ML073310706
ML073310706 | |
Person / Time | |
---|---|
Site: | Indian Point ![]() |
Issue date: | 08/16/2007 |
From: | Atlantic States Marine Fisheries Commission |
To: | Office of Nuclear Reactor Regulation |
References | |
07-01 | |
Download: ML073310706 (238) | |
Text
Stock Assessment Report No. 07-01 (Supplement) of the Atlantic States Marine Fisheries Commission American Shad Stock Assessment Report for Peer Review Volume I August 2007 Working towards healthy, self-sustaining populations for all Atlantic coast fish species or successful restoration well in progress by the year 2015
Stock Assessment Report No. 07-01 (Supplement) of the Atlantic States Marine Fisheries Commission American Shad Stock Assessment Report for Peer Review Volume I Conducted on July 16 - 20, 2007 And Accepted by The Shad and River Herring Management Board August 16, 2007 A publication of the Atlantic States Marine Fisheries Commission pursuant to National Oceanic and Atmospheric Administration Award No. NA05NMF4741025.
PREFACE The American Shad Stock Assessment Report analyzes the status of 31 stocks of American shad along the Atlantic coast. Due to the large volume of material contained within the report (1200+ pages), it is organized into three volumes. Volume I contains a comprehensive look at all of the stocks, including an introduction to the science and management of the species, summaries of coastwide indices, summaries of the state or river system assessments, conclusions and recommendations, and a look at hypothesized causes of decline. Volumes II and III provide an in-depth exploration of American shad stock status by state or river system. These volumes provide stand-alone assessments of stocks and serve as a reference for material contained in Volume I. The contents of the three volumes follow:
- Volume I: Introduction Coastwide Summaries State and River Stock Assessment Summaries Conclusions and Recommendations Causes of Decline
- Volume II: Maine New Hampshire Merrimack River Rhode Island Connecticut River Hudson River Delaware Bay and River Minority Report for Connecticut River
- Volume III: Maryland Susquehanna River Potomac River Virginia North Carolina South Carolina Georgia Florida ii
ACKNOWLEDGEMENTS The Atlantic States Marine Fisheries Commission thanks all of the individuals who contributed to the development of the American shad stock assessment and its terms of reference. The Commission also thanks the ASMFC Shad and River Herring Technical Committee (TC) and American Shad Stock Assessment Subcommittee (SASC) members who developed the consensus stock assessment report, especially Andrew Kahnle (NY State Department of Environmental Conservation) for his exceptional work as the Chair of the SASC. Members of the SASC include: Andrew Kahnle (Chair), Russ Allen (NJ Division of Fish and Wildlife), Christine Burgess (NC Division of Marine Fisheries), Kathryn Hattala (NY State Department of Environmental Conservation), Mike Hendricks (PA Fish and Boat Commission), Joe Hightower (U.S. Geological Survey/North Carolina State University), Laura Lee (ASMFC/RI Department of Environmental Management), John Olney (Virginia Institute of Marine Science), and Bob Sadzinski (MD Department of Natural Resources). Members of the Shad and River Herring TC include: Mike Hendricks, Chair (PA Fish and Boat Commission), Russ Allen (NJ Division of Fish and Wildlife), Phillips Brady (MA Division of Marine Fisheries), Michael Brown (Department of Marine Resources), Ellen Cosby (Potomac River Fisheries Commission), Phillip Edwards (RI Department of Environmental Management), Ruth Haas-Castro (NMFS), Don Harrison (GA Department of Natural Resources), Kathryn Hattala (NY State Department of Environmental Conservation), Reid Hyle (FL Fish and Wildlife Conservation Commission), Wilson Laney (USFWS), Rich McBride (FL Fish and Wildlife Conservation Commission), Larry Miller (USFWS), John Olney (Virginia Institute of Marine Science),
Cheri Patterson (NH Fish and Game Department), Bill Post (SC Department of Natural Resources), Bob Sadzinski (MD Department of Natural Resources), Tom Savoy (CT Department of Environmental Protection), Jon Siemien (D.C. Fisheries and Wildlife), Michael Stangl (DE Department of Natural Resources and Environmental Control), and Sara Winslow (NC Division of Marine Fisheries). The Commission would also like to thank everyone individuals from outside of the SASC and TC who contributed to the American shad stock assessment, including Kristen Delano, A.C. Carpenter, and many other federal employees, state employees, and academics.
The Commission appreciates the efforts of the Commission staff who assisted in the development and review of the American shad stock assessment. A very special thanks goes to Erika Robbins and Patrick Kilduff for their outstanding efforts in coordinating the stock assessment process, writing and editing substantial parts of the assessment, and ensuring the delivery of a complete and thorough assessment for peer review.
iii
TERMS OF REFERENCE
- 1. Compile and determine adequacy of available life history data for each stock.
- 2. Compile and determine adequacy of available fishery-dependent and/or independent data as indices of relative abundance for each stock.
- 3. Determine most appropriate method of estimating natural mortality.
- 4. Determine which assessment analyses are most appropriate to available data for each stock.
- 5. Assessment methods will range from simple trend analysis to more complex models.
- 6. Estimate biological reference points for each stock where possible.
- 7. Determine current status of each stock where possible.
- 8. Develop recommendations for needed monitoring data and future research.
- 9. Describe the locations and amounts of shad and river herring bycatch in commercial fisheries for mackerel, sea herring, and other pelagic species and estimate the contribution of that bycatch to fishing mortality.
iv
TABLE OF CONTENTS Preface .......................................................................................................................................................... ii Acknowledgements......................................................................................................................................iii Terms of Reference...................................................................................................................................... iv List of Tables .............................................................................................................................................. vii List of Figures .............................................................................................................................................. ix List of Terms................................................................................................................................................ xi 1.0 Introduction to American Shad Stock Assessment for Peer Review ..................................................... 1 1.1 Introduction ........................................................................................................................................... 1 1.1.1 Assessment Approach......................................................................................................... 1 1.1.2 General Biology.................................................................................................................. 3 1.1.3 General Regulatory History ................................................................................................ 8 1.1.4 Assessment History........................................................................................................... 17 1.1.5 Benchmark Mortality Estimation...................................................................................... 19 1.1.6 Age Determination in American Shad .............................................................................. 38 1.1.7 Coastwide Commercial Landings of American Shad ....................................................... 42 1.2 Coastwide Summaries......................................................................................................................... 55 1.2.1 Juvenile Abundance Indices ............................................................................................. 55 1.2.2 Upriver Fish Passage ........................................................................................................ 56 1.2.3 Total Mortality.................................................................................................................. 56 1.2.4 Mean Length/Mean Age Summaries ................................................................................ 57 1.2.5 Coastwide CPUE Trends .................................................................................................. 58 1.3 State and River Stock Assessment Summaries ................................................................................... 87 1.3.1 Maine ................................................................................................................................ 89 1.3.2 New Hampshire ................................................................................................................ 91 1.3.3 Merrimack River............................................................................................................... 93 1.3.4 Rhode Island ..................................................................................................................... 96 1.3.5 Connecticut ....................................................................................................................... 98 1.3.6 Hudson River .................................................................................................................. 103 1.3.7 Delaware Bay and River ................................................................................................. 106 1.3.8 Maryland......................................................................................................................... 106 1.3.9 Susquehanna River ......................................................................................................... 113 1.3.10 Potomac River................................................................................................................. 115 1.3.11 Virginia ........................................................................................................................... 119 1.3.12 North Carolina ................................................................................................................ 128 1.3.13 South Carolina ................................................................................................................ 137 1.3.14 Georgia............................................................................................................................ 146 1.3.15 Florida............................................................................................................................. 149 1.4 Conclusions and Recommendations ................................................................................................. 153 1.5 Causes of Decline ............................................................................................................................. 155 1.5.1 Coastwide Perspective American Shad Stock Contributions in Mixed Stock Fisheries along the Atlantic Coast......................................................................................................................... 155 Relative Rates of Exploitation in Three Atlantic Coastal American Shad Stocks ... 170 Striped Bass Predation on Adult American Shad: Occurrence & Observed Effects on American Shad Abundance in Atlantic Coastal Rivers & Estuaries ................... 182 v
1.5.2 Stock-Specific Perspectives ............................................................................................ 195 Delaware River......................................................................................................... 195 Connecticut River..................................................................................................... 196 Hudson River............................................................................................................ 197 Appendix I..................................................................................................................................... 219 Appendix II ................................................................................................................................... 221 vi
LIST OF TABLES Table 1.1.3-1 Summary of mandatory fishery-independent monitoring programs for American shad as required by Addendum I to Amendment 1 of the Shad and River Herring FMP.................................................................................................................. 13 Table 1.1.3-2 Summary of mandatory fishery-dependent monitoring programs for American shad as required by Addendum I to Amendment 1 of the Shad and River Herring FMP ............................................................................................................................... 15 Table 1.1.5-1 American shad natural mortality rates used in Thompson Bell Biomass-per-recruit model ............................................................................................................................. 29 Table 1.1.5-2 River-specific life history and fishery data used as inputs to the Thompson Bell biomass-per-recruit model ............................................................................................. 30 Table 1.1.5-3 Results of benchmark determination (Z30) from the Thompson-Bell biomass per-recruit model, modified for egg-per-recruit, for regional American shad stocks. ......... 31 Table 1.1.5-4 Results of sensitivity of the Thompson-Bell biomass-per-recruit model to variation in M and river-specific inputs (maturity, fecundity and partial recruitment vector) ......... 32 Table 1.1.7-1 Existing mixed stock fisheries and past directed mixed stock fisheries for American shad along the Atlantic coast ......................................................................................... 46 Table 1.1.7-2 Summary of American shad commercial landings (kg) as reported for inland natal and mixed stock / ocean fisheries by National Marine Fisheries Service and/or respective state agencies ................................................................................................ 47 Table 1.2 List of rivers contained within this stock assessment, information on current American shad fisheries, and data use and availability ................................................. 60 Table 1.2.4 Linear regression statistics for comparison for relationships of mean age and mean total length in several Atlantic coastal rivers................................................................. 64 Table 1.2.5 Summary of trends in catch-per-unit-effort (CPUE) from fishery-dependent and independent datasets for adult American shad using linear regressions of CPUE against year by river and by dataset............................................................................... 65 Table 1.5.1-A1 Estimated American shad stock contributions (percent) in the mixed stock Atlantic coast fisheries, based on DNA and tag-recapture programs ........................................ 159 Table 1.5.1-A2 American shad landings reported for mixed stock fisheries........................................ 160 Table 1.5.1-A3 Landings (kg*1000) from Table 1.5.1-2 summed by area grouping ........................... 161 Table 1.5.1-A4 State stock estimates of harvest by mixed stock fisheries ........................................... 162 Supp Table A1 Tagging studies conducted off the coast of North Carolina ........................................ 163 Supp Table A2 Tagging indices conducted off the coast of South Carolina ........................................ 164 Supp Table A3 Summaries of genetic DNA studies (presented as percent of sample from river origin) conducted off the coasts of Virginia and Maryland (Brown and Epifano 1994) and New Jersey (1996) ...................................................................................... 165 Supp Table A4 Estimated stock composition (% by state/area) of American shad harvest for the 1992 and 1993 ocean-intercept fisheries off Maryland and Virginia, based on mtDNA studies (Brown and Epifano 1995) ................................................................ 166 Supp Table A5 Tagging studies conducted off the coasts of Maryland and Virginia .......................... 167 Supp Table A6 American shad tagging studies conducted in lower Delaware Bay and the Hudson River ............................................................................................................................ 168 Supp Table A7 American shad tagging studies summarized by Dadswell et al. (1987), conducted in the Connecticut River, Hudson River, New York Bight, Delaware River, and Chesapeake Bay........................................................................................................... 169 vii
Table 1.5.1-B1 Indices of adult abundance used in calculating relative exploitation for American shad of the Delaware, Hudson, and Connecticut Rivers.............................................. 173 Table 1.5.1-B2 Commercial harvest, adult abundance indices, and relative exploitation for the Delaware River American shad stock.......................................................................... 174 Table 1.5.1-B3 Commercial harvest, adult abundance indices, relative exploitation, and total mortality estimates for the Hudson River American shad stock ................................. 175 Table 1.5.1-B4 Commercial harvest, adult abundance indices, relative exploitation, and total mortality estimates for the Connecticut River American shad stock........................... 177 Table 1.5.1-C1 Relative abundance indices for adult American shad and striped bass in Atlantic coastal rivers and estuaries .......................................................................................... 186 Table 1.5.1-C2. Number and mean total length striped bass collected from the Hudson River examined in diet analyses, 1990-2006......................................................................... 187 Table 1.5.1-C3 Stomach contents of striped bass collected from the Hudson River Estuary, 1990-2006.................................................................................................................... 188 Table 1.5.1-C4 Relative abundance indices for adult American shad and striped bass in Atlantic coastal rivers and estuaries .......................................................................................... 189 Table 1.5.2 -C1 Survival estimates used as model inputs for Potential spawning stock abundance index for the Hudson River American shad stock ....................................................... 208 Table 1.5.2-C2 F30 based on Egg (EPR) or Biomass (BPR) per recruit for the Hudson River shad stock .................................................................................................................... 209 Table 1.5.2-C3 Various relative indices of spawning stock abundance for American shad of the Hudson River Estuary.................................................................................................. 210 viii
LIST OF FIGURES Figure 1.1.5-1 River-specific life history and fishery data, along with data from ASMFC 1998, used as inputs to the Thompson Bell Biomass-per-recruit model ................................ 33 Figure 1.1.5-2 Biomass-per-recruit and egg-per-recruit for regional American shad stocks................ 35 Figure 1.1.5-3 Sensitivity of changes in BPR model to variation in M, with river-specific input held constant .................................................................................................................. 36 Figure 1.1.5-4 Sensitivity of changes in BPR model to changes in river-specific inputs with M held constant at 0.5 ........................................................................................................ 37 Figure 1.1.7-1 American shad landings from natal Atlantic coast rivers and mixed stock fisheries, 1970-2005. Total landings not specified to natal or mixed ........................ 50 Figure 1.1.7-2 Comparison of tag returns of shad released near Rudee Inlet VA and results of DNA analyses for shad sampled in the same area, 1992 ............................................... 52 Figure 1.1.7-3 Tag returns from studies conducted along the Atlantic coast in lower Delaware Bay, MD-VA, North Carolina and South Carolina coasts............................................. 53 Figure 1.2.1 Normalized juvenile abundance indices by year for Atlantic coast rivers:
a) Merrymeeting Bay, Pawcatuck River, Connecticut River; b) Hudson River, Delaware River; c) Upper Chesapeake Bay, Nanticoke River, and Potomac River; and d) Rappahannock River, York River, James River, and Albemarle Sound............................................................................................................................. 68 Figure 1.2.2. Coastwide plots of fish-lift and fishway annual counts of American shad on rivers along the U.S. Atlantic coast from the Saco River, Maine to the Santee River, South Carolina. ........................................................................................................................ 70 Figure 1.2.3 Plots of scale based estimates of total mortality for American shad stocks by river system on the U.S. Atlantic coast from the Saco River, Maine to the Cape Fear River, North Carolina. ................................................................................................... 74 Figure 1.2.4 Plots of mean length versus mean age by river system from the Saco River, Maine to the Altamaha River, Georgia ..................................................................................... 79 Figure 1.3-1 Shad rivers assessed along the Atlantic coast................................................................ 87 Figure 1.3-2 Shad rivers assessed in the Northeast region................................................................. 88 Figure 1.3-3 Shad rivers assessed in the Mid-Atlantic region.......................................................... 102 Figure 1.3-4 Shad rivers assessed in the Southeast region............................................................... 127 Figure 1.5.1-B1 Relative exploitation of the Delaware River American shad stock using total harvest (ocean/mixed stock and in-river harvest) and the Lewis CPUE as a measure of stock abundance .................................................................................................................... 179 Figure 1.5.1-B2 Relative exploitation of the Hudson River American shad stock using total harvest (ocean / mixed stock and in-river harvest) and ESSB and Egg indices as a measure of stock abundance ...................................................................................................... 180 Figure 1.5.1-B3 Relative exploitation of the Connecticut River American shad stock using total harvest (ocean/mixed stock and in-river harvest) and the Holyoke Lift and In-river CPUE as measures of stock abundance.......................................................... 181 Figure 1.5.1-C1. Length frequency of striped bass, collected in the Hudson River, and used in diet analysis ........................................................................................................................ 190 Figure 1.5.1-C2 Relative abundance of American shad and striped bass in Atlantic coastal rivers and estuaries ................................................................................................................ 191 Figure 1.5.1-C3 Relationship between relative abundance of mature striped bass and mature American shad in Atlantic coastal rivers and estuaries ............................................... 193 Figure 1.5.2-C1 The potential spawning stock abundance (PSSA) and biomass (PSSB) indices for Hudson River American shad ................................................................................ 211 Figure 1.5.2-C2 Comparison of the Hudson River empirical (ESSB) and potential (PSSB) spawning stock biomass indices .................................................................................. 211 ix
Figure 1.5.2-C3 Comparison of the Hudson River empirical (ESSB) spawning stock biomass index with in-river harvest..................................................................................................... 212 Figure 1.5.2-C4 Comparison of the Hudson River potential (PSSB) spawning stock biomass index with estimated total (river and mixed stock) harvest, with correlation between the PSSB and harvest......................................................................................................... 213 Figure 1.5.2-C5 Comparison of observed fishing mortality rates based on catch curve on age and catch curve on repeat spawning versus selected overfishing rate................................ 215 Figure 1.5.2-C6 Outputs of ASPIC (surplus production) model run for the Hudson River American shad stock .................................................................................................................... 215 Figure 1.5.2-C7 Estimated a) stock size and harvest (kg-millions) of Hudson River American shad 1915-2005; b) 1970-2005; c) estimated stock size (N) 1970-2005 ............................. 216 Figure 1.5.2-C8 Observed data and stock reconstruction model predictions of abundance of eggs, YOY, predicted stock and CPUE, and exploitation rate for Hudson River American shad.............................................................................................................................. 217 Figure 1.5.2-C9 Equilibrium catch versus exploitation rate of Hudson River American shad.............. 218 Figure 1.5.2-C10 Comparison of fisheries harvest versus run size of American shad in the Columbia River, Oregon and Washington................................................................... 218 x
LIST OF TERMS Stock Assessment: An evaluation of a stock, including age and size composition, reproductive capacity, mortality rates, stock size, and recruitment.
Benchmarks: A particular value of stock size, catch, fishing effort, fishing mortality, and total mortality that may be used as a measurement of stock status or management plan effectiveness. Sometimes these may be referred to as biological reference points.
Bycatch: That portion of a catch taken in addition to the targeted species because of non-selectivity of gear to either species or size differences; may include non-directed, threatened, endangered or protected species.
Catch Curve: An age-based analysis of the catch in a fishery that is used to estimate total mortality of a fish stock. Total mortality is calculated by taking the negative slope of the logarithm of the number of fish caught at successive ages (or with 0, 1, 2... annual spawning marks).
Catch-Per-Unit-Effort The number or weight of fish caught with a given amount of (CPUE): fishing effort.
Cohort: See Year Class.
De minimis: Status obtained by states with minimal fisheries for a certain species and that meet specific provisions described in fishery management plans allowing them to be exempted from specific management requirements of the fishery management plan to the extent that action by the particular States to implement and enforce the plan is not necessary for attainment of the fishery management plan's objectives and the conservation of the fishery.
Discard: A portion of what is caught and returned to the sea unused. Discards may be either alive or dead.
Exploitation: The annual percentage of the stock removed by fishing either recreationally or commercially.
F30: The fishing mortality rate that will preserve 30% of the unexploited spawning biomass per recruit.
Fish Passage: The movement of fish above or below an river obstruction, usually by fish-lifts or fishways.
Fish Passage Efficiency: The percent of the fish stock captured or passed through an obstruction (i.e., dam) to migration.
Fishing Mortality (F): The instantaneous rate at which fish in a stock die because of fishing.
xi
Habitat: All of the living and non-living components in a localized area necessary for the survival and reproduction of a particular organism.
Historic Potential: Historic population size prior to habitat losses due to dam construction and reductions in habitat quality Iteroparous: Life history strategy characterized by the ability to spawn in multiple seasons.
Mortality: The rate at which fish die. It can be expressed as annual percentages or instantaneous rates (the fraction of the stock that dies within each small amount of time).
Natural Mortality (M): The instantaneous rate at which fish die from all causes other than harvest or other anthropogenic cause (i.e., turbine mortality). Some sources of natural mortality include predation, spawning mortality, and senescence (old age).
Ocean-Intercept Fishery: A fishery for American shad conducted in state or federal ocean waters targeting the coastal migratory mixed-stock of American shad.
Oxytetracycline (OTC): An antibiotic used to internally mark otoliths of hatchery produced fish.
Recovery: Describes the condition of when a once depleted fish stock reaches a self-sustaining or other stated target level of abundances.
Recruitment: A measure of the weight or number of fish that enter a defined portion of the stock, such as the fishable stock or spawning stock.
Relative Exploitation: An approach used when catch is known or estimated, but no estimates of abundance are available. For example, it may be calculated as the catch divided by a relative index of abundance. Long-term trends in relative exploitation are can be useful in evaluating the impact of fishing versus other sources of mortality.
Restoration: In this assessment, this describes the stocking of hatchery produced young-of-year American shad to augment wild cohorts and the transfer of adult American shad to rivers with depleted spawning stocks. Restoration also includes efforts to improve fish passage or remove barriers to migration.
Run Size: The magnitude of the upriver spawning migration of American shad.
Semelparous: Life history strategy in which an organism only spawns once before dying.
Spawning Stock Biomass: The total weight of mature fish (often females) in a stock.
Stock: A part of a fish population usually with a particular migration pattern, specific spawning grounds, and subject to a distinct fishery.
xii
Stock Status: The agreed perspective of the SASC of the relative level of fish abundance Sub-adult: Juvenile American shad which are part of the ocean migratory mixed-stock fish.
Total Mortality (Z): The instantaneous rate of removal of fish from a population from both fishing and natural causes.
Turbine Mortality: American shad mortalities that are caused by fish passing through the turbines of hydroelectric dams during return migrations to the sea.
Year Class: Fish of a particular species born during the same year.
Yield-per-Recruit: The expected lifetime yield per fish of a specific cohort.
Z30: The total mortality rate that will preserve 30% of the unexploited spawning biomass per recruit.
xiii
Section 1 Introduction to American Shad Stock Assessment for Peer Review
1.1 INTRODUCTION
This document provides a benchmark assessment of American shad stocks of the U.S. Atlantic Coast from Maine through Florida. It is organized into three major sections: (1) an assessment overview that provides summaries of the most meaningful data from a coastwide perspective and by state and major river system; (2) a hypothesis section where we explore various analyses suggesting reasons for observed change; and (3) state and river-specific source documents that provide background materials and summaries of all available data.
This document was prepared by the Stock Assessment Subcommittee of the Atlantic States Marine Fisheries Commission (ASMFC) Shad and River Herring Technical Committee. Data were obtained from U.S. Federal and State freshwater and marine resource management agencies, power generating companies, and universities. We routinely interacted with data collectors and managers to insure that application and interpretation of data were appropriate. Data summaries were developed during several regional data workshops whose participants included data generators, resource managers, and assessment scientists. Selection of the most useful data series occurred during a series of assessment workshops that also included data generators as participants.
1.1.1 Assessment Approach We opted to assess Atlantic coastal shad stocks on an individual basis. As an anadromous species, American shad should be assessed and managed by river system. American shad spawn in rivers along the entire U.S. Atlantic coast and there are gradual (latitudinal) differences among river systems in life history attributes. Of greater consequence are river-specific factors such as the presence of dams (with and without fish passage), water quality problems, and estuarine and in-river fisheries. All of these factors lead to river-specific variation in patterns of abundance and in restoration potential.
We used a simple index based approach in this assessment for several reasons. First is that American shad stocks have been exploited in oceanic and estuarine mixed-stock fisheries as well as river-specific fisheries. A few of the mixed-stock fisheries are adequately monitored but there is almost no information about how to allocate the mixed-stock harvest among stocks. Harvest is monitored for most in-river commercial fisheries but recreational harvest is less often monitored. Almost no information is available on discard or bycatch. The data gaps for American shad can be attributed partly to the many fisheries that harvest American shad (both within and between states) but also to the low priority the species receives in agency monitoring efforts. Declines in American shad abundance and in the importance of its commercial fisheries make the species a low priority for most marine fisheries agencies. This understandable prioritization results in there being few long-term fishery-independent indices, except on rivers with fish passage. Fishery-dependent indices provide some of the longest time series on American shad and may extend back to periods of higher abundance, but are often limited by concerns about changes in effectiveness (catchability) of gear over time. Some of the current fishery-independent surveys should be of sufficient length to be useful in assessments 5 to 10 years from now.
Other factors arguing for a simple index-based approach are uncertainties about the age data and the magnitude of various sources of mortality. The one recent age validation workshop (McBride et al. 2005) showed that errors in ageing from scales may result in both imprecision and bias. A detailed analysis of age data may be unproductive until further age validation studies are completed. Total mortality rates reflect the combined impact of intensive fisheries, spawning mortality, predation, and mortality associated with downstream passage at hydroelectric dams in some systems. Almost no stocks have sufficient information to separate mortality into these sources. Uncertainty about natural mortality is perhaps the 1
biggest limiting factor in drawing strong conclusions about the status of American shad. There are no empirical estimates of natural mortality associated with spawning. Inferences about its magnitude are based almost entirely on total mortality rates and spawning marks on scales. Although interpretation of spawning marks on scales needs an updated validation study, they may help in establishing the magnitude of spawning mortality. Unfortunately, a lack of spawning marks may simply be a reflection of intensive fishing; for example, if a high percentage of migrants are harvested fewer will return to spawn.
Considerable uncertainty also exists about the magnitude of predation on American shad. This predation could occur in rivers, estuaries, and in the ocean, and may be an important source of mortality for juvenile or adult American shad. Recent concern has focused on predation by striped bass, whose population has increased coastwide. There is a lot of diet information for striped bass, but the magnitude of predation mortality is difficult to assess because of uncertainty about the proportion of the striped bass population within different bodies of water. The final uncertainty about American shad mortality is the magnitude of ocean harvest. Directed ocean-intercept fisheries were closed in 2005, too recently for the impact of that closure to be evident in this assessment. Mortality due to bycatch in other ocean fisheries is basically unknown.
In this assessment, we focused on evaluations of selected indices of abundance for age-0 and mature shad from various fishery-dependent and independent sample programs, passage numbers at major dams, age and size data, and calculations of total mortality rates. We developed benchmark estimates of total mortality from simple regional biomass-per-recruit models. Data summaries were at the stock level where possible, although ancillary data from mixed stock sources were occasionally used as supporting information. We generally avoided use of tag-recapture population estimates of spawning stock abundance because their use requires many unverified assumptions and we avoided analyses requiring harvest-at-age by stock. We also avoided making sweeping adjustments of entire data time series based on one or two point estimates of adjustment factors. Complex model building and creative data analyses were confined to the hypotheses section.
There were several categories of data that we did not use in this assessment. We explored, but did not use data from the Marine Recreational Fisheries Statistics Survey (MRFSS) for several reasons. Recreational fishermen rarely catch American shad in marine waters. This leads to bias in catch estimates by the MRFSS because data on American shad are often missing from sampling units and must be substituted from larger units. There is also considerable concern about potential species misidentification. Anecdotal evidence from state biologists indicates that hickory shad, which is growing in abundance, have been misidentified as American shad, especially by anglers. Since we had no way to evaluate the potential for these biases, we did not pursue estimates of recreational harvest. We also evaluated, but did not use catches of American shad in the Northeast Fisheries Science Center bottom trawl survey as an abundance index. Again, incidence of shad in the data was low and resulting estimates of CPUE were highly variable and imprecise. Finally, we did not have time to provide an adequate analysis of shad bycatch in commercial fisheries in ocean waters. Data on incidence of American shad in commercial catches were available from National Marine Fisheries Service (NMFS) observer programs and commercial fishing trip reports. However, observations of American shad were a rare event and reasonable extrapolations of total bycatch require a thorough knowledge of NMFS databases and complex analyses to fill in the many missing values in sample units.
This assessment may not provide definitive answers to all the questions plaguing management of Atlantic coastal American shad (fishing, predation, other sources of natural mortality, fish passage), but it gives insight to managers on the complexity of the issues to assist them in their decision-making. It also lays the foundation for future assessments in terms of data sources and methods. Updating these datasets in future years should be straightforward and as datasets get longer, concerns about ageing methods are addressed, and the influence of ocean harvest diminishes these questions may be able to be answered. If the datasets are stronger in future assessments, they may support the use of more complex assessment models.
2
1.1.2 General Biology A broad overview of American shad and Alosa spp. status, biology, ecology, and population dynamics is available in Limburg and Waldman (2003). Additional information on life history can be found within the individual source documents for each assessment.
Migration and Genetic Information Migration Shad spend most of their life at sea, returning to their natal river system only when sexually mature. Prior to the 1950s, very little was known or understood about the ocean migratory routes of immature and adult fish. In one of the first attempts to study shad migration, Vladykov (1950) tagged shad on their spawning run in the St. Lawrence River. Out of system returns ranged from Cape Cod, Massachusetts to ocean waters off Nova Scotia. Talbot and Sykes (1958) conducted their own tagging studies and analyzed data from several tagging experiments of the U.S. Bureau of Commercial Fisheries. They found that both immature and mature shad from the Chesapeake and north summered in the Gulf of Maine, moved south along the coast in the fall, and over-wintered in large groups between North Carolina to Long Island, New York (Talbot and Sykes 1958). Juvenile shad that left their natal system in fall were assumed to have spent the winter in the mid-Atlantic area before joining the coastal migration to the Gulf of Maine.
Milstein (1981) reported that young (age-0 and age-1) American shad spend their time in near-shore estuaries supporting Talbot and Sykes (1958) model of juvenile American shad distribution. Talbot and Sykes (1958) also indicated some tag returns from fish tagged in the Gulf of Maine came from southern semelparous stocks. They surmised that these fish were immature when tagged while summering there in previous years.
More recent studies (Dadswell et al. 1987; Hattala et al. 1998) support these generalized migration patterns. Adult American shad spend summers in northwestern Atlantic waters as far north as the Gulf of Maine, the Bay of Fundy and off the coast of Nova Scotia. In the fall, they begin a southward migration to over-wintering areas off of Maryland, Virginia, and North Carolina.
Several researchers have found that shad prefer fairly narrow ranges of water temperatures, which appear to regulate migratory behavior (Leggett and Whitney 1972; Leggett 1977). These studies concluded that adult American shad migration generally followed a northerly route to the Gulf of Maine and Bay of Fundy in summer, before returning south in the fall following the coast to over-wintering areas near the mid-Atlantic region of the U.S. Neves and Depres (1978) refined this pattern by examining shad presence in NMFS trawl data versus water temperature. They concluded that American shad followed fairly specific temperature windows of 3 to 15ºC during their migration at sea. American shad over-winter from North Carolina to New Jersey; however, Cape Hatteras, North Carolina appears to serve as a dividing point for pre-spawning fish of southern and northern rivers. Water temperatures south of Cape Hatteras generally average about 15ºC. Southern stocks move inshore at or near Cape Hatteras to be inside the Gulf Stream. They follow near-shore southerly currents averaging 15ºC to reach their natal river. Parkers (1992) study tagging shad in southeastern North Carolina confirmed this southerly migration pattern. Pre-spawning fish of northern stocks also move inshore in areas, but north of Cape Hatteras, and move north as northern edge of water temperatures warm to about 4ºC to enter their spawning rivers (Neves and Depres 1978). Peak runs into rivers north of Cape Hatteras occur when water temperatures range from 10 to 15ºC.
Genetic Information The timing of the pre-spawning shad migration along the coast is important to understand as commercial fisheries from South Carolina to New Jersey and in Rhode Island have exploited this mixed stock 3
assemblage. The ocean-intercept fishery was closed in 2005, as states instituted regulations that followed ASMFC Amendment 1 to the Shad and River Herring Interstate Fishery Management Plan on January 1, 2005 (see Section 1.1.4).
Investigations were conducted to determine composition of shad captured in the mixed stock directed ocean intercept fisheries in Virginia through New Jersey from the late 1980s to the mid-1990s (Brown and Epifano 1994, Brown 1996). Chapman (1993) investigated the genetic composition of the Susquehanna River stock. Bentzen et al. (1989) studied the genetic differences in relation to life history strategies (semelparity versus iteroparity) of east coast American shad stocks.
All studies used restriction enzyme analysis on mitochondrial DNA to determine the genotype frequencies for each stock. Many of the genotypes were common to most stocks, with some geographic clumping. Results (uniqueness of stock) were based on maximum likelihood analysis of genotype frequency distributions. The conclusions of Brown (1996) came under scrutiny when genetic identity was questioned based on comparisons of the otolith marks for each fish sample, which indicated a different stock origin of the same sample.
Age Please refer to Section on Age Determination (1.1.6).
We do not know the maximum age American shad can attain in the absence of fishing, since all East Coast American shad stocks have been exploited. Most East Coast shad stocks have been exploited prior to the written record in the mid-1800s through the present. Maximum age of southern semelparous (Florida to South Carolina) stocks follows maximum age of maturity. Most fish are mature by age-7.
However, the maximum observed age for the Altamaha River, Georgia was age-8 (Godwin and McBay 1967). In the Hudson, shad grew to age-10 in the years following World War II after the stock had experienced a major collapse (Talbot 1954). Recent data from the Hudson River stock indicate that female American shad can reach age-13. Males reach about age-10. Reported maximum age in other east coast iteroparous stocks range from ages 9 to 11 (Connecticut, Leggett 1976; Virginia, Olney and OReilly 2002; and North Carolina, S. Winslow, pers. comm.). Maximum age reported for Maine rivers was age-12 (see Section 9.2).
Growth Growth data are presented in individual chapters.
Reproduction Spawning runs of American shad begin in Florida in mid-November or early December and sequentially move north with latitude (Mansueti and Kolb 1953). Runs in Georgia and South Carolina generally begin in January and peak in March. By April, American shad runs begin to peak in North Carolina, followed by the runs in Virginia and Maryland. In Delaware, Pennsylvania, and New York, spawning typically peaks in late spring (May) and may extend into early June. New England and Canadian shad spawning runs range from mid-May to late July.
Most American shad native to the rivers of the southeastern U.S. are semelparous, however, in rivers to the north, stocks are iteropaous (Carscadden and Leggett 1975). There is no distinct geographic dividing point between iteroparous and semelparous life-history strategies. Although available data suggest that the transition occurs in southern North Carolina, validation studies of repeat spawning mark determination on scales in this region have not been conducted.
4
American shad are broadcast spawnerstheir fertilized eggs are carried by river currents and hatch within 2 to 17 days depending on water temperatures (Jones et al. 1978). Larvae drift with the current until they mature into juveniles. The clinal variation in repeat spawning is considered an energy trade off in terms of fecundity to migration and perhaps growth. Since reproductive characteristics vary with latitude, discussion of percent of repeat spawning, maturity, and fecundity are presented separately in the individual chapters.
Leggett and Carscadden (1978) described reciprocal trends between the degree of repeat spawning in American shad and relative fecundity with increasing latitude. A recent investigation of batch fecundity of American shad from the Connecticut River (Connecticut), York River (Virginia), and St. Johns River (Florida) showed that batch fecundity does not vary significantly across latitudes (Olney and McBride 2003).
Natural Mortality American shad have a complex life history strategy and natural mortality likely varies by life stage.
Mortality while at sea is unknown but would represent a combination of natural death, bycatch, and directed ocean harvest. There is no information about at-sea survival except for analyses done on cohorts of stocked fish (Section 10). Natural mortality at sea should be similar to other clupeids.
Post-spawning mortality of adults has been observed in the Delaware River (Chittenden 1976). Weight loss during their freshwater stay varied by size and sex, with an average somatic weight loss of 42% for males and 50% for females (Chittenden 1976). The weight loss is attributed to the lack of suitable prey resources and energetic costs of migration, final gonadal development, and spawning (Chittenden 1976).
In other systems, post-spawning mortality is primarily inferred by the presence or absence of spawning marks on scales. Ages based on scales suggest that instantaneous total mortality rates (Z) are high for some populations, but the contribution of fishing (F) and natural (M) mortality to Z is poorly understood because of limitations in tagging programs and fishery records.
Please refer to Benchmark Mortality Calculations section (1.1.5).
Literature Cited Bentzen, P., G.C. Borwn, and W.C. Leggett. 1989. Mitochondrial DNA polymorphism, population structure, and life history variation in American shad (Alosa sapidissima). Canadian Journal of Fisheries and Aquatic Sciences. 46:1446-1454.
Brown, B.L. 1996. Mixed-stock analysis of American shad in New Jersey's coastal intercept fishery.
Final report for a two-year research project funded by the Hudson River Foundation. Ecological Genetics Labortaory, Virginia Commonwealth University. Richmond, VA.
Brown, B. L. and J. M. Epifanio. 1994. Mixed-stock analysis of American shad in Virginia's and Maryland's coastal intercept fisheries. Report to Virginia Marine Resources Commission, Report
- F-110-R, 115 pp.
Carscadden, J. E. and W. C. Leggett. 1975. Meristic differences in spawning populations of American shad, Alosa sapidissima: evidence for homing to tributaries in the St. John River, New Brunswick. Fisheries Research Board of Canada Bulletin 32:653-660.
Chapman, R.W. 1993. Genetic investigation of the resurgence of American shad in the Susquehanna River. Final report to Maryland Department of Natural Resources, Chesapeake Bay Research and Monitoring Division, Power Plant Topical Research Program.
5
Chittenden, M.E., Jr. 1976. Weight loss, mortality, feeding, and duration of residence of adult American shad, Alosa sapidissima, in fresh water. Fishery Bulletin 74:151-157 Dadswell, M., G. Melvin, P.J. Williams and D.E. Themelis. 1987. Influences of origin, life history, and chance on the Atlantic coastal migration of American shad. IN Common Strategies of Anadromous and Catadromous fishes, American Fisheries Society, Symposium 1.
Godwin, W.F. and L.G McBay. 1967. Preliminary studies of the shad fishery of the Altamaha River GA.
Georgia Game and Fish Division, Marine Fisheries Division, Brunswick, Georgia.
Hattala, K., R. Allen, N. Lazar and R. OReilly. 1998. Stock contributions for American shad landings in mixed stock fisheries along the Atlantic coast. ASFMC, American shad Stock Assessment Peer Review Report, Washington, D.C.
Jones, P. W., F. D. Martin, and J. D. Hardy, Jr. 1978. Development of fishes of the mid- Atlantic Bight.
An atlas of egg, larval, and juvenile stages. Vol 1. Acipenseridae through Ictaluridae. U.S. Fish and Wildlife Service, FWS/OBS-78/12.
Leggett, W.C. 1976. The American shad, Alosa sapidissima, with special reference to its migrations and population dynamics in the Connecticut River. American Fisheries Society, Merriman and Thorpe, eds. The Connecticut River Ecological Study: the impact of a nuclear power plant.
Monograph 1. Bethesda, Maryland.
Leggett, W.C. 1977. Ocean migration rates of American shad. J. Fish. Res. Board Can. Vol 34, pp 1422-1426.
Leggett, W.C. and J.E. Carscadden. 1978. Latitudinal variation in reproductive characteristics of American shad, Alosa sapidissima, evidence for population specific life history strategies. J. Fish.
Res. Board Can. Vol 35, pp 1469-1478. Ottowa, Canada.
Leggett, W.C., and R.R. Whitney. 1972. Water temperature and the migrations of American shad. Fishery Bulletin, U.S. 70 (3): 659-670.
Limburg, K.E., and J.R. Waldman, editors. 2003. Biodiversity, status and conservation of the worlds shads. American Fisheries Society, Symposium 35. Bethesda, Maryland.
Mansueti, R.J., and H. Kolb. 1953. A historical review of the shad fisheries of North America.
Chesapeake Biological Laboratory, Publication No. 97. 293 p.
Milstein, C. 1981. Abundance and distribution of juvenile Alosa species off southern New Jersey.
Transactions of American Fisheries Society 110:306-309. Bethesda, Maryland.
Neves, R.J. and L. Depres. 1978. The oceanic migration of American shad, Alosa sapidissima, along the Atlantic coast. Fishery Bulletin 77 (1): 199-211.
Olney, J., and R. OReilly. 2002. A summary of Virginias American shad fisheries in 2001 including monitoring and restoration program. Atlantic States Fisheries Marine Commission, Annual Repot, Washington D.C.
Olney, J.E., and R.S. McBride. 2003. Intraspecific variation in batch fecundity of American shad:
revisiting the paradigm of reciprocal latitudinal trends in reproductive traits. Pages 185-192 in 6
Biodiversity, Status, and Conservation of the World's Shads. American Fisheries Society Symposium 35.
Parker, J. A. 1992. Migratory patterns and exploitation of American shad in the nearshore ocean waters of southeastern North Carolina. North Amer. J. Fish. Management. 12:752-739.
Talbot, G. E. 1954. Factors associated with fluctuations in abundance of Hudson River shad. U.S. Fish and Wildlife Service Fish. Bull. 56:373-413.
Talbot, G. And J Sykes. 1958. Atlantic coast migrations of American shad. U.S. Fishery Bulletin 58(142),
pp 473-490.
Vladykov, V.D. 1950. Movements of Quebec shad (Alosa Sapidissiam) as demonstrated by tagging.
Naturaliste Canadien. 77(5-6):121-135.
7
1.1.3 General Regulatory History1 American shad have waned in their importance as a food fish since the turn of the 19th century, when they were among the top three species harvested on the Atlantic coast (U.S. Fish Commission 1872-1881). In the late 1880s through the early 1900s shad harvest was massive, nearly 50 million pounds coastwide per year. But in the years shortly following, most East Coast shad stocks experienced serious declines. Stock collapses were a coastwide event (see Section 1.1.7). Overfishing, habitat loss from dam construction, dredge and fill operations, as well as habitat degradation (pollution) are among the primary cause.
Change in human practices was slow to come and the same pattern of events occurred in most all states within another 50 years. Following WWII, some East Coast stocks experienced a second collapse, faulted primarily to overfishing during the war and the seven to ten year period that followed (Talbot 1954; Fredin 1954). Degenerative environmental conditions further contributed to the declines. Degradation that began in the 1800s grew worse: constructing major dams and destroying, providing little or no passage, filling shallow water spawning habitat, and worsening water quality problems associated with pollution created low and no-oxygen blocks in major portions of large rivers (two of the most infamous occurring in the Delaware and the Hudson).
By the early 1980s, stock sizes for most Atlantic coast shad stocks were, in reality, just remnant populations. While moratoriums were in place in several states (Maryland and New England),
commercial harvest of American shad continued in others. Despite stock declines and moratoria, the once traditional in-river spring fisheries for roethe eggs are considered a delicacyexpanded in recent years to include late winter and early spring fisheries in ocean waters and large coastal bay waters. These ocean and bay fisheries exploited the pre-spawning migration of American shad. Current fisheries are relatively small, compared to the magnitude of past fisheries, but it should be noted that current fisheries continue to operate on much smaller stocks than those present 40 or even 20 years ago.
The earliest records of regulations pertaining to management of American shad fisheries date back to the 1700s when Massachusetts, Maryland, and Virginia took legislative action toward managing their fisheries (ASHP 2006; A.C. Carpenter, pers. comm.). Since this time, individual state regulations have managed American shad stocks using gear and effort restrictions, and closed areas and seasons for recreational and commercial fisheries, and catch and effort data reporting for commercial fisheries. State and river-specific regulatory histories and current regulations are contained in the individual reports.
The written historical record varies and is often vague on the driving forces behind the development of fishery management strategies for East Coast states. Beginning about 1868, mid-Atlantic and New England states legislated the creation of fish commissions recognizing the tremendous importance of fish, particularly shad, as a food source. Accounts from the New York Forest, Fish and Game Commission (1908) reported that state commissioners from Maine to Pennsylvania met at a convention in New York City in 1868 to Apromote uniformity of law and action among the States@ and to discuss the regulation of nets and netting, the limitation of fishing seasons, and the selection of suitable fishways with most all of these rules pertaining to take of shad. This is one of the first references on actions taken to regulate fisheries. However, subsequent reference to management actions quickly disappeared, with the primary focus centered on development of hatchery culture.
The results of the 1868 convention are thought to have directly influenced the foundation of the US Commission of Fish and Fisheries (New York Forest, Fish and Game Commission 1908). This federal commission came into existence in 1871 and took on the task of increasing the supply of readily available fish, with American shad as the most prominent species, to meet the increasing demands of a 1
Portions of this section have been taken from the Atlantic States Marine Fisheries Commissions Amendment 1 to the Interstate Fishery Management Plan for Shad and River Herring (1999).
8
growing population in an expanding country. The Commission=s tenet of investment of time and money in hatchery programs was deemed necessary on the federal level so that Ano one state would be burdened with such an enormous task@ (U.S. Fish Commission 1875), even though many states had already made the commitment on their own.
Shad stocks were still quite large during this time period (1870-1890), although it appears from the writings of the U.S. Fish Commission (USFC) and New York Forest, Fish and Game Commission that runs were in the first serious decline (recorded in the written record). One reference from a New York report mentions there were few fish to speak of from southern rivers, referring specifically to the Potomac River. New York was in search of brood stock to assist in replenishing the Hudson stock, perceived to be in major decline. At the same time, the USFC reported landings from the Potomac at about 6 million fish per year. This number is relatively small, and hence the New York assertion that few fish were in mid-Atlantic rivers, given that another USFC report indicated that about 10 to 20 years earlier Potomac River landings were five times higher at about 33 million.
On the federal level, USFC reports rarely refer to any problems with overfishing, but state that hatchery production would solve the problems related to fishing practices and the Aother drawback of pollution@
in spawning areas. Some states, however, did try to address the changes in fishing practices legislating licensing of fishers, escapement periods and some gear restrictions (Legislature of the State of New York, 1868, cited by Harpers Weekly 1872). But in spite of these attempts at management, most states focused on hatchery production as the best solution to fix the problem of declining runs.
As J.L. McHugh (1970) stated, [i]t is easy, armed with hindsight, to be critical of the past, concerning fisheries management practices. He also indicated that the philosophy of leaders at the time, Spencer Baird of the USFC along with others in many East Coast states, placed great importance on hatchery culture as a solution to the problems of marine and freshwater fisheries. Under these leaders, the federal and state governments embarked on a vigorous and apparently completely futile program of fish culture for more than 60 years.
The lack of success of these culture programs can be attributed to the prevalent attitudes of the time.
Although some individuals recognized the need to regulate fishermen, the prevalent attitude was that fish commissions could not regulate business nor take away a person=s livelihood. In addition, the growing human population created a different set of problems, none of which could be controlled by fish commissions. The environmental degradation from pollution (i.e., sewage wastes, mill effluents, ash and cinder dumped on spawning areas) was tolerated and accepted as difficulties to be endured. So in the face of these obstacles, fisheries policy was developed for what could be donehatchery stockingand not what should be addressedpollution and fisheries (New York Forest, Fish and Game Commission 1908). Unfortunately this policy continued for the next 100 years.
It is not very clear when the emphasis on hatchery culture began to wane. Most references on hatchery culture practices began to disappear after World War II, a period when a slow but evident awakening in fisheries management was about to begin.
At the beginning of WWII, the pioneer experiment began with a consortium of seven states recognizing the need for cooperation to address the condition of coastal fish stocks. The creation of the Atlantic States Marine Fisheries Commission in 1941 initially focused on increasing the food fish supply as it was commonly thought that food will win the war (ASMFC 1942). Increasing and maintaining fisheries production was of utmost importance to the war effort. The ASMFC used U.S. Fish and Wildlife Service staff to serve as the primary research arm of the newly formed commission. Early work focused on the need to create a statistical catch reporting system so that consequences of states= actions could be tracked to avoid harm to production levels. Also recognized was the need for a [single] management plan to be implemented as soon as the war was over.
9
Some species were recognized as needing special attention: striped bass, flounder, and red drum. Other concerns focused on methods to increase fishing efficiency following the loss of vessels and crew to the war effort and allowing of expansion of fishing areas to inshore areas for trawling. During this period of cooperation, states were still responsible for regulating their own fisheries. Many states relaxed or suspended fishing rules during the war years from 1942 through then end of WWII. Although this temporarily increased production, fishing under relaxed regulations took its toll on shad stocks. The condition of American shad stocks, particularly those of the mid-Atlantic, began to decline throughout the war years, leading up to a second major collapse within a century.
In 1949, the Beaufort, North Carolina Laboratory, under the U.S. Bureau of Commercial Fisheries reactivated, following its closure during WWII, to investigate the declines in East Coast American shad stocks. Overfishing was still not thought of as an important factor on stock size. Mansueti and Kolb (1953) listed overfishing as a possible effect after pollution, siltation of spawning areas, and dam construction. The Lab=s Shad Project (Talbot, Sykes, Fredin, Walburg, and Nichols, among others) received their initial instruction by Dr. R.J.H. Beverton, the father of fish population dynamics analyses. After which they began a series of in-depth investigations in shad populations and factors influencing their abundance in several East Coast systems. The investigators produced what are now considered the Aclassics@ in shad research: studies from the Hudson River (Talbot 1954), Connecticut River (Fredin 1954), Maryland rivers (Walburg 1955), York River (Nichols and Massman 1963), Neuse River (Walburg 1957), Edisto River (Walburg 1956), and St. Johns River, Florida (Walburg 1960). These studies were among the first steps toward understanding population dynamics through analyzing statistical data. Although many factors were identified as influencing shad abundance the recurrent cause of stock decline in most systems during the WWII period was overfishing.
When the anadromous fish program ended in 1967, major shad research projects were completed in systems from the Connecticut River south to the St. Johns River, Florida. A James Sykes, a leader of some of the anadromous fish projects, retrospective (Wolfe 2000) stated that the greater accomplishment came by way of laying the ground work for future state management of fisheries through provision of baseline population data and methods of stock assessment that were made available to state marine research agencies, many of which were non-existent at the time.
By 1980, coastwide landings of American shad had fallen from approximately 50 million pounds at the beginning of the twentieth century to 3.8 million pounds. Large declines in commercial landings were perceived as an indication that management action would be required to restore American shad to their former levels of abundance. Therefore, the members of the Atlantic States Fishery Commission (ASMFC) recommended the preparation of a cooperative Interstate Fishery Management Plan (FMP) for American Shad and River Herring. The ASMFC adopted this recommendation in 1981 and the FMP was completed in 1985. The FMP specified recommended management measures, focused primarily on regulating exploitation and enhancing stock restoration efforts. At the time of the 1985 FMP, the ASMFC did not have authority over individual state fisheries and implementation of the Plan was at the discretion of the states. The ASMFC approved a supplement to the FMP in 1988, which included reports prepared by the Shad and River Herring Stock Assessment Subcommittee, summaries of material presented at the 1987 Anadromous Alosine Research Workshop, and changes to management recommendations and research priorities based on new research findings.
In 1994, American shad stocks were continuing to decline, which led the Shad and River Herring Plan Review Team and Management Board to determine that the 1985 FMP was not adequate for protecting or restoring the remaining shad and river herring stocks. The 1985 FMP did not require any specific management approach or monitoring requirements within the management unitit only asked that states provide annual summaries of restoration efforts and ocean fishery activity. To address the shortcomings of the 1985 FMP, the ASMFC implemented Amendment 1 to the Interstate Fishery Management Plan for 10
Shad and River Herring. The 1993 Atlantic Coastal Fisheries Cooperative Management Act helped in this effort by requiring states to adopt management guidelines in this and other approved Commission Plans.
Amendment 1 was approved in 1999 (ASMFC 1999).
The goal of Amendment 1 is to protect, enhance, and restore East Coast migratory spawning stock of American shad, hickory, shad, and river herring (alewife and blueback herring collectively) in order to achieve stock restoration and maintain sustainable levels of spawning stock biomass. To meet this goal, the Amendment identifies several objectives. The objectives listed below are pertinent to American shad:
- 1. Prevent overfishing of American shad stocks by constraining fishing mortality below F30.
- 2. Develop definitions of stock restoration, determine appropriate mortality rates, and specify rebuilding schedules for American shad populations within the management unit.
- 3. Promote improvements in degraded or historic alsoine habitat throughout the species range.
- 4. Establish criteria, standards, and procedures for plan implementation as well as determination of states compliance with management plan provisions.
Amendment 1 established a five-year phase out of the ocean-intercept fishery for American shad by January 1, 2005 and required fishing mortality targets for specific in-river fisheries (Connecticut River, Hudson River, Delaware River, Upper Chesapeake Bay, Edisto River, Santee River, and Altamaha River). It also required states to implement an aggregate 10-fish daily creel limit in recreational fisheries for American shad and hickory shad in all rivers except the Santee River, where an aggregate limit 20-fish (American and hickory shad) per day creel limit was enacted.
In addition to establishing fishing regulations, Amendment 1 established monitoring requirements for American shad, including juvenile abundance and adult spawning stock surveys and in-river creel surveys. The specific monitoring requirements, both fishery-dependent and independent, have been modified through a Technical Addendum in 2000 and an Addendum in 2002. The current monitoring requirements are contained in Tables 1.1.3-1 and 1.1.3-2. All states have implemented regulations that are compliant with Amendment 1.
Literature Cited ASHP (Atlantic salmon history project). 2006. Available: home.gwi.net ASMFC. 1999. Amendment 1 to the Interstate Fishery Management Plan for Shad and River Herring.
Report No. 35, 76 pp.
ASMFC (Atlantic States Marine Fisheries Commission). 1942. Proceedings of the First annual meeting, New York City, NY.
Fredin, R.A. 1954. Causes of fluctuations in abundance of Connecticut River shad. US Fishery Bulletin 88, Vol 54, pp 247-259.
Harpers Weekly. 1872. Seth Greens shad hatchery. Harpers Weekly, New York.
Mansueti, R. and H. Kolb. 1953. A historical review of the shad fisheries of North America.
Chesapeake Biological Laboratory Publication No. 97. 293 pp.
McHugh, J. L. 1970. Trends in Fishery Research. IN A Century of Fisheries in North America, N.
Benson, Editor, American Fisheries Society. Special publication No. 7, Washington, DC, USA.
New York Forest, Fish and Game Commission. 1908. Fourteenth annual report of the New York Forest, Fish and Game Commission to the NY Legislature, 1909. Albany, NY, USA.
11
Nichols, P.R. and W.H. Massmann. 1963. Abundance, age and fecundity of shad, York River, VA., 1953-59.
Talbot, G. 1954. Factors associated with fluctuations in abundance of Hudson River shad. US Fishery Bulletin 101, Vol 56, pp 373-413.
U.S. Commission of Fish and Fisheries.1872-1882. Report of the Commissioner of the United States Commission of fish and fisheries. Government Printing Office, Washington, DC, USA.
Walburg, C.H. 1956. Commercial and sport fisheries of the Edisto River South Carolina 1955. Special Scientific Report - No. 187, October 1956. United States Department of the Interior - Fish and Wildlife Service, Washington, D.C.
Walburg, C.H. 1960. Abundance and life history of shad St. John=s River Florida, Fishery Bulletin 177, from Fishery Bulletin of the Fish and Wildlife Service, Volume 60.
Walburg, C.H. 1957. Neuse River shad investigations, 1953. U.S. Fish and Wildlife Service, Fisheries No. 206, Washington, D.C.
Walburg, C.H. 1955. Relative abundance of Maryland shad, 1944-1952. Research Report 36, U.S. Fish and Wildlife Service, Washington D.C.
Wolfe, D.A. 2000. A History of the Federal Biological Laboratory at Beaufort NC 1899-1999. U.S.
National Oceanic and Atmospheric Administration, Beaufort, North Carolina.
12
Table 1.1.3-1 Summary of mandatory fishery-independent monitoring programs for American shad as required by Addendum I to Amendment 1 of the Shad and River Herring FMP.
State System Sampling Program (annual unless otherwise noted)
ME Androscoggin
- Annual spawning stock survey and representative sampling for biological data
& Saco Rivers
- Calculation of mortality and/or survival estimates
- Hatchery Evaluation NH Exeter River
- Annual spawning stock survey and representative sampling for biological data
- Calculation of mortality and/or survival estimates MA Merrimack
- Annual spawning stock survey and representative sampling for biological data River
- Calculation of mortality and/or survival estimates RI Pawcatuck
- Annual spawning stock survey and representative sampling for biological data River
- Calculation of mortality and/or survival estimates CT Connecticut
- Annual spawning stock survey and representative sampling for biological data River
- Calculation of mortality and/or survival estimates
- JAI: Juvenile abundance survey (GM)
NY Hudson River
- Annual spawning stock survey and representative sampling for biological data
- Calculation of mortality and/or survival estimates
- JAI: Juvenile abundance survey (GM)
- Annual spawning stock survey and representative sampling for biological data River
- Calculation of mortality and/or survival estimates
- JAI: Juvenile abundance survey (GM)
NJ Delaware
- Annual spawning stock survey and representative sampling for biological data River
- Calculation of mortality and/or survival estimates
- JAI: Juvenile abundance survey (GM)
PA Susquehanna
- Annual spawning stock survey and representative sampling for biological data River
- Calculation of mortality and/or survival estimates
- JAI: Juvenile abundance survey (GM)
- Hatchery Evaluation Lehigh River
- Annual spawning stock survey and representative sampling for biological data
- Calculation of mortality and/or survival estimates
- Hatchery Evaluation Delaware
- Annual spawning stock survey and representative sampling for biological data River
- Calculation of mortality and/or survival estimates
- JAI: Juvenile abundance survey (GM) 13
Table 1.1.3-1 (cont.) Summary of mandatory fishery-independent monitoring programs for American shad as required by Addendum I to Amendment 1 of the Shad and River Herring FMP (continued).
State System Sampling Program (annual unless otherwise noted)
- Annual spawning stock survey and representative sampling for biological data
- Calculation of mortality and/or survival estimates MD Upper Chesapeake
- Annual spawning stock survey and representative sampling for biological Bay data
- Calculation of mortality and/or survival estimates
- JAI: Juvenile abundance survey (GM)
- Hatchery Evaluation DC Potomac River
- Annual spawning stock survey and representative sampling for biological data
- Calculation of mortality and/or survival estimates VA James, York, and
- Annual spawning stock survey and representative sampling for biological Rappahannock data Rivers
- Calculation of mortality and/or survival estimates
- JAI: Juvenile abundance survey (GM)
- Hatchery Evaluation NC Albemarle Sound
- Annual spawning stock survey and representative sampling for biological and its tributaries, data Tar-Pamlico, Neuse,
- Calculation of mortality and/or survival estimates and Cape Fear
- Hatchery Evaluation Rivers SC Santee-Cooper
- Annual spawning stock survey and representative sampling for biological system, Eidsto River, data Winyah Bay and
- Calculation of mortality and/or survival estimates tributaries (Waccamaw and Pee
- State may elect to sample these systems on a rotational basis (i.e., one system Dee Rivers)* evaluated per year)
GA Altamaha River
- Annual spawning stock survey and representative sampling for biological data
- Calculation of mortality and/or survival estimates FL St. Johns River
- Annual spawning stock survey and representative sampling for biological data
- Calculation of mortality and/or survival estimates 14
Table 1.1.3-2 Summary of mandatory fishery-dependent monitoring programs for American shad as required by Addendum I to Amendment 1 of the Shad and River Herring FMP.
State System Sampling Program ME In-river
- Monitor recreational landings, catch and effort every 5 years.
- Mandatory reporting of catch (numbers and weight) and effort from commercial fisheries; sub-samples shall indicate size, age, and sex composition of catch.
NH In-river/coastal
- Monitor recreational landings, catch and effort every 5 years.
MA Merrimack
- Monitor recreational landings, catch, and effort every 5 years.
River and Connecticut River CT Connecticut
- Mandatory reporting of catch (numbers and weight) and effort from commercial River fisheries; sub-samples shall indicate size, age, and sex composition of catch.
- Monitor recreational landings, catch, and effort every 5 years.
RI Pawcatuck
- Monitor recreational catch and effort every 5 years.
River Atlantic Ocean
- Mandatory reporting of catch (numbers and weight) and effort from commercial fisheries; sub-samples shall indicate size, age, and sex composition of catch.
- Participate in Ocean landings stock composition study.
NY Hudson River
- Mandatory reporting of catch (numbers and weight) and effort from commercial fisheries; sub-samples shall indicate size, age, and sex composition of catch.
- Monitor recreational landings, catch, and effort every 5 years.
- Monitor recreational landings, catch, and effort every 5 years.
River*
NJ Delaware River
- Mandatory reporting of catch (numbers and weight) and effort from commercial and Bay* fisheries; sub-samples shall indicate size, age, and sex composition of catch.
- Monitor recreational landings, catch, and effort every 5 years.
- Mandatory reporting of catch (numbers and weight) and effort from commercial fisheries; sub-samples shall indicate size, age, and sex composition of catch.
- Participate in Ocean landings stock composition study.
- Mandatory reporting of catch (numbers and weight) and effort from commercial and Bay fisheries; sub-samples shall indicate size, age, and sex composition of catch.
- Monitor recreational landings, catch, and effort every 5 years.
Nanticoke River
- Mandatory reporting of catch (numbers and weight) and effort from commercial Chesapeake Bay fisheries; sub-samples shall indicate size, age, and sex composition of catch.
tributary
- Monitor recreational landings, catch, and effort every 5 years.
(upstream portion)
- Mandatory reporting of catch (numbers and weight) and effort from commercial fisheries; sub-samples shall indicate size, age, and sex composition of catch.
- Participate in Ocean landings stock composition study.
- Cooperative effort between New Jersey, New York, Pennsylvania, and Delaware.
15
Table 1.1.3-2 (cont.) Summary of mandatory fishery-dependent monitoring programs for American shad as required by Addendum I to Amendment 1 of the Shad and River Herring FMP (continued).
State System Sampling Program PA Delaware
- Monitor recreational landings, catch, and effort every 5 years.
River* (Cooperative effort between New Jersey, New York, Pennsylvania, and Delaware)
MD In-river
- Monitor recreational landing, catch, and effort every 5 years.
- Mandatory reporting of catch (numbers and weight) and effort from commercial fisheries; sub-samples shall indicate size, age, and sex composition of catch.
- Participate in Ocean landings stock composition study.
DC Potomac River
- Monitor recreational landings, catch, and effort every 5 years.
VA In-river
- Monitor recreational landings, catch, and effort every 5 years Atlantic Ocean
- Mandatory reporting of catch (numbers and weight) and effort from commercial fisheries; sub-samples shall indicate size, age, and sex composition of catch.
- Participate in Ocean landings stock composition study.
NC Albemarle
- Mandatory reporting of catch (numbers and weight) and effort from commercial Sound and its fisheries; sub-samples shall indicate size, age, and sex composition of catch.
tributaries, Tar-
- Monitor recreational landings, catch, and effort every 5 years Pamlico, Neuse, and Cape Fear Rivers Atlantic Ocean
- Mandatory reporting of catch (numbers and weight) and effort from commercial fisheries; sub-samples shall indicate size, age, and sex composition of catch.
- Participate in Ocean landings stock composition study.
SC Edisto River,
- Mandatory reporting of catch (numbers and weight) and effort from commercial Santee River, fisheries; sub-samples shall indicate size, age, and sex composition of catch.
Winyah Bay and
- Monitor recreational landings, catch, and effort every 5 years.
its tributaries (Waccamaw and
- State may elect to sample these systems on a rotational basis (i.e., one system Pee Dee Rivers) evaluated per year)
- Mandatory reporting of catch (numbers and weight and effort form commercial fisheries; sub-samples shall indicate size, age, and sex composition of catch.
- Participate in Ocean landings stock composition study.
GA Ogeechee
- Mandatory reporting of catch (numbers and weight and effort form commercial fisheries; sub-samples shall indicate size, age, and sex composition of catch.
- Monitor recreational landing, catch, and effort every 5 years.
FL St. Johns River
- Mandatory reporting of catch (numbers and weight) and effort from commercial fisheries; sub-samples shall indicate size, age, and sex composition of catch.
- Monitor recreational landings, catch and effort every 5 years.
- Cooperative effort between New Jersey, New York, Pennsylvania, and Delaware.
16
1.1.4 Assessment History Historical Reviews of the American Shad Fishery and Stock The first extensive report of the U.S. American shad fisheries was done in 1896 by Stevenson (1899).
Beginning in 1950, the U.S. Fish and Wildlife service conducted a series of investigations to examine causes of decline, determine factors favoring recovery, and provide management information, all of which was reported in Walburg and Nichols (1967). Major shad rivers in each Atlantic coast state were investigated to collect comprehensive catch and effort data by gear type. For historical perspective, these data were compared to that reported by Stevenson (1899).
ASMFC Atlantic Coast Stock Assessments Gibson et al. 1988 The ASMFC Shad and River Herring Technical Committee conducted its first coastwide assessment in 1988 (Gibson et al. 1988) on 12 Atlantic coast rivers. The Shepherd stock-recruitment model was used to estimate maximum sustainable yield (MSY) and maximum sustainable fishing rate (Fmsy). The status of American shad stocks was evaluated by comparing fishing mortality rates (F) in assessed rivers to Fmsy.
ASMFC 1998 The second coast-wide stock assessment conducted by the ASMFC was completed and reviewed in 1998 by the American Shad and River Herring Technical Committee (ASMFC 1998). Generally, assessments were conducted on a river-specific basis, but some grouping of river systems occurred (e.g., Maine rivers were examined collectively, Upper Bay Maryland, Albemarle Sound, and Waccamaw and Pee Dee rivers). The Thompson-Bell yield-per-recruit (YPR) model was used to derive the overfishing definition (F30) for some shad stocks where possible. The assessment examined catch and harvest data, exploitation rates, fish-lift counts, current and historic coastal (Fc) and in-river (Fr) fishing mortality rates, and other indicators of stock status for American shad from selected stocks or river systems located from Maine to the Altamaha River, Georgia, with special attention on recent (1992 to 1996) stock dynamics. The basis for choosing F30 as an overfishing definition was not provided.
Trends in total mortality (Z) were examined for the Pawcatuck River, Rhode Island, upper Chesapeake Bay, Maryland, and tributaries of Albemarle Sound, North Carolina, as well as trends in commercial landings for Maine rivers, North Carolina rivers (Albemarle Sound, Neuse, Pamlico, and Cape Fear rivers), and South Carolina rivers (Waccamaw-Pee Dee, Savannah, Edisto, and Santee rivers). Trends in relative adult stock abundance were examined in the Merrimack River (Massachusetts and New Hampshire) based on fishway counts and in Virginia rivers (James, York, and Rappahannock rivers) based on commercial catch-per-unit-effort.
The 1998 assessment concluded that there was evidence of recent (1992-1996) and persistent stock declines in the Hudson River, New York and York River, Virginia and that stock abundance increased in the Pawcatuck River and Connecticut River in the most recent years examined. The assessment concluded that the drop in commercial landings in the Edisto River was largely due to a reduction in fishing effort and did not reflect stock abundance. In addition, the assessment reported that there was no evidence of recent stock declines for the Merrimack River, Delaware River, upper Chesapeake Bay tributaries, Rappahannock River, James River, Santee River and the Altamaha River. Stock declines inferred from declining trends from river-specific commercial landings were evident for the Neuse River, Pamlico River, Cape Fear River, Waccamaw-Pee Dee River, Savannah River, Albemarle Sound tributaries, and rivers in the state of Maine.
17
Where estimation of recent F rates (1992-1996) was possible, all were below F30. The assessment also concluded that there was no evidence that the ocean-intercept fishery had an adverse impact on American shad abundance along the Atlantic coast and that there was no evidence of recent (1990-1996) recruitment failure for Maine rivers, Pawcatuck River, Connecticut River, Hudson River, Delaware River, Upper Chesapeake Bay tributaries, Altamaha River and Virginia rivers.
More detailed accounts of previous assessments are provided within each full assessment section. For discussion on how the current approach compares to the previous assessments, please see Section 1.1.1. A broad overview of American shad and other Alosa spp. status, biology, ecology, and population dynamics is available in Limburg and Waldman (2003). In that volume, Boreman and Friedland (2003) investigated reproductive potential of American shad in terms of eggs per recruit under different rates of fishing mortality for southern, northern, and Hudson River stocks and concluded that habitat restoration and stocking efforts would augment the rebuilding of stocks where fishing mortality had been reduced.
Limburg et al. (2003; also in Limburg and Walden (2003)) reviewed life history, habitat issues, exploitation history, and identifies information that would benefit successful management of American shad.
Literature Cited ASMFC (Atlantic States Marine Fisheries Commission). 1998. American shad stock assessment peer review report. Atlantic States Marine Fisheries Commission, Washington, D.C.
Boreman, J., and K.D. Friedland. 2003. Sensitivity of American shad to changes in fishing mortality.
Pages 267-273 in K.E. Limburg and J.R. Waldman editors. Biodiversity, status and conservation of the worlds shads. American Fisheries Society, Symposium 35. Bethesda, Maryland.
Gibson, M.R., V.A. Crecco, and D.L. Stang. 1988. Stock assessment of American shad from selected Atlantic coast rivers. Atlantic States Marine Fisheries Commission. Special Report No. 15.
Limburg, K.E., K.A. Hattala, and A. Kahnle. 2003. American shad in its native range. Pages 125-140 in K.E. Limburg and J.R. Waldman editors. Biodiversity, status and conservation of the worlds shads. American Fisheries Society, Symposium 35. Bethesda, Maryland.
18
1.1.5 Benchmark Mortality Estimation Using Biomass-per-Recruit Modeling Authors:
Kathryn A. Hattala and Andrew W. Kahnle Hudson River Fisheries Unit, New York State Department of Environmental Conservation 21 South Putt Corners Road, New Paltz, New York 12561 Kristen A. Delano, Robert J. Latour and John E. Olney School of Marine Science, Virginia Institute of Marine Science Route 1208 Greate Road, P.O. Box 1346, Gloucester Point, Virginia 23062 Introduction In this section, we describe modeling used to develop a benchmark total mortality. We also discuss sensitivity analyses on various model inputs. Since there are many competing theories on relative causes of mortality in Atlantic coastal American shad stocks (see Section 1.5), we decided to develop a benchmark rate for total instantaneous mortality (Z) rather than for an instantaneous rate of fishing (F).
This does not eliminate the issue of partitioning mortality into F and the instantaneous rate of natural mortality (M) in modeling, but it does avoid an emphasis on F when comparing the results to observed estimates of Z. The debate remains over causes of decline that is occurring in the majority of East Coast stocks. Limited evaluations of some hypotheses are given in Section 1.5.
We decided to use a Z30 as a biological reference point for the American shad. Z30 is defined as the total mortality rate that will preserve 30 percent of the unexploited spawning stock biomass per recruit. This is similar to the benchmark used by ASMFC (1998) in the last coast-wide assessment. We used a regional approach to estimate reference points because most individual stocks did not have all of the needed stock specific data. As a comparison we also calculated Z30 based on egg-per-recruit (EPR) using fecundity data from the Hudson and York rivers.
Methods Our analyses augment a basic biomass-per-recruit (BPR) model for females with estimates of egg production for information on EPR. Our model starts with recruits at age one. These recruits are decremented annually by natural mortality until they reach harvestable ages. They are then decremented by natural and fishing mortality through the maximum age observed in the stock or region (Table 1.1.5-1). The numbers of survivors-at-age are multiplied by weight-at-age for estimates of biomass-at-age. As survivors mature, the fraction of females of each age that is mature is multiplied by fecundity at that age.
Biomass and egg production by age are summed for all ages. In the final step, total biomass and total egg production are each divided by the number of initial recruits for an estimate of BPR and EPR. The model was run for a range of fishing rates (F) from zero to 0.7. Formulae used in model calculations are summarized in Appendix I of Section 1. River specific model inputs by age are listed in Table 1.1.5-2.
We used the assumption of a Type 1 fishery in our analyses. Estimates of Z30 were made for New England Rivers in aggregate, the Hudson River, the York River for the Chesapeake Bay region, and North Carolina rivers in aggregate.
We used results from the BPR modeling for our benchmark mortality rates because BPR modeling was possible for all regions. We added EPR analyses to provide perspective and confidence in results.
19
We included sensitivity analyses for runs of the biomass and egg-per-recruit model to illustrate effects of varying M, of a Type I versus a Type II fishery, and river-specific data inputs on model outputs.
Calculation of F30 for the York River Harvest of American shad in the York River, Virginia, is currently banned except for subsistence fishing by native Americans (the Mattaponi and Pamunkey tribal governments) allowed under a 17th century treaty agreementthe 1677 Treaty at Middle Plantation (www.baylink.org). Fishing is known to occur annually but effort and landings are not reported to the Virginia Marine Resources Commission. As a result, the impact of these removals on the York River stock cannot be assessed easily.
The biomass model run for the York River was configured for the Native American drift gill-net fishery to provide a baseline overfishing definition (F30) that might help evaluate the impact of such fisheries if an estimate of fishing rate were known.
Selection of Model Inputs One important point for the reader to note is that the stock-specific data inputs used in all the modeling have been obtained from fished stocks. These data may not reflect those characteristics from an unfished stock.
Natural Mortality The debate on the appropriate level of natural mortality (M) for American shad remains unresolved.
Previous assessments have used a variety of values for M. ASFMC (1998) used different values for M, for different age groups and stocks. These values were M=0.3 for ages one through three (all stocks),
M=2.5 for ages four through eight for southern semelparous stocks; and M=1.5 for ages four through 10 for northern iteroparous stocks, except the Hudson (Table 1.1.5-1). The high values of M for the older ages were calculated from age structure data of the Connecticut River shad stock and the mortality that stock was experiencing. For the Hudson, values used were M= 0.30 for ages one through three and 0.60 for ages six through 10 (ASMFC 1998).
Gibson et al. (1988) used age invariant (constant), river-specific natural mortality, calculated using a variety of age-based methods (Hoenig 1983; Pauly 1980; Leggett 1976). No specific values were provided in the report. Deriso et al. (2000) used an age invariant rate of 0.3 for Hudson River American shad. Carmichael (1999) used an M=0.5 for the assessment of Chowan River, North Carolina blueback herring. A constant M of 0.2 was used for Atlantic herring, with an age 10+ group (ASMFC 1999). Given that shad live longer than blueback herring, and as long, if not longer, than Atlantic herring, the value of M=0.3 appears reasonable.
In this assessment, we used age invariant values obtained by Hoenigs method (1983) based on the maximum observed age within region.
River-Specific Inputs: Maturity, Fecundity, and Partial Recruitment Vector The remaining inputs of maturity (percent mature-at-age)fecundity and partial recruitment (PR) vectorwere based on available data obtained from each river system or region (Table 1.1.5-2; Figure 1.1.5-1). We attempted to use river-specific data wherever possible. Data from a neighboring system were used if a specific input was not available (e.g., we used the Hudson PR vector was used for the New England region).
20
Comparison of inputs indicated several differences among systems. American shad of the Connecticut River (ASMFC 1998) matured at a faster rate than shad in the other regions or rivers (Table 1.1.5-2; Figure 1.1.5-1). Fecundity data were available for the Hudson and the York rivers. Full recruitment to fishing occurred at a younger age in shad of the Connecticut (ASMFC 1998) and York rivers than in shad of the Hudson River and Albemarle Sound (Figure 1.1.5-1). It is possible that fisheries in the Connecticut and the York rivers used smaller mesh gill nets than in the other rivers. PR vector data were based on in-river fisheries only as no data were collected on any of the directed mixed stock ocean or estuary fisheries before the closure.
The previous assessment (ASFMC 1998) used a von Bertalanffy growth curve based on data from the Connecticut River to estimate weight-at-age for the different stocks (northern, southern, and Hudson) by varying W. The other parameters (K and t0) remained constant. In each case, weight-at-age was over-estimated as compared to observed river-specific data (Table 1.1.5-2; Figure 1.1.5-1).
York River Inputs The model required an age-based partial recruitment vector for the drift gill nets used in the Native American fishery. Concurrent (1998-2000) drift gill net (DGN) data from the Native American fishery and pound net data from commercial nets (considered non-selective gears) were available and selectivity was estimated using an age-length key as reported in Appendix II.
If the drift gill-net fishery is the sole source of removals from the York River stock, then the dome-shaped recruitment vector developed solely from the fishery (Appendix II) might be considered appropriate as model input; however, known directed (ocean fishery) and bycatch removals (see Sections 1.1.7 and 12),
along with other unknown sources of fishing mortality make this scenario improbable. Therefore, the partial recruitment vector used in our modeling assumed full recruitment to the gear after age-5.
Other parameter inputs (maturity-at-age, fecundity-at-age) for the York River yield model were obtained from Maki et al. (2001) and Hyle (2005), respectively. Weight-at-age for York River fish was estimated from the best fit Linear von Bertalanffy model (see Section 12, Table 12-14) using the following relationship:
Weight = a[(498.2 + 6.2( Age))(1 e 0.46( Age0.07) )]b In the nonlinear regression solution, model parameter estimates were: a = 0.000056, b = 2.752. Model inputs are listed on Tables 1.1.5-1 and -2.
Sensitivity Analyses of Model Inputs We conducted various sensitivity analyses of model inputs using the Hudson and York Rivers BPR model to calculate an F30 using a variety of data inputs. Here we focused on an F30 rather than a Z30 to keep output curves on the same x-axis while varying M and other factors.
Natural Mortality A major issue during the last assessment was sensitivity of the BPR model to changes in various inputs.
The most contested debate centered on the selection of M. Since M remained in debate, Hattala and Kahnle (1998) presented an analysis of the sensitivity of F30 from BPR modeling to M inputs ranging from 0.2 to 1.4. Rather than repeat that analysis in this paper, we selected a subset of values to illustrate 21
effects. For this exercise, we ran the Hudson River BPR and EPR models using constant, age-invariant values of M= 0.3, 0.5, and 0.7 (Table 1.1.5-1). All other inputs (weight, maturity, and fecundity-at age, and partial recruitment vector) remained constant. To provide further contrast to the age invariant values, we also used age dependent values, which decreased with age, based on a method by Boudreau and Dickie (1989) and Dickie et al. (1987; Table 1.1.5-1). These authors related M to a specific rate of production (biomass) for each size group (age) in a population. The curve of M on age is an indication of the natural mortality pattern of a stock. Given that immature American shad and other herring are forage for many fish predators, it is likely that M is not age invariant and that it is higher at young ages. Since M remains uncertain, we conducted this sensitivity analysis to determine effects of various values of M on estimates of an overfishing definition.
Maturity, Fecundity, and Recruitment Vectors We included additional runs of the BPR and EPR model, using data inputs for two rivers to illustrate effects of using river-specific data inputs on model outputs. The data of concern were age-specific fecundity and weight, a maturity schedule, and a partial recruitment vector (vulnerability to a fishery). We did this to illustrate the inherent uncertainty associated with altering one systems data to mimic other stocks (ASMFC 1998) rather than using available, or adjacent river systems, stock-specific data. Maturity schedules were similar for each of the stocks; however, there were differences in input data for fecundity, partial recruitment vectors, and weight-at-age (Figure 1.1.5-2). Leggett and Carscadden (1978) suggested that fecundity-at-age varied with latitude along the East Coast. Values for the Hudson are less than for the York, a system to the south, following the pattern suggested.
Partial recruitment vectors are a function of the how each fishery operates and gear used to capture each size and age class of fish. Partial recruitment vectors, calculated from age structure data were obtained from in-river fisheries in each system. Even though mesh size used in the two river fisheries was similar (5.25 in versus 5.5 in), data from the York indicated a higher vulnerability (to the fishery) of age-4 through 7 fish (Figure 1.1.5-2). Shad caught in the Hudson River fishery were on average 45-48 mm larger than those caught in the York (see Sections 7 and 12). Weight-at-age was also smaller for the York than for the Hudson stock (Figure 1.1.5-2).
In the initial benchmark model runs, M varied for the two stocks as observed maximum age observed in the stock were different. For the sensitivity analyses model runs, M was held constant at 0.5 to examine differences the stock specific inputs (Table 1.1.5-2) would generate.
Fisheries Type Biomass-per-recruit modeling requires an assumption of how natural and fishing mortality overlap during the biological year. A Type 1 fishery assumes that natural morality occurs after fishing ends. A Type 2 fishery assumes that fishing and natural mortality operate concurrently (Ricker 1975). We calculated a Z30 for the Hudson using both Type 1 and Type 2 fisheries to examine if differences occur.
Results Benchmark Definitions Z30 - All Regions We selected our regional benchmark mortality values from the BPR analyses using age invariant M. From these analyses, Z30 ranged from 0.54 to 0.76 depending on region. The Z30 EPR calculated for comparison 22
of the Hudson and York rivers were almost identical to those obtained from BPR analyses (Table 1.1.5-4). Curves of BPR at increasing levels of Z showed highest values of BPR at each Z for the Hudson River, followed by the Albemarle, York River, and New England (Figure 1.1.5-2). Curves of EPR values at increasing levels of Z showed higher values for the York River, than for the Hudson River (Figure 1.1.5-2). Note that the curves start at different values of Z because natural mortality varied among systems.
Benchmark estimates from ASMFC (1998) were much higher; than our results if our Z30 was translated to F30 by subtracting M (Table 1.1.5-4).
York River F30 The benchmark fishing rate (F30) for the drift gill-net fishery in the York River is estimated to be 0.27 (Table 1.1.5-3).
A benchmark developed for the York River (17.44, the geometric mean of the catch index values observed in 1953-1957) is based on shad abundance in the 1950s and is higher than the geometric mean catch index of current monitoring data (see Section 12). Additional control rules should be the target fishing rate (F30 = 0.27) for the Native American fishery and target total mortality rate (Z30 = 0.62). To apply these rules, an estimate of F by the Native American fishery and a better understanding of natural mortality are required.
Sensitivity Results Natural Mortality The response of EPR, BPR, and F30 varied with the value of M used in the analyses. Curves of EPR and BPR at increasing levels of Z showed that highest values EPR and BPR were produced by lowest values of age invariant M (Figure 1.1.5-3).
For a constant, age-invariant M, the F30 increased with increasing M. The F30 values ranged from F30 =
0.22 (EPR) when M=0.3 to F30 = 0.36 for M=0.7 (Table 1.1.5-4). Estimates of F30, based on BPR, were similar and ranged from 0.24 to 0.38. For age-specific values of M, the resulting values of F30 were lower than age invariant M for either EPR or BPR. For age-specific M, F30 was equal to 0.19 for EPR, and 0.21 for BPR (Table 1.1.5-4).
River-Specific Inputs M was held constant at M=0.5 for the Hudson and the York River stocks. As expected, estimates for EPR were highest for the more southern York River stock, dropping slightly for the Hudson to the north (Table 1.1.5-4, Figure 1.1.5-4). The same pattern of south to north did not occur for the BPR curves (Figure 1.1.5-4). The Hudson River BPR estimates were higher at F estimates near 0; however, the York River=s curve was similar to the Hudson Rivers F values greater than of 0.4.
It should be noted that fecundity-at-age and biomass-at-age are both affected by fishing because fishing tends to remove the largest fish. These effects should be kept in mind when considering results of any EPR or BPR analyses. Fecundity-at-age data for the Hudson River stock were produced when stocks were recovering from the overfishing events following WWII. The weight-at-age data used for the Hudson River and perhaps the York River stocks showed recent declines. Ideal data inputs would be observations taken during a time period when fishing impacts were fairly low. Unfortunately, no data exists for any east coast shad stock under these conditions.
23
Type 1 and Type 2 Fisheries We used the assumption of a Type 1 fishery in our analyses, but added a Type 2 case to identify effects of these fishery assumptions on our analyses output. Given the stress of spawning experienced by American shad (movement from fresh to saltwater, little or no food consumption while spawning) and the lengthy exposure to fisheries (late February - ocean fisheries, to May - terminal in-river fishery), we feel a Type 2 fishery where both natural and fishing mortality are occurring simultaneously can be more realistic.
However, analyses results using Type I and Type II mortality assumptions were essentially the same (Table 1.1.5-4).
Discussion Models and analyses presented in this section were developed to provide Atlantic coastal states with an approach to assessing the status of American shad stocks. We used the EPR and BPR modeling to identify a benchmark definition of Z because we felt that the use of a simple approach would set a reasonable standard for evaluating current estimates of Z available for Atlantic coastal shad stocks comparable to reference points arrived at by other methods. However, even this simplified approach was limited by its reliance on data inputs that were obtained from fished stocks.
Another aspect of this section was to compare this assessment to the 1998 ASMFC American shad assessment since some issues remain unresolved. These assessment models are not overly complicated, but they do require attention to data inputs. In this document we attempt to clarify the data choices made and outline the supporting reasoning. We also hope that this paper will assist others in better understanding methodology and implications for other East Coast systems.
Natural Mortality The choice of a value for M is very important to all modeling work on American shad. Natural mortality rates of fishes are inversely related to longevity. American shad populations with no repeat spawning, such as in southern stocks, clearly have a higher natural mortality rate than those that exhibit iteroparity.
Fish in most southern stocks rarely reach a maximum age of eight, few fish reach age seven. Most northern stocks, north of North Carolina, have some degree of repeat spawning; maximum age falls within the range of eight to 11 or 12, with the exception of the Hudson River. Model runs, with a selected M, that generate the virgin stock size benchmark, should, at minimum, use maximum ages observed in the wild populations. If they do not, then virgin stock size can be underestimated.
Maximum age of Hudson shad most often equals 12, a few fish have been observed at age 13. These older ages in the Hudson stock suggest M should be fairly low to reflect the stock=s longevity. It is not clear how old shad can get since current data (collected within the last 27 years) reflect conditions present in shad populations at low stock size and the effects of fishing.
For comparative value, many other fish stocks have similar natural mortality rates. Age invariant M has been the choice of most assessments. For top end predators, natural mortality is fairly low (striped bass, M=0.15; ASMFC 2005). Shad, however, fall into the prey species category at younger ages, until they grow large enough to avoid predation. For a similar prey-type, though non-anadromous, clupeid species, Atlantic herring, the value selected for M is 0.2 (SAW 1996). For another anadromous species, Atlantic salmon, the value of M=0.12 is used (Freidland et al. 1994).
24
River-Specific Input Data Effects of using different river-specific biological input data were confounded in our analysis because we did not hold any one input constant while varying others. It appears that the modeling was more sensitive to changes in M than other biological inputs.
Literature Cited ASMFC. (Atlantic States Marine Fisheries Commission). 1998. American shad stock assessment peer review report. Washington, D.C.
ASMFC. (Atlantic States Marine Fisheries Commission). 2005. 2005 Stock Assessment Report for Atlantic Striped Bass: Catch-at-Age Based VPA & Tag Release/Recovery Based Survival Estimation A report prepared by the Striped Bass Technical Committee for the Atlantic Striped Bass Management Board. Washington, D.C.
Boudreau, P.R., and L.M. Dickie. 1989. Biological modeling of fisheries production based on physiology and ecological scalings of body size. Can. J. Fish. Aquat. Sci. 46:614-623.
Carmichael, J. 1999. Status of Blueback herring in the Chowan River, North Carolina, 1972-1999. North Carolina Department of Environment and Natural Resources, Division of Marine Fisheries, Morehead City.
Crecco, V., and M. Gibson. 1988. Methods of estimating fishing mortality rates on American shad stocks.
In 1988 Supplement to the American shad and river herring Management Plan, Fisheries Management Report No. 12 of the Atlantic States Marine Fisheries Commission, Washington, D.C.
Dadswell, M., G. Melvin, P. J. Williams, and D.E. Themelis. 1987. Influences of origin, life history, and chance on the Atlantic coastal migration of American shad. In Common Strategies of Anadromous and Catadromous fishes, American Fisheries Society, Symposium 1. Bethesda, Maryland.
Dickie, L.M. , S.R. Kerr, and P. Scwinghamer. 1987. An ecological approach to fisheries assessment.
Can. J. Fish. Aquat. Sci. 44(Suppl. 2):68-74.
Deriso, R. , K. Hattala, and A. Kahnle. 1999. Hudson River shad assessment and equilibrium calculations.
Prepared through ESSA Technologies, Ltd. for the NY State Dept. of Environmental Conservation. Toronto, Canada.
Freidland, K.D., R.E., Haas and T.F Sheehan. 1996. Comparative post smolt growth, maturation and survival in two stocks of Atlantic salmon. Fishery Bulletin 94:654-663.
Gibson, M., V. Crecco, and D. Stang. 1988. Stock assessment of American shad from selected Atlantic coast rivers. Special report #15 of the Atlantic States Marine Fisheries Commission, Washington, D.C.
Hoenig, J. 1983. Empirical use of longevity data to estimate mortality rates. Fishery Bulletin 82(1):898-903.
25
Hyle, A.R. 2005. Reproductive biology of American shad, Alosa sapidissima, in the Mattaponi River.
M.S. Thesis presented to the School of Marine Science, College of William and Mary.
Leggett, W.C. 1976. The American shad, Alosa sapidissima, with special reference to its migrations and population dynamics in the Connecticut River. In Merriman and Thorpe, editors, The Connecticut River Ecological Study: the impact of a nuclear power plant. American Fisheries Soc. Monograph 1. Bethesda, Maryland.
Leggett, W.C., and J.E. Carscadden. 1978. Latitudinal variation in reproductive characteristics of American shad (Alosa sapidissima): evidence for population specific life history strategies in fish. J. Fish. Research. Bd. Can. 35:1469-1478.
Maki, K.L., J.M. Hoenig, and J.E. Olney. 2001. Estimating proportion mature at age when immature fish are unavailable for study, with application to American shad (Alosa sapidissima) in the York River, Virginia. J. North American Fisheries Management 21:703-716.
Pauley, D. 1980. On the inter-relationship between natural mortality, growth parameters and, and mean environmental temperature in 175 fish stocks. J. Cons. Explor. Mer. 39:175-192.
Ricker, W.E. 1975. Computation and interpretation of biological statistics of fish populations. Fisheries Research Board of Canada Bulletin 191.
SAW. 1996. 21st Northeast Regional stock assessment workshop, Northeast Fisheries Science Center, Reference document 96-05d.
26
Table 1.1.5-1b American shad natural mortality rates used in Thompson Bell Biomass-per-recruit model.
Maximum Observed Age Invariant Region Age, Age, or Age Natural Mortality*
Group Current Assessment New England 11 0.38 Hudson River, NY 14 0.30 York River, VA 12 0.35 Albemarle Sound, NC 10 0.42 Sensitivity Analysis 14 0.30 8 0.50 6 0.70 Age Variant M**
1 0.51 2 0.37 3 0.29 4 0.25 5 0.23 6 0.21 7 0.20 8 0.20 9 0.19 10 0.19 11 0.19 12 0.19 13 0.19 14 0.19 ASMFC 1998 (Previous Assessment)
All rivers 1-3 0.3 Hudson 4-10 0.6 Northern rivers (NC-ME) 4-10 1.5 Southern rivers (SC-FL) 4-8 2.5
- Hoenig 1983
- Boudreau and Dickie (1989) and Dickie et al. (1987) 27
Table 1.1.5-2b River-specific life history and fishery data used as inputs to the Thompson Bell biomass-per-recruit model.
Maturity Weight-at-age New Hudson Albemarle Age York R. VA ASMFC 1998 Von B New Hudson York Albemarle ASMFC England R. NY S. NC England R. NY R. VA S. NC 1998 Linear von Gompertz Gompertz Gompertz North South Hudson Bertalanffy 1 0% 0% 0% 0% 0% 22 142 82 65 43 30 55 2 0% 0% 0% 0% 0% 164 391 363 299 353 247 459 3 0% 0% 2% 1% 20% 492 774 709 722 903 632 1174 4 2% 15% 33% 18% 60% 901 1228 1026 1203 1541 1079 2004 5 25% 63% 68% 70% 100% 1258 1676 1287 1614 2157 1510 2805 6 61% 91% 80% 98% 100% 1510 2068 1492 1912 2695 1886 3503 7 86% 98% 100% 100% 100% 1670 2383 1653 2109 3137 2196 4078 8 96% 100% 100% 100% 100% 1766 2623 1782 2231 3486 2440 4532 9 100% 100% 100% 100% 100% 1820 2798 1889 2305 3755 4882 10 100% 100% 100% 100% 100% 1851 2923 1981 2349 3959 5147 11 100% 100% 100% 1868 3010 2064 12 100% 100% 3071 2141 13 100% 3112 14 100% 3140 Age Partial Recruitment Vector Fecundity 1 0.00 0.00 0.00 0.00 - -
2 0.00 0.00 0.00 0.00 - -
3 0.14 0.03 0.00 0.00 95491 249268 4 0.17 0.64 0.04 0.45 157637 360980 5 0.40 1.00 0.32 0.90 219783 452682 6 0.70 1.00 0.65 1.00 281929 524888 7 0.87 1.00 0.88 1.00 344075 581546 8 0.95 1.00 0.98 1.00 406221 626905 9 1.00 1.00 1.00 1.00 468367 664476 10 1.00 1.00 1.00 1.00 530513 696871 11 1.00 1.00 592659 725932 12 1.00 1.00 654805 752922 13 1.00 716951 -
14 1.00 779097 -
Table 1.1.5-3b Results of benchmark determination (Z30) from the Thompson-Bell biomass per recruit model, modified for egg-per-recruit, for regional American shad stocks.
Max Age or Z30 Region M F30 Age Group EPR BPR New England 11 0.38 - 0.64 Hudson River, NY 14 0.30 0.52 0.54 York River, VA 12 0.35 0.62 0.62 0.27 Albemarle Sound, NC 10 0.42 - 0.76 ASMFC 1998 All rivers 1-3 0.3 Hudson 4-10 0.6 0.99 0.39 Northern rivers (NC-ME) 4-10 1.5 1.93 0.43 Southern rivers (SC-FL) 4-8 2.5 2.98 0.48 29
Table 1.1.5-4b Results of sensitivity of the Thompson-Bell biomass-per-recruit model to variation in M and river-specific inputs (maturity, fecundity and partial recruitment vector).
F30 Z30 Max Age or Type 1 Type 2 Type 1 Type 2 River Type of Analyses System M EPR BPR EPR BPR EPR BPR EPR BPR Natural mortality 14 0.30 0.22 0.24 0.23 0.25 0.52 0.54 0.53 0.55 river input = constant 8 0.50 0.29 0.31 0.30 0.32 0.79 0.81 0.80 0.82 6 0.70 0.36 0.38 0.37 0.39 1.06 1.08 1.07 1.09 Natural mortality age variant see Table 1.1.5-3 0.19 0.21 River specific inputs Hudson 0.50 0.29 0.31 0.79 0.81 M = constant York 0.50 0.32 0.32 0.82 0.82
Figure 1.1.5-1b River-specific life history and fishery data, along with data from ASMFC1998, used as inputs to the Thompson Bell Biomass -per-recruit model.
Maturity 100%
New England 80%
Hudson R. NY 60% York R. VA Albemarle S. NC 40%
ASMFC 1998 20%
0%
1 2 3 4 5 6 7 8 9 10 11 12 13 14 Age Fecundity 1000 800 Hudson R. NY Thousands York R. VA 600 400 200 0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 Age Partial recruitment vector 100%
90%
80%
70%
60%
50% Hudson R. NY 40% York R. VA 30% Albemarle S. NC 20%
10% ASMFC1998 0%
1 2 3 4 5 6 7 8 9 10 11 12 13 14 Age 31
Figure 1.1.5-1b (cont.) River-specific life history and fishery data, along with data from ASMFC1998, used as inputs to the Thompson Bell Biomass -per-recruit model.
A. Weight at age New England Hudson R. NY 5000 York R. VA Albemarle S . NC 4000 Grams 3000 2000 1000 0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 Age New England B. Weight at age York R. VA 5000 Albemarle S. NC ASMFC98-North 4000 Grams 3000 2000 1000 0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 Age C. Weight at age Hudson R. NY 5000 ASMFC98-Hudson 4000 Grams 3000 2000 1000 0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 Age 32
Figure1.1.5-2b Biomass-per-recruit and egg-per-recruit for regional American shad stocks.
2.50 New England 2.00 Hudson R.-NY York R.-VA Biomass per recrui Albemarle-NC 1.50 1.00 0.50 0.00 0.30 0.50 0.70 0.90 1.10 1.30 Z
0.45 0.40 0.35 Hudson R.-NY York R.-VA 0.30 Egg per recrui 0.25 0.20 0.15 0.10 0.05 0.00 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00 1.10 Z
33
Figure 1.1.5-3b Sensitivity of changes in BPR model to variation in M, with river-specific input held constant.
3.0000 M=0.3 2.5000 M=0.5 M=0.7 2.0000 M=var BPR 1.5000 1.0000 0.5000 0.0000 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 F
0.5000 0.4500 M=0.3 0.4000 M=0.5 M=0.7 0.3500 M=var 0.3000 EPR 0.2500 0.2000 0.1500 0.1000 0.0500 0.0000 0.00 0.20 0.40 0.60 0.80 F
34
Figure 1.1.5-4b Sensitivity of changes in BPR model to changes in river-specific inputs with M held constant at 0.5 0.20 0.18 0.16 0.14 Hudson York 0.12 EPR 0.10 0.08 0.06 0.04 0.02 0.00 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 F
0.70 0.60 0.50 Hudson York 0.40 BPR 0.30 0.20 0.10 0.00 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 F
35
Revisions to Benchmark Mortality Estimation Using Biomass-per-Recruit Modeling Based on Recommendations from the Peer Review Panel Authors:
Kathryn A. Hattala and Andrew W. Kahnle Hudson River Fisheries Unit, New York State Department of Environmental Conservation 21 S. Putt Corners Road, New Paltz, New York 12561 The following section addresses only a portion of the suggestions made by the American Shad Peer Review Panel. The Stock Assessment Subcommittee (SASC) Chair was asked to provide these revised estimates during the Panel's meeting. These changes were made during the meeting because they could be accomplished in a short period of time and would have an implication on the benchmarks used to determine the status of the American shad stocks. The Panel refers to these revised estimates in the American Shad Advisory Report, therefore the results are provided in this addendum. The Peer Review Panel made additional suggestions, but the SASC and Technical Committee will work cooperatively to explore those recommendations if directed by the Shad and River Herring Management Board.
The following materials respond to critiques by the Peer Review Panel on the formulation in our biomass per recruit modeling. The panel made suggestions regarding the application of mortality rates to immature and mature fish, the form of the recruitment or vulnerability vector, and the application of vulnerability to exploitation rate (µ) rather than fishing mortality (F).
Benchmark Model
- We applied a survival value (S) at age, which included F and natural mortality (M), to the Number at age of immature fish.
- We developed recruitment vectors from empirical age data from various gears and maturity schedules from repeat spawn data.
- We applied the recruitment vector and maturity schedule at age to µ at age.
- We used a Type I fishery.
Review Panel Suggestions
- Apply only M to number at age of immature fish.
- The maturity schedule may have confounded the recruitment vectors. To reduce these confounding effects, assume a recruitment of 1 for all ages.
- Apply the recruitment vector at age to F before converting to µ, and then apply the maturity schedule at age to µ.
- The fishing season fits between a Type I and a Type II fishery. Try both types to provide bounds on the resulting Z30^ estimates.
Results (Table 1.1.5-3rev)
- In all cases, the revised Z30 values were higher than the Z30 values calculated in our original assessment formulation. Relative change was greatest for New England and Albemarle Sound in North Carolina. All revised Z30 estimates were lower than the Z30 values (calculated from F30+M) calculated in the 1998 ASMFC assessment.
- In all cases the Type I fishery calculations produced a higher benchmark Z than the Type II fishery calculations; however, differences were small.
^
Z30 is the total mortality rate that will preserve 30% of the unexploited spawning biomass per recruit.
F30 is the fishing mortality rate that will preserve 30% of the unexploited spawning biomass per recruit.
36
Table 1.1.5-3rev Results of benchmark determination (Z30) from the Thompson-Bell biomass-per-recruit model, modified for egg-per-recruit, for regional American shad stocks.
Non-bold values are from original assessment and based on a Type I fishery.
Bold values were results of revised calculations for Type I (T1) and Type II (T2) fisheries.
Z30 Max Age or Region M F30 Age Group EPR BPR New England 11 0.38 - 0.64 -
revised T1 - 0.98 -
revised T2 - 0.91 -
Hudson River, NY 14 0.30 0.52 0.54 -
revised T1 0.68 0.73 -
revised T2 0.57 0.62 -
York River, VA 12 0.35 0.64 0.63 0.28 revised T1 0.85 0.85 0.50 revised T2 0.76 0.76 0.41 Albemarle Sound, NC 10 0.42 - 0.76 -
revised T1 - 1.01 -
revised T2 - 0.94 -
ASMFC 1998 All rivers 1-3 0.3 Hudson 4-10 0.6 0.99 0.39 Northern rivers (NC-ME) 4-10 1.5 1.93 0.43 Southern rivers (SC-FL) 4-8 2.5 2.98 0.48 37
1.1.6 Age Determination in American Shad Author:
John Olney School of Marine Science, Virginia Institute of Marine Science Route 1208 Greate Road, P.O. Box 1346, Gloucester Point, Virginia 23062 Age is estimated using scales and otoliths in American shad and other alosine fishes, including the river herrings (Cating 1953; Libby 1985; Marcy 1969; Rothschild 1963). Scale-based ageing of American shad is the predominant method in current coastwide monitoring programs from Maine to Florida. Otolith-based ageing is the method of choice of certain management agencies, especially those that are engaged in hatchery-based restoration activities (e.g., Pennsylvania Fish and Boat Commission, Virginia Department of Game and Inland Fisheries). Both scale- and otolith-based approaches have been used since the early 1900s (Leim 1924; Barney 1924). Scale-based methods have been validated for the Connecticut River (Judy 1961, but see McBride et al. 2005). Although otoliths now form the basis for most age and growth studies of fishes (Campana 1999), the validity of utilizing otoliths to age American shad, with the exception of larval shad (Savoy and Crecco 1987), has not been established.
Generally, scales are removed from the upper left lateral region just below the dorsal fin. Scales are cleaned and either pressed on acetate sheets or mounted on glass slides for viewing (see individual stock assessments for details of these methods). Cating (1953) provided criteria to enumerate presumptive annulus in American shad scales. The criteria are based on the assumption that annuli are consistently located in relation to the position of transverse grooves on the scale surface. Specifically, the first annulus is located within the first 4-7 transverse grooves; the second annuli is located between the 8th and 11th; and the third annulus is located between the 12th and 16th transverse grooves. Fish older than three years are aged using a combination of these criteria and the enumeration of spawning marks (persistent scars caused by erosion of the scale margin during spawning).
Both sagittal otoliths are removed from the auditory capsules and cleaned. In most monitoring programs that use otolith-based methods, one otolith is mounted, ground or sectioned and examined for tetracycline marks applied to larvae in the hatchery. Biologists identify the presence and pattern of the tetracycline tag using an epifluorescent microscope and can determine age in some cases using cohort-specific hatchery marks. The second otolith of the pair is examined as a whole structure (that is, not sectioned or ground) under a dissecting microscope (see individual stock assessments for details). Presumptive annuli are enumerated as one opaque zone and its successive hyaline zone.
Scale-based and otolith-based approaches to age American shad may not be interchangeable methods and there are few published studies that address this question. Olney and Hoenig (2000) compared ages determined by scales and otoliths in samples collected in 1998 and 1999 in Virginia samples. Scale and otolith methods were in agreement 41% of the time (125 of 305 comparisons) in 1998 samples and 45%
of the time (82 of 182 comparisons) in 1999 samples. Differences between methods were significant in each year. Aschenbach et al. (1996) reported low agreement between paired comparisons of ages determined by scales and sectioned otoliths. In 57 percent of their comparisons, scales yielded older ages than otoliths. In these studies, otolith and scale ages differed predominantly by one year.
The apparent capability to infer spawning history of an individual fish by enumerating spawning marks is considered a unique advantage of the scale-based method in comparison to otolith-based approaches. In their classic study of population-level variation in reproductive characteristics, Leggett and Carscadden (1978) used spawning marks on scales to describe variation in the degree of iteroparity in American shad across its native range. Spawning marks also have enabled the estimation of maturity schedules for American shad (Maki et al. 2001, 2002) and are used in current coastwide monitoring programs to estimate frequency of repeat spawning and total mortality.
38
Largely as a result of an early demonstration of the validity of Catings (1953) techniques, scale ageing has become the standard for shad assessments. Judy (1961) validated scale-based ageing methods for ages 3-5 in the Connecticut River. In the study, Judy marked 100,000 juvenile fish in 1952 by clipping pelvic fins and then recaptured survivors when they matured between 1956 and 1958. Ages were 98 percent accurate among the 129 recaptured fish and the sample included both virgin and previously spawned individuals. Thus, both annuli and spawning marks on scales were considered valid indicators of age.
In August 2004, an ASMFC age-determination workshop using 52 known-age fish from the Delaware River system was held to provide another test of the validity of scale-based ageing techniques. McBride et al. (2005) reported the results of these trials. Thirteen biologists from the ASMFC Shad and River Herring Technical Committee participated in the trials. All individuals were actively engaged in age determination of American shad in their respective states but some were more experienced in scale-ageing techniques than others. Each biologist read the scale sample twice. Precision (agreement between age estimates from the same scales) ranged from 50-76.5 percent. Accuracy (agreement between scale age and known age) was highest for ages 3-6 (33.7-48.5%) and lowest for older fish (3.9-12.1%, ages 7 and 8). Bias was detected in the trials. Ages of younger fish were sometimes overestimated and ages of older fish were typically underestimated. McBride et al. (2005) concluded that the scale-based method was not applicable to stocks in certain rivers and recommended against using age-based techniques to assess stocks of American shad until further age-validation studies were completed.
Both of these age validation studies (Judy 1961; McBride et al. 2005) are subject to criticisms and constraints. Judy (1961) did not describe a protocol that McBride et al. (2005) considered a blind trial.
That is, it is unknown if ages were determined by scale readers who were unaware of the ages of the sample. This could have occurred since the readers in Judys trials might have known which year marked juveniles were captured as virgin adults. Since known-age American shad are difficult to obtain, McBride et al. (2005) used cultured fish that had been recaptured in the Delaware River. In some cases, scale samples came from fish that were recaptured after long migrations (~ 300 km upriver). Several biologists in the McBride et al. (2005) trials noted that some scales were highly eroded and difficult to read. A serious problem could have resulted if the erosion on these scales obscured annuli, especially along the scale margin. McBride et al. (2005) acknowledged that other monitoring programs sample migrating shad in shorter rivers or at locations near the river mouth, thereby potentially avoiding the erosion effects observed in their trials.
Given the foregoing considerations, the American shad stock assessment subcommittee chose to moderate its use of age data in the following coastwide assessment but not abandon these data entirely. We evaluated age determination methods and results on a case-by-case basis in each jurisdiction by closely examining ageing methods, age structure, and spawning history data for signs of bias or inaccuracy.
These signs included extreme truncation of ages when size data suggested older fish in samples, absence of repeat spawners when iteroparity was expected, and cases in which state biologists acknowledged that they were inexperienced and lacked confidence in their scale reading techniques. We examined mean age and mean size data from individual states for consistency (Appendix B). In these plots, we looked for evidence that the ratio of mean age to mean size did not vary appreciably over the available time series.
We chose not to use age data from the Delaware system in light of the McBride et al. (2005) results. We had greater confidence in age data in the Connecticut River, the site of Judys (1961) validation, and in systems where fish were collected prior to long migrations. We also had greater confidence in ages that were determined by more experienced biologists, a consideration supported by the results of McBride et al. (2005).
Fifty years ago, Mansuetti and Kolb (1953) noted that great difficulty has been experienced by biologists in ageing shad. Since that time, shad biologists have not made significant advances in this field.
However, new tools are now available in the form of natural tags to test our methods and develop new 39
approaches. For example, in an ongoing study, biologists are using oxygen isotopes in otoliths as a natural tag to track recruitment of specific cohorts of American shad. The returning adults are available for study as known-age fish beginning in 2006 and will be used to evaluate multiple ageing techniques.
Literature Cited Aschenbach, E.F., III, G.C. Garman, and C.M. Conway. 1996. Approaches for ageing adult American shad (Alosa sapidissima) from scales and otoliths: a graphical and statistical comparison. Pages 31-39 in D. Mackinlay and J. Nelson, editors. High Performance Fish II Symposium Proceedings.
International Congress on the Biology of Fishes. San Francisco, California.
Barney, R.L. 1924. Confirmation of Borodins scale method of determining age in shad. Transactions of the American Fisheries Society 54:168-177.
Cating, J.P. 1953. Determining age of Atlantic shad from their scales. U.S. Fish Wildl. Serv. Fish. Bull.
54:187-199.
Campana, S.E. 1999. Chemistry and composition of fish otoliths: pathways, mechanisms and applications. Marine Ecology Progress Series 188:263-297.
Judy, M.H. 1961. Validity of age determination from scales of marked American shad. Fishery Bulletin 61:161-170.
Leggett, W.C., and J.E. Carscadden. 1978. Latitudinal variation in reproductive characteristics of American shad (Alosa sapidissima): evidence for population specific life history strategies in fish.
Journal of the Fisheries Research Board of Canada 35:1469-1478.
Libby, D. 1985. A comparison of scale and otolith ageing methods for the alewife, Alosa pseudoharengus. Fishery Bulletin 83:696-701.
Liem, A.H. 1924. The history of the shad, Alosa sapidissima (Wilson), with special reference to the factor limiting its abundance. Contributions to Canadian Biology, Nova Scotia 2:163-284.
Maki, K.L., J.M. Hoenig, and J.E. Olney. 2001. Estimating proportion mature at age when immature fish are unavailable for study, with application to American shad (Alosa sapidissima) in the York River, Virginia. J. North American Fisheries Management 21:703-716.
Maki, K.L., J.M. Hoenig, and J.E. Olney. 2002. Interpreting Maturation Data for American Shad in the Presence of Fishing Mortality - A Look at Historical Data from the York River, Virginia. J. North American Fisheries Management.
Mansuetti, R., and H. Kolb. 1953. A historical review of the shad fisheries of North America. Publication No. 97, Chesapeake Biological Laboratory, Solomons, Maryland.
Marcy, B.C. Jr. 1969. Age determinations from scales of Alosa pseudoharengus (Wilson) and Alosa aestivalis (Mitchill) in Connecticut waters. Transactions of the American Fisheries Society 98:622-630.
McBride, R.S., M.L. Hendricks and J.E. Olney. 2005. Testing the validity of Catings (1953) method for age determination of American shad using scales. Fisheries 30:10-18.
40
Olney, J.E., and J.M. Hoenig. 2000. Monitoring relative abundance of American shad in Virginias rivers. 1999 Annual report to the Virginia Marine Resources Commission, Contract No. F-116-R-2, 7 July 2000.
Rothschild, B.J. 1963. A critique of the scale method for determining the age of the alewife, Alosa pseudoharengus (Wilson). Transactions of the American Fisheries Society 92:409-413.
Savoy, T.F., and V.A. Crecco. 1987. Daily increments on the otoliths of larval American shad and their potential use in population dynamic studies. Pages 413-431 in R.C. Summerfelt and G.E. Hall, editors. Age and Growth of Fishes. Iowa State University Press, Ames.
41
1.1.7 Coastwide Commercial Landings of American Shad Authors:
Kathryn A. Hattala and Andrew W. Kahnle Hudson River Fisheries Unit, New York State Department of Environmental Conservation 21 S. Putt Corners Road, New Paltz, New York 12561 Introduction Ocean and mixed stock (OMS) fisheries have probably caught American shad for as long as the species has been fished. Beginning in the early 1980s, changes in interstate management focused on increasing restrictions to aid in the recovery of the striped bass stock. We suspect these restrictions forced fishers to look to other species to fill the gaps. American shad became a small, but important component of the ocean fishery during the late winter and early spring during the 1980s through 2005. During this 25 year period, changes in some American shad stocks began to occur which could not be explained by in-river harvest. Attention focused on the increased losses from the OMS fishery.
The OMS fisheries for American shad can be characterized as either a directed (targeted) fishery or a known bycatch fishery (documented by NMFS landings). Directed fisheries occurred in Rhode Island and states from New Jersey to South Carolina. These directed fisheries were closed in 2005. Known bycatch fisheries occur in New York and to a lesser degree, in the New England states (Maine to Massachusetts, Table 1.1.7-1). The few exceptions that have neither type of OMS fishery are the states of Georgia, Florida, Pennsylvania and, very minimally, Connecticut.
An additional mixed stock fishery continues to occur within the non-ocean waters of Delaware and New Jersey in lower Delaware Bay. This fishery has the potential to harvest fish from a wide range of shad stocks from South Carolina to Quebec.
A mixed stock bycatch fishery occurs in Chesapeake Bay. This bycatch is probably discarded and is not quantified. Harvest of American shad in the Chesapeake Bay is banned and there are no reporting requirements for discards by pound nets and other gears.
Lastly, one component of bycatch is more nebulous - and that is the fisheries where young shad are landed as unidentified bait. One fishery was observed on many occasions by individual states biologists, but could not be tracked well by any data reporting system since young fish were often difficult to identify and were lumped into the ubiquitous baitfish category in the NMFS reporting system.
Stock composition of OMS landings of American shad is important to assessing impacts of the losses to individual stocks. Good data on stock composition of shad OMS harvest are not available. However, published and unpublished tag release recapture studies of American shad have been conducted at several locations along the U.S. Atlantic coast. Studies include tag release locations in spawning estuaries and in mixed stock locations along the coast. There have also been a few DNA studies of stock composition of the OMS harvest. Both tag release recapture and DNA studies provide some insight on stocks that might have contributed to the mixed stock harvest.
In this section, we summarize available data on commercial in-river and OMS landings of American shad along the U.S. Atlantic coast. We also summarized available tag release recapture and DNA data on American shad.
42
Methods Landings data were obtained from the National Marine Fisheries Service (NMFS) and from individual states. In some cases, landings were provided by both sources. Since 1998, landings were reported in annual ASMFC state compliance reports. State biologists were queried about the accuracy of landings data and occasional adjustments were made for obvious errors. For example, Florida landings were occasionally inflated by the inclusion of gizzard shad from locations such as the west coast where American shad are not present. Whenever state and NMFS landings differed for a given state and the discrepancy could not be resolved, we selected the higher value. This was an infrequent occurrence.
We segregated landings data summaries into in-river losses and ocean/mixed stock losses. The latter category includes mixed stock losses in lower Delaware Bay but not losses in Chesapeake Bay. We did not include data on potential losses of American shad to the coastal bait fishery.
DNA analyses and tag release-recapture data for American shad along the U.S. Atlantic coast were obtained from available reports and publications.
Restults Landings The OMS fisheries occurred primarily on pre-spawn shad beginning in late winter in southern states (North Carolina to South Carolina), late February through April in mid Atlantic states, and from summer through fall in New England waters. OMS fisheries to the south of North Carolina were not very large and most shad landings were from within state natal rivers (Figure 1.1.8-1, Table 1.1.8-2). Total natal river landings from this region (North Carolina to Florida) declined from the late 1970s to the late 1980s and then stabilized at a low level. Total mixed stock landings from this region increased from the late 1970s through the late 1980s and then decreased through 2005. Total mixed stock landings in this region only exceeded total natal river landings in one year (1988). A very different pattern occurred in the mid Atlantic and north (Maine through Virginia). Harvest in the OMS fisheries in this region was equal to or greater than harvest in natal river systems in those states that allowed fisheries (Figure 1.1.8-1, Table 1.1.8-2). Since 1970, total natal river landings generally declined while total mixed stock landings generally increased. Total mixed stock landings in this region exceeded natal river landings in the late 1980s and remained relatively high until about 2003 (Figure 1.1.8-2). It should be noted here that Maine and Massachusetts have not allowed any American shad to be landed since the 1960s.
DNA and Tagging Data Only two DNA mixed stock analysis studies have been conducted on American shad. Brown and Epifano (1994) obtained genetic samples from fish in the commercial harvest off of the coasts of Maryland and Virginia in 1991-1993. Results indicated high variation in stock composition among locations and among years. Results also differed between the DNA study and a tagging study (Jesien 1992) that released fish at the same locations (Figure 1.1.8-2). The second DNA study was conducted off the NJ coast in 1996 (Brown 1996) and concluded that fish originated from Canadian, Hudson River and Susquehanna River stocks. The results of this study were deemed questionable by the ASMFC Shad and river herring Technical Committee, since the Susquehanna River is primarily comprised of hatchery produced shad of Hudson River or Delaware River origin. The true Susquehanna stock may still exist, but as a tiny remnant given that major component of the returning stock are from hatchery contributions.
The results of several tagging studies (Talbot and Sykes 1958, Miller 1982 and Leggett unpublished) were summarized by Dadswell et al. (1987). The pattern of tag returns described the coastal migration of American shad. Shad from all regions of the coast summered in the Bay of Fundy, off the St. Lawrence, 43
and off the Canadian Maritimes/Gulf of Maine. Three partially distinct wintering areas occurred off Florida, the mid-Atlantic Bight, and the Scotian shelf. Parker (1992) and McCord (1988) tagged fish off of North and South Carolina; most shad were caught in each of the respective states waters or in systems to the south (Figure 1.1.8-3C and D). Jesien (1992) tagged shad in the ocean fishery off the coasts of Virginia and Maryland. In each of the years, a different mix of returns came from a wide range of the coast (GA to MA). Most of the fish tagged in the Maryland-Virginia region were recaptured in either Virginia rivers or the Delaware River south (Figure 1.1.8-3B).
The most recent tagging data are from an ongoing study initiated in 1995 by New York and New Jersey (Figure 1.1.8-3A). Most (25%) of the released shad tagged in lower Delaware Bay were harvested in the mixed stock fishery within the Bay. An additional 22% were caught in ocean fisheries, the directed portion of which is now closed. Others were recaptured either in the Delaware, Hudson and Connecticut rivers along with an array of returns from the St. Lawrence River and Canadian rivers to the north, to the Santee River in the south.
Discussion Clearly, the OMS harvest has been a large component of total American shad harvest over the last 25 years and since the late 1980s, it was the dominant component of shad harvest from Virginia north.
Although directed harvest of American shad in U.S. Atlantic coastal ocean waters stopped in 2005, it cannot be ignored in any assessment of American shad stock change. This is especially true given the wide range of stocks harvested and potentially affected by the mixed stock fishery.
Landings data were affected by a variety of reporting issues. States collecting and reporting data stated that underreporting was common among commercial fishermen. Moreover, some segments of the fishery in inland waters went unreported because fish were sold in local markets and not captured in the reporting process. Given these problems, it is likely that summarized data were biased low. However, they still provided valuable insight into the general magnitude and trends of harvest along the U.S. Atlantic coast.
Information on stock composition of the mixed stock harvest remains imprecise. DNA stock identification studies on American shad are still in their infancy and need work on verification and refinement. Tag release - recapture methods were affected by poorly known recapture effort, unknown reporting rates, and small recapture sample sizes. However, the approach remains valuable because data from many studies are available and two studies have been ongoing for over ten years. Moreover, tag release and recapture locations and times are generally reliable. Otolith chemical signature analysis is a new and promising technique (Walther et al., in press), but no data on ocean shad stock composition have been produced.
Literature Cited Brown, B. 1996. Mixed stock analysis of American shad in New Jersey's coastal intercept fishery. Final report to the Hudson River Foundation, New York NY.
Brown, B. and J. Epifano. 1995. Mixed stock analysis of American shad in Virginia's and Maryland's coastal intercept fisheries. Final report to the Virginia Marine Resources Commission. Sport Fish Restoration Project F-110-R 1992-1994.
Dadswell, M.J., G.D. Melvin, P.J. Williams and D.E. Themelis. 1987. Influences of origin, life history and chance on the Atlantic coast migration of American shad. IN Common strategies of anadromous and catadromous fishes. American Fisheries Society, Symposium #1.
44
Epifano, Smouse, Kobak and Brown: 1995. Mitochondrial DNA divergence among populations of American shad: how much variation is enough for mixed stock analysis CanJ FishAquat. Sci 52:1688-1702.
Hattala, K, R. Allen, N. Lazar, and R. OReilly. 1998. Stock contributions for American shad landings in mixed stock fisheries along the Atlantic coast. pp 140-153 in ASMFC. (Atlantic States Marine Fisheries Commission). 1998 American shad stock assessment peer review report. Washington, D.C. USA.
Jesien, R., S. Gaichas, J. Serafy and C. Hocutt. 1992. Stock composition of coastal Maryland and Virginia American shad fishery: 1992 tagging study. Report to Maryland Dept. Of Natural Resources, Annapolis, MD, USA.
McCord, W. 1988, 1987,1986. Monitoring and assessment of South Carolinas commercial shad fishery and incidental catch of sturgeons. Completion reports to National Marine Fisheries Service.
Parker, J.A. 1992. Migratory patterns and exploitation of American shad in the near shore ocean of southeastern North Carolina. N.A. Jour. of Fish. Management 12 (4) 752-759.
Talbot G. B. and J.E. Sykes. 1958. Atlantic coast migrations of American shad Fishery Bulletin 142.
Walther, B., S. Thorrold and J. Olney. In press. Geochemical signatures in otoliths record natal origins of American shad. Transactions of the American Fisheries Society.
45
Table 1.1.7-1 Existing mixed stock fisheries and past directed mixed stock fisheries for American shad along the Atlantic coast.
Ocean Waters Estuarine Stock Fishery Directed Fishery Undocumented Potential Stocks State Bycatch Fishery Area (prior to 2005) Bycatch Affected Maine No landing allowed X New Hampshire EEZ landings only, X 5% trip limit Massachusetts No landing allowed X Rhode Island X No landing allowed X Connecticut X New York 5% trip limit X Pennsylvania New Jersey X X Lower Delaware Bay SC to Canada Delaware X X Lower Delaware Bay SC to Canada Maryland X X Chesapeake Bay VA, MD, PA, DE Virginia X X Chesapeake Bay VA, MD, PA, DE North Carolina X X South Carolina X X 46 Georgia X Florida X
Table 1.1.7-2 Summary of American shad commercial landings (kg) as reported for inland natal and mixed stock/ocean fisheries by National Marine Fisheries Service and/or respective state agencies.
(a) Inland Natal River Systems CT NY1 NJ DE MD VA YEAR ME NH MA RI NC SC GA FL TOTAL CT R. HR(+NJhr) DE R. DE R. a-1980 a-1994 1970 78,518 113,763 2,223 - 194,504 1971 109,182 84,959 1,179 - 195,320 1972 113,037 144,562 953 - 258,552 1973 116,847 121,020 726 - 238,594 1974 112,130 110,678 544 136 223,489 1975 75,071 114,806 - 862 190,739 1976 177,811 98,930 - - 213,777 490,519 1977 150,777 94,167 - 1,542 196,083 442,569 1978 138,938 206,207 - - 196,906 542,050 1979 93,804 232,833 - - 122,835 51,512 500,984 1980 140,843 595,622 - 2,815 45,266 398,243 88,571 52,382 85,499 1,409,240 1981 147,284 281,323 - 5,785 37,859 101,352 110,717 134,656 88,825 907,800 1982 128,369 171,869 499 10,262 22,633 139,872 157,795 69,556 97,607 798,461 1983 193,234 208,384 1,950 6,826 16,049 211,022 200,532 57,075 133,214 1,028,285 1984 180,896 318,155 3,357 6,450 19,135 283,681 259,156 65,053 120,051 1,255,934 1985 182,347 342,951 10,478 13,289 3,704 136,455 148,091 105,019 121,395 1,063,729 1986 146,059 362,321 8,029 12,983 4,732 98,670 140,937 116,224 78,459 6,203.9 974,618 1987 151,457 310,345 9,163 4,736 107,795 129,949 57,557 113,173 6,080.1 890,255 1988 85,776 355,138 7,847 11,074 24,683 105,672 50,415 91,981 55.8 732,642 1989 82,147 220,314 7,620 5,556 46,886 129,207 53,786 92,173 123.8 637,813 1990 117,675 210,257 18,309 7,264 58,509 125,414 28,909 59,058 54,165.3 679,560 1991 67,722 149,401 10,475 5,314 23,058 116,707 46,035 51,721 5,797.9 476,231 1992 65,454 120,475 18,945 4,194 20,866 97,617 61,662 65,657 74.8 454,946 1993 43,845 62,692 8,869 5,900 30,106 68,343 22,537 57,096 - 299,388 1994 47,174 90,072 4,112 6,508 34,963 22,793 66,834 17.2 272,475 1995 27,931 112,885 5,357 6,483 46,668 69,529 97,918 43.1 366,815 1996 30,281 83,690 499 4,579 63,942 145,437 104,136 - 432,564 1997 41,279 67,799 4,196 3,843 54,983 104,310 93,748 - 370,158 1998 40,526 105,484 34 3,650 95,046 161,589 95,636 - 501,964 1999 20,219 66,501 2,572 932 44,748 92,727 45,328 - 273,027 2000 48,724 69,555 19,640 3,115 84,850 201,293 60,939 - 488,117 2001 26,869 45,997 31,343 1,668 63,157 97,614 54,604 - 321,252 2002 49,034 59,242 14,854 1,139 120,784 204,179 28,734 - 477,964 2003 50,407 49,998 38,192 2,154 173,609 160,751 37,002 - 512,112 2004 30,086 33,040 41,764 1,368 118,997 152,635 17,213 - 395,103 2005 32,626 23,807 21,112 307 86,656 18,071 - 182,579
Table 1.1.7-2 (cont.) Summary of American shad commercial landings (kg) as reported for inland natal and mixed stock/ocean fisheries by National Marine Fisheries Service and/or respective state agencies.
(b) Mixed Stock/Ocean Fishery YEAR ME NH MA RI CT NY NJ DE MD VA NC SC GA FL2 TOTAL 1970 4,445 23,905 6,078 34,428 1971 1,361 16,965 3,447 21,773 1972 4,082 23,542 4,173 31,797 1973 1,497 19,505 3,447 24,449 1974 181 23,814 3,629 27,624 1975 136 38,556 7,620 46,313 1976 1,134 31,933 16,148 49,216 1977 544 60,873 32,251 93,668 1978 11,118 363 907 59,512 31,707 103,607 1979 544 3,674 40,280 43,047 37,879 125,424 1980 12,682 3,130 3,810 907 51,484 54,296 40,731 43,507 1,789 70,341 282,675 1981 41,096 2,540 7,575 14,243 26,445 59,286 83,711 125,048 48,723 67,837 476,504 1982 11,741 1,225 13,336 35,970 33,385 126,917 148,498 125,645 29,021 111,171 636,909 1983 17,554 1,542 6,124 10,660 14,969 88,860 98,773 9,092 94,216 1,718 93,225 436,732 1984 15,157 2,313 13,472 16,602 70 15,241 94,802 93,342 8,658 292,272 6,129 178,240 736,298 1985 7,258 3,311 10,115 41,187 - 42,548 97,615 78,245 68,054 150,666 1,433 62,395 562,826 48 1986 10,438 7,666 27,261 23,769 431 33,067 71,215 96,143 57,255 161,295 28,616 102,255 619,410 1987 11,975 18,734 18,507 47,129 - 5,262 83,689 111,614 54,116 179,275 18,671 163,122 64,423 776,517 1988 14,461 20,837 22,967 55,339 181 7,031 105,916 132,047 120,042 194,521 22,720 117,209 120,771 934,042 1989 21,091 13,882 6,178 19,038 533 10,342 181,077 98,268 221,272 181,332 17,485 103,528 74,771 948,798 1990 5,354 17,330 2,540 10,337 1,393 2,444 203,801 215,420 128,663 147,500 16,812 73,199 77,058 901,851 1991 903 8,584 289 12,617 445 11,923 174,343 231,542 106,139 181,274 8,717 65,433 26,676 828,885 1992 658 4,492 140 6,029 161 9,585 129,552 124,326 90,191 196,043 10,866 49,490 22,514 644,046 1993 - 2,971 181 18,394 110 3,510 145,194 134,309 35,329 220,801 12,756 29,455 11,115 614,124 1994 477 12,803 130 8,137 848 2,722 98,372 102,474 15,261 171,361 15,375 32,618 11,308 471,885 1995 173 13,862 206 12,683 27 6,552 126,971 86,653 27,491 68,453 46,714 60,021 12,152 461,959 1996 485 16,118 61 6,452 11,458 95,275 120,241 60,165 108,008 26,385 100,719 1,656 547,022 1997 88 11,538 341 16,674 17,100 102,278 74,174 75,357 162,306 44,594 51,400 24 555,876 1998 192 6,881 801 15,236 41,424 105,678 94,641 54,085 157,628 53,533 30,612 8 560,719 1999 77 1,667 101 20,076 31,130 118,437 102,676 27,703 105,034 14,951 8,768 218 430,837 2000 132 2,695 122 7,854 11,603 96,983 75,384 16,350 74,481 50,319 40,342 363 376,627 2001 216 368 477 30,777 14,173 91,200 127,882 26,028 114,588 5,370 41,942 453,021 2002 8 - 192 39,553 27,634 110,488 55,201 18,741 69,087 3,800 38,293 362,998 2003 2 1 503 17,548 11,100 68,843 40,035 4,869 23,369 5,677 16,823 188,769 2004 4 49 12 6,652 6,828 56,996 78,933 3,795 25,087 3,050 20,839 202,245 2005 88 11 - - 1,950 20,009 55,762 101 77,922
Table 1.1.7-2 (cont.) Summary of American shad commercial landings (kg) as reported for inland natal and mixed stock/ocean fisheries by National Marine Fisheries Service and/or respective state agencies.
(c) Total Landings YEAR ME NH MA RI CT NY NJ DE MD VA NC SC GA FL TOTAL 1970 78,518 118,208 26,127 6,078 432,372 67,087 241,088 99,021 1,068,500 1971 109,182 86,320 18,144 3,447 308,539 45,042 190,512 114,534 875,720 1972 113,037 148,645 24,494 4,173 212,504 72,122 155,993 54,432 785,401 1973 116,847 122,517 20,231 3,447 145,606 11,884 108,592 44,906 574,031 1974 112,130 110,860 24,358 3,765 167,303 11,022 73,347 45,360 548,145 1975 75,071 114,942 38,556 8,482 109,426 28,305 82,555 14,833 472,170 1976 177,811 100,064 31,933 16,148 213,777 75,837 14,697 42,049 12,519 684,836 1977 150,777 94,712 60,873 33,793 196,083 54,896 - 53,523 43,910 688,566 1978 11,118 363 138,938 207,114 59,512 31,707 196,906 182,355 - 107,934 59,441 995,387 1979 544 93,804 236,507 40,280 43,047 122,835 126,133 89,391 121,509 52,078 926,128 1980 12,682 3,130 3,810 907 140,843 647,106 54,296 43,546 45,266 441,750 90,360 122,723 85,499 - 1,691,916 1981 41,096 2,540 7,575 14,243 147,284 307,768 59,286 89,495 37,859 226,400 159,440 202,492 88,825 - 1,384,304 1982 11,741 1,225 13,336 35,970 128,369 205,254 127,416 158,760 22,633 265,517 186,816 180,727 97,607 - 1,435,370 1983 17,554 1,542 6,124 10,660 193,234 223,353 90,811 105,598 25,140 305,237 202,251 150,299 133,214 - 1,465,017 1984 15,157 2,313 13,472 16,602 180,966 333,396 98,159 99,792 27,793 575,952 265,285 243,294 120,051 - 1,992,232 1985 7,258 3,311 10,115 41,187 182,347 385,498 108,093 91,534 71,757 287,122 149,524 167,414 121,395 - 1,626,556 49 1986 10,438 7,666 27,261 23,769 146,490 395,389 79,244 109,126 61,987 259,965 169,553 218,479 78,459 6,204 1,594,029 1987 11,975 18,734 18,507 47,129 151,457 315,607 92,852 116,349 54,116 287,070 148,620 220,680 113,173 70,503 1,666,772 1988 14,461 20,837 22,967 55,339 85,957 362,169 113,763 143,121 120,042 219,204 128,391 167,624 91,981 120,827 1,666,684 1989 21,091 13,882 6,178 19,038 82,680 230,656 188,698 103,825 221,272 228,218 146,692 157,314 92,173 74,895 1,586,611 1990 5,354 17,330 2,540 10,337 119,068 212,701 222,110 222,684 128,663 206,008 142,226 102,108 59,058 131,223 1,581,412 1991 903 8,584 289 12,617 68,167 161,325 184,817 236,856 106,139 204,332 125,424 111,468 51,721 32,474 1,305,116 1992 658 4,492 140 6,029 65,616 130,060 148,497 128,520 90,191 216,909 108,484 111,153 65,657 22,588 1,098,993 1993 - 2,971 181 18,394 43,955 66,202 154,063 140,210 35,329 250,907 81,099 51,992 57,096 11,115 913,512 1994 477 12,803 130 8,137 48,023 92,794 102,484 108,981 15,261 171,361 50,338 55,411 66,834 11,325 744,360 1995 173 13,862 206 12,683 27,958 119,437 132,328 93,137 27,491 68,453 93,381 129,550 97,918 12,195 828,774 1996 485 16,118 61 6,452 30,281 95,148 95,774 124,820 60,165 108,008 90,327 246,156 104,136 1,656 979,586 1997 88 11,538 341 16,674 41,279 84,900 106,474 78,018 75,357 162,306 99,577 155,710 93,748 24 926,034 1998 192 6,881 801 15,236 40,526 146,907 105,712 98,291 54,085 157,628 148,579 192,200 95,636 8 1,062,683 1999 77 1,667 101 20,076 20,219 97,631 121,009 103,609 27,703 105,034 59,699 101,495 45,328 218 703,864 2000 132 2,695 122 7,854 48,724 81,159 116,624 78,499 16,350 74,481 135,168 241,635 60,939 363 864,744 2001 216 368 477 30,777 26,869 60,170 122,543 129,550 26,028 114,588 68,528 139,556 54,604 - 774,273 2002 8 - 192 39,553 49,034 86,876 125,341 56,340 18,741 69,087 124,584 242,472 28,734 - 840,962 2003 2 1 503 17,548 50,407 61,098 107,036 42,188 4,869 23,369 179,286 177,574 37,002 - 700,882 2004 4 49 12 6,652 30,086 39,868 98,760 80,301 3,795 25,087 122,047 173,473 17,213 - 597,348 2005 88 11 - - 32,626 25,757 41,121 56,069 - - 86,757 - 18,071 - 260,501
- 1. NY=Hudson River NY plus NJ Hudson landings, NJ= total for state- Hudson land.
- 2. FL's Ocean fishery on FL stocks
- 3. DE total includes Nanticoke River landings
Figure 1.1.7-1 American shad landings from natal Atlantic coast rivers and mixed stock fisheries, 1970-2005. Total landings not specified to natal or mixed.
Kg American shad commercial landings (NC to FL)
Thousands 900 Total 800 Mixed 700 Natal 600 500 400 300 200 100 0
1970 1973 1976 1979 1982 1985 1988 1991 1994 1997 2000 2003 50
Figure 1.1.7-1 (cont.) American shad landings from natal Atlantic coast rivers and mixed stock fisheries, 1970-2005. Total landings not specified to natal or mixed.
Kg American shad commercial landings (ME to VA)
Thousands 1600 Mixed 1400 Natal 1200 1000 800 600 400 200 0
1970 1973 1976 1979 1982 1985 1988 1991 1994 1997 2000 2003 Landings , by percent natal v. mixed, ME-VA Natal Mixed 100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
1970 1973 1976 1979 1982 1985 1988 1991 1994 1997 2000 2003 51
Figure 1.1.7-2 Comparison of tag returns of shad released near Rudee Inlet VA and results of DNA analyses for shad sampled in the same area, 1992.
Percent tag returns for releases at Rudee Inlet VA 1992 Merrimack R.
GA - Sav-Ogee 5%
10% Delaware R.
10%
Santee NJ - ocean 5%
5%
VA Rivers 10%
NC-rivers 30%
VA - ocean 25%
DNA stock ID (%) for samples off Rudee Inlet, VA 1992 St. John'sFL 2%
VA Rivers 24% St. Lawrence 30%
Delaware 4%
Connecticut R.
17% St. John CA 23%
52
Figure 1.1.7-3 Tag returns from studies conducted along the Atlantic coast in lower Delaware Bay, MD-VA, North Carolina and South Carolina coasts.
A Tag returns from A. shad tagged in Lower Delaware Bay, 1995-2005 Other* 6.8 %
Ocean
- Other rivers/areas:
22%
Cape Fear, NC Chesapeake Bay Hudson R. Indian, DE 20% Lehigh, PA Pawcatuck, RI Petitcodiac, NB 0 Santee, SC Connecticut R. Shubenacadie, NS 14% St. Lawrence, QB Susquehanna, MD Delaware R. Westfield, MA 16%
Delaware Bay 25%
B Tag returns from A. shad released off MD-VA coast, 1991-92 Merrimack 2%`
SC GA NC 3% 3% CT 3%Hudson 3%
10%
NC-ocean Delaware R. 22%
7%
VA-ocean Delaware Bay 3%
18%
DE-NJ ocean 3%
Chesapeake Bay 21%
53
Figure 1.1.7-3 (cont.) Tag returns from studies conducted along the Atlantic coast in lower Delaware Bay, MD-VA, North Carolina and South Carolina coasts.
C Tag returns for A. shad tagged off NC shore, 1989-1990 GA - Sav-Ogee 4%
SC - Edisto NC - ocean 4%
23%
SC - Waccamaw-PeeDee 19.3%
NC- Cape Fear 50%
D Tag returns from A. shad released off S. Carolina 1986-1988.
FL St. John's NC 2% 1%
SC ocean GA rivers 6%
10%
SC rivers 81%
54
1.2 COASTWIDE SUMMARIES Authors:
Andrew Kahnle and Kathryn Hattala Hudson River Fisheries Unit, New York State Department of Environmental Conservation 21 South Putt Corners Road, New Paltz, New York 12561 Patrick Kilduff Atlantic States Marine Fisheries Commission 1444 Eye Street, NW, 6th Floor, Washington, D.C. 20005 The following section provides a coastwide overview of juvenile abundance indices, upriver passage numbers at dams, total mortality estimates, mean length and age, and catch-per-unit-effort (CPUE) estimates evaluated in this assessment. An overview of the rivers contained within this stock assessment, information on current American shad fisheries, and data use and availability is provided in Table 1.2.
1.2.1 Juvenile Abundance Indices Approach All available juvenile abundance indices (JAI) were examined for temporal and geographic patterns in young-of-year production (Figure 1.2.1). No young-of-year data are available south of Albemarle Sound.
New England Young-of-year production in New England has varied without trend for the last 20 years with the exception of a dramatic rise for the Merrymeeting Bay and a moderate rise in the Connecticut River data since 2001 (Figure 1.2.1a). All indices (Maine, Connecticut, and Rhode Island) had synchronous declines in 1998 and 2001.
Hudson and Delaware Rivers Hudson River and non-tidal Delaware River JAIs increased though late 1980s. The Hudson River experienced, before undergoing a general decline with poor recruitment in the Hudson since 2002 (Figure 1.2.1b). The non-tidal Delaware River trend experienced a peak in 1996 followed by an alternating pattern of decrease (1996-2002) and increase (2002-2005). There was an increase in the tidal Delaware River JAI in 1988 followed by fluctuation without trend through the present. All three indices show a decrease in 2002.
Upper Chesapeake Bay (Maryland) and Potomac River Young-of-year production in the upper Chesapeake Bay has increased since 1995, but the index has shown wide year-to-year fluctuations (Figure 1.2.1c). JAIs for the Nanticoke River and Potomac River have increased steadily since 1995.
Lower Chesapeake Bay Rivers (Virginia) and Albemarle Sound All JAI indices from the lower Chesapeake Bay (James, York, and Rappahannock rivers) and Albemarle sound show have a period of relatively high values in early 1980s, a low period from the mid-1980s to the 55
mid-1990s, and then occasional high values until 2005 when all indices declined (Figure 1.2.1d). Virginia rivers show greater interannual fluctuation than the Roanoke River (Albemarle Sound). The James River JAI time series displays no measurable recruitment in most years with only 5 non-zero years since 1980.
Strong recruitment occurred in 1996 and 2003 on the York River. The Rappahannock JAI had moderate peaks in 1996, 2003 and 2004. In other years on these systems, recruitment was low to moderate. The York River and Rappahannock River JAIs had peaks in 1996 and 2003 with low values in between. There has been consistent young-of-year production on the Albemarle Sound since the mid-1990s, but poor young-of-year production occurred in 2000, 2004, and 2005.
1.2.2 Upriver Fish Passage Data on annual numbers of fish passed upriver at dams on several Atlantic coastal rivers exhibited a coastwide pattern of increase followed by a decrease (Figure 1.2.2). Interestingly, the timing of this pattern varied somewhat among rivers. Passage at dams from the Santee River in South Carolina to the Lehigh River (a tributary of the Delaware River in Pennsylvania), the Merrimack River in Massachusetts, and the Saco River in Maine peaked in 1999 through 2002. Passage on the Connecticut River peaked in 1992 and on the Pawcatuck River in 1985; however, even at these facilities there was a moderate increase prior to 1999 or 2002 followed by a decline. Fish passage facilities take some time (years) to optimize passage and it is expected that there would be an increase in passage as operational experience is gained at each facility. Still, it is unexpected that passage on so many rivers would decline at approximately the same time. This observation suggests a coastwide change in environmental conditions or mortality factors that affected stocks from South Carolina to Maine within the last five years.
1.2.3 Total Mortality This section discusses observed temporal and spatial patterns observed in total mortality rate (Z) estimates for Atlantic coast American shad stocks from Maine to North Carolina (Figure 1.2.3). Data were obtained from fishery-independent and fishery-dependent sources. Please see individual state assessments (Sections 2 to 16) for details of the datasets used. Total mortality estimates were not developed for the Delaware River (ageing issues) or for rivers from South Carolina through Florida (lack of age data and stocks are semelparous). Z30 benchmarks were developed for regions from New England through North Carolina (see Section 1.1.5), but not for stocks from South Carolina through Florida.
Contrast among years was observed in Z-values within several systems, especially in systems with long datasets (Figure 1.2.3). Annual Z-estimates for males and females in the Connecticut River fluctuated without trend from 1980 through the mid-1990s. Interannual variation for both sexes then increased, and Z-estimates for females remained at the same level while those for males increased dramatically during the last 10 years. Z-estimates for the Hudson River (both sexes) were slightly above the Z30 for that river for the later half of the 1980s. Estimates then climbed dramatically and remained high until the last couple of years when estimates began to decline. Annual Z-estimates for females in Virginia rivers remained above Z30, but have generally declined since the late 1990s when these time series begin. Trends were not obvious for males in Virginia Rivers because of small sample sizes.
The Z-estimates in several systems fluctuated without trend for the duration of the time series (Figure 1.2.3). Estimates for sexes combined in Maryland rivers of the upper Chesapeake Bay fluctuated but were well above Z30. Z-estimates for both sexes fluctuated just above Z30 in the Nanticoke River (Chesapeake Bay) and Albemarle Sound in North Carolina; annual fluctuations in Z were larger in the Nanticoke River than Albemarle Sound until recent years. Potomac River Z-estimates for American shad are available for the last four years and they have declined.
56
Small sample sizes and data gaps among years clouded interannual patterns in North Carolina rivers south of Albemarle Sound and in most New England rivers. Z-estimates for the Exeter River in New Hampshire have fluctuated around Z30 for the last 15 years.
Available Z-estimates generally exceeded Z30 for most years in rivers where data were suitable for catch curve analysis and where data supported spawning stock biomass-per-recruit (SSBPR) modeling (Figure 1.2.3). There is evidence in these data, in conjunction with other data summarized in this assessment, that Z-values at these levels have affected stock characteristics. In the Hudson River stock, abundance, mean size, and mean age were relatively high in the 1980s when Z-values were relatively low. As Z-values climbed in the 1990s, abundance, mean size, and mean age all have declined. In the Connecticut River, mean age and mean repeat spawn have declined as Z-values increased. The Z-estimates in Albemarle Sound have remained at levels just about Z30 and abundance of mature American shad in that system has increased. Finally, although Z-values in the upper Chesapeake Bay have fluctuated without trend, they have been very high and that stock has declined.
1.2.4 Mean Length and Mean Age Summaries Approach In this assessment we often used interannual change in mean age to corroborate change in recruitment or mortality. Given the uncertainty about ageing of American shad (see Section 1.1.7) and the possibility that interranual change in mean age could be caused by concomitant change in ageing techniques, we thought it important to evaluate interannual consistency of ageing in the age data developed for this assessment. Since fish generally get older as they grow in length, we expected that changes in annual mean length would be correlated with changes in mean age for the same samples unless ageing technique varied over the time series. Absence of a significant relationship was considered an indication that ageing technique had changed. We assumed that length was measured without error. Since over twenty years of data were available in several systems, many shad populations experienced change in recruitment or mortality that led to change in mean length and mean age. Some age and length datasets varied substantially over the time series and provide good contrast for these analyses. We did not consider this evaluation a test of ageing accuracy.
For this evaluation, we plotted annual mean length and mean age on year by sex where available for each sample program and system. We further analyzed these relationships with regressions of age on length. In most cases, we queried agencies that generated age datasets about experience and consistency in personnel making age determinations.
Results Results of our analyses were mixed. Plots of mean length and mean age suggested that ageing techniques were consistent for some datasets and not for others (Figure 1.2.4). In six of 23 datasets tested, the slope between age and length was significantly different from zero (P<0.05) for both sexes (Table 1.2.4).
Significant slopes also occurred in two datasets for males only and in five datasets for females only.
Slopes were not significant for either sex in eight datasets. There was no apparent geographic or regional pattern in significant slopes. However, those datasets in which trends were significant for both sexes tended to be from agencies where a single individual with many years of experience aged the scales for the entire time series. In some datasets without significant trends, a change had occurred in personnel during the time series. In others, aberrant ages appeared early in the time series as the person making age determinations was gaining experience. It was interesting that ageing of females was consistent in more 57
datasets than males. All of the significant female datasets were from regions where shad were iteroparous spawners and spawning marks were expected on scales. Such marks very likely aided in age determinations of scales. It should be noted that small sample sizes might have skewed age results in some datasets. There was no relationship between number of years in the dataset and significance. In this assessment, we generally avoided use of age data to infer change in stock characteristics from time series without significant slopes.
1.2.5 Coastwide CPUE Trends Approach We evaluated trends in fishery-dependent and independent CPUE indices of adult American shad along the Atlantic coast with linear regressions, or in the case of the Altamaha River, with a quadratic relationship. In some instances, the available data were sub-set into shorter periods to examine trends in different portions of the time series. These analyses and accompanying statistics facilitated an objective assessment of trends over time. Results were reported for river-specific datasets and for datasets that contained CPUE for mixed stock fisheries (Table 1.2.5). The term mixed stock includes those shad partaking in the coastal ocean migration as well as those that are caught in estuarine waters before beginning migrations up natal streams.
New England There were no in-river adult CPUE indices available for American shad from Maine to Rhode Island.
American shad abundance indices developed from surveys conducted by Connecticut Department of Environmental Protection in the Connecticut River and the Long Island Sound did not show a significant trend over the past two decades.
Hudson and Delaware Rivers There was a significant negative trend in the Hudson River commercial gill net CPUE from 1986 to 1999.
Although this time series continued through 2001, sample sizes are too low to provide confidence in later CPUE estimates. The longest time series of CPUE data for the Delaware River came from the Lewis haul fishery, which is a small commercial operation above tidal waters. The CPUE from this fishery increased significantly concurrent with marked water quality improvements associated with implementation of the Clean Water Act. Both the Lewis haul seine index and the Delaware commercial CPUE have declined significantly in the past 17 years.
Chesapeake Bay Hook and line CPUE of American shad in the Susquehanna River collected by the Maryland Department of Natural Resources increased significantly from 1984 to 2001, but CPUE has decreased in recent years.
Commercial pound net CPUE on the Nanticoke River has increased significantly since 1988, while there has been no trend in commercial fyke net CPUE. During the last 10 years, both the fishery-dependent pound net CPUE and fishery-independent gill net CPUE have increased significantly on the Potomac River. There has been no trend in staked gill net CPUE on either the York or the Rappahannock Rivers since 1998, while there has been a significant increase in staked gill net CPUE on the James River over the same time. There has been no trend in fishery-independent electrofishing or gill-net survey indices on the James River.
58
North Carolina In Albemarle Sound, there has been a significant increase in fishery-independent gill net CPUE over the past 16 years and in the fishery dependent gill net CPUE over the past 12 years. There are no significant trends in any other North Carolina river systems.
South Carolina, Georgia, and Florida The commercial drift gill net CPUE for American shad in Winyah Bay, South Carolina did not show a significant trend from 1981 to 1997. Commercial gill net CPUE in the Waccamaw River increased significantly from 1979 to 2005, while there was no trend in CPUE for American shad from the Pee Dee River from 1979 to 1999.
While both the upper set gill-net fishery and lower drift net fishery CPUE indices increased on the Santee River, only the increase in the lower drift net series was significant. In contrast, two of three commercial set gill-net CPUE series sets for the Edisto River showed a significant decline.
No significant trends were detected in the commercial gill-net CPUE indices for American shad in the Combahee River or the Savannah River, nor were any significant trends in CPUE series detected for the Altamaha River, Georgia (area under the curve of catch/net hour from fish tagging collections) or the St.
Johns River, Florida (recreational creel survey).
59
Table 1.2 List of rivers contained within this stock assessment, information on current American shad fisheries, and data use and availability. Please note that while rivers are listed in groups according to their geographic location, they are only assessed in one section (e.g., the Savannah River borders South Carolina and Georgia, it is only assessed in Section 14, which includes all South Carolina rivers and not in Section 15, which includes all Georgia rivers). American shad stocks assessed are bolded. If a recreational fishery is permitted, the creel limit for that fishery is indicated. Rivers with different creel limits in different states have the creels for each state separated by a / (e.g., St. Marys River: 8/10Georgia creel limit/Florida creel limit). Y = yes; Moratorium = commercial moratorium; = moratorium exists but bycatch is permitted; CR = catch and release; FD = fishery-dependent; FI = fishery-independent; JAI = juvenile abundance index; z = data available and used for assessment; l = data available but not used for assessment. Empty cells indicate a negative response.
Fishery Available Data State River* Assessed Moratorium Commercial Recreational FD - Comm FD - Rec FI - Adult JAI ME Nonesuch 2/day Presumpscot 2/day Royal 2/day Merrymeeting Bay 2/day l l Kennebec Y 2/day z 60 Cathance 2/day l l Androscoggin Y 2/day z Abbadagasett 2/day l Eastern 2/day l Sheepscot 2/day Penobscot 2/day Narraguangus 2/day l Pleasant 2/day Harrington 2/day Chandler 2/day Machias/East Machias 2/day Denny's 2/day St. Croix 2/day Saco Y 2/day z Mousam 2/day St. George 2/day Medomak 2/day Sebasticook 2/day
- River unless otherwise noted.
Table 1.2 (cont.) List of rivers contained within this stock assessment, information on current American shad fisheries, and data use and availability. Please note that while rivers are listed in groups according to their geographic location, they are only assessed in one section (e.g., the Savannah River borders South Carolina and Georgia, it is only assessed in Section 14, which includes all South Carolina rivers and not in Section 15, which includes all Georgia rivers). American shad stocks assessed are bolded.
If a recreational fishery is permitted, the creel limit for that fishery is indicated. Rivers with different creel limits in different states have the creels for each state separated by a / (e.g., St. Marys River: 8/10Georgia creel limit/Florida creel limit). Y
yes; Moratorium = commercial moratorium; = moratorium exists but bycatch is permitted; CR = catch and release; FD
fishery-dependent; FI = fishery-independent; JAI = juvenile abundance index; z = data available and used for assessment; l
= data available but not used for assessment. Empty cells indicate a negative response.
Fishery Available Data State River* Assessed Moratorium Commercial Recreational FD - Comm FD - Rec FI - Adult JAI NH Cocheco Y 2/day l Exeter Y Y 2/day z Oyster Y 2/day l Lamprey Y 2/day l Winnicut Y 2/day l 61 Great Bay Estuary Y 2/day l Salmon Falls Y 2/day Taylor Y 2/day l MA Merrimack Y Y 6/day z Neponset Y 6/day Charles Y 6/day Palmer Y 6/day Indianhead Y 6/day RI Runnins Y RI/CT Pawcatuck Y Y CR z z MA/CT Connecticut Y Y 6/day l l z z CT Thames Housatonic NY Hudson Y Y 6/day z z z z NY/PA/NJ/DE Delaware River and Bay Y Y 6/6/6/10 z z MD Nanticoke Y Y z z Choptank Y CR l l l
- River unless otherwise noted.
Table 1.2 (cont.) List of rivers contained within this stock assessment, information on current American shad fisheries, and data use and availability. Please note that while rivers are listed in groups according to their geographic location, they are only assessed in one section (e.g., the Savannah River borders South Carolina and Georgia, it is only assessed in Section 14, which includes all South Carolina rivers and not in Section 15, which includes all Georgia rivers). American shad stocks assessed are bolded.
If a recreational fishery is permitted, the creel limit for that fishery is indicated. Rivers with different creel limits in different states have the creels for each state separated by a / (e.g., St. Marys River: 8/10Georgia creel limit/Florida creel limit). Y
yes; Moratorium = commercial moratorium; = moratorium exists but bycatch is permitted; CR = catch and release; FD
fishery-dependent; FI = fishery-independent; JAI = juvenile abundance index; z = data available and used for assessment; l
= data available but not used for assessment. Empty cells indicate a negative response.
Fishery Available Data State River* Assessed Moratorium Commercial Recreational FD - Comm FD - Rec FI - Adult JAI MD (cont.) Patuxent Y CR l l l Pocomoke Y l Chesapeake Bay Y l PA/MD Susquehanna River and Flats Y Y CR l l z z MD/DC/VA Potomac Y CR z z z 62 VA York Y CR z l z z James Y CR z l z z Rappahannock Y CR z l z z NC Albemarle Sound Y Y 10/day z l z l Roanoke Y Y 10/day l z Chowan Y 10/day Pamlico Sound Y 10/day l Tar-Pamlico Y Y 10/day z l z Neuse Y Y 10/day z l z Cape Fear Y Y 10/day z l z SC Winyah Bay Y Y 10/day z l Waccamaw Y Y 10/day z l Great Pee Dee Y Y 10/day z Little Pee Dee Y 10/day l
- River unless otherwise noted.
Table 1.2 (cont.) List of rivers contained within this stock assessment, information on current American shad fisheries, and data use and availability. Please note that while rivers are listed in groups according to their geographic location, they are only assessed in one section (e.g., the Savannah River borders South Carolina and Georgia, it is only assessed in Section 14, which includes all South Carolina rivers and not in Section 15, which includes all Georgia rivers). American shad stocks assessed are bolded.
If a recreational fishery is permitted, the creel limit for that fishery is indicated. Rivers with different creel limits in different states have the creels for each state separated by a / (e.g., St. Marys River: 8/10Georgia creel limit/Florida creel limit). Y
yes; Moratorium = commercial moratorium; = moratorium exists but bycatch is permitted; CR = catch and release; FD
fishery-dependent; FI = fishery-independent; JAI = juvenile abundance index; z = data available and used for assessment; l
= data available but not used for assessment. Empty cells indicate a negative response.
Fishery Available Data State River* Assessed Moratorium Commercial Recreational FD - Comm FD - Rec FI - Adult JAI SC (cont.) Lynches Y 10/day l Black Y 10/day l Sampit Y 10/day l Bull Creek Y 10/day Santee Y Y 20/day z z z l Cooper Y Y 10/day z z l 63 Wateree Y 10/day l Congaree Y 10/day Broad Y 10/day Wando Y 10/day Ashely Y 10/day Ashepoo Y 10/day l Combahee Y Y 10/day z Edisto Y Y 10/day z l l Coosawatchie Y 10/day SC/GA Savannah Y Y 10/8 z l GA Altamaha Y Y 8/day z z l Ogeechee Y Y 8/day z z l Satilla Y 8/day l GA/FL St. Marys Y^ 8/10 l FL St. Johns Y Y^ 10/day l z z Nassau 10/day
- River unless otherwise noted.
^ Commercial fishery is not active due to net bans.
Table 1.2.4 Linear regression statistics for comparison for relationships of mean age and mean total length in several Atlantic coastal rivers. Significant trends (P-value < 0.05) are identified by: Bold = Positve trend; Bold Italics = Negative trend.
Male Female River R-Square P Slope R-Square P Slope Saco, ME 0.623 0.061 0.013 0.693 0.039 0.007 Exeter, NH 0.346 0.021 0.022 0.003 0.835 0.003 Pawcatuck, RI 0.037 0.566 0.004 0.061 0.489 0.005 Hudson, NY 0.753 0.000 0.021 0.720 0.000 0.021 Susquehanna, MD 0.722 0.001 0.017 0.539 0.010 0.009 Potomac, MD 0.357 0.287 0.004 0.851 0.025 0.010 York, VA 0.414 0.085 0.039 0.687 0.011 0.031 Rappahannock, VA 0.255 0.247 0.066 0.342 0.167 0.029 James, VA 0.285 0.217 0.021 0.756 0.011 0.039 Roanoke, NC (Inland) 0.217 0.351 0.031 0.331 0.232 -0.037 Albemarle, NC 0.401 0.177 0.027 0.267 0.293 0.017 Tar-Pamlico, NC (Inland) 0.286 0.274 -0.019 0.701 0.038 -0.042 Pamlico, NC 0.769 0.123 0.038 0.007 0.913 0.003 Neuse, NC (Inland) 0.649 0.053 0.022 0.854 0.008 0.024 Cape Fear, NC (Inland) 0.050 0.669 -0.008 0.203 0.369 0.061 Cape Fear, NC 0.814 0.014 0.016 0.702 0.037 0.015 Santee, SC 0.613 0.066 0.060 0.094 0.554 0.013 Waccamaw, SC 0.677 0.023 0.012 0.392 0.133 0.019 Edisto, SC 0.817 0.005 0.021 0.906 0.001 0.030 PeeDee, SC 0.713 0.017 0.014 0.660 0.026 0.025 Altamaha, GA 0.001 0.969 0.000 0.017 0.870 0.002 64
Table 1.2.5 Summary of trends in catch-per-unit-effort (CPUE) from fishery-dependent and independent datasets for adult American shad using linear regressions of CPUE against year by river and by dataset. Datasets that sample mixed stocks of American shad are found at the end of the table. Not all datasets had data for all years. Significant trends (P-value < 0.05) are identified by: Bold =
Positve trend; Bold Italics = Negative trend. FD = Fishery-dependent, FI = Fishery-independent.
State River Adult Index Trend Years N-obs Slope (CPUE) R-squared P-value FL St. Johns Roving Creel Survey-CPUE Non-significant 1992-2005 11 -0.009 0.140 >0.05 GA Altamaha FI-Drift Gill Net CPUE Nonlinear 1986-2005 20 a= -0.014, 0.3381 a= 0.012, b= 0.323, b= 0.009, c= -0.018 c= 0.971 FD-Gill Net CPUE Non-significant 1982-1991 10 0.080 0.002 0.899 FD-Drift Gill Net CPUE Non-significant 2000-2005 6 4.107 0.151 0.446 Savannah FD-Drift Gill Net CPUE Non-significant 2000-2005 6 -6.700 0.460 0.137 SC Waccamaw FD-Drift Gill Net CPUE Positive 1979-2005 25 0.095 0.476 <0.001 Pee Dee FD-Set Gill Net CPUE Non-significant 1979-1999 20 -0.006 0.017 0.580 Santee FD-Set Gill Net CPUE Non-significant 1992-2005 14 0.160 0.137 0.051 FD-Drift Gill Net CPUE Positive 1980-2003 23 0.049 0.329 0.004 Edisto FD-Set Gill Net CPUE Negative 1980-2004 23 -0.035 0.186 0.040 65 FD-Set Gill Net CPUE Non-significant 1979-2004 21 -0.062 0.097 0.169 FD-Set Gill Net CPUE Negative 1980-2004 23 -0.040 0.184 0.041 Combahee FD-Set Gill Net CPUE Non-significant 1993-2004 11 0.012 0.019 0.683 Savannah FD-Set Gill Net CPUE Non-significant 1980-2005 24 -0.010 0.011 0.632 NC Roanoke FI-Electrofishing CPUE Non-significant 2000-2005 6 3.569 0.294 0.266 Pamlico FD-Gill Net CPUE Non-significant 1994-2005 12 -1.840 0.040 0.570 Tar FI-Electrofishing CPUE Non-significant 2000-2005 6 -8.617 0.063 0.685 Neuse FD-Gill Net CPUE Non-significant 1994-2005 12 4.450 0.170 0.190 FI-Electrofishing CPUE Non-significant 2000-2005 6 1.738 0.503 0.114 Cape Fear FD-Gill Net CPUE Non-significant 1994-2005 12 -2.110 0.100 0.330 FI-Electrofishing CPUE Non-significant 2000-2005 6 -21.620 0.218 0.428
Table 1.2.5 (cont.) Summary of trends in catch-per-unit-effort (CPUE) from fishery-dependent and independent datasets for adult American shad using linear regressions of CPUE against year by river and by dataset. Datasets that sample mixed stocks of American shad are found at the end of the table. Not all datasets had data for all years. Significant trends (P-value < 0.05) are identified by: Bold = Positve trend; Bold Italics = Negative trend. FD = Fishery-dependent, FI = Fishery-independent.
State River Adult Index Trend Years N-obs Slope (CPUE) R-squared P-value VA James FD-Staked Gill Net CPUE Positive 1998-2005 8 0.750 0.646 0.016 FI-Electrofishing CPUE Non-significant 2000-2004 5 0.047 0.044 0.689 FI-Gill Net CPUE Non-significant 2000-2005 6 0.900 0.138 0.538 York FD-Staked Gill Net CPUE Non-significant 1998-2005 8 -0.597 0.175 0.302 Rappahannock FD-Staked Gill Net CPUE Non-significant 1998-2005 8 0.688 0.476 0.058 PRFC Potomac FI-Gill Net CPUE Positive 1996-2005 10 0.417 0.874 <0.001 FD-Pound Net CPUE Positive 1988-2005 18 0.717 0.822 <0.001 MD Susquehanna FD-Hook and Line CPUE Positive 1984-2005 22 0.547 0.630 <0.001 Postive 1984-2001 18 0.740 0.780 <0.001 Non-significant 2002-2005 4 -2.014 0.624 0.210 66 Nanticoke FD-Pound Net CPUE Positive 1988-2005 17 0.034 0.247 0.042 FD-Fyke Net CPUE Non-significant 1988-2005 18 0.001 0.001 0.917 DE, PA, NJ, NY Delaware FD-Lewis Haul Seine CPUE Positve 1925-2005 81 0.217 0.159 <0.001 Negative 1989-2005 17 -2.391 0.579 <0.001 FD-Delaware Gill Net CPUE Negative 1989-2005 17 -2.592 0.392 0.009 NY Hudson FD-Male Gill Net CPUE Negative 1986-1999 14 -1.430 0.590 0.001 FD-Female Gill Net CPUE Negative 1986-1999 14 -2.960 0.700 0.000 CT Connecticut FD-Gill Net CPUE Non-significant 1981-2005 25 -0.034 0.019 0.511 RI Pawcatuck No CPUE data na na na na na na MA Merrimack No CPUE data na na na na na na NH Exeter, Lamprey, Cocheco No CPUE data na na na na na na ME Androscoggin, Kennebec, Cathance, Eastern No CPUE data na na na na na na
Table 1.2.5 (cont.) Summary of trends in catch-per-unit-effort (CPUE) from fishery-dependent and independent datasets for adult American shad using linear regressions of CPUE against year by river and by dataset. Datasets that sample mixed stocks of American shad are found at the end of the table. Not all datasets had data for all years. Significant trends (P-value < 0.05) are identified by: Bold = Positve trend; Bold Italics = Negative trend. FD = Fishery-dependent, FI = Fishery-independent.
State Mixed Stock Adult Index Trend Years N-obs Slope (CPUE) R-squared P-value SC Winyah Bay FD-Drift Gill Net CPUE Non-significant 1981-1997 17 0.060 0.050 0.390 NC Albemarle Sound FD-Gill Net CPUE Positive 1994-2005 12 7.487 0.443 0.018 FI-Gill Net CPUE Positive 1991-2005 16 0.005 0.418 0.009 MD Upper Chesapeake Bay FD-Pound Net CPUE Non-significant 1988-2000 13 1.006 0.220 0.124 CT Long Island Sound FI-Trawl Survey (Spring) Non-significant 1984-2005 22 -0.016 0.103 0.146 FI-Trawl Survey (Fall) Non-significant 1984-2004 21 -0.037 0.063 0.274 67
Figure 1.2.1-1 Normalized juvenile abundance indices by year for Atlantic coast rivers: (a)
Merrymeeting Bay, Pawcatuck River, Connecticut River; (b) Hudson River, Delaware River (striped bass recruitment surveytidal, shad recruitment surveynon-tidal); (c)
Upper Chesapeake Bay, Nanticoke River, and Potomac River; and (d) Rappahannock River, York River, James River, and Albemarle Sound.
(a) 6 Merrymeeting Bay ME Pawcatuck RI 5 Connecticut CT Normalized CPUE 4
3 c
2 1
0 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 (b) 4.5 Hudson NY Delaware NJ-Shad 4
Delaware NJ-bass 3.5 Normalized CPUE 3
2.5 2
1.5 1
0.5 0
1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 68
Figure 1.2.1-1 (cont.) Normalized juvenile abundance indices by year for Atlantic coast rivers: (a)
Merrymeeting Bay, Pawcatuck River, Connecticut River; (b) Hudson River, Delaware River (striped bass recruitment surveytidal, shad recruitment surveynon-tidal); (c) Upper Chesapeake Bay, Nanticoke River, and Potomac River; and (d) Rappahannock River, York River, James River, and Albemarle Sound.
(c) 10 Potomac MD 9 Upper Ches. Bay MD Nanticoke MD 8
7 Normalized CPUE 6
5 4
3 2
1 0
1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 (d) 12 Rappahannock VA 10 James VA York VA Albemarle S NC 8
Normalized CPUE 6
4 2
0 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 69
Figure 1.2.2. Coastwide plots of fish-lift and fishway annual counts of American shad on rivers along the U.S. Atlantic coast from the Saco River, Maine to the Santee River, South Carolina.
All count data are plotted on the same x-axis to ease temporal comparisons, but note that y-axis scales differ.
Cataract Dams - lowermost (S aco) 6000 5000 Numbers of fish 4000 3000 2000 1000 0
1950 1952 1954 1956 1958 1960 1962 1964 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 Year Essex Dam (Merrimack) 100000 80000 Numbers of fish 60000 40000 20000 0
1950 1952 1954 1956 1958 1960 1962 1964 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 Year Potter Hill Dam (Pawcatuck) 5000 4000 Numbers of fish 3000 2000 1000 0
1950 1952 1954 1956 1958 1960 1962 1964 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 Year 70
Figure 1.2.2 (cont.) Coastwide plots of fish-lift and fishway annual counts of American shad on rivers along the U.S. Atlantic coast from the Saco River, Maine to the Santee River, South Carolina. All count data are plotted on the same x-axis to ease temporal comparisons, but note that y-axis scales differ.
Holyoke Dam (Connecticut) 800000 700000 Numbers of fish 600000 500000 400000 300000 200000 100000 0
1950 1952 1954 1956 1958 1960 1962 1964 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 Year Easton Dam (Lehigh/ Delaware) 5000 4000 Numbers of fish 3000 2000 1000 0
1950 1952 1954 1956 1958 1960 1962 1964 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 Year Conowingo Dam #1 (Susque hanna) 250000 200000 Numbers of fish 150000 100000 50000 0
1950 1952 1954 1956 1958 1960 1962 1964 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 Ye ar 71
Figure 1.2.2 (cont.) Coastwide plots of fish-lift and fishway annual counts of American shad on rivers along the U.S. Atlantic coast from the Saco River, Maine to the Santee River, South Carolina. All count data are plotted on the same x-axis to ease temporal comparisons, but note that y-axis scales differ.
Holtwood Dam #2 (Susquehanna) 120000 100000 Numbers of fish 80000 60000 40000 20000 0
1950 1952 1954 1956 1958 1960 1962 1964 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 Year S afe Harbor Dam #3 (S usquehanna) 100000 80000 Numbers of fish 60000 40000 20000 0
1950 1952 1954 1956 1958 1960 1962 1964 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 Year York Haven Dam #4 (S usquehanna) 20000 Numbers of fish 15000 10000 5000 0
1950 1952 1954 1956 1958 1960 1962 1964 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 Year 72
Figure 1.2.2 (cont.) Coastwide plots of fish-lift and fishway annual counts of American shad on rivers along the U.S. Atlantic coast from the Saco River, Maine to the Santee River, South Carolina. All count data are plotted on the same x-axis to ease temporal comparisons, but note that y-axis scales differ.
Boshe rs Dam (Jame s) 800 700 600 Numbers of fish 500 400 300 200 100 0
1950 1952 1954 1956 1958 1960 1962 1964 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 Ye ar St. Ste phe n Dam (Sante e )
700000 600000 500000 Numbers of fish 400000 300000 200000 100000 0
1950 1952 1954 1956 1958 1960 1962 1964 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 Ye ar 73
Figure 1.2.3. Plots of scale based estimates of total mortality for American shad stocks by river system on the U.S. Atlantic coast from the Saco River, Maine to the Cape Fear River, North Carolina. Z30 estimates are plotted for systems compared to the benchmark. FI = fishery-independent, FD = fishery-dependent.
Merrimack R. MA & Saco R. ME MA-Male MA-Female 2.50 ME-Females ME-Sex comb Z30-NE 2.00 1.50 Z
1.00 0.50 0.00 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Exeter River, NH Male 2.50 Female Z30-NE 2.00 1.50 Z
1.00 0.50 0.00 1990 1992 1994 1996 1998 2000 2002 2004 Pawcatuck River, RI Male 3.00 Female Sex comb.
2.50 Z30-NE 2.00 Z 1.50 1.00 0.50 0.00 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 74
Figure 1.2.3 (cont.) Plots of scale based estimates of total mortality for American shad stocks by river system on the U.S. Atlantic coast from the Saco River, Maine to the Cape Fear River, North Carolina. Z30 estimates are plotted for systems compared to the benchmark. FI = fishery-independent, FD = fishery-dependent.
Connecticut River, CT Male Female Z30-NE 6.00 5.00 4.00 Z 3.00 2.00 1.00 0.00 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 Hudson River NY Male Female Z30-HR 1.60 1.40 1.20 1.00 Z 0.80 0.60 0.40 0.20 0.00 1980 1983 1986 1989 1992 1995 1998 2001 2004 Nanticoke R. MD Male Female Z30-VA 2.50 2.00 1.50 Z
1.00 0.50 0.00 1980 1983 1986 1989 1992 1995 1998 2001 2004 75
Figure 1.2.3 (cont.) Plots of scale based estimates of total mortality for American shad stocks by river system on the U.S. Atlantic coast from the Saco River, Maine to the Cape Fear River, North Carolina. Z30 estimates are plotted for systems compared to the benchmark. FI = fishery-independent, FD = fishery-dependent.
Maryland Rivers Susquehanna Z30-VA Potomac 4.00 3.50 3.00 2.50 Z 2.00 1.50 1.00 0.50 0.00 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 Virginia Rivers- Female York 2.00 Rappahannock James 1.80 Z30-VA 1.60 1.40 1.20 Z 1.00 0.80 0.60 0.40 0.20 0.00 1997 1998 1999 2000 2001 2002 2003 2004 2005 Virginia Rivers- Male York Rappahannock 2.00 James 1.80 Z30-VA 1.60 1.40 1.20 Z 1.00 0.80 0.60 0.40 0.20 0.00 1997 1998 1999 2000 2001 2002 2003 2004 2005 76
Figure 1.2.3 (cont.) Plots of scale based estimates of total mortality for American shad stocks by river system on the U.S. Atlantic coast from the Saco River, Maine to the Cape Fear River, North Carolina. Z30 estimates are plotted for systems compared to the benchmark. FI = fishery-independent, FD = fishery-dependent.
Albemarle Sound NC Male-FD Female-FD Male-FI Female-FI 3.50 Z30-NC 3.00 2.50 2.00 Z
1.50 1.00 0.50 0.00 1980 1983 1986 1989 1992 1995 1998 2001 2004 Pamlico R. NC Male-FD 2.50 Female-FD 2.00 1.50 Z
1.00 0.50 0.00 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 Neuse R. NC Male-FD 3.00 Female-FD 2.50 2.00 Z 1.50 1.00 0.50 0.00 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 77
Figure 1.2.3 (cont.) Plots of scale based estimates of total mortality for American shad stocks by river system on the U.S. Atlantic coast from the Saco River, Maine to the Cape Fear River, North Carolina. Z30 estimates are plotted for systems compared to the benchmark. FI = fishery-independent, FD = fishery-dependent.
Cape Fear R. NC Male-FD Female-FD 3.00 Male-FI Female-FI 2.50 2.00 Z 1.50 1.00 0.50 0.00 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 78
Figure 1.2.4. Plots of mean length versus mean age by river system from the Saco River, Maine to the Altamaha River, Georgia.
Saco River, ME 7.5 560 M-mean age F-mean age 7.0 M-mean TL F-mean TL 6.5 520 6.0 Mean Age Mean TL 5.5 480 5.0 4.5 440 4.0 3.5 400 1994 1995 1996 1997 M-mean 1998 age 1999 F-mean age Exeter River, NH M-mean TL F-mean TL 8.5 560 7.5 520 Mean Age Mean TL 6.5 480 5.5 440 4.5 3.5 400 1991 1993 1995 1997 1999 2001 2003 2005 Merrimack River, MA M-mean age F-mean age M-mean TL F-mean TL 7.5 600 560 6.5 Mean Ag e Mean TL 520 5.5 480 4.5 440 3.5 400 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 79
Figure 1.2.4 (cont.) Plots of mean length versus mean age by river system from the Saco River, Maine to the Altamaha River, Georgia.
Pawcatuck River, RI M-mean age F-mean age M-mean TL F-mean TL 7.00 600 6.50 560 6.00 Mean Age Mean TL 5.50 520 5.00 480 4.50 440 4.00 3.50 400 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 Hudson River, NY M-mean age F-mean age M-mean TL F-mean TL 7.5 600 7.0 560 6.5 Mean Age Mean TL 6.0 520 5.5 5.0 480 4.5 440 4.0 3.5 400 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 80
Figure 1.2.4 (cont.) Plots of mean length versus mean age by river system from the Saco River, Maine to the Altamaha River, Georgia.
Susquehanna River, MD M-mean age F-mean age M-mean TL F-mean TL 6 560 5.5 520 Mean Age 5
4.5 480 Mean TL 4
440 3.5 3 400 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 Potomac River M-mean age F-mean age M-mean TL F-mean TL 7.5 540 500 6.5 Mean Age Mean TL 460 5.5 420 4.5 380 3.5 340 1998 1999 2000 2001 2002 2003 2004 2005 81
Figure 1.2.4 (cont.) Plots of mean length versus mean age by river system from the Saco River, Maine to the Altamaha River, Georgia.
Rappahannock River, VA M-mean age F-mean age M-mean TL F-mean TL 7.5 540 6.5 500 Mean Age 5.5 460 Mean TL 4.5 420 3.5 380 1998 1999 2000 2001 2002 2003 2004 2005 York River, VA M-mean age F-mean age M-mean TL F-mean TL 7.5 540 500 6.5 Mean Age Mean TL 460 5.5 420 4.5 380 3.5 340 1998 1999 2000 2001 2002 2003 2004 2005 James River, VA M-mean age F-mean age M-mean TL F-mean TL 7.5 540 6.5 500 Mean Age Mean TL 5.5 460 4.5 420 3.5 380 1998 1999 2000 2001 2002 2003 2004 2005 82
Figure 1.2.4 (cont.) Plots of mean length versus mean age by river system from the Saco River, Maine to the Altamaha River, Georgia.
Roanoke R./ Albemarle S. NC ELE-mean age GN-mean age Male ELE-meanTL GN-meanTL 7.0 470 6.5 460 6.0 5.5 450 Mean Age Mean TL 5.0 440 4.5 4.0 430 3.5 420 3.0 410 2.5 2.0 400 2000 2001 2002 2003 2004 2005 Roanoke R./ Albemarle S. NC ELE-mean age GN-mean age Female ELE-meanTL GN-meanTL 8.0 530 7.0 520 6.0 510 Mean Age Mean TL 5.0 500 4.0 490 3.0 480 2.0 470 2000 2001 2002 2003 2004 2005 Tar-Pamlico R. NC ELE-mean age GN-mean age Male ELE-meanTL GN-meanTL 6.5 500 6.0 480 5.5 460 5.0 Mean Age Mean TL 4.5 440 4.0 420 3.5 400 3.0 2.5 380 2.0 360 2000 2001 2002 2003 2004 2005 83
Figure 1.2.4 (cont.) Plots of mean length versus mean age by river system from the Saco River, Maine to the Altamaha River, Georgia.
Tar-Pamlico R. NC ELE-mean age GN-mean age Female ELE-meanTL GN-meanTL 7.0 540 6.5 530 6.0 5.5 520 Mean Age Mean TL 5.0 510 4.5 4.0 500 3.5 490 3.0 480 2.5 2.0 470 2000 2001 2002 2003 2004 2005 Neuse River NC M-mean age F-mean age electro-fishing M-mean TL F-mean TL 6.5 540 6.0 500 5.5 5.0 460 Mean Age Mean TL 4.5 420 4.0 3.5 380 3.0 340 2.5 2.0 300 2000 2001 2002 2003 2004 2005 Cape Fear River NC ELE-mean age ELE-mean age Male ELE-meanTL ELE-meanTL 6.0 540 500 5.0 460 Mean Age Mean TL 4.0 420 380 3.0 340 2.0 300 2000 2001 2002 2003 2004 2005 84
Figure 1.2.4 (cont.) Plots of mean length versus mean age by river system from the Saco River, Maine to the Altamaha River, Georgia.
Cape Fear River ELE-mean age ELE-mean age Female ELE-meanTL ELE-meanTL 7.0 560 6.0 520 Mean Age Mean TL 5.0 480 4.0 440 3.0 400 2000 2001 2002 2003 2004 2005 Santee River, SC M-mean age F-mean age M-mean TL F-mean TL 7.0 600 560 6.0 Mean Ag e Mean TL 520 5.0 480 4.0 440 3.0 400 1979 1980 1981 1982 1983 1984 Waccamaw River, SC M-mean age F-mean age M-mean TL F-mean TL 6.0 600 560 5.0 Mean Ag e Mean TL 520 480 4.0 440 3.0 400 1979 1980 1981 1982 1983 1984 1985 85
Figure 1.2.4 (cont.) Plots of mean length versus mean age by river system from the Saco River, Maine to the Altamaha River, Georgia.
PeeDee River, SC M-mean age F-mean age M-mean TL F-mean TL 6.0 600 560 5.0 Mean Age Mean TL 520 480 4.0 440 3.0 400 1979 1980 1981 1982 1983 1984 1985 Edisto River, SC M-mean age F-mean age M-mean TL F-mean TL 6.0 560 520 5.0 Mean Age 480 Mean TL 4.0 440 3.0 400 1979 1980 1981 1982 1983 1984 1985 Altamaha River, GA M-mean age F-mean age M-mean TL F-mean TL 6.0 520 5.0 480 Mean Age Mean TL 4.0 440 3.0 400 1979 1980 1981 1982 1983 1984 86
Figure1.3-1. Shad rivers assessed along the Atlantic coast.
400 200 0 400 km 87
Figure 1.3-2 Shad rivers assessed in the northeast region.
r R
ebec ive Kenn o ggin Androsc River Sa r co Riv e Rive r
88 Conn Cocheco River im Merr ecticu t Lamprey River a ck Rive er Exeter River r Riv so n Hud Pawcatuck River 200 100 0 200 km
Status of American Shad Stocks in Maine Distribution, Biology and Management Over 20 coastal rivers in Maine once supported runs of American shad. Since colonial times, overfishing and habitat loss from dam construction and pollution contributed to the disappearance and dramatic declines of shad stocks throughout the state. Current management and monitoring of shad stocks occur in the Saco, Androscoggin, and Kennebec rivers. These and several other spawning rivers are tributaries of the Merrymeeting Bay estuary. Juveniles from all tributaries use the Bay as a nursery area.
Adult American shad enter Maine Rivers from mid-May to the end of June. Peak spawning can occurs during June or as late as July or August. Tagging studies of mature fish indicate that post spawning shad from Maine move as far north as New Brunswick in the summer and south to North Carolina in winter.
The stock is iteroparous with the oldest observed fish at age 11. Recent ages ranged from two to nine, but most adult fish are four through seven years old.
There are seven dams on the Saco River that block migration starting at head-of-tide. The first four dams have upstream passage, but passage at the third dam (1 km above head-of-tide) does not pass shad well. There are currently no downstream fish passage facilities on the Saco River. Three dams on the main stem Androscoggin River block passage starting 10 km above head-of-tide. Spawning is suspected below the lowermost dam. There is ineffective upstream fish passage at the first dam and upstream and downstream passage at the next two dams. The lowermost dam on the Kennebec River was located at head-of-tide; it was removed in 1999, opening 70 km of river, up to the next impassible dam, which does not have fish passage.
The restoration goal is to restore access of shad to historic spawning reaches in the main stem and major tributaries of each system. Restoration targets exceed 200,000 adult for the Saco and Androscoggin Rivers and 725,000 adults for the Kennebec River. Restoration activities include Weight (kg in thousands) 1,600 stocking larvae in the Saco River and larvae 1,400 and pre-spawn adults to the Androscoggin 1,200 and Kennebec rivers. Stocked larvae have 1,000 OTC marks. Between two and ten percent of 800 juveniles collected in Merrymeeting Bay 600 had marks (2003-2005); two to 17 percent 400 200 had marks in the Kennebec (2001-2005). 0 1887 1898 1909 1932 1943 1954 1965 1976 1987 1998 The Fishery Shad were historically harvested from all the major rivers along the coast of Maine. Statewide landings peaked at 1.5 million kg in 1912, dropped to mean of 51,400 kg in 1928-1933, and became commercially extinct through 1940. Landings increased to a high of 502,044 kg in 1945 and remained at a relatively low level from 1948 through 1976. From 1978 89
6,000 to 1990, landings averaged 14,369 kg. Since Number of Fish Passed 5,000 the directed fishery closed in 1995, annual 4,000 landings have been less than 200 kg. A limited recreational fishery occurs in the Saco and 3,000 several small coastal rivers.
2,000 Indices 1,000 Adult passage at the lowermost dam on the 0 Saco River peaked in 1998 and has since 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 gradually declined. Passage has remained well below the restoration target.
Relative abundance of juvenile American shad from Merrymeeting Bay was relatively high in the 1980s, low during the 1990s and has generally increased since 2000.
1.20 Juvenile abundance was positively related to the number of larvae stocked in Merrymeeting Bay 1.00 Juvenile Abundance Index tributaries. The increase in juvenile abundance 0.80 followed the removal of the first dam on the Kennebec River (1999) and coincided with an 0.60 increase in larval stocking.
0.40 Assessment Results 0.20 Catch curve estimates of Z for the Saco River population (sexes combined) ranged from 0.94 0.00 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 to 1.63.
Saco River, Maine Benchmarks 2 The Z30 for New England was Z=0.64. Estimates for 1.5 the Saco River exceeded this value.
Z 1 0.5 Summary 0 Abundance of managed stocks of American shad in 94/95 95/96 96/97 97/98 98/99 Maine is well below their historic potential and current restoration goals. Stocks in the Saco and Androscoggin rivers have been denied access to most of the historic spawning reaches for over 100 years. Limited upstream and almost no downstream passage occurs past existing barriers on both rivers. Access to 70 km of spawning habitat was made available for shad of the Kennebec in 1999 with the removal of the first barrier. Stocks are maintained by supplemental stocking of larvae, limited passage on the Saco, and limited spawning below the first dam on the Androscoggin. High estimates of total mortality on the Saco stock are a concern because directed harvest has been banned. The Kennebec stock may increase now that access to some spawning habitat has been restored, but it will not reach full potential until access to the entire spawning reach has been provided.
Stocks of the Saco and Androscoggin rivers are not likely to improve until effective upstream and downstream passage have been developed at all barriers within the spawning reaches.
Summary Table River Source Data Year Range Trend Saco Adult trap at fishway 1993-2005 General decrease since 1999 Total mortality 1993-2005 Stable above Z30 Kennebec Juvenile abundance index 1984-2005 General increase since 1998 90
Status of American Shad Stocks in New Hampshire Distribution, Biology and Management New Hampshires coastal rivers once supported abundant runs of anadromous fish including American shad, river herring, and Atlantic salmon. All diadromous species have been denied access to historical, freshwater, spawning habitat since the construction of milldams as early as the 1600s but more dramatically during the nineteenth century textile boom. Restoration began with construction of fishways in the late 1960s and early 1970s in the Cocheco, Exeter, Oyster, Lamprey, Winnicut rivers in the Great Bay estuary, and the Taylor River in the Hampton-Seabrook estuary. A downstream passage facility is present on the Cocheco; on all others downstream passage is over the spillway.
Shad are iteroparous and are mature at age five or six for females and four through six for males. The oldest observed fish was age ten; the highest number of repeat spawn marks was four.
Restoration efforts include the movement of fish over barriers and supplemental stocking of adults in the Exeter.
The Fishery Shad were not reported for New Hampshire waters back to the turn of the 20th century. Recorded commercial harvest available since 1975 are fish of mixed stock origin and are landed from ocean waters outside of state waters in the Exclusive Economic Zone (EEZ) and caught with gill nets or trawls. Peak landings occurred in 1988 and again 25000 in 1996. Current landings are Commercial Landings minimal. In inland waters recreational 20000 catch-and-release fishing is permitted; a two fish take is allowed.
Kilograms 15000 Indices 10000 Recent restoration efforts focus on the Exeter River. Adult return data are 5000 poor. A small residual run of shad persists in the Lamprey and Cocheco 0
rivers, although no stocking has 1975 1978 1981 1984 1987 1990 1993 1996 1999 2002 2005 occurred in these rivers since 1988.
The low number of returns may be 91
due to difficulties associated with upriver or downriver passage over the dam, increased mortality from other sources, or both. Passage data were not used as abundance indices as many passage changes occurred throughout the times Fish Stocking in Exeter River series.
1600 Returns Assessment Results 1400 Stocked Total mortality estimates calculated 1200 from adult age structure in the Exeter River have varied without Number of Fish 1000 trend over the time series.
800 600 Benchmarks Sample sizes are too low to make 400 conclusions on total mortality rates 200 relative to a benchmark of 0
Z30=0.64.
80 82 84 86 88 90 92 94 96 98 00 02 04 19 19 19 19 19 19 19 19 19 19 20 20 20 Summary Year New Hampshires rivers are under restoration. Returns to the Exeter are extremely small relative to the Exeter River number stocked five to six years 2.50 earlier. As restoration activities Males continue (stocking spawning adults 2.00 Females or larvae), efforts should be made to obtain effective upstream and 1.50 downstream passage at all barriers, Z and improve water quality. Efforts should also be made to identify and 1.00 reduce all sources of mortality whether during ocean residency or 0.50 in-river.
0.00 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Year 92
Status of American Shad Stocks in the Merrimack River Distribution, Biology and Management American shad were formerly an important anadromous fish in Massachusetts. Shad were historically abundant in the larger rivers of the Commonwealth including the Connecticut, Merrimack, Neponset, and Charles Rivers, and also in a few smaller rivers including the Palmer and Indianhead.
Early historical records indicate fisheries were highly valued as the Commonwealth enacted legislation to protect anadromous runs. Over the last century, however, most runs were extirpated or reduced to extremely small populations due to construction of dams, water pollution at the spawning grounds, and overfishing.
The major coastal run in Massachusetts occurs in the Merrimack River. Shad used to move up the entire Merrimack up to km 185. Presently, the main stem Merrimack Mainstem dam with has five hydroelectric dams; fish passage is passage provided on the lower three dams. Recent First mainstem dam without passage upgrades at Essex (km 46, #1) and Pawtucket (km 65 #2) dams occurred within the past 25 years. Completion of a passage facility at the Essex Dam occurred in 1982. In 1986, a fish-lift and a modified Ice Harbor ladder were installed at the Pawtucket Dam. For the first time in nearly a century, adult shad were able to reach the Amoskeag Dam (km 120, #3) in New Hampshire. Through cooperative efforts of state and federal agencies, in 1989, a fish ladder was constructed at the Amoskeag Dam. This fishway allows shad access to the Hooksett Dam (km 135, #4). Resolution of problems associated with upstream and downstream passage problems is ongoing at each fishway.
Merrimack shad are iteroparous. Females mature at age five or six, and males at four through six. The oldest observed fish was age ten; the highest number of repeat spawn marks was five for 1200000 females and six for males.
1000000 Massachusetts reinstituted a restoration program for the Charles River in 2005 Landings (kg) 800000 stocking up to three million hatchery raised American shad larvae. 600000 400000 The Fishery Historical records for commercial landings in 200000 Massachusetts begin in the late 1880s. Annual reporting began about 1928. Landing prior to 0 1945 may be from inland or ocean waters. 1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 93
Since 1950, all landings were from ocean waters. Total kilograms of fish varied between Essex Dam Fish Passage several thousand kilograms to nearly 133,000 90,000 kg. The highest catch of 957,000 kg occurred 80,000 in 1957, landed by purse seine. This fishery 70,000 60,000 was short lived as the gill net and pound net Number 50,000 became the primary gear. After 1967, catches 40,000 became more sporadic; one exception was a 30,000 period of increased landings from 1981 to 20,000 1989. Since 1987, a moratorium has been in 10,000 0
place on commercial shad harvest in all waters 1983 1986 1989 1992 1995 1998 2001 2004 of the Commonwealth of Massachusetts.
Current landings are minimal and Merrimack R., MA M-mean age F-mean age most likely from fisheries in the EEZ.
M-mean TL F-mean TL In state waters, a recreational catch of 7.5 600 six fish per day by hook and line only is permitted. Most shad fisheries are 560 6.5 believed to be predominantly catch-and-release efforts.
Mean Age Mean TL 520 5.5 480 Indices 4.5 Annual fish passage counts at the first 440 dam at Essex are not indicative of the 3.5 400 size of the Merrimack shad 1983 1986 1989 1992 1995 1998 2001 2004 population. The number of shad counted at the Essex Dam fish lift varies annually and may be related to environmental conditions, fish passage effectiveness, and the size of the population entering the river and reaching Essex Dam. Adult shad are known to spawn in the river downstream from the Essex Dam. As an example, in 2005 a significant high water event resulted in a closure of the lift for most of the spring season, 2.50 severely impacting passage.
Males Samples of mean age and mean total length were 2.00 Females obtained infrequently before 1999. Since 1999, a brief Z30 spike in larger older fish occurred in 2003. 1.50 Z
Assessment Results 1.00 Total mortality estimates are only available since 1999 and have been quite variable.
0.50 Benchmarks Sample sizes were adequate in most years. Estimates of 0.00 total mortality are at or exceed the benchmark of Z30=0.64. Such high estimates of mortality are of 19 20
- 20
- 20 20 99 00 20 02 20 04 05 01 03 concern in a stock in recovery.
The Merrimack River Anadromous Fish Restoration Program developed an interim objective of an average of 35,000 adults. However, the overall target is to develop and maintain a self-sustaining 94
population of shad in their historical spawning habitat. The interim objective was met for some years (1999 to 2004), but the target is far from being achieved until all passage issues at all dams are resolved.
Summary The Merrimack River shad stock is not recovered, but is subject to periodic restoration stocking events.
Recommendations include:
- 1. Addressing upstream and downstream fish passage problems at all dams.
- 2. Improving monitoring of the adult run below Essex dam to develop an index of adult abundance to evaluate passage efficiency.
- 3. Increasing sample size at the lift to better characterize the run.
- 4. Developing a juvenile abundance index for this system.
- 5. Increasing effort to identify and reduce all sources of mortality whether during ocean residency or in-river.
95
Status of American Shad in Rhode Island Distribution, Biology and Management Fisheries in Rhode Island rivers garnered much attention in the post colonial era through the 1700s, but soon lost their importance as the growing textile industry used the river as a source of waterpower. Decreased water quality from factory discharges and increased municipal waste from the growing population accompanied the construction of dams. Anadromous fish runs either disappeared quickly or were heavily fished. A few remnant runs of American shad persist in the Runnins and the Pawcatuck Rivers.
The Pawcatuck River is the only remaining monitored run of American shad in Rhode Island.
By 1896, upstream shad passage in the Pawcatuck River was completely blocked by dams (dams are indicated on map with black bars). Shad were extirpated from the Pawcatuck for nearly 100 years. The run was reestablished by stocking efforts that occurred from 1975 through 1985. Since 1986, the stock has relied on wild reproduction.
Rhode Island shad are iteroparous. Shad are fully mature at age-5 or 6 for females and 4 through 6 for males. The oldest observed fish was age-8. The occurrence of repeat spawning is very low, which may be due to difficulties associated with upriver or downriver passage over the dam, increased mortality from other sources, or both. Shad are limited to the lower 44 km of river; the first barrier at the Potter Hill Dam (km 22) has a fishway to Commercial Landings provide access above the dam.
Spawning occurs in the 22km 60000 above this dam and perhaps in a portion of the river below.
50000 The Fishery 40000 Recorded commercial harvest Landings (kg) in Rhode Island waters 30000 indicates landings occurred since the turn of the 20th 20000 century. All harvest since the 1950s and up to 2005 occurred 10000 on mixed stocks in ocean 0
waters. Recreational catch and 1887 1892 1897 1902 1907 1912 1917 1922 1927 1932 1937 1942 1947 1952 1957 1962 1967 1972 1977 1982 1987 1992 1997 2002 release fishing is permitted; no Year take is allowed.
Reported commercial landings of American shad peaked in 1896, and again in 1940 at 24 thousand kg. The fishery took an upswing in 1981 and continued to increase to the peak of 55 thousand kg in 1989, most landings from floating traps.
In 1990, landings dropped by two-thirds and remained there until another small peak of 40 thousand kg in 2002. Rhode Islands fishery was included in the closure of directed ocean fisheries as specifed by ASMFC Amendment 1 and landings dropped to near zero by 2005.
96
Indices Adult Passage at Potter Hill Dam Adult passage data from the trap at 4500 Potter Hill Dam on the Pawcatuck River 4000 characterize the adult population, but 3500 with some uncertainty as not all fish 3000 enter the fishway to the trap. A Number 2500 decreasing trend was evident from 1999 2000 to the present.
1500 1000 Other data present conundrums: there is 500 an increasing trend in mean total length, 0 yet a declining trend in mean age. In 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 addition, sex ratio (M:F), once Year dominated by males, is now 7.00 600 dominated by females, inferring 6.50 that smaller fish are not present 560 in the samples. This lack of 6.00 smaller fish combined with the Mean Age small sample size, could cause Mean TL 5.50 520 5.00 mean length to increase.
480 4.50 Total mortality estimates 440 calculated from adult age 4.00 structure were not used as 3.50 400 problems were identified with 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 the aging process.
M-mean age F-mean age M-mean TL F-mean TL Assessment Results Examination of most indices Sex Ratio indicate a declining trend in the stock.
6.0 30 5.0 1980-2005 25 Benchmarks Sex Ratio (1980-Sex Ratio (1979) 1979 The Pawcatuck River run was nonexistent 4.0 20 until 1979. It experienced brief relief in 3.0 15 the early 1980s due to stocking efforts to recreate the run. Left on its own, the run 2005) 2.0 10 has persisted at varying levels since 1986 1.0 5 and relies solely on wild production. The 0.0 0 current estimates of the adult run have 1979 1982 1985 1988 1991 1994 1997 2000 2003 declined dramatically since 1999 and are Year far below the long-term average of 1100 fish at the lift.
Summary The Pawcatuck River stock of American shad is in recovery. The stock relies solely on wild reproduction.
Decreasing numbers of shad have appeared at the fishway in the Pawcatuck since 1999. The greatest uncertainty lies in an unknown level of ocean mortalityeither from fisheries that could impact the stock (see bycatch section) or possibly predation.
Source Data Year Range Trend Adults Trapped at Fishway 1985-2005 Decreasing over all years Mean Size 1993-2205 Increasing Sex Ration (M:F) 1996-2005 Decreasing (favors F)
Mean Age 1981-1987, 2000-2005 Decreasing over all years BENCHMARK 1100 adults Currently (151)well below 97
Status of American Shad in the Connecticut River Distribution, Biology and Management American shad support sport and commercial fisheries on the Connecticut River. Data on relative abundance, age structure and sex ratio of the run, and the size and extent of the sport and commercial fisheries have been obtained annually since 1974.
The Connecticut River, the largest river in New England, extends about 400 miles from its source in New Hampshire, just south of the Canadian border, to the mouth at Old Saybrook, Connecticut. The river drainage encompasses 11,250 square miles within New Hampshire, Vermont, Massachusetts, and Connecticut. The Connecticut River flow is heavily regulated; there are at least 125 reservoirs within the basin used for power generation and 16 flood control reservoirs.
The Holyoke Dam in Massachusetts is the first barrier to upriver fish migration on the Connecticut River. A fish passage facility began operation at Holyoke Dam in 1955. Major technological improvements to the Holyoke lift were made in 1969, 1975, and 1976. No further improvements were until 2006.
Commercial fishing for American shad occurs in the main stem Connecticut River in the State of Connecticut The commercial season runs from April 1 to June 15 (5 min. stretched mesh). American shad may also be taken commercially in marine waters. In-river effort is restricted by two mandatory rest days per week, as well as gear restrictions. Angling is the only legal method for recreational harvest of American shad in Connecticut. The season runs from April 1 through June 30 in rivers and streams open to fishing all year; otherwise, the season runs from the 3rd Saturday in April through June 30. There is a recreational daily possession limit of six American and hickory shad in the aggregate, per person, in inland and marine waters. There have been no changes to Connecticut Statutes or regulations pertaining to American shad angling since 1999, when the existing six fish recreational creel limit was modified to include hickory shad as an aggregate creel limit for the two species.
Connecticut River American shad are iteroparous. Annual reproductive success is monitored through the collection of juvenile American shad and the calculation of an annual juvenile index of relative abundance.
The Fishery A commercial gill net fishery and a recreational hook and line fishery have harvested American shad in the Connecticut River since the late 1800s. The commercial shad fishery in the Connecticut River is a spring gill-net fishery (April-June) that extends from the river mouth to Glastonbury, Connecticut (river km 62); landings data date back to 1880. The fishery has changed little since the adoption of outboard-powered vessels, with the exception of the change to drift gill nets from haul seine, fixed gill nets, traps, and pound nets. The number of commercial American shad fishing licenses (effort) has declined since 98
peaks during and after World War II and is expected to stay low or further decrease as fishermen retire and are not replaced.
There is the potential for significant bycatch losses of American shad in the Atlantic herring fisheries in the Gulf of Maine, which lands annually more than 60 million pounds of herring annually.
Commercial fishermen report in-river commercial landings and effort. Commercial landings (numbers) of Connecticut River shad varied greatly from 1981 to 2005 (Table 1). Commercial landings in the River fell steadily from 1981 through 1999 before rebounding and then declining again. In-river commercial effort declined from 1981 through the present (Table 1). Studies indicate that in-river commercial fishermen might have underreported their landings by 35 to 67% annually from 1966 to 1983.
Recreational American shad landings in numbers were estimated from 1980 to 1996 and periodically thereafter by a roving creel census. No biological samples have been collected since 1978. Prior to 1994, recreational landings comprised up to 82% of annual total in-river landings (Table 1). Recreational landings fell dramatically thereafter and estimates became unreliable and imprecise as reflected by high
(>80%) proportional standard errors (PSE). Discard mortality of recreationally caught American shad was assumed to be 100%.
Both commercial and recreational in-river landings remained relatively high from 1981 to about 1992 with peak total landings of 159,000 fish occurring in 1986 (Table 1).
Estimates from the Marine Recreational Fisheries Statistics Survey of American shad harvested in ocean had excessive PSE estimates (>80%) and were considered unreliable.
Indices 12.00 Figure 1 CT River Commercial CPUE Commercial CPUE (catch per gill net day) peaked in 1986, declined, and 10.00 then peaked at a lower level in 2003. 8.00 There was no significant linear trend CPUE 6.00 through this time series (Table 1; Figure 1). 4.00 2.00 Annual lift counts at Holyoke Dam and Fish-lift generally increased 0.00 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 through 1992 when numbers peaked at Year 720,000 fish (Figure 2). Numbers declined sharply in 1993 and have fluctuated without trend since. A regression of lift numbers on year showed a significant increase from 1976 through 1992, but no trend from 1993 through the present. The fish-lift counts have two biases: (1) since it is at river km 140, all 800,000 Figure 2 Shad Passage at Holyoke Dam removals from commercial harvest 700,000 and most removals from recreational 600,000 downriver of the dam; and (2) the fractions of the population that Number of Fish 500,000 spawn above the dam and below the 400,000 dam are not known. The Connecticut 300,000 River stock of American shad 200,000 persisted since closure of the 100,000 Holyoke Dam in 1849 suggesting that spawning occurred below the 0
1955 1959 1963 1967 1971 1975 1979 1983 1987 1991 1995 1999 2003 dam.
Year 99
Relative abundance of age zero 14.00 Figure 3 Age-0 Abundance in CT River American shad in the Connecticut River 12.00 (geometric mean catch per seine haul) has varied without trend since 1981 10.00 Geometric Mean (Figure 3). 8.00 6.00 Assessment Results 4.00 Age structure and spawning history of mature American shad were derived 2.00 from scale samples collected at the 0.00 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 Figure 4 Males Year 7
Holyoke fish passage facility in Holyoke, 6
Massachusetts. Males and females were 5
predominately ages four through six (Figures 4 Age and 5). Age-structure of male and female 4 American shad showed a loss of older fish in recent years. Mean percent of repeat spawners 3 declined throughout the time series (Figure 6).
2 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 Estimates of total mortality (Z) for adult American shad were determined from age structure as the 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 Year negative slope of the linear regression of ln percent at age on age. Annual Z-estimates Figure 5 Females 60%
9 50%
Figure 6 Percent of Repeat Spawners 8
40%
7 Age 30%
6 5 20%
4 10%
3 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 0%
1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1970 1971 1972 1973 1976 1977 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1996 2001 2002 2003 2004 1996 1974 1975 1978 1979 1980 1981 1993 1994 1995 1997 1998 1999 2000 1997 1998 1999 2000 2001 2002 2003 2004 2005 Year Year Male Female increased throughout the time series for both 6 Figure 7 Z-Estimates for CT River males and females (Figure 7). Males 5 Females Benchmarks 4 Annual estimates of Z have exceeded the Z30 for the New England of Z = 0.64 for most years Z 3 since 1970. 2 1
Summary Juvenile indices and commercial CPUE varied 0 without trend since 1981. Lift counts at the first 1970 1974 1978 1982 1986 Year 1990 1994 *1998 2002 barrier increased through 1992, dropped drastically the following year, and have since varied without trend. In addition, there has been a systematic increase in Z since 1970 (exceeding the reference point in most years), the percentage of repeat spawners returning to the Holyoke fish-lift has decreased, and age-structure has eroded.
100
Summary Table Source Data Year Range Trend Adult trap at fishway 1976-2005 Increase to 1992, decline, no trend since 1993 Commercial CPUE 1981-2005 No trend Adult age structure 1970-2005 Loss of older fish Total mortality 1970-2005 Increasing and generally above Z30 Age zero index 1981-2005 No trend Table 1. Annual losses (numbers in thousands) of Connecticut River American shad.
In-river In-river Commercial Commercial Total In-river Commercial Ocean Year commercial Recreational Effort1 CPUE2 landings landings3 landings landings 1981 49 907 5.4 69 118 33 1982 40.5 790 5.13 44 84.5 42.5 1983 49.5 840 5.89 99 148.5 26.5 1984 39.5 575 6.87 71 110.5 35 1985 38 590 6.44 41 79 41 1986 54 525 10.29 105 159 39.5 1987 31.5 350 9 93 124.5 42 1988 31 450 6.89 53 84 51 1989 30.5 400 7.63 60 90.5 39 1990 22.5 500 4.5 38 60.5 39.5 1991 24 500 4.8 85 109 38.5 1992 25.5 410 6.22 120 145.5 25 1993 17 400 4.25 65 82 27.5 1994 16 350 4.57 45 61 16 1995 10.5 400 2.63 14 24.5 23 1996 12 300 4 11 23 24 1997 16 300 5.33 16 24.5 1998 16 300 5.33 16 30.5 1999 8 225 3.56 8 28 2000 17.5 225 7.78 17.5 17.5 2001 11 200 5.5 11 26.5 2002 21 250 8.4 21 26.5 2003 20 250 8 20 14 2004 12 225 5.33 12 12.5 2005 11 200 5.5 11 4 1 - gill net days, 2 - catch/gill net day*100, 3 - 1995 value may be incomplete 101
Figure 1.3-3 Shad rivers assessed in the mid-Atlantic region.
er er Riv na Riv han que Sus 102 Delaw are ma c Poto Pa Rive tux r en tR ive r
Choptank River Rap pah ann ock R iver Nanticoke River Jame s River York River -
200 100 0 200 km
Status of American Shad in the Hudson River, New York Distribution, Biology and Management Federal Dam Troy (km 245)
Anadromous fish have been harvested from the Hudson River since the 1600s when the Dutch first colonized the valley. The immigrants brought with them the skill and tools of fishing, and learned from the Native Americans about the numerous fish that arrived every spring.
Kingston Regulation of the fishery came early, beginning in the mid (km 146) 1800s. Gear restrictions and the first of the escapement periods were created by the New York legislature in 1868.
The Hudson River was also the location of the one of the first American shad hatcheries on the Atlantic coast. Seth Green, a New York fish commissioner and conservationist, worked with Federal biologists, of the then U.S. Bureau of Fisheries, to promote stocking shad to reverse perceived declines in shad stocks on the mid-Atlantic coast. Despite these efforts, the Hudson River American shad stocks New York City demise continued because of an unfortunate series of overfishing events, complicated by anthropogenic impacts that carried on for nearly 140 years.
The Hudson is a highly productive system. It is tidal for 254 km to the first impassable barrier. Shad spawn in fresh water in the upper two-thirds of river above Kingston, New York. In ocean waters, the coastal migratory range extends from the Bay of Fundy, Nova Scotia south to coastal waters of Virginia.
The oldest American shad observed was age 13. The stock is iteroparousolder fish have up to eight spawning marks. Fish can weigh up to nearly 4.5 kg, but average about 1.2 kg.
Current management is through a cooperative interstate fishery management plan coordinated through the Atlantic States Marine Fisheries Commission.
KG*1000 Hudson River Commercial Landings The Fishery 2500 Commercial harvest records begin in the late 2000 1800s. The highest peak of 1.9 million kg occurred in 1889. It is not known if fishing 1500 continued in the following five years; however, 1000 a second peak occurred in 1904 at 1.5 million kg. From the turn of the 20th century until 1936, 500 landings were relatively low. Over the period 1937-1948, just prior to, during, and after World 0 1880 1895 1910 1925 1940 1955 1970 1985 2000 War II, Hudson shad became a very important commodity on the worlds food market.
Sustained landings ranged from 1.1 to 1.9 million kg annually. By 1949 and through the 1950s, the stock rapidly collapsed. Fishing slowed for 20 years only to be followed by another resurgence in the 1980s.
Landings have declined since. During this last resurgence, a mixed-stock fishery developed in ocean waters (landings not indicated on graph). Two major ocean fishery regulations were implemented after 2000. The first, the Harbor Porpoise Take Reduction Program (HPTRP) forced ocean shad gill netters to use smaller mesh sizes. A regulated 40% decrease in effort occurred in 2002, followed by a complete closure in 2005. Current ocean bycatch remains undocumented.
103
Fixed gill nets Indices 60.00 The in-river fishery is monitored by onboard 50.00 Male observers and provides accurate catch-effort data.
Female We used the data from the passive fixed gear gill-Weekly c/f sum 40.00 net fishery in the lower Hudson. This CPUE series 30.00 indexes adult shad as they migrate through the 20.00 lower river to upriver spawning areas. Data are 10.00 usable until 2001, when sample size became too low to provide confidence in estimates. CPUE 0.00 1980 1983 1986 1989 1992 1995 1998 2001 2004* declined from a high in 1986 to a low in the late Year 1990s. Female CPUE spiked in the last two years, coincident with the HTPRP Mean TL - commercial fishery & spawing stock M-cf F-cf implementation. 600.0 M-ss F-ss A decline in fish size occurred 580.0 from the mid-1980s through the 560.0 late 1990s concurrent with 540.0 declining CPUE and landings. 520.0 mm 500.0 Samples from the fishery and the 480.0 independent spawning stock 460.0 survey indicated fish size was 440.0 dropping. Larger, older fish 420.0 disappeared through the late 400.0 1980s and mid 1990s. Some 1980 1983 1986 1989 1992 1995 1998 2001 2004 larger fish began to appear in 7.5 Spawning Stock Survey 2.5 2000 after ocean restrictions M-mean repeat F-mean repeat 7.0 banning large mesh gill nets M-mean age F-mean age 2.0 6.5 occurred.
Mean repeat spawn 6.0 1.5 Mean age In 2004 and 2005, larger fish 5.5 appeared and smaller, 5.0 1.0 younger fish disappeared, 4.5 most likely a result of poor 4.0 0.5 recruitment. 3.5 3.0 0.0 Mean age and mean repeat 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 spawning showed the same trend. Mean age and repeats declined until 1994, then stabilized until 2000, with some improvement in 2003 through 2005.
Young of year indices (normalized values) The juvenile abundance indices, 3.5 Geometric (beach seine and post-yolk sac 3
Arithmetic larval), increased through the early PYSL 1980s when the spawning stock was 2.5 high. It has since declined Number per hau 2 erratically hitting an all time low in 2001. It presently remains at this 1.5 low level.
1 0.5 0
1974 1977 1980 1983 1986 1989 1992 1995 1998 2001 2004 104
Assessment Results Total mortality rates (Z) M-age F-age M-rpt spwn F-rpt spwn Spawning Stock Survey All indices described above Z30 indicate a declining trend in the stock through the late 1980s and 1.80 the mid 1990s. Recruitment has 1.60 been poor since 2001. 1.40 Total mortalityZ 1.20 Total mortality estimates 1.00 calculated for the spawning stock 0.80 show that Z for the Hudson River 0.60 0.40 stock was stable in the late 1980s 0.20 ranging from 0.4 to 0.8. Z began 0.00 to climb through the 1990s and 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 remained high, but variable, for the rest of the time series. 120 ESSA We also calculated empirical 100 ESSB indices of spawning stock 80 ESSA-ESSB abundance (ESSA) and biomass (ESSB). Indices were high in the 60 1980s, then decreased at a varying 40 rate to the present low. The ESSB is at its lowest level in 20 years. 20 Benchmarks 0 We calculated a biological 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 reference point (Z30) from a Thompson-Bell biomass-per-recruit model using Hudson River inputs for weight-, maturity- and vulnerability-at-age, and M=0.3, based on maximum age of 13 observed in the stock. Z30 was 0.54.
Current Z values are well above this reference point.
Summary The Hudson River shad stock has experienced a series of major declines over the past 140 years. Current data indicate the stock is at its historic low. The rise in mortality in the last 20 years coincided with a decrease in mean age, mean size, and stock size. Mortality rates on the adult stock remain high, well above reference point. Recent poor recruitment is a major concern. Shad are vulnerable to a host of fisheries on the Atlantic coast during the entire duration of the ocean residency. Total ocean bycatch estimates remain unknown. Mortality on the stock needs to be reduced.
Source Data Year Range Trend Adult Fishery CPUE 1980-2001 Decreasing over all years, small spike in last two years ESSB 1985-2005 Decreasing Mean Size 1980-2205 Declined until 1999 with slight increase to present Mean Age 1980-2005 Declined until late 1990s with slight increase in past three years Mean Repeat 1980-2005 Decreasing over all years Juvenile Index 1980-2005 Decline since 1986, past four year at historic low Z 1984-2005 Increased in the early 1990s and then remained high Relative Exploitation 1985-2005 Increased in the early 1990s and then remained high Reference Point Z30=0.54 Current Z=0.87 105
Status of American Shad in the Delaware River Distribution, Biology and Management The Delaware River stretches for 330 miles from the East and West branches above Hancock, New York, to the mouth of Delaware Bay. The Delaware Drainage Basin includes the states of Delaware, New Jersey, New York, and Pennsylvania.
New Jersey and Delaware have adopted regulations for limited entry fisheries and gear specific limitations on commercial fishing while Pennsylvania and New York do not permit commercial harvest within the Delaware River Basin.
All Basin states have enacted six fish American shad recreational bag limits except the State of Delaware, which has a ten fish limit, combined American and hickory shad.
The Delaware River had the largest annual commercial shad harvest of any river on the Atlantic coast in the late 1890s with estimates ranging up to 19 million pounds in a year. The harvest declined rapidly in the early 1900s due to overfishing, water pollution, and dams on major tributaries, and the shad stock eventually collapsed. Water quality in the Delaware River improved following the passage of the Federal Clean Water Act in 1972. Although the Delaware River shad population had shown signs of recovery during the 1980s into the early 1990s, recent estimates of the adult stock have been well below the target of 750,000 fish. Delaware River and Bay Commercial Landings 100000 700000 90000 Lower Bay Landings (lbs)
River/Upper Bay Landings 600000 The Fisheries 80000 The commercial fishery for American 70000 500000 shad in the Delaware River occurs during 60000 400000 the spring spawning migration. Landings 50000 300000 data should be used with caution since 40000 (lbs) 30000 New Jersey landings are likely 20000 River/Upper Bay 200000 underreported prior to 1999 and there are 10000 Low er Bay 100000 no estimates of harvest for the state of 0 0 Delaware before 1985. Shad harvested in 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 the Delaware River or Upper Bay area are considered to be Delaware stock while those from the Lower Bay areas are mixed stock and the origin of these fish vary annually. Harvest numbers have increased in recent years due to mandatory reporting.
Many recreational surveys have been conducted within the Delaware River since 1965. Recreational catch estimates have fluctuated with a peak in 1992 at 83,141 shad caught throughout the Delaware River. Other recreational fishery data are limited to logbook data during years of population estimates and some recent trends from shad guide fishermen in the upper Delaware River.
Indices The longest time series of commercial effort is the Lewis haul seine fishery in the non-tidal section of the Delaware River above Trenton, New Jersey. The fishery employs seine nets of different length depending on the water flow and height. Although this may be problematic, the length of the time series still gives a good indication of spawning run strength. The fishery CPUE began steadily increasing from 1969 to 1992 before 106
dropping off significantly through recent years. Effort data from gill-net catches in Delaware (1989-2005) also show a peak in 1992 with a decreasing trend to low levels in recent years.
Lewis Haul Seine Fishery CPUE Juvenile abundance data are collected in two 60.00 surveys in non-tidal (upper) and tidal (lower) 50.00 areas during August, September, and October. Juvenile production has remained Fish/Haul 40.00 30.00 fairly stable except for a dominant year class 20.00 in 1996 and a poor year class in 2002.
10.00 0.00 Adult abundance estimates are limited to an 1925 1929 1933 1937 1941 1945 1949 1953 1957 1961 1965 1969 1973 1977 1981 1985 1989 1993 1997 2001 2005 index of population derived from mark-Year recapture studies and a hydroacoustic survey. The population index shows an DE River Mean Juvenile Abundance Index 7.00 increase in adult abundance throughout the 6.00 1970s and 1980s with a gradual descent from 5.00 1989 to 1999. The stock rebounded from 4.00 2000 to 2003 but faltered again in recent Z+2 3.00 years.
2.00 1.00 Assessment Results 0.00 1980 1982 1984 1986 Recent findings have determined that ageing 1988 1990 1992 1994 1996 1998 2000 2002 2004 Year of scales from Delaware River American shad can not be substantiated. While otolith technology is considered better, the principal scientists behind otolith ageing are not satisfied with the process at this time. Without confidence in these ageing techniques, the SASC agreed that alternative methods would be preferred to assess the Delaware River stock of American shad.
Indices were standardized (Z-transformed) with two added to eliminate any negative values. Two estimates of adult shad abundance, the Lewis haul seine fishery CPUE and population index were compared with each other and also averaged together as an indicator of the spawning run. The results show an adult abundance increase from the lows of the late 1970s to a peak abundance level in 1992. This was followed by a significant decrease in abundance through 1994. The adult population Estimates of Population Indices has continued to decline and is 5.000 poplew isavg currently at levels equivalent to the 4.500 popest early 1980s. 4.000 lew is cpue 3.500 There does not seem to be 3.000 recruitment problems with juvenile Z+2 2.500 production within the Delaware 2.000 River and therefore, based on YOY 1.500 abundance, the shad population 1.000 should have remained stable 0.500 throughout the time series.
0.000 1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 Year 107
Estimates of relative exploitation from commercial fisheries were developed for the 1985 to 2005 time period when reliable in-river estimates of harvest from the Delaware commercial fishery were available to ascertain if potential overfishing occurred during the period of low adult abundance in the early to mid-1990s. The analysis has shown a 6
decrease in relative exploitation since relflewis 1985 using the Lewis CPUE data .
5 relfrev relfnj The other two estimates are from 4 Upper Delaware Bay commercial fisheries and have also shown a Z+2 3
general downward trend since the beginning of their respective time 2
series. All three relative exploitation 1 estimates tract fairly well and show that in-river fishing mortality is 0
1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 probably not the cause of any decline of the Delaware River shad stock in recent history.
The analysis has shown an increase in relative exploitation in recent years, beginning in 2000, but the extent of this trend is unknown. In-river exploitation in the Lewis CPUE data varied without trend from 1985 to 1999 but increased significantly in recent years. This increase is likely the result of New Jerseys mandatory reporting enacted in 2000 and not an actual increase in exploitation.
The SASC also looked at potential interactions with striped bass within the Delaware Estuary to determine if the shad decline was a direct result a predator-prey relationship. The analyses indicate that striped bass numbers fluctuate similar to American shad and there is potential for striped bass to be a limiting factor to shad stock growth but there is no evidence that striped bass has caused a decline in the shad population.
Benchmarks The SASC discussed the development of benchmarks for relaxation of harvest regulations, in the event of stock improvement, and reduction of harvest to prevent recruitment failure, in the event of continued stock collapse. Although there are many potential indices for use as benchmarks the SASC could not come to consensus on which indices would be beneficial at this time. Examples of CPUEs for the Delaware River included in this document:
- Six commercial fishery CPUEs from the Delaware River and Bay
- Three CPUE estimates from the recreational fishery
- Estimates of YOY abundance
- Index of population
- Abundance CPUEs from hydroacoustic monitoring
- CPUE from the Smithfield Beach gillnetting and the Lehigh River fish ladder at Easton Since the SASC was unable to reach consensus on what could be considered the best scientific benchmark or benchmarks, the SASC considers all options for benchmarks to be on the table and asks that the Peer Review Panel provide comments on the scientific pros and cons of each for use when the Management Board and Technical Committee meet to discuss the issue in the future.
108
Summary After evaluating the data on the Delaware River shad stock, the SASC recommends the following:
- No relaxation of the current regulations or sampling requirements take effect until the shad population is estimated to be at least 750,000 fish throughout the entire spawning reach of the Delaware Basin for more than two consecutive years. This recommendation is taken from the original Delaware River Basin Plan and would be dependent on the Delaware Basin States determination of a reliable estimator of the population throughout the entire spawning reach of the Delaware Basin.
- Undertake a more thorough investigation into predator-prey relationships to determine if predation on shad by striped bass or other predators is a significant problem.
- Determine fishing mortality on the Delaware River stock from out of Basin activities including bycatch discard in other fisheries.
- Initiate investigations to ensure that habitat quality and suitability within the Delaware Basin is adequate to restore the American shad stock in the Delaware River and its tributaries.
- Obtain annual estimates of the recreational catch, harvest, and CPUE.
- Require all commercial shad fisheries within the Delaware Basin to sample for hatchery-marked restoration fish.
109
Status of American Shad in Maryland Distribution, Biology, and Management The Chesapeake Bay is a dynamic estuary with thousands of miles of tributaries that are both tidal and non-tidal.
Anadromous species once spawned in every major river system and in most of the smaller tributaries. American shad stocks were a key food source for early settlers and it was not Susq ueha nna Rive r
Delaw Riv er until gear improvements in the late 1800s that stocks took a are noticeable decline. Potom ac Riv e
r Pa tux en tR ive r
Choptank River Rap pa Adult American shad enter the lower Chesapeake Bay in hann ock R iver Nanticoke River February; peak spawning is during May but may occur - James River York River 200 100 0 200 km through June. This iteroparous stock has spawned up to four times and age structure indicates that most adult fish are four or five years old but age structure ranges from 2-11. Juvenile American shad hatch approximately one week after fertilization and most leave the Bay in late Fall.
The American shad restoration program in Maryland began in 1984 and has focused on restoring historic runs to the major tributaries. Presently, three tributaries are annually stocked with millions of juvenile American shad. Commercial Landings 3500 Metric Tons Landed The Fishery 3000 American shad have long been important in 2500 the cultural and economic development of the Chesapeake Bay region. Commercial 2000 fishing increased beyond a subsistence 1500 fishery after the revolutionary war and by 1000 the late 1800's, American shad supported 500 the most important commercial fishery in 0 the Chesapeake Bay. American shad 80 97 20 32 38 49 55 61 67 73 79 landings in Maryland peaked in 1890, 18 18 19 19 19 19 19 19 19 19 19 declined to the late 1940's, increased Year through the 1950's and then declined to precarious levels by the late 1970s. The commercial American shad fishery in the Chesapeake Bay was virtually unregulated with only limited gear, area, and time restrictions and was the most important fishery in 2
1.8 the Chesapeake Bay until the 1950s.
1.6 1.4 Female American shad were targeted by the fishery Mortality Rate 1.2 because of their roe. The consistent demand for 1
0.8 American shad for flesh and roe demand dropped in the 0.6 1960s as consumers sought other fish because of the 0.4 Catch Curve Cohort Analysis scarcity of American shad. Limited effort during World 0.2 War II allowed the stock to rebuild but the illimitable 0
commercial fishery depleted stocks until Maryland 1997 1998 1999 2000 2001 2002 Year closed its commercial and recreational fishery in 1980.
Indices Limited American shad data exists for the Nanticoke, Choptank, Patuxent, and Pocomoke rivers and the coastal commercial fishery. The lack of absolute abundance estimates, relative abundance indicators, and long-term age-structure only allows index-based assessments for several areas.
110
Pound Net CPUE (GM) with Trend Line The Nanticoke River is the only assessed system in Maryland besides the Susquehanna River that 1.8 Expon. (Trend Line) 1.6 has sufficient data to allow an index-based 1.4 assessment of adult American shad stocks. The 1.2 CPUE (GM)
Choptank, Patuxent, and Pocomoke rivers each 1 had viable stocks but overfishing likely caused 0.8 0.6 their decline and these stocks based on limited 0.4 data, appear to have very low abundance. 0.2 0
Fishery-dependent data from Nanticoke River 19 19 19 19 19 19 19 19 19 19 19 19 20 20 20 20 88 89 90 91 92 93 94 95 96 97 98 99 00 01 02 pound nets have trended up during the last few 20 2003 04 05 Year years but this may be hatchery driven (analysis on adult otoliths is incomplete). The trend in repeat spawners, which is an inverse recruitment indicator, shows a very high number compared to the number of virgin fish, indicating overfishing or predation may be occurring on sub-adult fish. Mortality estimates for Nanticoke River American shad based on repeat spawning marks averaged 1.16 (catch curve) and 1.17 (cohort analysis) and there appears to be no trend Baywide Juvenile American Shad Indices Marylands American shad coastal 3.5 commercial fishery, closed in 2005, was a Geometric Mean CPUE 3
2.5 mixed stock fishery. Catch-per-unit-effort 2 based on the oceanside commercial gill net 1.5 fishery dropped dramatically after 1999 1 when regulations increased and price 0.5 dropped.
0 195 9 2 196 5
196 8
196 197 197 1 4
0 198 3
198 Baywide juvenile American shad indices, 6
198 9
198 199 199 2 5
200 200 1
although limited in some tributaries of the 197 7 199 8 4 Year Chesapeake Bay has increased since the 1990s but has trended down in the last two years if 2006 is included, although this index is heavily weighted by large catches in the Potomac and upper Chesapeake Bay.
In the Nanticoke River, supplemental stocking may have caused a behavioral shift resulting in the lower river being quickly bypassed especially during low spring flows (high salinity) because of the homing to the upper reaches of the river.
Non-hatchery juvenile American shad are prevalent in the upper Nanticoke River and indicate significant natural reproduction. Restoration stocking is also supplementing adult stocks but American shad adult catch by cooperative watermen in the Nanticoke River is well below historical landings.
In over twenty-five years of the American shad recreational and commercial closures and supplemental stocking, stocks have not rebounded to self-sustaining levels for most major tributaries of the Chesapeake Bay (the exception is the Maryland portion of the Susquehanna River). The introduction of hatchery-produced American shad into the Susquehanna, Patuxent, and Choptank rivers had initially rebounded in these systems until 2002 when adult relative abundance indicators declined and this trend has continued into 2006. Recent age and spawning history data showed fewer larger and older fish and a decrease in virgin spawners indicating increased mortality and recruitment failure.
Benchmarks In most years, Z estimates from the repeat spawning marks are higher than benchmark Z30 for the mid-Atlantic region. Since there is no directed fishery in the Chesapeake Bay and bycatch mortality is thought to be minimal, evaluation of the causes of increased American shad mortality is essential.
111
Summary Landings data from the Nanticoke, Patuxent, Choptank, and Pocomoke rivers and from the main Bay and ocean commercial shad fisheries demonstrate the decline of the American shad stock. Adult indices in the Choptank, Patuxent, and Pocomoke rivers were at historically low levels based on landings data and more recent empirical data suggests that stocks are still depressed. Baywide juvenile indices have increased since the late 1990s indicating increased juvenile survival. Nanticoke River adult indices are low, even though it is supplemented by hatchery fish.
112
Status of American Shad in the Susquehanna River & Susquehanna Flats Distribution, Biology and Management The American shad stock in the Susquehanna River and Flats was spatially restricted by the construction of hydroelectric dams during the early 1900s, blocking access to hundreds of miles of natal habitat in Pennsylvania and New York.
Overfishing likely impacted the stock as well.
Fish-lifts were retrofitted to the four main stem Susquehanna River dams, beginning in 1972, and have reopened upriver habitat.
Adult American shad enter the Susquehanna River in March; peak spawning is during May but may extend through June. The stock is iteroparous with individuals exhibiting as many Percent Hatchery Adult Shad at 100 90 as three spawn marks. Spawning fish range in 80 70 age from 2 to 11, but most adult fish are four or 60 five years old.
50 40 Presently, each of the four main-stem dams has Conowingo Dam Lifts lifts or ladders to pass adult American shad but 30 20 10 downstream fish passage is limited to turbine 0 passage or spill.
19 19 19 19 19 19 20 20 20 89 91 93 95 97 99 01 03 05 The annual stocking of millions of juvenile American shad, mainly by Pennsylvania, has driven the recent recovery of the stock as is indicated by the high percentage of hatchery adults collected at Conowingo Dam. However, there is no significant correlation between juvenile abundance indices and returning adults. Upper Chesapeake Bay, Susquehanna River 800 and Flats American Shad Landings (lbs in thousands)
The Fishery 19 44 700 The historic commercial American 19 47 600 shad fisheries in the Susquehanna 19 50 500 River targeted American shad until 19 53 Maryland closed the directed fishery 19 56 400 in 1980. After 1980, a significant 19 59 300 recreational catch-and-release 19 62 American shad fishery has developed 19 200 65 below Conowingo Dam. 19 68 100 19 71 0
19 Indices 74 Index-based methods were employed 19 77 in this assessment. Adult American shad relative abundance indicatorshook and line and Conowingo Dam fish lift countshad dramatically increased from 1972 to 2002, but have decreased since 2002. Significant changes in attractant flows have occurred at most fish-lifts on the Susquehanna River in order to maximize catchability of American shad.
113
Cumulative adult shad passage efficiency 18 90 (the number passed by the fourth dam 16 Hook and Line GM 80 divided by the number passed at the first 14 Lift GM 70 dam) is presently less than 3% and the Hook and Line GM 12 60 lack of downstream passage for adults Lift GM 10 50 minimizes repeat spawning. Fish passage 8 40 numbers at the three dams above Conowingo Dam indicate that the goal of 6 30 2 million American shad will be difficult 4 20 to obtain unless stocks increase 2 10 significantly and lift efficiencies are 0 0 maximized. 19 19 19 19 19 19 19 19 20 20 20 84 86 88 90 92 94 96 98 00 02 04 Upriver American shad hatchery juvenile abundance indices are not correlated with stocking but wild juvenile production is correlated with adult abundance above Safe Harbor Dam, indicating that access to pristine spawning habitat is limiting upriver reproduction.
200,000 175,000 Benchmarks 150,000 In most years our Z-estimates from the No. of Fish 125,000 100,000 repeat spawning marks are higher than the 75,000 benchmark Z30 for the mid-Atlantic region.
50,000 25,000 Summary 0 American shad stocks have rebounded since 19 19 19 20 20 20 20 20 20 the closure of the directed fishery in 1980.
97 98 99 00 01 02 03 04 05 Pennsylvania established a restoration program that triggered the recovery of the Conowingo Holtwood Safe Harbor York Haven stock until 2002 when abundance began declining. Z-estimates from 9 1200 Susquehanna River stocks are higher Haul Seine than the Z30 benchmark and the cause 8
needs to be identified. Fishway Lift Net Area Under the 1000 GM Combined CPUE -
Lift Net 7 efficiency must be significantly 6 800 improved to support a self-sustaining 5 population. Identification of the factors 600 causing the present decline is critical 4
Haul Seine Curve for recovery of this stock.
3 400 2
200 1
0 0 1990 1992 1994 1996 1998 2000 2002 2004 Year 114
Status of the Potomac River American Shad Stock Distribution, Biology and Management The Potomac River is a 4,500,000 major tributary of the 4,000,000 Figure 1.
Chesapeake Bay that 3,500,000 historically supported a 3,000,000 Estimated Pounds large fishery for American shad (Figure 1). Its 2,500,000 watershed ranks fourth in 2,000,000 area among all Atlantic east 1,500,000 coast rivers. The estuary 1,000,000 extends 113 miles from its mouth to just below Little 500,000 Falls. At Little Falls, a low 0 head dam was built in the 1878 1884 1890 1896 1902 1908 1914 1920 1926 1932 1938 1944 1950 1956 1962 1968 1974 1980 1986 1992 1998 2004 2005 19th century and traditionally had an ineffective fish passage way. A newly installed fish passage now allows shad to extend their range an additional 10 to 12 miles upstream to Great Falls, a natural barrier to all anadromous species. Shad spawning area extends from about Stump Neck, Maryland and Cockpit Point, Virginia upriver to Great Falls, a distance of about 40 miles. Upstream, spawning occurs in a 10 to 12 mile reach between Little Falls and Great Falls. A cooperative restoration program using hatchery-reared larvae began in 1995. During the eight-year stocking phase of the project over 15.8 million shad fry were stocked into the Potomac River. However, the prevalence of adult American shad with hatchery marks has remained very low since 1998, suggesting that the stocking of larvae has not yet influenced run size. The Potomac River serves as a source of brood stock for American shad recovery efforts in the Rappahannock River, other Maryland rivers, and the Susquehanna River.
Since 1963, the Potomac River Fisheries Commission (PRFC) has regulated the American shad fishery in the Potomac River. However, there are five fishery management authorities on the river, each with separate management areas and some with shad monitoring programs. The PRFC is the Maryland-Virginia bi-state Commission with fisheries management authority for the main stem, exclusive of the tributaries on either side, from the Chesapeake Bay to the southern Maryland-District of Columbia boundary line; the District of Columbia with authority for the Potomac to the Virginia shore and other waters within D.C.; the Maryland Department of Natural Resources (MD DNR) with authority for the tributaries of the Potomac on the Maryland side of the river and the fluvial portion of the river upstream of D.C.; the Virginia Marine Resources Commission with authority for commercial fisheries in all tidal Virginia tributaries and for recreational fisheries in the saltwater portions of the tidal Virginia tributaries below the Route 301 Bridge; and the Virginia Department of Game and Inland Fisheries with authority for recreational fisheries in the freshwater portions on the Virginia tributaries.
The Fishery A moratorium on the taking of shad was established in 1982. Currently, bycatch of American shad is permitted by pound nets and gill nets set for other fishes. Fishers are limited to a one bushel (approximately 60 lbs) per licensee, per day. Gill nets are fished from November through March 25 and pound nets can operate from February 15 to December 15 each year. Both gear types are limited entry fisheries such that no new licenses are sold. All licensed fishermen are 115
required to submit reports of their daily harvest of all species by gear type on forms supplied by the PRFC. The recreational fishery for American shad is currently closed.
Survey Indices Since 1985, the MD 14 DNR has employed 12 Figure 2.
multi-panel drift gill nets to monitor the 10 Geometric Mean Chesapeake Bay 8
component of the Atlantic coast 6 striped bass population. The 4 primary objective of 2 this survey is to generate estimates 0 of relative 19 19 19 19 19 19 19 59 62 65 68 71 74 77 abundance-at-age 19 19 19 19 19 1980 83 86 89 92 19 20 2095 98 01 04 for striped bass.
American shad are caught as bycatch and Current Potomac Pound Net Index MD DNR personnel have collected catch Bycatch + Discards and biological data on this species since 1997. MD DNR has sampled by seine for 100 juvenile shad abundance on the Potomac River since 1954. This survey was Figure 3.
Pounds per net-day originally intended to collect YOY striped 80 bass but it serves to generate indices on many fish species. 60 Assessment Approaches 40 We examined trends in estimates of total mortality (Table 1), juvenile abundance (Figure 2), landings plus discards in 20 commercial pound nets (Figure 3), and catches in the MD DNR fishery- 0 independent gill-net survey (Figure 4). To 1998 1999 2000 2001 2002 2003 2004 2005 2006 assess stock status and set a management Year benchmark, we compared current pound net landings during 1999-2005 (bycatch plus discards) with historic data from the 1970s and 4 CPUE (N/1000 yds net/hr) the 1940s to 1950s. Catch-per-unit-effort in Figure 4.
1944 to 1952 was estimated from landings 3 data provided by Walburg and Sykes (1957; 2 Table 2).
2 Assessment Results The gill-net index (Figure 4), the pound net 1
index (Figure 5) and the juvenile abundance index depict strongly increasing trends in relative abundance. Since 2002, estimates of 0 total mortality are declining. Recent total 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Year 116
mortality estimates are within the range of reference values derived from a yield model exercise (Z30= 0.62 to 0.85, depending on the estimate of natural mortality used as input).
16 Benchmarks Potomac River 14 A benchmark for American shad in the Pound Net Bycatch Pounds per net-day Potomac River is the geometric mean of 12 pound net landings reported in Walburg and 10 Sykes (1957) for the years 1944 to 1952 or 8 31.1 pounds per net-day (Figure 6). 6 4
Conclusions Recent total mortality estimates are within the 2 range of reference values derived from a 0 Figure 5.
Chesapeake Bay yield model. From 1944 to 1956, Potomac River landings of American 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 shad were relatively stable, averaging Year approximately 850,000 pounds annually, and ranging from about 500,000 to 1,300,000 pounds. In the late 1970s, total landings of American shad decreased sharply from 120,000 pounds in 1976 to 16,000 pounds in 1980. A moratorium on the taking of shad was established in 1982. The geometric mean of the 1940s to 1950s pound net Potomac River Pound Net Indexes landings is 31.1 pounds per net-day. The Geometric Means geometric mean of the 1970s data is 2.9 pounds 100 per net-day. The geometric mean of the current 90 data is 13.6 pounds per net-day. The mean of the Figure 6.
Pounds per net-day 80 current pound net catch (bycatch plus discard) is 70 Walburg and Sykes 1957 well below the 1950s average (when catches were 60 sustainable), but is greater than the 1970s average 50 (when landings were not sustainable) and is 40 Benchmark increasing. Among Chesapeake Bay stocks of 30 PRFC Discards American shad, the Potomac River population 20 + Bycatch data shows the most promising signs of recovery. 10 PRFC Total Catch 0
1940 1950 1960 1970 1980 1990 2000 2010 Year Literature Cited Walburg, C.H. and J.E. Sykes. 1957. Shad fishery of Chesapeake Bay with special emphasis on the fishery of Virginia. U.S. Fish Wildlife Service, Research Report 48, 26p.
117
Table 1. Estimates of total mortality (Z) of mature American shad in the Potomac River using catch-at-age and repeat spawning data, 2002-2005. Catch-at-age data is based on scale ages.
Year Ages Catch-at-Age Data Repeat Spawning Data 2002 8-Jun 1.31 1.02 2003 9-Jun 1.05 0.92 2004 8-Jun 0.78 0.74 2005 10-Jun 0.82 0.66 Table 2. Historic landings data and CPUE calculated from Walburg and Sykes (1957) for 1944-1952.
Year Effort Virginia Catch Maryland Catch Total Catch CPUE (lbs/net-day) 1944 8,615 670,000 9,041 679,041 78.82 1945 15,413 294,200 8,359 302,559 19.63 1946 11,019 268,000 11,142 279,142 25.33 1947 11,403 992,900 22,697 1,015,597 89.06 1948 16,813 351,200 13,494 364,694 21.69 1949 22,778 356,400 27,055 383,455 16.83 1950 21,367 455,200 20,396 475,596 22.26 1951 13,792 424,000 5,658 429,658 31.15 1952 15,653 451,674 25,636 477,310 30.49 118
Status of American Shad Stocks in Virginia Distribution, Biology and Management Mixed stocks of American shad enter the lower Chesapeake Bay in late winter-early spring and segregate into river-specific populations. Historical in-river landings data in Virginia are from the three primary spawning runs in the Rappahannock, York, and James rivers. Each river population is considered to represent a unit stock with little or no mixing of other stocks in areas upstream of the river mouth. The York River population migrates to either of two adjacent tributaries (the Pamunkey and Mattaponi rivers). American shad spawn in freshwater portions of the rivers, usually beginning in March and ending in June with peaks in April. The annual spawning run consists of virgin fish 3 to 7 years in age (based on analysis of scales) plus repeat spawners (age-4 through age-12). American shad age-9 and older are rare; maximum age recorded is 12 years.
Shad have historically ascended farther upriver than at present within tributaries that are obstructed. Recent construction of the Boshers Dam fish way on the James River and breaching of the Embrey Dam on the Rappahannock River are intended to restore these historic habitats. In spring 1994, the Virginia Department of Game and Inland Fisheries and the U.S. Fish and Wildlife Service began hatchery-restocking efforts in the James and Pamunkey rivers. Adult shad from the Pamunkey River are used as brood stock for the James River releases. In spring 2004, stocking was initiated in the Rappahannock River using adult shad from the Potomac River as brood stock. The success of the restoration program in the James River was evidenced by increasing adult catch rates by monitoring gear in 1998 through 2002 as large numbers of mature hatchery fish returned to the spawning grounds.
The Fishery Figure 1 The Virginia Marine Resources Commercial Landings Commission imposed a moratorium 1,600,000 James on the taking of American shad in 1,400,000 York Virginia rivers and the Chesapeake 1,200,000 Rappahannock Bay in 1994 in response to sharp 1,000,000 declines in commercial landings Pounds 800,000 (Figure 1). The ocean-intercept 600,000 fishery in Virginia coastal waters was closed in December 2004. 400,000 Drift-net fishing by two Native 200,000 American tribal governments and 0 the taking of brood stock by the 1973 1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 Year Virginia Department of Game and Inland Fisheries on the spawning grounds of the York River system for stock restoration in the James River are permitted. An active catch and release recreational fishery exists on the James and Rappahannock rivers and to a lesser extent, the York River (especially the Mattaponi River).
Survey Indices The primary fishery-independent source of data on adult shad abundance is a staked gill net (SGN) monitoring program in each river. When the in-river fishing moratorium was imposed in Virginia, commercial fishermen who held permits for existing SGN stands were allowed to retain priority rights for the locations of those stands. One of these locations on each river (James, York, and Rappahannock) was selected to monitor catch rates in a sentinel fishery. The historic performance of these SGN stands relative to other fishing locations, the amount of fishing effort that would be required to mimic past performance, and the possible influence of fishing activity downstream of the historic locations on catch rates were evaluated by Olney and Hoenig (2001).
119
Three commercial fishermen were contracted to fish these stands two days in each week in the company of a scientist. Catch data from each river are summarized as a standardized catch index (the area under the curve of daily catch rate versus time of year). The longest available time-series of juvenile abundance of American shad results from the Virginia Institute of Marine Science striped bass seine survey (multiple stations in all rivers, 1980-2005).
Figure 2 Assessment Approaches Assessment approaches are (1) evaluating current status by comparing contemporary 20 catch indexes (the area under the curve of York River Kellum's Catch Index daily catch rates versus time in staked gill-net 15 monitoring) to those recorded in voluntary commercial logbooks of fishers in the 1950s and from 1980 to 1993, and (2) examining 10 patterns of juvenile recruitment and hatchery restoration success (prevalence of hatchery- 5 marked mature fish). A yield-per-recruit model using York River stock-specific growth parameters was constructed for the Native 0 1998 2000 2002 2004 2006 American tribal drift gill-net fishery to Year provide a benchmark total mortality rate.
Assessment Results During the seven years of monitoring on the 20 York River, the catch index has been variable James River Brown's Catch Index with higher values (>12) in 1998 and 2001 15 and lower values (<9) in other years. The data suggest a trend towards decreasing catch rates during the period of monitoring (Figure 2). On 10 the James River, catch index values in 2000 through 2005 are higher than those in 1998 5 and 1999 (2.57 and 2.99, respectively). This increase in abundance is due to the first influx of mature hatchery fish into the spawning 0 1998 2000 2002 2004 2006 population (Olney et al. 2003). The data Year depict a trend of increasing catch rates during the period of monitoring (Figure 2). The 2003 and 2004 values of the VIMS catch index on the Rappahannock are higher than any previous year of monitoring. The data depict a 20 Rappahannock Catch Index trend of increasing catch rates since 1998 Rappahannock River (Figure 2). 15 In recent years of VIMS SGN monitoring (2000-2005), mean age of females has 10 increased as a result of lower proportions of age-4 fish in the monitoring catch (Figure 3). 5 On the York River, estimates of total mortality (Table 1) ranged from 0.72 to 1.43 (catch-curve method) and 0.68 to 1.67 (repeat 0
1998 2000 2002 2004 2006 spawning method). On the James River, Z- Year 120
estimates ranged from 0.98 to 1.59 (catch-curve method) and 0.98 to 1.62 (repeat spawning method). On the Rappahannock River, Z-estimates ranged from 0.77 to 1.89 (catch-curve method) and 0.71 to 1.36 (repeat Figure 3 spawning method).
7.0 0.6 Seine survey data on the James River Proportion Age 4 in catch James River Mean Age show no measurable recruitment Mean age of Females 6.5 Proportion Age 4 0.5 during most years (Figure 4). With the 6.0 0.4 exception of 2003 data, juvenile abundance index values are 5.5 0.3 consistently higher on the Mattaponi River than they are on the Pamunkey 5.0 0.2 River and the York River (Figure 4, 4.5 0.1 also Wilhite et al. 2003). In the time series, recruitment is highest (>7.0 on 4.0 0.0 1998 2000 2002 2004 2006 the Mattaponi River and >3.0 on the York River) in 1982, 1984, 1985, Year 1996, and 2003. Years of apparent 7.0 0.6 recruitment failure are 1991, 2001, York River Proportion Age 4 in catch Mean Age Mean Age of Females 6.5 0.5 and 2002. On the Rappahannock Proportion Age 4 River, the highest juvenile abundance 6.0 0.4 index (JAI) values (>0.5) were recorded in 1982, 1989, 2003, and 5.5 0.3 2004 (Figure 4). The Rappahannock 5.0 0.2 River time series suggests recruitment failure in 1980, 1981, 1985, 1988, 4.5 0.1 1991, 1992, 1995, and 2002.
4.0 0.0 1998 2000 2002 2004 2006 Benchmarks Year The benchmark total mortality rate (Z30) ranged from 0.62 to 0.85 depending on the estimate of natural 7.0 Rappahannock River 0.6 Proportion Age 4 in catch mortality used as input (Table 2). A Mean Age Mean Age of Females Proportion Age 4 6.5 0.5 catch rate benchmark of 17.44 (the geometric mean of the catch index 6.0 0.4 values observed in 1953-1957) is 5.5 0.3 appropriate to assess the York River stock since American shad abundance 5.0 0.2 in the 1980s was insufficient to support the fishery. The geometric 4.5 0.1 means of the catch index values 4.0 0.0 observed between 1980 and 1993 on 1998 2000 2002 2004 2006 the James River (6.4) and Year Rappahannock River (1.45) are not benchmarks but are considered interim targets for restoration. Additional studies are needed to relate current catch rates in the James River to historical data that are available in the form of 1950s commercial logbooks. There are no older logbook data available for the Rappahannock River.
121
Figure 4 James River 0.4 0.3 0.2 CPUE (GM) 0.1 0
1980 1985 1990 1995 2000 2005
-0.1
-0.2 Year Rappahannock River 2.5 2
1.5 CPUE (GM) 1 0.5 0
-0.5 1980 1985 1990 1995 2000 2005
-1 Year Mattaponi River 20 15 CPUE (GM) 10 5
0 1980 1985 1990 1995 2000 2005
-5 Year Chart Title 20 15 CPUE (GM) 10 5
0 1980 1985 1990 1995 2000 2005
-5 Year York River 12 10 8
CPUE (GM) 6 4
2 0
1980 1985 1990 1995 2000 2005
-2 Year 122
Conclusions Figure 5 James River (Figure 5): The geometric 35 mean of the historical catch index during Hatchery Prevalence James River 30 the 1980s on the James River is 6.40. The 8.2 % (n= 14) 1998 James Catch Index 3.6 % (n= 7) 1999 average of the current catch index is lower 25 40.3 % (n= 156) 2000 (5.39), indicating that the James River 40.2 %
46.4 %
(n= 103)
(n= 150) 2001 2002 stock has not recovered from the severe 20 51.4% (n= 142) 2003 32.5% (n= 68) 2004 declines in the 1980s and early 1990s. 15 23.8% (n= 40) 2005 Although densities of larval shad are often high on the spawning grounds, there is 10 little evidence of recruitment success on 5
the James River, and the stock is dependent on hatchery inputs. Current 0 estimates of total mortality (Z) for the 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 stock using catch-at-age and repeat Year spawning data (Table 1b) are higher than the estimates of Z30 generated by Figure 6 the yield model. 35 York River York River (Figure 6): The geometric 30 mean of the historical catch index York Catch Index 25 during the 1980s on the York River is 3.22. The average of the current catch 20 index is higher (8.34), indicating some recovery from the severe declines in the 15 1980s and early 1990s. The 1950s data 10 include two years of a high index (26-33), two years of a moderate index (14) 5 and one low index year (8.7, 1955).
VIMS monitoring in 1998 through 2005 0 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 suggests that the York River stock has recovered to a level that is close to its Year abundance during the 1980s. However, the stock level was low during that period and incapable of supporting an active fishery. The York River stock is currently well below the geometric mean of the 1950s data (Figure 7) when abundance of American shad was higher and harvest was apparently sustainable (Nichols and Massmann 1963). Catch indexes have Figure 7 40 been trending downward in recent years.
York River In addition, low juvenile production in 1995 and 1997 through 1999 has York River Index 30 reduced recruitment of young fish to the spawning population in recent years. 20 Proposed RestorationTarget Current estimates of total mortality (Z) M. Green Current for the stock using catch-at-age and repeat spawning data (Table 1a) usually 10 exceed the estimates of Z30 generated by the yield model. 0 R. Kellum Rappahannock River (Figure 8): The 2003 to 2004 values of the VIMS catch 1950 1960 1970 1980 1990 2000 2010 index on the Rappahannock are higher Year 123
Figure 8 than all years in the historical data. The geometric mean of the historical catch index 30 during the 1980s on the Rappahannock Rappahannock Catch Index Rappahannock River River is 1.45. The geometric mean of the 25 current VIMS catch index is higher (3.20).
20 Low juvenile production in 1995 and 1997 through 1999 has resulted in an increase in 15 mean age since fewer young fish are recruiting to the spawning population.
10 Historical data from the 1950s that are 5
directly comparable to the current monitoring location at the mouth of the river 0 are not available. Thus, an interim 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 restoration target for the stock is based on Year the 1980s data. It should be noted that since the catch index for the Rappahannock River is historically lower than the York and James rivers, there is uncertainty about what an appropriate target level should be for this stock. There is little evidence of severe stock decline in the Rappahannock River. On the basis of historical and current catch rates, the present status of the Rappahannock River stock is stable with recent evidence of increasing abundance. Current estimates of total mortality (Z) for the stock using catch-at-age and repeat spawning data (Table 1c) usually exceed the estimates of Z30 generated by the yield model.
Although harvest of American shad in Virginia has been banned since 1994, the index-based assessment suggests that stock abundance remains low relative to historic logbook data, especially in the James and York rivers. The reasons for this slow recovery are unknown but probably include low levels of recent recruitment, unreported removals and discard mortality.
Literature Cited Nichols, P.R., and W.H. Massmann. 1963. Abundance, age and fecundity of shad, York River, Virginia, 1953-1959.
Olney, J.E., and J.M. Hoenig. 2001b. Managing a fishery under moratorium: assessment opportunities for Virginias stocks of American shad (Alosa sapidissima). Fisheries 26(2):6-12.
Olney, J.E., D.A. Hopler, Jr., T.P. Gunther Jr., K.L. Maki, and J.M. Hoenig. 2003. Signs of recovery of American shad, Alosa sapidissima, in the James River, Virginia. American Fisheries Society Special Symposium 35:323-329.
124
Table 1 Estimates of total mortality (Z) of mature American shad in (a) the York River, (b) the James River, and (c) the Rappahannock River calculated using catch-at-age and repeat spawning data.
Catch-at-age data is based on scale ages, 1998-2005. Asterisk indicates scale age data was unavailable.
(a)
Catch-at-Age Repeat Year Ages Data Spawning Data 1998 5-9 1.43 0.97 1999 5-9 1.04 0.99 2000 5-10 0.96 0.95 2001 5-9 1.07 1.17 2002 5-8 1.41 1.67 2003 6-10 1.2 1.07 2004 6-10 1.34 1.2 2005 6-10 0.72 0.68 (b)
Catch-at-age Repeat Year Ages data spawning data 1998 * *
- 1999 5-8 0.98 0.98 2000 5-8 1.31 1.44 2001 5-9 1.35 1.06 2002 5-8 1.59 1.62 2003 5-10 1.09 1.31 2004 5-10 0.98 1.06 2005 6-9 1.17 1.06 (c)
Catch-at-age Repeat Year Ages data spawning data 1998 * *
- 1999 5-8 1.89 0.71 2000 4-8 1.02 0.84 2001 5-9 1.29 1.36 2002 5-9 1.03 0.98 2003 5-9 0.77 0.92 2004 6-10 1.08 1.03 2005 6-10 0.98 0.97 125
Table 2 F30 and Z30 yield model estimates for the York River stock using Hudson River and VIMS partial recruitment vectors.
F30 Z30 Hudson VIMS Hudson VIMS Age M Benchmark Benchmark Recruitment Recruitment Recruitment Recruitment Values Values Vector Vector Vector Vector 12 0.35 0.27 0.27 0.3 0.62 0.62 0.65 10 0.42 0.29 0.29 0.32 0.71 0.71 0.74 8 0.53 0.32 0.33 0.36 0.85 0.86 0.89 126
Figure 1.3-4 Shad rivers assessed in the southeast region.
Roan ok e Rive r
Tar R ive r Albemarle Sound us Ca pe Ne e River e r Pamlico River Fea Riv Riv er Sa v e Dee h Pe na r an Sante Ri ve e River ee r Waccamaw River Og ch e e Little Pee Dee River Riv er Cooper River Alta m aha Edisto River 127 Satilla Ri ve Combahee River River r Saint Marys River in Sa t s
hn Jo ve r Ri 200 100 0 200 km
Status of American Shad Stocks in North Carolina Distribution, Biology and Management American Shad ascend all coastal rivers in North Carolina and are most abundant in Albemarle and Pamlico sounds and the Roanoke, Chowan, Tar-Pamlico, Neuse, Northeast Cape Fear, and Cape Fear rivers (Figure 1). The life history of shad from Albemarle Sound tends to closely mirror those stocks to the north. American shad populations in Fig. 1 North Carolina river systems south of Albemarle Sound form a region where stocks transition from the iteroparous stocks of the Albemarle Sound Roanoke River north to the semelparous stocks seen in the south. American shad are fully mature at ages Tar River Pamlico River Neuse River 7 or 8 in Albemarle Sound, while shad in Neuse River other North Carolina systems reach full Cape Fear River maturity at ages 6 or 7.
Prior to 1987, few limits were placed on commercial fishing (e.g., no mesh size or yardage limits, seasons, or closed areas). In 1988, an area closure was instituted for a 50 25 0 50 Kilometers portion of Albemarle Sound, along with gill -
net mesh restriction in other areas. In 1995, further rules were adopted that made it unlawful to take American shad for commercial purposes by any method from April 15 through January 1. North Carolinas ocean-intercept fishery was closed in 2005. Recreational fishermen can harvest up to 10 shad per person per day by hook and line. 4,500,000 4,000,000 Fig. 2 The Fishery 3,500,000 North Carolina landings of American shad Commercial Landings (kg) 3,000,000 peaked in 1897 at at four million kg and 2,500,000 decreased to 0.7 million kg by 1918 (Figure 2,000,000 2). A second peak of just over 1.4 million kg 1,500,000 was reached in 1928. Landings declined and 1,000,000 stabilized from 1930 to 1970 averaging 404,000 kg. Landings have declined since the 500,000 early 1970s and have remained relatively 0
1880 1888 1896 1904 1912 1920 1928 1936 1944 1952 1960 1968 1976 1984 1992 2000 stable with an average of 128,000 kg from Year 1973 to 2005.
Pound nets were a large component of the commercial harvest in the 1970s but now the vast majority of shad are harvested with gill nets. Landings fluctuate greatly over time, but are currently around the long-term average of 58,263 kg for Albemarle Sound and 13,167 kg for the Neuse River. Current landings are below the long-term averages for the Pamlico River (mean = 10,494 kg) and the Cape Fear River (mean =
13,568 kg). Landings have been declining for the past 3 to 4 years for all systems except the Pamlico River, which remains low but consistent (Table 1).
Indices Commercial gill-net catch-per-unit-effort (CPUE) data for directed shad trips (>100 lbs) were available after the inception of the North Carolina Trip Ticket Program (1994 to 2005; Figure 3). These CPUEs showed an increasing trend for Albemarle Sound but were without trend for the Pamlico, Neuse, and Cape Fear Rivers.
128
Fig. 3 Albemarle Sound Increasing trends were evident in 300 the fishery-independent CPUE CPUE (lb/trip) 250 200 estimates from the NCDMF gill-net 150 survey in Albemarle Sound as well 100 as the NCWRC electrofishing 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 survey in the Roanoke River (Figure 4). No trends were evident Pamlico River 300 from NCWRC electrofishing surveys from any other systems CPUE (lb/trip) 250 200 (Figure 5). Fishery-independent 150 survey indices are not available for 100 any systems other than Albemarle 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 Sound prior to 2000. A summary of Neuse River all indices is given in Table 2.
300 CPUE (lb/trip) 250 Assessment Results 200 For all systems, catch curve 150 100 estimates of total mortality rates (Z) 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 were highly variable and generally ranged from 1 to 2 for both males 300 Cape Fear River and females (Figure 6 Albemarle Sound commercial gill net, Figure 7 CPUE (lb/trip) 250 200 Albemarle Sound commercial and 150 fishery independent gill, 2000-100 2005, and Figure 8 Roanoke River 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 Year fishery independent electrofishing; Tables 3-6; please the North Carolina source document (§13) for figures of Z-estimates from the other systems).
Benchmarks Fig. 4 Catch curve estimates of total Albemarle Sound Fishery Independent Gill Net 2.50 mortality rate were higher than 2.00 the recommended level (Z30) for CPUE 1.50 Albemarle Sound. Adequate data 1.00 0.50 were not available to calculate 0.00 reference points for any other 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 systems.
Fishery Independent Electrofishing 1.50 Summary Albemarle Sound/Roanoke River: 1.00 CPUE Mean statewide commercial 0.50 harvest from Albemarle Sound 0.00 since 1973 has been about three 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 percent of the high reported in 1897. However, it should be Fishery Dependent Gill Net realized that harvest levels in the 2.00 late 1800s are useful as an 1.50 indicator of stock size, but they CPUE 1.00 were likely not sustainable and 0.50 should not be viewed as a goal 0.00 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 for future harvest. Since landings 129
from the Albemarle Sound fishery have made up a significant portion of statewide 350 Roanoke landings since the late 1980s, it is Fig. 5 Tar 300 Neuse reasonable that current abundance in Cape Fear 250 Albemarle Sound and its tributaries is well CPUE (Fish/hr) below the historic potential for these stocks. 200 Current landings are much less than the 150 MSY of 1-2 million kg estimated for these 100 stocks by Hightower et al. (1996). Estimates 50 of Z based on commercial monitoring 0 samples from the Albemarle Sound fishery 2000 2001 2002 2003 2004 2005 suggest that total mortality on stocks of Year Albemarle Sound and its tributaries have generally exceeded the target value since the early 1970s, especially for males. Catch-per-trip in the commercial fishery since the mid 1990s and CPUE from more recent fishery-independent sampling programs have all increased slightly suggesting a recent improvement in stocks of Albemarle Sound and the Roanoke River; although, preliminary estimates from a hydroacoustic survey currently underway indicate that adult abundance is still low (J. Hightower, pers. comm.). High mortality rates may have affected stocks in the 1970s and 3 Figure 6. Males Age 1980s, but a recent stock increase RS suggests that mortality levels have 2.5 Z30 not affected stock levels in the last 2 15 years; however, these improvements may be a result of Z 1.5 artificial enhancement via the 1 ongoing stocking program in the 0.5 Roanoke River. Harvest and presumably stock levels remain 0 very low in the historical context.
1971 1973 1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 Tar-Pamlico: Current status of Year American shad of the Tar-Pamlico Females Age River remains unknown. Landings 3.5 from the Pamlico River were much RS 3
Z30 higher 20 years ago than in recent 2.5 years. Current landings have been 2
less than 10,000 kg since the late Z 1980s; however, we do not know if 1.5 the decline in landings is related to 1 change in effort. Historical data are 0.5 needed to provide perspective on 0 the potential harvest from this system. Estimates of total mortality 1971 1973 1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 have been relatively high since the Year mid-1970s. Gill-net CPUE and total effort have remained low and stable since 1994. Electrofishing CPUE on the spawning grounds, however, has been higher in the Tar than in other North Carolina rivers since 2000, which may be a function of stream size and physical configuration of the sampling sites. Apparently, mortality levels are high enough to keep the stock depressed, but not high enough to lead to stock collapse.
130
Neuse River: Adequate historic harvest data 3.5 Males FD-Age Figure 7.
specific to the Neuse River are not available to 3 FD-RS provide perspective to current landings. 2.5 FI-Age FI-RS Landings displayed several peaks since 1972, 2 Z30 but peaks were progressively lower. Effort Z 1.5 data are not available for the entire time series 1 making it difficult to determine whether 0.5 declining stock size or effort caused the 0 reduction. Years of increased effort generally 1999 2000 2001 2002 2003 2004 2005 2006 corresponded to years of increased harvest Year with the exception of the last few years when Females FD-Age 2.5 effort remained high but catch and CPUE FD-RS declined. The recent decline in CPUE also 2 FI-Age FI-RS corresponded to relatively high estimates of 1.5 Z30 total mortality and a decline in mean length. Z 1
Cape Fear River: Current abundance of 0.5 American shad of the Cape Fear River is 0
unknown as is abundance relative to the 1999 2000 2001 2002 2003 2004 2005 2006 maximum potential for this stock. Adequate Year historic landings are not available for comparison with recent landings. Estimated mortality for Cape Fear shad appear high relative to desired levels. The CPUE from the commercial fishery suggest that stock levels in recent years have been about what it was in the mid-1990s. The CPUE from both the commercial gill-net fishery and fishery-Fig. 8 independent electrofishing suggest a stock Males increase from about 2000 or 2001 through 2004.
2.0 Age Since effort also increased during this time 1.6 RS period, it would appear that recent levels of Z30 fishing mortality have been high enough to keep Z
1.2 the stock from increasing.
0.8 0.4 0.0 2000 2001 2002 2003 2004 2005 Year Females 2.0 Age 1.6 RS Z30 1.2 Z
0.8 0.4 0.0 2000 2001 2002 2003 2004 2005 Year 131
Table 1 North Carolina commercial landings (kg), 1995-2005. Landings were prohibited from April 15 to December 31 of each year.
Cape Atlantic Albemarle Pamlico Pamlico Neuse Year Fear Ocean Sound Sound River River River 1995 46714 27,648 2,373 4,342 6,945 5,071 1996 26385 29,916 4,135 3,934 11,086 12,165 1997 44594 28,844 7,264 5,392 9,891 7,069 1998 53533 76,406 2,272 5,306 5,314 5,055 1999 14955 31,784 2,746 3,139 3,501 3,086 2000 50307 58,779 7,173 9,708 4,182 5,034 2001 5370 43,094 4,440 2,911 4,842 5,708 2002 3800 79,427 6,306 6,792 18,224 8,702 2003 5677 127,320 5,174 7,838 16,379 15,667 2004 3050 82,362 741 7,819 15,257 12,667 2005 101 57,557 2,479 6,667 11,014 7,852 Table 2 Summary of abundance trends by river system. FI = fishery-independent.
System Source Data Year Range Trend Albemarle Commercial gill net 1994-2005 Increasing FI gillnet 1991-2005 Increasing Roanoke FI electrofishing 2000-2005 Increasing Tar-Pamlico Commercial gill net 1994-2005 No Trend FI electrofishing 2000-2005 No Trend Neuse Commercial gill net 1994-2005 No Trend FI electrofishing 2000-2005 No Trend Cape Fear Commercial gill net 1994-2005 No Trend FI electrofishing 2000-2005 No Trend 132
Table 3 Total instantaneous mortality estimates for American shad in the Albemarle Sound, NC.
Age = catch curve from ages, RS = catch curve using repeat spawn marks.
Year Fishery-Dependent Gill Net Fishery-Independent Gill Net M-age F-age M-RS F-RS M-age F-age M-RS F-RS 1972 1.62 1.61 1.15 0.83 1973 1.17 2.46 1.12 1.52 1974 1.43 2.23 2.62 2.13 1975 1.50 1.53 1.83 1.52 1976 2.19 1.80 1.53 1.56 1977 1.71 1.58 2.47 2.53 1978 2.62 1.31 2.79 2.87 1979 1.84 1.89 1980 0.72 1.35 0.81 1.05 1981 0.78 0.57 1.42 0.84 1982 1.50 1.30 1.23 1.06 1983 1.57 1.40 1.41 1.36 1984 0.77 0.80 0.94 1.18 1985 1.12 0.85 1.27 1.71 1986 1.16 0.90 1.31 0.56 1987 1.14 1.39 1.66 1.08 1988 1.14 1.14 1.06 0.98 1989 1.26 1.06 1.53 1.09 1990 1.07 1.63 1.27 0.88 1991 0.84 1.24 1.42 1.29 1992 0.78 1.77 1.10 0.85 1993 1.01 0.73 1.44 1.19 1994 1995 1996 1997 1998 1999 2000 0.79 1.43 0.95 0.71 1.25 1.57 2.30 0.67 2001 0.89 1.81 1.35 2.09 1.10 1.47 0.62 1.56 2002 0.34 2.30 1.90 1.38 2.89 0.29 1.76 1.84 2003 0.89 0.34 0.85 0.66 2.14 0.94 1.43 0.65 2004 0.95 0.94 1.24 1.26 0.86 0.74 1.72 0.62 2005 0.39 1.45 1.15 0.95 0.92 0.49 1.86 1.20 133
Table 4 Total instantaneous mortality estimates for American shad in the Pamlico River, NC.
Age = catch curve from ages, RS = catch curve using repeat spawn marks.
Fishery-Dependent - Gill Net Fishery-Independent - Gill Net Year M-age F-age M-RS F-RS M-age F-age M-RS F-RS 1972 1973 1974 1975 1.49 2.67 2.09 2.82 1976 1.10 0.34 0.35 0.26 1977 1978 3.56 1.76 2.08 1979 0.35 1.72 3.46 1980 0.97 2.03 4.36 1981 1.68 1.99 1.97 3.47 1982 1983 1.70 0.95 0.95 1.35 1984 1.37 0.92 1.67 1.55 1985 0.21 1.15 1.22 2.12 1986 1.65 1.87 1.95 2.25 1987 1.39 0.90 1.66 2.71 1988 1.53 1.32 1.13 1.41 1989-1999 No samples 2000 1.70 1.79 0.69 0.66 2001 0.89 1.81 1.35 2.09 2002 0.66 1.40 1.35 1.48 0.69 0.92 2003 1.32 1.44 1.39 1.41 0.35 1.58 1.61 1.52 2004 0.69 1.98 1.61 1.06 1.32 0.80 0.46 2005 1.10 0.99 1.79 0.93 035 1.39 0.69 134
Table 5 Total instantaneous mortality estimates for American shad in the Neuse River, NC. Age = catch curve from ages, RS = catch curve using repeat spawn marks.
Fishery-Independent Gill Net Fishery-Independent Gill Net Year M-age F-age M-RS F-RS M-age F-age M-RS F-RS 1972 1973 1974 1975 1976 1977 1.04 1.73 1.14 2.00 1978 2.24 2.09 3.03 4.16 1979 1.39 2.06 4.61 3.89 1980 1.45 2.46 1981 1.18 2.46 2.08 2.71 1982 1983 1.34 1.23 1.25 1.88 1984 1.30 0.61 1.19 1.96 1985 2.64 1.84 1.99 3.01 1986 2.08 0.94 1.41 2.43 1987 0.69 1988 1.07 1.90 1.73 1989-1999 No sampling 2000 0.00 1.68 2.48 1.15 2001 1.56 1.79 1.24 1.01 1.10 0.69 2002 0.56 0.15 2.43 1.95 2003 0.28 1.99 2.51 1.73 0.92 2004 0.54 1.43 1.73 1.18 1.10 2.67 1.10 0.47 2005 1.84 0.87 135
Table 6. Total instantaneous mortality estimates for American shad in the Cape Fear River, NC.
Age = catch curve from ages, RS = catch curve using repeat spawn marks.
Fishery-Dependent Gill Net Fishery-Independent Electrofishing Year M-age F-age M-RS F-RS M-age F-age M-RS F-RS 1972 1973 1974 1975 1976 1.27 1.78 1977 1.17 1.63 1978 1.57 1.32 1979 1.87 1.32 1980 0.88 0.94 1981 1982 1983 1.70 1.79 1.22 1.39 1984 1.90 1.17 2.30 1.85 1985 1.61 0.66 2.75 2.22 1986 0.41 0.98 2.20 2.51 1987 0.92 1.39 3.97 1988 2.64 1.18 1.95 2.54 1989-1999 No sampling 2000 0.88 0.89 4.80 1.08 2001 1.59 0.80 1.87 1.58 1.59 1.79 1.87 1.58 2002 0.78 2.33 2.15 0.78 2.33 2.15 2003 0.55 1.99 2.40 1.63 1.15 0.41 1.73 2.08 2004 1.44 0.94 3.61 2.13 1.79 0.69 1.35 2005 0.35 1.82 2.64 1.45 0.59 1.98 136
Status of American Shad Stocks in South Carolina Distribution, Biology and Management Commercial fisheries for American shad occur in Winyah Bay and the Waccamaw, Pee Dee, Black, Santee, Edisto, Combahee, and Savannah rivers, while the Lynches River, Sampit River, Ashepoo River, Ashley River, and Cooper River no longer support large, if any, commercial fisheries (see map of southeast U.S. rivers for overview of South Carolinas major rivers). Recreational fisheries exist in the Cooper, Savannah, Edisto, and Combahee rivers, as well as the Santee River Rediversion Canal. Data were available to assess trends in fishery and stock status of American shad for the following river systems in South Carolina: Winyah Bay and its major tributaries (i.e., Waccamaw and Great Pee Dee rivers), Santee River, Cooper River, Edisto River, Combahee River and Savannah River. Georgia Department of Natural Resources (GADNR) provided additional data for the Savannah River.
The South Carolina Department of Natural Resources (SCDNR) manages its commercial shad fisheries using a combination of seasons, gear restrictions, and catch limits. All licensed fishermen were required to report their daily catch and effort to the SCDNR beginning in 1998. South Carolina has a 10-fish aggregate daily creel limit in recreational fisheries for American and hickory shad in all rivers except the Santee River, which has an aggregate creel limit of 20-fish per day.
American shad returning to South Carolina rivers are generally believed to be semelparous. SCDNR reports that no repeat spawning marks have been observed in their sampling since 2001. Approximately 200 fish were sampled each year from both river and coastal ocean locations; however, a low degree of repeat spawning was reported in 1985 (3% for males and 2% for females).
The Fishery The commercial gill-net fishery targets females. SCDNR has collected landings data by river system since 1979 and instituted mandatory catch and effort reporting in 1998; however, mandatory reporting has not been fully implemented. Questions remain regarding the integrity of the reportsirregular or infrequent fishing by license holders and year-to-year variability in river-wide recordshave not permitted successful development of total catch and effort statistics by river.
Historical commercial shad landings data from the National Marine Fisheries Service (NMFS) are available for South Carolina back to 1880 with the highest reported landings occurring in 1896 (304,819 kg). NMFS port agents compiled landings data until 1979. Landings generally declined from the late 1800s throughout the twentieth century reaching a low in the 1970s (11,022-28,305 kg). Total American shad landings in South Carolina decreased from 191,098 kg in 2004 to 92,194 kg in 2005 (Table 1). From 1979 to the middle 1990s, ocean-intercept landings were typically greater than in-river landings. Since then, the ratio of ocean landings to in-river landings has declined, culminating with the ocean fishery closing in 2005. In 2000, 2,727 commercial shad fishing trips were reported to SCDNR, decreasing to 2,132 trips in 2005. With the ocean-intercept fishery closure in 2005, the Santee River and Winyah Bay are now the largest commercial shad fisheries in South Carolina, with Santee River landings comprising 58% and Winyah Bay landings 38% of the 2005 statewide total.
Winyah Bay (Waccamaw, Great Pee Dee, Little Pee Dee, Lynches, Sampit, and Black rivers): Winyah Bay has supported South Carolinas largest commercial shad fishery in the last few years. Fisheries were centered in the lower 64 km (40 miles) of the Waccamaw River. Both drift and anchored nets are used throughout this area. The Winyah Bay fishery was not included in the closure of the ocean-intercept fishery. Shad gillnetting on the Sampit River was generally limited to the first 16 km of the river above its confluence with the Winyah Bay. Most American shad fishing occurs in the lower 97 km of Black River.
Gillnetting for American shad on the Great Pee Dee River occurs to rkm 105, with activity up to at least 137
rkm 240. Fish migrating to the Little Pee Dee and Lynches Rivers must have successfully by-passed fisheries in the Great Pee Dee before entering their respective tributary streams at approximately rkm 72 and rkm 113, respectively. Shad gillnetting is generally restricted to the lower 32 km of the Little Pee Dee River. The Lynches River shad fishery is prosecuted in the lower 24 km.
Winyah Bay landings averaged 37,695 kg a year since 1979, with a period of below average landings from 1987 to 2000 (Table 1). American shad landings reached 114,104 kg in 1981. Recent peaks in landings came in 2002 and 2004. Winyah Bay landings (32,797 kg) were below average for the time series in 2005. Landings for the major tributaries of Winyah Bay (Waccamaw River and Pee Dee River) are available since 1999. Reported landings for the Pee Dee River averaged 11,935 kg from 1999 to 2005.
Reported landings for the Waccamaw River averaged 11,935 kg from 1999 to 2005. The smaller shad fisheries of other Winyah Bay tributaries have not faired well. By 1960, the Sampit River no longer supported a commercial shad fishery. Shad fisheries in both the Lynches River and Black River have experienced large declines since 1960.
Santee River: The Santee River was historically one of the largest watersheds on the Atlantic coast and supported spawning stocks of American shad as far inland as rkm 483. With the impoundment of the Santee-Cooper lakes in the late 1940s, this system was closed to anadromous fish migrations above rkm 121. This situation persisted until 1985, when the Santee-Cooper Rediversion Canal and fish-lift at St.
Stephen Dam were completed. The fish-lift passes pre-spawning adult shad into the lakes and provides access to historical spawning grounds in portions of the Wateree and Congaree rivers. Since completion of the Rediversion Project, the shad and river herring gill-net fisheries have been restricted to protect the Santee River striped bass population from incidental catches. The entire Rediversion Canal and Santee channel below the Santee dam are closed to gill nets.
American shad landings on the Santee River were 5,183 kg in 1896 and 24,610 kg in 1960. Since 1979, Santee River commercial shad harvest has averaged 42,260 kg a year, with 53,901 kg landed in 2005 (the lowest since 1995; Table 1). Note that the landings include before and after Rediversion landings, where annual harvest averaged 2,554 kg from 1979 to 1985; however, since the completion of the Rediversion Canal in 1985 landings have averaged 56,157 kg a year. Annual fishing effort has averaged 710 trips since 1999. The number of trips in 2005 dropped to 577 from 696 in 2004.
A recreational creel survey conducted by SCDNR in the Santee River before and after completion of the Rediversion Canal showed that total annual effort for all species targeted in recreational fisheries increased by 52.1% in the post-Rediversion survey, while landings increased by 77.5% from the earlier period.
Cooper River: The Cooper River is navigable for approximately 80 km and is impounded by the Pinopolis Dam, which was part of the original Santee-Cooper Project, at approximately rkm 88. The Cooper River likely supported a small shad stock before the creation of the lakes and rerouting of the Santee River. Hook and line is the only legal gear on the Cooper River.
Historically, the Cooper River has not supported a large commercial fishery, with only 823 kg landed in 1896 and 2,859 kg landed in 1960. Commercial landings reported from the Cooper River have been intermittent (Table 1). The Cooper River supports an active recreational fishery below the Pinopolis Dam tailrace in the late winter to early spring. A recreational creel survey conducted by SCDNR on the Cooper River estimated shad catch (in numbers) from 2001 to 2005 to be 3,864, 3,199, 6,856, 5,529, and 14,629.
A recreational creel survey investing all recreational fisheries on the Santee River and Cooper River conducted by SCDNR before (1981 to 1982) and after (1991 to 1993) completion of the Rediversion Canal showed that, although effort increased slightly after Rediversion on the Cooper River, landings of all fish decreased over 50% from the earlier period.
138
Edisto River: The Edisto River is approximately 356 km long and is open to the shad gill-net fishery (both set and drift nets) for its entirety, and continues to support a gill-net fishery to approximately rkm 161. The Edisto River has supported commercial shad fishery for over 100 years and a recreational fishery since the late 1960s. Historically, commercial fishing effort was concentrated between rkm 30 and 50, with gillnetting, bow netting, and hook and line fishing occurring to rkm 170. Sporadic recreational netting extended into the North and South Forks for at least an additional 50 km. Current fisheries occur in the same areas, but at reduced levels of effort.
The magnitude of the Edisto River commercial fishery has declined since 1896 when landings were 58,732 kg. Landings dropped to 15,145 kg in 1960 and from 1979 to 2005 Edisto River commercial shad landings averaged 2,934 kg a year (Table 1). The lowest landings in the time series occurred from 1994 to 1997. Since 2000, landings have averaged 2,211 kg a year with 1,686 kg landed in 2005. This assessment does not account for an unknown recreational harvest.
Combahee River: The Combahee River is approximately 72 km long and both drift and set gill nets are legal. Nearly all activity in the fishery occurs between approximately rkm 40 and rkm 80. There is a very small recreational hook and line (trolling) fishery. Drift nets are rarely, if ever, used on the Combahee River.
In 1896, 6,419 kg of shad were harvested on the Combahee River; landings dropped to 878 kg in 1960.
The Combahee has supported a small fishery that has landed an average of 715 kg shad per year since 1979 (Table 1). Since 1998, landings have been below the time series average, but have been stable, and, in 2005, 403 kg of American shad were landed. 160,000 140,000 Figure 1.
Savannah River: The Savannah River is open to commercial fishing 120,000 SC (kg) with set and drift gill nets up to GA (kg)
Landings (kg) 100,000 about rkm 322. There is a substantial recreational hook and 80,000 line fishery below New Savannah 60,000 Bluff Lock and Dam at Augusta, Georgia. Drift nets are used 40,000 primarily in tidal portions of the 20,000 river, but set nets are the principal gear used throughout the river. -
1964 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 American shad harvested from the Year Savannah River are landed in both Georgia and South Carolina. Landings decreased by an order of magnitude from 1896 (94,074 kg) and 1960 (74,671 kg) to 2005 (9,766 kg). From 1964 to 1979, annual landings in Georgia averaged approximately 30,000 kg. South Carolina shad landings were stable from 1979 to 1987 averaging 16,689 kg per year, but have decreased with landings not exceeding 10,000 kg since 1997 and reaching a time series low of 1,150 kg in 2002 (Table 1; Figure 1). Savannah River shad landings for South Carolina in 2005 were 3,407 kg. Georgia landings have accounted for 71% of Savannah River shad landings when both states have reported landings. A recreational creel survey provides a snapshot of estimated recreational catch and effort for 1999.
139
Indices Overview of Indices: SCDNR identified a group of reliable fishermen who volunteered their catch and effort data on female American shad during investigations into South Carolinas commercial American shad fishery from 1979 to 1985. These records were used to develop fishery-dependent catch-per-unit-effort (CPUE) from 1979 to 2000 and were collected from fishermen from Winyah Bay and Waccamaw, Pee Dee, Santee, Edisto, Combahee, and Savannah rivers. With the initiation of mandatory reporting in 1998, SCDNR decided to use the mandatory reporting records from these reliable fishermen to continue the CPUE time series beginning in 2001 to maintain consistency with data from earlier years.
The CPUE data collected from the reliable fishermen throughout the state are used only to make general observations on changes in perceived stock status since 1979. Many variables, such as water temperature, water levels, flow rates, and fuel and shad prices affect observed CPUE values and these parameters are highly variable between seasons and might have substantial impacts on catchability, and even effort, particularly in certain rivers.
Fish passage at the St. Stephen Dam was monitored by hydroacoustic sampling from 1986 to 1987, real-time human counts from 1988 to 1994, and time-lapse video recording from 1994 to 2005. Since the proportion of Santee River American shad that entered the Rediversion Canal and the efficiency of the Fish Lock both are unknown and appear to vary among years, fish passage at this facility can only be used to document general abundance trends.
Relative exploitation of American shad on the Santee River was estimated using commercial gill-net landings data and fish-lift counts from the St. Stephen fish-lift. A minimum population bound was calculated by summing landings and fishway counts in numbers for each year. This is considered a minimum bound because commercial landings are and do not include recreational removals, and fish passage is unknown. Relative exploitation rates were estimated by dividing catch by the minimum population bound for the year.
The Cooper River supports an active Winyah Bay recreational fishery below the Pinopolis Figure 2.
8.0 Dam tailrace in the late winter to early 7.0 spring. SCDNR has conducted a creel fish/92-m net-hr 6.0 survey from 2001 to 2005 to estimate 5.0 catch, harvest, and CPUE in this 4.0 3.0 recreational fishery. 2.0 1.0 Winyah Bay: Data from the volunteer 0.0 fishermen in the drift gill-net fishery in 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 Winyah Bay produced a continuous Year dataset from 1981 through 1997. Fish moving through the lowermost portion Waccamaw River of this complex system may be bound for any Winyah Bay tributary (i.e., a 6.0 Figure 3.
mixed stock). The CPUE fluctuated 5.0 fish/92-m net-hr without trend throughout the time series 4.0 (Table 2; Figure 2). No data were 3.0 collected from 1998 through 2000. The 2.0 cooperating fisherman who provided 1.0 drift-net catches for this area switched 0.0 to set nets beginning in 2001.
1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 Year 140
Pee Dee River The commercial drift gill-net CPUE in 1.4 Figure 4.
the lower Waccamaw River had a 1.2 fish/92-m net-hr significant increasing slope (Table 2; 1.0 Figure 3). CPUE data for the Pee Dee 0.8 River did not show a significant trend 0.6 0.4 over the years 1979 to 1999 (Table 2; 0.2 Figure 4). 0.0 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 Santee River: CPUE records for 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Year commercial gill net fisheries on the Santee include a lower drift gill-net series and upper set-net series (Figures 5 and 6). Both CPUE series Santee River (Lower) increase with time, but only the lower 3.0 Figure 5.
set-net CPUE trend is significant 2.5 fish/92-m net-hr (Table 2). The lower set gill-net 2.0 fishery consistently increases over the 1.5 whole time series including the years 1.0 before completion of the Rediversion 0.5 Canal. 0.0 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 From 2001 to 2005, the St. Stephen 2006 Year fish-lift passed 965,804 American shad into the Santee-Cooper lake system. Annual American shad passage Santee River (Upper) decreased during this time compared to 8.0 Figure 6.
7.0 the previous six years. Counts peaked fish/92-m net-hr 6.0 5.0 from 1995 through 2000, ranging from 4.0 306,493 to 592,321 shad passed per 3.0 year. Since 2001, annual counts 2.0 1.0 averaged 193,161 shad and in 2005 a 0.0 total 215,438 shad were passed through 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 the fish-lift. Year The relative exploitation estimates, derived from the commercial landings and fish count data, peaked in 1988 and 1989 when minimum population size was at its lowest. This might partially explain catch increases in the lower set nets, if shad were more catchable those two years. Relative exploitation rates decreased to time series lows in 1990, since then both exploitation and the minimum population bound increased through 2000. From 2001 to 2005, commercial gill-net landings and fish-lift counts have decreased (i.e., the minimum population bound), while relative exploitation rates remained near 2000 levels.
Cooper River: Recreational catch-per-man-hour (CPMH) averaged 0.96 and ranged from 0.59 in 2002 to 1.60 in 2005. CPMH increased slightly during the five-year period. The 10-fish per day creel limit has been in effect for the duration of this study. Twenty-two percent of the catch was released in 2005.
Edisto River: Three CPUE time series are available for the Edisto River: the lower 24-hour set gill net, the tide set gill net, and the Jacksonboro set gill neteach has declined (Figures 7-9) with significant negative slopes for both the lower 24-hour set gill net and the Jacsksonboro set gill net datasets (Table 2).
141
Combahee River: CPUE data for the Combahee River set gill-net fishery had no discernable trend (Table 2; Figure 10).
Edisto River (lower set nets)
Figure 7. Savannah River: The SCDNR 2.5 commercial set gill-net CPUE series 2.0 for the Savannah River did not fish/92-m net-hr 1.5 show a discernable trend (Table 2; 1.0 Figure11). GADNR effort data 0.5 show that an average 66.7 drift gill-0.0 net trips per year were made on the 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 Savannah River since 2000. Earlier effort records from GADNR were 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 Year not reliable, because records for individual trips often included Edisto River (tide set nets) records from multiple trips or from 9.0 8.0 Figure 8. partial trips. CPUE was not 7.0 calculated from GADNR catch and fish/92-m net-hr 6.0 trip records for set gill nets because 5.0 4.0 the number of reported trips 3.0 fluctuated inconsistently since 2000 2.0 1.0 (from 1 to 52 trips per year) and the 0.0 1978 1979 1980 1981 1982 1983 time series is short.
1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 Other Studies: SCDNR has also 2003 2004 2005 2006 Year conducted tag-return studies in the Edisto River (Jacksonboro set nets) gill-net fisheries for several rivers.
3.5 Figure 9. These studies rotate among rivers 3.0 and have run 2 to 5 years per river fish/92-m net-hr 2.5 before changing to a different river.
2.0 During these studies, SCDNR has 1.5 collected biological information to 1.0 support other studies (e.g., age, 0.5 repeat spawning, length, and weight 0.0 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 data). The ASMFC American Shad Stock Assessment Subcommittee 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 Year did not use these exploitation rates to evaluate the American shad stock status in South Carolina because no Figure 10.
Combahee River Set Net data were available to determine if 1.2 the following assumptions of tag-1.0 return studies were violated: (1) no fish/92-m net-hr 0.8 tag loss, (2) no tag mortality; (3) all 0.6 recovered tags are reported; (4) age 0.4 and size distribution of tagged fish 0.2 mimics that of the populations; (5) 0.0 tagged fish randomly mix with 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 untagged fish; and (6) all tagged Year fish continue up river and do not stop their migrations after tagging.
142
Juvenile Surveys: Trawl sampling studies were conducted for juvenile American shad in the fall of 1985 in the Edisto River and Winyah Bay. These programs were discontinued after a single sampling season.
Figure 11. Savannah River Set Net Assessment Results 3.5 Winyah Bay
Conclusions:
3.0 When considering the status of American shad in Winyah Bay, fish/92-m net-hr 2.5 2.0 signals of the mixed stock 1.5 within the Bay may be driven 1.0 by the contributions of the 0.5 larger shad fisheries (and 0.0 1978 1979 1980 1981 1982 1983 populations) in the Waccamaw and Pee Dee rivers, potentially 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 Year masking decreases in smaller components of the Winyah Bay shad fishery tributaries (e.g., Lynches, Sampit, and Black rivers). That is, even with no effort in the smaller tributaries, fisheries prosecuted in Winyah Bay may hinder the rebuilding of American shad populations in these rivers. The available evidence suggests that the larger shad stocks in Winyah Bay have remained stable or increased slightly since the late 1970s, but these trends cannot be substantiated for the status of the smaller tributaries.
Santee River
Conclusions:
The Santee River American shad stock appears to have increased since the completion of the Rediversion Canal (Table 2); however, the large decrease in American shad counts at the St. Stephen fish-lift since 2000 is cause for concern. Decreases could indicate a reduction in stock size or a reduction in the proportion of fish entering the diversion canal and the lift. Moreover, age and size distribution of shad in the Santee declined since the diversion, which could indicate increased mortality. If annual exploitation rates increase, then there could be cause to closely monitor this American shad stock, especially if effort is redirected to the Santee River from the closed ocean-intercept fishery.
Cooper River
Conclusions:
The recreational creel survey provides a short time series of harvest and catch rate estimates (2001-2005) and provides evidence that the river has been able to sustain the current level recreational harvest. The recreational fishery appears to be healthy and the increasing catch-per-man-hour (CPMH) observed in the creel survey may indicate that the population can withstand current harvest levels without any decline; however, additional years of the survey will be need to confirm these observations.
Edisto River
Conclusions:
Recent estimates of commercial CPUE have declined in all three available time series, significantly for two of them (Table 2), and landings have been below the time series (1979-2005) average for 13 of the last 15 years. Given the low landings and declining commercial set gill-net CPUE, continued harvest of Edisto River American shad could prolong the recovery of this stock or lead to continued stock declines.
Combahee River
Conclusions:
This relatively small river is perceived to have undergone significant American shad stock declines over the past 25 years. The Ashepoo River, an even smaller sister river to the Combahee and Edisto, has apparently followed a similar trend in stock status and is no longer known to support a gill-net fishery. No CPUE data are available for the Ashepoo, and the commercial set-net CPUE for Combahee River American shad showed no significant trend since 1993 (Table 2). Continued harvest of American shad on the Combahee River could reduce the chance of recovery of spawning run and prolong the perceived depleted status of shad on this river.
143
Savannah River
Conclusions:
Over the past century, the magnitude of shad landings from the Savannah River has declined by an order of magnitude. The commercial set-net CPUE data available since 1979 indicates some stability in the current harvest levels (Table 2).
Other Rivers: There is a little or no information on the shad stocks of the Lynches, Black, Sampit, Wateree, Catawba, Wando, Ashley, Ashepoo, and Coosawhatchie rivers and Bull Creek that provide insights on stock status. Since these stocks are perceived as small, removals from these stocks could prolong or prevent successful rebuilding of these stocks. Threats to small stocks harvested within the Winyah Bay mixed stock fishery are discussed above.
Benchmarks No benchmarks were established for evaluating American shad status in South Carolinas rivers.
Summary Available information on the river specific stocks shows different states of stock status. The largest stocks (Winyah Bay (mixed stock), Waccamaw River, and Santee River) appear stable or to have increased since the late 1970s; however, in the case of the Winyah Bay mixed stock this cannot be substantiated for its smaller tributaries (Lynches River, Black River and Sampit River). The Santee River fishery and stock appears to have increased since completion of the Rediversion project. Recent decreases in fish-lift counts, declines in mean age and size, and the potential for increased effort suggests that continued monitoring of this stock is warranted to ensure the long-term health of the Santee River stock. The Cooper River appears to be able to support current recreational harvest levels, and continuing the creel survey to monitor this fishery is advised. On the Savannah River, there has been no trend in commercial CPUE since the late 1970s.
The decline in commercial CPUE and low landings on the Edisto River provides evidence of a declining, if not depleted, American shad stock. Similarly, the Combahee River American shad stock is likely depressed, but the only available catch and effort index for the stock begins in 1993. There is a lack of information on the American shad stocks of other rivers in South Carolina. These stocks are perceived as small and removals from these stocks could prolong or prevent successful rebuilding of these stocks.
144
Table 1. American Shad landings by river in South Carolina since 1996.
Winyah Bay Santee Cooper Edisto Combahee Savannah River Ocean- Statewide Year Total In-river Complex River River River River (SC) Intercept (All Waters) 1996 9275.5 126952.7 NR 1032.7 NR 8479.5 145740.5 100929.1 246669.5 1997 17024.5 80195.5 126.4 353.6 NR 6827.7 104527.7 51507.3 156035.0 1998 12238.2 150188.6 170.9 3603.6 573.6 2876.4 169651.4 28193.6 197845.0 1999 8889.5 65531.4 NR 1739.1 831.8 4409.5 81401.4 4595.5 85996.8 2000 37331.8 162386.8 NR 3340.9 667.3 2873.2 206600.0 47243.6 253843.6 2001 57288.2 83265.0 NR 1525.5 127.7 3379.5 145585.9 41391.4 186977.3 2002 85501.8 77650.9 NR 2038.6 401.8 1150.0 166743.2 46711.4 213454.5 2003 65074.1 69701.4 NR 1874.1 304.1 1842.7 138796.4 15011.4 153807.7 2004 77167.3 85662.3 5.0 2800.0 392.3 4185.0 170211.8 20886.4 191098.2 2005 32796.8 53901.4 NR 1686.4 402.7 3407.3 92194.5 0.0 92194.5 Table 2. Summary of abundance trends in from commercial gill net CPUE in South Carolina by river.
System Dataset Gear Years Mixed Stock* Slope P-value R-square Conclusion No discernable trend in Winyah Bay Drift Gill Net 1981-1997 Yes 0.060 0.390 0.050 dataset Waccamaw River Drift Gill Net 1979-2005 No 0.095 <0.001 0.476 Significant positive slope No discernable trend in Pee Dee River Set Gill Net 1979-1999 No -0.006 0.580 0.017 145 dataset 1979-180,1986, 1992- No discernable trend in Santee River Upper Santee Set Gill Net No 0.160 0.137 0.175 2005 dataset Lower Santee Drift Gill Net 1980-2003 No 0.049 0.004 0.329 Significant positive slope Edisto River Lower Set Gill Net 1980-2004 No -0.035 0.040 0.186 Significant negative slope 1979-1982,1985,1986,1988- No discernable trend in Tide Set Net Set Gill Net No -0.062 0.169 0.097 1994, dataset 1996,1997,1999-2000 Jacksonboro Set Gill Net 1980-2005 No -0.040 0.041 0.184 Significant negative slope No discernable trend in Combahee River Set Gill Net 1989, 1993-2004** No 0.012 0.683 0.019 dataset No discernable trend in Savannah River Set Gill Net 1980-2005 No -0.010 0.632 0.011 dataset
- Mixed stock: shad caught in Winyah Bay could be bound for any of its tributaries or perhaps to other river systems.
- Trend for 1993 to 2004
Status of American Shad Stocks in Georgia Distribution, Biology and Management The Georgia Department of Natural Resources (GADNR) manages American shad by river system. The Altamaha, Ogeechee, and Savannah rivers each support commercial American shad fisheries (see map of southeast U.S. rivers). There are small shad runs in the Satilla and St. Marys rivers, and no commercial landings have been reported from these rivers since the 1980s. In addition to commercial landings data, GADNR actively monitors American shad fisheries on the Altamaha and Ogeechee rivers. Monitoring of the Savannah River American shad stock was assigned to South Carolina by the Atlantic States Marine Fisheries Commission (ASMFC) and the Savannah River assessment is included in Chapter 14.
Commercial set and drift gill-net fisheries are conducted in the Altamaha, Ogeechee, and Savannah rivers, with regulations including mesh size and lead length restrictions, and closed season. The Altamaha River has the largest commercial shad harvest in Georgia followed by the Savannah River and then the Ogeechee River. Before 1980, a shad license was required to fish commercially for American shad and hickory shad in Georgia, since then commercial shad fishing has been covered by a general commercial fishing license. The commercial shad season is open each year from January 1 to March 31. Each river has weekly closures in effect. Set and drift gill nets are the only legal commercial gear for American shad in Georgia. Set gill nets were banned in the lower Savannah River in 1990. All sturgeon, catfish, and non-shad gamefish must be immediately released. There is no commercial shad fishing on the St. Marys River, due to the Florida net ban enacted in 1995.
Recreational fisheries currently exist in the Ogeechee and Savannah rivers, with a creel limit in both rivers of 8 fish per angler. Shad are infrequently caught as bycatch in Altamaha River recreational fishery.
Historically, the Ogeechee River had the largest recreational fishery in the state. However, in recent years the recreational fishery on the Savannah River, according to anecdotal evidence, has become larger than that on the Ogeechee River.
Shad in Georgia are believed to be semelparous. There are no records of repeat spawning marks in 1,311 male specimens and 2,452 female specimens examined by GADNR that were collected from the 1960s to recent years.
The Fishery Studies in the late 1970s and the 1980s of the commercial fishery for American shad documented substantial underreporting, as shad that were sold to inland dealers or kept for personal use were not included in National Marine Fisheries Service landings data. Mandatory catch and effort reporting was instituted in 2000. Before mandatory reporting, catch and effort data were required, but were not always collected due to difficulties in enforcement and reported individual trips were often summaries of multiple trips or represented partial trips. Underreported landings and a short time series of mandatory effort reporting, limit the extent that commercial fishery data can be used in this assessment.
Statewide records of American shad landings in Georgia date back to 1880 and peaked in the early 1900s. In the 1960s, landings reached a second peak of consistently high landings, sustaining average annual harvests of nearly 200,000 kg. Since the 1970s, landings have consistently declined to current low levels. An all-time low of 11,579 kg of American shad was landed in 2002 and 18,071 kg were landed in 2005. Landings from 1996 to present for each river are in Table 1.
Reported landings of American shad in the Altamaha River peaked in 1968 at 213,963 kg and then declined steadily to the early 1980s. From 1983 through 1988, landings average 122,150 kg before declining to an average of 44,675 kg from 1989 to 1994. Landings increased briefly from 1995 to 1998, averaging 106,646 146
kg annually and then declined to a mean of 30,535 kg a year from 1999 to 2005. Landings in 2005 were 25,653 kg. A roving commercial gill-net survey was conducted by GADNR from 1982 to 1991 to collect catch and effort for the entire Altamaha River shad fishery. Catch and effort data were available for the commercial drift gill-net fishery since 2000.
A total of 94,689 kg of shad was harvested from the Ogeechee River in 1896, all of which was harvested using drift gill nets. Landings averaged 1,483 kg per year from 1989 to 1997, with a low of 122 kg in 1992.
Since 1998, landings averaged 268 kg annually, with lows of 17 kg in 2003 and 69 kg in 2005. Recreational harvest of American shad and hickory shad in the Ogeechee River was estimated through an access creel survey conducted in 1996 (1,239 fish), 2000 (295 fish), and 2005 (442 fish).
Historically, harvests greater than 10,000 kg per year occurred on both the Satilla and St. Marys rivers; however, no commercial harvest has been reported from either river since the late 1980s.
3.50 Indices Figure 1 An area under the curve index (AUC) of 3.00 y=-0.0143x2 + 0.3229x - 0.0184 daily catch (numbers) of drift gill nets Area Under Curve 2.50 R2 = 0.3381 per net (ft)-hour. for Altamaha River 2.00 spawning runs of American shad was 1.50 developed for each season since 1986 using data collected by GADNR during 1.00 collections for their tag-return study 0.50 (Figure 1). The AUC approach was 0.00 utilized because it accounts for both the 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 magnitude of daily catch rates and season Year duration.
The tag-return study on the Altamaha has been used by GADNR to develop estimates of exploitation and population size. The ASMFC American Shad Stock Assessment Subcommittee (SASC) did not use these exploitation rates to evaluate the Altamaha River American shad stock status because no data were available to determine if assumptions of tag-return studies were violated: (1) no tag loss, (2) no tag mortality; (3) all recovered tags are reported; (4) age and size distribution of tagged fish mimics that of the populations; (5) tagged fish randomly mix with untagged fish; and (6) all tagged fish continue up river and do not stop their migrations after tagging.
GADNR conducts a recreational creel survey on the Ogeechee River every five years to estimate the harvest and effort. There are no monitoring efforts on either the Satilla River or St. Marys River.
A juvenile survey was conducted from 1982 to 1991 in the Altamaha, but was discontinued because the juvenile abundance index did not relate to the subsequent adult spawning stock or the parent stock that produced it. A juvenile survey was conducted from 1982 to 1985 on the Ogeechee River.
Assessment Results The lone CPUE series spanning sufficient time to provide insight of recent population trends was from the GADNR tag-return study collections and was used to develop an area under the curve estimate of the seasonal catch in numbers per net (ft)-hour. A rise in the area under the curve index begins in 1989 and peaks in the mid to late 1990s before decreasing to low levels since 1999 (Figure 1). Other indices are of short duration (recent commercial drift gill-net CPUE) or were terminated in the early 1990s (roving commercial gill-net survey and did not have significant trends. In summary, the Altamaha River American shad fishery and stock are at depressed levels compared to 1960s and earlier. Data are insufficient to quantitatively assess the Ogeechee, Satilla, and St. Marys rivers.
147
Benchmarks No benchmarks were developed by the SASC for Georgia rivers.
Summary The Altamaha River American shad fishery and stock are at depressed levels compared to the 1960s and earlier. The AUC index indicates a brief increase in the stock from 1995 through 1998, before decreasing to current low levels. The American shad fishery on the Altamaha River and other Georgia rivers is believed to be declining. Until signs of increased abundance are evident, either from the decline in commercial effort within the river and from the closure of the ocean-intercept fishery, harvest of American shad should not be increased on the Altamaha River. Developing a benchmark based on a desirable level of the AUC index might be challenging, since the fishery is declining and the long history of underreporting harvest.
Increased harvest of American shad on the Ogeechee River, and harvest Satilla and St. Marys rivers, is not warranted until usable indicators of stock status can be developed to guide further management activities.
Continued harvest of shad on Georgias shad producing rivers may prolong successful rebuilding of their natal shad runs.
Table 1. American Shad landings by river in Georgia since 1996.
Total Altamaha Savannah Ogeechee St. Marys Year Satilla River Landings River River River River 1996 120,781 50,750 14,859 1,789 1997 106,484 42,763 11,680 1,058 1998 148,540 44,733 52,557 350 1999 50,752 14,383 4,983 442 2000 66,029 18,730 4,657 434 2001 60,704 18,150 5,750 351 2002 30,813 8,491 1,729 351 2003 38,798 11,916 1,780 17 2004 29,416 13,648 3,410 134 2005 41,529 11,636 6,366 69 Table 2. Summary of abundance trends for the Altamaha River in Georgia.
R-System Index Gears Years Slope P-value Conclusion square Nonlinear trend:
a= -0.0143, a= 0.0121, b=
Altamaha Drift Gill 1986 to Increases to mid-AUC b= 0.3229, 0.0091, c= 0.3381 River Net 2005 1990s and then c= -0.0184 0.9708 declines to present Roving Set and Non-significant 1982 to Commercial Drift Gill 0.080 0.899 0.002 positive slope and 1991 Creel Survey Net poor fit Non-significant Commercial Drift Gill 2000 to 4.107 0.446 0.151 positive slope and CPUE Net 2005 poor fit 148
Status of American Shad Stocks in Florida Distribution, Biology and Management American shad occur in the St. Johns, Nassau, and St. Marys rivers of northeast Florida, and each are considered a separate stock. Only the St. Johns River has sufficient data to conduct a stock assessment and it is defined as a separate 82°0'0"W 81°0'0"W management unit for American Jacksonville 0km 0mi American shad !
Electrofishing sites shad. Adults are present during by year 2001-2002 (
the winter and spring (
! 2002-2003 (November to May) and Orange Park §
¨ ( 2003-2004
! 295
( 2004-2005 juveniles occur from the spring Black StJohnsRiver_Route !
to autumn and into the winter 30°0'0"N Creek 50mi Index Grid 30°0'0"N in some years. Florida landings Kilometers 10 5 0 10 20 Miles of shad expanded rapidly in the 100km 10 5 0 10 20 late 1800s, peaked early in the 1900s, and declined markedly Gainesville during the twentieth century Palatka (McBride 2000). Available Dunns Creek evidence supports that 100mi Crescent ATLANTIC Lake American shad native to Florida Ocklawaha River OCEAN rivers are semelparous. Lake George Floridas commercial shad Ocala 200km Lake Dexter landings were severely reduced §
¨ 95 De Land after January 1, 1992, by a 29°0'0"N 29°0'0"N regulation to increase mesh 150mi Lake size, as well as adding net- Monroe Lake tending and soak-time §
¨ Wekiva River Lake Jesup 75 300km Harney !
regulations to the ocean- Puzzle Lake intercept shad fishery (Williams Econlockhatchee 200 mi 32°0'0"N Locator Map River 1996). An amendment to AL Orlando GA Atlantic Floridas Constitution, which Ocean
§
¨ 4 30°0'0"N took effect in July 1995, Cocoa prohibited the use of entangling 28°0'0"N Lake Poinsett nets within 3 miles of the Gulf Of Mexico FL 400 km 250 mi !
Atlantic coast, effectively 26°0'0"N Lake Washington eliminating shad fishing within Melbourne 24°0'0"N state waters. Since January 1, §
¨ 28°0'0"N 4
28°0'0"N Lake Hellen Blazes 1997, hook and line has been 86°0'0"W 84°0'0"W 82°0'0"W 82°0'0"W 80°0'0"W 81°0'0"W the only allowable fishing gear for American shad, Alabama shad, and hickory shad in the state, and it has been unlawful to possess more than 10 fish of any combination of these species. Historical regulations are covered in detail in Chapter 16.
The St. Johns has only a 9.1 m elevation change over its entire length (499 km). There is one dam on the main stem of the St. Johns River, at Lake Washington, and its tributary, the Oklawaha River, is dammed.
The primary concerns regarding habitat on the St. Johns River are the competing demands for water resources between human development and wildlife and fisheries needs.
149
The Fishery Florida's shad landings peaked in the late 1800s at 1 to 3 million pounds and fluctuated between 200,000 and 900,000 pounds from the 1920s to the 1960s. Landings have declined further, from less than 200,000 pounds in the early 1970s to zero in recent years. Floridas landings of shad dropped dramatically in the 1990-1991 fishing year, continued to drop during the 1990s, and no landings have been reported since 2000.
In the late 1800s, Floridas shad were caught Ocean Landings Total Landings primarily in drifting gill nets, secondarily in Fishing Year (pounds) (pounds) haul seines, and thirdly in anchored or staked 1986-1987 142,026 155,430 gill nets. By the 1950s, most landings of American shad were made by haul seine 1987-1988 266,251 266,374 followed by gill nets. Haul seining was 1988-1989 164,839 165,112 discontinued during the early 1970s in the St.
Johns River, and gill nets were used into the 1989-1990 169,881 289,293 1990s. Commercial shad fishing grounds have 1990-1991 58,810 71,592 also shifted geographically. In the 1950s, the 1991-1992 49,633 49,798 dominant mode of harvest was by set gill nets in the lower river and by haul seine in the middle 1992-1993 24,503 24,503 river (near Palatka, rkm 127) and, by the early 1993-1994 24,930 24,968 1990s, nearly all the shad harvested came from 1994-1995 26,791 26,886 gill nets fished in coastal waters offshore of Mayport, Florida. The Marine Recreational 1995-1996 3,650 3,650 Fisheries Statistics Survey does not appear to 1996-1997 54 54 intersect with the American shad recreational 1997-1998 18 18 fishery on the St. Johns River, because the fishery is concentrated well upstream. Today, 1998-1999 480 480 Floridas shad fishery is composed primarily of 1999-2000 800 800 recreational anglers fishing on the spawning grounds and most anglers practice catch and 2000-2005 0 0 release.
Indices A roving creel survey was conducted to monitor catches of American and hickory shad during 11 years of the 13-year period from 1992 to 2005. Angler interviews were completed together with instantaneous counts of the number of anglers along an 11.9 km stretch of the river between the mouth of Lake Jessup and the north end of Lake Harney. In average or poor fishing years fishing success was typically below 1.0 fish per angler hour, while in better than average years fishing success was typically above 1.0 fish per angler hour.
Anglers within the creel survey area caught from 1,260 (2004-2005) to 12,592 (1998-1999) American and hickory shad each year and averaged 5,879 shad per year (+ 3,676 s.d.; n=11 years). Catch and release is commonly practiced in this fishery, with 79% of the catch released in 2001-2002, 77% in 2002-2003, 71%
in 2003-2004, and 79% in 2004-2005.
An electrofishing survey was completed during a 4-year period, 2001 to 2005, to determine spawning seasonality and distribution, and to generate independent estimates of spawner abundance to compare to the creel survey. The electrofishing survey covered a broad range of shad spawning locations and months. Effort was allocated broadly along the St. Johns River and within its major tributaries, covering several key areas of the shad spawning grounds and bracketing the full spawning period for American shad.
150
Assessment Approach This assessment examines catch and effort data, using a falsification approach to test the following hypotheses:
- 1. Ho: Fishing success (i.e., angler catch rates) is not related to fisheries-independent (i.e.,
electrofishing) estimates of abundance. This hypothesis will be tested by correlating fishing success (ratio-of-mean estimates from the creel survey) and geometric mean abundance (shad collected by fisheries-independent electrofishing), in order to validate that creel survey catch rate estimates are related to true shad abundance.
- 2. Ho: The annual time-series trend for shad abundance is not different than zero. This hypothesis will be tested with regression analysis of the creel survey abundance time series for the period 1992 through 2005.
- 3. Ho: American shad mean sizes are not different over time. There is a limited amount of data available for this, and although it is not rigorously tested here, we look to see if fork lengths are similar between different decades.
FWC Shad Creel, Periods 3 -- 8
- 4. Ho: Sex ratio of American shad is not different over time. There is a 2.0 Fishing success (fish/hr) y = -0.0092x + 19.2 limited amount of data available for r = 0.14 ns this, and although it is not rigorously 1.5 tested here, we look to see if sex ratios are similar to 50:50 between 1.0 different decades.
0.5 Assessment Results 0.0 There was a statistically significant and 1990 1995 2000 2005 positive relationship between shad Spaw ning year abundance measured by electrofishing versus that measured by a roving creel survey. This was observed when the creel survey estimates were correlated to electrofishing in either the creel areas (i.e., Lake Monroe to Lake Harney; r=0.41; n=39; P ~ 0.01) or an adjacent, upstream section (Lake Harney to Puzzle Lake; r=0.69, n=16; P<0.01), demonstrating that the time series of annual measures of angler catch rates can be used as a proxy for Creel Area Efishing Estimates Puzzle Lake Efishing Estimates population size. r = 0.41**, n = 39 3.5 r = 0.69**, n = 16 2.5 3.0 Efishing rate (shad/hr) Efishing rate (shad/hr)
The time series of angler 2.0 2.5 catch rates does not show 1.5 2.0 a statistically significant 1.5 trend over time, 1993 to 1.0 1.0 2005. These relatively 0.5 stable estimates of catch 0.5 rate have occurred during 0.0 0.0 0.0 0.5 1.0 1.5 2.0 2.5 0.0 0.5 1.0 1.5 2.0 2.5 a period of generally Fishing success (shad/hr) Fishing success (shad/hr) declining effort in the recreational shad fishery.
Todays (2001-2005) male American shad were about five percent shorter and the females were about eight percent shorter than 50 years ago.
151
The proportions of females, relative to females and males combined, have changed notably in the last several decades. During the complete 1957- 1958 spawning run, females were only slightly more abundant than males (n=63,692; prop. females=0.53). During recent (2001-2005) spawning runs, females were considerably less abundant compared to males throughout the year. Across all years, males dominated the catch (n=1,786; prop. females=0.36).
C_6 FWC Shad Creel, Periods 3 -- 8 Benchmarks E_6 The benchmark used in this stock assessment is Mean catch (n), effort (hr) 2000 derived from catch and effort data. Specifically, a sustained catch rate greater than 1.0 fish per 1500 angler hour during the years 2001 to 2005 was proposed by McBride (1999a) as the initial 1000 restoration goal. Sustained angler catch rates 500 greater than 1.0 fish per hour was considered as a criteria for accepting that population size had 0 increased because of netting regulations. 1990 1995 2000 2005 Spaw ning year Summary This report describes stable catch rates during a period of declining fishing effort for the shad recreational fishery in Floridas St. Johns River. Also, the average size of shad and the proportions of females have declined markedly during the last several decades. In general, these features do not describe a desirable status for the fishery or for its rebuilding. American shad in Floridas St. Johns River are at historically depressed levels, and at best, can be described as at low but currently stable population sizes.
152
1.4 CONCLUSION
S AND RECOMMENDATIONS 1.4.1 Conclusions This section contains conclusions based on observations from the overview of the coast-wide summary materials presented in this introductory section. Stock specific conclusions are contained in individual state sections (Sections 2-16) and in Section 1.4 - Stock Assessment Summaries.
- Ocean mixed stock harvest has been a large component of total American shad harvest over the last 25 years and since the late 1980s it was the dominant component of shad harvest from north of Virginia. Some segments of the American shad harvest remain unknown (e.g., inland commercial, recreational harvest, under-reporting for each segment, and unreported bycatch),
likely resulting in underestimated landings. However, reported landings still provide valuable insight into the general magnitude and trends of harvest along the U.S. Atlantic coast. The expected benefits resulting from the ocean intercept fishery closure were not obvious in this assessment and might take one or more generations of American shad before they are realized.
- Available total mortality estimates generally exceeded Z30 for most years in rivers where data were suitable for catch curve analysis and where data supported SSBPR modeling. There is some evidence in these data, in conjunction with other data summarized in this assessment, that these Z values have affected characteristics of some stocks.
- Data on annual number of fish passing upriver at dams on several Atlantic coastal rivers exhibited a coast-wide pattern of an increase followed by a decrease. Interestingly, most fish passage numbers declined at about the same time (late 1990s to early 2000s). This synchronous decline suggests a coast wide change in environmental conditions or mortality factors that affected stocks from South Carolina to Maine within the last five years.
- Continuous fishery dependent and independent catch-per-unit-effort series generally only provide insight into recent stock dynamics, except for the Delaware River Lewis haul seine index.
- Trends in juvenile production do not show consistent patterns coast-wide; however, regional patterns and some local trends were noteworthy:
o Recruitment has increased in the upper Chesapeake Bay, including the Potomac River, and Merrymeeting Bay, Maine in recent years.
o Recruitment patterns in the lower Chesapeake Bay (James, York, and Rappahannock rivers) and in Albemarle Sound have been similar.
o Relatively low young-of-year production was observed in all New England juvenile surveys in 1998 and 2001 o There has been consistent low recruitment in the Hudson River since 2002 1.4.2 Recommendations While some of the recommendations made in this assessment are specific or applicable to one system only, the recommendations listed below are applicable coastwide. They were developed as themes emerged in the assessment of all 31 rivers. Here we classify them according to whether they relate to fisheries and fishery assessments or to habitat. Please note that they are not prioritized.
153
Recommendations for Fisheries and Fishery Assessments:
- 1. Do not increase directed fisheries for American shad.
- 2. Restrict fisheries operating on stocks where total mortality is increasing and relative abundance is decreasing.
- 3. Identify all fisheries where bycatch occurs, then quantify the amount and disposition of bycatch.
In fisheries where bycatch is allowed, quantify the discards.
- 4. Employ observer coverage to verify the reporting rate of commercial catch and harvest as well as bycatch and discards.
- 5. Identify directed harvest and bycatch losses of American shad in ocean and bay waters of Atlantic Maritime Canada.
- 6. Employ microchemistry techniques to identify stock composition in mixed stock harvest.
- 8. Collect annual estimates of recreational catch, total harvest, CPUE, age, size, and sex composition of fish in each fishery.
- 9. Do not continue in-river tagging programs (conducted in Georgia, South Carolina, and Maryland) used to estimate exploitation and population size unless methods to identify reporting rate, tag mortality and loss, and movement (fallback), which are needed to estimate exploitation, are developed.
- 10. Continue tagging using Brownie-type models to estimate survival.
- 11. Monitor juvenile production in semelparous stocks. Such monitoring may indicate when recruitment failure has occurred.
- 12. Mark stocked larvae with OTC marks that allow age and year-class identification in mature fish.
This is critical for verification of various aging techniques.
- 13. Characterize passage-associated efficiency, mortality, migration delay, and sub-lethal effects on American shad at hydroelectric dams.
- 14. Annually update all summary data tables of on-going data collection for use in the next assessment in the format used in this stock assessment for use in ASMFC stock assessments only.
Recommendations for Habitat
- 1. Develop safe, timely and effective upriver and downriver passage for adults and downriver passage for juvenile at all barriers within spawning reaches.
- 2. Maintain water quality and suitable habitat for all life stages of American shad in all rivers with shad populations. Refer to Amendment 1 for habitat issues pertaining to American shad and the ASMFC Anadromous Species Habitat Source Document (in prep).
- 3. In rivers with flow regulation, maintain flows at levels that ensure adequate fish passage, water quality, and habitat protection.
- 4. All rivers systems assessed in this document should have shad management (e.g., recovery and restoration) plans. Review and update these plans on a regular basis.
154
1.5 CAUSES OF DECLINE: HYPOTHESES Results of data summarized elsewhere in this document suggest declines in many Atlantic coastal stocks of American shad, especially in the mid-Atlantic and southern New England states. Many causes for this decline have been suggested. They include bycatch in ocean fisheries, former mixed stock harvest from coastal and estuarine locations, increased predation on juvenile and adult shad by piscine and mammalian predators, and losses from down river passage at dams and hydroelectric facilities. This section explores two of these potential causes: losses from mixed stock harvest and predation. Papers included in this section used creative analyses and complex models to provide coastwide and stock specific perspectives on these two hypotheses. This section does not provide definitive answers, but does set the stage for crafting future analyses and investigations.
1.5.1 Coastwide Perspective Papers in this section include:
A. American Shad Stock Contributions in Mixed Stock Fisheries Along the Atlantic Coastan Update B. Relative Rates of Exploitation in Three Atlantic Coastal American Shad Stocks C. Striped Bass Predation on Adult American Shad: Occurrence and Observed Effects on American Shad Abundance in Atlantic Coastal Rivers and Estuaries A. American Shad Stock Contributions in Mixed Stock Fisheries Along the Atlantic Coastan Update Authors:
Kathryn A. Hattala and Andrew W. Kahnle Hudson River Fisheries Unit, New York State Department of Environmental Conservation 21 S. Putt Corners Road, New Paltz, New York 12561 Introduction In Section 1.1.7, we summarized available data on commercial in-river and ocean and mixed stock (OMS) fishery landings of American shad along the U.S. Atlantic coast. Clearly, ocean harvest was large and cannot be ignored in any assessment of American shad stock change over the last 20 years. To be most useful, ocean harvest needs to be partitioned to source stock. Available techniques for partitioning harvest include otolith chemistry, tag release recapture analyses, and DNA methodology. Otolith chemical signature analysis is a new and promising technique (Walther et al., in press), but no data on ocean shad stock composition have been produced. Published and unpublished tag release recapture studies of American shad have been conducted at several locations along the U.S. Atlantic coast with tag release locations in both spawning estuaries and in mixed stock locations along the coast. There have also been a few DNA studies of stock composition of the OMS harvest. Both tag release recapture and DNA studies provide some insight on stocks that might have contributed to the mixed stock harvest. Tag release recapture studies were available from the late 1950s through the present. DNA studies were conducted in 1991-1995. A few studies of ocean harvest stock composition are available from the early 1990s. Hattala et al. (1998) attempted to partition ocean landings to stock using available tag release - recapture and DNA data available at the time. We updated that study through 2005 using additional and more current tag recapture data.
155
Methods We used tag recapture data from studies that released tagged fish in mixed stock areas and DNA data sampled from mixed stock fisheries harvest to develop estimates of stock composition in these fisheries.
We partitioned the mixed stock fisheries into regions such that each region had either a tag release location or a DNA study. Regions were also based on timing of fisheries along the coast and the stocks that were potentially affected by them. Percent recapture from given states or rivers (stock) were then applied to total annual mixed stock harvest within region. Stock composition within region remained constant among years because tag-recapture studies only occurred at discrete intervals through the time period.
Regions Mixed stock landings came from Section 1.1.7. Landings were grouped into the following regions:
southern (South Carolina and North Carolina), mid-Atlantic (Virginia to Maryland), Delaware Bay, the New Jersey Coast, and northern (New York through New England; Table A1). Percent contributions of the affected stocks were developed as follows:
North and South Carolina: Stock and state percent compositions in this region were derived from two tagging studies conducted in ocean waters off North Carolina (Parker 1992; Supplemental Table A11) and South Carolina (McCord 1986-1988; Supplemental Table A2). Estimates of percent recaptured by state or river were averaged among years and between studies Maryland and Virginia: MtDNA studies were used to apportion harvest in this region (Supplemental Tables A3 and A4). The DNA data were used rather than tagging data because the tagging study conducted in ocean waters off Maryland and Virginia (Jesien 1992; Supplemental Table A5) was limited by access to fishing vessels through the season. Subsequent tag releases were not distributed evenly over the fishing season.
For the 1980-1988 period, percentages in Table A1 represent an average of three sample areas over two years: the 1992 and 1993 harvests from Virginia fisheries off Rudee Inlet and Wachapreague and the 1993 Ocean City, Maryland harvest of coastal shad. For 1989 to present, percent composition is based on average stock composition determined for the 1992 and 1993 Wachapreague fishery and 1992 Ocean City fishery (three sample areas).
Delaware Bay and New Jersey Coast: Results from the most recent tagging study, conducted in 1995 through 2005 in lower Delaware Bay, were used for this region (Supplemental Table A6). We refined the estimates for the Delaware and Hudson stocks for this assessment by readjusting the harvest that was assigned to the Delaware-New Jersey portion of the mixed stock fishery made by Hattala et al. (1998).
This newer version (as compared to the 1998 version) segregates Delaware ocean and Delaware Bay landings from that of New Jersey ocean landings. We allowed a higher percentage of New Jersey ocean harvest to be attributed to the Hudson stock (Table A1) since shad caught along the coast were assumed to be headed north, and not south to the Delaware or other southern rivers. This agrees with the migratory pattern described by Dadswell et al. (1987). We recognize that the resulting attributed harvests may be biased higher but they provide a reasonable upper bound on estimated harvest.
New York and New England: The only data available for this region were from early studies conducted in the New York Bight (Talbot and Sykes 1958, summarized in Dadswell et al. 1987, Supplemental Table A7).
1 Supplemental Tables immediately follow Section 1.5.1 tables.
156
Apportionment to Affected Stocks The apportionment of mixed stock fishery landings occurred in a step-wise progression. Percents attributed to affected stocks were calculated from tagging or mtDNA studies, as described above (Table A1). Landings (Table A2) of American shad include those from mixed stock fisheries only (see Section 1.1.7). Annual landings from each states fishery were totaled within region (i.e., South Carolina and North Carolina were added to form the SC-NC group; Table A3). The affected stocks (or groups) estimated portion was calculated by multiplying each percent listed in the column under each affected stock (Table A1) by the landings harvested by each of the five groups (Table A3). The total harvest for an affected stock was the sum of the regional harvest estimates attributed to that stock (Table A4).
Landings in the earlier 1998 version were adjusted for underreporting of 50 percent. We chose not to adjust landings for this assessment since the 1998 under-reporting rate was a best guess. No study on reporting rate in the mixed stock fishery has ever been conducted.
Results Losses to the mixed stock fishery tended to be highest for stocks in North Carolina, South Carolina, the Delaware River, the Hudson River, and the Connecticut River. Losses to rivers in Chesapeake Bay and New England were much smaller.
Discussion There were several problems associated with the methodology used in this section. However results do provide a general overview of stock, or state, specific losses to the mixed stock fishery. Better estimates will have to await data from additional tagging, DNA, or other (otolith micro-chemistry) studies from mixed stock areas. Although the ocean-intercept fishery closed in 2005, the issue remains relevant because ocean bycatch continues. A summary of problems with the methodology used in the present analysis follows.
Within regional stock composition, percents were kept constant within and among years. This is probably an unrealistic assumption. Stock composition in a given area likely varies seasonally as shad from different stocks migrate through the area. Changes in stock size may also have occurred for some or all stocks for the period 1980 to the present and such changes would have altered stock composition among years. Stock composition of harvest probably also changed over time as fisheries altered timing and location of operations.
Estimates of stock composition were also affected by the studies used to partition harvest. DNA stock identification studies on American shad are still in their infancy and need work on verification and refinement. Tag-recapture methods were affected by poorly known recapture effort, unknown reporting rates, and small recapture sample sizes. However, use of tag release recapture data remains valuable because data from many studies are available and two studies have been ongoing for over ten years.
Moreover, tag-recapture locations and times are generally reliable.
Conclusion We did not preclude the usefulness of past tagging or DNA studies to examine the composition of the mixed stock fishery. We use the data only to attempt a first order approximation of potential fishing pressure of the mixed stock fisheries on a particular stock or region.
157
In the Relative Exploitation discussion, we illustrate the possible impact of the estimated harvest on individual stocks.
Literature Cited Brown, B.L., and J.M. Epifano. 1995. Mixed stock analysis of American shad in Virginias and Marylands coastal intercept fisheries. Report to Virginia Marine Resources Commission, Newport News, Virginia.
Dadswell, M.J., G.D. Melvin, P.J. Williams, and D.E. Themelis. 1987. Influences of origin, life history and chance on the Atlantic coast migration of American shad. In Common strategies of anadromous and catadromous fishes. American Fisheries Society, Symposium #1. Bethesda, Maryland.
Hattala, K., R. Allen, N. Lazar and R. OReilly 1998. Stock contributions for American shad landings in mixed stock fisheries along the Atlantic coast. IN Atlantic States Marine Fisheries Commission 1998. American shad Stock Assessment Peer Review Report. Washington DC USA.
Jesien, R., S. Gaichas, J. Serafy, and C. Hocutt. 1992. Stock composition of coastal Maryland and Virginia American shad fishery: 1992 tagging study. Report to Maryland Dept. Of Natural Resources, Annapolis.
McCord, W. 1988, 1987, 1986. Monitoring and assessment of South Carolinas commercial shad fishery and incidental catch of sturgeons. Completion reports to National Marine Fisheries Service.
Parker, J.A. 1992. Migratory patterns and exploitation of American shad in the near shore ocean of southeastern North Carolina. N.A. Jour. of Fish. Management 12(4):752-759.
Talbot, G.B., and J.E. Sykes. 1958. Atlantic coast migrations of American shad. Fishery Bulletin, U. S.
58:472-490.
Walther, B., S. Thorrold and J. Olney. In press. Geochemical signatures in otoliths record natal origins of American shad. Transactions of the American Fisheries Society.
158
Table 1.5.1-A1 Estimated American shad stock contributions (percent) in the mixed stock Atlantic coast fisheries, based on DNA and tag-recapture programs.
Affected Stocks Mixed Stock Fishery Area Delaware Hudson Connecticut New England -
FL GA SC NC VA MD R. R. R. Canada SC-NC 1.00% 7.20% 54.30% 37.50%
VA-MD (1980-1988) 1.00% 23.60% 20.20% 16.40% 2.60% 11.50% 3.40% 6.40% 14.60%
VA-MD (1989-present) 1.00% 31.70% 26.70% 9.30% 4.30% 11.00% 11.00% 3.00% 3.00%
Delaware Bay 2.10% 39.00% 42.00% 15.00% 2.10%
NJ Coast 13.00% 70.00% 15.00% 2.00%
NY-NE 0.50% 0.50% 1.00% 6.00% 50.00% 42.00%
159
Table 1.5.1 -A2 American shad landings reported for mixed stock fisheries (kg*1000; Sources: NMFS and state reports). Note: inland landings are not included (see Section 1.1.7).
Year SC NC VA MD DE NJ NY RI MA NH ME 1980 70.3 1.8 43.5 0.0 40.7 54.3 51.5 0.9 3.8 3.1 12.7 1981 67.8 48.7 125.0 0.0 83.7 59.3 26.4 14.2 7.6 2.5 41.1 1982 111.2 29.0 125.6 0.0 148.5 126.9 33.4 36.0 13.3 1.2 11.7 1983 93.2 1.7 94.2 9.1 98.8 88.9 15.0 10.7 6.1 1.5 17.6 1984 178.2 6.1 292.3 8.7 93.3 94.8 15.2 16.6 13.5 2.3 15.2 1985 62.4 1.4 150.7 68.1 78.2 97.6 42.5 41.2 10.1 3.3 7.3 1986 102.3 28.6 161.3 57.3 96.1 71.2 33.1 23.8 27.3 7.7 10.4 1987 163.1 18.7 179.3 54.1 111.6 83.7 5.3 47.1 18.5 18.7 12.0 1988 117.2 22.7 194.5 120.0 132.0 105.9 7.0 55.3 23.0 20.8 14.5 1989 103.5 17.5 181.3 221.3 98.3 181.1 10.3 19.0 6.2 13.9 21.1 1990 73.2 16.8 147.5 128.7 215.4 203.8 2.4 10.3 2.5 17.3 5.4 160 1991 65.4 8.7 181.3 106.1 231.5 174.3 11.9 12.6 0.3 8.6 0.9 1992 49.5 10.9 196.0 90.2 124.3 129.6 9.6 6.0 0.1 4.5 0.7 1993 29.5 12.8 220.8 35.3 134.3 145.2 3.5 18.4 0.2 3.0 0.0 1994 32.6 15.4 171.4 15.3 102.5 98.4 2.7 8.1 0.1 12.8 0.5 1995 60.0 46.7 68.5 27.5 86.7 127.0 6.6 12.7 0.2 13.9 0.2 1996 100.7 26.4 108.0 60.2 120.2 95.3 11.5 6.5 0.1 16.1 0.5 1997 51.4 44.6 162.3 75.4 74.2 102.3 17.1 16.7 0.3 11.5 0.1 1998 30.6 53.5 157.6 54.1 94.6 105.7 41.4 15.2 0.8 6.9 0.2 1999 8.8 15.0 105.0 27.7 102.7 118.4 31.1 20.1 0.1 1.7 0.1 2000 40.3 50.3 74.5 16.3 75.4 97.0 11.6 7.9 0.1 2.7 0.1 2001 41.9 5.4 114.6 26.0 127.9 91.2 14.2 30.8 0.5 0.4 0.2 2002 38.3 3.8 69.1 18.7 55.2 110.5 27.6 39.6 0.2 0.0 0.0 2003 16.8 5.7 23.4 4.9 40.0 68.8 11.1 17.5 0.5 0.0 0.0 2004 20.8 3.0 25.1 3.8 78.9 57.0 6.8 6.7 0.0 0.0 0.0 2005 0.0 0.1 0.0 0.0 55.8 20.0 2.0 0.0 0.0 0.0 0.1
Table 1.5.1-A3 Landings (kg*1000) from Table 1.5.1-2 summed by area grouping (see Table 1.5.1-1).
Year SC-NC VA-MD Delaware Bay NJ Ocean NY-NE 1980 72.1 43.5 63.7 31.3 72.0 1981 116.6 125.0 114.4 28.6 91.9 1982 140.2 125.6 208.8 66.6 95.7 1983 94.9 103.3 121.1 66.5 50.8 1984 184.4 300.9 112.3 75.8 62.8 1985 63.8 218.7 100.4 75.4 104.4 1986 130.9 218.5 125.1 42.2 102.2 1987 181.8 233.4 161.2 34.1 101.6 1988 139.9 314.6 168.7 69.3 120.6 1989 121.0 402.6 125.7 153.6 70.5 1990 90.0 276.2 312.8 106.4 38.0 1991 74.1 287.4 299.7 106.2 34.3 1992 60.4 286.2 176.1 77.8 20.9 1993 42.2 256.1 190.3 89.2 25.1 1994 48.0 186.6 121.2 79.6 24.3 1995 106.7 95.9 114.6 99.0 33.5 1996 127.1 168.2 128.2 87.3 34.6 1997 96.0 237.7 89.8 86.6 45.7 1998 84.1 211.7 101.1 99.2 64.5 1999 23.7 132.7 140.3 80.8 53.0 2000 90.7 90.8 110.8 61.5 22.4 2001 47.3 140.6 140.1 78.9 46.0 2002 42.1 87.8 62.3 103.4 67.4 2003 22.5 28.2 42.9 66.0 29.2 2004 23.9 28.9 81.4 54.6 13.5 2005 0.1 0.0 74.6 1.2 2.0 161
Table 1.5.1-A4 State stock estimates of harvest by mixed stock fisheries (kg; Table 1.5.1-3 landings
- Percent in Table 1.5.1-1). Estimates are not adjusted for under-reporting.
Year FL GA SC NC VA MD Delaware R. Hudson R. Connecticut R. New England - Canada 1980 1.2 5.2 49.4 35.8 7.5 2.8 34.6 54.5 53.0 38.6 1981 2.4 8.4 92.8 69.0 21.0 6.1 63.6 77.8 75.4 59.8 1982 2.7 10.1 105.8 78.0 21.1 8.1 105.5 144.3 97.2 64.2 1983 2.0 6.8 75.9 56.5 17.2 5.5 68.3 104.0 60.2 40.3 1984 4.9 13.3 171.1 129.9 49.7 10.5 88.9 114.2 78.9 74.2 1985 2.8 4.6 86.3 68.1 36.4 8.3 75.2 108.7 92.6 79.4 1986 3.5 9.4 122.6 93.2 36.4 8.8 80.4 95.7 90.2 78.3 1987 4.2 13.1 153.8 115.3 38.8 10.0 95.2 105.6 95.0 80.8 1988 4.5 10.1 150.2 116.0 52.2 12.3 112.2 137.3 116.1 101.5 1989 5.2 8.7 193.3 152.9 37.8 20.3 114.0 208.9 89.2 47.4 1990 3.7 6.5 136.4 107.5 25.9 18.6 166.6 238.5 90.2 32.9 1991 3.6 5.3 131.4 104.5 26.9 18.8 162.6 233.9 86.7 31.5 162 1992 3.5 4.3 123.5 99.1 26.7 16.1 110.5 161.2 57.1 22.6 1993 3.0 3.0 104.1 84.2 23.9 15.1 114.2 172.0 62.1 24.0 1994 2.3 3.5 85.2 67.8 17.5 10.7 78.4 128.6 47.9 19.9 1995 2.0 7.7 88.4 65.6 9.1 6.7 68.5 130.0 51.7 21.3 1996 3.0 9.2 122.3 92.6 15.8 10.1 80.2 135.5 54.7 24.0 1997 3.3 6.9 127.5 99.5 22.3 12.3 72.9 127.2 56.5 30.0 1998 3.0 6.1 112.8 88.1 20.0 11.5 76.3 139.1 68.7 37.6 1999 1.6 1.7 55.0 44.3 12.6 8.9 80.4 133.3 63.7 30.8 2000 1.8 6.5 78.0 58.2 8.6 6.3 61.4 101.0 39.8 15.7 2001 1.9 3.4 70.3 55.3 13.3 9.2 80.8 132.3 60.1 28.1 2002 1.3 3.0 50.7 39.2 8.5 5.4 48.1 112.2 61.2 34.3 2003 0.5 1.6 21.2 16.0 2.8 2.3 28.7 69.1 31.8 15.3 2004 0.5 1.7 22.1 16.7 2.8 3.0 42.1 76.4 28.0 9.4 2005 0.0 0.0 0.1 0.0 0.0 1.6 29.3 32.3 12.4 2.5
Supplemental Table A1 Tagging studies conducted off the coast of North Carolina.
Study: J. Parker 1992, Ocean Tagging Study - Southeast NC Release Location: Southeast NC* S.E. NC (Wrightsville) TOTAL Release Period: January-March December 1989-April Year: 1989 1990 N-caught: 320 N-tagged: 301
% tagged: 94%
River & State(s) Returns % Returned Returns % Returned Returns % Returned Group 1989 1990 Average NC - Ocean 7 17.10% 6 37.50% 13 22.80% NC 73.20% 75.00% 74.10%
NC - Albemarle Sound NC- Neuse River NC- Cape Fear River 23 56.10% 6 37.50% 29 50.90%
163 SC - Waccamaw-Pee Dee R. 8 19.50% 3 18.80% 11 19.30% SC 24.40% 18.80% 21.60%
SC - Santee R.
SC - Edisto R. 2 4.90% 2 3.50%
GA - Savannah-Ogeechee R. 1 2.40% 1 6.30% 2 3.50% GA 2.40% 6.30% 4.30%
Total returns 41 16 57 0.18937
% of releases 13.60% 5.30% 18.90%
Additional information:
Mean TL: 484.9 494.9 Gill Net Mesh Size (in): 5.5 Percent-at-age:
4 15.6 4.4 5 76.6 53.2 6 7.8 one repeat 42.4
- Southeast NC = south of Pamlico Sound and near the Cape Fear River.
Supplemental Table A2 Tagging indices conducted off the coast of South Carolina.
Study: W. McCord, SC Ocean Tagging Release Location: SC Ocean SC Ocean SC Ocean SC Ocean Release Period:
Year: 1986 1987 1988 All years N-tagged: 485 24.70% 125 31.20% 275 24.70% 885 25.60%
River & State(s) Returns % Returned Returns % Returned Returns % Returned Returns % Returned Groups % Returned NC- Cape Fear R. 1 1.50% 1 0.40%
NC - Ocean 1 0.80% 1 0.40% NC 0.90%
SC - Ocean 4 3.30% 5 12.80% 4 5.90% 13 5.70%
SC - Waccamaw-Pee Dee R. 82 68.30% 25 64.10% 37 54.40% 144 63.40%
SC - Georgetown 1 2.60% 1 1.50% 2 0.90%
SC - Santee R. 9 7.50% 9 13.20% 18 7.90%
164 SC - Cooper R. 1 0.80% 1 1.50% 2 0.90%
SC - Edisto R. 4 3.30% 1 2.60% 3 4.40% 8 3.50%
SC - Savannah R. 6 5.00% 1 2.60% 3 4.40% 10 4.40% SC 86.80%
GA - Ogeechee R. 2 1.70% 2 0.90%
GA - Altamaha R. 11 9.20% 6 15.40% 4 5.90% 21 9.30% GA 10.10%
FL - St. Johns R. 4 5.90% 4 1.80% FL 1.80%
Unknown 1 1.50% 1 0.40%
Total Returns 120 39 68 227 99.60%
Supplemental Table A3 Summaries of genetic DNA studies (presented as percent of sample from river origin) conducted off the coasts of Virginia and Maryland (Brown and Epifano 1994) and New Jersey (1996).
Study: Brown & Epifano 1994 (Table 11)
Sampling Location: Rudee Inlet, VA Wachpreague, VA Ocean City, MD Year: 1992* 1993 1992 1993 1992 River of Origin: % % mean % % mean %
St. Lawrence R. 30 0 15.0 0 0 0.0 Shubenacadie R. 0 0 0.0 7 0 3.5 St. John R. 23 2 12.5 0 11 5.5 Connecticut R. 17 6 11.5 3 6 4.5 Hudson R. 0 0 0.0 14 3 8.5 Delaware R. 4 15 9.5 0 12 6.0 23 Susquehanna R. 0 0 0.0 13 0 6.5 Rappahannock R. 10 8 9.0 0 0 0.0 Pamunkey R. 0 7 3.5 0 0 0.0 York R. 10 0 5.0 0 11 5.5 0 James R. 4 15 9.5 0 17 8.5 Chowan R. 0 0 0.0 1 35 18.0 44 Cape Fear R. 0 21 10.5 0 0 0.0 Santee R. 0 23 11.5 9 0 4.5 0 Savannah R. 0 0 0.0 49 5 27.0 32 St. John's R. (FL) 2 0 1.0 3 0 1.5 0 Summary of Composition:
Study: Brown & Epifano 1994 (Table 11) Brown 1996 Year: 1992-1993 1992-1993 1992 1991 1994 1995 Sampling Location: Rudee Wachpreague Ocean City Virginia New Jersey Coast River of Origin: % % % % % % mean Canadian 27.5 9.0 21 47 34.0 Connecticut R. 11.5 4.5 32 Hudson R. 0.0 8.5 21 22 21.5 Delaware R. 9.5 6.0 23 9 Susquehanna R.** 0.0 6.5 31 55 31 43.0 VA Rivers 27.0 14.0 2 3 1.5 NC Rivers 10.5 18.0 44 5 Santee R. 11.5 4.5 19 Savannah R. 0.0 27.0 32 St. Johns (FL) 1.0 1.5 2
%- max. likelihood estimates of proportion
- trace amounts of other systems: Columbia (of Hudson origin in 1870s) and Nanticoke
- *NOTE: most shad returning to the Susquehanna River are of Hudson River or Delaware River parent broodstock.
165
Supplemental Table A4 Estimated stock composition (% by state/area) of American shad harvest for the 1992 and 1993 ocean-intercept fisheries off Maryland and Virginia, based on mtDNA studies (Brown and Epifano 1995).
Time Period State/Area Affected 1980-1988 1989-1996 Florida 1.0 1.0 South Carolina 23.6 31.7 North Carolina 20.3 26.7 Viginia rivers 16.4 9.3 Maryland rivers 2.6 4.3 Delaware River 11.5 11.7 Hudson River 3.5 5.7 Connecticut R. 6.4 3.0 Canadian rivers 14.6 6.0 166
Supplemental Table A5 Tagging studies conducted off the coasts of Maryland and Virginia.
Study Jesien (1992 Report to State of MD) Ocean Tagging Study, MD & VA Release Location: Rudee Inlet, VA Ocean City, MD Rudee Inlet, VA TOTAL MD-VA TOTAL MD-VA Release Period: February-March March-April March-April Year: 1991 1992 1992 1992 1991 & 1992 N-caught: 3199 648 3127 3775 6974 N-tagged: 569 260 430 690 1259
% Tagged: 18% 40% 14% 18% 18%
River & State(s) Returns % Returns Returns % Returns Returns % Returns 1992 Returns % Returns 1991-1992 Returns % Returns MA - Merrimack R. 1 5.00% 1 2.60% 1 1.72%
CT, MA - Connecticut R. 1 5.30% 1 5.30% 1 2.60% 2 3.45%
NY,NJ - Hudson R. 2 10.50% 2 3.45%
PA, DE, NJ - Delaware R. 2 10.50% 9 47.40% 2 10.00% 11 28.20% 13 22.41%
DE, NJ - Delaware Bay 2 10.50% 2 5.10% 2 3.45%
DE - Ocean 1 5.30% 1 2.60% 1 1.72%
NJ - Ocean 1 5.00% 1 2.60% 1 1.72%
MD - Chesapeake Bay 1 5.30% 1 1.72%
VA - Chesapeake Bay rivers 9 47.40% 2 10.00% 2 5.10% 11 18.97%
VA - Ocean 1 5.30% 4 21.10% 5 25.00% 9 23.10% 10 17.24%
NC - Ocean 3 15.80% 1 5.30% 1 2.60% 4 6.90%
NC - Albemarle Sound 3 15.00% 3 7.70% 3 5.17%
NC- Neuse R. 1 5.00% 1 2.60% 1 1.72%
NC - Cape Fear R. 2 10.00% 2 5.10% 2 3.45%
SC - Waccamaw-Pee Dee R. 1 5.30% 1 2.60% 1 1.72%
SC - Santee R. 1 5.00% 1 2.60% 1 1.72%
GA - Savannah-Ogeechee R. 2 10.00% 2 5.10% 2 3.45%
Total Returns 19 19 20 39 58
% of Releases 3.30% 7.30% 4.70% 5.70% 4.60%
Additional Information:
Mean TL 513 541 517 Gill net mesh 5.5 6 5.5
Supplemental Table A6 American shad tagging studies conducted in lower Delaware Bay and the Hudson River.
Study Hattala & Allen, Cooperative American Shad Tagging Program Time Period: 1995-present. Data summarized and reported in annual ASMFC reports.
Recaptures by Release Location Releases Lower Delaware Bay N % Hudson R. NY N % Year Lower DE Bay Hudson R.
Ocean 49 20.70% Delaware Bay 5 3.00% 1995 107 2648 CT - Connecticut R. 34 14.30% NY - Hudson R. 148 89.20% 1996 294 1579 Delaware Bay 56 23.60% NS - Ocean 1 0.60% 1997 509 2046 DE, NJ, PA - Delware R. 38 16.00% RI - Ocean 1 0.60% 1998 555 249 NY - Hudson R. 44 18.60% MD - Ocean 2 1.20% 1999 764 413 NC - Cape Fear R. 1 0.40% NJ - Ocean 5 3.00% 2000 430 493 Chesapeake Bay 2 0.80% VA - Ocean 3 1.80% 2001 663 542 DE - Indian R. 1 0.40% NJ - Raritan Bay 1 0.60% 2002 273 66 PA - Lehigh R. 1 0.40% 2003 170 604 168 RI - Pawcatuck R. 1 0.40% 2004 51 766 NB - Petitcodiac R. 1 0.40% 2005 220 519 SC - Santee R. 1 0.40%
NB - Shubenacadie R. 1 0.40%
QB - St.Lawrence R. 1 0.40%
MD - Susquehanna R. 5 2.10%
MA - Westfield R. 1 0.40%
Total 237 166
Supplemental Table A7 American shad tagging studies summarized by Dadswell et al. (1987), conducted in the Connecticut River, Hudson River, New York Bight, Delaware River, and Chesapeake Bay.
Studies: Talbot and Sykes 1958 Miller 1982 Leggett (unpublished)
Release Location: Hudson R/NY Bight Rudee Inlet/Ches. Bay Delaware Connecticut N-tagged: 11579 4775 2920 18374 River & State(s) Returns % Returns Returns % Returns Returns % Returns Returns % Returns CAN - St. Lawrence R. 2 1.50%
CAN - Miramichi R. 4 3.00%
CAN - Bay of Fundy 13 0.70% 3 2.60% 18 13.30%
Gulf of Maine 9 0.50% 5 4.30% 3 2.20%
MA - Cape Cod 7 0.40% 4 3.50%
MA - Merrimack 1 0.70%
CT, MA - Connecticut R. 146 7.90% 21 18.30% 36 26.70% 311 83.40%
NY - Ocean 23 1.30% 5 4.30% 2 1.50% 19 5.10%
NY, NJ - Hudson R. 1582 86.10% 31 27.00% 25 18.50%
169 PA, DE, NJ - Delaware R. 45 33.30%
DE, NJ - Delaware Bay 20 1.10% 9 7.80% 2 0.50%
DE - Ocean 1 0.70%
NJ - Ocean 13 0.70% 5 4.30% 4 1.10%
MD - Chesapeake Bay 10 0.50%
VA - Chesapeake Bay rivers 7 0.40% 2 0.50%
VA - Ocean 1 0.70% 10 2.70%
NC - Ocean 5 0.30% 14 12.20% 3 2.20% 23 6.20%
NC - Rivers 2 0.10% 16 13.90% 1 0.30%
SC - Rivers 1 0.90%
GA - Rivers 1 0.90% 1 0.30%
Total Returns 1837 15.90% 115 135 373
% of Releases 15.90% 2.40% 4.60% 2.00%
Aggregation Areas: Summer - off Fundy, St. Lawrence estuary, off Newfoundland and Labrador Winter - off Florida, Middle Atlantic Bight, Scotian Shelf, and Gulf of Maine
B. Relative Rates of Exploitation in Three Atlantic Coastal American Shad Stocks Authors:
Andrew W. Kahnle and Kathryn A. Hattala Hudson River Fisheries Unit, New York State Department of Environmental Conservation 21 S. Putt Corners Road, New Paltz, New York 12561 Introduction Rates of total mortality have increased in several American shad stocks along the U.S. Atlantic coast in recent years (Section 1.2.3). In many of these stocks, concurrent declines in abundance, mean age, and mean size have occurred. Understanding the causes of increased mortality is important to managing stocks currently in decline. Total mortality reflects the combined impact of fisheries, spawning mortality, predation, and mortality associated with downstream passage at hydroelectric dams. Recent interest has focused on effects of predation on American shad stocks (Section 1.5, Savoy and Crecco 2004). Earlier interest focused on effects of fishing and in particular, effects from the growing harvest of American shad from ocean waters (ASMFC 1998).
Commercial harvest of American shad occurs within natal rivers and in various mixed stock fisheries in ocean waters and bays. Harvest from ocean waters has been a growing component of Atlantic coastal shad harvest over the last 25 years. Growth of this fishery was most pronounced in the region from Virginia through Maine. Ocean harvest in this region increased substantially from 1970 through the late 1980s and early 1990s (Section 1.1.7). It then declined through 2005 when directed harvest from ocean waters was closed. Natal river harvest in this region declined through the entire time series. Ocean harvest exceeded natal river harvest from the late 1980s through 2004.
In this section, we use trends in relative exploitation to evaluate the possibility that increased mortality observed in coastal shad stocks was caused by increased fishing mortality. We confined our analyses to American shad of the Delaware, Hudson, and Connecticut rivers because stocks in these rivers have been well studied for many years and all of them show some effects of the rise in mortality. Moreover, most of the ocean harvest from these three stocks occurred in the region from Maine through Virginia where the mixed stock harvest was large (Section 1.1.7).
Methods We calculated relative exploitation rates as total annual harvest (kg) divided by river specific annual indices of adult abundance. Exploitation rates were developed from in-river harvest alone and from in-river harvest plus that part of the mixed stock ocean harvest attributed to the stock (termed total stock harvest). Stock composition of the ocean harvest was developed in Section 1.5.1. River specific adult abundance indices and estimates of total mortality came from river specific assessments in this volume Specifics of the adult indices used are given in Table B1.
Results Annual patterns of relative exploitation differed somewhat among stocks.
Delaware River: Relative exploitation rates for the Delaware River stock based on total stock harvest increased gradually from 1970 through 2005 (Table B2; Figure B1). CPUE in the Delaware River commercial gill net fishery declined from 1990 through the present. Relative exploitation rates for in-river harvest (i.e., not including any mixed stock harvest) are given in Sections 1.5.2 and 8.
170
Hudson River: Relative exploitation rates from in-river harvest suggested a peak of relatively high exploitation in 1980 followed by a decline and then fluctuation at low values through the present (Table B3; Figure B2). Rates from total stock harvest suggested the same peak in 1980, a low period through the rest of the 1980s, an increase in 1990, followed by fluctuation at a relatively high level from 1990 through the present (Table B3; Figure 2). Patterns of change in rates were similar using either adult abundance index. Estimates of total instantaneous mortality (Z) for the Hudson River stock (Section 7) were relatively low in the mid to late 1980s, climbed in the early 1990s, and have fluctuated at relatively high levels since that time (Table B3; Figure B2).
Connecticut River: Relative exploitation rates for Connecticut River shad from in-river harvest and total stock harvest and based on lift counts generally declined from the late 1970s through the present (Table B4; Figure 3B). Relative exploitation rates from total stock harvest and based on CPUE in the river fishery generally declined from 1981 through the present (Table B4; Figure 3B). Estimates of Z for the Connecticut River stock generally increased from 1970 though the present (Table B4; Figure B). The increase was most pronounced in males.
Discussion The increase in relative exploitation for the Delaware River stock suggests a rise in fishing rate. Since the increase was coincident with the rise in mixed stock harvest and relative exploitation from in-river fishing did not increase (Sections 1.5 and 8), mixed stock harvest appears to have been the more important source of fishing mortality in recent years. The decline in CPUE in the Delaware River commercial fishery since 1990 suggests that increased fishing rate on this stock may have affected abundance. Age data were not available for estimates of Z so it is not known if changes in relative exploitation rates were related to changes in total mortality.
As in the Delaware River, in-river rates of relative exploitation remained relatively constant in the Hudson River stock while rates from total stock harvest increased. This again suggests some influence of ocean harvest on fishing rate. The spike in relative exploitation rate in the Hudson stock from in-river and total stock harvest that occurred in 1980 appeared to be driven by in-river harvest because ocean harvest was still a small component of total stock harvest at that time. The rise in relative exploitation rate from total Hudson harvest is coincident with the rise in Z suggesting that fishing may have been an important factor in the rise in Z in this stock.
Fishing did not appear to be a factor in the increased total mortality in the Connecticut River stock. In the Connecticut River, relative exploitation rate from in-river harvest and from total stock harvest declined as the mortality rates increased.
Results of these analyses suggest that the rise in mixed stock harvest that occurred in the 1980s may have increased fishing rates on the Delaware and Hudson River stocks of American shad. Conversely, fishing in general does not appear to have been a factor in the changing age structure and abundance of American shad in the Connecticut River. The directed fishery for American shad was closed in all Atlantic coastal states in 2005. Results of these analyses suggest that the closures may have been warranted for some stocks.
Literature Cited ASMFC (Atlantic States Marine Fisheries Commission). 1998. American shad stock assessment peer review report. Atlantic States Marine Fisheries Commission, Washington, DC.
Leggett, W. C., T. F. Savoy, and C. A. Tomichek. 2004. The impact of enhancement initiatives on structure and dynamics of the Connecticut River population of American shad. Pages 391-405 in 171
P.M. Jacobson, D.A. Dixon, W.C. Leggett, B.C. Marcy, Jr., and R.R. Massengill, editors. The Connecticut River Ecological Study (1961-1973) revisited: ecology of the lower Connecticut River 1973-2003. American Fisheries Society Monograph 9, Bethesda, Maryland.
Savoy, T. F., and V. A. Crecco. 2004. Factors affecting the recent decline of blueback herring and American shad in the Connecticut River. Pages 361-377 in P.M. Jacobson, D.A. Dixon, W.C.
Leggett, B.C. Marcy, Jr., and R.R. Massengill, editors. The Connecticut River Ecological Study (1961-1973) revisited: ecology of the lower Connecticut River 1973-2003. American Fisheries Society Monograph 9, Bethesda, Maryland.
172
Table 1.5.1-B1. Indices of adult abundance used in calculating relative exploitation for American shad of the Delaware, Hudson, and Connecticut Rivers.
Years River Index Evaluated Source Delaware CPUE in the commercial Lewis Haul seine Allen et al. (Section 8, this fishery 1970 - 2005 volume)
CPUE-revised in commercial gill net river fishery 1989-2005 Hudson ESSB - combined fishery independent / Hattala and Kahnle (Section dependent index 1985 - 2005 7, this volume)
EGG - fishery independent index of egg abundance 1974 - 2005 Connecticut Holyoke Dam fish lift passage 1976 - 2005 SAC (Section 6, this volume)
In river commercial gill net CPUE 1981 - 2005 173
Table 1.5.1-B2. Commercial harvest, adult abundance indices, and relative exploitation for the Delaware River American shad stock.
Relative Commercial Landings (kg) Adult Index Exploitation Estimated Delaware R. Total In-river Total Lewis Year Delaware R. Com. Fishery Delaware R Delaware Delaware Seine Portion of CPUE Harvest/
R. R. CPUE OMS* (revised) Lewis CPUE 1970 7606 2223 9829 4.9 2.0 1971 5162 1179 6341 12.3 0.5 1972 7031 953 7983 5.4 1.5 1973 5804 726 6529 7.2 0.9 1974 6923 680 7603 8.5 0.9 1975 11647 862 12509 14.9 0.8 1976 12138 0 12138 12.0 1.0 1977 23491 1542 25034 10.2 2.5 1978 23014 0 23014 10.1 2.3 1979 21052 0 21052 18.7 1.1 1980 34634 2815 37449 13.0 2.9 1981 63626 5785 69411 54.2 1.3 1982 105493 10761 116254 29.8 3.9 1983 68276 8776 77052 14.4 5.3 1984 88904 9807 98711 15.7 6.3 1985 75170 23767 98937 29.3 3.4 1986 80445 21012 101457 30.7 3.3 1987 95167 13898 109066 16.5 6.6 1988 112166 18921 131087 35.6 3.7 1989 113992 13177 127169 52.2 56.4 2.4 1990 166592 25574 192165 25.3 30.5 7.6 1991 162640 15788 178428 30.4 11.1 5.9 1992 110472 23139 133611 51.0 78.6 2.6 1993 114237 14769 129006 10.5 53.3 12.3 1994 78396 10620 89016 7.9 34.9 11.3 1995 68457 11841 80298 19.0 26.5 4.2 1996 80196 5078 85274 3.7 15.6 23.3 1997 72899 8039 80938 12.0 27.8 6.8 1998 76254 3684 79939 13.2 28.6 6.1 1999 80365 3504 83869 4.6 7.3 18.2 2000 61438 22755 84193 4.1 20.8 20.7 2001 80847 33011 113858 6.8 5.8 16.6 2002 48075 15992 64067 3.8 7.7 16.7 2003 28707 40346 69052 5.2 17.0 13.2 2004 42141 43132 85273 4.1 9.6 20.9 2005 29256 21419 50675 2.9 3.9 17.5
- OMS - Ocean mixed stock fishery 174
Table 1.5.1-B3. Commercial harvest, adult abundance indices, relative exploitation, and total mortality estimates for the Hudson River American shad stock. ESSB is a fishery dependent index of ages 5-7 expanded to all ages by fishery independent age data and mean weight at age.
Egg is an annual index of eggs per 1000 m3 from river-wide sampling.
Commercial Landings (kg) Adult Indices Relative Exploitation Total Mortality Estimated Total Total Hudson In- In-river In-river Year Total HR HR Female R. river ESSB Egg Harvest/ Harvest/ Male Z Hudson Harvest/ Harvest/ Z Portion of Hudson ESSB Egg ESSB Egg OMS*
1974 15609 110678 126288 0.097 114.1 130.2 1975 26255 114806 141061 0.060 191.3 235.1 1976 27398 98930 126328 0.037 267.4 341.4 1977 52964 94167 147132 0.036 261.6 408.7 1978 51903 206207 258110 0.044 468.7 586.6 1979 47583 232833 280416 0.045 517.4 623.1 175 1980 54488 595622 650110 0.046 1294.8 1413.3 1981 77838 281323 359161 0.161 174.7 223.1 1982 144343 171869 316212 0.123 139.7 257.1 1983 103989 208384 312372 0.356 58.5 87.7 1984 114242 318155 432397 0.472 67.4 91.6 0.72 0.53 1985 108684 342951 451635 64.3 0.262 53.3 70.2 130.9 172.4 0.70 0.46 1986 95678 362321 457999 112.9 0.770 32.1 40.6 47.1 59.5 0.68 0.63 1987 105597 310345 415942 89.3 0.349 34.8 46.6 88.9 119.2 0.60 0.56 1988 137284 355138 492422 80.2 0.259 44.3 61.4 137.1 190.1 0.79 0.63 1989 208860 220314 429174 69.2 0.327 31.8 62.0 67.4 131.2 0.72 0.54 1990 238521 210257 448777 25.9 0.270 81.1 173.2 77.9 166.2 0.83 0.63 1991 233884 149401 383286 26.6 0.086 56.1 143.9 173.7 445.7 0.98 0.97 1992 161160 120475 281635 20.9 0.075 57.5 134.5 160.6 375.5 1.37 0.98 1993 172047 62692 234739 14.2 0.120 44.0 164.7 52.2 195.6 1.41 0.67 1994 128638 90072 218710 30.1 0.227 29.9 72.6 39.7 96.3 1.24 1.06 1995 130010 112885 242895 19.1 0.121 59.0 127.0 93.3 200.7 0.81 1.42
Commercial Landings (kg) Adult Indices Relative Exploitation Total Mortality Estimated Total Total Hudson In- In-river In-river Year Total HR HR Female R. river ESSB Egg Harvest/ Harvest/ Male Z Hudson Harvest/ Harvest/ Z Portion of Hudson ESSB Egg ESSB Egg OMS*
1996 135536 83690 219226 41.5 0.262 20.2 52.8 31.9 83.7 1.29 0.78 1997 127247 67799 195047 13.6 0.036 49.8 143.4 188.3 541.8 0.65 0.43 1998 139083 105484 244567 21.5 0.086 49.1 113.8 122.7 284.4 1.03 0.99 1999 133268 66501 199769 16.6 0.085 40.0 120.2 78.2 235.0 1.33 1.15 2000 100962 69555 170517 25.3 0.119 27.5 67.5 58.4 143.3 1.29 1.20 2001 132345 45997 178341 9.3 0.039 49.6 192.2 117.9 457.3 0.97 1.11 2002 112240 59242 171482 8.3 0.034 71.5 207.1 174.2 504.4 2003 69057 49998 119055 15.9 0.072 31.5 75.0 69.4 165.4 1.13 1.13 2004 76355 33040 109395 8.1 0.033 40.9 135.4 100.1 331.5 0.80 0.91 2005 32286 23807 56093 9.9 0.042 24.1 56.8 56.7 133.6 0.90 0.83 176 *OMS - Ocean mixed stock fishery
Table 1.5.1-B4. Commercial harvest, adult abundance indices, relative exploitation, and total mortality estimates for the Connecticut River American shad stock.
Commercial Landings (kg) Adult Indices Relative Exploitation Total Mortality Total Estimated In- In-Total Ct/In-Year Connecticut In-river Total river Holyoke river Male Ct/H river Female Z R. Portion of Connecticut Connecticut Com. Lift Ct/H Z Lift CPUE OMS* CPUE lift (1000s) 1970 6720 78518 85238 66000 1.03 1.14 1971 3742 109182 112924 53000 0.85 1.17 1972 6198 113037 119236 26000 0.99 1.00 1973 4191 116847 121039 25000 0.98 0.67 1974 4207 112130 116337 53000 0.83 0.87 1975 6995 75071 82065 110000 1.21 1.07 1976 7779 177811 185590 350000 0.51 0.53 1.47 1.30 177 1977 14241 150777 165017 200000 0.75 0.83 1.83 1.46 1978 14136 138938 153074 140000 0.99 1.09 1.84 1.29 1979 14336 93804 108141 260000 0.36 0.42 2.27 1.42 1980 53045 140843 193888 380000 0.37 0.51 2.15 1.61 1981 75402 147284 222686 5.40 380000 0.39 0.59 41.22 2.37 1.27 1982 97182 128369 225551 5.13 290000 0.44 0.78 44.00 2.14 1.58 1983 60181 193234 253414 5.89 530000 0.36 0.48 43.00 1.24 1.70 1984 78873 180896 259769 6.87 500000 0.36 0.52 37.81 1.64 1.15 1985 92586 182347 274934 6.44 480000 0.38 0.57 42.69 1.68 1.31 1986 90192 146059 236251 10.29 350000 0.42 0.68 22.97 1.20 2.02 1987 95036 151457 246493 9.00 280000 0.54 0.88 27.39 1.29 1.53 1988 116144 85776 201920 6.89 290000 0.30 0.70 29.31 2.61 1.59 1989 89246 82147 171393 7.63 350000 0.23 0.49 22.48 1.37 1.84 1990 90171 117675 207846 4.50 360000 0.33 0.58 46.19 2.00 2.01 1991 86663 67722 154386 4.80 520000 0.13 0.30 32.16 1.63 1.31
Commercial Landings (kg) Adult Indices Relative Exploitation Total Mortality Total Estimated In- In-Total Ct/In-Year Connecticut In-river Total river Holyoke river Male Ct/H river Female Z R. Portion of Connecticut Connecticut Com. Lift Ct/H Z Lift CPUE OMS* CPUE lift (1000s) 1992 57120 65454 122575 6.22 720000 0.09 0.17 19.71 1.81 2.02 1993 62137 43845 105982 4.25 340000 0.13 0.31 24.94 1.82 2.02 1994 47860 47174 95034 4.57 181000 0.26 0.53 20.79 0.69 1.82 1995 51660 27931 79591 2.63 190000 0.15 0.42 30.32 1.34 1.57 1996 54660 30281 84941 4.00 276000 0.11 0.31 21.24 2.35 1.87 1997 56468 41279 97747 5.33 299000 0.14 0.33 18.33 4.42 2.98 1998 68666 40526 109192 5.33 316000 0.13 0.35 20.47 3.28 1.25 1999 63674 20219 83893 3.56 194000 0.10 0.43 23.59 2.30 1.81 2000 39783 48724 88507 7.78 225000 0.22 0.39 11.38 0.69 178 2001 60086 26869 86955 5.50 273000 0.10 0.32 15.81 2.02 1.11 2002 61182 49034 110216 8.40 375000 0.13 0.29 13.12 5.56 2.39 2003 31756 50407 82163 8.00 287000 0.18 0.29 10.27 4.14 2.07 2004 28028 30086 58115 5.33 191290 0.16 0.30 10.90 4.24 2.12 2005 12390 32626 45016 5.50 117756 0.28 0.38 8.18
- OMS- Ocean mixed stock fishery
Figure 1.5.1-B1. Relative exploitation of the Delaware River American shad stock using total harvest (ocean/mixed stock and in-river harvest) and the Lewis CPUE as a measure of stock abundance.
Total Del.R harvest/ Lewis CPUE Delaware River Del River com. fishery CPUE 25.0 ( i d) 90.0 80.0 20.0 70.0 Del R. CPUE 60.0 Relative u 15.0 50.0 40.0 10.0 30.0 5.0 20.0 10.0 0.0 0.0 1970 1974 1978 1982 1986 1990 1994 1998 2002 179
Figure 1.5.1-B2. Relative exploitation of the Hudson River American shad stock using total harvest (ocean/mixed stock and in-river harvest) and ESSB and Egg indices as a measure of stock abundance. ESSB is a fishery-dependent index of ages 5-7 expanded to all ages by fishery independent age data and mean weight at age. Egg is an annual index of eggs per 1000 m3 from river-wide sampling.
Hudson River Total HR harvest/ Egg 1600.0 In-river harvest/ Egg 1400.0 1200.0 Relative u (Egg) 1000.0 800.0 600.0 400.0 200.0 0.0 1974 1977 1980 1983 1986 1989 1992 1995 1998 2001 2004 Hudson River 250.0 Total HR harvest/ ESSB In-river harvest/ ESSB 200.0 Relative u (ESSB) 150.0 100.0 50.0 0.0 1984 1987 1990 1993 1996 1999 2002 2005 Hudson River 1.60 1.40 Male Z 1.20 Female Z Total mortality-1.00 0.80 0.60 0.40 0.20 0.00 1984 1987 1990 1993 1996 1999 2002 2005 180
Figure 1.5.1-B3 Relative exploitation of the Connecticut River American shad stock using total harvest (ocean/mixed stock and in-river harvest) and the Holyoke Lift and in-river CPUE as measures of stock abundance.
Connecticut River Total Ct / H Lift In-river Ct /H lift 1.20 1.00 0.80 Relative u 0.60 0.40 0.20 0.00 1976 1979 1982 1985 1988 1991 1994 1997 2000 2003 Connecticut River Total Ct/In-river CPUE (1000s) 50.00 45.00 40.00 35.00 Relative u 30.00 25.00 20.00 15.00 10.00 5.00 0.00 1981 1984 1987 1990 1993 1996 1999 2002 2005 Connecticut River 6.00 Male Z Female Z 5.00 Total mortality-Z 4.00 3.00 2.00 1.00 0.00 1970 1973 1976 1979 1982 1985 1988 1991 1994 1997 2000 2003 181
C. Striped Bass Predation on Adult American Shad: Occurrence and Observed Effects on American Shad Abundance in Atlantic Coastal Rivers and Estuaries Authors:
Andrew W. Kahnle and Kathryn A. Hattala Hudson River Fisheries Unit, New York State Department of Environmental Conservation 21 South Putt Corners Road, New Paltz, NY 12561 Introduction The possibility that increased striped bass abundance could negatively affect American shad and other potential prey has been discussed for some time (Walter et al. 2003). As mortality mortality rates of mature American shad have increased, researchers have focused on the possibility that mature striped bass could be preying on mature American shad. Clearly mature bass and mature American shad overlap in time and space in spawning rivers and along the Atlantic coast. Large striped bass (>90 cm TL) do prey on small mature American shad (Crecco et al. 2007, this volume) and such predation could affect American shad abundance. We evaluated the potential for adult predation by evaluating diet data for adult striped bass in the Hudson River estuary and inspecting the relationship between abundance indices of mature American shad and striped bass in selected coastal rivers.
Methods Hudson River Diet Data We collected approximately 200 mature striped bass annually from the Hudson River estuary from 1990 through 2006 for contaminant analyses. Fish were taken by electrofishing, haul seine, and fishery-independent and dependent gill nets in spring, summer, and fall. Fish were generally placed in a cooler of ice shortly after collection. Most fish were collected in the spring during the spawning migration. In the laboratory, fish were measured to the nearest millimeter total length (TL). Stomachs were removed and contents were identified to the lowest taxonomic level possible and enumerated. Contents were not preserved. Diets were summarized as frequency of occurrence (number and percent of stomachs containing a given item) by season and general diet category (crabs, fish, other, and empty) and by taxon for the entire sample, all seasons combined. We added diet summaries for striped bass >900 mm TL because that size striped bass was proposed by (Crecco et al. 2007, this volume) as being the size most likely to prey on adult American shad.
Striped Bass and American Shad Abundance (Selected Rivers)
We obtained indices of abundance for mature American shad and striped bass for the Connecticut River, Hudson River, Delaware River, Susquehanna River and upper Chesapeake Bay, Potomac River, and Albemarle Sound. Indices came from a variety of fishery independent and dependent sample programs conducted by state natural resource agencies and published in agency or interstate reports (Table C1).
Sampling occurred in spring. The two species were generally collected from within the same geographic area within each river or estuary, but by different sample gears within system. It was our working hypothesis that in systems where striped bass and American shad abundance consistently varied in opposite directions, striped bass may have affected the abundance of American shad. Abundance correlations between the two species were tested by a simple linear regression of annual abundance of American shad on annual abundance of striped bass.
182
Results Hudson River Striped Bass Diet We examined stomachs from 1,859 mature striped bass in the Hudson River during the 16-year study period, of which 1,646 came from spring collections (Table C2). Fish up to 1165 mm TL were taken, but modal length interval was 650 to 700 mm TL (Figure C1). Forty-seven fish (2.5%) were greater than 900 mm TL. Most stomachs were empty (84 % of total, 74 % of fish >900 mm TL; Table C3). Of stomachs containing food items, those with fish predominated in the total sample (66%) and in bass >900 mm TL (100%; Table C3). The two most common food items in stomachs were unidentified fish (36 % of total, 42 % of bass over 900 mm TL) and herring (18 % of total, 33 % of bass over 900 mm TL; Table C3).
American shad were observed in the stomachs of two striped bass (<1% of total, 17 % of bass over 900 mm TL) and both of these striped bass were greater than 1000 mm TL.
Striped Bass and American Shad Abundance (Selected Rivers)
There was some indication in some systems that abundance of mature striped bass and American shad was inversely correlated, at least for short periods of time (Table C4; Figure C2). In the last few years in the Connecticut River, it appears that shad abundance declined when striped bass abundance went up. A similar pattern occurred in the late 1980s in the Hudson River, and in the last few years in upper Chesapeake Bay. However, there were no long stretches of time in any system in which striped bass increased and American shad decreased. On the contrary, there were long stretches of time in almost all systems when abundance of the two species appeared to change in the same direction. This was most apparent in the Delaware River where data suggested a concurrent decline over the time series, in the Hudson River during the 1990s, and in the Potomac River where data suggested a concurrent increase. A simple regression of shad abundance on bass abundance was significant for the entire time series in the Delaware River (r2 = 0.44, slope = 2.02, P = 0.04; Figure C3) and in 1990 to 2001 in the Hudson River (r2
= 0.74, slope = 0.47, P = <0.01.). Regressions were not significant between species abundances in any other system.
Discussion Hudson River Striped Bass Diet Results of our diet analyses indicated that mature striped bass of the Hudson River Estuary do prey on river herring and mature American shad; however, American shad were only observed in the largest striped bass and only in two fish during the 16-year study period. Results of our study differed somewhat from those observed in a recent study of striped bass diet in the Connecticut River (Crecco et al. 2007, this volume). In the Hudson study, two of 47 striped bass >900 mm TL contained American shad. In the Connecticut study, 19 of 28 striped bass >900 mm TL contained American shad. American shad were not an important prey item in Connecticut River American shad less than 900 mm TL and not prey at all in smaller Hudson River striped bass. Only the smallest mature shad were consumed by striped bass in the Connecticut River. Sixteen percent of the striped bass in the Hudson had food items in their stomachs, while fifty one percent of the bass in the Connecticut study had food items.
Differences in percent full stomachs and frequency of American shad predation between striped bass of the Hudson and Connecticut rivers might be explained by differences in biota of the two systems. Striped bass spawn in the Hudson River, but not in the Connecticut River, so one would expect striped bass present in the Connecticut to be more focused on feeding. Furthermore, striped bass in any abundance are a new addition to the Connecticut River and prey, including American shad, may not have adjusted to the presence of such a large predator. Mature striped bass and American shad have occupied overlapping habitats in the Hudson River for hundreds of years and could be expected to have developed coping 183
strategies. Finally, it is likely that large striped bass of the Connecticut River have shifted to American shad as blueback herring abundance decreased as suggested by Savoy and Crecco (2004).
In any diet study of fish, the observed frequency of a given food item may be affected by rate of digestion rendering the item unrecognizable. This is most likely to occur with small, soft-bodied foods. We did not consider this for mature American shad in the Hudson and Connecticut River striped bass diet studies because mature shad are relatively large compared to their predators and because the relatively cool water temperatures during spring when most fish were collected would have slowed digestion.
It should be pointed out that striped bass >900 mm TL are a relatively minor component of striped bass present in Atlantic coastal rivers in spring. In haul seine sample of Hudson River striped bass spawning stocks, fish >900 mm TL made up an average of 2.5% of the males and 3.1% of the females in 1985 through 2006. These proportions compare favorably with the 2.5% of the Hudson River diet sample that were >900 mm TL. Size composition of mature striped bass in the Connecticut River is not known.
About 6.1% of striped bass in the diet study in that river were >900 mm TL.
Mature American shad do not appear to be an important food item for striped bass along the Atlantic coast. Walter et al. (2003) surveyed over 35 published and unpublished reports on striped bass diets in Atlantic coastal rivers, estuaries, and ocean waters. They found that menhaden, anchovies, river herring, and Atlantic herring were reported as dominating the diets of striped bass older than age one. The review summary was not detailed enough to identify if American shad had been identified as a minor prey item by any author.
Striped Bass and American Shad Abundance (Selected Rivers)
We did not find strong evidence from adult abundance data over the last 20 years that increased abundance of striped bass concurred with decreased abundance of American shad. Both American shad and striped bass are relatively long-lived fish with several age classes in their spawning populations. If striped bass fed on the smallest and youngest American shad, one would expect to see reduced recruitment to the shad spawning population and slow erosion of shad abundance as bass abundance remained high or increased. We did not see evidence of such long-term changes in available abundance data. Short-term increases in bass occasionally coincided with short-term declines in shad, but such change would not likely be caused by predation on recruiting fish unless only a few age classes of prey were present. More importantly, there were several datasets in which both American shad and striped bass changed in the same direction for all, or part of, the time series. Such concurrent changes suggest change in environmental conditions or mortality factors that affected both species in a similar manner.
Literature Cited ASMFC (Atlantic States Marine Fisheries Commission). 2005. 2005 Stock assessment report for Atlantic striped bass: Catch at age based VPA & tag release/recovery based survival estimates, Report prepared by the Striped Bass Technical Committee for the Atlantic Striped Bass Management Board. Washington, DC. USA.
Savoy, T. 2005. Stock Assessment Data for American Shad in the Connecticut River: Data Description of American shad in the Connecticut River for Stock Assessment by the Atlantic States Marine Fisheries Commission. Connecticut Department of Environmental Protection, Marine Fisheries Division. Old Lyme, Connecticut.
Savoy, T and V. Crecco. 2004. Factors affecting the recent decline of blueback herring and American shad in the Connecticut River. Pages 361-377 in P.M. Jacobson, D.A. Dixon, W.C. Leggett, B.C.
Marcy, Jr., and R.R. Massengill, editors. The Connecticut River Ecological Study (1961-1973) 184
revisited: ecology of the lower Connecticut River 1973-2003. American Fisheries Society Monograph 9, Bethesda, Maryland.
Walter, III, J.F., 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.
185
Table 1.5.1-C1. Relative abundance indices for adult American shad and striped bass in Atlantic coastal rivers and estuaries River Species Data Source Connecticut American Shad Total annual lift numbers at the Holyoke Dam on the Connecticut River (Savoy 2005).
Striped Bass Relative adult abundance from annual electrofishing collections (Crecco et al. 2007, this volume).
Hudson American Shad Adult index based on index of egg abundance and age structure in the haul seine fishery-independent sampling (Hattala and Kahnle 2007, this volume).
Striped Bass Adult index based on index of egg abundance and age structure in the haul seine fishery-independent sampling.
Delaware American Shad Catch per haul in commercial haul seine collections (Allen et al. 2007, this volume).
Striped Bass Electrofishing CPUE for ages 3-13 (ASMFC 2006).
Susquehanna American Shad Catch per angler hour in fishery-independent sampling in the Conowingo Dam tailrace (Sadzinski et al. 2007, this volume).
Striped Bass Index of spawning biomass March - May (B. Sadzinski, pers. comm.).
Potomac American Shad Pounds per net day in commercial pound (Carpenter et al.
2007, this volume).
Striped Bass Index of spawning biomass March-May (B. Sadzinski, pers.
comm.).
Albemarle Sound American Shad Pounds per trip in commercial gill net (Burgess et al. 2007, this volume).
Striped Bass CPUE in spring fishery-independent gill net survey (Burgess, pers. comm.).
186
Table 1.5.1-C2. Number and mean total length striped bass collected from the Hudson River examined in diet analyses, 1990-2006.
Primary Food Season Item N Mean TL SD Min Max All Fish Spring Crabs 39 666.5 79.1 535 875 Fall Crabs 2 525.5 30.4 504 547 ALL 41 659.6 83.1 504 875 Fall Fish 32 579.2 114.7 356 815 Spring 140 691.3 122.3 476 1165 Summer 26 661.9 88.9 484 838 Unknown 2 568.0 9.9 561 575 ALL 200 668.3 123.5 356 1165 Fall Other 1 672.0 Spring 59 624.1 89.7 467 995 Summer 2 586.0 35.4 561 611 ALL 62 623.6 88.0 467 995 Fall Empty 74 621.4 113.3 390 935 Spring 1408 649.6 94.8 425 1006 Summer 51 667.7 153.8 368 1054 Unknown 21 588.9 66.4 496 727 ALL 1554 648.0 98.3 368 1054 Fish > 900 mm TL Spring Fish 12 980.0 73.9 900 1165 Fall Empty 1 935.0 Spring 30 947.0 35.1 902 1006 Summer 4 956.8 67.8 904 1054 ALL 35 947.8 38.3 902 1054 187
Table 1.5.1-C3. Stomach contents of striped bass collected from the Hudson River Estuary, 1990-006.
Category All Fish Fish > 900 mmTL Total number of stomachs examined 1859 47 Empty 1555 83.6% 35 74.5%
Total with food items 304 16.4% 12 25.5%
Food Items N-Stomachs N-Stomachs Crab, blue 41 13.5% -
Crab, Mud 2 0.7% -
Crab, unidentified 6 2.0% -
TOTAL Crabs 49 16.1% 0 Fish, Unidentified 108 35.5% 5 41.7%
American eel 3 1.0% 0 American shad 2 0.7% 2 16.7%
Atlantic menhaden 14 4.6% -
Catfish/bullhead 3 1.0% 1 8.3%
Cunner 1 0.3% -
Herring, total 55 18.1% 4 33.3%
Alewife (7) -
Blueback herring (12) (1)
Needlefish 2 0.7% -
Pipefish 1 0.3% -
Sand lance 1 0.3% -
White perch 11 3.6% -
TOTAL Fish 200 65.8% 12 100.0%
Amphipods 3 1.0% -
Arthropod 4 1.3% -
Crustacean 1 0.3% -
Grass shrimp 1 0.3% -
Isopods 13 4.3% -
Miscellaneous* 14 4.6% -
Rocks 4 1.3% -
Squid 3 1.0% -
Unidentified digested material 22 7.2% -
TOTAL-Other 48 15.8% 0
- Sticks, worms, bones, bologna sandwich, mollusk parts, clam shell, plastic straw, mussel 188
Table 1.5.1-C4. Relative abundance indices for adult American shad and striped bass in Atlantic coastal rivers and estuaries.
Connecticut River Hudson River Delaware River Upper Bay Potomac River Albemarle Sound Shad Striped Holyoke Striped American Striped Striped Shad Shad Shad Year FD Striped Striped Bass Shad Bass Shad Bass Bass Catch Pound FD Haul Bass Bass FI Gill Lift Electro Adult Adult Electro Per Net Gillnet Seine SSB SSB Net (N*1000) CPUE Index Index CPUE Angler hr CPUE CPUE CPUE CPUE 1985 31.82 0.42 2.1 64.9 25.9 1986 55.40 0.72 3.0 152.0 45.7 1987 45.03 1.51 3.7 400.5 88.8 1988 39.30 1.34 6.0 250.3 0.9 63.6 1989 36.57 3.93 6.5 120.3 0.7 80.5 1990 15.60 4.67 5.0 98.4 1.7 62.5 1991 17.83 5.69 5.7 109.4 1.6 138.7 1.2 1992 14.49 4.43 6.1 275.0 2.1 379.4 0.5 1993 720 11 11.95 5.25 5.6 278.5 4.8 420.9 0.3 1994 181 44 23.23 5.45 5.6 87.3 1.7 156.8 0.2 1995 190 13 12.63 2.60 7.4 547.7 3.6 293.8 169.5 1.0 189 1996 276 86 26.75 11.06 3.7 3.4 10.1 347.9 4.1 391.6 144.8 0.7 1997 299 142 9.12 3.40 12.0 4.1 10.5 256.9 6.9 369.6 167.8 0.9 1998 316 110 13.87 4.04 13.2 3.7 14.2 157.4 6.7 217.0 206.6 1.0 1999 194 59 11.79 9.90 4.6 2.6 15.8 161.4 5.1 275.2 166.8 1.2 2000 225 198 37.71 17.84 4.1 2.1 20.4 169.9 5.8 301.8 172.9 0.9 2001 273 126 29.08 14.17 6.8 1.9 14.8 490.2 11.2 273.2 151.9 1.0 2002 375 113 3.8 1.6 13.5 266.4 7.9 380.7 226.1 0.9 2003 287 162 6.26 4.89 5.2 2.5 11.2 566.2 14.9 118.5 280.6 0.4 2004 191 203 34.90 4.43 4.1 2.9 12.2 389.8 10.3 578.8 232.9 1.0 2005 5.12 2.9 0.0 7.1 469.7 12.2 196.1 201.0 0.7
Figure 1.5.1-C1 Length frequency of striped bass, collected in the Hudson River, and used in diet analysis.
600 Fall 500 Summer 400 Spring Number 300 200 100 0
300 400 500 600 700 800 900 1000 1100 1200 Total length (mm) 190
Figure 1.5.1-C2 Relative abundance of American shad and striped bass in Atlantic coastal rivers and estuaries.
Connecticut River 800 250 700 Shad Bass 200 600 500 150 Shad 400 Bass 300 100 200 50 100 0 0 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Hudson River 60.00 20.00 Shad 18.00 50.00 16.00 Bass 40.00 14.00 12.00 Shad 30.00 10.00 Bass 8.00 20.00 6.00 10.00 4.00 2.00 0.00 0.00 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 Delaware River 14.0 4.5 Shad 4.0 12.0 Bass 3.5 10.0 3.0 8.0 2.5 Shad Bass 6.0 2.0 1.5 4.0 1.0 2.0 0.5 0.0 0.0 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 191
Figure 1.5.1-C2 (cont.) Relative abundance of American shad and striped bass in Atlantic coastal rivers and estuaries.
Upper Chesapeake Bay 25.0 Shad 600.0 Bass 500.0 20.0 400.0 15.0 Shad 300.0 Bass 10.0 200.0 5.0 100.0 0.0 0.0 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 Shad Potomac River Bass 16.0 700.0 14.0 600.0 12.0 500.0 10.0 400.0 Shad 8.0 Bass 300.0 6.0 4.0 200.0 2.0 100.0 0.0 0.0 1985 1987 1989 1991 1993 1995 1997 1999 2001 Shad Albemarle Sound Bass 300.0 1.4 250.0 1.2 1.0 200.0 0.8 Shad 150.0 Bass 0.6 100.0 0.4 50.0 0.2 0.0 0.0 1991 1993 1995 1997 1999 2001 2003 2005 192
Figure 1.5.1-C3 Relationship between relative abundance of mature striped bass and mature American shad in Atlantic coastal rivers and estuaries.
Striped bass vs A. shad abundance - Connecticut River 800 600 y = -0.7474x + 372.8277 2
A.shad R = 0.1099 400 200 0
0 50 100 150 200 250 Striped bass Striped bass vs A. shad abundance - Hudson River 60.00 50.00 y = -0.2266x + 26.173 2
40.00 R = 0.0056 A. sahd 30.00 20.00 10.00 0.00 0.00 5.00 10.00 15.00 20.00 Striped bass Striped bass vs A. shad abundance - Delaware River 14.0 y = 2.0212x + 1.048 12.0 2 R = 0.4385 10.0 A. shad 8.0 6.0 4.0 2.0 0.0 0.0 1.0 2.0 3.0 4.0 5.0 Striped bass 193
Figure 1.5.1-C3 (cont.) Relationship between relative abundance of mature striped bass and mature American shad in Atlantic coastal rivers and estuaries.
Striped bass vs A. shad abundance - Potomac River 20.0 15.0 y = 0.0071x + 4.0073 A. shad 2
10.0 R = 0.0598 5.0 0.0 0.0 100.0 200.0 300.0 400.0 500.0 600.0 700.0 Striped bass Striped bass vs A. shad abundance-Upper Chesapeake Bay 25.0 y = 0.0055x + 7.3992 20.0 2 R = 0.0317 A. shad 15.0 10.0 5.0 0.0 0.0 100.0 200.0 300.0 400.0 500.0 600.0 Striped bass Striped bass vs A. shad abundance-Albemarle Sound 300. 0 250. 0 200. 0 A. shad 150. 0 100. 0 y = -21.456x + 207.33
- 50. 0 R2 = 0. 0214
- 0. 0 0.0 0.2 0.4 0.6 0.8 1. 0 1.2 1.4 Striped bass 194
1.5.2 Stock-Specific Perspectives A. Delaware River The SASC examined data to determine if the decline in adult shad abundance during the1990s was caused by overfishing. In the 1998 assessment, estimates of fishing mortality were determined based on landings data and stock estimates derived from landings and population estimates. Controversy arose due to lack of confidence in the population estimates and the probability of their accuracy to perform such calculations.
Estimates of relative exploitation were developed from commercial CPUE data to ascertain if potential overfishing occurred during the period of low adult abundance in the early to mid-1990s. Relative exploitation estimation is a basic approach with minimum assumptions that reveals trend in exploitation (annual harvest divided by an index of relative abundance) rather than absolute estimates of fishing mortality. The SASC developed estimates from the New Jersey and Delaware gill-net fisheries as well as the Lewis haul seine fishery. For this assessment, estimates of relative exploitation were developed by dividing annual in-river harvest (river and upper bay commercial landings) by the CPUE.
Relative exploitation rates were developed for the 1985 to 2005 time period when more reliable in-river estimates of harvest are available (see Section 8, Figure 8.19). All estimates were standardized (Z-transformed) with a value of two added to eliminate any negative numbers for easier comparison. All estimates of relative exploitation were fairly similar throughout the 1990s. The analysis has shown an increase in relative exploitation in recent years beginning in 2000 but the extent of this trend is unknown.
This increase may be a direct result of the mandatory reporting enacted by New Jersey starting in 2000 and might not be an actual increase in exploitation but rather an increase in reported harvest.
Alternatively, the increase may be a result of current low population size and could be potentially harmful to stock restoration. Further study is needed to determine if recent trends in exploitation are of a magnitude to necessitate concern.
To test the hypothesis of the mandatory reporting effect on relative exploitation, the analysis was repeated using only State of Delaware landings instead of all in-river landings. The results of this analysis indicate that the relative exploitation has actually decreased in recent years (see Section 8, Figure 8.20). This is also no indication that overfishing was a major factor in any adult population decline in the 1990s.
Two estimates of relative exploitation of American shad were calculated (see Section 8, Figure 8.21) using the CPUE from the Lewis fishery combined with harvest data from the Delaware Estuary (1954-2005) and in-river fisheries (1985-2005). All estimates were standardized (Z-transformed) with a value of two added to eliminate any negative numbers for easier comparison. Although Estuary landings data are potentially biased with mixed stock landings from the lower Delaware Bay area, they are considered useful in determining how current estimates compare in magnitude to those of the past. Estimates of relative exploitation from the estuary harvest were very high from 1954 to 1968 when compared to the 1990s, as well as when compared to estimates of in-river relative exploitation in the 1990s.
The relative exploitation derived from the Lewis fishery varied without trend from 1985 to 1999, but increased dramatically in recent years. Again it is likely that this increase is a direct result of the mandatory reporting enacted by New Jersey starting in 2000 and is not an actual increase in exploitation, reinforcing the need to explore exploitation, if not actual harvest, within the Delaware River Basin.
Since overfishing did not seem to be a major factor in stock decline during the 1990s, other data were analyzed to determine the potential cause. The SASC also looked at potential interactions with striped bass within the Delaware Estuary to determine if the shad decline was a direct result of a predator-prey 195
relationship. Analysis of American shad young-of-year (YOY) and age 1+ indices were compared to various striped bass indices from 1980 to 2005 including:
- NJ seine age 1+, arithmetic and geometric means (1980-2005)
- NJ seine YOY (1980-2005)
- Delaware Bay (DE) trawl 2-8 aggregate (1982-2005)
- Delaware River spawning stockages 2-8 separate and 3-12 aggregate (1996-2004)
The only significant correlation found was between the Delaware River striped bass spawning stock age-3 and the upper Delaware River American shad YOY (r2 = 0.5747) with no time lag. When lagged, there is no correlation (r2 = - 0.0009). Because the DE trawl did not catch many striped bass from 1982 to 1990, the SASC looked at the DE trawl compared to the averaged Delaware River YOY for the period 1991 to 2005 (see Section 8, Figure 8.22). The resulting correlation (r2 = 0.5488) was exceeded only by the age-3 Delaware River spawning stock analysis.
After eliminating recent years to determine if there was any correlation in the 1990s when shad abundance was declining, a stronger correlation (r2 = 0.7415) was found from 1991 to 2000 (see Section 8, Figure 8.23). Analyses with the YOY indices and Delaware Bay trawl striped bass indices were not time lagged due to fact that the Delaware Bay trawl is an aggregate index. An important detail to note is that the relationship is dominated by the 1996 year-class of American shad, which may have an effect on all assumptions of striped bass-American shad interactions.
The striped bass-American shad analyses show that when shad YOY production is high, there are ample striped bass present in the Delaware system of the 2-8 year-old age classes and the opposite seems to be true if shad YOY production is low. For both species this seems to be dictated by environmental conditions. This may mean that striped bass was a limiting factor during the 1990s in years when shad production was high. Additional empirical evidence, such as stomach content analysis, is necessary to determine if the correlation has a factual basis or is just due to the opposing directions in which the two species have been heading in the Delaware since the mid-1990s.
B. Connecticut River American shad landings and run sizes have fallen steadily in the Connecticut River since 1993 despite relatively high and persistent juvenile production since 1990. Several analyses have been conducted to determine the effects of fishing and predation on the recent stock decline of Connecticut River shad that have incorporated both ocean recreational shad landings and ocean commercial discards in the development of fishing mortality estimates. In addition, equilibrium and non-steady state overfishing thresholds (Fmsy, Nmsy) were estimated for Connecticut River shad.
Connecticut Department of Environmental Protection (CTDEP) estimates of fishing mortality on American shad declined after 1995, which did not support the hypothesis that the drop in shad run size since 1993 was due to overfishing. Additionally, trends in juvenile production have been at or above the long-term median index despite the persistent drop in adult stock size from 1996 to 2005, providing evidence that there has been a recruitment bottleneck in the Connecticut River American shad stock. This scenario led CTDEP to hypothesize that this recent failure in shad productivity was largely due to a systematic rise in striped bass predation on adult shad in the River (Savoy and Crecco 2004).
An age-aggregated Steele and Henderson (S-H) production model was constructed on data for the Connecticut River stock of American shad. Results indicated that striped bass consumption rates on adult 196
shad rose in the River four-fold after 1994, coincident with a steady rise in striped bass abundance.
Estimates of annual adult shad consumption by striped bass from 1999 to 2005 were 5 to 15 times greater than the in-river landings (sport plus commercial) during those years. The approximate equilibrium Fmsy level of 0.39 for shad based on the S-H model easily exceeded the total fishing mortality levels on adult shad in all years from 1989 to 2005. Non-equilibrium Fmsy levels approached 0.50 in most years from 1981 to 1993, but annual Fmsy levels fell steadily thereafter to 0.02 in 2004 following a steady rise in striped bass consumption rates. Statistical and empirical evidence given here strongly suggest that the recent emergence of a recruitment bottleneck and the subsequent decline in adult shad run size in the Connecticut River were linked mainly to predation effects from striped bass.
For the complete analysis of the Connecticut River predation hypothesis, please see the Appendix A:
Stock Assessment of American Shad in Connecticut (Minority Opinion).
Literature Cited Savoy, T.F. and V.A. Crecco. 2004. Factors affecting the recent decline of blueback herring and American shad in the Connecticut River. In P.M. Jacobsen, D.A. Dixon, W.C. Leggett, B.C.
Marcy, Jr. and R.R. Massengill (eds), The Connecticut River Ecological Study (1965-1973)
Revisited: Ecology of the Lower Connecticut River 1973-2003. American Fisheries Society, Monograph 9, Bethesda, Maryland. pp. 361-377.
C. Hudson River Authors:
Kathryn A. Hattala and Andrew W. Kahnle Hudson River Fisheries Unit, New York State Department of Environmental Conservation 21 South Putt Corners Road, New Paltz, New York 12561 Introduction Currently the Hudson River American shad stock is at its lowest point in its 125-year history since record keeping began. The Hudson is not alone in this position; other stocks, from the Mid-Atlantic states to Maine, are there too. So how did the Hudson stock get here, and more importantly, why?
The sections preceding this have outlined the sporadic history of data collection on shad stocks, including those with coastwide consequences for many stocks: the inadequate knowledge of bycatch in coastal fisheries, the unknown component of shad in bait fisheries, and stock composition in known current and past fisheries. This lends a level of uncertainty to any analyses to quantify the effect of fishing. However, the trends observed in the Hudson stock over the past 25 years led us to one conclusion.
Methods To realistically estimate the effects of fishing on the Hudson shad stock, we cannot ignore the harvest from the ocean mixed stock fishery. Effects of mixed stock harvest on selected individual stocks were debated during preparation of the last assessment (ASMFC 1998). We updated estimates of stock composition of mixed stock harvest in Section 1.5.1.
During the mid-1980s we noticed several changes in the Hudson River stock. Fewer shad were returning (change in CPUE in the in-river fixed gill-net fishery, see Section 7), plus older fish were disappearing.
197
Mixed stock harvest combined with in-river harvest make up most of the removals for a single stock. An unknown, and unaccounted, bycatch kill adds to the uncertainty of the total kill amount. Once total harvest is estimated it can then be compared to a spawning stock index of the affected population. We attempt to draw this comparison to estimate the effect of mixed stock and in-river harvest on the Hudson stock.
Spawning Stock Abundance and Biomass Indices Spawning stock abundance (SSA) and biomass (SSB) indices can be approached in two ways. The first is an empirical approach, the calculation of an index that measures the current spawning stock equivalent to the annual production. The methods for the calculating our empirical spawning stock abundance and biomass indices (ESSA/ESSB) are described in Section 7. Here we include a second approach that predicts future stock potential. This approach begins an abundance index at age zero, then uses a process that Agrows them up@ to generate an estimate of current or future spawning stock abundance and biomass.
We developed both empirical and predictive indices to compare expected versus observed returns of adults.
Potential Production Index Our approach begins with an abundance index at age zero, growing them up over their lifetime time to estimate each year-class=s potential contribution to the spawning stock. For the measure at age zero, we chose the PYSL index because the long times series of available data includes most of the year-classes present at the adult ages in the spawning population.
Each annual age-0 measure was decremented by assumed annual survival rates for pre-recruits from age zero through age four, then by survival derived from annual mortality rates observed in the adult stock for ages five through 14. Values used were S=0.19 for age zero to age one, S=0.37 for ages one and two, and S=0.74 (from M=0.3) for age three to recruitment to the fishery (ASMFC 1998). Estimates of S for the adult stock for 1974 to 1983 are from Deriso et al. (2000). S for 1984 to 2005 are from the annual mortality rates calculated from the spawning stock age structure (Table C1). A maturity schedule is then applied to each age of each year-class. The annual estimate of the forward projection of the PYSL to spawning stock abundance (PSSA) equals the sum of year-classes from ages three through ten predicted to be present each year. Ages three through ten represent most of the adult spawning population with ages greater than age ten comprising a small number of the adults present. To calculate the biomass index (PSSB), the fraction at-age is multiplied by observed annual weight-at-age (WAA in kilograms, Section 7 appendix) before summation.
Available data on age zero allowed the calculation of the PSSA and PSSB from1984 through 2008.
Estimates of Current F We attempted to use several modeling approaches as well as catch-curve analyses (Section 7) to estimate current F on the Hudson stock. These techniques included attempts at using virtual population analysis (VPA), statistical catch-at-age, and surplus production models. The goal was to arrive at some convergence in model-generated mortality rates with mortality estimates calculated from catch-curve or tag based methods. We also developed a stock reconstruction model, which utilized all available stock data. The Hudson River American shad stock has relatively long data series, unlike most other east coast shad stocks.
We made an effort to try a variety of models to understand the nuances of model behavior and to find useful tools in our assessment. Convergence of model outputs would, at minimum, increase the level of confidence in estimates of F.
198
Surplus Production The most commonly used surplus production model used in east coast fisheries is ASPIC, available from its author Dr. Michael Prager (2005) at the NMFS Beaufort Laboratory. NMFS staff, at Woods Hole, offers a Windows based version of Dr. Prager=s model, for those not versed in running DOS based software (NOAA 2005). Prager (2005) developed this model for use when accurate age data is lacking. Primary model inputs include a long-term data series of catch and effort or CPUE. With these inputs, the model attempts to calculate a series of exploitation rates, and estimates population growth parameters (K and r) and fishing mortality rates, MSY, etc., along with diagnostic tools (residuals).
For the Hudson shad stock, in-river catch data was available from 1915 to the present. Several effort data series are available. Talbot (1954) estimated fishing effort for the period 1915 to 1951 in standard fishing units. A SFU equals the number of licensed nets in both New York and New Jersey waters, multiplied by the allowed fishing hours per week. The number of nets for New Jersey is adjusted by a correction factor of five as Talbot (1954) calculated that their fishing power was five times greater than nets in New York.
Using his methodology and data up until the present, we input data from the entire time series from 1915 to 2005. Another effort data series available was calculated by Klauda et al. (1976). It is similar to Talbot=s, but instead uses the total amount of square yard hours of net multiplied by allowable fishing hours. We also calculated this time series for the last 25 years so the complete times series is from 1931 to 2005 (see Section 7).
Virtual Population Analysis We attempted to fit Hudson River catch-at-age data to two different VPA models, the ADAPT version produced by Woods Hole (NOAA 2005) and the ICES VPA versions (G. Shepherd, pers. comm.). Use of a VPA requires accurate catch-at-age data from all sources of fishing mortality along with indices of abundance at various life stages for tuning.
The only accurate catch-at-age data for known harvest of Hudson River stock is from the in-river harvest of American shad. Ocean catch-at-age was impossible to generate as little data has been collected to describe age-structure of the catch, let alone stock composition. In order to work in a VPA, multiple year age data summaries would need to be generated, summarized by each individual fishery (both directed and bycatch) and stock. To keep things simple, we used the method described in Section 1.5.1 as a first order approximation of the Hudson River component of ocean catch. Then we applied the age structure observed in the Hudson River to the estimated ocean catch.
The only indices of abundance that exist to tune the VPA are those that are measured in the Hudson at adult ages and at age zero (see Section 7).
Age Structured We also input the data described above into ASAP (Age Structured Assessment Program, NFT version, NOAA 2005). ASAP is a forward projecting model that requires age based catch data, the same as described for the VPA above.
Stock Reconstruction The stock reconstruction model is the result of cooperative work of the authors and Dr. C. Walters (pers.
comm.) and Dr. R. Deriso ( pers. comm.). This model builds an annual age structured population in biomass over a series of years of available data, 1915 to the present. The model begins with an estimated natural recruitment rate (millions of age ones), and then applies a survival rate generated from a constant M to 199
recruitment-at-age for ages one through 14. Weight-at-age data is input to generate a vulnerable stock size in biomass. Inputs for this model are the same as those used for the Hudson in Section 1.1.5.
Another input to the model is losses to the stock. This includes in-river and (1915 to the present, see Section 7), and estimated ocean harvest (1970 to present, see Section 1.5.1) of adults which are used to estimate annual exploitation from the calculated vulnerable stock size.
Maturity, fecundity and vulnerability-at-age are applied to the numbers-at-age to calculate annual egg production, which is also decremented by annual exploitation. The resulting annual egg production, as proxy for the adult stock, is used in a Beverton-Holt stock recruitment (S-R) function to estimate annual recruit production. Included in the S-R are estimated entrainment losses (1952 to present) at age zero (see Sections 1.1.5 and 7).
The user is allowed to select other input parameters: proportion of entrainment mortality before compensation, future harvest (exploitation) rate, future entrainment mortality rate (proportion of larvae lost), and annual natural survival rate (exp(-M)). The model estimates recruitment efficiency (proportion of natural biomass needed for one half the normal recruitment), the natural (unfished) recruitment rate (millions of age ones) and the ratio of 1915 (beginning year) recruitment to unfished recruitment.
The stock production model is set up as a Microsoft Excel spreadsheet for ease in use and can be tuned to observed data in two different ways using the SOLVER function. Observed data measured over the past 25 years are used as tuning indices: PYSL, YOY, adult relative abundance indices and mortality rates (see Section 7). One tuning method uses a maximum likelihood function that attempts to fit the model to the observed data series. The second method tunes the model to past and present exploitation rates using least squares analysis. Comparisons can also be made to using observed versus model generated age structure.
This model was further refined (Deriso et al. 2000). It utilized a similar approach as the stock production model, but in a much more detailed method that includes the use of age and repeat spawn data to calculate estimates of survival and impact of entrainment loss.
The stock-recruitment function allows for a predictive component to estimate the effects of future management strategies.
Assessment Results Overfishing Definition In Section 1.1.5, we selected a Z30 = 0.54, using age invariant natural mortality (M=0.3), a Type I fisheries, and a simple biomass per recruit (BPR) analysis. For this section however, we converted the benchmark to an F30=0.24 (BPR) by subtracting M (Table C2).
Spawning Stock Abundance and Biomass Potential SSA and SSB The forward projection or potential spawning stock abundance index (PSSA) attempts to predict what the spawning stock should be, based on production of young that occurred previously. The index was relatively low in 1984, then increased to high level for the period 1987 to 1990 (Table C3; Figure C1). The index began to decrease in stepwise drop/increase fashion until 1997 after which the index dropped and continued to decline through the present. The predicted outlook is bleak, as the index remains low through 2008. A 200
similar pattern occurred for the potential spawning stock biomass index (PSSB), but the patterns of high and low are much more pronounced than for the PSSA (Figure C1).
Comparison of the Empirical versus Potential The abundance indices (ESSA or PSSA) correlated with their respective related biomass index (ESSB or PSSB; Figure C2). However, when the potential index is compared to the empirical index a gap appears through the years 1990 to about 1998 (Figure C2). Focusing on the biomass indices, we compared the ESSB to the in-river harvest. As expected, the two track each other well (r2=0.84, P=6.06E-09). They both measure what returned to the Hudson (Figure C3). We also compared the PSSB to the total harvest (estimated ocean mixed stock fishery harvest plus in-river harvest) for the Hudson (Figure C4). The estimated total harvest correlates well (r2=0.70, P=1.2E-06; Figure C4) with the PSSB. These correlations suggest that the ocean harvest may have come from the predicted stock that did not return to the river.
Current F Given that Z = M + F, estimates of Z from catch curve analyses of age (Section 7) and assumed values of M can be used to generate estimates of F. Using the age invariant estimate of M=0.30, estimates of F for females remained relatively low (range of 0.16 to 0.33) for the period 1984 to 1990 (Figure C5). F estimates based on spawning marks were similar for the same period. From 1991 through 1995, values of F increased and peaked at 1.12 in 1995. F varied, but remained high, with the exception of the drop in 1997, until 2003.
The last two years F declined to 0.53. A similar pattern in F occurred for males, but was higher than females in most years.
Deriso et al. (2000) found fishing mortality rates (F) for older shad of 0.4 to 0.5 for the period 1974 to 1992, with rates higher for female shad than for males. Average exploitation was 0.33 (F=0.4) for the same period. These estimates of F, and those generated from recent estimates of Z (see Section 7), exceed all estimates of overfishing we recommend in our analyses (Figure C5). It should be noted that the beginning of the decline in the PSSB coincides with the increase in F beginning in 1991.
Model Estimates Surplus Production Several model runs were made with ASPIC. Each run resulted in an error where r, the intrinsic rate of stock growth, was forced to the minimum constraint making results trivial. Examination of the effort residuals suggests that effort was underestimated during periods of the peak harvest (e.g., WWII; Figure C6). The catch data series also includes a period where dramatic changes were taking place in stock size, from record peak harvests (high stock size) of approximately 1.4 million kg per year in the 1940s to a record low catches (low stock size) of less than 90,000 kg per year in the last 10 years.
Surplus production models require that all mortality to a stock be included. This presents a problem when trying to include a more accurate estimate of ocean catch. Our estimate of ocean catch is only a first order approximation of kill. This model also requires a corresponding estimate of effort, which is not available.
This precludes the model=s usefulness of as an effective tool if only one component of catch and effort (for the Hudson River) is included and the ocean kill and associated effort is left out.
Virtual Population Analysis We were unable to obtain good VPA results in several attempts to run either the ADAPT or ICES version.
VPA modeling requires a large amount of detailed catch-at-age data, and abundance indices for all ages:
201
age one to the assumed oldest age, which is 13, for Hudson River American shad. In order to get either model to run, a series of adjustments to the input data were necessary. These included collapsing of age structure data from age three to 13 down to ages three to 8+, and realigning age zero data to reflect abundance at age one and two.
Interpretation of results proved just as problematic. The two VPA types varied widely in resulting F estimates. Patterns of residuals suggested bias. Most all of these problems can be attributed to the lack of necessary catch-at-age data, collapsing the age structure data for model input, the assumption of applying the Hudson River age structure to the estimated ocean catch, and the lack of tuning indices. Only five tuning indices are available for shad, all of which are from observed data in the Hudson. In contrast, the VPA for striped bass uses nearly 50 tuning indices covering ages zero to 15+ during both in-river and ocean residency (ASMFC 2005). The VPA method is not worth pursuing for American shad further until detailed catch-at-age data are available.
Statistical Catch-at-Age The ASAP did not fare any better than either the VPA or ASPIC models. The same problems occurred in that there are no age data to associate with ocean mixed stock harvest. We suspect there may be an issue with fishery selectivity. In addition to no age data for the mixed stock harvest, there are no estimates of effort or fishery selectivity. We tried several times series of data (1915 to present, 1931 to present and 1970 to present). Residuals were extremely skewed, with very low estimates of F, a result that is counterintuitive given other measures of F (see above).
Stock Reconstruction After exploring the use of the models above, many problems were identified. Most were on the input side of incomplete records of effort and/or accurate data of catch and kill-at-age. Little data were recorded when the mixed stock/ocean fisheries operated leaving wide margins for error. We attempted to simplify assumptions by using a model that did not require a high level of detailed data. The stock reconstruction approach uses general life history data, as would be included in a yield model and complements it with observed fishery dependent and independent programs.
Dramatic changes in stock size occurred for the Hudson River stock over the past nearly one hundred year period from 1915 to the present. The long sustained high harvest from 1936 to 1955 indicates that the Hudson River shad stock size was several orders of magnitude larger than anything observed since then. In looking at all long-term data, the stock reconstruction model estimated initial stock size at nearly 11.3 million kg (Figure C7.A). The recent estimated vulnerable stock size (mature fish) for the Hudson averages, for 2001 to 2005, about 475,000 kg or about 245,000 fish (Figures C7.B and C7.C), lower than the range of the 1995 and 1996 population estimates reported by Hattala et al. (1998).
Both model fits, least squares fit of patterns of observed exploitation and use of maximum likelihood function tuned to recent observed indices are similar. The model fit to the exploitation patterns of the1950s and 1974 to 2005 tracks the trend in the observed indices of egg, PYSL and YOY abundance, CPUE of returning adults and exploitation rates (Figure C8). However, the maximum likelihood estimation (MLE) fit tends to exacerbate patterns of exploitation that occurred in the years preceding the collection of the observed indices. The MLE model fits the years from 1974 to 2005 reasonably well, but estimates excessive, and unrealistic, exploitation rates (exceeding100%) during the late1950s.
The stock reconstruction model estimates equilibrium yield of the stock to be approximately 590,000 kg at a
µmsy = 0.22 (Figure C9), equivalent to Fmsy of 0.25. This is essentially equal to the overfishing definition value F30=0.24 (BPR) from the Thompson-Bell model results. Estimated exploitation rates increased during 202
the early 1980s, peaked in the late 1990s and have since declined (Figure C8). Exploitation rates exceeded
µmsy starting in the mid 1980s.
Discussion Z-Estimates The most recent observed estimates of Z are higher than those observed in mid 1980s and result in F values that exceed our over fishing definition at most reasonable values of M. The possible weakness in our Z estimates is that they are based on age composition generated from scale samples. Ageing of scales remains an art and estimates have not been verified by known age fish. However, the same staff and methods have been used to age shad for the entire time period. Thus any bias should be consistent. The reduction in average age led to increased mortality estimates regardless of size of bias. Our estimates of current Z (see Section 7) and F are close to those generated by Deriso et al. (1995) by a stock reconstruction analyses.
Estimates of M used in the last coastwide assessment (ASMFC 1998) are considerably higher than what we suggested in this assessment (See Section 1.1.5). Estimates of Z in ASMFC (1998) were derived from cohort-based survival estimates from the age structure of fish collected only in the Connecticut River (Savoy and Shake 1993, 1994 and Savoy 1995). F was estimated from exploitation (the in-river commercial fishery harvest divided by an estimated population size). F was then subtracted from Z to obtain M.
However, it should be noted that these M estimates included all other sources of mortality other than the in-river fishery harvest of the Connecticut. The population estimate for the Connecticut stock has not been verified since the mid 1980s (see Section 6).
Ocean Harvest There are no data to estimate reporting rates for ocean harvest. Moreover, few sampling programs were conducted to directly identify stocks in the mixed stock harvest. We used tag data and knowledge of east coast migratory habits of shad for a first order approximation of harvest composition (see Section 1.5.1).
The resulting numbers are not, and will never be, without uncertainty, but they do provide a general perspective on stocks that were affected by mixed stock harvest.
Our potential spawning stock index suggests that losses occurred between the juvenile and adult stage.
Certainly, the mixed stock harvest is one possible source of such losses. This possibility is strengthened by the fact that differences between the predicted and observed Hudson River adult abundance indices increased at about the same time that the size of losses to the mixed stock harvest increased.
The PSSB we calculated is a relative measure of potential spawning stock biomass, an estimate of expected return to the Hudson. Over the course of the last fifteen years, dramatic increases in mortality occurred on the Hudson River=s adult shad stock. The ESSA bears out the difference observed between what returned to the Hudson versus what was expected (Figure C2). The two indices are in general agreement for the first two years, 1985 and 1986, but then the indices begin to diverge slowly, then more dramatically through the mid to late 1990s. The divergence of the indices suggests that the mortality occurred in ocean waters. The PSSB closely tracks the combined estimated ocean and Hudson River harvest - the Amissing@ difference being ocean harvest.
An easy mistake that can be made is under-estimating the effects of what could be considered a small amount of fishing. Total harvest of American shad from the mid-Atlantic to New England was barely 140,000 kg in 2005. Ocean harvest made up greater than 50% of the total harvest from 1988 to 2005. From Virginia to Maine only three systems, (Delaware, Hudson and Connecticut rivers) support in-river fisheries, 203
all other states allow no harvest. With stocks at historic lows, even small amounts of harvest have the ability to continue to drive stocks lower.
Other investigators (Savoy and Crecco 2004) suggested the major change in returning adults to the Connecticut River shad was due to increased predation by striped bass on the adults as they returned to the Connecticut each spring. We agree that predation could be a factor increase, but we feel it would be on young fish and not adults. The size of an adult American shad would mean that the predator on the order of 40 to 50 pounds or more to be able to consume a four to five pound fish as prey. Analyses conducted in Section 1.5.1-c indicated that adult American shad were only a minor dietary component of striped bass in the Hudson River and that there was no evidence for a long-term decline in adult shad abundance that coincided with an increase in striped bass.
Predation cannot explain the disappearance of older adult shad as observed in the Hudson. However, predation on young, over time could erode recruitment into the adult population. Crecco and Savoy (2004) observed that relative survival of juvenile shad to recruitment as virgins declined as striped bass abundance increased. They concluded that the decline was caused by predation on pre-recruits.
Herring are a more likely target and well-known prey item of striped bass. The blueback herring run dropped precipitously in the Connecticut system, prompting a closure on all take in 2002. The states of Massachusetts and Rhode Island closed their herring fisheries in 2006 due to decline in their runs. In the Hudson River and the Chesapeake Bay, changes in herring have only been recently recorded. The fish-lift counts of blueback herring at the Conowingo Dam fish-lift increased until 2001, but have since declined (R.
Sadzinski, pers. comm.).
Bycatch Another area of high uncertainty is the kill that occurs on immature, sub-adult fish that become part of the Atlantic coast=s Abait@ fishery. Removals of American shad at early ages (ages one through three) still remain unaccounted for. Hattala (memo to the ASMFC Shad & River Herring Technical Committee) reported to the shad technical committee that vessel trip reports obtained from NMFS more often have no shad listed as bycatch. Small shad landed as bait in state waters essentially disappear into any (NMFS) database recording unclassified bait. Given the amount of known bycatch of more lucrative species (i.e.
striped bass) in various fisheries on the coast, the likelihood of shad bycatch being equal to zero is low.
Stock Characteristics We feel that the changes we have seen in the Hudson River American shad stock are a result of over fishing. We base our conclusion on observed changes in size and age structure and on recent rates of mortality relative to acceptable levels.
Size and mean age decreased and remained low after 1991, relative to that in the mid 1980s. These changes could be caused by changes in year class production resulting in more young fish or in decreased survival of older fish. Increased fishing is the logical cause of any survival decrease. We tested effects of year class fluctuation on age structure by normalizing catch-at-age data by relative abundance of the same cohort at age zero. Resulting mean age (Section 7, Figure 7.13) continued to be lower in the most recent data suggesting that change was caused by actual losses of older fish rather than on year class fluctuation. A size-at-age decrease is one of the classic signs of overfishing where the fastest growing fish of each age class are removed, reducing overall size at a particular age (Hilborn and Walters 1992).
The changes in mean size-at-age is not limited to the loss of older fish, but also the result of fishing pressure over time altering what remains available at age. Fishing, especially gill-net fisheries, which are the primary 204
means of harvest of shad on the Atlantic coast, tend to remove the larger of any fish from a group of individuals. Weight-at-age data from the commercial fishery indicated a shift at age to smaller fish, the same shift in size was much more pronounced in spawning stock samples (see Section 7, Figures 7.11 and 7.12). If current data are used, the importance of river specific data, as used in yield modeling, becomes complicated by the greater over-whelming influence exerted by chronic fishing pressure that continually alters stock characteristics.
Future Projections Examples of an Unexploited Shad Stock: The Columbia River One real world example of stock growth of American shad is available from the Columbia River, on the Pacific coast. In 1871, Seth Green, a New York fish culturist, transported Hudson River American shad larvae across the United States and released the fingerlings into the Sacramento River (Ebbesmeyer and Hinrichsen 1997). The establishment, expansion and colonization of American shad along the Pacific coast is not all that well documented. However, shad reached the Columbia River system by about 1880.
Consistent record keeping has occurred from 1938 to the present for fish passage numbers at the major dams on the Columbia (ACOE and www.fpc.org). In addition to fish passage records, we also obtained data on the Columbia commercial fishery for shad (R. Hinrichsen and R. Beamesderfer pers. comm.), which allowed us to do some estimation of stock exploitation and growth. Figure C10 shows that from 1938 though 1960 the fishery sustained a fairly high level of exploitation on the stock, u~ 0.89 kept the stock at very low levels. Shad were actually shipped east and sold as Atlantic shad, during the time period when most east coast stocks were declining after being over-fished during WWII (McPhee 2002). In the early sixties, salmon became a more popular item in the markets and fishing pressure eased on shad.
The most dramatic increase in the Columbia shad stock occurred in the 1970s and 1980s as exploitation declined. We used the fish passage counts at the two major dams on the Columbia River, Bonneville and the Dalles Dams, to calculate the growth rate of the Columbia shad stock that use the upper river. The growth rate was about 6 percent per year and the passage numbers have grown to nearly three million fish.
Shad increasingly became known as the Atrash fish@ compared to the more lucrative salmon.
The dynamics of this west coast shad stock are poorly known. Shad are considered an exotic nuisance with most all resources focused on native salmon runs. Water temperatures increase slowly from 7ºC through 21ºC from April through early August and the runs can last three to four months, quite unlike the east coast.
Passage counts give us an estimate of numbers of fish that utilize the upper Columbia, above Bonneville Dam. However, below the dam is approximately 215 km of river in which shad could possibly spawn. The number of shad spawning in the lower river (below Bonneville Dam) is unknown. It is also not clear how many adult shad get out, or if they get out, of the Columbia=s dammed upper river. If adult fish do not get out, this may be part of the reason why the growth rate is so low.
One thing is for certain, the stock did not respond (in growth) until exploitation was reduced to a very low level. The current age structure of the Columbia stock might also provide a clue how old shad can get in a lightly unexploited population, if the fish are iteroparous and allowed to repeat spawn, and get out of the Columbia system after spawning.
Response of Hudson River Shad It is not known if the Hudson River shad stock will respond in a similar way as the Columbia stock did when exploitation was decreased. The Columbia stock=s response was slow, but may be dependent on a host of unknown factors about fish passage issue (i.e., downstream return of Columbia stock mentioned above).
205
Clupeid stocks are notorious for lack of response to restrictions in fish management changes (Hilborn and Walters 1992).
Conclusions and Recommendations We feel that the Hudson River American shad stock was overfished during the late 1980s and early 1990s and that this overfishing altered stock characteristics and reduced abundance. Moreover, overfishing appeared to be caused by the ocean mixed stock harvest that occurred beginning in the early 1980s. The Hudson River American shad stock is a shared resource along its entire migratory range, from North Carolina to Maine and Canada. As long as fisheries continue to operate in coastal waters, decisions on the fishery, and the direction it will take are also a shared process.
We recommend that fisheries that potentially affect the Hudson be restricted to reduce further damage to this stock. These fisheries include those in the Hudson River, both commercial and recreational take, as well as known remaining mixed stock fisheries outside of the Hudson River system. The remaining mixed stock fisheries include those in lower Delaware Bay. Other known bycatch fisheries should be regulated to minimize catch of American shad and a concerted effort made to identify bycatch in the other numerous fisheries that are implicated (e.g. Atlantic herring fishery).
Literature Cited Applied Science Associates. 2006. 2005 Year Class Report for the Hudson River Estuary Monitoring Program. Prepared for Dynegy, Mirant and Entergy . USA.
ASMFC. (Atlantic States Marine Fisheries Commission). 1998. American shad stock assessment peer review report. Washington, D.C.
ASMFC. (Atlantic States Marine Fisheries Commission). 2005. 2005 Stock assessment report for Atlantic striped bass. Washington, D.C.
Deriso, R., K. Hattala, and A. Kahnle. 2000. Hudson River shad assessment and equilibrium calculations.
Prepared through ESSA Technologies, Ltd. for the NY State Dept. of Environmental Conservation.
Toronto, Canada.
Ebbesmeyer, C., and R. Hinrichsen. 1997. Oceanography of the Pacific shad invasion. Shad Journal Vol 2 (1). Available: http://www.cqs.washington.edu/~hinrich/foundation.html.
Freidland, K.D., R.E. Haas, and T. F Sheehan. 1996. Comparative post smolt growth, maturation and survival in two stocks of Atlantic salmon. Fishery Bulletin 94:654-663.
Gibson, M., V. Crecco, and D. Stang. 1988. Stock assessment of American shad from selected Atlantic coast rivers. Special report #15 of the Atlantic States Marine Fisheries Commission, Washington, D.C.
Hattala, K., A. Kahnle, D.R. Smith, R. Jesien, and V. Whalon. 1998. Total mortality, population size and exploitation rates of American shad in the Hudson River Estuary. Special interim report to the Atlantic States Marine Fisheries Commission from the NY State Dept. of Environmental Conservation, New Paltz.
206
Kahnle, A., D. Stang, K. Hattala, and W. Mason. 1988. Haul seine study of American shad and striped bass spawning stocks in the Hudson River Estuary. Summary report for 1982-1986. NY State Dept. of Environmental Conservation, New Paltz.
Klauda, R.J., M. Nittel, K.P. Campbell. 1976. Commercial fishery for American shad in the Hudson River: fish abundance and stock trends. In: (USFWS) U.S. Fish and Wildlife Service and (NMFS) National Marine Fisheries Service. Proceedings of a workshop on American shad.
Amherst, Massachusetts.
McPhee, J. 2002. The Founding Fish. Published by Farrar, Straus and Giroux, New York.
NOAA Fisheries Toolbox Version 2.8, 2005. Virtual Population Analysis (VPA), A Stock Production model incorporating covariates (ASPIC), and an Age Structured Assessment Program (ASAP),
Available: http://nft.nefsc.noaa.gov.
Prager, M. 2006. Surplus production model. Available: http://www.sefsc.noaa.gov/mprager/.
Savoy, T. 1995. Anadromous fish studies in Connecticut Waters, Annual progress report, AFC-22-2, to US Dept of Commerce, NOAA, 1994-1995. Old Lyme, Connecticut.
Savoy, T., and D. Shake 1994. Anadromous fish studies in Connecticut Waters, Annual progress report, AFC-21-2, to US Dept of Commerce, NOAA, 1993-1994. Old Lyme, Connecticut.
Savoy, T., and D. Shake 1993. Anadromous fish studies in Connecticut Waters, Annual progress report, AFC-21-2, to US Dept of Commerce, NOAA, 1992-1993. Old Lyme, Connecticut.
Talbot, G. 1954. Factors associated with fluctuations in abundance of Hudson River shad. US Fishery Bulletin 101(56):373-413.
207
Table 1.5.2-C1. Survival estimates used as model inputs for Potential spawning stock abundance index for the Hudson River American shad stock.
Pre-recruits Recruits Age S at age Year S for Age 5+
1 0.19 1974 0.41 2 0.37 1975 0.35 3 0.74 1976 0.36 4 0.74 1977 0.43 1978 0.43 1979 0.44 1980 0.37 1981 0.43 1982 0.44 1983 0.44 1984 0.59 1985 0.63 1986 0.53 1987 0.57 1988 0.53 1989 0.58 1990 0.53 1991 0.38 1992 0.38 1993 0.51 1994 0.35 1995 0.24 1996 0.46 1997 0.65 1998 0.37 1999 0.32 2000 0.30 2001 0.33 2002 0.33 2003 0.32 2004 0.40 2005 0.44 208
Table 1.5.2-C2. F30 based on Egg (EPR) or Biomass (BPR) per recruit for the Hudson River shad stock.
Type 1 Type 2 F30 F30 Type of model run M Fmax EPR BPR EPR BPR Hudson River max age =14 0.30 0.20 0.22 0.24 0.23 0.25 Age specific 0.18 0.19 0.21 0.20 0.22 209
Table 1.5.2-C3. Various relative indices of spawning stock abundance for American shad of the Hudson River Estuary.
Sum of CPUE Spawning stock abundance Year weekly Age Age Age Ages ESSA ESSB ESSB PSSA PSSB CPUE Age five Age six Age seven Five six seven 5&6 Age (5-7) Age (5-7) est* (3-10) (3-10) 1980 19.59 7.62 7.70 2.74 1981 14.47 3.50 5.38 2.93 1982 8.02 2.19 2.28 1.71 1983 9.16 3.16 2.95 0.94 1984 9.49 2.03 3.53 1.71 1.17 2.05 1985 26.65 6.95 8.17 5.19 40.82 28.45 28.71 33.05 31.82 64.3 1.92 3.47 1986 52.27 15.84 16.57 8.92 55.74 50.23 67.62 52.78 55.40 112.9 2.28 4.28 1987 47.37 8.22 13.57 9.66 25.47 55.76 73.01 38.49 45.03 89.3 2.51 4.89 1988 42.22 9.03 12.86 10.42 30.52 37.58 56.59 34.30 39.30 80.2 2.33 4.71 1989 33.79 7.47 12.24 8.88 25.83 46.36 38.88 35.62 36.57 69.2 2.36 4.85 1990 16.61 3.54 5.14 4.05 10.84 22.91 15.25 15.76 15.60 25.9 2.61 5.04 1991 18.31 5.27 5.54 4.02 16.15 15.47 27.27 15.80 17.83 26.6 1.96 3.82 210 1992 14.61 6.25 4.86 2.15 16.05 13.09 13.95 14.61 14.49 20.9 1.43 2.65 1993 13.02 3.36 4.83 2.31 8.43 13.49 18.99 10.82 11.95 14.2 2.04 3.65 1994 24.24 6.46 10.62 5.54 10.16 43.04 60.95 19.34 23.23 30.1 2.14 3.64 1995 11.49 4.35 4.43 1.89 9.12 15.14 25.04 11.41 12.63 19.1 1.32 2.02 1996 20.25 8.80 6.96 2.39 20.23 40.02 34.30 25.89 26.75 41.5 1.36 2.01 1997 7.11 3.28 2.05 0.87 7.69 11.82 10.88 8.88 9.12 13.6 1.87 2.84 1998 12.23 5.75 3.20 1.29 11.50 17.36 23.93 13.08 13.87 21.5 1.22 1.83 1999 10.81 4.77 3.44 1.28 8.17 22.74 18.21 11.17 11.79 16.6 18.5 0.74 1.09 2000 31.61 14.29 10.40 3.40 24.28 79.84 130.57 34.35 37.71 58.0 25.3 0.51 0.73 2001 25.74 9.97 9.01 3.82 24.49 31.74 41.06 27.47 29.08 43.7 9.3 0.43 0.59 2002* 8.3 0.31 0.47 2003* 15.9 0.27 0.39 2004* 8.1 0.32 0.45 2005* 9.9 0.45 0.64
- Data from 2002-2005 insufficient to generate ESSA index, see Section 7
Figure 1.5.2-C1. The potential spawning stock abundance (PSSA) and biomass (PSSB) indices for Hudson River American shad.
6.00 PSSA 5.00 PSSB 4.00 PSSA/PSSB 3.00 2.00 1.00 0.00 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 Figure 1.5.2-C2. Comparison of the Hudson River empirical (ESSB) and potential (PSSB) spawning stock biomass indices.
PSSB 3.50 ESSB 3.00 PSSA ESSA 2.50 2.00 SSB 1.50 1.00 0.50 0.00 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 Year 211
Figure 1.5.2-C3. Comparison of the Hudson River empirical (ESSB) spawning stock biomass index with in-river harvest.
400000 120 River harvest 350000 ESSB 100 300000 80 250000 Harvest 200000 60 ESSB 150000 40 100000 20 50000 0 0 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 Year 450000 y = 3412x + 22648 400000 R2 = 0.8366 p=6.06E-09 In River-Harvest-kg 350000 300000 250000 200000 150000 100000 50000 0
0.0 20.0 40.0 60.0 80.0 100.0 120.0 ESSB 212
Figure 1.5.2-C4. Comparison of the Hudson River potential (PSSB) spawning stock biomass index with estimated total (river and mixed stock) harvest, with correlation between the PSSB and harvest.
6.00 600000 PSSB 5.00 500000 Total harvest 4.00 400000 Harvest-kg SSB 3.00 300000 2.00 200000 1.00 100000 0.00 0 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 Year 600000 500000 Total Harvest-kg 400000 300000 y = 67668x + 106965 R2 = 0.6994 200000 p=1.2E-06 100000 0
0.00 1.00 2.00 3.00 4.00 5.00 6.00 PSSB 213
Figure 1.5.2-C5. Comparison of observed fishing mortality rates based on catch curve on age and catchcurve on repeat spawning versus selected overfishing rate.
Observed fishing rates based on CC on age, M=0.3 1.40 Male 1.20 Female F30(EPR)
F30(BPR) 1.00 0.80 F
0.60 0.40 0.20 0.00 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 Observed fishing rates based on CC on spawning marks, M=0.3 1.40 Male 1.20 Female F30(EPR)
F30(BPR) 1.00 0.80 F
0.60 0.40 0.20 0.00 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 214
Figure 1.5.2-C6. Outputs of ASPIC (surplus production) model run for the Hudson River American shad stock.
70 60 Observed CPUE Estimated CPUE 50 CPUE 40 30 20 10 0
1931 1936 1941 1946 1951 1956 1961 1966 1971 1976 1981 1986 1991 1996 2001 2
1.5 1
CPUE residuals (log scal 0.5 0
-0.5 1931 1936 1941 1946 1951 1956 1961 1966 1971 1976 1981 1986 1991 1996 2001
-1
-1.5
-2
-2.5
-3 215
Figure C7. Estimated a) stock size and harvest (kg-millions) of Hudson River American shad 1915-2005; b) 1970-2005; c) estimated stock size (N) 1970-2005.
14.00 A 12.00 Stock 10.00 Harvest Kg-millions 8.00 6.00 4.00 2.00 0.00 B
1.60 1915 1922 1929 1936 1943 1950 1957 1964 1971 1978 1985 1992 1999 1.40 1.20 Stock Harvest Kg-millions 1.00 0.80 0.60 0.40 0.20 0.00 1970 1973 1976 1979 1982 1985 1988 1991 1994 1997 2000 2003 900000 800000 C 700000 600000 Stock size - N 500000 400000 300000 200000 100000 0
1970 1973 1976 1979 1982 1985 1988 1991 1994 1997 2000 2003 216
Figure 1.5.2-C8. Observed data and stock reconstruction model predictions of abundance of eggs, YOY, predicted stock and CPUE, and exploitation rate for Hudson River American shad.
Eggs Juveniles obs obs 6.00 2.5 pred pred 5.00 2
4.00 1.5 3.00 1
Egg Abund Age 0 abund 2.00 1.00 0.5 0.00 0 1974 1977 1980 1983 1986 1989 1992 1995 1998 2001 2004 1974 1977 1980 1983 1986 1989 1992 1995 1998 2001 2004 217 Exploitation Stock biomass obs 0.80 obs 3.5 pred 0.70 pred 3
0.60 2.5 0.50 2
u 0.40 SSB 1.5 0.30 1 0.20 0.5 0.10 0 0.00 1974 1977 1980 1983 1986 1989 1992 1995 1998 2001 2004 1974 1977 1980 1983 1986 1989 1992 1995 1998 2001 2004
Figure 1.5.2-C9. Equilibrium catch versus exploitation rate of Hudson River American shad.
1.40 Equilibirum catch (millions- kg 1.20 1.00 0.80 0.60 0.40 0.20 0.00 0.00 0.10 0.20 0.30 0.40 0.50 0.60 Exploitation rate Figure 1.5.2-C10. Comparison of fisheries harvest versus run size of American shad in the Columbia River, Oregon and Washington.
6000000 1.20 Run size Fishery 5000000 1.00 u
4000000 0.80 N: fishery & run 3000000 0.60 u 2000000 0.40 1000000 0.20 0 0.00 1938 1943 1948 1953 1958 1963 1968 1973 1978 1983 1988 1993 1998 2003 u=0.89 u=0.29 u=0.15 u=0.08 F=2.17 F=0.35 F=0.16 F=0.08 218
APPENDIX I Formulae Used in Biomass Model Analyses Authors:
Kathryn A. Hattala and Andrew W. Kahnle Hudson River Fisheries Unit, New York State Department of Environmental Conservation 21 South Putt Corners Road, New Paltz, New York 12561 The Thompson Bell Yield- per-recruit (YPR) model, modified for biomass-per-recruit with the addition of weight-at-age and fecundity for egg-per-recruit, was calculated as follows:
N * µ* W n
j j j=t YPR =
R (1)
Where: YPR = lifetime yield (kg)-per-recruit n = Maximum age in the population t = Age of first recruitment (age 3 for females)
Nj = Number of individuals at the start of year j Wj = Mean weight (kg) of individuals at the start of year j
µ = Exploitation rate R = Number of recruits at age one Mortality was modeled using:
Nj+1 = Nj * µ Where: Nj+1 = Number of fish alive at age j+1 Nj = Number alive at age j Exploitation, µ, was calculated as follows:
Type 1 fishery: µ = =1-exp (-F j)
Type 2 fishery: µ = (F j
- Aj) / Zj Where: Fj = Fishing mortality rate from j to j+1 Aj = Total mortality rate from j to j+1, calculated as 1-S, where S = exp(-Zj)
Zj = Fj + Mj (Total mortality)
Mj = Natural mortality rate from j to j+1 Vulnerability to the fishery was age based, calculated from observed data obtained from various river systems.
Age invariant natural mortality was obtained from the formula from Hoenig (1983):
Loge M = 1.46 - 1.01
- Loge (TMAX)
Where: M = instantaneous rate of natural mortality TMAX = maximum age of the fished stock 219
Natural mortality-at-age was also calculated from observed weight-at-age data for the Hudson, as part of the sensitivity analysis, using methods of Boudreau and Dickie 1989 and Dickie 1987. Weight (in lbs)-at-age was converted to kcal by multiplying by 592.
M = 2.88 * (weight-kcal-at-age)0.33 The model was run at fishing rates (Fj) of zero to 0.7 in 0.01 increments.
Number harvested-at-age was converted to weight by multiplying numbers by weight-at-age. Weight-at-age was estimated using the various growth functions. Gompertz (Hudson only):
Wt = W0
- exp {G * [1-exp(-g*t)]}
Where: Wt = Weight at age t W0 = Weight at time t0 G = Instantaneous growth rate at time t0 g = rate of decrease of G Egg-per-recruit (EPR) was calculated as:
N
- P
- p *G n
j=t j j j j (7) 10
-6 EPR =
R Where: EPR = Lifetime egg deposition-per-recruit n = Maximum age in the population t = Age of first maturity in females Nj = Number of females at age j Pj = Proportion of females mature at age j Gj = Mean fecundity of age j females R = Number of recruits at age one Biomass-per-recruit (BPR) was calculated as:
(8)
N
- P *W
- G n
j=t j j j j 10
-6 BPR =
R Where: BPR = Lifetime biomass of spawning stock-per-recruit n = Maximum age in the population t = Age of first maturity in females Nj = Number of females at age j Pj = Proportion of females mature at age j Wj = Mean weight (kg) of individuals at age R = Number of recruits at age one Maturity schedule for female American shad were calculated from observed age and or repeat spawning data, to estimate proportion mature-at-age.
220
APPENDIX II York River, Virginia Partial Recruitment Vector Analyses Kristen A. Delano, Robert J. Latour and John E. Olney School of Marine Science, Virginia Institute of Marine Science Route 1208 Greate Road, P.O. Box 1346, Gloucester Point, Virginia 23062 Introduction There are two types of recruitment vectors, or selectivity values, that can be calculated. The first calculation is a relative selectivity that estimates the selectivity of the gear for each age group with respect to the most abundant age group, so that the age group caught with the greatest frequency has a selectivity of 1 and all other groups are some fraction of that. An absolute selectivity value can be determined by correcting for the effects of mesh size on the relative proportions of the different age groups. This selectivity value is calculated using catch-at-age data from a selective gear, drift gill net (DGN), and a non-selective gear, pound net (PN), determining the proportion-at-age for each age group in each gear, dividing each age groups proportion in the selective gear by the proportion found in the non-selective gear, and finally, dividing each age groups corrected proportion-at-age by the value of the group with the highest value.
Datasets Concurrent York River DGN and PN data were available for 1998-2000. Drift gill-net samples were obtained from tribal fisherman on the Mattaponi and Pamunkey Rivers, which constitute the headwaters of the York River system. Mattaponi and Pamunkey samples were combined within years to represent an overall York River sample since it is unclear if the stocks are separate down to the tributary level. Pound net samples were obtained from Mr. Green at the mouth of the York River. Mr. Green fished five separate pound nets at the mouth of the York River, but all nets were within the same river kilometer. Catches from all five nets were combined within years to represent an overall Green pound net sample for that year. One caveat of using Mr. Greens nets was the possibility of mixed stock catches. These nets were located at the mouth of the York River where it empties into the main stem Chesapeake Bay and upper Bay stocks could possibly be caught along with York River fish as they migrate past the nets to more northern rivers. Since the drift gill-net mesh sizes target maturing female fish, only female fish were used from the pound net samples.
None of our DGN or PN samples were aged so an age-length key was developed for each of the gears. As much as possible, care was taken to select datasets for the keys from gears that were located in spatially and temporally similar conditions to those of the samples to which they were being applied. Fork length was used instead of total length since total length can be greatly affected by fin injury and degradation.
Age-length key To determine whether fork length of fish varied by year, necessitating individual analyses by year, a 1-way ANCOVA was run for the years 1998 to 2003 on the female York River stake gill-net monitoring data.
Table AII.1 One-way ANCOVA results for York River American shad fork length (mm) versus year.
Dependent variable=year, independent variable=fork length (mm), covariate=ScaleAge.
Source DF Seq SS Adj SS Adj MS F P ScaleAge 1 12802 8452 8452 14.86 0.000 Year 5 99193 99193 19839 34.89 0.000 Error 2172 1235119 1235119 569 Total 2178 1347115 221
After removing the effects of varying age structure on size (P<0.0005), the 1-way ANCOVA was significant for year effects (P<0.00005). Due to significant year effects, age-length keys were created for each year separately.
We developed individual age-length keys for 1998, 1999, and 2000 York River Indian DGN samples using VIMS stake gill-net (SGN) data from the same years. Stake gill-net samples were collected downriver from the DGN sites at the VIMS monitoring net and aged by one reader. In some cases, fish were dropped from the DGN sample when age-length values were not available in the SGN key.
Additionally, when the age-length key required a DGN sample to be separated into more age groups than there were fish in the sample, we attempted to allocate the samples between the most abundant age groups. If age groups had the same abundance, the older age group was preferred. Since mesh sizes of nets used in the drift gill-net fishery were the same in 1998 through 1999 (5.5 mesh), but not 2000 (5.25), selectivity values for 2000 should not be used alongside or combined with the 1998 and 1999 values.
Since no aged pound net samples were available from 1998, 1999, and 2000, a pound net age-length key for Mr. Greens PN samples was developed from aged samples collected in Mr. Kellums PN in 2001 and 2002. Mr. Kellums PN samples were collected upriver from Mr. Greens PN samples, had two different scale readers (2001=Maki, 2002=Watkins), and since Mr. Kellums samples were from different years than Mr. Greens samples, they were combined to generate the key values applied to Mr. Greens samples. Again, some fish were dropped when age-length values were not available in the key and when the age-length key required a sample to be separated into more age groups than there were fish in the sample, we attempted to allocate the samples between the most abundant age groups; if age groups had the same abundance, the older age group was selected.
The age-length keys were applied to the Indian drift gill-net samples and the Green pound net samples from 1998 to 2000 and age-frequencies were determined for each gear and year.
Selectivity Relative selectivity- The first selectivity calculation involved determining the percent contribution of each age class to the total sample and then dividing that value by the maximum of those values, so that the age class constituting the highest proportion of the catch had a selectivity value equal to 1 and all other age classes were some fraction of that:
Table AII.2 Number-at-age for the Mattaponi and Pamunkey River combined drift gill-net dataset.
Mesh size of the DGN is included in parentheses next to the year.
Age (years)
Year 3 4 5 6 7 8 9 10 1998 (5.5) 0 29 38 18 9 6 1 0 1999 (5.5) 2 17 88 48 10 4 0 0 2000 (5.25) 9 113 158 108 21 1 1 1 Table AII.3 Proportion-at-age for the Mattaponi and Pamunkey River combined drift gill-net dataset.
Mesh size of the DGN is included in parentheses next to the year.
Age (years)
Year 3 4 5 6 7 8 9 10 1998 (5.5) 0 0.29 0.38 0.18 0.09 0.06 0.01 0 1999 (5.5) 0.01 0.10 0.52 0.28 0.06 0.02 0 0 2000 (5.25) 0.02 0.27 0.38 0.26 0.05 0 0 0 222
Table AII.4 Partial recruitment vector values for the York River drift gill-net fishery, by age class and year. Mesh size of the DGN is included in parentheses next to the year. (Figure B1)
Age (years)
Year 3 4 5 6 7 8 9 10 1998 (5.5) 0 0.76 1.00 0.47 0.24 0.16 0.03 0 1999 (5.5) 0.02 0.19 1.00 0.55 0.11 0.05 0 0 2000 (5.25) 0.06 0.72 1.00 0.68 0.13 0.01 0.01 0.01 Absolute selectivity- The second selectivity calculation involved correcting the drift gill-net proportion-at-age values for gear mesh size selectivity using concurrent, non-selective pound net data. The proportion-at-age for all age classes present in the Green pound net data was calculated:
Table AII.5 Number-at-age for Mr. Greens York River pound net data. Mesh size of the DGN is included in parentheses next to the year.
Age (years)
Year 3 4 5 6 7 8 9 10 1998 (5.5) 0 69 88 40 29 3 0 0 1999 (5.5) 0 168 272 159 96 11 0 0 2000 (5.25) 0 24 41 28 16 2 0 0 Table AII.6 Proportion-at-age for Mr. Greens York River pound net data. Mesh size of the DGN is included in parentheses next to the year.
Age (years)
Year 3 4 5 6 7 8 9 10 1998 (5.5) 0 0.30 0.38 0.17 0.13 0.01 0 0 1999 (5.5) 0 0.24 0.39 0.23 0.14 0.02 0 0 2000 (5.25) 0 0.22 0.37 0.25 0.14 0.02 0 0 For a given age class, the proportion-at-age from the DGN sample was divided by the proportion-at-age from the PN sample to adjust the DGN proportions to the non-selective proportions found in the pound net samples:
Table AII.7 Corrected proportion-at-age for the York River drift gill-net fishery, calculated by dividing the DGN proportion-at-age by Greens PN proportion-at-age. Mesh size of the DGN is included in parentheses next to the year. - indicates no value due to lack of age class in pound net sample.
Age (years)
Age 3 4 5 6 7 8 9 10 1998 (5.5) - 0.95 0.98 1.02 0.70 4.53 - -
1999 (5.5) - 0.42 1.35 1.26 0.44 1.52 - -
2000 (5.25) - 1.27 1.04 1.04 0.35 0.13 - -
A final, corrected selectivity value was determined by dividing each of the adjusted proportion-at-age values by the maximum of those values, so that the age group constituting the highest proportion of the catch was equal to 1 and all other age classes were some fraction of that (Table B8). Note the absence of values in the smallest and largest age classes due to truncation of the sample during the age-length key conversion.
223
Table AII.8 Absolute selectivity values for the York River drift gill-net fishery, by age class and year.
Mesh size of the DGN is included in parentheses next to the year. - indicates no value due to lack of age class in pound net sample. (Figure B1)
Age (years)
Year 3 4 5 6 7 8 9 10 1998 (5.5) - 0.21 0.22 0.23 0.16 1.00 - -
1999 (5.5) - 0.28 0.89 0.83 0.29 1.00 - -
2000 (5.25) - 1.00 0.82 0.82 0.28 0.11 - -
Figure AII.1 Relative and absolute selectivity values for the York River drift gill-net fishery using year specific age-length keys for ageing samples.
Relative selectivity 1.2 1998 (5.5) 1 1999 (5.5)
Selectivity 0.8 2000 (5.25) 0.6 0.4 0.2 0
3 4 5 6 7 8 9 10 Age Category Absolute Selectivity 1998 (5.5) 1999 (5.5) 1.2 2000 (5.25) 1 Selectivity 0.8 0.6 0.4 0.2 0
3 4 5 6 7 8 9 10 Age Category 224