13

Table 1. Abundance Indices and Associated Standard Errors ("SO'), Based on Long River Survey Data (continued).

River Herring Atlantic Tomcod Year Post Yolk-Sac Larvae Post Yolk-Sac Larvae Class Index SE Index SE 1974 1,925,093,580 1,073,772,004 128,306,743 19,426,263 1975 2, 177,549,296 197,088,426 67,024,797 19,768,962 1976 1,590,931,203 156,327,051 42,77.7,042 13,470,065 1977 1,789,369,237 309,551,598 164,621,663 70,515,234 1978 2,483,545,195 230,530,412 54,313,088 1.0,307,482 1979 1,492,563,623 65,281,612 18,127,435 3,099,375 1980 1,451,864,997 82,238,743 95,402,234 13,128,146 1981 2,097,039,055 238,479,765 74,140,778 13,052,007 1982 2,761,588,726 248,286,854 28,419,800 7,665,326 1983 3 398,542,430 247,313,066 42,683,202 8,311,722 1984. 2,263,857,937 168,138,864 147,133,069 25,916,525 1985 2,360,908,396 138,470,331 109,664,584 11,132,251 1986 3,060,453,736 212,481,475 53,404,268 4,770,519 1987 945,121,604 62,594,106 138,570,516 12,594,732 1988 1,205,794,912 101,740,608 78,376,300 10,680,903 1989 1,515,234,476 181,441,810 185,450,859 23,858,579 1990 1,296,493,803 106,557,985 107,915,374 25,158,013 1991 1,105,840,600 89,654,766 116,333,462 14,859,973 1992 1,592,451,980 119,021,893 32,021,214 4,889,565 1993 957,005,646 76,057,902 126,394,886 .20,139,893 1994, 1,006,699,048 57,426,960 85,456,373 22,227,930 1995 745,594,402 44,387,051 79,816,881 6,641,688 1996 2,092,537,070 119,641,340 51,571,386 5,696,759 1997 338,336,798 21,073,725 110,409,961 28,829,551 1998 599,669,094 37,989,853 53,594,909 8,409,591 1999 658,448,983 38,493,738 17,392,702 2,076,588 2000 1,736,751,090 110,473,230 11,120,807 1,442,773 200.1 941,430,470 69,923,386 93,816,691 8,320,053 2002 798,010,496 43,842,607, 4,382,650 649,979 2003 608,369,228 39,023,677 38,715,789 3,683,762 2004 681,555,090 40,476,571 115,401,578 16,005,570 14

Table B-2. Abundance Indices and Associated Standard Errors C'SE"), Based on Beach Seine Survey Data.

White Perch Striped Bass Spottail Shiner Spottail Shiner Year Young-of-Year Young-of-Year Young-of-Year Egg Class Index SE Index SE Index SE Index SE 1974 566,346 61,280 1,373,138 264,598 658,945 87,448 1,128,997 107,867 1975 2,342,937 440,999 1,367,496 242,374 1,286,297 193,361 1,578,455 195,841 1976 1,944,220 255,910 864,743 70,734 1,324,434 203,989 0 0 1977 953,799 87,722 1,375,537 124,595 495,690 66,445 0 0 1978 2,675,700 402,374 3,042,920 614,048 1,363,313 148,541 0 0 1979 2,921,393 285,862 794,022 91,389 956,236 97,330 0 0 1980 1,884,895 231,650 1,265,254 147,121 633,323 72,196 312,488 80,635 1981 1,862,222 160,903 1,827,767 152,481 1,865,058 216,442 627,176 96,220 1982 1,967,754 287,490 934,550 97,768 477,090 62,605 173,130 25,821 1983 1,803,266 399,823 1,642,536 191,103 1,070,822 104,909 197,639 51,127 1984 703,959 145,133 1,300,754 173,872 616,182 128,367 222,054 41,973 1985 757,003 82,536 238,259 21,226 543,246 66,532 116,419 17,690 1986 1,036,321 97,303 298,745 31,415 388,736 69,297 276,641 48,687 1987 1,169,236 121,876 2,976,381 314,807 470,267 74,827 234,226 45,133 1988 1,738,310 255,364 1,172,303 68,239 419,874 49,588 276,581 49,087 1989 1,105,280 278,101 1,238,434 116,464 623,204 95,526 272,136 61,641 1990 588,162 75,727 1,486,911 89,409 808,662 101,694 144,012 31,435 1991 580,165 76,201 1,125,126 64,076 855,292 110,557 833,354 126,276 1992 463,555 53,444 1,046,654 53,265 726,888 124,009 453,069 112,051 1993 806,848 97,157 1,640,132 90,969 655,117 95,425 391,317 97,925 1994 315,662 39,618 1,136,106 63,179 1,624,997 289,784 168,358 27,009 1995 425,062 49,042 1,404,935 89,202 603,130 94,204 229,394 41,809 1996 44,925 10,283 299,997 30,506 174,026 39,053 58,663 15,101 1997 571,160 114,812 1,892,597 169,399 1,197,799 170,583 140,490 33,758 1998 270,835 51,992 1,384,364 85,327 273,165 53,055 147,082 40,400 1999 1,411,184 169,447 1,715,282 142,568 2,040,399 243,244 154,889 21,463 2000 304,950 52,787 580,006 52,449 303,081 52,956 164,945 29,160 2001 1,019,516 119,666 2,392,216 170,860 2,143,066 610,761 16,919 5,028 2002 699,145 80,612 1,145,686 60,295 1,132,479 146,862 174,197 50,311 2003 2,177,013 228,303 2,282,684 118,276 2,102,568 257,006 565,369 131,279 2004 632,961 89,075 807,661 70,743 1,031,399 152,802 436,330 79,667 15

Table 2. Abundance Indices and Associated Standard Errors ("SE"), Based on Beach Seine Survey Data (continued).

Alewife Blueback Herring Year Young-of-Year Young-of-Year Class Index SE Index SE 1974 583,238 74,805 3,647,758 502,857 1975 572,550 107,585 10,888,524 1,249,788 1976 352,263 96,375 21,621,271 3,075,761 1977 517,792 49,081 31,795,371 4,717,652 1978 1,027,891 174,698 22,993,451 4,200,939 1979 340,271 59,099 8,221,314 1,461,758 1980 93,783 17,894 8,892,467 2,207,337 1981 477,348 84,403 32,066,440 9,586,015 1982 116,606 24,817 10,164,307 1,750,817 1983 214,922 42,154 16,326,879 2,278,723 1984 49,776 10,864 3,577,323 786,742 1985 119,509 22,024 3,323,511 664,762 1986 119,468 48,899 1,555,182 357,032 1987 80,611 13,768 6,188,101 773,111 1988 87,080 15,727 5,887,963 1,008,925 1989 43,711 12,956 3,230,116 497,839 1990 157,159 25,580 9,436,487 1,274,900 1991 335,535 63,111 3,530,392 596,059 1992 40,507 9,371 6,642,282 1,599,250 1993 69,438 11,826 4,234,168 531,496 1994 148,030 30,079 9,584,696 1,308,960 1995 91,731 22,716 3,202,735 892,613 1996 47,371 14,912 4,044,353 890,186 1997 291,323 54,177 12,075,530 2,541,612 1998 40,865 30,194 155,761 32,365 1999 445,167 79,622 5,691,570 776,702 2000 76,445 37,606 2,342,499 572,561 2001 330,876 70,451 5,268,663 704,402 2002 60,954 13,491 1,438,577 299,230 2003 452,292 87,223 10,203,281 1,459,824 2004 218,118 35,902 5,091,421 620,888 16

Table B-3. Abundance Indices and Associated Standard Errors ("SE"), Based on Fall Shoals Survey and Atlantic Tomcod Mark Recapture Data.

Bay Anchovy American Shad Atlantic Tomcod Year Young-of-Year (FSS) Young-of-Year (FSS) Ages 1 and 2 (ATMR)

Class Index SE Index SE Index SE 1974 - - 3,666,156.2 667,339 1975 - - - 3,680,086.9 375,142 1976 - - - 19,210,329.2 2,767,571.7 1977 - - - 2,434,397.0 458,488.1 1978 - - - 5,894,583.8 917,687.4 1979 - - - 9,128,535 1,692,155.4 1980 - - - 4,747,440 3,355,405.2 1981 - - - 25,066,665.0 14,468,003 1982 - - - 12,983,676.9 2,899,705 1983 - - - 6,657,331.2 1,302,504.2 1984 - - - -

1985 218,612,898 21,269,766 1,591,435 190,139 2,093,677 171,796 1986 132,925,173 13,133,411 3,104,605 640,844 -

1987 246,910,112 26,982,497 647,070 157,299 3,526,907.2 570,280 1988 422,678,791 38,213,532 997,871 144,252 5,897,656.7 524,801.4 1989 349,952,337 26,107,654 2,754,815 198,752 6,804,809.4 1,239,300.2 1990 161,039,442 14,450,450 1,139,272 235,276 3,208,815.0 615,208.4 1991 190,474,265 11,540,891 680,209 72,781 388,763.0 84,175.2 1992 185,902,303 13,738,226 1,306,732 147,744 2,553,778.3 319,857.2 1993 249,913,241 19,475,645 .464,702 48,446 663,439.1 155,295.9 1994 206,642,043 14,141,476 1,036,782 88,932 2,384,183 659,618.4 1995 439,617,793 28,732,239 471,444 75,896 88,492.5 50,523.4 1996 102,941,191 5,959,974 2,859,373 451,439 3,277,909.3 1,637,090 1997 283,382,412 17,014,202 913,970 107,851 1,291,980.5 302,916.5 1998 189,541,611 9,166,785 232,260 56,459 592,891.0 241,105.3 1999 165,375,818 9,972,244 853,411 135,639 181,179.0 59,983.3 2000 57,208,944 3,577,181 878,405 100,807 2,504,266 624,327.3 2001 109,701,139 8,052,515 1,006,787 162,014 40,875 28,743.1 2002 171,692,430 10,652,063 497,537 57,524 108,528.0 76,363 2003 148,898,706 11,753,477 351,278 47,131 1,653,319 425,310 2004 218,178,981 17,899,774 336,973 63,105 --

17

Table B-4. Estimates of Indian Point Conditional Mortality Rate (CMR) for entrainment.

Year White Perch Striped Bass Spottail Shiner Bay Anchovy American Shad River Herring Atlantic Tomcod Class CMR CMR I CMR I CMR CMR I CMR I CMR 1974 7.45 5.65 0.87 7.31 0.22 0.83 3.65 1975 8.65 7.78 1.04 6.61 0.35 1.42 6.75 1976 3.22 4.73 1.38 3.45 0.33 1.85 8.76 1977 7.27 13.89 1.41 13.78 0.38 2.47 10.15 1978 5.28 8.55 2.32 12.54 0.24 1.26 10.6 1979 8.02 11.92 1.62 10.8 0.2 2.24 18.8 1980 3.36 11.87 1.66 18.44 0.03 0.48 25.47 1981 6.54 4.17 3.43 18.56 0.2 0.57 11.68 1982 4.33 6.99 2.06 4.19 0.44 0.81 17.47 1983 17.23 7.36 3.17 9.04 0.09 3.05 7.69 1984 8.92 17.25 1.58 6.26 7.5 5.34 16.58 1985 0.55 3.97 1.77 10.06 0 0.02 34.5 1986 4.07 16.26 1.55 5.07 3.56 0.92 11.36 1987 0.66 2.3 1.53 9.99 0 0.04 14.61 1988 7.94 11.63 4.1 17.73 0.15 0.51 23.94 1989 4.03 5.96 8.32 7.96 0.28 1.41 4.49 1990 3.48 6.12 2.18 20.85 0.43 2.94 5.52 1991 1.4 4.95 3.92 9.09 0.07 0.41 6.99 1992 2.7 6.16 0.99 7.12 0.05 0.41 14.11 1993 2.34 5.6 0.89 7.08 0.13 0.23 3.67 1994 3.14 6.81 1.1 5.94 0.12 0.49 7.57 1995 1.92- 4.22 2.54 14.99 0.1 0.12 5.77 1996 4.S8 12.01 1.89 15.55 0.42 0.49 8.47 1997 1.29 1.42 0.64 6.62 0.05 0.6 10.35 1998 4.87 8.46 0.45 7.82 0.12 0.59 10.01 1999 4.16 11.35 2.57 13.81 0.23 3.66 21.54 2000 7.31 4.03 1.63 7.77 1.86 4 11.23 2001 5.69 8 2.56 15.4 0.3 .1.82 20.97 2002 11.96 13.77 3.03 10.57 1.23 4.84 23.25 2003 7.67 12.26 1.21 12.97 0.19 1.85 20.43 2004 18

Table B-5. Zebra Mussel Index.

Year [ Zebra Mussel Class I Index 1974 0 1975 0 1976 0 1977 0 1978 0 1979 0 1980 0 1981 0 1982 0 1983 0 1984 0 1985 0 1986 0 1987 0 1988 0 1989 0 1990 0 1991 0 1992 0 1993 1 1994 1 1995 1 1996 1 1997 1 1998 1 1999 1 2000 1 2001 1 2002 1 2003 1 2004 1 19

Table B-6. Striped Bass Predation Index.

Striped Bass Year PYSL Class Index 1974 116,793,360 1975 167,352,740 1976 55,463,017 1977 147,319,974 1978 113,088,409 1979 111,789,357 1980 193,067,215 1981 565,580,988 1982 214,574,357

.1983 134,838,042 1984 200,167,635 1985 93,874,968 1986 171,163,020 1987 405,324,057 1988 351,072,816 1989 1,071,325,339 1990 1,295,596,696 1991 1,896,058,025 1992 1,436,836,717 1993 2,008,989,233 1994 2,009,527,814 1995 939,209,970 1996 3,629,518,187 1997 1,252,166,315 1998 1,413,117,919 1999 3,468,043,472 2000 5,803,754,734 2001 5,258,385,169 2002 '587,019,561 2003 1,853,946,447 2004 1,646,077,551 20

Table B-7. Temperature Indices.

FSS PWW Year ITemperature Degree-Day Class [ Index Index 1974 1975 1976 18.8 1977 57.7 1978 60.8 1979 22.5 61.3 1980 22.4 128.1 1981 19.8 98.0 1982 64.3 1983 24.0 107.9 1984 22.8 91.2 1985 21.5 63.1 1986 21.5 61.1 1987 19.9 111.1 1988 24.6 121.1 1989 22.2 65.2 1990 22.7 68.4 1991 21.5 108.9 1992 20.2 6.5 1993 22.2 97.1 1994 22.2 103.6 1995 22.6 94.9 1996 22.3 28.6 1997 22.4 63.7 1998 23.5 94.1 1999 23.2 136.8 2000 21.7 0.9 2001 23.1 98.9 2002 23.5 121.6 2003 22.6 106.8 2004 22.5 18.8 21

Table B-8. Survival Indices.

White Perch Striped Bass Spottail Shiner Bay Anchovy American Shad River Herring Atlantic Tomcod Year PYSL to YOY PYSL to YOY Egg to YOY PYSL to YOY PYSL to YOY PYSL to YOY -Egg to Age-i Age-1 to Age-2 Class Index Index Index Index Index Index index I Index 1974 0.0041 0.0118 0.5837 - 0.0030 -

1975 0.0056 0.0082 0.8149 - 0.0053 - 0.2008 1976 0.0034 0.0156 -- - 0.0138 0.4411 0.0103 1977 0.0015 0.0093 - - 0.0184 0.0371 0.0249 1978 0.0031 0.0269 - - 0.0100 0.2826 0.0460 1979 0.0033 0.0071 - - - 0.0077 0.1731 -

1980 0.0026 0.0066 2.0267 - 0.0064 -

1981 0.0021 0.0032 2.9737 - 0.0155 -

1982 0.0013 0.0044 2.7557 - 0.0039 - 0.0699 1983 0.0026 0.0122 5.4181 - 0.0049 0.0613 1984 0.0011 0.0065 2.7749 - 0.0018 1985 0.0005 0.0025 4.6663 0.6248 0.0189 0.0015 1986 0.0005 0.0017 1.4052 1.1231 0.0204 0.0006 1987 0.0012 0.0073 2.0077 1.3025 0.0232 0.0068 - 0.2014 1988 6 0.0023 0.0033 1.5181 2.7801 0.0128 0.0050 0.1235 0.3714 1989 0.0012 0.0012 2.2900 24.7590 0.6318 0.0023 0.1186 0.1251 1990 0.0008 0.0011 5.6152 180.9377 0.0105 0.0084 0.0298 0.0448 1991 0.0006 0.0006 1.0263 0.0340 0.0157 0.0035 0.0055 1.3636 1992 0.0004 0.0007 1.6044 2.4038 0.0131 0.0042 0.3153 0.1078 1993 0.0007 0.0008 1.6741 0.4355 0.0139 0.0045 0.0154 0.4661 1994 0.0003 0.0006 9.6520 0.3539 0.0273 0.0097 0.3110 -

1995 0.0007 0.0015 2.6292 0.5237 0.0189 0.0044 - -

1996 0.0000 0.0001 2.9665 0.2540 0.0919 0.0043 - 0.2314 1997 0.0019 0.0015 8.5259 0.2806 0.0468 0.0366 0.0148 0.2933 1998 0.0005 0.0010 1.8572 10.0496 0.0214 0.0003 0.0173 0.1004 1999 0.0021 0.0005 13.1733 0.5749 0.0428 0.0093 0.0160 1.0951 2000 0.0003 0.0001 1.8375' 42.1978 0.0865 0.0015 0.7792 -

2001 0.0014 0.0005 126.6690 2.1385 0.0206 0.0065 -

2002 0.0012 0.0020 6.5011 0.9887 0.0433 -0.0019 2003 0.0031 0.0012 3.7189 22.8254 0.0302 0.0186 2004 0.0009 0.0005 2.3638 0.3039 0.0255 0.0079 22

Table B-9. Growth Rate Indices 1 11 T American 1 Blueback Year Class 1974 White Perch Index 0.0972 j

Striped Bass Index 0.0727 j

Spottail Shiner Index 0.0844 j

Bay Anchovy Index { Index Shad Alewife Index 0.0265 J Index Herring 0.08 10 1975 0.0605 0.0495 0.0624 0.0420 0.0563 1976 0.0873 0 .05 42

-1977 1978 1979 0.0725 0.0697 0.0768 0.057 1 0.0894 1980 0.0790 0.0729 0 .0742 0.0337 0.0658 1981 0.0578 0.0501 0.0651 0.0350 0.0632 1982 0.0769 0.0460 0.0733 0.0454 0.0591 1983 0.0845 0.09 19 0. 1417 0.09 16 0. 1037 1984 0.1142 0.0942 0 .0 824 0.0752 0.0669 1985 0.06 11 0.1245 0.0520 0.0288 0.0234 0.0525 0.0304 1986 0.0640 0.0433 0.0534 0.0703 0.07 16 0.0459 0.0604 1987 0.0750 0.0685 0.0864 0.03 11 0.0466 0.0630 0.0555 1988 0.0589 0.0532 0.0691 0.0928 0.08 13 0.0520 0.0573 1989 0.0973 0.07 12 0.0788 0.0870 0.066 1 0.08 15 0.0858 1990 0. 1081 0.0866 0.0998 0. 1000 0.07 11 0.0585 0.0603 1991 0.0620 0.0591 0.0552 0.0505 0.0572 0.05 10 0.0808 1992 0.0933 0.0840 0.0616 0.0617 0.0759 0.04 12 0.058 1 1993 0.0732 0.0589 0.0621 0.0475 0.0346 0.0271 0.0200 1994 0.0362 0.03 72 0.0502 0.0890 0.0546 0.0425 0.0204 1995 0. 1088 0.0823 0.0793 0.0668 0.0460 0.0471 0.0845 1996 0. 1073 0. 1070 0.1168 0.0642 0.0853 0.0729 0.0384 1997 0.0764 0.0657 0.07 16 0.0997 0.0756 0.046 1 0 .0322 1998 0.08 13 0 .0802 0.0603 0.0732 01.0520 0.0670 0.0454 1999 0.457 0.0671 0.04 14 0.0256 0.0320 0.0086 0.03 16 2000 0.08 13 0.0773 0.0732 0.0781 0.0824 0.0797 0.06 10 2001 0.0961 0.0652 0.0978 0.0637 0.07 10 0.0686 2002 0.0624 0.0625 0.0637 0.0400 0.0445 0.0366 0.0982 2003 0.0732 0.0517 0.0863 0.0841 0.0493 0.0536 0.0465 2004 0.05 15 0.0474 0.0592 .0.1006 0.0601 0.04 11 0.0715 23

Table B- 10. Spatial DistributionIndices -- The Fractionof Standing Crop that is Downriverof rkm 100.

Year White Perch Striped Bass Spottail Shiner Bay Anchovy American Shad PYSL 1 YOY PYSL YOY YOY PYSL YOY PYSL YOY Class Index Index Index Index Index Index Index Index Index 1974 0.4102 0.3501 0.6199 0.8947 0.0783 1.0000 0.0209 1975 0.4373 0.7000 0.7998 0.9192 0.0772 1.0000 0.1802 1976 0.1782 0.5473 0.7834 0.9109 0.1804 1.0000 0.0380 1977 0.2008 0.3872 0.7088 0.8765 0.0668 0.9999 0.0139 1978 0.2638 0.6703 0.8044 0.9554 0.1594 1.0000 0.0274 1979 0.3384 0.6210 0.8876 0.9027 0.2137 1.0000 0.0351 1980 0.2276 0.6592 0.7788 0.8260 0.0709 0.9998 0.0198 1981 0.2585 0.6813 0.5834 0.9247 0.0874- 0.9998 0.0267 1982 0.3628 0.7975 0.8013 0.9668 0.2880 1.0000 0.0461 1983 0.4220 0.5556 0.8632 0.8634 0.1347 0.9997 0.0293 1984 0.2366 0.7919 0.8475 0.9402 0.0794 0.9997 0.3433 1985 0.1420 0.6204 0.6800 0.9004 0.0749 0.9982 0.8978 0.0015 0.3707 1986 0.2147 0.7541 0.8164 0.9115 0.0962 1.0000 0.9178 0.0104 0.1426 1987 0.0984 0.4309 0.4985 0.9110 0.0145 0.9964 0.9547 0.0012 0.1960 1988 0.3191 0.7514 0.7726 0.8233 0.1086 0.9249 0.8584 0.0032 0.3732 1989 0.4646 0.7267 0.7884 0.9188 0.1493 0.9557 0.8974 0.1272 0.1777 1990 0.3406 0.4131 0.5434 0.8682 0.0743 1.0000 0.9365 0.0539 0.3500 1991 0.2109 0.3581 0.7037 0.6287 0.0165 0.9835 0.6000 0.0036 0.2074 1992 0.2616 0.5105 0.8321 0.8619 0.0344 0.9964 0.8679 0.0154 0.3391 1993 0.1911 0.3349 0.7026 0.8189 0.0593 0.9966 0.7392 0.0029 0.2788 1994 0.2156 0.4619 0.8595 0.8084 0.0767 0.9995 0.9240 0.0077 0.3255 1995 0.2054 0.3869 0.7445 0.8986 0.0143 0.9888 0.7635 0.0049 0.3529 1996 0.1587 0.7707 0.7570 0.7614 0.1261 0.9978 0.9603 0.0062 0.2600 1997 0.2799 0.4857 0.8852 0.8555 0.0774 1.0000 0.8117 0.0078 0.1259 1998 0.2646 0.5741 0.8162 0.8603 0.0351 0.9986 0.8190 0.0202 0.0674 1999 0.1919 0.6035 0.7352 0.7392 0.0220 0.9987 0.8487 0.0235 0.2024 2000 0.6546 0.5040 0.9908 0.7759 0.1723 0.9797 0.8889 0.1399 0.2930 2001 0.1508 0.4677 0.7024 0.8177 0.0193 1.0000 0.9302 0.0438 0.2072 2002 0.2851 0.2743 0.8712 0.7682 0.0008 1.0000 0.7100 0.0879 0.0657 2003 0.3001 0.4981 0.8249 0.8803 0.0572 1.0000 0.9507 0.0132 0.1721 2004 0.2150 0.1672 0.8196 0.6875 0.0407 0.9997 0.9363 0.0364 0.1225 24

Table 10. Spatial Distribution Indices -- The Fraction of Standing Crop that is Downriver of rkm 100 (continued).

Year Alewife Biueback Herring Atlantic Tomcod PYSL YOY PYSL YOY PYSL Class Index* Index Index* Index Index 1974 0.0448 0.9065 0.0448 0.2928 0.9903 1975 0.0650 0.8709 0.0650 0.1996 0.9902 1976 0.1571 0.6064 0.1571 0.1818 0.9912 1977 0.0575 0.5622 0.0575 0.4164 0.9953 1978 0.0985 0.5909 0.0985 0.1202 0.9854 1979 0.1189 0.4444 0.1189 0.1452 0.9860 1980 0.0193 0.5528 0.0193 0.0663 0.9528 1981 0.0844 0.4460 0.0844 0.3646 0.9853 1982 0.0704 0.7575 0.0704 0.2143 0.9663 1983 0.1715 0.2247 0.1715 0.1088 0.9960 1984 0.2939 0.3330 0.2939 0.2982 0.9778 1985 0.0086 0.4559 0.0086 0.3012 0.9496 1986 0.0776 0.3842 0.0776 0.1475 0.9741 1987 0.0077 0.3363 0.0077 0.2725 0.8921 1988 0.0545 0.7762 0.0545 0.2218 0.9609 1989 0.0894 0.7374 0.0894 .0.1058 0.9980 1990 0.1879 0.4526 0.1879 0.0988 0.9712 1991 0.0228 0.0304 0.0228 0.0101 0.9837 1992 0.0595 0.4622 0.0595 0.5121- 0.9976 1993 0.0097 0.2508 0.0097 0.2744 0.9950 1994 0.0265 0.5730 0.0265 0.3236 0.9915 1995 0.0184 0.1994 0.0184 0.1357 0.9411 1996 0.0186 0'.4721 0.0186 0.6749 0.9852 1997 0.1830 0.2906 0.1830 0.0769 0.9935 1998 0.0448 0.8889 0.0448 0.0846 0.9928 1999 0.1857 0.2304 0.1857 0.2034 0.9732 2000 0.2224 0.1696 0.2224 0.1666 0.9024 2001 0.0698 0.1830 0.0698 0.0800 0.9721 2002 0.2350 0.0914 0.2350 0.3404 0.9938 2003 0.1196 0.5519 0.1196 0.2539 0.9934 2004 0.1376 0.5527 0.1376 0.1861 0.9849 25

Table B-11. Striped Bass Stressor Response Indian Point Yearclass Metric Entrainment Zebra Striped Bass Temperature Mortality Mussels Predation (CMR)

PYSL-to-YOY -0.69-08

-0.84 Survival PYSL Abundance YOY Abundance Growth Rate n

%PYSL Downriver * "

orm100 Downriver -0.63 -0.68 of rkm 100

• Yearclass +0.84 +0.84 Correlation coefficients significant at the 0.05 level are printed in black and correlation coefficients significant at the 0.10 level are printed in gray. A blank cell in the table indicates that the correlation coefficient was not significant at a probability level of 0.10 or lower. Cells shaded gray indicate pairs of indices that were not considered relevant, based on impact hypothesis and/or life history considerations.

r 26

Table B-12. Striped Bass Response Metric Response Yearclass Metric PYSL-to- PYSL YOY YY Abundance Abundance Survival

.O 0.84 Survival PYSL +0.84 Abundance YOY L Abundance Yearclass -0.84 +0.84 Correlation coefficients significant at the 0.05 level are printed in black and correlation coefficients significant at the 0.10 level are printed in gray. A blank cell in the table indicates that the correlation coefficient was not significant at a probability level of 0.10 or lower. Cells shaded gray indicate pairs of indices that were not considered relevant, based on impact hypothesis and/or life history considerations.

27

Table B-13. White Perch Stressor Response Indian Point Yearclass Metric Entrainment Zebra Striped Bass Temperature Mortality Mussels Predation (CMR)

PYSL-to-YOY Survival +0.44 -0.36 -0.57 +0.42 -0.53 PYSL -03 Abundance -04 YOYii! *** -0.54 Abundance *** -0.51 Growth I of rkm 100

%YOY*

Downriver **,

' -0.40 -0.37 ofrkm 100 aA Yearclass +0.84 +0.84 Correlation coefficients significant at the 0.05 level are printed in black and correlation coefficients significant at the 0.10 level are printed in gray. A blank cell in the table indicates that the correlation coefficient was not significant at a probability level of 0.10 or lower. Cells shaded gray indicate pairs of indices that were not considered relevant, based on impact hypothesis and/or life history considerations.

28

Table B-14. White Perch Response Metric Response Yearclass Metric PYSL-to- PYSL YOY YOY Abundance Abundance Survival PYSL-to-YOY +0.76 -0.53 Survival PYSL Abundance YOY +0.76 -0.51 Abundance Yearclass -0.53 -0.51" Correlation coefficients significant at the 0.05 level are printed in black and correlation coefficients significant at the 0.10 level are printed in gray. A blank cell in the table indicates that the correlation coefficient was not significant at a probability level of 0.10 or lower. Cells shaded gray indicate pairs of indices that were not considered relevant, based on impact hypothesis and/or life history considerations.

29

Table B-15. American Shad Stressor Response Indian Point Yearclass Metric Entrainment Zebra Striped Bass Mortality Mussels Predation Temperature (CMR) 1 PYSL-to-YOY +0.58 +0.55 Survival AbnacPYSL ,.**

Abunanc -0.31 -0.46 Abundance, .- 05 Growth Rate

% PYSL Downriver J of rkm 100 i Downriver -0.48 of rkm 100*  %!'i'*

Yearclass; +0.84 +0.84 Correlation coefficients significant at the 0.05 level are printed in black and correlation coefficients significant at the 0.10 level are printed in gray. A blank cell in the table indicates that the correlation coefficient was

'not significant at a probability level of 0.10 or lower. Cells shaded gray indicate pairs. of indices that were not considered relevant, based on impact hypothesis and/or life history considerations.

30

Table B-16. American Shad Response Metric Response Yearclass Metric PYSL-to- PYSL YOY YOY Survval Abundance Abundance Survival YOY *+0.55 Survival Abndnc SL: +0.75 -0.46 YOY +0.75

• 0.57 Abundance Yearclass +0.55 -0.46 -0.57 Correlation coefficients significant at the 0.05 level are printed in black and correlation coefficients significant at the 0.10 level are printed in gray. A blank cell in the table indicates that the correlation coefficient was not significant at a probability level of 0.10 or lower. Cells shaded gray indicate pairs of indices that were not considered relevant, based on impact hypothesis and/or life history considerations.

31

Table B-17. Atlantic Tomcod Stressor Response Indian Point T eu Yearclass/

Metric Entrainment Zebra Striped Bass Year Mortality Mussels Predation (PWW (CMR) degree-days)

Egg-to-Agel -0.59 Survival Agel-to- '

Survival Egg "-0.42 Abundance Abundance -0.36 Age !v -0.65 -0.72

%PYSL Downriver ofrkm 100 Yearclass/ +0.84 +0.84 Year Correlation coefficients significant at the 0.05 level are printed in black and correlation coefficients significant at the 0.10 level are printed in gray. A blank cell in the table indicates that the correlation coefficient was not significant at a probability level of 0.10 or lower. Cells shaded gray indicate pairs of indices that were not considered relevant, based on impact hypothesis and/or life history considerations.

32

Table B-18. Atlantic Tomcod Response Metric Response Yearclass/

Metric Egg-to- Agel-to- Egg Agel Year Agel Age2 Abundance Abundance Survival Survival Egg-to-Agel+06 Survival+06 Agel-to- +0.56 Age2 Survival Egg -0.42 Abundance Age]1 +.1 07 Abundance +.107 Yearclass/

Year+0.56 Year -0.42 -0.72 I I I Correlation coefficients significant at the 0.05 level are printed in black and correlation coefficients significant at the 0.10 level are printed in gray. A blank cell in the table indicates that the correlation coefficient was not significant at a probability level of 0.10 or lower. Cells shaded gray indicate pairs of indices that were not considered relevant, based on impact hypothesis and/or life history considerations.

33

Table B-19. Alewife Stressor Response Indian-Point Yearclass Metric Entrainment Zebra Striped Bass Temperature Mortality Mussels Predation (CMR)

PYSL-to-YOY Survival 0 "0.q PYSL Abundance -0.56 -0.70

+0.84 +0.84 Abundance YOY *"

Growth Rate Downriver

• of rkm 104 Downriver -. 3-0.4 of rkm 140 Yearclass +0.84 +0.84 Correlation coefficients significant at the 0.05 level are printed in black and correlation coefficients significant at the 0.10 level are printed in gray. A blank cell in the table indicates that the correlation coefficient was not significant at a probability level of 0.10 or lower. Cells shaded gray indicate pairs of indices that were not considered relevant, based on impact hypothesis and/or life history considerations.

34

Table B-20. Alewife Correlation coefficients significant at the 0.05 level are printed in black and correlation coefficients significant at the 0.10 level are printed in gray. A blank cell in the table indicates that the correlation coefficient was not significant at a probability level of 0.10 or lower. Cells shaded gray indicate pairs of indices that were not considered relevant, based on impact hypothesis and/or life history considerations.

35

Table B-21. Blueback Herring Stressor Response Indian Point Yearclass Metric Entrainment Zebra Striped Bass Temperature Mortality Mussels Predation (CMR)

PYSL-to-YOY Survival PYSL q' Abundance -0.56 -0.70 YAbundans -0..318 -0+45 Growth Rate Downriver of rkm 100 V Downriver i N of rkm 100 Yearclass +0.84 +0.84 Correlation coefficients significant at the 0.05 level are printed in black and correlation coefficients significant at the 0.10 level are printed in gray. A blank cell in the table indicates that the correlation coefficient was not significant at a probability level of 0.10 or lower. Cells shaded gray indicate pairs of indices that were not considered relevant, based on impact hypothesis and/or life history considerations.

36

Table B-22. Blueback Herring Response Metric

Response

PYSL-to- Yearclass Metric PYSL YOY YOY Abundance Abundance Survival PYSL-to-YOY Survival PYSL

-0.70 Abundance YOY

-0.45 Abundance Yearclass -0.70 -0.45 Correlation coefficients significant at the 0.05 level are printed in black and correlation coefficients significant at the 0.10 level are printed in gray. A blank cell in the table indicates that the correlation coefficient was not significant at a probability level of 0.10 or lower. Cells shaded gray indicate pairs of indices that were not considered relevant, based on impact hypothesis and/or life history considerations.

37

Table B-23. Bay Anchovy Stressor Response Indian Point Yearclass Metric Entrainment Zebra Striped Bass Mortality Mussels Predation Temperature (CMR)

PYSL-to-YOY Survival PYSLAbundnc Abundance YOY ,* *-0.53 YOY

• N k Growth I .FIR, .

Rate -N..*i Downriver  !,Ni *

• of rkm 100 * ***

Downriver * **i,*** '

ofrkm 100 i *; ...

Yearclass +0.84 +0.84 Correlation coefficients significant at the 0.05 level are printed in black and correlation coefficients significant at the 0.10 level are printed in gray. A blank cell in the table indicates that the correlation coefficient was not significant at a probability level of 0.10 or lower. Cells shaded gray indicate pairs of indices that were not considered relevant, based on impact hypothesis and/or life history considerations.

38

Table B-24. Bay Anchovy Response Metric Response Yearclass Metric PYSL-to-PYSL YOY YOY Survival Abundan .ce Abundance FIYSL-to- 21 YOY Survival PYSL Abundance YOY Abundance ý'PF AM Ilk- kA "i Yearclass Correlation coefficients significant at the 0.05 level are printed in black and correlation coefficients significant at the 0.10 level are printed in gray. A blank cell in the table indicates that the correlation coefficient was not significant at a probability level of 0.10 or lower. Cells shaded gray indicate pairs of indices that were not considered relevant, based on impact hypothesis and/or life history considerations.

39

Table B-25. Spottail Shiner Correlation coefficients significant at the 0.05 level are printed in black and correlation coefficients significant at the 0.10 level are printed in gray. A blank cell in the table indicates that the correlation coefficient was not significant at a probability level of 0.10 or lower. Cells shaded gray indicate pairs of indices that were not considered relevant, based on impact hypothesis and/or life history considerations. ,1 40

Table B-26. Spottail Shiner Response Metric

Response

Metric Yearclass Egg-to-YOY PYSL YOY Survival Abundance Abundance Egg-to-YOY

+0.40 Survival PYSL Abundance YOY Abundance Yearclass +0.40 Correlation coefficients significant at the 0.05 level are printed in black and correlation coefficients significant at the 0.10 level are printed in gray. A blank cell in the table indicates that the correlation coefficient was not significant at a probability level of 0.10 or lower. Cells shaded gr'ay indicate pairs of indices that were not considered relevant, based on impact hypothesis and/or life history considerations.

41

APPENDIX C 1

2 3 Potential Effects of Striped Bass Predation on 4 Juvenile Fish in the Hudson River 5

6 7

8 9

10 Douglas G. Heimbuch 11 AKRF, Inc.

12 Hanover, Maryland, U.S.A. 21076 13 dgheimbuch@verizon.net 14 15 16 I

17 Abstract 18 This study addressed the question of whether the increase in striped bass (Morone 19 saxatilis)abundance in the Hudson River that began after 1990, and the associated increase in 20 predatory demand, could have been responsible for observed declines in juvenile abundance of 21 river herring (i.e., blueback herring (Alosa aestivalis)and alewife (Alosapseudoharengus)),

22 Atlantic tomcod (Microgadustomcod) and white perch (Morone americana),and the apparent 23 decline in juvenile survival of striped bass, in the Hudson River. Seasonal (August through 24 October) predatory demand of Hudson River striped bass (ages 1 through 13) was estimated to 25 have increased from an average of 3.4 million kg yrl for the period 1982-1990 to an average of 26 15.0 million kg yr-1 for the period 1991-2004. Juvenile river herring average abundance declined 27 60% since 1990, juvenile Atlantic tomcod average abundance declined 69%, juvenile white 28 perch average abundance declined 59%, and juvenile striped bass survival declined 87%. It was 29 estimated that the observed declines in juvenile abundance and the apparent decline in striped 30 bass juvenile survival could be explained by the increase in striped bass predatory demand if: 1) 31 3.3% of the seasonal predatory demand of age 1 through age 13 Hudson River striped bass was 32 satisfied by consumption ofjuveniles of the four taxa, or 2) 11.1% of the seasonal predatory 33 demand of age 1 and age 2 Hudson River striped bass was satisfied by consumption of juveniles 34 of the four taxa. Historical information on the fraction of the Hudson River striped bass stock 35 that inhabits the Hudson River from August through October, combined with historical 36 information on dietary preferences of Hudson River striped bass, appear consistent with these 37 levels of consumption.

2

38 39 Introduction

40 Background

41 The Atlantic coast population of striped bass (Morone saxatilis)experienced a major 42 increase in abundance over the past decade in response to changes in fishery regulation (Richards 43 and Rago 1999). The average biomass of the population (age I and older) increased over five-44 fold from 16,800,000 kg to 87,900,000 kg for the period 1983-1990 to the period 1991-2004 45 (ASMFC 2005). The increase in abundance of the population raised concerns that the predatory 46 demand of the restored stock might deplete stocks of some forage species (Hartman 2003, 47 Uphoff 2003, and Savoy and Crecco 2004).

48 In the Hudson River, one of three major spawning estuaries of the Atlantic coast 49 population of striped bass (ASMFC 2005), the abundances ofjuvenile blueback herring (Alosa 50 aestivalis)and alewife (Alosapseudoharengus),collectively referred to as river herring, and 51 Atlantic tomcod (Microgadustomcod) and white perch (Morone americana)have declined since 52 about 1990 (Central Hudson Electric and Gas Corporation et al. 1999 and Hurst et al. 2004).

53 During the same period, striped bass juvenile abundance has remained fairly stable while the 54 abundance of larval striped bass abundance has increased substantially. White perch, river 55 herring and striped bass spawning occurs in late May and June in the Hudson River. Juvenile 56 striped bass, white perch and river herring are collected by beach seines from late July through 57 October (Central Hudson Electric and Gas Corporation et al. 1999). Atlantic tomcod hatching 58 occurs in late February and early March (Dew and Hecht 1994), and juveniles are present by late 59 April (Central Hudson Electric and Gas Corporation et al. 1999). These five species comprised 3

60 85% of the average catch of estuarine and diadromous species collected by beach seines from 61 1980 through 2000 (Hurst et al. 2004).

62 Pre-spawning striped bass enter the lower Hudson River estuary in mid- to late fall and 63 overwinter in the lower Hudson River (McLaren et al. 1981, and Clark 1968). In April, adult 64 striped bass, including some immature fish, begin to migrate to the upriver spawning grounds 65 (Bear Mountain Bridge (river km 74) to Newburgh-Beacon Bridge (river km 98)), often with 66 immatures migrating first followed by older mature fish (McLaren et al. 1981). The peak period 67 of spawing is typically between April and May After spawing, most adult striped bass migrate to 68 the lower river and then out of the river to the Atlantic coast (McLaren et al. 1981). However, 69 some portion of the adult population remains in the river, perhaps year-round (Secor and Piccoli 70 1996). Recaptures of tagged age 2 (immature) striped bass in the Hudson River have been 71 reported in each month, April through November, and in each year, 1987 through 1992 (Dunning 72 et al. 2006), providing positive evidence of their presence in the river through the fall.

73 The historical commercial fishery for striped bass in the Hudson River was open from 74 May through November prior to its closure after 1975 over concerns of PCB contamination 75 (McLaren et al. 1988). Commercial fishing generally was conducted with gill nets from the 76 George Washington Bridge (river km 19) to Hudson, NY (river km 181). In 1976, 1977 and 77 1978, a study was conducted to simulate the commercial fishery from April through June with 78 three commercial fishers fishing two days per week each week. The catch rate of striped bass 79 greater than 250 mm declined each month from an average of 659 fish in April, to an average of 80 342 fish in May, to an average of 258 in June (Texas Instruments 1980), indicating that perhaps 81 as much as 39% of the adult stock were still present in the river in June. A 2001 recreational 82 fishery survey of the Hudson River (Normandeau Associates, Inc. 2003) estimated striped bass 4

83 catch per unit effort (CPUE) for shore-based fishing of 13.5 (fish per 100 angling hours) in 84 spring (mid-March through mid-June) and 3.3 in late summer (August through September),

85 suggesting that late summer abundance could have been 24% of the spring abundance. Shore-86 based fishing was the predominant fishing mode in the portion of the Hudson River downriver of 87 the striped bass spawning grounds. That study also estimated striped bass harvest (mean total 88 length of 727 mm) per unit effort (HPUE) for shore-based fishing of 1.1 (fish per 100 angling 89 hours0.00103 days <br />0.0247 hours <br />1.471561e-4 weeks <br />3.38645e-5 months <br />) in spring and 0.2 in late summer, suggesting that late summer abundance of larger striped 90 bass could have been 18% of the spring abundance. In fall (October through November) the 91 shore-based fishing CPUE for striped bass increased to 29.9 (fish per 100 angling hours) and the 92 HPUE increased to 1.1, possibly due to the arrival of over-wintering pre-spawners.

93 Hudson River striped bass in their first year of life are primarily consumers of 94 invertebrates but become largely piscivorous during their second year of life (Walter et al. 2003, 95 and Gardinier and Hoff 1982), at which time they grow to exceed 200 mm (Texas Instruments 96 1980). Stomach content studies of adult striped bass in the Hudson River were conducted in 97 1974, 1976 and 1977 (Gardinier and Hoff, 1982) and from 1990 through 2006 (Kahnle and 98 Hattala, 2007). In 1976 and 1977, 380 striped bass from 200 mm to over 800 mm were collected 99 with a 900 foot haul seine in April and May; 102 contained recognizable food items. In 1974, 100 317 striped bass (including 13 between 200 mm and 275 mm) were collected with beach seines 101 and otter trawls from April through November. The only recognizable finfish present in 102 stomachs of striped bass larger than 200 mm were Atlantic tomcod, white perch, striped bass, 103 spottail shiner and unidentified clupeids (likely blueback herring, alewife and American shad, 104 which are common in the Hudson River). From 1990 through 2006 stomach contents of 1859 105 mature striped bass (modal length 659-700 mm TL) were examined, 89% of which were 5

106 collected in the spring. Approximately 15% of the stomachs from spring collected striped bass 107 contained food items, and 33% of stomachs from the fall and summer collected striped bass 108 contained food items. The dominant food items were unidentified fish (35.5%), crabs (16.10/6),

109 herring (18.1%), Atlantic menhaden (4.6%), isopods (4.3%) and white perch (3.6%). A stomach 110 content study conducted in winter months of 1991-1992 (with water temperature less than 100 C) 111 collected 137 striped bass larger than 200 mm (Dunning et al. 1997). The primary finfish 112 identified were blueback herring, clupeids, white perch, and striped bass.

113 Objective and Analysis Approach 114 The objective of this study was to determine whether the increase in predatory demand of 115 Hudson River striped bass, accompanying the increase in abundance of the recovered striped 116 bass stock, could have been responsible for the observed changes in abundance of juvenile 117 Atlantic tomcod, river herring, white perch and striped bass. The approach used to address this 118 objective was developed in response to the availability of relevant historical data. Estimates of 119 year- and age-specific abundances (age 1 through age 13+) and instantaneous mortality rates for 120 the coastwide striped bass stock from 1982 through 2004, and an estimate of the fractional 121 contribution of Hudson River striped bass to the coastwide stock, were available from the 122 Atlantic States Marine Fisheries Commission ("ASMFC") stock assessment (ASMFC 2005).

123 Estimates of the annual abundance of larval and juvenile life stages of the five species in the 124 Hudson River for 1977 through 2004 were available from a series of annual reports referred to as 125 Hudson River Year Class Reports (e.g., EA 1996), which document sampling results from the 126 Hudson River Monitoring Program ("HRMP") funded by electric generators on the Hudson 127 River. Season- and age-specific estimates of abundance of age 1 and older striped bass 128 inhabiting the Hudson River were not available for the period of interest. Furthermore, with the 6

I

129 exception of the studies cited in a previous paragraph, season- and age-specific characterizations 130 of diets of Hudson River striped bass also were not available.

131 The analysis approach contained four steps. The first was the development of a method 132 that would be supported by the available data for estimating instantaneous mortality rates that 133 might be due to predation. Existing multispecies virtual population analysis methods and 134 ecosystem balancing methods (Magnusson 1995, Whipple et al. 2000, and Christensen et al.

135 2005), which can generate separate estimates of mortality rate due to predation, were not selected 136 due to their extensive data requirements. The second step was the estimation of the changes in 137 juvenile abundances for two stanzas of years (1977 to 1991 was referred to as Period 1, and 1991 138 to 2004 was referred to as Period 2), and estimation of the changes in annual predatory demand 139 of Hudson River striped bass for the two stanzas of years. Over the 28 years of interest, August 140 through October has been the consistent sampling season for juvenile fish by the HRMP; 141 therefore, estimates of juvenile abundance were restricted to that three month season. These 142 estimates of change, expressed in terms of ratios, were used as the primary inputs to the analysis.

143 The third step was estimation of the instantaneous mortality rates that might be due to predation.

144 The final step was a comparison of the potential juvenile biomass consumed by striped bass 145 predation (kg yr'1), which was computed using the estimated mortality rates for possible 146 predation, to the estimated predatory demand of Hudson River striped bass. The purpose of the 147 final step was to confirm that the magnitude of predation required to produce the observed 148 change in juvenile abundance was no greater than the predatory demand of Hudson River striped 149 bass.

150 To address the possibility that different age classes of striped bass might exert different 151 levels of predation on juvenile fish in the Hudson River, the assessment was conducted 7

152 separately for two age groups of possible predators: ages 1 through 13 striped bass, and age 1 153 and age 2 striped bass only. Secor and Piccoli (1996) found evidence of size-dependent 154 dispersion of striped bass from the Hudson River with male age 2 striped bass spending most of 155 their year in mesohaline portions of the estuary.

156 Methods and Data 157 UnderlyingSystem ofEquations 158 For the purpose of estimating instantaneous mortality rates that were possibly due to 159 predation in the two periods, three ratios were defined. Ratios (of a variable in Period 2 to the 160 same variable in Period 1) were selected as the basic inputs to the analysis because scaling 161 factors that are common to the two periods (e.g., gear efficiency) would cancel out in ratios; this 162 can help eliminate possible biases that otherwise could arise due to possible errors in specifying 163 those scaling factors. Because the focus of the study was the overall change in predatory 164 demand and juvenile abundance between the two periods, and not detailed inter-annual 165 variability, the underlying system of equations was defined in terms of average conditions (rates) 166 for each period.

167 168 The first ratio was the potential change in juvenile biomass consumed by striped bass, 169 defined as a ratio of average biomass possibly consumed in Period 2 (C 2 ) to the average 170 possibly consumed in Period 1 (C1 ):

171 R, = . (1) 8

172 The second was the change in average juvenile abundance, defined as the ratio of average 173 abundance during the juvenile sampling season (i.e., August through October) in Period 2 (N 2) 174 to the average in Period 1 (N1 ):

175 R,=N2.N, (2) 176 The third was the change in the number of fish entering the juvenile life stage, defined as the 177 ratio of the average number entering the juvenile life stage in Period 2 (L 2 ) to the average in 178 Period 1 (l):

179 RI =_LLI (3) 180 The three ratios were expressed in terms of the mortality rates of interest through the 181 following standard equations from fishery science (Ricker 1975). For each period, the annual 182 seasonal consumption of juvenile biomass by predation (which is directly analogous to the 183 fishery yield) in periodj was defined as:

t

• i - e~~(gj " -*l) 184 C1 = impJe - J t'J (4) 185 where gj is the daily growth rate during the season; rnj is the background daily mortality rate (i.e.

186 all mortality except mortality due predation); mpj is the additional daily mortality rate due to 187 predation during the season; and t is the duration of the season (days). The biomass at the 188 beginning of the season, Bj, was defined as:

189 B,=w1 N. (m + mPJt (5) 190 where wj is the weight per fish at the beginning of the season. The average annual juvenile 191 abundance during the sampling season was defined as:

9

192 N = f.e-(mJ+mp.J)' Il-e' m- (6)

• ~ ~(mi + m oJ~t(6 193 where m'j is the background daily mortality rate from the beginning of the juvenile life stage' to 194 the beginning of August, Lj is the average abundance at the beginning of the juvenile life stage 195 during periodj, and t' is the duration (days) from the beginning of the juvenile life stage to the 196 beginning of the juvenile sampling season.

197 Combining equations (1) through (6) gives the following two equations which form the 198 basis for the analysis:

199Rc((in 2t)w2 (n 2 +mrn,2 )t 1-(gz -in-", )

- 2)t I (m2 +mp' - - 2 1.99 R,(mp2ý2Ie(2+p21 +2 pM L/2 (7)

R, ( ( m + m, )t I -e-1-e(mt+mP.I)l (m) + M,, -g,)t 1

200 and 201 R- =m 2 + mp,2 (8)

R, e-(m'I +mp'), I -e(M',_ '

(mI +m , )t 202 The right hand sides of equations (7) and (8) contain only underlying rates (and initial weight per 203 fish for equation (7)), and the left hand side of the equations contain the measurable quantities.

204 Approximations 205 Estimates of the instantaneous mortality rates due to possible predation for Period 1 and 206 Period 2 can be identified through an exhaustive search (by computer) for values of rnp and mp,2 207 that satisfy the non-linear equations (7) and (8), given input values for the two ratios of ratios and 208 estimates for the growth rates and background mortality rates. Alternatively, equations (7) and 10 II

209 (8) can be linearized, and approximate closed-form solutions for mpa, and mp,2 can be derived (see 210 Appendix A). The closed-form solutions provide a more convenient method for conducting the 211 analysis and also provide a basis for developing variance estimates (see Appendix B).

212 The approximation for the ratio of ratios in equation (7) is:

213 R, _, mp 2 (

214 where a is the ratio (Period 2 to Period 1) of the average juvenile weight per fish at the mid-215 point of the season. The logarithm of the ratio of ratios in equation (8) is approximately:

216 In R (m- ,-mp, 2 t1+ 2 +'8 (10) 217 where 8 is the difference between the juvenile background mortality rates for Period 1 and 218 Period 2.

219 Combining equations (9) and (10) provides approximate solutions for the potential 220 predation mortality rates in the two periods expressed in terms of functions of the two ratios of 221 ratios:

222 min- (11) 223 and 224 mp, 2 -R '" (12) 225 11

226 Changesin Juvenile and LarvalAbundances 227 The ratio of abundances of post yolk-sac-larvae (Table 2) was used as a surrogate for the 228 ratio of abundance of fish entering the juvenile life stage (RI) because field data on the number of 229 fish entering the juvenile life stage were not available. Average abundance indices for post yolk-230 sac larvae were computed as the average of weekly standing crop estimates from Hudson River 231 Year Class Reports. Weekly standing crop estimates for post yolk-sac larvae were based on data 232 collected by the HRMP's Longitudinal River Survey ("LRS") which sampled with 1 mi 233 ichthyoplankton nets attached to epibenthic sleds (to sample the bottom stratum) and Tucker 234 trawls (to sample the mid-water stratum). Annual abundance indices based on LRS data were 235 computed for 1977 through 2004, based on data from stratified random sampling from the 236 George Washington Bridge north to the Federal Dam at Troy, NY during May and June.

237 Alewife and blueback herring were treated as a single taxonomic group (river herring) because 238 they could not be reliably identified to species as post yolk-sac larvae.

239 The ratios of average abundances (Rn) ofjuvenile river herring, Atlantic tomcod, white 240 perch and striped bass (Table 3) were based on annual indices of juvenile abundance. Annual 241 juvenile abundance was computed as the average of weekly standing crop estimates from 242 Hudson River Year Class Reports (e.g., EA 1996). Weekly standing crop estimates for juvenile 243 fish inhabiting the beach zone of the Hudson River were based on data collected by the HRMP's 244 Beach Seine Survey ("BSS"), which sampled with 100 ft beach seines from the George 245 Washington Bridge to the Federal Dam at Troy, NY. Weekly standing crop estimates for 246 juvenile fish inhabiting the shoals, bottom and channel of the Hudson River were based on data 247 collected by the HRMP's Fall Shoals Survey ("FSS"), which sampled with beam trawls (to 12

248 sample the bottom stratum) and Tucker trawls (to sample the mid-water stratum) from the 249 George Washington Bridge to the Federal Dam at Troy, NY.

250 Annual abundance indices based on BSS data were computed for 1977 through 2004, 251 using data from biweekly sampling in August through October. Annual abundance indices based 252 on FSS data were computed for 1985 through 2004, using data from biweekly sampling in 253 August though October. The FSS was conducted from 1979 to 1984; however, beam trawls 254 replaced epibenthic sleds for sampling the bottom and shoal strata in 1985. To avoid possible 255 confounding effects of the gear change, FSS data prior to 1985 were not included in the analysis.

256 However, because BSS and FSS indices of abundance (1985-2004) were significantly correlated, 257 juvenile abundance indices for a given species from the BSS from 1979 through 1984 were used 258 to predict FSS abundance indices (as if beam trawl sampling had occurred in those years) for the 259 years prior to 1985.

260 For each species, annual average (August through October) juvenile abundance estimates 261 (Table 3) were computed by adjusting the annual average standing crop estimates from the BSS 2'62 and FSS for gear efficiency and summing the resulting abundance estimates:

263 - Assy + AFssy (13) qBss qFss 264 where ABss~ y and 4ssy are the reported average (August through October) standing crop 265 estimates from the two programs for year y, and qBss and qFss are gear efficiencies for the two 266 sampling programs. Gear efficiency estimates used for this computation are those reported in 267 Central Hudson Electric and Gas Corporation et al. (1999), which were based on gear efficiency 268 studies (Normandeau Associates Inc. 1984, Kjelson and Colby 1977, and Loesch 1976) and on 269 comparisons of striped bass BSS catch rates to striped bass mark-recapture estimates of 13

270 abundance. For the BSS, the gear efficiency was assumed to be 4%; and for the FSS, the gear 271 efficiency was assumed to be 8.85, the average of the reported beam trawl gear efficiency (15%)

272 and the reported Tucker trawl gear efficiency (2.7%).

273 The estimates of juvenile abundance computed as described above are generally 274 consistent with other estimates reported in the literature. Young et al. (1988) reported estimates 275 ofjuvenile white perch abundance in the Hudson River based on mark-recapture studies from 276 1974 through 1979. The estimates ranged from 13 million to 205 million with and average of 74 277 million. The estimated average juvenile white perch abundance for Period 1 of 65.5 million 278 from this study is consistent with those mark-recapture estimates. McLaren et al. (1988) 279 reported mark-recapture estimates of abundance for one year old (roughly mid-February) 280 Hudson River Atlantic tomcod for 1975 to 1980 which ranged from 2.5 to 8.9 million, with an 281 average of 5.8 million. To be consistent with the Period 1 estimate (Table 2) of 54 million 282 juveniles, the mortality rate from mid-September to mid-February would have to be 283 approximately Z=2.2 (5 months). Although estimates of survival rates for juvenile Hudson River 284 Atlantic tomcod could not be found in the literature, McLaren et al. (1988) reported annual 285 mortality rates from age I to age 2 for Atlantic tomcod. The average for 1975 through 1979 was 286 Z=2.8 (12 months), which is not inconsistent if both the difference in age and the difference in 287 duration are considered.

288 Changesin PredatoryDemand 289 For the purpose of assessing whether the change in predatory demand could have been 290 responsible for the observed changes in juvenile abundance, the ratio of potential consumption of 291 juvenile biomass by striped bass (Re) was assumed to be the same as the ratio (Period 2 to Period 292 1) of predatory demands of striped bass:

14

293 RP = _

H2 (14) 294 where Hi is the average of annual estimates of predatory demand during periodj.

295 Estimates of the annual predatory demand exerted by the Hudson River stock were based 296 on estimates of annual production by the Hudson River stock and an assumed trophic efficiency 297 between striped bass and their prey. Age-specific estimates of annual production, H,,, (kg yr1),

298 of age-I and older striped bass were based on the production formulation from Ricker (1975):

299 H ay =GayBay - G yGB(15)'Y-z (Zay -G,y (15) 300 where SBay is the estimated abundance of age a striped bass in year y, Wa is the average weight 301 of age a striped bass at the beginning of the year, Gay is the annual growth rate for age a striped 302 bass in year y, and Zay is the annual mortality rate for age a striped bass in year y. Annual 303 predatory demand, Pay, was estimated by dividing annual production by trophic efficiency, 304 assumed to be 10% (Pauly and Christensen 1995, Jennings and Mackinson 2003, and Jennings et 305 al. 2002).

306 -Estimates of the coastwide abundance of age 1 through age 13 striped bass for 1982 to 307 2004 (SBay) were from the 2005 Stock Assessment (Table 18a, ASMFC 2005). Because striped 308 bass post yolk-sac larval abundance (an indicator of spawning stock abundance) was relatively 309 stable from 1977 through 1990, the average age-specific abundances from 1982 through 1990 310 were assumed to be representative of the averages for all years in Period 1 (1977 through 1990).

311 For each age class (age 1 and older) and year the total striped bass mortality rate (Zay) was 312 computed as the sum of reported age- and year-specific fishing mortality rate (Table 16, ASMFC 313 2005) and a constant natural mortality rate of 0.15 (ASMFC 2005). The fraction of the 314 coastwide abundance of striped bass that was of Hudson River origin was assumed to be 13%

15

315 (ASMFC 2005). Age- and year-specific annual growth rates (Ga,y) were estimated from reported 316 average weights at age (Table 13, ASMFC 2005) assuming approximately exponential growth 317 (Ricker 1975) over successive two-year intervals:

318 Gay = 0.5 InW__+,y+I (16) 319 where W*y is the reported average weight for age a striped bass in year y, and the initial weight 320 for each age group and year, Wa,y ,was estimated as:

321 ,= Gay (17) 322 Estimates of coastwide predatory demand of striped bass computed using these methods 323 (Table 1) are consistent with other published estimates. Hartman (2003) estimated the annual 324 coastwide predatory demand of the striped bass population to be 17.9 mt in 1982 and 147.9 mt in 325 1995. His estimates were based on age- and year-specific coastwide striped bass abundance and

-326.- survival estimates from ASMFC (2000). Using those same inputs and the methods described 327 above for this study, the estimates of coastwide predatory demand of striped bass are 17.3 mt in 328 1982 and 135.7 mt in 1995. The estimates listed in Table 1 used updated abundance and survival 329 estimates from ASMFC (2005), which account for the difference in comparison to Hartman's 330 estimates. Uphoff (2003), also using ASMFC abundance estimates from 2000, estimated the 331 annual coastwide potential consumption of Atlantic menhaden by striped bass to be 26 mt in 332 1982-1983, and 190 to 200 mt from 1994 to 1998.

333 The seasonal pattern of predatory demand by striped bass was characterized based on 334 average monthly water temperatures in the Hudson River and the consumption component of a 16

335 bioengergetics model for striped bass (Hartman and Brandt 1995). The fraction of the annual 336 consumption (,r) that occurred from August through October was approximated as:

10 ZCR, 337 n8 m=

12 (18)

Z CR,,,

338 where CRin is the predicted consumption rate (gm gm" day-) for the average water temperature 339 in month m. This approximation does not account for possible month-specific variability in 340 growth and mortality rates of striped bass. Estimates of month-specific water temperature, 341 required for the bioenergetics model of the seasonal pattern of consumption, were from 342 Poughkeepsie Water Works data (Table B-4, EA 1996). The consumption from August through 343 October was estimated to be 41.8% of the annual total. The average seasonal predatory demand 344 (Table 1) for each period was estimated as the product of the average annual predatory demand 345 for the period and the fraction of the annual consumption that occurred from August through 346 October.

347 Estimationof InstantaneousMortalityRates Due to PossiblePredation 348 Instantaneous mortality rates for possible predation, that were consistent with the 349 estimated ratios (R,, Rb, Rp), were identified through exhaustive search (by computer) of 350 candidate values of mp, and mp,2 using equations (7) and (8). Because the question being 351 addressed was whether the increase in striped bass predation could have caused the observed 352 changes in juvenile abundance, all other things being equal, background mortality rates, growth 353 rates, and initial weights were assumed to have remained the same for the two periods.

354 Approximate estimates also were computed using the equations (11) and (12); and for the reason 355 noted above, the parameter a was set equal to 1, and the parameter f8 was set equal to 0.

17

356 Variance estimates for the approximations were computed using the methods described in 357 Appendix B.

358 Estimation ofPotentialConsumption ofJuvenile Biomass 359 The potential juvenile biomass consumed by striped bass was computed using equation 360 (4) with the estimates of instantaneous mortality due to potential predation and the estimates of 361 average seasonal juvenile abundance. Also required for estimating potential juvenile biomass 362 consumed by striped bass were estimates of daily background mortality rates and growth rates of 363 the juvenile fish, and initial weights of the juvenile fish.

364 For each species, the background daily mortality rates (Table 4) for the three month 365 sampling season (August through October) were estimated as a power function of dry weight 366 (Peterson and Wroblewski, 1984):

367 m = 0.00525(0.2we 0.25 (19) t 1 368 where dry weight is assumed to be 20% of wet weight (Peterson and Wroblewski, 1984).

369 Similarly, the background daily mortality rate for the interval from the start of the juvenile life 370 stage to August was estimated as:

371 m= 0-.00525(0.2weg) 0 2 5 (20) 372 The duration of the juvenile sampling season (t) was set to 90 days (August through October),

373 and (based on life history considerations discussed in the Introduction) the interval from the 374 beginning of the juvenile stage to the beginning of the juvenile sampling season (t) was set to 15 375 days for white perch, river herring and striped bass, and set to 90 days for Atlantic tomcod.

18

376 For each species, the daily juvenile growth rate through October (Table 4) was estimated 377 from the beginning and ending weights, assuming approximate exponential growth during that 378 interval, as (Ricker 1975):

379 In(wmd) g= nwta" D (21) t+tl 380 and the weight of species s at the beginning of August (Table 4) was estimated as:

381 w = w., ar~eg' (22) 382 Estimates of the average weight per fish at the beginning and end of the juvenile life stage were 383 derived from reported lengths and length-weight relationships. For river herring, the lengths at 384 the beginning and end of the juvenile stage were set to 25mm and 92mm (Mullen et al. 1986),

385 respectively, and the length-weight relationship was from PSEG (2006). For Atlantic tomcod, 386 the initial length (for mid-May) and the final length (for the end of October) were set to 25mm 387 and 120mm, respectively, (McLaren et al. 1988); and the length-weight relationship was from 388 Dew andHecht (I 994). For whi te perch; the engtIhs at the beginning and end of the juvenile-389 stage were set to 25mm and 80mm, respectively (Texas Instruments 1980); and the length-390 weight relationship was from Klauda et al. (1988). For striped bass, lengths at the beginning and 391 end of the juvenile stage were set to 30mm and 95mm, respectively (Dey 1981); and the length-392 weight relationship was from Fay et al. (1983).

393 Sensitivity Analysis to Address Assumptions 394 A sensitivity analysis was conducted to address: 1) the possible effects of density 395 dependent mortality occurring between the larval and juvenile life stages, 2) the effects of 396 possible errors in the estimation of background mortality rates on the predicted juvenile biomass 19

397 'to predation, and 3) an alternative assumption regarding the fraction of the coastwide stock that 398 was from the.Hudson River. Other input parameters, which did not require formal sensitivity 399 analyses, but which could affect results are discussed at the end of this section.

400 For Atlantic tomcod, river herring and white perch, the historical data indicated a decline

.401 in larval abundance from Period I to Period 2, and for striped bass an increase was indicated.

402 The results presented above assume the ratio of abundance (Period 2 to Period 1) of fish entering 403 the juvenile life stage is the same as the ratio of larval abundance. However, if density 404 dependent effects were present, the ratio of abundance of fish entering the juvenile stage could 405 have been closer to unity. To address this possibility, the analyses were re-run with values for 406 the ratio of abundance of fish entering the juvenile stage (RI) ranging from the estimated value 407 (rn) based on post yolk-sac larval abundances to a value of RI=I (i.e. constant recruitment to the 408 juvenile life stage). An index of the degree of density dependent effects (1)was defined as:

409 1- (R, - r,) (23)

(1- rt) 410 with arange from 0 (for R =r,) to I (for RI=1).

411 The equation used to estimate the background mortality rate for juvenile fish (equations 412 (19) and (20)) is a theoretically derived relationship for pelagic marine ecosystems (Peterson and 413 Wroblewski 1984). Other authors (e.g. McGurk (1993), Lorenzen (1996) and Houde (1997))

414 have reported natural mortality rates of fish in marine and other ecosystems also as power 415 functions of weight, but with empirical estimates for the coefficients that differ somewhat from 416 those of Peterson and Wroblewski (1984). To address the effects of possible errors in the 417 assumed background mortality rate, the analyses were re-run with the background mortality rates 418 set to 0 and with the background mortality rates set to 2 times of the initial estimates.

20

419 Estimates of the coastwide abundance of age 1 striped bass, combined with indices of 420 juvenile abundance from the major spawning areas of striped bass (ASMFC 2005) indicate that 421 the proportion of the coastwide population of age 1 striped bass that is from the Hudson River 422 has changed from Period 1 to Period 2 (see Appendix C). The average estimated contributions 423 from the Hudson River for Periods 1 and 2 are 20.9% and 8.9% respectively. Assuming these 424 proportions apply to age 1 and age 2 striped bass, then the ratio of predatory demands (Rp) for 425 age 1 and age 2 striped bass would decline from 3.44 (Table 1) to 1.46. To address the effects of 426 this alternative assumption regarding the contribution of Hudson River striped bass to the 427 coastwide stock, the analyses were re-run the analysis with the alternative estimate for Rp for age 428 1 and age 2 striped bass.

429 Other input parameters of concern were the trophic conversion efficiency, the fraction of 430 the annual predatory demand exerted during the three month fall season, and gear efficiencies.

431 Selection of alternative values for these parameters would not affect estimates of instantaneous 432 mortality rates possibly due to predation because, as noted above, the inputs to the analyses are 433 ratios in which scaling factors that are common to both periods cancel out. However, if one of 434 these factors varied substantially between the two periods, then the degree of change in that 435 factor would determine the effect on estimates of instantaneous mortality rates possibly due to 436 predation.. The possible effects of changes in these factors between the two periods were viewed 437 as second order considerations for this study; and therefore, sensitivity analyses of those possible 438 changes were not undertaken.

439 Because the estimates of juvenile biomass possibly consumed by predation use these 440 input parameters directly (not in ratios) estimates of juvenile biomass possibly consumed by 441 predation would be affected by assumed gear efficiencies. A change of the assumed gear 21

442 efficiency (e.g. doubling) would cause an inversely proportional change (i.e., halving) of the 443 estimate ofjuvenile biomass possibly consumed. Similarly, a change of the assumed trophic 444 conversion efficiency (e.g. doubling) would cause an inversely proportional change (i.e.,

445 halving) of the estimate of predatory demand. A change of the assumed fraction of the annual 446 predatory demand exerted during the three month fall season (e.g., doubling) would cause a 447 directly proportional change (i.e., doubling) of the estimate of predatory demand. Because the 448 sensitivities of the estimates to these assumptions were clear, no additional analyses were 449 conducted to address them.

450 Results 451 Estimates of InstantaneousMortalityRates PossiblyDue to Predation 452 Estimates of the seasonal instantaneous mortality rates possibly due to predation by 453 striped bass (Tables 5 and 6) were higher for juvenile striped bass than for juveniles of the other 454 three taxa. The estimated rates were slightly higher under the assumption that predation was by 455 age 1 and age 2 striped bass only, than under the assumption that predation was by age 1 through 456 age 13 striped bass. The estimated instantaneous mortality rates for Period 2 were 12 to 15 times 457 higher than for Period I assuming predation was by all age classes; and were 10 to 12 times 458 higher than Period 1 assuming predation by age 1 and age 2 striped bass only. For river herring, 459 Atlantic tomcod and white perch, the estimates based on the approximations were very similar to 460 the estimates based on exhaustive search; however, for striped bass the approximations 461 underestimated the Period 2 rate and overestimated the Period 1 rate. The bias in the 462 approximations for larger mortality rates was expected because the Paloheimo approximation 463 works best with small mortality rates (Paloheimo 1961). Coefficients of variation for the 22

464 estimates (based on the approximate standard errors) were 3-12% for striped bass, 31-39% for 465 river herring, 10-13% for Atlantic tomcod, and 9-14% for white perch.

466 ComparisonofJuvenile Biomass Possibly Consumed by StripedBass to Hudson River 467 StripedBass PredatoryDemand 468 The estimated juvenile biomass possibly consumed by striped bass during the three 469 month season (Tables 7 and 8) was 148,000 kg in Period I and 509,000 kg in Period 2 assuming 470 predation by age 1 and age 2 striped bass only, and was 112,000 kg in Period I and 498,000 kg 471 in Period 2 assuming predation by age 1 through age 13 striped bass. Assuming predation by age 472 1 and age 2 striped bass only, the juvenile biomass possibly consumed by striped bass was

-473 11.11% of the estimated seasonal predatory demand, and assuming predation by age I through 474 age 13 striped bass, the juvenile biomass possibly consumed was 3.33%. Estimated consumption 475 of juvenile striped bass was higher than the estimated consumption of the other three taxa, 476 approximately 2 times higher than river herring, 4 times higher than Atlantic tomcod, and over 5 477 times higher than white perch.

478 Effects of Changesin Assumptions-- Sensitivity Analyses 479 Reducing the assumed background mortality rate had the effect of increasing the 480 estimates ofjuvenile biomass possibly consumed by striped bass (Figures 2 and 3); increasing 481 the assumed background mortality rate reduced the estimates of juvenile biomass'possibly 482 consumed by striped bass. Increases in the assumed degree of density dependent effects up to an 483 index value between 0.5 and 0.75 caused the estimates of the juvenile biomass possibly 484 consumed by striped bass to increase. Further increases in the assumed degree of density 485 dependent effects, with the index increasing to 1, caused estimates of the juvenile biomass 23

486 possibly consumed by striped bass to decrease (Figures 2 and 3). Changing the assumed 487 proportion of the coastwide stock of age 1 and age 2 striped bass from 13% in both periods to 488 20.9% in Period 1 and 8.9% in Period 2 caused estimates of seasonal juvenile biomass possibly 489 consumed by striped bass to increase. For Period 1 the estimate increased from 148,000 kg to 490 409,000 kg, and for Period 2 the.estimate increased from 509,000 kg to 600,000 kg.

491 Considering the combined effects of alternative assumptions for background mortality 492 rates and degree of density dependent effects, estimates of the percent of seasonal predatory 493 demand potentially satisfied by consumption of juveniles of the four taxa were less than 18% for 494 predation by age I and age 2 striped bass only, and were less than 6% for predation by age 1 495 through age 13 striped bass. Under the assumption that 20.9% (in Period 1) and 8.9% (in Period

___ 496 2)_ of the coastwide stock of age 1 and age2 striped bass were Hudson River fish, the maximum 497 estimate of the percent of seasonal predatory demand potentially satisfied by consumption of 498 juveniles of the four taxa increased from 18% to 28% (Figure 4).

499. DiscussiOn 500 The percent of the seasonal predatory demand that could be satisfied by juvenile biomass 501 consumed by striped bass has two components: 1) the fraction of the Hudson River striped bass.

502 population that inhabits the river from August through October, and 2) the contribution of the 503 juvenile target species to the diet of striped bass in the river during those months. For example, 504 if 75% of age 1 and age 2 striped bass from the Hudson River stock were present in the river 505 from August through October, and 40% of their diet while in the river was satisfied by juveniles 506 of the target species, then 30% of the predatory demand would be satisfied by those juvenile fish.

507 The estimated percents of seasonal predatory demand that would be needed to explain the 508 observed declines in juvenile abundance appear consistent with what is known about the fraction 24

509 of the stock that inhabits the river in fall, and with what is known about Hudson River striped

• 510 bass dietary preferences. The findings of Secor and Piccoli (1996) demonstrated that some 511 fraction of the adult stock inhabits the river year-round; the simulated commercial fishery study 512 indicated that more than one third of the spawning stock may have remained in the river in June; 513 and the 2001 recreational fishery survey indicated that as much as 18%-24% of the striped bass 514 abundance present in the river during the spring was present in the river by late summer. The 515 available stomach content studies (Gardinier and Hoff 1982, Dunning et al 1997, and Kahnle and 516 Hattala 2007) found clupeids, Atlantic tomcod, white perch, and striped bass among the 517 dominant identifiable food items in age I and older in the Hudson River.

518 This study focused on the decline in juvenile abundance of four forage taxa as measured 519 by sampling that occurred from August through October, and did not explicitly address possible 520 reductions in spawning stock biomass that could have been caused by the reductions in juvenile 521 abundance. However, the data on post yolk-sac larvae river herring, Atlantic tomcod and white 522 perch abundance suggest that a reduction in spawning has occurred for these taxa, which may be 523 due, in part, to the increased mortality during the juvenile stage. The reduction in spawning 524 might also be due to increased mortality in older life stages of these taxa - possibly due, in part, 525 to striped bass predation on age I or older fish. For striped bass, estimates of post yolk-sac larval 526 abundance suggest a six-fold increase in larval abundance from Period 1 to Period 2, which is 527 consistent with the apparent increase in adult abundance. However, the data on striped bass 528 juveniles shows no corresponding increase in juvenile abundance. The analysis presented in this 529 paper demonstrated that striped bass predation alone could have kept the juvenile abundance 530 from increasing. Other possible explanations include a drastic reduction in the juvenile 531 background mortality rate, or density dependent out-migration ofjuveniles.

25

532 The results from this study indicate that the increase in predatory demand of Hudson 533 River striped bass could have been responsible for the decline in juvenile abundance of river 534 herring, Atlantic tomcod and white perch, and responsible for the apparent decline in survival of 535 striped bass from the post yolk-sac larvae to juveniles. The required magnitude of consumption

-536 of juvenile biomass to account for the declines in juvenile abundance appears to be well below 537 the estimated predatory demand of Hudson River striped bass, whether considering all ages, or 538 only age I and age 2 striped bass. The sensitivity analyses suggest this result is fairly robust to 539 possible violations in assumptions and to possible errors in input parameter values. However, a 540 field survey to estimate the biomass of juvenile fish consumed by Hudson River striped bass in 541 the fall would be needed to confirm the proposition that Hudson River striped bass, in fact, were 542 responsible for the declines in juvenile abundance.

26

543 Acknowledgements 544 I gratefully acknowledge the very thoughtful comments and suggestions on the first draft 545 of this manuscript from four anonymous reviewers and an associate editor of the Transactions of 546 the American Fisheries Society. I also acknowledge the many field crews and scientists who, 547 since the mid 1970's, collected the data on age-0 Hudson River fishes that I relied on for this 548 study; and I acknowledge the scientists of the ASMFC its member states for their many years of 549 collecting and analyzing data on the Atlantic coastal striped bass stock, the results of which were 550 key inputs to this study.

551 552 Literature cited 553 Atlantic States Marine Fisheries Commission (ASMFC). 2005. 2005 Stock Assessment Report

...... for Atlantic Striped Bass: Catch-at-Age Based VPA & Tag Release/Recovery Based Survival Estimation. Washington, D.C.

Central Hudson Gas and Electric Corporation, Consolidated Edison Company of New York, Inc.,

New York Power Authority, and Southern Energy New York. 1999. Draft environmental impact statement - state pollutant discharge elimination system permits

.for Bowline Point, Indian Point 2 & 3, and Roseton steam electric generating stations.

Pearl River, New York.

Clark, J. 1968. Seasonal movements of striped bass contingents of Long Island Sound and the New York Bight. Transaction of the American Fisheries Society. 97:320-343.

27

Cristensen, V., C.J. Walters, and D. Pauly. 2005. Ecopath with Ecosim: a user's guide.

Fisheries Centre, University of British Columbia, Vancouver. November 2005 edition.

154 p. Vancouver, British Columbia.

Dew, B.C. and J.H Hecht. 1994a. Hatching, estuarine transport, and distribution of larval and early juvenile Atlantic tomcod, Microgadustomcod, in the Hudson River. Estuaries Vol.

17, No. 2, p. 472-488.

Dew, B.C. and J.H. Hecht. 1994b. Recruitment, growth, mortality, and biomass production of larval and early juvenile Atlantic tomcod in the Hudson River estuary. Transaction of the American Fisheries Society. 123:681-702.

Dunning, D.J., J.R. Waldman, Q.E. Ross, and M.T. Mattson. 1997. Use of Atlantic tomcod and other prey by striped bass in the lower Hudson River estuary during winter. Transactions of the American Fisheries Society. 126:857-861.

Dunning, D.J., J.R. Waldman, Q.E. Ross, and M.T. Mattson. 2006. Dispersal of age-2+ striped bass out of the Hudson River. Pages 287-294 in J.R. Waldman, K.E. Limburg, and D.L.

Strayer, editors. Hudson River fishes and their environment. American Fisheries Society, Symposium 51, Bethesda, Maryland.

EA Engineering, Science and Technology (EA). 1996. 1995 year class report for the Hudson River estuary monitoring program. Prepared for Consolidated Edison Company of New York, Inc. New York, New York.

28

Fay, C.W., R.J. Neves, and G.B. Pardue. 1983. Species profiles: life histories and environmental requirements of coastal fishes and invertebrates (Mid-Atlantic) - striped bass. U.S. Fish and Wildlife Service, Division of Biological Services, FWS/OBS-82/11.8. U.S. Army Corps of Engineers, TR EL-82-4. 36 pp. U.S. Fish and Wildlife Service, National Wetlands Research Center, Washington, D.C., and U.S. Army Corps of Engineers, Waterways Experiment Station, Vicksburg, Mississippi.

Gardinier, M. and T.B. Hoff. 1982. Diet of striped bass in the Hudson River estuary. New York Fish and Game Journal. Vol. 29, No. 2. p 15 2 - 165 .

Hartman K.J. and S.B. Brandt. 1995. Comparative energetics and the development of bioenergetics models for sympatric estuarine piscivores. Canadian Journal of Fisheries and Aquatic Sciences 52:1647-1666.

Hartman, K.J. 2003. Population-level consumption by Atlantic coastal striped bass and the influence of population recovery upon prey communities. Fisheries Management and Ecology. 10:281-288.

Hartman, K.J., and S.B. Brandt. 1995. Predatory demand and impact of striped bass, bluefish and weakfish in the Chesapeake Bay: applications of bioenergetics models. Canadian Journal of Fisheries and Aquatic Science. Vol 52. p.1667-1687.

Hurst, T.P., K.A. McKown, and D.O. Conover. 2004. Interannual and long-term variation in the nearshore fish community of the mesohaline Hudson River estuary. Estuaries. Vol 27, No. 4. p. 659-669.

29

Jennings, S., K.J. Ware, and S. Mackinson. 2002. Use of size-based production and stable isotope analyses to predict trophic transfer efficiencies and predator-prey body mass ratios in food webs. Marine Ecology Progress Series. Vol. 240:11-20.

Kahnle A.W., and K.A. Hattala. 2007. Striped bass predation on adult American shad:

occurrence and observed effects on American shad abundance in Atlantic coastal rivers and estuaries. In American shad stock assessment report for peer review. Volume I.

Stock Assessment Report No. 07-01 (Supplement) of the Atlantic States Marine Fisheries Commission. August 2007. Washington, D.C.

Kendall, M. and A. Stuart. 1977. The advanced theory of statistics. Volume 1. Distribution Theory. 4th Edition. Macmillan Publishing Co., Inc. New York, New York. 472 p.

Kjelson, M.A., and D.R. Colby. 1977. The evaluation and use of gear efficiencies in the estimation of estuarine fish abundance. In M. Wiley (ed.) Esuarine Processes, Vol. II, Circulation, Sediments, and Transfer of Material in the Estuary, p. 416-424. Academic Press, New York.

Klauda, R.J., J.B. McLaren, R.E. Schmidt, and W.P. Dey. 1988. Life History of White Perch in the Hudson River estuary. American Fisheries Society Monograph 4:69-88.

Loesch, H., J. Bishop, A. Crowe, R. Kuckyr, and P. Wagner. 1976. Technique for estimating trawl efficiency in catching brown shrimp (Penaeusaztecus), Atlantic croaker (Micropogon undulates), and spot (Leiostomus xanthurus). Gulf Research Reports 2:29-33.

y 30

Magnusson, K.G. 1995. An overview of the multispecies VPA - theory and applications.

Reviews in Fish Biology and Fisheries, 5:195-212.

McLaren, J.B., J.C. Cooper, T.B. Hoff, and V. Lander. 1981. Movements of Hudson River striped bass. Transaction of the American Fisheries Society. 110:681-702.

McLaren, J.B., Klauda, R.J., Hoff, T.B., and Gardinier, M. 1988. Commercial fishery for striped bass in the Hudson River, 1931-80. In C.L. Smith (ed.) Fisheries Research in the Hudson River. State University of New York Press. Pages89-123. Albany, New York.

McLaren, J.B., T.H Peck, W.P. Dey and M. Gardinier. 1988. Biology of Atlantic tomcod in the Hudson River estuary. American Fisheries Society Monograph 4:102-112.

Mullen, D.M., C.W. Fay, and J.R. Moring. 1986. Species profiles: life histories and environmental requirements of coastal fishes and invertebrates (North Atlantic) -

alewife/blueback herring. U.S. Fish and Wildlife Service. Biological Report (82)(11.56).

U.S. Army Corps of Engineers, TR EL-82-4. 21 pp. U.S. Fish and Wildlife Service, National Wetlands Research Center, Washington, D.C., and U.S. Army Corps of Engineers, Waterways Experiment Station, Vicksburg, Mississippi.

Normandeau Associates, Inc. 1984. Relative catch efficiency of a 3 m beam trawl and a 6.2 m high-rise trawl for sampling young of the year fishes in the Hudson River estuary.

Prepared for New York Power Authority, White Plains, New York.

Normandeau Associates, Inc. 2003. Assessment of Hudson River recreational fisheries.

Prepared for New York State Department of Environmental Conservation and Bureau of Marine Resources Hudson River Fisheries Unit. 69 pp. New Paltz, New York.

31

Paloheimo, J.E. 1961. Studies on estimation of mortalities: I. Comparison of a method described by Beverton and Holt to a new linear formula. Journal of the Fisheries Research Board of Canada 18:645-662.

Pauly, D, and V. Christensen. 1995. Primary production required to sustain global fisheries.

Nature. Vol. 374:255-257.

Peterson, I., and J. S. Wroblewski. 1984. Mortality rate of fishes in the pelagic ecosystem.

Canadian Journal of Fisheries and Aquatic Sciences 41(77):1117-1120.

PSEG. 2006. New Jersey Pollution Discharge Elimination System Permit Application for Salem Generating Station.

Richards, R.A., and P.J. Rago. 1999. A case history of effective fishery management:

Chespeake Bay striped bass. North American Journal of Fisheries Management. 19:356-375.

Ricker, W.E. 1975. Computation and interpretation of biological statistics of fish populations.

Bulletin of the Fisheries Research Board of Canada. Bulletin 191. 382 p. Department of the Environment, Fisheries and Marine Service. Ottawa.

Savoy, T.F., and V.A. Crecco. 2004. Factors affecting the recent decline of blueback herring and American shad in the Connecticut River. Pages 361-378. In P.M Jacobson, D.A.

Dixon, W.C. Leggett, B.C. Marcy, Jr., and R.R. Massengill, editors. The Connecticut River ecological study (1965-1973) revisited: ecology of the lower Connecticut River 1973-2003. American Fisheries Society, Monograph 9, Bethesda, Maryland.

32

Secor, D.H. and P.M. Piccoli. 1996. Age- and sex- dependent migrations of striped bass in the Hudson River as determined by chemical microanalysis of otoliths. Estuaries. Vol. 19, No. 4, p. 778-793.

Texas Instruments. 1980. 1978 Year class report for the multiplant impact study of the Hudson River estuary. Appendix B. Prepared for Consolidated Edison of New York, Inc. 846 pp. New York, New York.

Uphoff, J.H. Jr. 2003. Predator-prey analysis of striped bass and Atlantic menhaden in upper Chesapeake Bay. Fisheries Management and Ecology. 10:313-322.

Walter, J.F. III., A.S. Overton, K.H. Ferry, and M.E. Mather. 2003. Atlantic coast feeding habits of striped bass: a synthesis supporting a coast-wide understanding of trophic biology. Fisheries Management and Ecology. 10:349-360.

Whipple, S.J., J.S. Link, L.P. Garrison, and M.J. Fogarty. 2000. Models of predation and fishing mortality in aquatic ecosystems. Fish and Fisheries, 1:22-40.

Young, J.R., R.J. Klauda, and W.P. Dey. 1988. Population estimates for juvenile striped bass and white perch in the Hudson River. American Fisheries Society Monograph 4:89-101.

554 33

555 556 Table 1. Estimates of average predatory demand (P,) of striped bass populations for the two 557 periods of years (/) with estimated standard errors (in parentheses) and ratios of estimated 558 predatory demands (Period 2 to Period 1).

559 Stock Season Ages _a. Ratio of P, P2 Average (kg) (kg)

Predatory Demands (Rp)

Atlantic January- 1 - 13+ 61,829,229 274,937,594 Coastwide December (2,031,616) (5,853,828)

Hudson August- 1 - 13+ 3,363,749 14,957,667 4.45 River October (110,398) (318,097)

Hudson August- 1 and 2 1,332,950 4,583,020 3.44 River October Only (89,354) (246,129) 560 34

561 562 Table 2. Average index values for post yolk-sac larval ("PYSL") abundance (Lj) for the two 563 periods of years (j) with estimated standard errors (in parentheses) and ratios of average 564 PYSL abundances (Period 2 to Period 1).

565 Taxon L2 Ratio of Average PYSL Abundance (RI)

Striped Bass 362,055,919 2,371,617,937 6.55 (13,868,061) (76,310,566)

River Herring 2,008,741,295 990,192,857 0.49 (50,041,415) (19,314,422)

Atlantic Tomcod 92,730,226 66,887,806 0.72 (6,329,898) (3,548,958)

White Perch 985,594,499 855,285,920 0.87 (15,678,812) (15,857,150) 566 567 35

568 569 Table 3. Estimates of average seasonal (August through October) juvenile abundance (Nj) for 570 the two periods of years (j) with estimated standard errors (in parentheses) and ratios of 571 average estimated juvenile abundances (Period 2 to Period 1).

572 Taxon N1 N2 Ratio of Average Juvenile Abundances (R.)

Striped Bass 68,372,839 57,132,380 0.84 (1,312,794) (903,684)

River Herring 1,118,600,941 448,416,556 0.40 (30,380,270) (14,130,644)

Atlantic Tomcod 54,150,749 16,8.59,655 0.31 (2,671,040) (995,044)

White Perch 65,493,845 26,860,369 0.41 (1,782,169) (779,541) 573 36

574 575 Table 4. Life history parameter estimates for juvenile striped bass, river herring, Atlantic 576 tomcod and white perch.

577 Parameter Taxon Striped Bass River Atlantic White Perch Herring Tomcod Juvenile Growth Rate, g (day-) 0.032 0.047 0.030 0.034 Initial Weight of Juvenile Fish, w 0.286 0.034 0.095 0.179 (gm)

Background Mortality Rate -- August 0.606 0.847 0.470 0.669 through October, m Background Mortality Rate -- 0.151 0.250 0.922 0.169 Beginning of Juvenile Life (15 days) (15 days) (90 days) (15 days)

Stage to the Beginning of August, m' (duration in parentheses) 578 37

579 580 Table 5. Estimates of average seasonal (August through October) instantaneous mortality rates 581 possibly due to predation by age I and age 2 striped bass for Period 1 (1977-1990) and 582 Period 2 (1991-2004). For estimates based on approximation, estimated standard errors 583 are listed (in parentheses).

584 Prey Taxon Estimates Based on Estimates Based on Exhaustive Search Approximation hp,l mmp,2 mP, m p,2 Striped Bass 0.611 6.172 1.157 4.760 (0.137) (0.199)

River Herring 0.049 0.475 0.048 0.410 (0.015) (0.159)

Atlantic Tomcod 0.103 1.287 0.112 1.232 (0.014) (0.150)

White Perch 0.149 1.737 0.178 1.489 (0.018) (0.134) 585 38

586 587 Table 6. Estimates of average seasonal (August through October) instantaneous mortality rates 588 possibly due to predation by age 1 through age 13 striped bass for Period 1 (1977-1990) 589 and Period 2 (1991-2004). For estimates based on approximation, estimated standard 590 errors are listed (in parentheses).

591 Prey Taxon Estimates Based on Estimates Based on Exhaustive Search Approximation mn,, mp, 2 mpr, 1 m, 2 Striped Bass 0.449 5.894 0.834 4.437 (0.048) (0.141)

River-Herring 0.037 0.457 .0.03.6 -P0.1398....

(0.011) (0.154)

Atlantic Tomcod 0.077 1.255 0.084 1.205 (0.008) '(0.146)

White Perch 0.113 1.712 0.133 1.445 (0.018) (0.134) 592 39

593 594 Table 7. Estimates of average seasonal (August through October) juvenile biomass possibly 595 consumed by predation by age 1 and age 2 striped bass (C1 ) for Period 1 (1977-1990) 596 and Period 2 (1991-2004), and corresponding percent of seasonal predatory demand of 597 age 1 and age 2 Hudson River striped bass.

598 Prey Taxon Percent C1 C2 of (kg) (kg)

Seasonal Predatory Demand Striped Bass 76,652 263,547 5.75%

River Herring 39,804 136,821 2.99%

Atlantic Tomcod 18,073 62,137 1.36%

White Perch 13,420 46,147 1.01%

Total 147,949 508,652 11.11%

599 600 40

601 602 Table 8. Estimates of average seasonal (August through October) juvenile biomass possibly 603 consumed by predation by age 1 through age 13 striped bass (Cj) for Period 1 (1977-604 1990) and Period 2 (1991-2004), and corresponding percent of seasonal predatory 605 demand of age I through age 13 Hudson River striped bass.

606 Prey Taxon Percent C1 C2 of (kg) (kg)

Seasonal Predatory Demand Striped Bass 58,215 258,862 1.73%

River Herring 29,741, 132,319 0.88%

Atlantic Tomcod 13,671 60,787 0.41%

White Perch 10,281 45,713 0.31%

Total' 111,908 497,681 3.33%

607 608 41

609 610 Figure 1. Estimates of annual predatory demand of the Atlantic coast striped bass stock, ages 611 through 13.

612 613 400 614 350 --

615 300 -

616

• 250 617 200 -

618 0 150 619- 100 00 620 50 621 0 ,

1980 1985 1990 1995 2000 2005 622 Year 623 624 625 42

626 627 Figure 2. Estimates of the percent of seasonal predatory demand of age 1 and age 2 Hudson 628 River striped bass potentially satisfied by consumption of juveniles of the four taxa, as 629 functions of the index of density dependent effects (see text) and assumed background 630 mortality rates. Curve A is for the estimated background mortality rates (see text), curve 631 B is for background mortality rates of zero, and curve C is for two times the estimated 632 background mortality rates. The proportion of the coastwide population of age 1 and 633 age 2 striped bass that were Hudson River fish was assumed to be 13% in Period 1 and 634 Period 2.

635 636 20%

637 a 15%

638 0 I.

639 10%

0 640 U,

641 0 5%

642 0%

643 0 0.25 0.5 0.75 1 Index of Density Dependent Effects 644 43

645 646 Figure 3. Estimates of the percent of seasonal predatory demand of age I and age 2 Hudson 647 River striped bass potentially satisfied by consumption ofjuveniles of the four taxa, as 648 functions of the index of density dependent effects (see text) and assumed background 649 mortality rates. Curve A is for the estimated background mortality rates (see text), curve 650 B is for background mortality rates of zero, and curve C is for two times the estimated 651 background mortality rates. The proportion of the coastwide population of age 1 and age 652 2 striped bass that were Hudson River fish was assumed to be 20.9% in Period 1 and 653 8.9% in Period 2.

654 655 30%

656 25%

657 0 20%

I-658 15%

0 659 10%

0 660 U 5%

661 0%

662 0 0.25 0.5 0.75 1 Index of Density Dependent Effects 663 664 44

665 666 Figure 4. Estimates of the percent of seasonal predatory demand of age 1 through age 13 667 Hudson River striped bass potentially satisfied by consumption ofjuveniles of the four 668 taxa, as functions of the index of density dependent effects (see text) and assumed 669 background mortality rates. Curve A is for the estimated background mortality rates (see 670 text), curve B is for background mortality rates of zero, and curve C is for two times the 671 estimated background mortality rates. The proportion of the coastwide population of age 672 1 through age 13 striped bass that were Hudson River fish was assumed to be 13% in 673 Period 1 and Period 2.

674 6%

675 5%

676 2 4%

677 3%

678 679 rA 2%

4.

Q 680 01 1%

681 0%

0 0.25 0.5 0.75 1 682 Index of Density Dependent Effects 683 684 685 45

686 687 Appendices 688 689 Appendix A: Derivation ofApproximations 690 The approximations were based on the following equivalences:

691 R* -§Ij - -* (Al)

R1? N2~

692 and 6 93 R._.I E N= _ ( L= (A 2 )

694 The first order Taylor series approximation (evaluated at mpj=O) for the numerator (with 695 j=2) or denominator (withj=1) of equation (Al), that expresses that term as a function'ofthe 696 mortality rate for predation, is:

-- rl~~~~e(-m,,) -¢l e ( -s 697 - At hlt erns-gt (A3) 698 which, using the approximation from Paloheimo (1961) can be written as:

699 m- m) tw e (A4) 700 Therefore, an approximation for the ratio of ratios in equation (Al) is:

46

701 (A5) 702 where a is the ratio (Period 2 to Period 1) of the average juvenile weight per fish at the mid-703 point of the season.

704 Again using the approximation from Paloheimo (1961), the numerator (with]j=2) or 705 denominator (withj1=1) of equation (A2) was approximated as:

- (m 1+ma. 1 )t 706 N1--

Li (m' 1 +mp,~re 2 (A6) 707 Therefore, the logarithm of the ratio of ratios in equation (A2) is approximately:

708 ln-i) (mR - mp,2 Qt'+2.t) +'8 (A7) 709 where )6 is the difference between the juvenile background mortality rates for Period 1 and 710 Period 2.

711 Appendix B: Formulaefor VarianceEstimates 712 Formulae for variance estimates for the approximate estimates of instantaneous mortality 713 rates due to possible predation were derived using a Taylor series approximation (Kendall and 714 Stuart 1977). Because the variances were intended to represent imprecision due to sampling

(

715 error, and data for the three component ratios are from independent sampling programs, all 716 covariance terms were set to zero. Lower case symbols (e.g. r,,) indicate estimates of 717 corresponding paramters (e.g. R,).

718 For the approximate estimate of the instantaneous mortality rate for Period 1:

47T

Inel 719 m, J +In (BI) 720 the formula for the variance estimate is:

32 (dmpi .. (din ,1,*2 ap 721 ar,) )2 var(r,) +, drp)

M-'- 2-.var(rp) var(mf,,) dr.p1 vartr,) + ---- (B2) 722 where 723 dr [ r 1p Lp r.d)r

- 1r F -2 rpjJ(t'+-t (B3) drop, ( rf*l p - )t+t -

724 dr, . r t ,2 (B4) 725 and 726 (B5) drdp = r, t'+2rIn(

727 For the approximate estimate of the instantaneous mortality rate for Period 2:

728 rn1,2m"2= -r1)¢t,+t/ (B6)

Lrp 22) 729 the formula for the variance estimate is:

730 var(mP2 ) ._ddr.

mp2-) 2 p2 (B7) var(r.) +(_dmp2 dr, )

)2 var(r,)+(dmp2dr1 , )) 2var(rp) 731 where 48

732 dm, 2 = FrI )/.lY' (L 2 in(IY12)(138) 733 dmp,2 =- (r,-' _r', *), )t (139) dr, r. 2) 734 and 735* d",, ,.-,.n".-

dM 2=r,,r,( , _1 -

),,"(t,--' (Bo) 736 Estimated variances for the component ratios (r,,, rp, and ri) were computed using the 737 following formulation (using r, as an example):

, 2.

2 var(n2 ) var((B1 738 var(r,) - 2 2 (B11) 739 where 740 var(ij) =-.4 (se(nii)) (B12) 1k 741 ,-= n (B13) 742 for year i within periodj; and 743 r, = "2 (B 14) 744 Estimates of standard errors (for equation (B 12)) for estimates of juvenile and post yolk-sac 745 larval abundance were from the annual Year Class Reports (e.g. EA 1996). For estimates of 746 predatory demand, estimates of standard errors were based on reported coefficients of variation 747 for estimates of age-specific abundance of Atlantic coast striped bass (ASMFC 2005).

748 Parameters other than abundance were treated as constants in the variance estimates.

49

749 Appendix C: Estimatesof the Proportionof the CoasiwidePopulationofAge 1 Striped 750 Bass from the Hudson River 751 The proportion of the coastwide population of age-i striped bass that was of Hudson 752 River origin was estimated from: 1) the time series of estimates of age-1 abundance (NIy), and 2) 753 the time series ofjuvenile abundance indices for four major spawning areas: Chesapeake Bay 754 Maryland (CBM), Chesapeake Bay Virginia (CMV), Hudson River (HR), and Delaware River 755 (DR). For each year, y, the proportion was estimated as:

7HRXHRy 756 XCBMcy y + 8CBVXCBVy + AHRXHR,y + 8DRXDR,y (C1) 757 where the fl's are the estimated regression coefficients from a multiple regression of age-i 758 coastwide abundance against the year-specific juvenile indices (XcBmy, XcBvy, XHRly, XDR.y) from 759 the four spawning areas (ASMFC, 2005):

760 N',y = /JCBMXCBM y + 8cBVXCBV,y + /tHRXHRy + / 3 DRXDRy (C2) 761 The R2 for this multiple regression was 0.96 (p<0.0001).

762 50

APPENDIX D Appendix D Prepared by:

Webster Van Winkle Van Winkle Environmental Consulting Co.

John Young ASA Analysis & Communication, Inc.

Entrainment Susceptibility at Indian Point and Change in YOY Abundance Cooling water withdrawals impose some incremental mortality on species susceptible to entrainment. The effect of this incremental mortality may be inconsequential to the populations and communities in the water body, or, if the increment is large enough, could potentially lead to either a decrease or a reduced rate of increase in the affected populations. However, in addition to cooling water withdrawals, there are many other factors that can affect population trends, including changes in prey and predator populations, climatic effects, harvesting intensity, habitat modification, invasive species, and water quality. Thus, over any given time period, populations of some species can be expected to increase, while others decrease, regardless of cooling water withdrawals.

If entrainment at IP2 and IP3 were having an adverse impact on the Hudson River fish community, then species with high susceptibility to entrainment would be expected to have decreased, or increased less in abundance, over the past 32 years than would species with low susceptibility. This possibility can be evaluated by examining the relationship between a measure of entrainment susceptibility and a measure of population change derived by comparing the mean abundance of young-of-year ("YOY") fish belonging to various species from 1974-1989 to the mean abundance of the same species of fish from 1990-2005. YOY is selected for the metric because the effects of entrainment have been realized by the time fish reach the YOY stage, and this age group is still within the estuary and can be sampled for most species. The periods 1974-1989 and 1990-2005 were selected so that the two periods of comparison would include equal numbers of years.

Evaluating the relationship between entrainment susceptibility and change in YOY abundance requires selecting those species for which adequate data are available for both variables. Entrainment susceptibility can be characterized quantitatively by evaluating the distribution of entrainable life stages in the Regions from which IP2 and IP3 withdraw water in comparison to all the Regions sampled. The expected effect of continued annual entrainment losses of early life stages, if losses are severe enough to affect population size, is a negative relationship between entrainment susceptibility and the ratio of YOY abundance from the early part of the time series (1974-1989) to the latter part (1990-2005).

I

METHODS The process for evaluating the relationship between entrainment susceptibility and changes in YOY abundance is summarized in Figure D-1. The process involves three steps:

(1) Calculate a species-specific metric of entrainment susceptibility based on larval abundance data from the LRS; (2) Calculate a species-specific metric of change in YOY abundance based on data from the BSS/FSS; and (3) Determine if entrainment susceptibility is negatively related to change in YOY abundance.

Step 1. Entrainment Susceptibility Based on Larval Distribution (EntSus)

A species-specific metric of entrainment susceptibility is calculated from the utilities' LRS for the 32-year period 1974-2005.1 Species using the Hudson River estuary as a spawning and nursery area vary by season within a year. In addition, the geographic and temporal extent of the LRS sampling varies among years, and some species occur in two or three seasonal periods. These realities are addressed by dividing the LRS database into three seasonal periods and considering only those weeks that were sampled:

• Winter & early spring: Years 1975-1980 and 1995-2005; Weeks 8-16; Regions 1-6

• Late spring:*Years 1974-2005; Weeks 17-27; Regions 1-12

• Summer: Years 1991-2005; Weeks 28-41; Regions 1-7 Identification of larvae to species level is not always practical, in which case larvae are classified by genus or family. Differences in taxonomic level of EntSus and YOY abundance data are resolved in one of two 'ways: (a) if BSS/FSS data are adequate at species level but LRS data are not, then use the same genus or family EntSus value for each species, or (b) if BSS/FSS An index of standing crop (the number of fish in an area or volume at a particular time) is estimated by life stage and species. Standing crop indices are calculated for each habitat (shorezone, benthic, water column) in each region and each week by taking the product of the average density in a habitat during that week and the area (shorezone habitat) or volume (benthic and water column habitats) contained in that region. The standing crop index for each region and week is then estimated as the sum of the habitat index values. This value is an index rather than an absolute standing crop value because no adjustment is applied for differences in collection efficiency between sampling gears (ASA, 2005; Chapter 2, Materials and Methods, 2004 Year Class Report).

2

data are not adequate at species level but LRS data are, then pool species-level LRS abundance data to the genus or family taxonomic level.

Relative abundance of larvae in Regions 3-5, EntSus, is the index of entrainment susceptibility. For each sampled year (and each seasonal period when possible), EntSus is estimated for each species as the ratio of standing crop in Regions 3-5 to standing crop in all sampled regions. For those species occurring in more than one of the three seasonal periods, annual EntSus values are calculated as an average across periods, p, weighted by abundance for each period:

• pSCpEntSusip EntSus, = _PSCP where EntSusi = fraction of species in the Hudson River estuary in Regions 3-5 in year i SCip = sum of abundance of the species within seasonal period p in year i EntSusip = value of EntSus for seasonal period p in year i Annual EntSus values are estimated for each species for each year in which the species occurred during 1974-2005. Mean entrainment susceptibility and its variance are calculated for 2

each species based on its annual EntSus values.

Step 2. Change in YOY Abundance (R)

The utilities' Beach Seine Survey (BSS) and Fall Shoals Survey (FSS) programs are selected as the best measures of change in abundance of YOY fish. These programs have sampled the estuary using similar gear and methodology since 1974, although there have been variations in the Regions sampled and in time of initiation and end of the sampling across the years. To maintain consistent sampling effort and maximize comparability of results, data are restricted to Regions 1-12 and weeks 31-42, approximately corresponding to August through October.

Abundance data by species are categorized into two salinity zones, three habitats, and two time periods. The two salinity zones are brackish (Regions 1-6; river miles 12-61) and freshwater (Regions 7-12; river miles62-152). The three habitats sampled by these surveys are:

2 Entrainment susceptibility at Indian Point will change during extreme water years. In wet years some freshwater and anadromous species will be more at risk, while in dry years some marine species will be more at risk.

3

(a) shorezone (bottom area in water 10 ft or less in depth), sampled with the 100-ft beach seine in the BSS from 1974-2005; (b) benthic (volume of water between river bottom and 3 ft above the bottom), sampled with the beam trawl in the FSS from 1985-2005; and (c) water column (water volume not included in either the shorezone or benthic habitats), sampled with the Tucker trawl in the FSS from 1979-2005. Except for weekly BSS sampling in the 1970s, all of the sampling was done on an alternate week basis.

Time series of abundance data are divided into two periods: Period 1 = 1974-1989; Period 2 = 1990-2005. This division results in equal number of years in the two periods for shorezone habitat (16 years), but unequal number of years for benthic habitat (five years and 16 years) and water column habitat (11 years and 16 years).

The available data for measuring change in abundance provide the potential for six independent estimates of relative abundance change for each species (two salinity zones and three habitats). However, some species may be concentrated in particular habitats or salinity zones. Due to the strong salinity preferences of freshwater and marine fish, only sampling from their preferred salinity zone (freshwater zone for freshwater fish, brackish zone for marine fish) was used. In addition, it is difficult to accurately measure abundance changes for species that occur only occasionally. Thus, species data from a salinity zone-habitat combination are included in the analysis only if the total catch meets a minimum level of catch in at least one of the two periods (see Step 3 below). To adjust for the unequal number of years for benthic and water column habitats mentioned above, the Period 1 catch is adjusted upward by a factor based on the number of years sampled, i.e., 3.20 (=16 yr/5 yr) for benthic and 1.45 (=16 yr/il yr) for water column.

For each selected salinity zone-habitat, the weighted mean YOY abundance for Period 1, Period 2, and Periods 1 and 2 combined are calculated with the GLM procedure in SAS. Mean abundance for each of these three time intervals is calculated as the weighted mean abundance across the sampling Regions within a salinity zone, where the weight is the proportion of the total amount of a habitat in that salinity zone that occurs within each of its six Regions.

Relative change in YOY abundance for each species, Ri, and its standard error, se(R), are calculated based on (Cochran 1977, pp. 30-34)3. Since Ri is bounded on the lower side by 0 for Let:

XIk= weighted mean cpue in Period I for species i in habitatj in salinity zone k 4

decreases in abundance, is 1 if mean abundance is unchanged, and is unbounded above 1 for 4

increases in abundance, a loglo transformation is used to normalize the distribution of R values.

j /i y2

_ Yk = relative change in species i abundance from Period I to Period 2 jk se(R,)-o, + R EYJk -2Rij_*Y1,kY k 2 Uk) se(Ri) =anY-=l.#n-1 n, -

Step 3. Association between'Entrainment Susceptibility and Change in YOY Abundance Three correlation methods (Pearson, Spearman, and Kendall) are used to evaluate the association between EntSus and YOY abundance change using the CORR procedure in SAS.

There is no simple mathematical relation between any two of these three methods. When the true correlation coefficient is not zero, it is likely that each coefficient is sensitive to different types of departures from independence (Sokal and Rohlf, 1995).

Availability of data varies among species, and results of correlation analysis could be sensitive to how many species are included in the analysis. Thus a limited sensitivity analysis is performed to evaluate to what extent the correlation results depend on selection criteria. The approach to this sensitivity analysis is to define two cases, Case A and Case B. The species in X 2 Yk = weighted mean cpue in Period 2 for species i in habitat j in salinity zone k x.Uk.= weighted mean cpue over both Periods for species i in habitat j in salinity zone k Y-Iyk = XIUk / 5-.uk = relative mean cpue in Period 1 for species i in habitatj in salinity zone k Y20k = X2Uk / Y-U* = relative mean cpue in Period 2 for species i in habitat j in salinity zone k nli = number of salinity zone-habitat combinations selected for species i in Period 1.

n2j = number of salinity zone-habitat combinations selected for species i in Period 2.

4 The effectiveness of estimating change in YOY abundance from Period I to Period 2 based on BSS/FSS data is limited for some species because these surveys do not sample some habitats that are primary habitats for YOY (i.e., tributaries, bays, wetlands, or shorezone habitat with structure). Although R integrates BSS/FSS YOY abundance data from benthic, water column, and shorezone habitats, the growth and survival of larvae and YOY fish that are most common in these unsampled habitats may be determined by factors that are largely irrelevant for species in the sampled habitats. Examples of such factors are micro-habitats suitable for parental nest building and guarding of young, protection from predators, and availability of food not present in open water habitats. Although species that frequent these habitats exclusively or primarily are not adequately sampled compared to other Hudson River species, there is a relatively small amount of such unsampled habitats in the estuary, and these species are not likely to be affected by IP entrainment because of their preference for these unsampled habitats.

5

Case B are a subset of the species in Case A. Species in Case A are selected based on LRS data criteria for EntSus and on BSS/FSS data criteria for YOY abundance. Species are excluded from Case A to create Case B based on more restrictive criteria for both larval and YOY abundance data. Species selection decisions are made independently for each of these two variables. Thus, a species can be excluded from this evaluation even if data are adequate for one variable but not the other variable.

Species selection criteriafor entrainmentsusceptibility based on larval abundance Cases A and B. EntSus > 0, i.e., minimum of one larva in LRS samples from Regions .3-5 during 1974-2005.

Case A. Minimum average of 100 larvae per year of occurrence collected in LRS samples from Regions 1-12 during 1974-2005.

Case B. Minimum average of 1,000 larvae per year of occurrence collected in LRS samples from Regions 1-12 during 1974-2005.

5 Species and salinity-zone habitatselection criteriafor change in YOY abundance Case A. Minimum of 100 YOY collected in BSS/FSS samples in at least one SZ-habitat in at least one of the two time periods.

Case B. Minimum of 1,000 YOY collected in BSS/FSS samples in at least one SZ-habitat in at least one of the two time periods.

RESULTS Entrainment Susceptibility (EntSus)

EntSus is a measure of the proportion of larvae in those habitats sampled by the LRS that were collected in Regions 3-5 compared to Regions. 1-12.6 Twenty four (24) species meet the Case A selection criterion for EntSus.7 For these 24 species, mean EntSus scores range from 0.45 for striped bass to 0.02 for American shad.8 Number of SZ-habitats selected can vary from 1 to 6 for anadromous and estuarine species and from I to 3 for freshwater and marine species. If a SZ-habit is selected for Period 1 (or 2), Period 2 (or 1) is included also.

6 The LRS does not sample in some habitats that are critical for many Hudson River fish species for spawning and larval life stages, e.g., tributaries, bays, wetlands, and shorezone habitat with structure.

' Five of these 24 species are not selected for correlation analysis because they do not meet the Case A selection criterion for YOY abundance.

8 The list of species collected during the intensive entrainment study at Indian Point (1983-1987) was compared with the list of species collected during the 1974-2005 LRS in Regions 1-12. Four species, all marine, were collected only in the Indian Point entrainment study and not in the LRS. These species are not selected for the 6

Mean annual EntSus values for the representative species varied by more than an order of magnitude: striped bass (0.45), bay anchovy (0.42), Atlantic tomcod (0.26), white perch (0.16),

alewife and blueback herring (0.05), and American shad (0.02). Most of these seven species were collected as larvae every year, although the average number of larvae collected per year of occurrence varied by two orders of magnitude from alewife/blueback herring (3 x 105) to American shad (2 x 103). Spottail shiner had fewer than 100 larvae/yr occurrence, and no EntSus value is calculated.

Change in YOY Abundance Forty-six (46) species are selected based on the Case A criterion for YOY abundance.

However, only 19 of these species are also selected based on the Case A criterion for larval abundance, and thus only these 19 species are selected for the EntSus-R correlation analysis.

Correlation Analysis Table D- 1 shows the correlation coefficients and probability values, for both Case A and Case B, for all three correlation indices. Figures D-2 and D-3 provide plots of mean entrainment susceptibility vs. the normalized index of relative change in YOY abundance from Period 1 to Period 2 for both Case A and Case B. For both Cases A and B, all three estimates of the correlation between Loglo(R) and EntSus are not statistically significantly different from zero (Table D-1). This result is opposite the expected significant negative correlation if Indian Point entrainme'nt were adversely affecting the population trends of susceptible species. Therefore, the effect of Indian Point entrainment on abundance patterns of the fish community, if there is one, is not large enough to be statistically detectable in the 32 years of monitoring data.

Nineteen (19) taxa, representing 31 species, four of the five guilds, 13 taxonomic families, and a broad range of both EntSus and R values (Table D-2, Figures D-I and D-2) are selected for Case A. Eleven (11) of these taxa, representing 17 species, are retained in Case B. 9 Plots of EntSus vs. Loglo(R) illustrate that more species decreased than increased in YOY EntSus- R analysis. The species (and number of larvae collected) are Atlantic needlefish (3), smallmouth flounder (1), striped searobin (1), and northern searobin (1).

9 Eight taxa are excluded from Case A in creating Case B. The eight taxa are: Atherinid spp., banded killifish, gizzard shad, centrarchid spp, northern pipefish, rainbow smelt, winter flounder, and yellow perch. These taxa are excluded because of not meeting the more restrictive Case B selection criterion for larvae, YOY, or both.

7

abundance for both cases (Figures D-2 and D-3), but the change in abundance values (R) was only weakly associated with the magnitude of EntSus values.

DISCUSSION AND CONCLUSIONS EntSus is a quantitative index bounded by 0.00 and 1.00. It is based on LRS data for larval abundance in water column and benthic habitats sampled in Regions 3-5 relative to larval abundance in these habitats sampled in Regions 1-12 of the Hudson River estuary. Thus, EntSus is an index of risk of entrainment of larvae at Indian Point. It is not an index of impact on the population.

The low correlations observed between EntSus and Logo(R) are counter to the expected more negative correlations if Indian Point entrainment were a significant factor influencing population dynamics of the fish community. Although the number of taxa (19) for which both variables could be measured is small, these taxa represent approximately 94% (Case A) and 88%

(Case B) of all YOY fish captured in the BSS/FSS programs from 1974-2005.'

In conclusion, 32 years of monitoring data do not support the hypothesis that entrainment at Indian Point has caused substantial harm to the fish community of the Hudson River estuary.

Although more species have decreased than increased in YOY abundance over this time period, changes in abundance are unrelated to species susceptibility to entrainment at IP2 and IP3.

8

REFERENCES ASA Analysis & Communication. 2005. 2003 Year Class Report for the Hudson River Estuary Monitoring Program. Prepared for Dynegy Northeast Generation, Entergy Nuclear Indian Point 2 LLP, Entergy Nuclear Indian Point 3 LLC, Mirant Bowline.

Cochran, W. G. 1977. Sampling Techniques. John Wiley & Sons, New York, New York Sokal, R.R., and F.J. Rohlf. 1995. Biometry-The Principles and Practice of Statistics in Biological Research. 3 rd Edition. W. H. Freeman and Company, San Francisco, CA.

9

Table D-1. Pearson, Spearman, and Kendall correlation coefficients for the association between Logjo(R) and mean EntSus. A value ofp represents the probability of a sample correlation coefficient larger than the observed sample correlatiorn coefficient, if the true correlation coefficient is zero.

Case N ....Pearson Spearman Kendall 19 r 0.225 0.182 0.129 A

P 0.355 0.457 0.442 12 r 0.157 -0.042 -0.046 B

p 0.625 0.897 0.837 10

Table D-2. EntSus and LogjoR values for Figures 1 and 2, including standard errors. Case A, 19 taxa; Case B, 12 taxa. Sorted by EntSus, low to high, for each case.

Case Family Guild Taxon/Species ,EntSus SE EntSus R SE R Logio R A CLUP A American shad 0.023 0.009 0.480 0.091 -0.318 A PLEU M Winter flounder 0.030 0.007 0.440 0.374 -0.357 A CLUP A Alewife 0.051 0.008 1.133 0.337 0.054 A CLUP A Blueback herring 0.051 0.008 0.582 0.101 -0.235 A CLUP F Gizzard shad 0.072 0.049 2.011 0.671 0.303 A SYNG M Northern pipefish 0.079 0.024 0.774 0.058 -0.111 A CYPR F Cyprinid unid 0.107 0.013 1.154 0.076 0.062 A PERC F Tesselated darter 0.109 0.012 0.971 0.149 -0.013 A CENT F Centrarchid unid 0.116 0.015 2.271 1.609 0.356 A MORO E White perch 0,158 0.013 0.440 0.072 -0.357 A PERC F Yellow perch 0.201 0.024 0.551 0.197 -0.259 A CYPD E Banded killifish 0.210 0.096 0.306 0.242 -0.515 A OSME A Rainbow smelt 0,260 0.030 0.633 0.087 -0.198 A GADI E Atlantic tomcod 0.263 0.042 0.400 0.134 -0.398 A CLUP M Atlantic menhaden 0.300 0.046 80.026 35.284 1.903 A SCIA M Weakfish 0.302 0.050 0.516 0.265 -0.287 A ATHE E Atherinid sp. 0.339 0.032 3.509 2.487 0.545 A ENGR M Bay anchovy 0.417 0.032 0.720 0.200 -0.142 A MORO A Striped bass 0.454 0.020 1.236 0.380 0.092 B CLUP A American shad 0.023 0.009 0.527 0.109 -0.278 B CLUP A Alewife 0.051 0.008 1.267 0.574 0.103 B CLUP A Blueback herring 0.051 0.008 0.582 0.101 -0.235 B CYPR F Cyprinid unid 0.107 0.013 1.233 1.432 0.091 B PERC F Tesselated darter 0.109 0.012 0.971 0.149 -0.013 B MORO E White perch 0.158 0.013 0.459 0.094 -0.338 B OSME A Rainbow smelt 0.260 0.030 0.821 0.129 -0.086 B GADI E' Atlantic tomcod 0.263 0.042 0.346 0.157 -0.461 B CLUP M Atlantic menhaden 0.300 0.046 80.026 35.284 1.903 B SCIA M Weakfish 0.302 0.050 0.398 0.294 -0.400 B ENGR M Bay anchovy 0.417 0.032 0.720 .0.200 -0.142 B MORO A Striped bass 0.454 0.020 0.976 0.364 -0.011 ll

Figure D-1. Analysis Flow Chart for Entrainment Susceptibility Select taxa that meet larvae Select taxa that meet YOY selection criteria I I selection criteria Calculate annual entrainment Calculate relative change in susceptibility index abundance from Period 1 (1974-1989) to Period 2 (1990-2005) for each selected taxon in each salinity zone-habitat I

Calculate mean entrainment I

Calculate combined relative susceptibility index across years change for each taxon for each taxon Select taxa meeting both larvae and YOY criteria II Determine correlation of entrainment susceptibility and relative change in abundance 12

Figure D-2. Association between change in YOY abundance from Period 1 to Period 2, Loglo(R), and entrainment susceptibility, EntSus, for the 19 taxa selected for Case A. Zero on the logarithmic Y axis corresponds to no change in YOY abundance. Use Table 2 as an aid in determining which species is associated with which point in the figure.

N=19; r=0.16; P=0.51 CASE A 2

• Anadromous 0 Estuarine A Freshwater e Marine C) 0 V-J 0 0*

} I I

-2 0 0.1 0.2 0.3 0.4 0.5 Mean Entsus

Figure D-3. Association between change in YOY abundance from Period I to Period 2, Loglo(R), and entrainment susceptibility, EntSus, for the 11 fish taxa selected for Case B. Zero on the logarithmic Y axis corresponds to no change in YOY abundance. Use Table 2 as an aid in determining which species is associated with which point in the figure.

CASE B 2

0 *Anadromous n Estuarine A Freshwater

[l Marine C')

0) 0

-J I 0

-2 0 0.1 0.2 0.3 0.4 0.5 Mean Entsus