Project Title : Genetic Identification of Atlantic Cod Spawning Stocks in U . S . Waters

In this collaborative research project, we formed a partnership of commercial fisherman and scientists from UNH and NYU to investigate stock definitions for Atlantic cod using DNA markers. Cod in U.S. waters are currently managed as two stocks: 1) a Gulf of Maine stock and 2) a Georges Bank and south stock. This designation is decades old and warrants reevaluation in light of concerns that fisheries management units may not reflect biologically meaningful population units. To address this, we used 10 microsatellite and 6 SNP markers to characterize the population genetic structure of cod in U.S. waters. We found significant differentiation among temporally and spatially divergent populations of cod (global FST = 0.0044, G ́ST = 0.0144), primarily stemming from two non-neutral loci, and strong evidence for a population structure that contradicts the current two-stock management model. Our results indicate that cod in U.S. waters are broadly structured into three groups: 1) a northern spring spawning coastal complex in the Gulf of Maine (GOM), 2) a southern complex consisting of winter-spawning inshore GOM, offshore GOM and sites south of Cape Cod, MA, and 3) a Georges Bank population. The strongest differentiation occurs between populations in the northern and southern complex (FST = 0.0054 0.0156), some of which spawn in the same bays in different seasons. This population genetic structure is stable over a 5-year period. We suggest a model of population structure that is maintained by geographic and seasonal differences in larval recruitment and differential adult life history strategies; local ecological adaptations may also be important. Our findings contribute to a growing body of knowledge that cod and other marine fish populations are structured on a finer scale than previously thought and that this structure supports biocomplexity and locally adapted populations. As such, it may be warranted to re-evaluate current management units and tailor management plans toward this finer scale. Introduction: Atlantic cod (Gadus morhua) is a principal component of the Northeast groundfish assemblage, and as such has supported tremendously productive commercial and recreational fisheries in the region for years. The social and economic importance of this fish is highlighted by Kurlansky (1997), who titled his history of the North Atlantic cod fisheries: “Cod: A Biography of the Fish That Changed the World.” Cod’s very desirability has led to rampant over harvesting across its range. North American cod stocks are in particularly poor shape. The recent collapse of the Canadian cod fisheries (Hutchings and Myers 1994) led to massive unemployment and social upheaval in Newfoundland. Within U.S. waters, cod stock biomass, commercial landings, and recruitment have fluctuated dramatically over the past four decades, with some of the lowest estimates to date recorded by the late 1990s. Today, many New England fishermen who rely heavily on cod are struggling to remain financially solvent as fishery management plans aimed at rebuilding stock biomass restrict harvesting. A significant complication in the management of the Atlantic cod, and many other fishes, is uncertainty of their stock structure (i.e., number of stocks within a region, boundaries of stocks, and extent of stock mixing). In general, informed fisheries management should be built on detailed knowledge of the stock structure of harvested species, and should be based on discrete unit stocks (Cushing 1968). Management regimes should be tailored specifically for each stock to reflect differences in recruitment, growth, and mortality. This strategy has been called for Atlantic cod (Ames 2002), but has not yet been implemented because of a void of information on its stock structure in U.S waters. To this end, the goal of this project is to address the uncertainty that exists about the stock structure and number of discrete spawning populations of Atlantic cod in U.S. waters. Interim Final Report to Northeast Consortium, September 2009: Berlinsky et al. Genetic Identification of Atlantic Cod Spawning Stocks in U.S Waters 3 Management of the cod fishery in U.S. waters usually assumes a two stock model: 1) the Gulf of Maine (GOM), and 2) Georges Bank and south. The empirical verification of this model is limited, however, and the studies upon which it is founded are decades old. This division is primarily based on limited and dated tagging studies reported in 1902, the 1930s and 1962 (Serchuk and Cohen 1990), frequencies of parasitic infestation (Sherman and Wise 1961), maximum spawning time data (Colton et al. 1979), growth rate variation (Penttila and Gifford 1976), and differences in age and length at sexual maturity (Fahay 1998). Additionally, patterns of recruitment of the Georges Bank stock are generally different from those observed in the GOM stock (Serchuk and Cohen 1990). These diagnostic biological characteristics are ecophenotypic in nature, however, and the markrecapture results were generated when these stocks were much larger and were not based on actively spawning fish tagged on the spawning grounds. During the past 50 years, cod stocks on Georges Bank and within the GOM have undergone tremendous fluctuations in abundances, which could compromise the applicability of earlier studies to current management. In recent years, both stocks have demonstrated decreases in landings, spawning stock biomass, and year class strength, although specific trends in these indices often differ between the two putative stocks. There is also much uncertainty regarding fine scale structuring of stocks within the GOM. Spawning within this stock occurs throughout the year with the exception of the summer months. Many areas have two or more spawning events within a single year (Berrien and Sibunka 1999, Howell et al. 2008) affording the possibility of more than one stock at a single geographic locale. Ames (2002) and others have reported that historically there were 91 spawning grounds in the GOM extending from Cape Ann in the west to Lurcher Shoals to the east. Today, spawning is only observed at approximately 40 of these locales, with the greatest reduction in activity observed at those sites in Downeast Maine (Berrien and Sibunka 1999). Ames (2002) proposed the historical existence of four major subpopulations of cod within the GOM, including a Western subpopulation from northern Massachusetts Bay to Sheepscot Bay, a Midcoast subpopulation in the area of Penobscot Bay, a Downeast subpopulation in the Mt. Desert Island area, and a Bay of Fundy subpopulation. However, evaluation of the genetic discreteness of these subpopulations has not been attempted prior to this study. Genetic stock identification provides a sensitive approach to evaluate the stock structure of species in which spawning populations have been reproductively isolated for considerable times and between which contemporary gene flow is very low or non-existent. Genetic stock identification at selectively neutral loci (e.g. microsatellite DNA and SNP markers) offers a major advantage over other approaches in that genetic stock signatures are not subject to environmental influences and therefore remain stable over many generations. Once diagnostic genetic markers have been identified and their temporal stability over several generations confirmed, spawning stocks need not be revisited on a regular basis to reconfirm genotype frequencies. After the stock structure of spawning aggregations has been defined, mixed stock analysis can be performed to describe mixing and migration of fishes from spawning areas to other areas. Furthermore, the genetic data can complement other biological data collected from the same populations, such as acoustic monitoring and tagging data, to better understand population structuring processes, such as migrations and spawning site fidelity. Interim Final Report to Northeast Consortium, September 2009: Berlinsky et al. Genetic Identification of Atlantic Cod Spawning Stocks in U.S Waters 4 With the goal of characterizing the genetic stock structure of cod in US waters, our project responded directly to recent stock assessments that reported overfishing and decline of both the GOM and Georges Bank stocks (May and Col 2002, O’Brien et al. 2002, NEFMC 2005). The depressed state of cod populations necessitates stock-specific management plans to maintain fishery sustainability; better information about cod stocks is crucial to effective rebuilding plans. The need for applying genetic and other biological data to re-evaluate stock boundaries and management units of commercially important fish species worldwide has been highlighted recently (Cadrin and Secor 2008, Hauser and Carvahlo 2008, ICES 2009, Reiss et al. 2009). This project has involved commercial fisherman, researchers from UNH and NYU, and other partners (UMass-Dartmouth, Mass Division of Marine Fisheries) in a collaborative effort to plan and implement a robust sampling regime for Atlantic cod. Commercial fishermen’s knowledge of the location and method of capture of spawning cod was critical to the success of this project. Both the goals and approaches of this project were directly relevant to programmatic goals of the Northeast Consortium. Our project addresses the topic area “commercial harvest and species sampling.” Most specifically, our project addresses one of the highest priority research needs of the New England Fishery Management Council, which calls for "investigations into stock definition, stock movements, mixing, and migration, such as through tagging studies, DNA markers, morphological characteristics, etc" for Northeast Multispecies (Groundfish) assemblages. Our project also contributes to a number of the priorities outlined in the Northeast Fisheries Science Center section of the NMFS FY 2004-2009 strategic research plan (NMFS, 2004): “Determine the biological, environmental, and habitat processes controlling the reproductive success of important fishery resources.” “An outgrowth of the stock assessment peer review is continuing investigations of the appropriate assessment methodologies for Georges Bank/GOM cod stocks.” Project Objectives: The overarching goal of our project is to identify the spatial and temporal stock structure of Atlantic Cod (Gadus morhua) in US waters, using genetic analyses. Specifically, our objectives are to 1) Identify and sample major spawning aggregations throughout the Gulf of Maine, Georges Bank, and areas south of Georges Bank. 2) Determine if cod stocks on Georges Banks, GOM, and south of Georges Bank are genetically unique. 3) Characterize spatial or temporal genetic heterogeneity among spawning aggregations within these broad geographic areas (GOM, Georges Bank, and south). Overall Hypothesis: Cod spawning aggregations in northeast U.S. waters are genetically differentiated. Participants: Scientists 1) David Berlinsky, Associate Professor, Department of Biological Sciences, University of New Hampshire, 38 College Road, Durham, NH 03824; (603) 862-0007; [email protected] 2) Adrienne Kovach, Assistant Research Professor, Department of Natural Resources, University of New Hampshire, Durham, NH 03824; (603) 862-1603; [email protected] Interim Final Report to Northeast Consortium, September 2009: Berlinsky et al. Genetic Identification of Atlantic Cod Spawning Stocks in U.S Waters 5 3) Isaac Wirgin, Associate Professor, Dept. of Environmental Medicine, NY University School of Medicine, 57 Old Forge Road, Tuxedo, NY 10987; (845) 731-3548; [email protected] 4) Lorraine Maceda, Dept. of Environmental Medicine, NY University School of Medicine, 57 Old Forge Road, Tuxedo, NY 10987; (845) 731-3548; [email protected] Steering Committee 5) Steven Cadrin, NOAA Northeast Fisheries Science Center, Cooperative Marine Education & Research, School for Marine Science & Technology, 200 Mill Road, Suite 32, Fairhaven, MA 02719 U.S.A. (508) 910-6358; [email protected] 6) Ken LaValley, Extension Specialist Commercial Fisheries Technology, Nesmith 219; University of New Hampshire, Durham, NH 03824; (603) 862-4343; [email protected] 7) Ted Ames, Penobscot East Resource Center, 43 School Street Room 1E PO Box 27, Stonington, Maine 04681; (207) 367-2708; [email protected] Commercial Fishermen (advised and/or assisted with sample collection): 8) David Goethel, F/V Ellen Diane, Hampton, NH & New England Fishery Management Council member 9) Proctor Wells, F/V Tenacious, Phippsburg, ME 10) Frank Mirarchi, F/V Barbara L. Peters, Scituate, MA 11) Chris Odlin F/V Lydia and Maya, Portland ME 12) Carl Bouchard F/V Stormy Weather, Hampton, NH 13) Jeff Carver, F/V Saint Joseph, Scituate, MA Methods: Sampling: During December 2005 – July 2008, 1581 adult cod were captured via otter trawl, gill net or hook and line, in collaboration with commercial and recreational fishermen and fishery biologists. Spawning condition was assessed by visual inspection of the gonads or observations of running milt or eggs using the National Marine Fisheries Service ovarian staging criteria. Caudal fin clip samples (1 cm) were taken from each fish and preserved in 95-100% ethanol. Spawning cod were sampled from the following sites: northeast peak of Georges Bank, the inshore Gulf of Maine in Ipswich Bay, Massachusetts Bay, and Bigelow Bight, ME, the offshore Gulf of Maine at Jeffrey’s Ledge and Stellwagen Bank, and south of Cape Cod from Nantucket Shoals, and Cox Ledge. At Ipswich Bay, Massachusetts Bay and Cox Ledge, distinct spawning aggregates were identified and sampled in both the spring and winter. Additionally, adult fish not in spawning condition (resting individuals) were sampled from Ipswich Bay, Platts Bank and New York Bight. Six of the spawning aggregates were sampled in 2 subsequent years, enabling us to test for stability in the structure (see Table 1 in Appendix and Fig. 1, below). Interim Final Report to Northeast Consortium, September 2009: Berlinsky et al. Genetic Identification of Atlantic Cod Spawning Stocks in U.S Waters 6 Figure 1. Sample site locations for Atlantic cod collected in this study. Abbreviated sample site names refer to Table 1 of Appendix. Project Design: The collection sites targeted in this project followed our initial proposed sampling scheme, which built upon our prior NEC-funded research (Wirgin et al. 2007): actual collection sites were influenced by discussions with steering committee members (led by Dr. Steve Cadrin, NMFS/SMAST) and availability of spawning cod. This led to some additional collection opportunities beyond our original proposal, including a site on Nantucket Shoals and separate spring and winter spawning aggregations on Cox Ledge (resulting in 3 collection sites in the south, rather than the 1 originally proposed). We also identified and sampled an unanticipated spring spawning aggregate in Massachusetts Bay (in addition to the winter spawning fish in that bay). Offshore sites of Stellwagen Bank and Jeffreys Ledge were also included in our GOM collections (although not in our original proposal), due to the potential for differentiation among inshore and offshore populations. Limitations: We were unable to locate spawning (or any adult) fish in eastern Maine, due to a lack of fishing effort on these depleted populations. Our northern/eastern most collection site, therefore, was Bigelow Bight. We were also initially unable to obtain a sufficient sample of spawning cod from the western Georges Bank area, despite targeted collection trips; in 2009 we were successful in sampling this site and analyses of these samples are ongoing. We also found that the winter spawning site in Ipswich Bay, which we characterized in our previous Northeast Consortium-sponsored research (Wirgin et al. 2007), has not been an active spawning site in the last several years; instead, these fish are thought to be aggregating and spawning in Massachusetts Bay during this time (December-January; D. Goethel, personal communication). We therefore sampled winter-spawning fish in Massachusetts Bay and resting fish in Ipswich Bay in the winter. NYS, NYW GBW