Mapping the Distribution of Atlantic Cod Spawning on Georges Bank Using Fishermen’S Ecological Knowledge and Scientific Data

Mapping the Distribution of Atlantic Cod Spawning on Georges Bank Using Fishermen’s Ecological Knowledge and Scientific Data

Co-Principal Investigators: Dr. Gregory DeCelles1,2*, David Martins1,2, Dr. Douglas R. Zemeckis1, Ted Ames3, and Dr. Steven X. Cadrin1

1: School for Marine Science and Technology, University of Massachusetts Dartmouth, 200 Mill Road, Suite 325, Fairhaven, MA 02719 2: Massachusetts Division of Marine Fisheries, 1213 Purchase Street, New Bedford, MA 02740 3: Penobscot East Resource Center, 13 Atlantic Ave, Stonington, ME 04681 *Contact Author: [email protected], 508-990-2860 x 115

Project Dates: 7/1/2015 through 11/30/2016 Project Amount: $74,536

Table of Contents

Abstract 1 Introduction 2 Georges WINTERBank Cod FLOUNDER STOCK STRUCTURE 2 Purpose WINTERof This Study FLOUNDER STOCK STRUCTURE 3 Fishermen's Ecological Knowledge 6 Project Objectives 6

Phase 1: Analysis of Scientific Data Sources 8 Analysis WINTERof Historical FLOUNDER Reports STOCK STRUCTURE 8 Northeast WINTER Fisheries FLOUNDER Science Center STOCK Trawl STRUCTURE Survey 10 Canadian Deprtment of Fisheries and Oceans Trawl Survey 17 Canadian Fisheries Observer Program 21 GLOBEC Ichthyoplankton Sampling Program 23 Marine Resource Monitoring and Assessment Program 25 Phase 2: Interviews With Fishermen 27 Interview Methods 27 Results: Cod Spawning Grounds 30 Results: Common Themes and Observations 37 Phase 3: Synthesis of Scientific Data and Fishermen's 40 Ecological Knowledge Coverage of the Consensus Spawning Grounds by the NEFSC 43 and DFO Trawl Surveys Discussion and Future Research 46 Project Participants 48 Partnerships 48 Future Deliverables and Reports 49 Impacts and Applications 49 Presentations 49 Acknowledgments 50 References 51 Appendix A: Maps of Spawning Cod from Scientific Data Sources 58 Section 1: Data from the Northeast Fisheries Science Center 58 Biannual Trawl Survey Section 2: Data from the Canadian Depratment of Fisheries 74 and Oceans Trawl Survey Section 3: Data from the Canadian Fisheries Observer Program 83 Section 4: Data from the GLOBEC Sampling Program 92 Section 5: Data from the MARMAP Sampling Program 95 Appendix B: Phase 2 Figures - Fishermen's Ecological Knowledge 97 Appendix C: Phase 3 Figures - Synthesizing Fishermen's Ecological 124 Knowledge and Scientific Data Appendix D: UMass Dartmouth Interview Consent Form 127 Appendix E: Draft Manuscript that was submitted Attached to the ICES Journal of Marine Science in January, 2017 Abstract The spawning dynamics of Atlantic cod (Gadus morhua) on Georges Bank are not well understood. To address this uncertainty, we combined Fishermen’s Ecological Knowledge with traditional scientific data to develop a more holistic understanding of cod spawning on Georges Bank. Data from historical reports, trawl surveys, fisheries observers, and ichthyoplankton surveys were used to describe the spatial and temporal distribution of cod spawning activity. We also collected Fishermen’s Ecological Knowledge regarding cod spawning dynamics through semi-structured interviews (n=40). These fishermen had fine-scale spatial and temporal knowledge of cod spawning, and identified 210 spawning grounds on Georges Bank and Nantucket Shoals. Many of these spawning grounds were small (<50km2), and associated with particular bottom habitat features such as deep holes, or the edge of a shoal. The spawning seasons and locations identified by fishermen generally agreed with information from traditional scientific data, but it was evident that seasonal scientific surveys lack the spatial and temporal resolution needed to fully characterize the distribution of cod spawning activity. Our results provide a basis for further investigation of spawning dynamics, stock identity, and connectivity of cod on Georges Bank. In addition, the findings presented in this study can be used to guide future management measures intended to promote the rebuilding on Georges Bank cod.

Introduction Georges Bank Cod The Atlantic cod (Gadus morhua) is a demersal gadoid that inhabits relatively shallow waters (0- 600 m) in the North Atlantic, ranging from the Mid-Atlantic Bight off the northeast coast of the United States to the Barents Sea off Norway and Russia (Sundby, 2000; ICES, 2005). For many centuries, cod have supported major fisheries throughout the North Atlantic (e.g., Grand Banks: Rose, 2007), including off the coast of New England (Serchuk and Wigley, 1992; Rosenberg et al., 2005).

The development of the Georges Bank cod fishery has been well chronicled. The earliest records of fishing on Georges Bank by European settlers were trips made by vessels out of Marblehead, MA in the mid 1700’s (Goode, 1887a). By 1835, the Gloucester fleet consistently targeted halibut and cod on Georges Bank (Goode, 1887a,b). Around 1850 the catches of halibut began to decline sharply, and the fishery began to primarily target cod (Goode, 1887b). Historical landings data demonstrate that current cod landings (and Annual Catch Limits) represent a small fraction of the historical harvest levels for this stock (Figure 1). In 1880, over 12,000mt of cod were landed on Georges Bank by 163 vessels from Gloucester (Goode, 1887a). From 1891 to 1914, the fleet landed between 2,800 and 9,000mt of cod a year from Georges Bank, Nantucket Shoals, and the Great South Channel (Alexander et al., 1914). These catches were predominately taken on handlines and longlines, but the trawl fishery began in New England in 1905, and the trawl fleet gradually accounted for an increasing proportion of the cod catch (Alexander et al., 1914). Bigelow and Schroeder (1953) reported that the cod catches from Georges Bank, Nantucket Shoals, and the Great South Channel amounted to approximately 11,000mt in both 1935 and 1945. From 1960 through 1965, cod landings increased rapidly from approximately 11,000mt to 53,000mt, as effort from distant water factory trawl vessels intensified on Georges Bank (Serchuk and Wigley, 1992).

However, cod landings declined sharply by 1976, and many were worried about the magnitude of fishing effort from foreign vessels on Georges Bank. After the United States implemented the 200 mile Exclusive Economic Zone, effort by domestic trawl vessels increased significantly, and in 1982 landings reached a record level of approximately 62,000mt (NEFSC, 2012). Annual landings of Georges Bank cod exceeded 20,000mt throughout the 1980’s and early 1990’s, but the establishment of stricter effort control measures and implementation of closed areas caused cod landings to decline in 1994 (NEFSC, 2012). In recent decades, cod stocks have decreased in abundance, and catches are a small fraction of historical landings and estimates of Maximum Sustainable Yield (NEFSC, 2013; Palmer, 2014; Wang et al., 2015). Since the implementation of sector management in 2009, the annual catch of Georges Bank cod has not exceeded 5,000mt (NEFSC, 2012; NEFSC, 2015). The recent low catches and depleted status of Georges Bank cod results in considerable foregone economic yield to the fishery and the economies of New England and the Canadian Maritimes, which is estimated to be in the tens of millions of dollars per year (Edwards and Murawski, 1993).

Figure 1 - Reported cod catches on Georges Bank and Nantucket Shoals from 1879 through 2014. Catch information specific to Georges Bank and Nantucket Shoals was not available in all years. Cod catches from 1879 and 1880 were obtained from Goode (1887b). Catch data from 1891 to 1914 was obtained from Alexander (1914). Catch statistics from 1935 and 1945 were taken from Bigelow and Schroder (1953). Catch data from 1960 through 2011 was obtained from the 2012 Georges Bank stock assessment (NEFSC, 2013), and catch data from 2012 through 2014 was taken from the 2015 operational update assessment for Georges Bank cod (NEFSC, 2015).

Georges Bank cod has not rebuilt as anticipated despite substantial cuts in total allowable catches and directed management actions (NEFMC, 2009; NEFSC, 2013). Factors contributing to the lack of rebuilding include low egg production, few cod surviving to age-1 (poor age-1 recruitment index), age truncation, and persistently low spawning stock biomass. There is also evidence for recent increases in natural mortality for cod on eastern Georges Bank (Wang, 2016). The rebuilding of cod stocks in New England may also be hindered by unfavorable environmental conditions (e.g., Pershing et al., 2015; Meng et al., 2016). The most recent stock assessment of Georges Bank cod concluded that the stock was overfished, and that overfishing was occurring, although the assessment was not accepted as a basis for model advice (NEFSC, 2015). The most recent assessment for eastern Georges Bank also concluded that the stock was depleted, and that recruitment has been poor for the last 25 years (TRAC, 2016). Due to the poor status of the resource and the considerable uncertainty in the assessment, the assessment advised low catches (< 700mt) of eastern Georges Bank cod in 2017 (TRAC, 2016). Purpose of This Study

In this study, we sought to investigate cod spawning dynamics on Georges Bank using existing scientific data and fishermen’s ecological knowledge. Improving our understanding of cod spawning behavior can help to inform future stock assessments and fishery management actions intended to promote cod rebuilding on Georges Bank. Despite continuously evolving fishery management strategies (NEFMC, 2009; NEFSC, 2013), persistent difficulties remain with respect to ending overfishing (Rothschild et al., 2014) and managing the rebuilding of cod

(NEFSC, 2013; Palmer, 2014; Wang et al., 2015). Therefore, additional research is required to improve our understanding of cod population dynamics and spawning distribution on Georges Bank to inform future fishery management decisions intended to support rebuilding of the cod resource. Cod spawning activity also presents important implications for fishery management (Zemeckis et al., 2014a). For example, cod have been observed to exhibit spawning site fidelity by returning to the same spawning sites during the same time periods each season (Robichaud and Rose, 2001; Skjæraasen et al., 2011). Spawning site fidelity therefore limits reproductive connectivity among spawning components and supports the formation and maintenance of cod metapopulations (Zemeckis et al., 2014b). Given that cod spawning components are semi- discrete and can have limited connectivity among them (Smedbol and Stephenson, 2002), there is a great deal of risk with respect to the depletion of these population segments, which has been well documented in many cod stocks, including those within the Gulf of Maine (Ames, 2004) and the North Sea (Svedäng et al., 2010). The failure of past management plans to recognize cod population structure in the Gulf of Maine is thought to have contributed to the depletion of historical spawning components that have not rebuilt, and difficulties in achieving rebuilding targets (Ames, 2004; Kovach et al., 2010; Armstrong et al., 2013). Spawning aggregations in the Gulf of Maine have been protected using fine scale time-area closures, but a decade of scientific research was needed to inform the timing and location of these closures. Comparable fine-scale information on cod spawning grounds is not currently available for Georges Bank.

Multiple aspects of cod spawning dynamics can increase their vulnerability to overexploitation. For example, in addition to forming dense aggregations, individual cod spawn for multiple weeks (Zemeckis et al., 2014b) and spawning appears to occur on relatively fine spatial scales over multiple months in locations that are often predictable and close to shore (Siceloff and Howell, 2013; Dean et al., 2014). Furthermore, cod spawning behavior can be disrupted by fishing activity (Morgan et al., 1997; Dean et al., 2012). As a result, cod spawning dynamics present important implications for consideration by fishery managers (Zemeckis et al., 2014a), and more reliable data on the distribution of spawning would be valuable for consideration during the development of future fishery management plans.

The spatial and temporal distribution of spawning is also critically important for understanding recruitment dynamics and stock productivity. For example, spawning time and location chosen by adults initiates the transport of pelagic eggs and larvae, which influences their dispersal and survival (Huret et al., 2007). Depending on the location of spawning, early life stages could either be transported great distances or retained locally (Lough et al., 2006; Churchill et al., 2011), thus influencing population connectivity and the scale upon which population processes operate (Hare and Richardson, 2014). Furthermore, the time and location of spawning determines the likelihood of a match with positive environmental conditions required for the survival and growth of early life history stages (Cushing, 1990). Understanding spawning and recruitment dynamics is particularly important for helping to meet rebuilding objectives, because recruitment in these stocks has been consistently below average (NEFSC, 2013;

Palmer, 2014; TRAC, 2016) and understanding recruitment can lead to improved fishery management.

Understanding where and when cod spawn can also aid in stock identification, because the variability in spawning location and timing often function as mechanisms that contribute to the development of metapopulations by limiting the reproductive connectivity among subpopulations (Grabowski et al., 2011; Zemeckis et al., 2014b). Population segments of cod are the most geographically discrete and exhibit minimal mixing while spawning (Cushing, 1981). Therefore, studying the spatial and temporal distribution of cod spawning, as well as the connectivity among subpopulations, can provide insights into cod population structure. From 1972 onwards, cod have been assessed managed in U.S. waters as two units: the Georges Bank and Gulf of Maine stocks (Serchuk and Wigley, 1992) (Figure 2). Cod on the eastern portion of Georges Bank are assessed and managed jointly between the United States and Canada by the Transboundary Resources Assessment Committee (Wang et al., 2015). However, a recent synthesis of information, suggests that current management units do not accurately represent the underlying biological population structure of all cod subpopulations in U.S. waters (Zemeckis et al., 2014c). In particular, there is an apparent lack of connectivity between cod subpopulations on eastern Georges Bank (i.e. Northeast Peak) and western portions of the Georges Bank management unit (i.e. Nantucket Shoals, Great South Channel). In 2012, the Scientific and Statistical Committee of the New England Fishery Management Council identified cod stock structure as a key source of scientific uncertainty that should be investigated to support cod rebuilding efforts (SSC, 2012).

Figure 2 - Current management units for cod in U.S. waters, including the Gulf of Maine and Georges Bank stocks. Cod on eastern Georges Bank (outlined in bold) are managed as a transboundary resource by the United States and Canada. The black numbered circles depict the ports where we interviewed fishermen: 1 = Montauk, NY, 2 = New Bedford, MA, 3 = Hyannis, MA, 4 = Chatham, MA, 5 = Nantucket, MA, 6 = Boston, MA, 7 =

Gloucester, MA, 8 = Portland, ME, 9 = Yarmouth, Nova Scotia, 10 = Pubnico, Nova Scotia, and 11 = Lunenburg, Nova Scotia. Fishermen’s Ecological Knowledge Fishermen’s Ecological Knowledge (FEK) can be defined as the experiential knowledge that fishermen accumulate as they observe the marine environment over an extended period of time (Hind, 2015). FEK is often shared in real time between fishermen at sea, and passed down over multiple generations (Bergmann et al., 2004), and the transmission of knowledge between fishermen allows them to “scale up” their observations across time and space (Murray et al., 2008a). Fishermen are the closest to the resource, and they often keep detailed records of their fishing activity and catch (Neis et al., 1999a; Bergmann et al., 2004). Therefore, fishermen acquire a unique perspective that encompasses a range of spatial scales. For example, fishermen understand how fish move seasonally over a wide geographic range to feed or spawn. At the same time, fishermen also understand how small scale bathymetric features like specific boulder piles, and edge features along the sea bottom may each influence fish behavior or distribution. While fishermen are keenly aware of these fine-scale habitats, our scientific data collection methods often lack the spatial resolution to examine abundance and distribution at such a small scale. Following years of observation, fishermen are cognizant of long-term trends in fish abundance and changes in size structure (Pederson and Hall Arber, 1999; Macdonald et al., 2014). In addition, fishermen also understand how the distribution and abundance of target species can change across tidal, diel, lunar, and seasonal scales (Neis et al., 1999a; Johannes et al., 2000). After years of directed research, the scientific benefits that can be attained by soliciting FEK have become apparent. FEK can serve as a valuable and cost-effective supplement to existing institutional data, and incorporating FEK into the scientific process can lead to more robust management decisions (Bergmann et al., 2004; Scholz et al., 2004). FEK can offer insights into long-term trends in the abundance of fish populations (Macdonald et al., 2014), identify Essential Fish Habitat (Pederson and Hall-Arber, 1999; Bergmann et al., 2004), investigate stock structure (Neis, 1998; Murray et al., 2008b), and inform the location of Marine Protected Areas (Yates, 2014). FEK has also proven to be especially useful for identifying the location and timing of spawning activity (e.g., Ames, 1998; Neis, 1999b; Johannes et al., 2000; Silvano et al., 2006), and to discover spawning grounds that were previously unknown to scientists (e.g., Neis, 1998; Maurstad, 2002). Further, FEK has been used to identify spawning grounds that have become extirpated (Neis, 1998; Pederson and Hall-Arber, 1999; Ames, 2004), which can be critical for understanding long-term changes in stock productivity and recruitment. Project Objectives By soliciting FEK and analyzing historical reports, Ames (1998, 2004) was able to create detailed maps of the spawning grounds and migratory patterns of Atlantic cod (Gadus morhua) in the Gulf of Maine. His results provided valuable insight into cod stock structure and population dynamics, and helped inform future research to investigate spawning dynamics (Armstrong et al., 2013; Gurshin et al., 2013; Siceloff and Howell, 2013; Dean et al., 2014; Zemeckis et al., 2014b). In contrast, much less is known about the spatial and temporal distribution of cod

6 spawning on Georges Bank, including the Great South Channel and Nantucket Shoals regions. Therefore, our goal was to improve our understanding of cod spawning on Georges Bank, including comparison of historical and contemporary data to investigate potential shifts in spawning activity. In order to meet this goal, we accomplished the following research objectives: I) Map the spatial and temporal distribution of Atlantic cod spawning on Georges Bank, and compare historical and contemporary data to identify potential shifts in spawning activity.

II) Use data on the distribution of cod spawning to investigate population structure on Georges Bank (i.e. connectivity among subpopulations, fine-scale structuring).

III) Work closely with regional fishery scientists and managers in order to ensure maximum dissemination and application of project results.

IV) Identify critical information gaps relating to cod spawning activity and population structure on Georges Bank, which will be beneficial for developing future research projects (e.g., resource surveys, stock composition sampling).

V) Secondarily, collect information pertaining to the spawning distribution of other groundfish species (e.g., haddock, pollock, flounders) during interviews with fishermen and while reviewing historical accounts.

Phase One: Analysis of Scientific Data Sources We analyzed several existing scientific datasets and published reports to gather information related to cod spawning activity on Georges Bank. This scientific information was compiled and synthesized to document the current state of scientific knowledge regarding cod spawning activity on Georges Bank. Analysis of Historical Reports There is a large volume of historical information on the major cod spawning and feeding grounds on Georges Bank and in the Gulf of Maine. This information was primarily collected though conversations with fishermen, tagging studies, and field observations. The Gloucester fleet often targeted cod on the “Winter Fishing Grounds”, an area of approximately 1,100 square miles on eastern Georges Bank (Goode, 1887b; Rich, 1929). The Winter Fishing Ground was reported to serve as a major spawning ground in February, March, and April, when dense aggregations of spawning cod would predictably form every year (Goode, 1887b). The Winter Fishing Ground was described as lying eastward and southeastward of the North Shoal between the parallels of 41° 30' and 42° 00' north latitude and 66° 38' and 67° 30' west longitude (Goode, 1887b; Rich, 1929; Figure 3). The depth in this area ranged from 30-40 fathoms, and the bottom was described as being “broken, sharp, and rocky” (Rich, 1929). The Winter Fishing Grounds are marked on current nautical charts, and fishermen still consider this area to be a major cod spawning ground. Bigelow and Schroeder (1953) also described major cod spawning activity that occurs in February, March and April, and they defined the grounds as occurring from about 41 21oN to 41 31oN and 66 50oW to 67oW in depths of approximately 35 fathoms (Figure 3). In a literature review, Colton et al. (1979) reported that the spawning season for cod on Georges Bank ranged from December to April, with peak spawning occurring in March.

Nantucket Shoals has long been known to serve as a spawning and feeding ground for cod. Goode (1887b) described a number of important cod fishing grounds on Nantucket Shoals and east of Cape Cod including Outer Crab Ledge, Pollock Rip, Great Rip, Fishing, Rip, and Phelps Bank. Goode (1887b) described the seasonality of cod fishing in each of these locations, but unfortunately he did not make any mention of cod spawning activity at these locations. Smith (1902) collected cod from Nantucket Shoals and held them in a laboratory at Woods Hole for a tagging study. He observed that the eggs matured over a period of several weeks, and that the spawning season for cod taken on Nantucket Shoals ranged from the middle of November until the end of January.

Figure 3 – Location of cod spawning grounds in February, March and April. Goode (1887b) and Rich (1929) used accounts from fishermen to define the winter fishing grounds. Bigelow and Schroeder (1953) identified cod spawning activity using first hand observations and reports from fishermen.

Schroeder (1930) tagged over 24,000 cod on Nantucket Shoals between 1923 and 1929 to investigate their movement patterns and spawning behavior. Schroeder concluded that cod spawn on Nantucket Shoals from November to April, with the peak spawning activity occurring in December and January (Figure 4). Bigelow and Schroeder (1953) stated, “the broken bottom of Nantucket Shoals…has long been known as a center of abundance for ripe cod fish in late autumn and early winter.” Tagging studies documented two migratory behaviors of cod on Nantucket Shoals, with a group of fish that would remain resident on the Shoals throughout the year, and a second group of cod that would migrate to the south and east in the fall as water temperatures declined. These migratory fish would overwinter and spawn in several locations in southern New England and the mid-Atlantic including Coxes Ledge, eastern Long Island, Cholera Bank, and off of Atlantic City before migrating back to Nantucket Shoals the following spring (Smith, 1902; Schroeder, 1930; Wise, 1958).

Figure 4 - Cod spawning grounds on Nantucket Shoals as defined by Schroeder (1930). The four spawning grounds (from west to east) appear to refer to areas that are currently known as the Fingers, Old Man Shoal, Old South Shoal, and Davis Bank.

Northeast Fisheries Science Center Bottom Trawl Survey The NOAA Northeast Fisheries Science Center (NEFSC) has conducted a biannual bottom trawl survey in the northwest Atlantic, including Georges Bank, since the 1960’s (Despres-Patanjo et al., 1988). We obtained data from spring and fall survey tows from 1970 – 2014. Burnett et al. (1989) provide details on maturity staging and report that the NEFSC began collecting spawning stage information for cod in 1970. Spring survey tows were in March, April, and May, with most of the tows completed in April (70.9%) and March (22.9%). Fall survey tows were in September, October, November, and December, with most tows in October (80.1%) and November (10.1%). The analysis was limited to cod that were captured in offshore survey strata 9-25 (Figure 5), which are the strata included in the Georges Bank cod stock assessment (NEFSC, 2013).

Figure 5 - NEFSC offshore strata that are included in the Georges Bank cod stock assessment, and were included in our analysis. Cod captured by the NEFSC bottom trawl survey are routinely sampled for length, weight, sex, maturity stage, and stomach contents. Cod maturity is classified according to seven stages: unknown, immature, developing, ripe, ripe and running, spent, and resting (Table 1, Burnett et al., 1989). Cod were commonly subsampled during the 1970’s, and approximately 39% of the cod captured during this period were sampled for maturity. In subsequent years the maturity stage was determined for nearly all cod (>99%). Cod that were classified as ripe, ripe and running, or spent were considered to be in spawning condition. A large proportion of cod that were captured during the NEFSC bottom trawl survey were immature (Table 2). Many cod (n= 3,947) were observed to be in spawning condition, but only 169 were classified as ripe and running in either the spring or fall trawl surveys. More spawning cod were sampled on Georges Bank during the spring survey (n= 3,412) than during the fall survey (n= 535). The relatively high proportion of spent fish observed during the spring survey (Table 2) suggests that March, April, and May might represent the end of the spawning season for cod on Georges Bank.

Stage Sex Description and Criteria Immature Females small, translucent membrane usually colorless or pink; very thin, ribbon-like tissue lying along dorsal wall of gut cavity. Males testes colorless to gray, often more opaque than ovaries and appear more flattened Developing Females ovaries opaque and enlarged with blood vessels becoming prominent; small, opaque eggs present as ovary develops to occupy 2/3 of ventral cavity. Flatfish: ovary extending posteriorly from gut cavity; bright yellow and fine texture Males testes opaque with lobed or wavy appearance; color variable from red or pink to gray or white; milt may or may not be present in small amounts; will occupy up to 1/2 body cavity. Ripe Females this stage is to be used when female is ready to spawn, is spawning, or has not yet completed spawning; ovary may fill body cavity; eggs abundant and visible through ovary wan; a few to many transparent eggs may be present; as part of eggs are spawned, ovary may have bloodshot appearance, but eggs still numerous, color variable from bright orange to red. If eggs run from vent under slight pressure to abdomen, designate as Ripe and Running Males testes large, 2/3 of gut cavity filled, color white or pinkish or edges turning brown; milt present when testes squeezed if cut open Ripe and Females this stage denotes a ripe female with eggs flowing from vent with little or no Running pressure on abdomen Males same criteria as Ripe, but milt flows freely from vent with little or no pressure on abdomen Spent Females ovaries flaccid, usually reddish to purple, and sac-like; interior often quite fluid with a few translucent eggs present; ovarian wall opaque Males testes reduced in size, very little or no milt present when cut; colored gray to brown Resting Females ovaries much reduced in size, color purple to pink; interior jelly-like with no eggs visible; no prominent blood vessels Males testes small, opaque, and shrunken in appearance; color may be brownish to gray Table 1 - Six-stage maturity classification scheme used in NEFSC surveys for gadids (excerpted from Burnett et al., 1989)

NEFSC Spring NEFSC Fall Maturity Stage % % Survey Survey Immature 6734 24.54% 7282 40.00% Developing 1237 4.51% 3031 16.65% Ripe 415 1.51% 305 1.68% Ripe and Running 160 0.58% 9 0.05% Spent 2837 10.34% 221 1.21% Resting 10836 39.49% 5386 29.58% Unknown 5222 19.03% 1973 10.84% Total 27441 18207 Table 2 - Number of cod observed at each maturity stage during NEFSC spring and fall bottom survey tows from 1970 through 2014. During the spring survey from 1970 to 2014 there were 27,441 cod captured in 2,016 survey tows completed in offshore strata 9-25. However, the NEFSC spring survey rarely encountered large aggregations of spawning cod (Figure 6). For example, only 21 tows caught >20 cod that were in spawning condition.

Figure 6 - Histogram depicting the number of tows completed during the NEFSC spring survey tows (with positive cod catches) from 1970 through 2014 with different amounts of spawning cod (ripe, ripe and running or spent) and ripe and running cod. A major strength of the NEFSC bottom trawl survey database is that the time series of observations enables an investigation into long term trends in distribution and abundance. Investigating the distribution of ripe and running cod would be the most informative for determining whether cod spawning activity has shifted over time, but there were not enough observations of ripe and running cod in the database to permit such an investigation (See Appendix A, Figures A1.1 through A1.15). Therefore, we plotted the location of spawning

(defined as ripe, ripe and running or spent) cod observed in each decade of the spring and fall survey to look for evidence of changes in the distribution of spawning activity from the 1970’s to the present, during which time the estimated biomass of Georges Bank cod declined substantially (NEFSC, 2013). During spring surveys in the 1970’s and 1980’s, spawning cod were relatively widespread throughout Georges Bank (Appendix A, Figures A1. 1 and A1.4), but by the 1990’s, there were fewer observations of spawning cod across the central and southern portions of Georges Bank (Appendix A, Figure A1.7). In the 2000’s, observed spawning activity decreased substantially in the Great South Channel and across the central portion of Georges Bank, and the majority of spawning cod were observed on the Northeast Peak and the Eastern Part of Georges Bank (Appendix A, Figure A1.7). The transition from the R/V Albatross to the R/V Bigelow occurred in 2009, and the change in survey gear and subsequent reallocation of survey stations makes comparisons of cod catches from past and present surveys difficult. Spring survey observations from 2010 through 2014 show that the majority of spawning cod were observed on the Northern Edge and the Northeast Peak, but relatively few spawning cod were observed across the central portion of Georges Bank or in the Great South Channel (Appendix A, Figure A1.13). Spawning cod were also observed in a few tows made in close proximity to Chatham during this period. Cod spawning grounds on Nantucket Shoals are mainly outside of offshore strata 9-25, and survey tows made in this region would not be included in the assessment for this stock. Tows completed during the NEFSC spring survey were aggregated by stratum, and the proportion of tows with spawning cod was calculated for each stratum in each year (Table 3). Throughout the entire time series, spawning cod were most commonly sampled in survey strata 16, 20, 21, and 23, which correspond to eastern Georges Bank, the northern edge, and Northeast Peak of Georges Bank, as well as the deeper waters east of Cape Cod. The trawl observations suggest that most spawning activity in March and April occurs in waters <100m. For example, a relatively high proportion of spawning cod were observed in strata 16 and 21, where depths are primarily <100m. Strata 17 and 22 are adjacent to strata 16 and 21, but are >100m, and the proportion of spawning cod observed in these strata was much lower over the time series. In recent years, a lower proportion of tows completed in strata 9, 23, and 25 have encountered spawning cod, which may be indicative of a reduction in spawning activity east of Cape Cod and south of Nantucket (Appendix A, Figure A1.16). Conversely, the proportion of tows with spawning cod in strata 22 has increased over time, which may be indicative of a shift in spawning activity to deeper waters (Appendix A, Figure A1.16). Spawning cod have never been observed in the deep water strata (11, 12, 14, and 15) on southern Georges Bank, indicating that cod do not likely spawn in this region in March and April (Table 3).

Annual Year 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 Mean 1970 0.50 0.25 0.00 0.00 0.00 0.00 0.00 0.38 0.00 0.00 0.00 1.00 0.00 0.00 0.33 0.25 0.00 0.16 1971 0.50 0.22 0.00 0.00 0.22 0.00 0.00 0.33 0.50 0.00 0.66 0.83 0.40 0.00 0.40 0.29 0.25 0.33 1972 0.00 0.38 0.00 0.00 0.25 0.00 0.00 0.58 0.00 0.00 0.56 0.50 0.75 0.40 0.80 0.14 0.75 0.35 1974 0.25 0.00 0.00 0.00 0.25 0.00 0.00 0.45 0.00 0.00 0.33 0.50 0.60 0.00 0.80 0.17 0.66 0.28 1975 0.00 0.00 0.00 0.00 0.22 0.00 0.00 0.30 0.25 0.00 0.44 0.40 0.00 0.00 0.20 0.33 0.25 0.18 1976 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.10 0.33 0.00 0.33 0.83 0.50 0.00 0.66 0.17 0.33 0.20 1977 0.00 0.11 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.22 0.66 0.75 0.00 0.50 0.29 0.66 0.18 1978 0.00 0.00 0.00 0.00 0.33 0.00 0.00 0.33 0.00 0.00 0.00 0.00 0.25 0.00 0.17 0.00 0.00 0.09 1979 0.07 0.00 0.00 0.00 0.17 0.00 0.00 0.40 0.00 0.00 0.11 0.17 0.25 0.00 0.20 0.08 0.14 0.14 1980 0.06 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.11 0.00 0.25 0.00 0.00 0.33 0.00 0.05 1981 0.00 0.13 0.00 0.00 0.33 0.00 0.00 0.20 0.25 NA 0.00 0.00 0.00 0.00 0.25 0.00 0.25 0.11 1982 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.20 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.02 1983 0.20 0.00 0.00 0.00 0.00 0.00 0.00 0.30 0.00 0.00 0.22 0.83 0.25 0.00 0.60 0.17 0.00 0.18 1984 0.20 0.00 0.00 0.00 0.11 0.00 0.00 0.18 0.00 0.00 0.33 0.50 0.00 0.00 0.00 0.00 0.00 0.11 1985 0.00 0.14 0.00 0.00 0.55 0.00 0.00 0.50 0.25 0.00 0.66 0.83 0.33 0.40 0.25 0.20 0.50 0.34 1986 0.00 0.00 0.00 0.00 0.25 0.00 0.00 0.40 0.50 0.00 0.44 0.66 0.25 0.25 0.60 0.00 0.50 0.26 1987 0.40 0.13 0.00 0.00 0.11 0.00 0.00 0.10 0.00 0.00 0.11 0.17 0.00 0.00 0.60 0.00 0.00 0.11 1988 0.20 0.00 0.00 0.00 0.22 0.00 0.00 0.40 0.33 0.00 0.66 0.50 1.00 0.25 0.75 0.00 0.25 0.32 1989 0.00 0.00 0.00 0.00 0.11 0.00 0.00 0.50 0.00 0.00 0.33 0.00 0.36 0.00 0.25 0.17 0.50 0.17 1990 0.00 0.00 0.00 0.00 0.11 0.00 0.00 0.10 0.00 NA 0.33 0.33 0.14 0.25 0.30 0.00 0.00 0.13 1991 0.20 0.13 0.00 0.00 0.00 0.00 0.00 0.50 0.00 1.00 0.22 0.00 0.50 0.00 0.30 0.00 0.00 0.17 1992 0.00 0.00 0.00 NA 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.25 0.00 0.25 0.00 0.00 0.03 1993 0.00 0.00 0.00 NA 0.00 0.00 0.00 0.10 0.33 0.00 0.17 0.17 0.00 0.25 0.00 0.17 0.00 0.08 1994 0.00 0.00 0.00 NA 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.25 0.00 0.00 0.00 0.00 0.01 1995 0.20 0.00 0.00 0.00 0.00 0.00 0.00 0.33 0.00 0.00 0.00 0.17 0.25 0.25 0.50 0.00 0.33 0.13 1996 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.40 0.00 0.00 0.22 0.17 0.75 0.00 0.50 0.00 0.00 0.16 1997 0.00 0.00 0.00 NA 0.00 0.00 0.00 0.30 0.33 0.00 0.22 0.17 0.00 0.50 0.40 0.17 0.00 0.15 1998 0.00 0.00 0.00 0.00 0.43 0.00 0.00 0.20 0.00 0.00 0.00 0.00 0.50 0.25 0.20 0.00 0.50 0.13 1999 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.50 0.00 0.00 0.00 0.17 0.00 0.25 0.00 0.00 0.00 0.09 2000 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.50 0.00 0.00 0.44 0.50 0.50 0.50 0.00 0.50 0.25 0.25 2001 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.20 0.33 0.00 0.22 0.33 0.25 0.00 0.00 0.00 0.00 0.10 2002 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.50 0.00 0.00 0.11 0.17 0.00 0.00 0.60 0.00 0.50 0.15 2003 0.00 0.00 0.00 0.00 0.22 0.00 0.00 0.30 0.00 0.00 0.00 0.25 0.50 0.50 0.20 0.33 0.00 0.17 2004 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.44 0.00 1.00 0.00 0.17 0.75 0.00 0.00 0.00 0.00 0.11 2005 0.00 0.00 0.00 0.00 0.11 0.00 0.00 0.60 0.33 0.00 0.22 0.33 0.50 0.50 0.20 0.00 0.00 0.21 2006 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.40 0.00 0.00 0.22 0.17 0.50 0.00 0.40 0.00 0.50 0.16 2007 0.00 0.13 0.00 0.00 0.00 0.00 0.00 0.55 0.33 0.00 0.11 0.17 0.50 0.25 0.20 0.17 0.00 0.22 2008 0.00 0.00 0.00 0.00 0.11 0.00 0.00 0.40 0.33 0.00 0.22 0.17 0.25 0.50 0.00 0.17 0.00 0.19 2009 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.36 0.00 0.33 0.13 0.25 0.75 0.25 0.11 0.13 0.00 0.14 2010 0.00 0.00 0.00 0.00 0.20 0.00 0.00 0.50 0.33 0.00 0.63 0.83 0.75 0.66 0.33 0.13 0.00 0.30 2011 0.00 0.11 0.00 0.00 0.10 0.00 0.00 0.33 0.00 0.00 0.00 0.00 0.25 0.50 0.33 0.00 0.50 0.13 2012 0.00 0.00 0.00 0.00 0.22 0.00 0.00 0.50 0.25 0.00 0.33 0.80 0.25 0.66 0.17 0.00 0.00 0.22 2013 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.25 0.00 0.00 0.44 0.40 0.50 0.50 0.00 0.11 0.00 0.13 2014 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.08 0.00 0.00 0.25 0.50 0.50 0.00 0.00 0.17 0.00 0.09 Mean 0.06 0.04 0.00 0.00 0.10 0.00 0.00 0.33 0.11 0.03 0.23 0.31 0.35 0.17 0.27 0.10 0.17 0.16 Table 3 - Proportion of tows where spawning cod were observed in each strata during the NEFSC spring trawl survey from 1970 through 2014. For the NEFSC spring trawl survey, we performed a G* hotspot analysis (Getis and Ord, 1992) to identify tow locations where the catches of spawning cod were significantly greater than average (p < 0.05) and were surrounded by other tows with above average catches of spawning cod, following the methods outlined by the Closed Area Technical Team (NEFMC, 2013). The number of hotspots within a 100km2 grid was summarized to identify spatial clusters of spawning activity. Nearly all cod spawning hotspots (p < 0.05) were observed on the northern and eastern portions of Georges Bank from 41o40’N to 42o12’N and between 66oW and 68oW, with the greatest concentrations of hotspots located on the Northeast Peak (Figure 7).

Figure 7 - Distribution of cod spawning hotspots observed during the NEFSC spring trawl survey from 1970 through 2014. The number of hotspots within a 100km2 grid was summarized to identify spatial clusters of spawning activity. During fall surveys completed in the 1970’s, the majority of survey tows did not contain spawning cod, but spawning cod were observed in the Great South Channel, the Northern Edge, eastern Georges, and in some central portions of Georges Bank (Appendix A, Figure A1.3). During fall surveys completed between 1980 and 1989 spawning cod were observed in the Great South Channel and Closed Area I, as well as Eastern Georges Bank. However, unlike the previous decade, spawning cod were mostly absent from the Northern Edge and the Northeast Peak (Appendix A, Figure A1.6). The distribution of spawning cod from fall surveys completed between 1990 and 2009 was mostly restricted to the Great South Channel and Closed Area I, and fewer spawning cod were observed on eastern Georges than in the previous two decades (Appendix A, Figure A1.9, and A1.12). From 2010 to 2014 there were relatively few tows during the fall survey with positive cod catches, and observations of spawning cod were rare (Appendix A, Figure A1.15).

Canadian Department of Fisheries and Oceans Trawl Survey The Canadian Department of Fisheries and Oceans (DFO) has conducted an annual trawl survey on Georges Bank since 1987 (Stone and Gross, 2012). The survey is conducted in February and March, and is well-timed to coincide with previous observations of cod spawning across much of Georges Bank (Berrien and Sibunka, 1999). The maturity stages of cod are examined and classified according to the following eight categories; immature, developing 1, developing 2, ripe, running, spent, recovering, and resting (Morrison, 1990). We obtained data from DFO survey tows that were completed from 1987 through 2015, which included sampling of 16,668 cod. Cod that were classified as ripe, running, or spent were considered to be in spawning condition. Of the 16,668 cod in the database, 7,238 were classified as being in spawning condition, and 2,412 were observed to be actively spawning (running) at the time of capture. Relatively large (>10 cod/tow) tows of spawning cod were more commonly observed during the DFO survey than on the NEFSC spring survey (Figure 8). Male skewed sex ratios have been used as an indicator of cod spawning grounds, and a skewed sex ratio was observed on DFO survey tows (Morgan and Trippel, 1996). The sex ratio of spawning cod was heavily skewed towards males (n males= 5,520; n females= 1,718; ratio = 3.21:1). For running cod, the sex ratio was even more skewed towards males (n males= 2,057; n females= 355; ratio = 5.8:1).

Figure 8 - Histogram depicting the number of spawning and ripe and running cod that were observed during DFO bottom trawl survey tows between 1987 and 2015. The abundance and distribution of spawning cod and running cod observed during the DFO survey were plotted by decade to examine spatial and temporal trends in cod spawning activity. The DFO survey does not sample the entirety of Georges Bank. The majority of sampling effort is focused in Canadian waters and in the survey strata used in the eastern Georges Bank TRAC stock assessment (Figure 9). Sampling extends westward into U.S. waters as time allows.

Samples in the U.S. portion of Georges Bank are typically taken along the Southwest Part, and relatively few samples have been collected along the Northern Edge or in the Great South Channel.

Figure 9 - Stratification scheme used for the Canadian Department of Fisheries and Oceans (DFO) annual trawl survey. From 1987 through 1989, spawning cod were relatively widespread throughout eastern Georges Bank, in Closed Area II, and along the Southwest Part (Appendix A, Figure A2.1). The largest catches of spawning cod were taken east of Closed Area II, between 67oW and 66 30oW and 41 30oN and 42oN. From 1990 through 1999, cod in spawning condition (ripe, running, or spent) were commonly observed in all of the regions sampled by the DFO trawl survey (Appendix A, Figure A2.3). However, the vast majority of ripe and running cod were observed on eastern Georges Bank from 67oW and 66oW and 41 30oN and 42oN, and running cod were relatively rare in waters with depths > 50 fathoms (Appendix A, Figure A2.4). From 2000 through 2009, fewer spawning cod were captured along the Southwest Part and in Closed Area II (Figure A2.5). However, spawning cod were present in most of the locations sampled in Canadian waters over this time. The distribution of running cod was more restricted, and running cod were largely absent from survey tows completed in areas where the depth exceeded 50 fathoms (Appendix A, Figure A2.6). Running cod were also relatively scarce in Closed Area II from 2000 through 2009. From 2011 through 2015, the DFO survey primarily sampled in Canadian waters. Spawning cod were routinely sampled across all of eastern Georges Bank, and in Closed Area II (Appendix A, Figure A2.7). Again, the distribution of

18 running cod was somewhat restricted, with the greatest concentration of running cod observed on eastern Georges Bank in waters < 50 fathoms (Appendix A, Figure A2.8). In order to investigate locations where cod may be aggregating to spawn in February and March, we examined a subset of the DFO survey data with large catches of cod (n > 20) and plotted the proportion of cod that were observed to be in spawning condition in that tow (Appendix A, Figure A2.9). Between approximately 67oW and 66 10oW and 41 30oN and 42oN, the majority of tows contained a high proportion of cod in spawning condition, whereas the proportion of spawning cod was lower in survey tows that were completed on the eastern and northern edges of Georges Bank, where depths ranged from 50 to 100 fathoms. Tows completed during the DFO trawl survey were aggregated by stratum, and the proportion of tows with spawning cod was calculated for each stratum in each year (Table 4). Strata 5Z1, 5Z2, 5Z3, and 5Z4 were sampled consistently throughout the entire time series, while the other strata (5Z5 to 5Z9) were only sampled in a subset of years. Over the entire time series, the proportion of tows with spawning cod remained fairly constant. Tows made on the Northeast Peak, in strata 5Z2, consistently had the highest proportion of spawning cod (Appendix A, Figure A2.10). The proportion of tows with spawning cod on the southwest part of Georges Bank, in strata 5Z4, generally declined over time. A similar trend was observed in strata 5Z3, which encompasses the northern edge of Georges Bank, as well as portions of Georges Shoals. Interestingly, the DFO survey has sampled in the strata 5Z9 in recent years, and spawning cod were commonly observed in this stratum, suggesting that spawning activity may occur in the Northeast Channel, where depths exceed 200m. Spawning cod were never encountered in strata 5Z8, which is the relatively deep water stratum on southern Georges Bank.

Annual Year 5Z1 5Z2 5Z3 5Z4 5Z6 5Z7 5Z8 5Z9 Mean 1987 0.22 0.57 0.50 0.18 0.17 0.00 0.23 1988 0.25 0.71 1.00 0.19 0.27 0.00 0.41 1989 0.48 0.63 0.66 0.44 0.44 0.50 0.00 0.49 1990 0.34 0.81 0.91 0.44 0.35 0.00 0.50 1991 0.43 0.68 0.64 0.29 0.00 0.28 0.00 0.41 1992 0.45 0.72 0.66 0.73 0.73 0.60 0.00 0.64 1993 0.40 0.86 0.18 0.50 0.00 0.62 1994 0.17 0.63 0.29 0.00 0.00 0.42 1995 0.33 0.47 0.86 0.17 0.00 0.50 0.00 0.41 1996 0.22 0.79 0.63 0.23 0.66 0.00 0.57 1997 0.22 0.69 0.29 0.47 0.00 0.00 0.47 1998 0.30 0.70 0.18 0.08 0.00 0.25 0.00 0.42 1999 0.25 0.68 0.50 0.60 0.00 0.00 0.52 2000 0.30 0.71 0.70 0.20 0.40 0.00 0.51 2001 0.56 0.70 0.33 0.31 0.00 0.00 0.51 2002 0.30 0.85 0.50 0.27 0.00 0.00 0.00 0.53 2003 0.00 0.76 0.50 0.31 0.00 0.25 0.00 0.48 2004 0.10 0.69 0.44 0.00 0.00 0.33 0.00 0.42 2005 0.32 0.63 0.12 0.33 0.00 0.00 0.41 2006 0.26 0.66 0.45 0.13 0.00 0.00 0.44 2007 0.30 0.80 0.13 0.07 0.00 0.00 0.45 2008 0.38 0.92 0.25 0.18 0.00 0.00 0.54 2009 0.50 0.58 0.00 0.08 0.25 0.33 2010 0.30 0.83 0.33 0.20 0.00 0.67 0.49 2011 0.39 0.77 0.29 0.13 0.20 0.00 0.17 0.45 2012 0.31 0.39 0.56 0.07 0.00 0.13 0.30 2013 0.50 0.53 0.27 0.20 0.00 0.00 0.14 0.38 2014 0.17 0.88 0.33 0.09 0.00 0.00 0.40 0.47 2015 0.43 0.77 0.29 0.00 0.29 0.48 Mean 0.33 0.70 0.43 0.25 0.10 0.25 0.00 0.25 0.46 Table 4 - Proportion of tows where spawning cod were observed in each strata during the DFO trawl survey from 1987 to 2015. Using the same methods that were described above for the NEFSC spring survey, we performed a G* hotspot analysis (Getis and Ord, 1992) to identify survey tow locations where the catches of spawning cod were significantly greater than average (p < 0.05) and were surrounded by other tows with above average catches of spawning cod. The majority of cod spawning hotspots (p < 0.05) were observed in Canadian waters on eastern Georges Bank, between 66o5’W and 66o50’W and 41o35’N and 42o5’N, although some spawning hotspots were also present in Closed Area II (Figure 10).

Figure 10 - Distribution of cod spawning hotspots observed during the DFO trawl survey from 1987 through 2015. The number of hotspots within a 100km2 grid were summarized to identify spatial clusters of spawning activity.

Canadian Fisheries Observer Program Data from the Canadian Fisheries Observer Program was provided by Irene Andruschenko at the Canadian DFO. Unlike the bottom trawl survey datasets, the geographic distribution of the observer data is relatively limited, because all observations were collected in Canadian waters on eastern Georges Bank (east of 67oW). The trawl fishery on Georges Bank primarily targets haddock and pollock, but cod are often taken as a bycatch. The spatial and temporal characteristics of the data collected through the Fisheries Observer program is reflective of patterns of fishing effort by the trawl fleet, and seasonal fishery regulations. The majority of cod maturity samples were collected in January and February (Table 5), and relatively few samples were taken in March, April and May because eastern Georges Bank is typically closed to the Canadian groundfish fleet during these months to prevent the overexploitation of spawning cod. Canadian fishery observers are instructed to sample the maturity stages of cod using the same classification schemes that are employed on the DFO survey. The data collected by the observers are used to inform time/area closures on Georges Bank to protect cod

21 spawning activity. When >30% of the observed cod are in spawning condition, the Canadian portion of Georges Bank is closed to groundfishing until the end of May. This closure generally occurs at the beginning of February each year (Wang et al., 2015). The Canadian Observer data was collected from 1983 through 2015, and includes information from 1,566 tows with a positive cod catch. From these tows, 18,804 cod were sampled for maturity. Observers subsampled cod for maturity in many large tows. Therefore, there may be some tows with spawning cod that were not subsampled by the observers. The data have not been standardized with respect to factors such as tow duration or gear type, so the data cannot be used to make direct inferences about the magnitude of cod spawning at each location. However, the data are informative for investigating the timing and location of cod spawning on Georges Bank. For this analysis, cod that were ripe, running, or spent were considered to be in spawning condition.

Number of Tows Number of Ripe % of Cod that Number of Tows Month Number of Cod with Ripe and and Running were Ripe and with Cod Running Cod Cod Running January 852 9537 274 876 9% February 240 3983 145 825 21% March 6 28 4 14 50% April 0 0 0 0 May 18 362 1 1 0% June 47 496 6 10 2% July 85 1089 3 5 0% August 132 1466 5 6 0% September 86 721 4 4 1% October 30 366 0 0 0% November 29 372 5 20 5% December 41 384 5 19 5% Sum 1566 18804 452 1780 Table 5 - Monthly summary of cod maturity samples collected by the Canadian Fisheries Observer program from 1983 through 2015. A total of 7,713 of the 18,804 cod that were sampled by Canadian fisheries observers were in spawning condition, and 1,780 of those cod were running at the time of capture. Spawning cod were sampled year round on eastern Georges Bank, but a relatively small proportion of cod were observed to be in spawning condition from July through October (Table 6). Spawning activity appears to increase from November through January, before peaking in February and March. A clear understanding of these seasonal trends is confounded to some extent by the paucity of observations in March, April, and May, but data collected in March by the DFO bottom trawl survey indicates that cod are still actively spawning on eastern Georges Bank in early spring. Although there are few observations, running fish were not commonly recorded

22 by observers in May and June, and a greater proportion of cod were spent or recovering during these months (Table 6).

Month Immature Developing Developing Ripe Ripe and Spent Recovering Resting Total 1 2 Running January 3% 14% 29% 38% 9% 3% 3% 2% 100% February 2% 8% 20% 31% 20% 7% 9% 3% 100% March 0% 4% 4% 39% 50% 4% 0% 0% 100% April May 2% 36% 2% 1% 0% 17% 25% 16% 100% June 10% 22% 9% 3% 2% 23% 22% 8% 100% July 11% 20% 7% 2% 0% 4% 15% 40% 100% August 8% 20% 6% 1% 0% 5% 11% 48% 100% September 8% 35% 13% 5% 1% 0% 5% 34% 100% October 3% 32% 17% 3% 0% 4% 10% 30% 100% November 5% 33% 30% 8% 5% 2% 4% 12% 100% December 3% 20% 34% 22% 5% 2% 3% 12% 100% Table 6 - Monthly patterns of cod maturity reported by the Canadian Fisheries Observer program from 1983 through 2015. Maps depicting the monthly locations where running cod were sampled by the Canadian fisheries observer program are provided in Appendix A (Figures A3.1 through A3.9). In January and February, running cod are relatively widespread on eastern Georges Bank (Figures A3.1 and A3.2). From June through September, running cod were uncommon (Figures A3.3 through A3.6). The running cod sampled during these months were primarily distributed in deeper waters on the northern and eastern edges of Georges Bank, and there were large portions of eastern Georges Bank where running cod were absent in these months. No running cod were observed in October (Figure A3.7), and the distribution of running cod was constrained to the northern and eastern edges of Georges Bank in November and December (Figures A3.8 and A3.9). Maps depicting the proportion of cod in each that were observed to be in spawning condition during each month are also provided in Appendix A (Figures A.10 through A3.18). GLOBEC Ichthyoplankton Sampling Program As part of the United States Global Ocean Ecosystems Dynamics (GLOBEC) program, oblique plankton tows were completed across Georges Bank using a bongo net with a 333-um mesh (Sibunka et al., 2006). Sampling was completed monthly from February to July in 1995, and from January to June from 1996 to 1999. Cod eggs were classified to three stages, and stage- one (“early”) eggs represented the period from “spawned to just before blastopore closure” (Sibunka et al., 2006). A flowmeter was attached to the net to estimate the volume of water sampled during each tow, and a Standard Haul Factor was reported for each tow (Sibunka et al., 2006). The number of cod eggs observed at each developmental stage was multiplied by the Standard Haul Factor to calculate the concentration (#/10m2) of cod eggs for that tow.

The distribution and abundance of stage-one cod eggs that were observed in each month from 1995 through 1999 were plotted and examined. Cod eggs are pelagic, and are advected passively from the spawning grounds (Laurence and Rogers, 1976). Therefore, we chose to examine the distribution of stage-one eggs, because these eggs had been adrift for the shortest period of time and would presumably be closer to the spawning grounds than other stages. The developmental rate and stage duration of cod eggs is related to temperature, and the stage one duration of cod eggs in the North Sea varies from 7.6 to 8.1 days at 3.35oC, and 4.9 to 5.3 days at 7.9oC (Thompson and Riley, 1981). Page et al. (1999) reported that the stage one duration of cod eggs is five days at a temperature of 5oC. Given the strong currents that are common across much of Georges Bank, stage-one cod eggs could be advected considerable distances away from the spawning grounds during their development. Therefore, the locations where cod eggs have been sampled cannot necessarily be considered as spawning grounds. Nevertheless, the spatiotemporal distribution and concentration of stage one eggs can help infer geographic patterns of egg production. Cod eggs were present in the majority of tows completed from January through April, while cod eggs were rarely present in samples collected in May and June (Table 7; Appendix A, Figures A4.1 through A4.5). Stage one cod eggs were absent from the 38 samples collected in July of 1995. Stage-one cod eggs were present in the majority of tows completed in January, February and March, which suggests that cod spawning activity is widespread across Georges Bank during these months. Stage-one cod eggs were sampled less often in April, and by May the occurrence of stage-one eggs was greatly reduced.

# of Tows With Cod # of Tows With Month # of Tows Eggs (all stages) Stage One Cod Eggs January 229 148 (65%) 115 (50%) February 338 260 (77%) 212 (63%) March 342 223 (65%) 171 (50%) April 334 179 (54%) 118 (35%) May 374 55 (15%) 34 (9%) June 200 7 (4%) 3 (2%) July 38 0 (0%) 0 (0%) Table 7 - Summary of the monthly sampling intensity for the GLOBEC program on Georges Bank from 1995 through 1999. For each month the number (and %) of tows with positive catches of cod eggs and the number (and %) of tows with stages-one cod eggs is shown. For tows with positive catches of cod eggs, the average monthly concentration of eggs at each developmental stage was calculated (Table 8). The monthly distribution and concentration of stage one cod eggs were plotted to examine spatial and temporal patterns (Appendix A, Figures A4.1 through A4.5). In January, the greatest concentrations of stage-one cod eggs were observed on the eastern part of Georges Bank, in waters less than 50 fathoms (Appendix A, Figure A4.1). Stage-one cod eggs were rarely sampled across central Georges Bank in January.

A similar distribution was observed in February, but the concentration of eggs had increased in many locations (Appendix A, Figure A4.2). Stage-one eggs were also more commonly observed across the Northern Edge in February than in January. In March, the concentration of stage one eggs remained high on eastern Georges Bank, but high concentrations of eggs were also observed in the southern portion of Closed Area II (Appendix A, Figure A4.3). It is unclear if these eggs were spawned in this region, or were advected from an upstream location. By April, the concentration of stage-one cod eggs decreased, and eggs were rarely sampled along the Northern Edge (Appendix A, Figure A4.4). In May, stage one cod eggs were still present on Georges Bank, but at low concentrations and in scattered locations (Appendix A, Figure A4.5).

mean conc. mean conc. mean conc. Month of stage 1 SD of stage 2 SD of stage 3 SD eggs eggs eggs January 49.1 151.8 6.1 16.0 19.2 25.9 February 173.1 379.7 32.2 80.4 88.6 211.4 March 157.9 405.4 39.3 103.4 123.9 265.4 April 17.9 46.2 6.1 11.5 47.5 100.9 May 11.9 29.5 1.9 3.3 8.5 10.1 June 2.2 2.8 0.8 2.1 4.6 2.5 Table 8 - Mean monthly concentration and standard deviation of cod egg densities at each developmental stage that were observed during the GLOBEC sampling program from 1995 to 1999. Marine Resource Monitoring Assessment and Prediction Program The Marine Resource Monitoring Assessment and Prediction program (MARMAP) collected monthly bongo net samples of fish eggs and larvae in the Northwest Atlantic from 1977 to 1987 (Berrien and Sibunka, 1999). Eighty one surveys were completed over the 11 year period, and 214 standard survey locations were sampled on each trip (Berrien and Sibunka, 1999). Composite maps depicting the monthly distribution and concentration of cod eggs observed during the MARMAP survey are shown in Appendix A (Figures A5.1 and A5.2), and the average monthly concentration of cod eggs observed throughout the entire study area (Figure 11). The MARMAP reports showed the abundance and distribution of cod eggs that were aggregated across all stages, and it is not possible to investigate the abundance and distribution of stage one eggs from the data contained in the reports. The composite maps show that cod eggs were rarely observed on Georges Bank in July and August, and that cod eggs were not sampled anywhere on the Bank in September. The concentration of cod eggs increased across the Bank from October to December, and it appears that cod eggs were primarily concentrated in the vicinity of Georges Shoals in November, which may be indicative of late fall spawning activity in that region (Appendix A, Figure A5.2). In December and January, cod eggs were relatively abundant on Georges Shoals, on eastern Georges Bank, and south of Nantucket Shoals. In February, cod eggs were sampled throughout the Bank, with the greatest concentrations of cod eggs occurring on the Northeast Peak and eastern Georges. In March, cod eggs were

25 distributed across the Bank with the greatest concentrations occurring on the Southwest Part, around Closed Area II, and on eastern Georges. However, it is uncertain whether the large numbers of eggs sampled on the Southwest Part originated in this location, or were spawned elsewhere and advected to this region. Cod eggs are relatively abundant across much of the Bank in April, with high concentrations occurring across the eastern and northern portions of the Bank. In May and June the concentration of cod eggs was reduced, and the greatest concentrations were observed in a relatively small area on the Northeast Peak (Appendix A, Figure A5.1).

Figure 11 - Monthly concentration of cod eggs (all stages) observed during the MARMAP sampling program across the entire study area (includes Georges Bank and the Gulf of Maine). Figure obtained from Berrien and Sibunka (1999).

Phase Two: Interviews with Fishermen Interview Methods We used a semi-structured interview approach to gather fishermen’s observations on the timing and location of cod spawning on Georges Bank. In this flexible format, the interview was guided by a series of standardized questions, but the fishermen had latitude to discuss their observations and perspectives. In almost every interview, two or more scientists conducted the interview, with one scientist leading the discussion, and the other scientists recording answers and marking spatial information by hand on paper charts. An audio recording was also captured for each interview. Current and retired captains (n = 40) with experience fishing for cod on Georges Bank were interviewed. We started by interviewing knowledgeable captains known from previous collaborative research projects, and used a snowball sampling process (Babbie, 1989), where captains were asked to refer other fishermen who also targeted cod on Georges Bank. Project co-PI David Martins is fluent in Portuguese, which allowed us to interview Portuguese speaking fishermen who are typically unable to provide their input into fisheries science and management. The interviews were done at times and locations that were convenient for fishermen, and were typically conducted aboard their vessels in port, in cafes, their homes, or in our office nearby the port of New Bedford. Participation was voluntary, and fishermen were not required to respond to questions if they were uncomfortable disclosing certain information. Each fisherman was provided with a $50 Visa gift card as compensation for their participation, although many of the fishermen were not aware of the compensation until after the interview was completed. This amount was thought to provide sufficient motivation to participate, but not be too much to encourage participation for the wrong reasons (i.e., just the reward). Each fisherman was required to read and sign an interview consent form, developed and approved through the University of Massachusetts Dartmouth Office of Institutional Compliance and Ethics (Appendix D). The duration of interviews ranged from forty five minutes to three hours, depending upon each fisherman’s knowledge and their willingness to provide information. We interviewed fishermen throughout New England and Nova Scotia in order to gain the perspectives of a diverse group of fishermen with experience fishing across all regions of Georges Bank. A conscious effort was made to interview fishermen without the presence of other fishermen, so that participants could answer survey questions freely, and to the best of their knowledge, without influence from others. In one instance, three fishermen were interviewed together on Cape Cod due to logistical reasons that could not be avoided. Each interview began with a series of demographic questions designed to gauge the fishermen’s experience fishing for groundfish, and their experience specifically targeting cod on Georges Bank (Part 1 of Table 9). Next, fishermen were asked to recall times and locations where they had observed spawning cod (Part 2 of Table 9). Fishermen were asked to identify and delineate these locations on NOAA nautical chart 13200 which encompasses the Georges Bank cod stock area from southern New England to eastern Georges Bank. Many fishermen also provided information from their own charts, logbooks, and plotters, and used this information to precisely identify the locations and time of year where they captured spawning

27 cod over the course of their careers. For each spawning ground that fishermen identified, we asked them a series of questions to classify the timing of spawning at that location, and to collect information on the biological characteristics and habitat associated with that spawning site (Table 9). Question #3 from Part #2 was asked to identify what factors fishermen used to classify an area as a spawning ground. Some fishermen used the presence of eggs or milt on the deck as an indication of a cod spawning ground, while other fishermen claimed that the appearance of cod on the sounder could be used as an indication they were spawning. However, the majority based their reports on the maturity stages of cod in the catch. U.S. fishermen on Georges Bank dress their cod at sea and bring them to market without their gonads and guts, which gives them a great look at cod maturity stages. When fishermen indicated that they were basing their reports on observed cod maturity stages, then a photo guide was shown to them in an attempt to identify specific maturity stages that they remembered in the catch. (Appendix B, Figures B1 and B2). Having the photos of spawning cod helped to distinguish cod feeding grounds from spawning grounds, and record them as such. The images of cod spawning stages were obtained from Richard McBride (NEFSC), and are the same as those used to classify maturity on trawl surveys (Burnett et., al 1989). Finally, as time allowed, we asked the fishermen questions related to the biology and stock structure of cod on Georges Bank (Table 9). The questions asked during the interview process were developed early on in the project with input from scientists at NOAA, Canadian DFO, and Ted Ames who conducted a similar study on cod spawning grounds in the Gulf of Maine.

Table 9 - Questions that we asked fishermen during the interviews. After each interview, the spawning grounds identified by the fishermen were digitally mapped in ArcGIS. A unique shapefile was created for each spawning ground and a layer file containing all of the spawning grounds identified by each captain was created. The spawning grounds were also grouped in space and time for visualization purposes. For example, layer files were created which contained all of the spawning grounds that were identified in each month. Spawning grounds were classified according to different regions of Georges Bank (Figure 12), and layer files were created to display all of the spawning grounds identified in each region. We used geoprocessing tools in ArcGIS to quantify the amount of overlap between spawning grounds (i.e., shapefiles), in order to quantify the number of fishermen that identified spawning activity at a given location. We identified “consensus spawning grounds” as locations where spawning activity was reported by three or more fishermen.

Figure 12 – Eight regions that were used to classify and summarize cod spawning activity across Georges Bank. We convened a workshop with fishermen, fisheries scientists, and managers on August 8th, 2016 at the United Fishermen’s Club in New Bedford, MA. The purpose of the workshop was to summarize and disseminate the results of the interviews and scientific data, and to solicit feedback on the project. Results: Cod Spawning Grounds Of the 52 fishermen we contacted for the project, 40 agreed to complete an interview. All of the fishermen we interviewed were captains, and 39 of the 40 had been captains for the majority of their career. These fishermen collectively had 1,566 total years of fishing experience, including 1,373 years fishing specifically for cod on Georges Bank. On average, each fishermen had 34 years of experience targeting Georges Bank cod (range = 12 to 52 years). Eight of the 40 fishermen used multiple gear types to target Georges Bank cod. The majority of fishermen (n = 35) used an otter trawl, while others used gillnets (n = 8), longline (n = 7), and rod and reel (n = 2). The captains we interviewed fished from New Bedford, MA (n = 21), Chatham, MA (n = 7), Hyannis, MA (n = 1), Gloucester, MA (n = 1), Nantucket, MA (n = 1), Boston, MA (n = 1), Montauk, NY (n = 1), Portland, ME (n = 1), Pubnico, Nova Scotia (n = 3), Yarmouth, Nova Scotia (n = 2), and Lunenburg, Nova Scotia (n = 1) (Figure 2). The fishermen identified 210 cod spawning grounds on Georges Bank and Nantucket Shoals, but the same spawning grounds were often identified independently by multiple fishermen, including at varying spatial scales (Figure 13).

Figure 13 - Cod spawning grounds that were identified by fishermen. Each polygon represents a spawning ground that was identified by a single fisherman. The shading is used to identify areas where cod spawning activity was identified independently by multiple fishermen. On average, each fisherman individually identified five cod spawning grounds (range = 2 to 25 spawning grounds). It was evident during the interview process that captains were attentive to the reproductive condition of the cod they had caught. Fishermen would observe the reproductive condition of the fish when they gutted them, and often observed large amounts of milt or eggs on the deck when they encountered spawning cod. However, Canadian fishermen do not dress their fish at sea, and instead land them round. Therefore, Canadian fishermen typically could not tell that they caught spawning cod unless they observed milt or eggs flowing from the fish. Other factors fishermen described were high catch rates associated with targeting spawning aggregations, and some noted they could identify cod spawning aggregations using their sounders. Other fishermen would pay attention to the reproductive state of cod because they liked to eat the eggs, and they preferred to eat firm ovaries from cod that were developing. The fishermen stated that once cod became ripe or ripe and running the soft texture of the eggs was unpleasant, and they would no longer eat the ovaries. Thus, their ability to identify spawning cod was determined in part by their culinary practices at sea. “Nantucket Shoals and the Channel” was the region most commonly identified as having cod spawning grounds, with twenty seven fishermen identifying 81 spawning grounds in this region (Table 10). Nantucket Shoals was a popular winter fishing ground for the New Bedford, MA trawl fleet, and fishermen reported they would often fish here because of the high catch rates of cod and the close proximity to land, which would enable them to quickly find shelter during a storm. It should be noted that we interviewed 21 fishermen from New Bedford, and several

31 from Cape Cod, which may have influenced the spatial distribution of the spawning grounds that the fishermen reported. Georges Shoal and the Northern Edge were also identified as important regions for cod spawning, while relatively few spawning grounds were reported on the Southwest Part and Eastern Georges Bank.

Number of Number of Number of Fishermen that Number of Spawning Fishermen that Region Identified Feeding Grounds Grounds Identified Spawning Identified Identified Feeding Grounds Grounds Nantucket Shoals and Channel 27 84 8 13 Closed Area 1 7 7 2 2 Northern Edge 18 28 1 1 Georges Shoal 20 41 7 8 Southwest Part 8 9 2 2 Closed Area 2 12 21 1 2 Northeast Peak 12 13 5 15 Eastern Georges Bank 4 7 2 4 Table 10 – Number of cod spawning and feeding grounds that were identified in each region of Georges Bank. Twenty six consensus spawning grounds (i.e., areas independently identified by three or more fishermen) were identified during the interviews (Figure 14). These spawning grounds were widespread throughout the Georges Bank stock area. Many of the spawning grounds are discrete and associated with specific bathymetric features such as channels between shoals, edge habitats adjacent to shoals, complex rocky bottom, or areas with steep bathymetric contours. These consensus spawning grounds were familiar to most fishermen that were interviewed, and often had common names that fishermen used to identify each location. Nine of these spawning grounds were located on Nantucket Shoals and two were located just east of Cape Cod. Consensus spawning grounds were also identified on Georges Shoals, the Northern Edge of Georges Bank, and in Closed Area II. Only two consensus spawning grounds were identified in Canadian waters on eastern Georges Bank and none were identified on the Southwest Part of Georges Bank.

Figure 14 - Consensus spawning grounds that were identified independently by at least three fishermen. The fishermen had detailed knowledge about the timing of cod spawning and the seasonal availability of cod on the fishing grounds. Fishermen reported that there is little cod spawning activity on Georges Bank between July and September, and that cod spawning activity increases steadily from October to December (Figure 15). Over half of the participating fishermen reported active spawning grounds in January, February, and March. Fishermen noted that cod spawning declined from relatively high levels in April to lesser amounts in May and June.

Figure 15 - The number of fishermen who reported spawning activity in each month, and the number of spawning sites identified in each month throughout the entire study area. Fishermen reported that the timing of spawning varied between western and eastern regions of Georges Bank. Although at least one fisherman described spawning activity on Nantucket Shoals and the Great South Channel in each month, most fishermen described the spawning period in the western region as occurring from October to April, with peak spawning in November and December, which is earlier than other regions further to the east (Figure 16). On Georges Shoal, the spawning season was described as lasting from October through June, with the majority of spawning taking place between December and May. Peak spawning in the relatively deep waters of the Northern Flank was reported in April and May, which is later than peak spawning reported in other regions.

July Aug. Sept. Oct. Nov. Dec. Jan. Feb. Mar. Apr. May June Nantucket Shoals and Channel (n = 27) 0% Closed Area 1 (n = 7) 1 - 10% Northern Flank (n = 18) 10.1 - 20% Georges Shoal (n = 20) 20.1 - 50% Southwest Part (n = 8) >50% Closed Area 2 (n = 12) Northeast Peak (n = 12) Eastern Georges (n = 4) Figure 16 - Proportion of fishermen that reported spawning activity each month in the different geographic regions of Georges Bank. Fishermen typically identified one or more habitat characteristics associated with each cod spawning ground, and these habitat characteristics included both abiotic and biotic attributes. Cod spawning grounds were most commonly characterized as occurring in areas with sandy substrates, and fishermen often described sand “lumps” as the preferred spawning habitat (Table 11). Rocky, hard bottom habitats and areas with gravel substrate were also identified as

34 important for cod spawning activity. The fishermen often observed that cod spawning grounds were in areas with complex bathymetric features such as ridges, valleys, and deep holes. Fishing in these complex habitats is difficult, particularly with mobile gear, and requires a priori knowledge that the fishermen acquire through direct experience and information sharing among captains. The fishermen also reported that cod spawning grounds were often in areas with high concentrations of shellfish, including surf clams, quahogs, and mussels. Cod spawning aggregations were also reported in some instances to be associated with areas that held high concentrations of forage fish, such as herring, mackerel, or sand lance. We primarily interviewed trawl fishermen, which should be kept in mind when considering the habitat characteristics that were reported by the fishermen. The trawl fleet has historically represented the majority of cod fishing effort on Georges Bank. While some trawl fishermen did report targeting spawning cod on hard and rocky habitats, complex hard-bottom habitats may have been reported more frequently if we had interviewed more fixed gear fishermen.

Abiotic Habitat Attributes Biotic Habitat Attributes Sand or sand lumps 105 Shellfish 50 Rocks and "hard" bottom 67 Herring 9 Gravel 36 Mackerel 3 Pebble and cobble 11 Sand lance 8 Mud 9 Macroalgae 5 Shipwrecks 7 Crabs 1 Strong tides and currents 9 Worms 2 "Broken" bottom 3 Squid 1 Silver hake 1 Sponges 1 Sea Stars 3 Small haddock 1 Table 11 – Habitat characteristics associated with cod spawning grounds. The table shows the frequency with which each habitat characteristic was reported. Fishermen reported that cod spawning occurs across a wide range of depths on Georges Bank and Nantucket Shoals (Figure 17). Although the bulk of spawning activity was reported to occur in depths ranging from 20 to 91 meters, fishermen also reported that cod also spawn in shoal habitats (< 20m) and in deep water (> 165m) off the Northern Edge of Georges Bank.

Figure 17 – Depth range (m) of cod spawning grounds reported by fishermen. Of the 210 cod spawning grounds identified in this study, in 28 instances fishermen indicated that the spawning ground was no longer active. Fishermen identified 60 spawning grounds that they thought were still active. However, in the majority of cases (122 of 210 spawning grounds, 58%), fishermen were unsure whether the spawning ground they identified was still active or not. Many fishermen participating in this study expressed concern that the current management units for cod in U.S. waters do not match the biological population structure of the resource. We asked fishermen if they thought there was any connectivity between the spawning sites they identified (Part 3 of Table 9), with the largest focus on inshore versus offshore areas. Sixteen fishermen who had experience fishing in both areas answered this question, and 15 of the 16 fishermen believed that cod on eastern Georges Bank were distinct from those on Nantucket Shoals and western Georges Bank. In particular, fishermen noted that cod on eastern Georges Bank often attained larger sizes than those taken inshore. Fishermen also consistently described differences in the color, shape, and fillet quality of cod between the two areas. Many fishermen identified a longitudinal divide that they believe separate the two groups of cod on eastern and western Georges Bank cod, and most suggested a boundary line of either 68o W or 69oW, with many suggesting that the Great South Channel acts as a boundary between inshore and offshore cod groups. Ten fishermen remarked that there is connectivity between cod in the Gulf of Maine and Nantucket Shoals, and observed that cod make seasonal migrations from the western Gulf of Maine to Nantucket Shoals in search of food. Fishermen also distinguished 47 feeding grounds that had high catch rates of cod during certain times of the year (Appendix B, Figure B26). While cod feeding grounds were reported year round, the greatest number of feeding grounds was reported in June, July, and August, after the spawning season had ended (Appendix B, Figure B27). Many of the feeding grounds were reported by Canadian fishermen on Eastern Georges Bank (n = 15), and the timing of their

36 reports were likely influenced by seasonal fishery regulations, which prohibit fishing on Eastern Georges Bank from February to May to protect spawning cod (Wang et al., 2015). Haddock spawning grounds (n = 13) were identified by eight fishermen (Appendix B, Figure B28). These fishermen reported spawning to occur from October to July on Georges Bank, with the greatest number of spawning grounds (n = 5) reported in May, June, and July. Three fishermen also identified winter flounder spawning grounds (Appendix B, Figure B29). Only one spawning ground was identified for both pollock and yellowtail flounder. Results: Common Themes and Observations Although the primary goal of this research was to solicit FEK related to the timing and distribution of cod spawning on Georges Bank, other topics often arose during the interviews. The detailed notes we kept allowed us to summarize some of these recurring themes. Fishermen commonly discussed climate change as a major factor that drives the distribution and abundance of fish stocks on Georges Bank. Many fishermen commented that the distribution of fish has changed significantly over the course of their careers. Some fishermen remarked that warm water species such as striped bass, black sea bass, and fluke were rarely encountered on Georges Bank at the start of their careers, but that those species have become increasingly common over time. Many fishermen are concerned that warmer water temperatures are contributing to the poor stock status of Georges Bank cod. One fisherman commented, “Cod love the cold water, and the water now is different than it used to be.” A similar sentiment was repeated across many of the interviews. Many fishermen, especially those from New Bedford, expressed concern that clam fishing has had a negative impact upon cod spawning activity. The fishermen noted that clam fishing began on Nantucket Shoals in the late 1980’s, and that clam fishing was coincident with a decline in the abundance of spawning cod on Nantucket Shoals. The fishermen were also concerned that clam fishing has had a negative impact upon cod spawning activity across the shallow waters of central Georges Bank (e.g., Georges Shoals and Cultivator Shoals). Many fishermen felt that shellfish (e.g., clams, quahogs, and mussels), as well as the worms that are often abundant at these shellfish beds, are an important food source for cod, and that cod would seek out shellfish beds as spawning habitats so that they would have a readily available food source immediately after spawning. One fisherman commented, “When we caught the cod with their mouths all busted and cut up, we knew they were feeding on the clams.” The fishermen perceived that clam fishing caused widespread habitat damage, and eliminated an important food source for the post-spawning cod, which may have led to the extirpation of cod spawning grounds. One fisherman remarked, “The clam boats systematically destroy the clam beds and it definitely affects the fish. If the clams are gone the cod leave.” Another fisherman commented, “After the food disappears for a couple of generations, it is not in their instinct to go back to those places.” Another topic that the fishermen very often discussed was the interaction between seals and cod spawning activity. Nearly all fishermen perceived that the abundance of seals had increased markedly over the course of their careers, and the fishermen reported that they have

37 observed seals foraging on Georges Shoals, Browns Bank, and on the Northeast Peak in recent years. Many fishermen, especially those fishing from ports on Cape Cod and Canada, expressed concern that seals disrupt cod from schooling and spawning, and commented that seals can hunt cod effectively in water depths < 50 fathoms. These fishermen felt that cod have been driven out of their traditional inshore spawning grounds on the back side of Cape Cod by seal predation. One fisherman commented, “Until we address the issue of marine mammals we will never rebuild the cod fishery. We can’t keep ignoring the impacts that seals are having.” Many fishermen expressed their support for short term closures to protect cod spawning activity, provided that these closures are well-informed, and do not unduly restrict access to harvesting other stocks. One fisherman commented, “It matters when you catch them, you should let them spawn first before you catch them.” Others argued that spawning closures were ineffective, and that the most important responsibility of management was to set the quota at the appropriate level. Another common topic that was discussed during the interviews was the catchability of trawl gear in relation to the tide. Many fishermen commented that the tide has a large influence on the efficiency of trawl gear, and noted that the greatest catches of cod and flatfish occur during slack tide. This phenomenon was especially important on Georges Bank, where there are often strong tidal currents. One fisherman noted that his cod catch rates were typically increased by 300% to 400% during slack tide, and other fishermen commented that they would plan their cod fishing to ensure they made their favorite (i.e., highest CPUE) tow locations during slack tide. Fishermen also discussed the decline in the number of active fishing vessels, and fishing effort that they observed over the course of their careers. One fisherman noted, “It used to be like a city on Georges Shoals, now it is empty,” while another fisherman lamented, “I can fish for six or seven days and not have anyone to talk to on the radio.” In addition, many fishermen commented that their fishing behavior has totally changed in response to reductions in the cod quota, and they noted that very few fishermen actually target cod anymore. Many have changed their gear and fishing tactics, and now fish primarily in deeper waters where they target haddock, hake, monkfish, and pollock. Finally, a number of fishermen talked about technological advances in the fishery and how that changed their ability to target cod. Prior to the advent of Loran C, the fleet used Loran A which was not very accurate, so the fishermen would have to rely on visual aids such as buoys to avoid running aground in shallow areas. Many fishermen noted that the advent of Loran C and GPS allowed them to fish with more confidence in dangerous areas, such as Nantucket Shoals. New Bedford fishermen commented that the cod fishery on Nantucket Shoals really accelerated in the 1970’s as the development of Loran C allowed them to safely fish on the shoals. In particular, fishermen were then able to target cod in small scale areas like the “Lagoon” and “Old Man Shoal” that were flanked by dangerous shoals (Figure 14). In addition, fishermen discussed the importance of advances in depth sounders and fish finders, and how this increased their ability to target schooling fish like cod. Some fishermen thought that this new technology allowed fishing to become unsustainable, and one fisherman commented, “The

38 predictability of the fish, combined with the advent of Loran C and GPS was the downfall of the fishery.”

Phase 3: Synthesis of Scientific Data and Fishermen’s Ecological Knowledge The timing and location of cod spawning reported by fishermen was compared to the available scientific data that was summarized in Phase 1. The timing of spawning reported in each region of Georges Bank in the scientific literature is also summarized below in Table 12. Fishermen’s reported timing and location of cod spawning activity generally agreed with the scientific literature, but spawning dynamics are not well represented by seasonal trawl surveys that do not routinely sample Nantucket Shoals and Georges Shoals. Across all of Georges Bank and Nantucket Shoals, the timing of spawning reported by the fishermen (Figure 15) was in close agreement with results from the MARMAP ichthyoplankton sampling program (Berrien and Sibunka, 1999, Figure 11). Fishermen reported peak spawning activity from January to March, which matches closely with data collected by the DFO trawl survey and the Canadian Fisheries Observer program. Many of the spawning grounds identified by fishermen had been noted from prior scientific research, but the reports from fishermen suggest that spawning grounds are more geographically complex than was previously recognized.

Geographic Region Timing Criteria Reference Nantucket Shoals November - April Direct observation Schroeder, 1930 November - January Holding study Smith, 1902 Late autumn and early winter Unknown Bigelow and Schroeder, 1953 November Direct observation Kovach et al., 2010 November Direct observation Wirgin et al., 2007 November - March Eggs present (all stages) Berrien and Sibunka, 1999 Georges Shoals November - April Eggs present (all stages) Berrien and Sibunka, 1999 March - April Small catches of spawning cod NEFSC spring trawl survey Closed Area 2 February - April Direct observation Goode, 1884; Rich, 1929 February - April Direct observation Bigelow and Schroeder, 1953 March - April Small catches of spawning cod NEFSC spring trawl survey February - March Small catches of spawning cod DFO trawl survey February - April Stage one cod eggs GLOBEC sampling program Southwest Part March - April Stage one eggs GLOBEC sampling program March - April Very small catches of spawning cod NEFSC spring trawl survey February - March Very small catches of spawning cod DFO trawl survey Northern Edge March - April Small catches of spawning cod NEFSC spring trawl survey January - April Stage one cod eggs GLOBEC sampling program Northeast Peak February - March Large catches of spawning cod DFO trawl survey March - April Large catches of spawning cod NEFSC spring trawl survey January - February Direct observation Canadian fisheries observer program January - April Stage one cod eggs GLOBEC sampling program February Direct observation Wirgin et al., 2007 February - March Direct observation Kovach et al., 2010 Eastern Georges Bank February - March Large catches of spawning cod DFO trawl survey March - April Moderate catches of spawning cod NEFSC spring trawl survey January - May Stage one cod eggs GLOBEC sampling program January - February Direct observation Canadian fisheries observer program January - May Eggs present (all stages) Berrien and Sibunka, 1999 Table 12. Summary of the timing of cod spawning on different regions of Georges Bank as reported in the scientific literature.

Although fishermen reported that the spawning season on Nantucket Shoals is protracted, they identified November and December as the peak spawning season in this area (Figure 16). This timing aligns well with the scientific information. Sampling conducted during the MARMAP program observed a relatively high density of cod eggs in the waters around Nantucket Shoals in November and December, which may indicate spawning activity in the region (Berrien and Sibunka, 1999). Schroeder (1930) documented spawning on Nantucket Shoals from November to April, with peak spawning in December and January, while Smith (1902) reported that spawning lasted from November until January, and Bigelow and Schroeder (1953) documented spawning in “late autumn and early winter”. The four spawning grounds on Nantucket Shoals described by Schroeder (1930) (Figure 4 from section 1) were identified by three or more fishermen during the interviews, as were a number of other spawning grounds in the region. The NEFSC trawl survey occasionally samples the consensus spawning grounds that were identified on eastern Nantucket Shoals (“Asia Rip”, ”Middle Rip” and “Fishing Rip”), and small catches of spawning cod have been observed near these spawning grounds during both the NEFSC fall and spring surveys. However, consensus spawning grounds on western Nantucket Shoals (e.g., “The Lagoon”, “The Fingers”) are outside of offshore strata 9 to 25, and NEFSC survey tows made in these locations are not included in the Georges Bank cod stock assessment. Schroeder (1930) discussed the seasonal migration of cod on Nantucket Shoals in detail, and documented a year-round resident population of cod. Bigelow and Schroeder (1953) also claimed that there were both resident and migratory cod on Nantucket Shoals. Three fishermen with extensive experience fishing for cod in this region made similar observations, and referred to the cod inhabiting the area year round as “groundskeepers”. These fishermen also described migratory cod that visited Nantucket Shoals in the late fall and winter, and observed that cod migrate from the deeper waters of Great South Channel and potentially from the Gulf of Maine to Nantucket Shoals in the late fall. The cod would first appear on the eastern portion of Nantucket Shoals (e.g., Davis Shoal) and then would migrate further to the western shoals (e.g., Old Man Shoal) in the winter, where they would mix with the resident group of fish. Some fishermen we interviewed thought that this westward migration occurred after the first Nor’easter of the season, the effect being that the stormy north easterly winds would mix and bring in cool water that cod preferred. The fishermen observed that some of these cod would spawn on Nantucket Shoals, while others would continue migrating further to the west to spawn in other locations (presumably Coxes Ledge and south of Martha’s Vineyard). Fishermen reported widespread spawning activity on Georges Shoals that occurs from December to April, and identified several spawning grounds in this region (Figures 13, 14 and Appendix B). Interestingly, Georges Shoals had not been identified as a cod spawning ground in prior scientific reports. Cod eggs were sampled on Georges Shoals during the MARMAP program, primarily from November to April (Berrien and Sibunka, 1999), but were not commonly observed on Georges Shoals during the GLOBEC program. However, the MARMAP dataset included cod eggs of all stages, and it is possible that those eggs could have been spawned elsewhere and advected onto Georges Shoals. Due to their small size and the

41 complex bathymetry in this region, the consensus spawning grounds identified on Georges Shoals are not routinely sampled during the NEFSC trawl survey. However, relatively small catches of spawning cod were observed in close proximity to the “Northeast of Georges Shoal” and “Northeast Rip” consensus spawning grounds during the NEFSC spring trawl survey. Goode (1887b) and Rich (1929) described large aggregations of spawning cod on the “Winter Fishing Grounds” that occurred predictably every year in February, March, and April (Figure 3). A number of the consensus spawning grounds identified in the present study are within the “Winter Fishing Grounds” including “The Leg”, “Billy Doyle’s Hole”, “The Clover”, “East of the Leg”, “Northeast Rip”, and “Closed Area 2” (Figure 14). The timing of spawning reported by Goode (1887b) and Rich (1929) matches closely with the contemporary reports from the fishermen who indicated that most of the spawning activity occurs at “The Leg” from January to March and “Billy Doyle’s Hole” from January to March. Spawning cod were routinely caught in this region during the NEFSC spring trawl survey, and many of the cod spawning hotspots from the NEFSC spring survey were located in the “Winter Fishing Grounds”. Spawning cod were also observed in the “Winter Fishing Ground” during DFO trawl surveys, and stage one cod eggs were sampled in this area from January to April during the GLOBEC program. Only one of the forty fishermen identified spawning activity in the middle of Closed Area II that overlaps with the spawning grounds on Georges Bank that were reported by Bigelow and Schroeder (1953; Figure 3 in Section 1). The greatest concentrations of spawning cod observed during both the NEFSC spring and DFO trawl surveys were on the Northeast Peak of Georges Bank, and the majority of cod spawning hotspots identified by the DFO trawl survey were on the Northeast Peak. Spawning cod were routinely sampled in this area from January to March by the Canadian Fisheries Observer Program. Cod eggs were also abundant on the Northeast Peak during the MARMAP and GLOBEC sampling programs, with peak spawning noted in February and March (Berrien and Sibunka, 1999; Sibunka et al., 2006). The Northeast Peak has long been known to serve as a cod spawning ground (Rich, 1929), and spawning cod sampled on the Northeast Peak were genetically distinct from those sampled on Nantucket Shoals (Weiss et al., 2005; Kovach et al, 2010). Although there is ample scientific evidence of cod spawning on eastern Georges Bank, relatively few fishermen identified spawning grounds on the Canadian side of Georges Bank, and only two consensus spawning grounds were identified on the Northeast Peak, while none were observed on Eastern Georges Bank. These results are likely due in part to fishery regulations. U.S. fishermen have been prohibited from fishing on eastern Georges Bank since the implementation of the Hague Line in 1984. Some of the older US captains we interviewed reported fishing on eastern Georges Bank at the start of their careers, but many of the captains we interviewed had not fished in this area. We made an effort to interview retired trawl captains who fished on eastern Georges Bank in the 1960’s and 1970’s, but many of these captains are deceased, or have moved out of New England. We had hoped to gain more information about cod spawning on eastern Georges Bank during our trip to interview fishermen in Nova Scotia, but Canadian fishermen are prohibited from fishing on eastern

Georges Bank during the peak spawning season (February to May), and as such their ability to identify cod spawning grounds on eastern Georges Bank was limited. Few fishermen identified cod spawning activity on the Southwest Part, and no consensus spawning grounds were observed in this region. Relatively small catches of spawning cod were observed on the Southwest Part during both the NEFSC spring and DFO trawl surveys. Cod eggs were routinely sampled on the Southwest Part during the MARMAP and GLOBEC ichthyoplankton surveys (Berrien and Sibunka, 1999; Sibunka et al, 2006), although it is unclear if these eggs were spawned on the Southwest Part, or advected from other spawning locations. Fifteen fishermen felt that cod on eastern Georges Bank are separate from those on western Georges Bank and Nantucket Shoals. Fishermen’s perception of distinct groups of cod on eastern and western Georges Bank, with a longitudinal boundary near 68o W, is consistent with the boundary that Wise (1963) suggested from tagging observations. Tallack (2009) and Loehrke (2013) also observed that few cod moved between Georges Bank and southern New England. Schroeder (1930) also noted that few cod tagged on Nantucket Shoals moved eastward to Georges Bank. Genetic differences have been observed between cod on Georges Bank and those on Nantucket Shoals (Lage et al., 2004; Weiss et al., 2005; Kovach et al., 2010). Fishermen also described connectivity between cod on Nantucket Shoals, the Great South Channel, and the western Gulf of Maine. These observations agree with a genetic study completed by Kovach et al., (2010), which identified a “southern complex” of winter spawning cod in the Gulf of Maine and locations south of Cape Cod, Massachusetts. Based on conventional tag returns, Tallack (2009, 2011) observed considerable movement of cod between Cape Cod and the inshore Gulf of Maine, although some cod were also reported to migrate from Cape Cod to Georges Bank. Coverage of the Consensus Spawning Grounds by the NEFSC and DFO Trawl Surveys Data from the DFO and NEFSC trawl surveys were examined to evaluate whether the trawl surveys could corroborate the “consensus” spawning grounds identified by the fishermen. The location of tows completed during the NEFSC and DFO trawl surveys were plotted, and the “Spatial Join” tool in ArcGIS was used to identify tows that occurred within any of the “consensus” spawning grounds. For the NEFSC trawl survey, the analysis was limited to survey tows completed in offshore strata 9 through 25, which are the strata included in the Georges Bank stock assessment (NEFSC, 2013). For NEFSC survey tows completed from 1970 through 1995, only the starting location of the tow was available in the database that we obtained from the NEFSC. Therefore, the analysis was limited to survey tows whose starting point fell within one of the “consensus” spawning grounds, which likely underestimates the number of survey tows that actually occurred within a spawning ground. Both the start and end locations were available for survey tows completed from 1996 through 2014. These tow tracks were plotted in ArcGIS using the “XY to Line” tool, and the “Spatial Join” tool was used to identify tows that intersected any of the “Consensus” spawning grounds. This method assumes that all tows were completed in a straight line, which may underestimate the spatial extent of the tow. The same method was used to identify tows completed during the DFO trawl survey that occurred within

43 the “consensus” spawning grounds. For tows completed from 1987 through 1989 only the starting locations were provided in the data we obtained from the DFO. Both the start and end locations were provided for survey tows completed from 1990 through 2015. The same procedures were performed as used with the NEFSC trawl survey data when end locations were or were not available. For tows completed in offshore strata 9 through 25 during the NEFSC spring trawl survey, 97 of the 3,979 tows occurred within one of the “consensus” spawning grounds, while 106 of the 4,151 tows completed during the fall survey occurred in a “consensus” spawning ground (Table 12, Appendix C, Figure C1 and C2). Nine “consensus” spawning grounds have never been sampled by the NEFSC spring and fall trawl survey, while 14 have been sampled infrequently (n < 10 tows) over the time series of the spring and fall survey. Only two of the “consensus” spawning grounds we identified, the Northern Edge (East) and Northern Edge (40-60 fathoms), were routinely sampled during the spring and fall NEFSC surveys. For the Northern Edge (East) spawning ground, spawning cod were observed in 54% of the spring survey tows, but only 3% of the fall survey tows. The 12 fishermen that identified cod spawning activity on the Northeast Peak indicated that the spawning season in this region is protracted, and may extend from January to August, although a peak spawning season was not identified. The trawl survey data support the fishermen’s observations that the Northern Edge (East) is an active cod spawning ground in March, April, and May. For the Northern Edge (40-60 fathoms) spawning ground, spawning cod were observed on 15% of spring survey tows, and 5% of fall survey tows. Fishermen identified that the peak spawning in this region occurs in April and May (Figure 16), which coincides with the timing of the NEFSC spring trawl survey. The trawl survey suggests that some cod spawning activity occurs within the Northern Edge (40-60 fathoms) spawning ground that was identified by the fishermen, but spawning may occur within discrete portions of this area. For the DFO trawl survey, only 59 tows were completed in the “consensus” spawning grounds, and most (n = 41) were completed in the Northern Edge (East) spawning grounds (Table 12; Appendix C, Figure C3). Survey tows were also completed infrequently in three other spawning grounds (Northern Edge (40-60 fathoms), The Clover, the Baseball Bat, and The Leg). The majority of DFO trawl survey tows (61%) completed in the Northern Edge (East) spawning ground captured spawning cod, and ripe and running cod were observed on 29% of these tows. The DFO trawl survey data appears to confirm the fishermen’s observation that the Northern Edge (East) area is an active cod spawning ground in February and March.

NMFS Spring Trawl Survey NMFS Fall Trawl Survey DFO Trawl Survey # Tows # Tows # Tows with # Tows with Without # Tows with # Tows with Without # Tows with # Tows with Consensus Spawning Ground Name # Spawning Ripe and Maturity # Spawning Ripe and Maturity # Spawning Ripe and Tows Cod Running Cod Data Tows Cod Running Cod Data Tows Cod Running Cod Old Man Shoal 0 - - - 0 ------The Lagoon 0 - - - 0 - - - 0 - - Old South Shoal 0 - - - 0 - - - 0 - - Davis Shoal 0 - - - 0 - - - 0 - - The Fingers (North) 0 - - - 0 - - - 0 - - The Fingers (South) 0 - - - 0 - - - 0 - - The Lemons 2 1 0 0 1 0 0 1 0 - - Crab Ledge and Pollock Rip 7 4 0 - 6 1 0 1 0 - - Fishing Rip and the Shoals 5 0 0 1 5 1 1 - 0 - - Asia Rip 2 2 0 - 4 1 1 1 0 - - Middle Rip 1 1 0 - 2 0 0 - 0 - - Megabucks 1 0 0 - 0 - - - 0 - - Cultivator Shoal 0 - - - 0 - - - 0 - - Georges Shoal 1 0 0 - 1 0 0 - 0 - - Northeast of Georges Shoal 4 1 1 - 1 0 0 - 0 - - The Corner 3 0 0 - 2 0 0 - 0 - - Northern Edge (40-60 fathoms) 33 5 0 - 37 2 0 - 8 1 0 Northern Edge (25-40 fathoms) and The Fence 2 0 0 - 3 0 0 - 0 - - Northern Edge (East) 24 13 1 2 30 1 0 1 41 25 12 Closed Area 2 2 1 0 1 2 0 0 - 0 - - Northeast Rip 3 0 0 - 2 0 0 - 0 - - Billy Doyles Hole 1 0 0 - 0 - - - 0 - - The Clover 0 - - - 2 0 0 - 5 1 0 The Leg 3 2 0 - 5 0 0 1 2 1 0 East of the Leg 3 1 0 - 2 0 0 - 0 - - Table 12. Summary of NEFSC and DFO survey tows that were completed within each of the “consensus” spawning grounds identified by the fishermen.

Discussion and Future Research There was substantial agreement between the available scientific information and the information reported by fishermen about the timing and location of cod spawning on Georges Bank. Our results demonstrate that FEK can enhance our scientific understanding of cod spawning behavior on Georges Bank. Although the major spawning grounds reported by fishermen were identified in the scientific literature (e.g., the Northeast Peak and Nantucket Shoals), the fishermen we interviewed had extensive, fine-scale knowledge related to the timing and location of cod spawning, and identified some cod spawning grounds that were not previously described in scientific reports. Multiple fishermen who were interviewed independently indentified several fine-scale geographic regions (often < 50km2) where cod spawning occurs, and they had detailed knowledge of the habitat characteristics of these fine- scale locations. This level of spatial resolution could not typically be obtained using traditional scientific data collection approaches. One objective of this study was to use FEK to examine long-term shifts in the distribution of cod spawning activity on Georges Bank and Nantucket Shoals. However, meeting this objective proved to be surprisingly difficult. Nearly all of the fishermen we interviewed reported that they no longer target cod because of the low quotas that have been allocated in recent years. The majority of fishermen noted that they no longer fish in areas where cod are known to be abundant in order to target stocks with higher allocations such as haddock and to avoid exceeding their cod quota. Fishing is now prohibited in many areas where the US fleet used to target spawning cod. Beginning in 1994 closed areas were implemented to reduce fishing mortality (e.g., Closed Areas I and II), and additional closures have been enacted to protect habitat (e.g., Nantucket Shoals habitat closure). In addition, US fishermen have been banned from fishing on eastern Georges Bank since the Hague Line was established in 1984 (Murawski et al., 2000). On the Canadian portion of eastern Georges Bank, the trawl fleet is required to use selective fishing gear such as the haddock separator trawl, which is designed to reduce cod bycatch, and fishing is typically prohibited from February through May to protect spawning cod (Wang et al., 2015). These regulations, and resulting shifts in fishing behavior, make it difficult to evaluate which spawning grounds remain active on Georges Bank. These regulations are also likely an important factor for explaining why so few fishermen were able to identify cod spawning grounds on eastern Georges Bank, despite the scientific evidence which indicates that major spawning activity occurs in this region. The FEK collected in this study can serve as a valuable guide for future research efforts to better understand the stock dynamics and reproductive capacity of cod on Georges Bank. For example, a directed survey could be initiated to sample at the spawning locations that were identified by fishermen to assess which spawning grounds remain active. Such auxiliary studies are particularly helpful given the presently low biomass of cod on Georges Bank, because programmatic surveys conducted by the National Marine Fisheries Service and Canadian Department of Fisheries and Oceans may not have a sufficient density of observations to precisely monitor spawning activity during stock rebuilding. Furthermore, these auxiliary approaches can provide data on finer spatial and temporal scales, which can be used to inform future fishery management decisions.

In addition, the location and timing of cod spawning grounds that were identified by fishermen could be used to inform the starting conditions for Individual Based Modeling experiments designed to examine the transport and survival of cod eggs and larvae. Finally, these putative spawning areas could be used to guide the design of future tagging studies to investigate movement patterns, and stock structure, including connectivity among spawning components. FEK can serve as an important source of information for stock identification, because fishermen can identify the spawning locations, movement patterns, and morphological characteristics of their target species (Neis, 1998; Murray et al., 2008b). In this study, fishermen’s accounts suggested that cod frequently move across stock boundaries (e.g., from Cape Cod and Nantucket Shoals to the Gulf of Maine), and also identified persistent morphological differences and migratory patterns between cod on eastern and western Georges Bank. These findings are in agreement with a recent multidisciplinary investigation of cod stock structure (Zemeckis et al., 2014c), and have important implications for the assessment and management of the resource. We propose that FEK should routinely be collected as part of multidisciplinary stock identification studies along with traditional stock identification techniques such as genetics, analysis of life history traits, and applied marks. Collecting FEK is an effective way to increase dialogue between fishermen and scientists, and FEK can provide valuable information relevant to fishery management (Macdonald et al., 2014; Hedeholm et al., 2016). There is increasing recognition that FEK is part of the best available science, and as such, it should play a larger role in fishery management (Johannes et al., 2000; Stephenson et al., 2016). However, FEK is not typically integrated into management, often because of the difficulties associated with standardizing and synthesizing FEK effectively (Anuchiracheeva et al., 2003; Hind, 2015). In this study, we relied on nautical charts and GIS mapping to standardize the spatial information provided by the fishermen, and we set a criterion that consensus spawning grounds needed to be identified independently by three or more fishermen. By employing a semi-structured interview approach that asked each fisherman the same set of questions, we were able to collect information that could be tabulated and compared across interviews. By following these steps, we believe the information collected in this study can help guide management actions and future research efforts. The high participation rates in this study were a result of several factors. First, through prior cooperative research projects we had a pre-existing relationship with many of these fishermen, enabling us to gain their familiarity and trust. Secondly, we conducted the interviews at times and locations that were convenient to the fishermen, which didn’t cause them to lose fishing opportunities, and showed them that we valued their perspectives. Many fishermen were thankful for the opportunity to share their knowledge and expertise, and appreciated the ability to contribute to scientific research. In addition, one of the scientists was fluent in Portuguese, which allowed us to interview several fishermen that have extensive experience fishing for Georges Bank cod, but often do not participate in fishery management because of a language barrier. Many of the fishermen that participated were retired, and these fishermen may have been more willing to share their information because they would not be affected by any

47 resultant management measures. Finally, given the depleted status of Georges Bank cod, some fishermen may have been motivated to participate out of a desire to share information that could be used to help rebuild the resource.

Project Participants - Scientific Collaborators Micah J. Dean, William Hoffman, and Dr. Michael P. Armstrong - Massachusetts Division of Marine Fisheries, Annisquam River Marine Fisheries Station, Gloucester, MA 01930

Dr. Richard McBride - National Marine Fisheries Service, Northeast Fisheries Science Center, 166 Water Street, Woods Hole, MA 02556

Dr. Robert Stephenson, Dr. Yanjun Wang, and Irene Andrushchenko - Canadian Department of Fisheries and Oceans, St. Andrews Biological Station, 531 Brandy Cove Road, New Brunswick E5B 2L9

Project Participants - Industry Collaborators and Fishing Organizations Bill and Jason Amaru, F/V Joanne A III, Chatham, MA Rodney Avila, F/V Trident, New Bedford, MA Peter Cura, F/V Fisherman, New Bedford, MA Eric Hesse, F/V Tenacious, Chatham, MA Toby Lees, F/V Seel, Fairhaven, MA Ted Ligenza, F/V Rienna Marie, Chatham, MA Richie Canastra, New Bedford Seafood Display Auction, New Bedford, MA

Partnerships This project would not have been possible without the support and participation of fishermen throughout New England and Nova Scotia. This project provided an excellent opportunity for the Principal Investigators to interact with fishermen throughout the region, and provided the fishermen with a unique opportunity to document and discuss their ecological knowledge and observations. In particular, this project allowed Greg DeCelles and Crista Bank to travel to Nova Scotia, where they heard the perspectives and observations of Canadian fishermen, who work under a different set of regulations, and fish in different regions than U.S. fishermen. Many of the fishermen enjoyed the interviews, and valued the opportunity for their knowledge to be considered as part of the best available science. We hope that the partnerships that were formed as part of this study can be built upon through future research and field work to better understand the spawning dynamics of cod on Georges Bank and Nantucket Shoals.

Future Deliverables and Reports A manuscript that discusses the major findings of this study has been submitted for publication in the ICES Journal of Marine Science in January 2017. David Martins is using research completed as part of this project for his thesis to earn a Master’s Degree from the School for Marine Science and Technology, University of Massachusetts Dartmouth. David expects to complete his degree by the end of 2017. David’s thesis will further investigate the biological and habitat characteristics associated with the cod spawning grounds. A peer-reviewed manuscript will be submitted at the conclusion of Dave’s thesis research.

Impacts and Applications This research will also be made available to interested scientific parties, including the New England Fishery Management Council’s Groundfish Committee and the Groundfish Plan Development Team. In addition, this report will be provided for consideration at future cod stock structure workshops.

Presentations DeCelles, G., Martins, D., Zemeckis, D., and Cadrin, S.X. Using Fishermen’s Ecological Knowledge and Scientific Data to Map Cod Spawning Grounds on Georges Bank and Nantucket Shoals. Southern New England Chapter of the American Fisheries Society Winter Meeting. January, 2016. Avery Point, CT. DeCelles, G., Martins, D., Zemeckis, D., and Cadrin, S.X. Using Fishermen’s Ecological Knowledge and Scientific Data to Map Cod Spawning Grounds on Georges Bank and Nantucket Shoals. University of New Hampshire Biology Department Seminar Series. February, 2016. Durham, NH. Martins, D., DeCelles, G., Zemeckis, D., Bank, C., and Cadrin, S.X. Utilizing Fishermen’s Ecological Knowledge and Scientific Data to Map Atlantic Cod Spawning Grounds on Georges Bank and Nantucket Shoals. University of Massachusetts Intercampus Marine Science Symposium. March 25th, 2016. Dartmouth, MA. DeCelles, G., Martins, D., Zemeckis, D., Bank, C., and Cadrin, S.X. Mapping the Distribution of Atlantic Cod Spawning Activity on Georges Bank and Nantucket Shoals. American Fisheries Society Annual Meeting. August 23rd, 2016. Kansas City, MO.

Acknowledgements Funding for this research was provided by the New England Fishery Management Council through a grant administered by the Northeast Consortium. We are indebted to the fishermen who generously shared their time and knowledge with us. We are especially grateful to Gilbert Donaldson, who helped us contact fishermen in Nova Scotia and schedule interviews. Data requests were fulfilled by Sean Lucey, and Jon Hare at the NOAA Northeast Fisheries Science Center. We would also like to thank Crista Bank who assisted with interviews, and who traveled to Nova Scotia to interview Canadian fishermen. Dr. David Pierce from the Massachusetts Division of Marine Fisheries provided valuable feedback on this project. We also thank Dr. Greg Skomal for his support of the project.

References Alexander, A.B., Moore, H.F., and Kendall, W.C. 1914. Otter-trawl fishery. Appendix VI to the Report of the U.S. Commissioner of Fisheries for 1914. Ames, T. 1998. Cod and haddock spawning grounds in the Gulf of Maine from Grand Manan Channel to Ipswich Bay. In The Implications of Localized Fisheries Stocks, pp. 55-64. Ed. by I.H. von Herbing, and M. Tupper. Natural Resource, Agriculture, and Engineering Service, Ithaca, New York. 200 pp.

Ames, E.P. 2004. Atlantic cod stock structure in the Gulf of Maine. Fisheries, 29(1): 10-28. Anuchiracheeva, S., Demaine, H., Shivakoti, G.P., and Ruddle, K. 2003. Systematizing local knowledge using GIS: fisheries management in Bang Saphan Bay, Thailand. Ocean and Coastal Management, 46: 1049-1068. Armstrong, M.P., Dean, M.J., Hoffman, W.S., Zemeckis, D.R., Nies, T.A., Pierce, D.A., Diodati, P.J., et al. 2013. The application of small scale fishery closures to protect Atlantic cod spawning aggregations in the inshore Gulf of Maine. Fisheries Research, 141: 62-69. Babbie, E. 1989. The Practice of Social Research, 5th Edition. Wadsworth, Belmont, California. Bergmann, M., Hinz, H., Blyth, R.E., Kaiser, M.J., Rogers, S.I., and Armstrong, M. 2004. Using knowledge from fishers and fisheries scientists to identify possible groundfish ‘Essential Fish Habitats’. Fisheries Research, 66: 373-379. Berrien, P.L., and Sibunka, P. 1999. Distribution patterns of fish eggs in the U.S. northeast continental shelf ecosystem, 1977-1987. NOAA Technical Memorandum NEFSC 145. Bigelow and Schroeder. 1953. Fishes of the Gulf of Maine. Fishery Bulletin of the Fish and Wildlife Service. Burnett, J., O'Brien, L., Mayo, R.K., Darde, J., and Bohan, M. 1989. Finfish maturity sampling and classification schemes used during Northeast Fisheries Center bottom trawl surveys, 1963- 1989. NOAA Tech. Mem. NEFSC-F/NEC, vol. 76, 14pp. Churchill, J.H., Runge, J., and Chen, C. 2011. Processes controlling retention of spring-spawned Atlantic cod (Gadus morhua) in the western Gulf of Maine and their relationship to an index of recruitment success. Fisheries Oceanography, 20(1): 32-46. Colton, J.B., Smith, W.G., Kendall, A.W., Berrien, P.L., and Fahay, M.P. 1979. Principal spawning areas and times of marine fishes, Cape Sable to Cape Hatteras. Fishery Bulletin, 76(4): 911-915. Cushing, D.H. 1981. Fisheries Biology: A study in population dynamics. University of Wisconsin Press, Madison, WI. Cushing, D.H. 1990. Plankton production and year-class strength in fish populations: an update of the match/mismatch hypothesis. Advances in Marine Biology, 26: 249-293.

Dean, M.J., Hoffman, W.S., and Armstrong, M.P. 2012. Disruption of an Atlantic cod spawning aggregation resulting from the opening of a directed gill-net fishery. North American Journal of Fisheries Management, 32(1): 124-134. Dean, M.J., Hoffman, W.S., Zemeckis, D.R., and Armstrong, M.P. 2014. Fine-scale diel and gender-based patterns in behaviour of Atlantic cod (Gadus morhua) on a spawning ground in the western Gulf of Maine. ICES Journal of Marine Science, 71(6): 1474-1489. Despres-Patanjo, L.I., Azaroviz, T., and Byrne, C.J. 1988. Twenty-five years of fish surveys in the northwest Atlantic: the NMFS Northeast Fisheries Science Center’s bottom trawl survey. Marine Fisheries Review, 50(4): 69-71. Edwards, S. F., and Murawski, S. A., 1993. Potential Economic Benefits from Efficient Harvest of New England Groundfish. North American Journal of Fisheries Management, 13: 437-449

Getis, A., and Ord, J.K. 1992. The analysis of spatial association by use of distance statistics. Geographical Analysis, 24(3): 189-206.

Goode, G.B. 1887a. The fisheries and fishing industry of the United States. Section II. The cod, haddock, and hake fisheries. United States Commission of Fish and Fisheries.

Goode, G.B. 1887b. The fisheries and fishing industry of the United States. Section III: The fishing grounds of North America with forty-nine charts. United States Commission of Fish and Fisheries.

Grabowski, T.B., Thorsteinsson, V., McAdam, B., and Marteinsd r, G. 2011. Evidence of segregated spawning in a single marine fish stock: Sympatric divergence of ecotypes in Icelandic cod? PLoS ONE, 6(3): 1-9.

Gurshin, C.W.D., Howell, W.H., and Jech, J.M. 2013. Synoptic acoustic and trawl surveys of spring-spawning Atlantic cod in the Gulf of Maine cod spawning protection area. Fisheries Research, 141: 44-61.

Hare, J.A., and Richardson, D.E. 2014. The use of early life stages in stock identification studies. In Stock Identification Methods, 2nd edn, pp. 329-364. Ed. by S.X. Cadrin, L.A. Kerr, and S. Mariani. Elsevier, San Diego. 566 pp. Hedeholm, R.B., Jacobsen, R.B., and Nielsen, E.E. 2016. Learning from ‘apparent consensus’ in TAC disputes: exploring knowledge overlaps in LEK and genetic categorization of Atlantic cod. Marine Policy, 69: 114-120. Hind, E.J. 2015. A review of the past, the present, and the future of fishers’ knowledge research: a challenge to established fisheries science. ICES Journal of Marine Science, 72(2), 341-358. Huret, M., Runge, J.A., Chen, C., Cowles, G., Xu, G., and Pringle, J.M. 2007. Dispersal modeling of fish early life stages: sensitivity with application to Atlantic cod in the western Gulf of Maine. Marine Ecology Progress Series, 347: 261-274.

ICES. 2005. Spawning and life history information for North Atlantic cod stocks. ICES Cooperative Research Report, No. 274. 152 pp. Johannes, R.E., Freeman, M.M.R., and Hamilton, R.J. 2000. Ignore fishers’ knowledge and miss the boat. Fish and Fisheries, 1: 257-271. Kovach, A.I., Breton, T.S., Berlinsky, D.L., Maceda, L., and Wirgin, I. 2010. Fine-scale spatial and temporal genetic structure of Atlantic cod off the Atlantic coast of the USA. Marine Ecology Progress Series, 410: 177-195. Lage, C., Kuhn, K., and Kornfield, I. 2004. Genetic differentiation among Atlantic cod (Gadus morhua) from Browns Bank, Georges Bank, and Nantucket Shoals. Fishery Bulletin, 102: 289- 297. Laurence, G.C., and Rogers, C.A. 1976. Effects of temperature and salinity on comparative embryo development and mortality of Atlantic cod (Gadus morhua L.) and haddock (Melanogrammus aeglefinus (L.)). Journal International pour l’Exploration de lar Mer, 36(3): 220-228. Loehrke, J.L. 2013. Movement patterns of Atlantic cod (Gadus morhua) spawning groups off New England. MS Thesis, University of Massachusetts, Dartmouth, MA, USA. Lough, R.G., Hannah, C.G., Berrien, P., Brickman, D., Loder, J.W., and Quinlan, J.A. 2006. Spawning pattern variability and its effect on retention, larval growth and recruitment in Georges Bank cod and haddock. Marine Ecology Progress Series, 310: 193-212. Macdonald, P., Angus, C.H., Cleasby, I.R., and Marshall, C.T. 2014. Fishers’ knowledge as an indicator of spatial and temporal trends in abundance of commercial fish species: megrim (Lepidorhombus whiffiagonis) in the northern North Sea. Marine Policy, 45, 228-239. Maurstad, A. 2002. Fishing in murky waters: ethics and politics of research on fisher knowledge. Marine Policy, 26: 159-166. Meng, K.C., Oremus, K., and Gaines, S. 2016. New England cod collapse and the climate. PLoS ONE, 11(7): e0158487. Morgan, M.J., and Trippel, E.A. 1996. Skewed sex ratios in spawning shoals of Atlantic cod (Gadus morhua). ICES Journal of Marine Science, 53: 820-826.

Morrison, C. 1990. Histology of the Atlantic cod, Gadus morhua: an atlas. Part Three. Reproductive Tract. Canadian Special Publications of Fisheries and Aquatic Sciences 100. 183 p.

Murawski, S.A., Brown, R., Lai, H.L., Rago, P. J., and Hendrickson L., 2000. Large-scale closed areas as a fishery-management tool in temperate marine systems: the Georges Bank experience. Bulletin of Marine Science, 66(3): 775–798

Murray, G., Neis, B., Schneider, D.C., Ings, D., Gosse, K., Whalen, J., and Palmer, C.T. 2008a. Opening the black box: methods, procedures, and challenges in the historical reconstruction of

53 marine social-ecological systems. In Making and Moving Knowledge: Interdisciplinary and Community-Based Research in a World on the Edge, pp. 100-120. Ed. by J.S. Lutz, and B. Neis. McGill-Queen’s University Press, Montreal. 360 pp. Murray, G., Neis, B., Palmer, C.T., and Schneider, D.C. 2008b. Mapping cod: fisheries science, fish harvesters’ ecological knowledge and cod migrations in the northern Gulf of St. Lawrence. Human Ecology, 36: 581-598. Neis. B. 1998. Fishers’ ecological knowledge and the identification and management of localized populations of Atlantic cod (Gadus morhua) in northeast Newfoundland. In The Implications of Localized Fisheries Stocks, pp. 177-188. Ed. by I.H. von Herbing, and M. Tupper. Natural Resource, Agriculture, and Engineering Service, Ithaca, New York. 200 pp.

Neis, B., Felt, L.F., Haedrich, R.L., and Schneider, D.C. 1999a. An interdisciplinary method for collecting and integrating fishers’ ecological knowledge into resource management. In Fishing Places, Fishing People: Traditions and Issues in Small-Scale Canadian Fisheries, pp. 217-238. Ed. by D. Newell and R.E. Ommer. University of Toronto Press, Toronto. 412 pp.

Neis, B., Schneider, D.C., Felt, L., Haedrich, R.L., Fischer, J., and Hutchings, J.A. 1999b. Fisheries assessment: what can be learned from interviewing resource users? Canadian Journal of Fisheries and Aquatic Sciences, 56: 1949-1963.

New England Fishery Management Council (NEFMC). 2009. Northeast Multispecies Fisheries Management Plan - Amendment 16, 905 pp. New England Fishery Management Council (NEFMC). 2013. Omnibus essential fish habitat amendment 2 draft environmental impact statement. Appendix E – synopsis of Closed Area Technical Team analysis of juvenile groundfish habitats and groundfish spawning areas, 67 pp.

NEFSC (Northeast Fisheries Science Center), 2013. 55th northeast regional stock assessment workshop (55th SAW). US Department of Commerce, Northeast Fisheries Science Center Reference Document 13-11; 845 p. NEFSC (Northeast Fisheries Science Center), 2015. Operational assessment of 20 northeast groundfish stocks, updated through 2014. US Department of Commerce, Northeast Fisheries Science Center Reference Document 15-24; 251 p. Palmer, M.C. 2014. 2014 Assessment update report of the Gulf of Maine Atlantic cod stock. US Department of Commerce, Northeast Fisheries Science Center Reference Document 14-14; 119 pp. Page, F.H., Sinclair, M., Naimie, C.E., Loder, L.W., Losier, R.J., Berrien, P.L., and Lough, R.G. 1999. Cod and haddock spawning times on Georges Bank in relation to water residence times. Fisheries Oceanography, 8(3): 212-226.

Pederson, J., and Hall-Arber, M. 1999. Fish habitat: a focus on New England fishermen’s perspectives. American Fisheries Society Symposium, 22: 188-211.

Pershing, A. J., Alexander, M.A., Hernandez, C.M., Kerr, L.A., Bris, A.L., Mills, K.E., Nye, J.A., Record, N.R., Scannel, H.A., Scott, J.D., Sherwood, G.D., and Thomas, A.C. 2015. Slow adaptation in the face of rapid warming leads to collapse of the Gulf of Maine cod fishery. Science, 350: 809-812.

Rich, W.H. 1929. Fishing grounds of the Gulf of Maine. U.S. Department of Commerce, Bureau of Fisheries, Report of the United States Commissioner of Fisheries.

Robichaud, D., and Rose, G.A. 2001. Multiyear homing of Atlantic cod to a spawning ground. Canadian Journal of Fisheries and Aquatic Sciences, 58: 2325-2329. Rose, G.A. 2007. Cod: The ecological history of the North Atlantic fisheries. Breakwater Books Ltd., St. John’s, NF. 592 pp. Rosenberg, A.A., Bolster, W.J., Alexander, K.E., Leavenworth, W.B., Cooper, A.B., and McKenzie, M.G. 2005. The history of ocean resources: modeling cod biomass using historical records. Frontiers in Ecology and the Environment, 3(2): 84-90. Rothschild, B. J., Keiley, E. F., and Jiao, Y. 2014. Failure to eliminate over-fishing and attain optimum yield in the New England groundfish fishery. ICES Journal of Marine Science, 71: 226- 233.

Scholz, A., Bonzon, K., Fujita, R., Benjamin, N., Woodling, N., Black, P., and Steinback, C. 2004. Participatory socioeconomic analysis: drawing on fishermen’s knowledge for marine protected area planning in California. Marine Policy, 28: 335-349.

Schroeder, W.C. 1930 Migrations and other phases in the life history of the cod off southern New England. Bulletin of the Bureau of Fisheries, Vol. XLVI: 1-136. Scientific and Statistical Committee (SSC), 2012. Gulf of Maine cod. Report to the New England Fishery Management Council. Serchuk, F.M., and Wigley, S.E. 1992. Assessment and management of the Georges Bank cod fishery: An historical review and evaluation. Journal of Northwest Atlantic Fishery Science, 13: 25-52.

Siceloff, L., and Howell, W.H. 2013. Fine-scale temporal and spatial distributions of Atlantic cod (Gadus morhua) on a western Gulf of Maine spawning ground. Fisheries Research, 141: 31-43. Silvano, R.A.M., MacCord, P.F.L., Lima, R.V., and Begossi, A. 2006. When does this fish spawn? Fishermen’s local knowledge of migration and reproduction in Brazilian coastal fishes. Environmental Biology of Fish, 76: 371-386. Smedbol, R. K., and Stephenson, R. 2001. The importance of managing within-species diversity in cod and herring fisheries of the north-western Atlantic. Journal of Fish Biology, 59: 109-128.

Sibunka, J.D., Johnson, D.L., and Berrien, P.L. 2006. Distribution and abundance of fish eggs collected during the GLOBEC broad-scale Georges Bank surveys, 1995-1999. NOAA Technical Memorandum NEFSC-NE-199. Skjӕraasen, J.E., Meager, J.J., Karlsen, O., Hutchings, J.A., and Fern , A. 2011. Extreme spawning-site fidelity in Atlantic cod. ICES Journal of Marine Science, 68: 1472-1477. Smith, H.M. 1902. Notes on the tagging of four thousand adult cod at Woods Hole, Massachusetts. U.S. Commission of Fish and Fisheries, Part XXVII. Report of the Commissioner for the year ending June 30, 1901. pp. 193-208. Stephenson, R.L., Paul, S., Pastoors, M.A., Kraan, M., Holm, P., Wiber, M., Mackinson, S., Dankel, D.J., Brooks, K., and Benson, A. 2016. Integrating fishers’ knowledge research in science and management. ICES Journal of Marine Science, doi:10.193/icesjms/fsw025. Stone, H.H., and Gross, W.E. 2012. Review of the Georges Bank research vessel survey program, 1987-2011. Canadian Manuscript Report of Fisheries and Aquatic Sciences 2988. 108 pp. Sundby, S. 2000. Recruitment of Atlantic cod stocks in relation to temperature and advection of copepod populations. Sarsia, 85: 277-298. Sved ng, H., St l, J. Sterner, T., and Cardinale, M. 2010. Consequences of subpopulation structure on fisheries management: cod (Gadus morhua) in the Kattegat and Oresund (North Sea). Reviews in Fisheries Science, 18(2): 139-150. Tallack, S.M.L. 2009. Proceedings from a workshop to identify future research priorities for cod tagging in the Gulf of Maine. Northeast Fisheries Science Center Reference Document 09-09. 82 pp.

Tallack, S.M.L. 2011. Stock identification applications of conventional tagging data for Atlantic cod in the Gulf of Maine. In Proceedings from the 2nd International Symposium on Advances in Fish Tagging and Marking Techniques, pp. 1-15. Ed. by J. McKenzie, B. Parsons, A.C. Seitz, R.K. Kopf, M. Mesa, and Q. Phelps. American Fisheries Society, Auckland, NZ. Thompson, B.M., and Riley, J.D. 1981. Egg and larval development studies in the North Sea cod (Gadus morhua L.). Rapports et Proces-Verbaux des Reunions du Conseil International pour l’Exploration de lar Mer, 178: 553-559. TRAC, 2016. Eastern Georges Bank cod. TRAC Status Report 2016/01. 14 pp. Wang, Y., L. O’Brien, and I. Andrushchenko. 2015. Assessment of Eastern Georges Bank Atlantic Cod for 2015. TRAC Reference Document 2015/03.

Wang, Y. 2016. Changes in natural mortality of Atlantic cod (Gadus morhua) on eastern Georges Bank. Journal of Ocean University of China, 15(5), 879-889.

Weiss, E., Sanderson, M., and Gerlach, G. 2005. Population structure and essential fish habitat mapping of western Georges Bank cod. Final Report for Cooperative Research Partners' Program, Northeast Regional Office, National Marine Fisheries Service, 16 December 2005. Wise, J.P. 1958. The world’s southernmost indigenous cod. Journal Cont. Exploration de la Mer. 23:208-212.

Yates, K.L. (2014). View from the wheelhouse: perceptions on marine management from the fishing community and suggestions for improvement. Marine Policy, 48: 39-50. Zemeckis, D.R., Dean, M.J., and Cadrin, S.X. 2014a. Spawning dynamics and associated management implications for Atlantic cod (Gadus morhua). North American Journal of Fisheries Management, 34(2): 424-442. Zemeckis, D.R., Hoffman, W.S., Dean, M.J., Armstrong, M.P., and Cadrin, S.X. 2014b. Spawning site fidelity by Atlantic cod (Gadus morhua) in the Gulf of Maine: implications for population structure and rebuilding. ICES Journal of Marine Science, 71(6): 1356-1365. Zemeckis, D.R., Martine, D., Kerr, L.A., and Cadrin, S.X. 2014c. Stock identification of Atlantic cod (Gadus morhua) in US waters: an interdisciplinary approach. ICES Journal of Marine Science, 71(6): 1490-1506.

Appendix A – Maps of Spawning Cod from Scientific Data Sources Section 1: Data from the NOAA Northeast Fisheries Science Center Biannual Trawl Survey

Figure A1.1 – Distribution of spawning cod (defined as ripe, running, and spent) observed on the NEFSC spring trawl survey from 1970 through 1979.

Figure A1.2 – Distribution of ripe and running cod observed on the NEFSC spring trawl survey from 1970 through 1979.

Figure A1.3 – Distribution of spawning cod (defined as ripe, running, and spent) observed on the NEFSC fall trawl survey from 1970 through 1979.

Figure A1.4 – Distribution of spawning cod (defined as ripe, running, and spent) observed on the NEFSC spring trawl survey from 1980 through 1989.

Figure A1.5 – Distribution of ripe and running cod observed on the NEFSC spring trawl survey from 1980 through 1989.

Figure A1.6 – Distribution of spawning cod (defined as ripe, running, and spent) observed on the NEFSC fall trawl survey from 1980 through 1989.

Figure A1.7 – Distribution of spawning cod (defined as ripe, running, and spent) observed on the NEFSC spring trawl survey from 1990 through 1999.

Figure A1.8 – Distribution of ripe and running cod observed on the NEFSC spring trawl survey from 1990 through 1999.

Figure A1.9 – Distribution of spawning cod (defined as ripe, running, and spent) observed on the NEFSC fall trawl survey from 1990 through 1999.

Figure A1.10 – Distribution of spawning cod (defined as ripe, running, and spent) observed on the NEFSC spring trawl survey from 2000 through 2009.

Figure A1.11 – Distribution of ripe and running cod observed on the NEFSC spring trawl survey from 2000 through 2009.

Figure A1.12 – Distribution of spawning cod (defined as ripe, running, and spent) observed on the NEFSC fall trawl survey from 2000 through 2009.

Figure A1.13 – Distribution of spawning cod (defined as ripe, running, and spent) observed on the NEFSC spring trawl survey from 2010 through 2014.

Figure A1.14 – Distribution of ripe and running observed on the NEFSC spring trawl survey from 2010 through 2014.

Figure A1.15 – Distribution of spawning cod (defined as ripe, running, and spent) observed on the NEFSC fall trawl survey from 2010 through 2014.

Figure A1.16 – Proportion of tows with spawning cod observed annually in each survey strata during the NEFSC spring bottom trawl survey.

Section 2: Data from the Canadian Department of Fisheries and Oceans Trawl Survey

Figure A2.1 – Distribution of spawning cod (defined as ripe, running, and spent) observed on the Canadian DFO trawl survey from 1987 through 1989.

Figure A2.2 – Distribution of ripe and running cod observed on the Canadian DFO trawl survey from 1987 through 1989.

Figure A2.3 – Distribution of spawning cod (defined as ripe, running, and spent) observed on the Canadian DFO trawl survey from 1990 through 1999.

Figure A2.4 – Distribution of ripe and running cod observed on the Canadian DFO trawl survey from 1990 through 1999.

Figure A2.5 – Distribution of spawning cod (defined as ripe, running, and spent) observed on the Canadian DFO trawl survey from 2000 through 2009.

Figure A2.6 – Distribution of ripe and running cod observed on the Canadian DFO trawl survey from 2000 through 2009.

Figure A2.7 – Distribution of spawning cod (defined as ripe, running, and spent) observed on the Canadian DFO trawl survey from 2010 through 2014.

Figure A2.8 – Distribution of ripe and running cod observed on the Canadian DFO trawl survey from 2010 through 2014.

Figure A2.9 – Proportion of cod observed to be in spawning condition (defined as ripe, running, and spent) in tows with >20 cod on the Canadian DFO trawl survey from 1987 through 2015.

Figure A2.10 – Proportion of tows with spawning cod observed annually in survey strata 5Z1 - 5Z4 during the DFO bottom trawl survey. The average proportion of tows with spawning cod in each year is also shown.

Section 3: Data from the Canadian Fisheries Observer Program

Figure A3.1 – Distribution of ripe and running cod observed in January by Canadian fisheries observers from 1987 through 2015.

Figure A3.2 – Distribution of ripe and running cod observed in February by Canadian fisheries observers from 1987 through 2015.

Figure A3.3 – Distribution of ripe and running cod observed in June by Canadian fisheries observers from 1987 through 2015.

Figure A3.4 – Distribution of ripe and running cod observed in July by Canadian fisheries observers from 1987 through 2015.

Figure A3.5 – Distribution of ripe and running cod observed in August by Canadian fisheries observers from 1987 through 2015.

Figure A3.6 – Distribution of ripe and running cod observed in September by Canadian fisheries observers from 1987 through 2015.

Figure A3.7 – Distribution of tows with cod catches that were recorded by Canadian fisheries observers in October from 1987 through 2015. No ripe and running cod were observed in October.

Figure A3.8 – Distribution of ripe and running cod observed in November by Canadian fisheries observers from 1987 through 2015.

Figure A3.9 – Distribution of ripe and running cod observed in December by Canadian fisheries observers from 1987 through 2015.

Figure A3.10 – Proportion of cod sampled in January that were observed to be in spawning condition (ripe, running, or spent) by the Canadian fisheries observer program from 1987 through 2015.

Figure A3.11 – Proportion of cod sampled in February that were observed to be in spawning condition (ripe, running, or spent) by the Canadian fisheries observer program from 1987 through 2015.

Figure A3.12 – Proportion of cod sampled in May and June that were observed to be in spawning condition (ripe, running, or spent) by the Canadian fisheries observer program from 1987 through 2015.

Figure A3.13 – Proportion of cod sampled in July that were observed to be in spawning condition (ripe, running, or spent) by the Canadian fisheries observer program from 1987 through 2015.

Figure A3.14 – Proportion of cod sampled in August that were observed to be in spawning condition (ripe, running, or spent) by the Canadian fisheries observer program from 1987 through 2015.

Figure A3.15 – Proportion of cod sampled in September that were observed to be in spawning condition (ripe, running, or spent) by the Canadian fisheries observer program from 1987 through 2015.

Figure A3.16 – Proportion of cod sampled in October that were observed to be in spawning condition (ripe, running, or spent) by the Canadian fisheries observer program from 1987 through 2015.

Figure A3.17 – Proportion of cod sampled in November that were observed to be in spawning condition (ripe, running, or spent) by the Canadian fisheries observer program from 1987 through 2015.

Figure A3.18 – Proportion of cod sampled in December that were observed to be in spawning condition (ripe, running, or spent) by the Canadian fisheries observer program from 1987 through 2015.

Section 4: Data from the GLOBEC Sampling Program

Figure A4.1 – Distribution and concentration (#/10m2) of stage one cod eggs that were sampled on Georges Bank in January by the GLOBEC program from 1996 through 1999.

Figure A4.2 – Distribution and concentration (#/10m2) of stage one cod eggs that were sampled on Georges Bank in February by the GLOBEC program from 1995 through 1999.

Figure A4.3 – Distribution and concentration (#/10m2) of stage one cod eggs that were sampled on Georges Bank in March by the GLOBEC program from 1995 through 1999.

Figure A4.4 – Distribution and concentration (#/10m2) of stage one cod eggs that were sampled on Georges Bank in April by the GLOBEC program from 1995 through 1999.

Figure A4.5 – Distribution and concentration (#/10m2) of stage one cod eggs that were sampled on Georges Bank in May by the GLOBEC program from 1995 through 1999.

Section 5: Data from the MARMAP Sampling Program

Figure A5.1 - Composite maps showing the monthly distribution and concentration of cod eggs (all stages combined) observed from March through August in the northwest Atlantic during the MARMAP program from 1977 through 1987. Figure taken from Berrien and Sibunka (1999).

Figure A5.2 - Composite maps showing the monthly distribution and concentration of cod eggs (all stages combined) observed from September through February in the northwest Atlantic during the MARMAP program from 1977 through 1987. Figure taken from Berrien and Sibunka (1999).

Appendix B. – Phase 2 Figures - Fishermen’s Ecological Knowledge

Figure B1 - Pictures of female cod at different maturity stages that were shown to fishermen during the interviews to help them identify which maturity stages they observed at the spawning ground. A) immature female, B) resting female, c) developing female, D) ripe female, E) ripe and running female, F) spent female. The figures were provided by Rich McBride at the NOAA Northeast Fisheries Science Center.

Figure B2 - Pictures of male cod at different maturity stages that were shown to fishermen during the interviews to help them identify which maturity stages they observed at the spawning ground. A) immature male, B) resting male, c) developing male, D) ripe male, E) spent male, and F) ripe and running male. The figures were provided by Rich McBride at the NOAA Northeast Fisheries Science Center.

Figure B3 - The 210 cod spawning grounds that were identified by fishermen in this study. Each spawning ground was given a unique code during the interviews (e.g., 15C), which are shown on the map.

Figure B4 - 210 cod spawning grounds that were identified by fishermen in this study. Each polygon represents a spawning ground that was identified by a single fisherman. The shading is used to identify areas where cod spawning activity was identified independently by multiple fishermen.

100

Figure B5 - Spawning grounds that were identified by fishermen in October.

101

Figure B6 - Spawning grounds that were identified by fishermen in November.

102

Figure B7 - Spawning grounds that were identified in November. Each polygon represents a spawning ground that was identified by a single fisherman. The shading is used to identify areas where cod spawning activity was identified independently by multiple fishermen.

103

Figure B8 - Spawning grounds that were identified on Nantucket Shoals in November. Each polygon represents a spawning ground that was identified by a single fisherman. The shading is used to identify areas where cod spawning activity was identified independently by multiple fishermen.

104

Figure B9 - Spawning grounds that were identified by fishermen in December.

105

Figure B10 - Spawning grounds that were identified in December. Each polygon represents a spawning ground that was identified by a single fisherman. The shading is used to identify areas where cod spawning activity was identified independently by multiple fishermen.

106

Figure B11 - Spawning grounds that were identified on Nantucket Shoals in December. Each polygon represents a spawning ground that was identified by a single fisherman. The shading is used to identify areas where cod spawning activity was identified independently by multiple fishermen.

107

Figure B12 - Cod spawning grounds that were identified by fishermen in January.

108

Figure B13 - Spawning grounds that were identified in January. Each polygon represents a spawning ground that was identified by a single fisherman. The shading is used to identify areas where cod spawning activity was identified independently by multiple fishermen.

109

Figure B14 - Spawning grounds that were identified on Nantucket Shoals in January. Each polygon represents a spawning ground that was identified by a single fisherman. The shading is used to identify areas where cod spawning activity was identified independently by multiple fishermen.

110

Figure B15 - Cod spawning grounds that were identified by fishermen in February.

111

Figure B16 - Spawning grounds that were identified in February. Each polygon represents a spawning ground that was identified by a single fisherman. The shading is used to identify areas where cod spawning activity was identified independently by multiple fishermen.

112

Figure B17 - Spawning grounds that were identified on Nantucket Shoals in February. Each polygon represents a spawning ground that was identified by a single fisherman. The shading is used to identify areas where cod spawning activity was identified independently by multiple fishermen.

113

Figure B18 - Cod spawning grounds that were identified by fishermen in March.

114

Figure B19 - Spawning grounds that were identified in March. Each polygon represents a spawning ground that was identified by a single fisherman. The shading is used to identify areas where cod spawning activity was identified independently by multiple fishermen.

115

Figure B20 - Spawning grounds that were identified on Nantucket Shoals in March. Each polygon represents a spawning ground that was identified by a single fisherman. The shading is used to identify areas where cod spawning activity was identified independently by multiple fishermen.

116

Figure B21 - Cod spawning grounds that were identified by fishermen in April.

117

Figure B22 - Spawning grounds that were identified in April. Each polygon represents a spawning ground that was identified by a single fisherman. The shading is used to identify areas where cod spawning activity was identified independently by multiple fishermen.

118

Figure B23 - Cod spawning grounds that were identified by fishermen in May.

119

Figure B24 - Spawning grounds that were identified in May. Each polygon represents a spawning ground that was identified by a single fisherman. The shading is used to identify areas where cod spawning activity was identified independently by multiple fishermen.

120

Figure B25 - Cod spawning grounds that were identified by fishermen in June.

121

Figure B26 - Map of the 47 cod feeding grounds that were identified by fishermen in this study.

Figure B27 – Number of cod feeding grounds identified each month by fishermen.

122

Figure B28 - Thirteen haddock spawning grounds that were identified by eight fishermen during the interviews.

Figure B29 - Winter flounder spawning grounds that were identified by fishermen during the interviews.

123

Appendix C. – Phase 3 Figures: Synthesizing Fishermen’s Ecological Knowledge and Scientific Data

Figure C1: Location of tows completed during the NEFSC spring trawl survey from 1970 through 2014. For tows completed from 1970 through 1995, only the starting location was provided in the data we obtained. Both the start and end coordinates were provided for tows completed after 1995.

124

Figure C2: Location of tows completed during the NEFSC fall trawl survey from 1970 through 2014. For tows completed from 1970 through 1995, only the starting location was provided in the data we obtained. Both the start and end coordinates were provided for tows completed after 1995.

125

Figure C3: Location of tows completed during the DFO trawl survey from 1987 through 2015. For tows completed from 1987 through 1989, only the starting location was provided in the data we obtained. Both the start and end coordinates were provided for tows completed after 1989.

126

Appendix D: UMass Dartmouth Interview Consent Form

IRB PROTOCOL # Title: Mapping the Distribution of Spawning Cod on Georges Bank Using Fishermen’s Ecological Knowledge and Scientific Data Conducted By: Steve Cadrin (508-910-6358; [email protected]), Greg DeCelles (508-910-6393; [email protected]), Doug Zemeckis (508-910-6336; [email protected]), and Dave Martins (508-990-2860; [email protected]) of The University of Massachusetts Dartmouth, Department of Fisheries Oceanography, School for Marine Science and Technology. You are being asked to participate in a research study. This form provides you with information about the study. The person in charge of this research will also describe this study to you and answer all of your questions. Please read the information below and ask any questions you might have before deciding whether or not to take part. Your participation is entirely voluntary. You can refuse to participate without penalty or loss of benefits to which you are otherwise entitled. You can stop your participation at any time and your refusal will not impact current or future relationships with UMD or participating sites. To do so simply tell the researcher you wish to stop participation. The researcher will provide you with a copy of this consent for your records. The purpose of this study is to improve our understanding of the spatial and temporal distribution of cod spawning on Georges Bank, including identification of Essential Fish Habitat, by combining fishermen’s ecological knowledge with other scientific data to map cod spawning grounds. The ecological knowledge of fishermen will be acquired through a series of interviews with 50 to 100 active and retired fishermen with experience fishing for cod on Georges Bank. This project would build upon previous work and fill data gaps with respect to cod spawning on Georges Bank, which would be helpful for improving the management of the resource.

If you agree to be in this study, we will ask you to do the following things:

 Answer questions related to your fishing career.  Discuss your observations of cod spawning grounds on Georges Bank, and identify these spawning grounds using nautical charts.  Compare your observations of cod spawning grounds to those described in the scientific literature.

Total estimated time to participate in study is 1 to 3 hours. Risks of being in the study

 The risk associated with this study is no greater than everyday life.

Benefits of being in the study

 We hope to use the information gathered during this project to help inform and improve the management and assessment of the cod resource on Georges Bank.

127

Compensation:

 Participating fishermen will be compensated with a $50 pre-paid Visa gift card.

Confidentiality and Privacy Protections:

 The data provided will be kept confidential. To protect you from breach of confidentiality, the identity of each participating fisherman will be kept private by using an “identification code”, which will only be known by the Principal Investigators, and stored in a separate password protected document. This code will be used to map all of the data provided by each fisherman, and will be used on the audio recordings as well.  The collected data will be made available to the Principal Investigators; Steve Cadrin (Professor and Research Advisor), Greg DeCelles (Research Associate), Doug Zemeckis (Research Associate) and David Martins (Master’s Candidate).  All audio recordings and transcripts will be free of your name.  The data resulting from your participation may be made available to other researchers in the future for research purposes not detailed within this consent form. In these cases, the data will contain no identifying information that could associate you with it, or with your participation in any study.  The confidentiality conditions will not be broken under any conditions.

The records of this study will be stored securely and kept confidential. Authorized persons from The University of Massachusetts Dartmouth, members of the Institutional Review Board, and the Northeast Consortium have the legal right to review your research records and will protect the confidentiality of those records to the extent permitted by law. All publications will exclude any information that will make it possible to identify you as a subject. Throughout the study, the researchers will notify you of new information that may become available and that might affect your decision to remain in the study. Audio Recordings:  With your consent, an audio recording of the interview will be created. The purpose of the audio recording is twofold. Firstly, it will provide a record of what was discussed during the interviews, which will be beneficial for analyzing and interpreting the data. Secondly, the recorded interviews will ensure that the ecological knowledge of the fishermen is preserved for future generations. With your consent, at the conclusion of the project, copies of the interviews will be provided to the New Bedford Whaling Museum Research Library and Archives so that the information will be publicly available for future researchers.  To protect your anonymity and confidentiality, the audio tapes will be free of your name and labeled with your secure identification code.  The tapes will be kept in a secure place throughout the study, (e.g., in the drawer of a locked desk in the office of the Principal Investigator).  The tapes will only be listened to for research purposes.

Contacts and Questions: If you have any questions about the study please ask now. If you have questions later, want additional information, or wish to withdraw your participation call the researchers conducting the study. Their names, phone numbers, and e-mail addresses are at the top of this page. If you have questions

128 about your rights as a research participant, complaints, concerns, or questions about the research please contact Andrew Karberg, The University of Massachusetts Dartmouth Office of Institutional Compliance at (508) 910-9880 or email: [email protected]. You will be given a copy of this information to keep for your records.

Statement of Consent:

I have read the above information and have sufficient information to make a decision about participating in this study. I consent to participate in the study.

Signature:______Date: ______

______Date: ______Signature of Person Obtaining Consent

Signature of Investigator:______Date: ______

We may wish to present some of the tapes from this study at scientific meetings. Please sign below if you are willing to allow us to do so with your tape.

I hereby give permission for the audio tape made for this research study to also be used for educational purposes, and for a copy of this tape to be archived at the New Bedford Whaling Museum Research Library.

Signature:______Date: ______

129

Manuscripts submitted to ICES Journal of Marine Science

Using F ishermen’sFor Review Ecological Knowledge Only to Map Atlantic Cod Spawning Grounds on Georges Bank

Journal: ICES Journal of Marine Science

Manuscript ID Draft

Manuscript Types: Original Article

Date Submitted by the Author: n/a

Complete List of Authors: DeCelles, Gregory; University of Massachusetts Dartmouth, School for Marine Science and Technology; Massachusetts Division of Marine Fisheries, Resource Assessment Martins, David; University of Massachusetts Dartmouth, School for Marine Science and Technology; Massachusetts Division of Marine Fisheries, Recreational Fisheries Zemeckis, Douglas; University of Massachusetts Dartmouth, School for Marine Science and Technology Cadrin, Steven; University of Massachusetts Dartmouth, School for Marine Science and Technology

Fishermen’s Ecological Knowledge, Atlantic Cod, Spawning Grounds, Stock Keyword: Structure, Gadus morhua, Georges Bank

http://mc.manuscriptcentral.com/icesjms Page 1 of 54 Manuscripts submitted to ICES Journal of Marine Science

1 2 3 1 Using Fishermen’s Ecological Knowledge to Map Atlantic Cod Spawning Grounds on Georges Bank 4 5 1,2 1,2 1,† 1 6 2 Greg DeCelles* , David Martins , Douglas R. Zemeckis , and Steven X. Cadrin 7 8 3 *Corresponding author: tel: +1 508 990 2860; fax: + 1 508 990 0449; e-mail: 9 4 [email protected] 10 11 5 1: School for Marine Science and Technology, University of Massachusetts Dartmouth,706 South Rodney 12 13 6 French Blvd., New Bedford, MA 02740, USA 14 15 7 2: Massachusetts Division of Marine Fisheries, 1213 Purchase Street, New Bedford, MA 02740, USA 16 17 8 † Present Address: Department of Marine and Coastal Sciences, Rutgers University, 71 Dudley Road, New 18 9 Brunswick, NJ 08901, USAFor Review Only 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 1

http://mc.manuscriptcentral.com/icesjms Manuscripts submitted to ICES Journal of Marine Science Page 2 of 54

1 2 3 10 Abstract 4 5 6 11 The spawning dynamics of Atlantic cod ( Gadus morhua ) on Georges Bank and Nantucket Shoals 7 8 12 are not well understood. To address this uncertainty, we combined Fishermen’s Ecological 9 10 11 13 Knowledge with traditional scientific data to develop a more holistic understanding of cod 12 13 14 spawning on Georges Bank. Data from historical reports, trawl surveys, fisheries observers, and 14 15 16 15 ichthyoplankton surveys were used to describe the spatial and temporal distribution of cod 17 18 16 spawning activity. WeFor also collected Review Fishermen’s Ecological Only Knowledge regarding cod spawning 19 20 21 17 dynamics through semi-structured interviews (n=40). The fishermen had detailed knowledge of 22 23 18 the spatial and temporal distribution of cod spawning, and identified persistent fine-scale (i.e., 24 25 2 26 19 <50km ) spawning grounds that were often associated with specific habitat features, including 27 28 20 spawning grounds that were previously unreported in the scientific literature. The spawning 29 30 31 21 seasons and locations identified by fishermen generally agreed with information from 32 33 22 traditional scientific data, but it was evident that seasonal scientific surveys lack the spatial and 34 35 36 23 temporal resolution needed to fully characterize the distribution of cod spawning activity. Our 37 38 24 results will help inform management measures designed to promote the rebuilding of Georges 39 40 41 25 Bank cod, and also provide a basis for further investigations of cod spawning dynamics and 42 43 26 stock structure. 44 45 46 47 27 Keywords : Fishermen’s Ecological Knowledge, Atlantic Cod, Spawning Grounds, Stock Structure, Gadus 48 49 28 morhua , Georges Bank 50 51 52 53 54 55 56 57 58 59 60 2

http://mc.manuscriptcentral.com/icesjms Page 3 of 54 Manuscripts submitted to ICES Journal of Marine Science

1 2 3 29 1. Introduction 4 5 6 30 Fishermen and scientists observe the marine environment from different perspectives. 7 8 31 Scientists use standardized data collection methods, such as trawl surveys, to collect random 9 10 11 32 samples representative of the population of inference. Although surveys may occur over the 12 13 33 course of days or weeks, observations collected at a single station only provide snapshots of 14 15 16 34 these complex ecosystems (Murray et al., 2008a; Macdonald et al., 2014). Surveys are typically 17 18 35 conducted across largeFor spatial scales Review and the data are aggregated Only over relatively coarse 19 20 21 36 temporal scales (e.g., seasonal or annual) to construct time series of information on the 22 23 37 abundance, distribution, and demographics of fishery resources. Modern stock assessment 24 25 26 38 methods have been designed to use these standardized time series of abundance data to 27 28 39 estimate population size, evaluate stock status, and provide a basis for management advice. 29 30 31 32 40 On the other hand, the objective of fishermen is to catch fish, and their observations are not 33 34 41 random. A fisherman’s decision about when and where to fish is dictated by a variety of 35 36 37 42 factors, including their expectation of catch per unit effort, weather, fishery regulations, fuel 38 39 43 prices, and market demands. As a result, fishermen’s observations are not standardized, 40 41 42 44 making it difficult to integrate this information with traditional scientific knowledge (Neis et al., 43 44 45 1999a; Anuchiracheeva et al., 2003). In turn, fishermen’s observations are frequently dismissed 45 46 47 46 as “anecdotal”, and their perspectives are often overlooked by scientists and policy makers 48 49 47 (Pederson and Hall-Arber, 1999; Hind, 2015). However, there is a growing recognition that 50 51 52 48 involving fishermen more directly in stock assessment and management can improve fishery 53 54 49 science and increase the credibility of management decisions (Bergmann et al., 2004; Yates, 55 56 57 50 2014; Stephenson et al., 2016). 58 59 60 3

http://mc.manuscriptcentral.com/icesjms Manuscripts submitted to ICES Journal of Marine Science Page 4 of 54

1 2 3 51 Fishermen typically keep detailed records of their fishing activity and catch, and this 4 5 6 52 information is often shared in real time and passed down over multiple generations (Neis et al., 7 8 53 1999a; Bergmann et al., 2004). The transmission of knowledge among fishermen allows them 9 10 11 54 to “scale up” their observations across time and space, enabling fishermen to acquire unique 12 13 55 perspectives encompassing a range of spatial and temporal scales (Murray et al., 2008a). For 14 15 16 56 example, fishermen commonly have a thorough understanding of how fish move seasonally 17 18 57 For Review Only 19 over a wide geographic range to feed or spawn, as well as how small-scale bathymetric features 20 21 58 (e.g., specific boulder piles, convex sand humps) may each influence fish behaviour or 22 23 24 59 abundance. While fishermen are keenly aware of these fine-scale habitats, our scientific data 25 26 60 collection methods often lack the spatial resolution to examine abundance and distribution at 27 28 29 61 such a small scale. Following years of observation, fishermen are also cognizant of long-term 30 31 62 trends in fish abundance and changes in size structure (Pederson and Hall Arber, 1999; 32 33 34 63 Macdonald et al., 2014). In addition, fishermen also understand how the distribution and 35 36 64 abundance of target species can change across tidal, diel, lunar, and seasonal scales (Neis et al., 37 38 39 65 1999a; Johannes et al., 2000). 40 41 42 66 There is increasing recognition that Fishermen’s Ecological Knowledge (FEK) is part of the best 43 44 67 available science, and as such, it should play a larger role in fishery management (Johannes et 45 46 47 68 al., 2000; Stephenson et al., 2016). FEK has proven to be a valuable and cost-effective 48 49 69 supplement to existing institutional data, and incorporating FEK into the scientific process has 50 51 52 70 led to more robust management decisions (e.g., Bergmann et al., 2004; Scholz et al., 2004). For 53 54 55 71 example, FEK has been used to offer insights into long-term trends in the abundance and 56 57 72 distribution of fish populations such as megrim (Lepidorhombus whiffiagonis ) in the North Sea 58 59 60 4

http://mc.manuscriptcentral.com/icesjms Page 5 of 54 Manuscripts submitted to ICES Journal of Marine Science

1 2 3 73 (Macdonald et al., 2014) and cod (Gadus morhua ) in Greenland (Hedeholm et al., 2016). 4 5 6 74 Bergmann et al. (2004) used the knowledge of fishermen to identify Essential Fish Habitat for 7 8 75 cod, haddock (Melanogrammus aeglefinus ), and whiting (Merlangius merlangus ) in the Irish 9 10 11 76 Sea. FEK has also offered insight into the stock structure of cod in the Gulf of St. Lawrence 12 13 77 (Murray et al., 2008b) and Newfoundland (Neis, 1998). Fishermen often know where and when 14 15 16 78 spawning activity occurs, and FEK has been used to identify spawning grounds for many 17 18 79 For Review Only 19 species, including cod in the Gulf of Maine (Ames, 1997) and Newfoundland (Neis, 1999b), 20 21 80 bonefish ( Albula glossodonta ) in Tarawa, Kiribati (Johannes et al., 2000), and coastal fishes in 22 23 24 81 Brazil (Silvano et al., 2006). FEK has also led to the discovery of cod spawning grounds 25 26 82 previously unknown to scientists in Newfoundland (Neis, 1998) and along the coast of Norway 27 28 29 83 (Maurstad, 2002). Furthermore, FEK has been used to identify cod spawning grounds that have 30 31 84 been extirpated in the Gulf of Maine (Pederson and Hall-Arber, 1999; Ames, 2004), and in 32 33 34 85 Newfoundland (Neis, 1998), which can be critical for understanding long-term changes in stock 35 36 86 productivity and recruitment. 37 38 39 87 In this study, we sought to gather FEK related to the spatial and temporal distribution of 40 41 42 88 Atlantic cod ( Gadus morhua ) spawning activity on Georges Bank and Nantucket Shoals. For 43 44 89 many centuries, cod have supported major commercial and recreational fisheries throughout 45 46 47 90 the North Atlantic (e.g., Rose, 2007), including off the coast of New England (Serchuk and 48 49 91 Wigley, 1992; Rosenberg et al., 2005). However, in recent decades, cod stocks off New England 50 51 52 92 have decreased in abundance, and recent catches are a small fraction of historical landings and 53 54 55 93 estimates of Maximum Sustainable Yield (NEFSC, 2013; NEFSC, 2015; Wang et al., 2015). From 56 57 94 1972 onwards, cod have been managed in U.S. waters as two units: the Georges Bank and Gulf 58 59 60 5

http://mc.manuscriptcentral.com/icesjms Manuscripts submitted to ICES Journal of Marine Science Page 6 of 54

1 2 3 95 of Maine stocks (Serchuk and Wigley, 1992) (Figure 1). Cod on the eastern portion of Georges 4 5 6 96 Bank are assessed and managed jointly between the United States and Canada by the 7 8 97 Transboundary Resources Assessment Committee (Wang et al., 2015). Despite continuously 9 10 11 98 evolving fishery management strategies (NEFMC, 2009; NEFSC, 2013), persistent difficulties 12 13 99 remain with respect to ending overfishing (Rothschild et al., 2014) and managing the rebuilding 14 15 16 100 of cod (NEFSC, 2013; NEFSC, 2015; Wang et al., 2015). Some of the factors contributing to the 17 18 101 For Review Only 19 lack of rebuilding include poor recruitment, low weights at age, age truncation, and high natural 20 21 102 mortality rates (Wang et al., 2015). Therefore, additional research is required to improve our 22 23 24 103 understanding of cod population and spawning dynamics on Georges Bank to inform future 25 26 104 fishery management decisions intended to support rebuilding. 27 28 29 105 Understanding where and when cod spawn can aid in stock identification, because the 30 31 32 106 variability in spawning location and timing often function as mechanisms that contribute to the 33 34 107 development of metapopulations by limiting the reproductive connectivity among 35 36 37 108 subpopulations (Grabowski et al., 2011; Zemeckis et al., 2014a). Cod spawning activity also 38 39 109 presents important implications for fishery management (Zemeckis et al., 2014b). Cod spawn 40 41 42 110 over an extended period of time at locations that are often predictable and close to shore 43 44 111 (Siceloff and Howell, 2013; Dean et al., 2014), and fishing activity can disrupt cod spawning 45 46 47 112 behaviour (Morgan et al., 1997; Dean et al., 2012). In addition, given that cod spawning 48 49 113 components are semi-discrete and there is often limited connectivity among them (Smedbol 50 51 52 114 and Stephenson, 2001), there is a great deal of risk with respect to the extirpation of these 53 54 55 115 population segments. In fact, declines in spawning diversity have been well documented in 56 57 116 many cod stocks, including in the Gulf of Maine (Ames, 2004) and the North Sea (Svedäng et al ., 58 59 60 6

http://mc.manuscriptcentral.com/icesjms Page 7 of 54 Manuscripts submitted to ICES Journal of Marine Science

1 2 3 117 2010). The failure of past management plans to achieve rebuilding targets in the Gulf of Maine 4 5 6 118 has been considered to be in part due to their lack of consideration of the complex 7 8 119 metapopulation structure (Ames, 2004; Kovach et al ., 2010; Armstrong et al ., 2013). However, 9 10 11 120 relatively little is currently known about the spawning dynamics of cod on Georges Bank, and it 12 13 121 is uncertain whether declines in spawning diversity have influenced stock productivity. 14 15 16 17 122 By soliciting FEK and analyzing historical reports, Ames (1997, 2004) created detailed maps of 18 For Review Only 19 123 the spawning grounds and migratory patterns of Atlantic cod in the Gulf of Maine. His results 20 21 22 124 provided valuable insight into cod stock structure and population dynamics, and helped inform 23 24 125 future research investigating cod spawning dynamics in the Gulf of Maine (Armstrong et al., 25 26 27 126 2013; Gurshin et al., 2013; Siceloff and Howell, 2013; Dean et al., 2014; Zemeckis et al., 2014a). 28 29 127 However, similar fine-scale information on cod spawning grounds is not available for Georges 30 31 32 128 Bank, including the Great South Channel and Nantucket Shoals regions. Given the depleted 33 34 129 status of Georges Bank cod and the low stock productivity (NEFSC, 2013), additional directed 35 36 37 130 research is needed to improve our understanding of cod spawning activity and population 38 39 131 dynamics on Georges Bank, which can help inform future management decisions intended to 40 41 42 132 promote rebuilding. In this study we sought to combine FEK with traditional scientific data to 43 44 133 improve our understanding of cod spawning on Georges Bank, and compare historical and 45 46 47 134 contemporary data to investigate potential shifts in spawning activity. 48 49 50 135 2. Methods 51 52 53 136 2.1 Analysis of Scientific Data 54 55 56 57 58 59 60 7

http://mc.manuscriptcentral.com/icesjms Manuscripts submitted to ICES Journal of Marine Science Page 8 of 54

1 2 3 137 We analyzed several scientific datasets and published reports to gather information related to 4 5 6 138 cod spawning activity on Georges Bank to provide context and comparison to FEK. Historical 7 8 139 reports of cod fishing and spawning grounds on Georges Bank were reviewed (Goode, 1884; 9 10 11 140 Rich, 1929; Bigelow and Schroeder, 1953), along with information from historical tagging 12 13 141 studies (Smith, 1902; Schroeder, 1930; Wise 1963). 14 15 16 17 142 The abundance and distribution of cod were investigated using the NOAA Northeast Fisheries 18 For Review Only 19 143 Science Center’s (NEFSC) spring and fall bottom trawl survey data. Our analysis was limited to 20 21 22 144 tows completed from 1970 through 2014, and in NEFSC offshore strata 9 through 25 to match 23 24 145 the strata included in the Georges Bank cod stock assessment (NEFSC, 2013). Burnett et al. 25 26 27 146 (1989) provide details on maturity staging guidelines and report that the NEFSC began 28 29 147 collecting spawning stage information for cod in 1970. The timing of NEFSC surveys varied from 30 31 32 148 year to year, but all spring trawl survey tows were conducted in March, April, May, and June, 33 34 149 with most of the tows completed in April (70.9%) and March (22.9%). The fall trawl survey tows 35 36 37 150 occurred in September, October, November, and December, and most tows were completed in 38 39 151 October (80.1%) and November (10.1%). Similarly, the abundance and distribution of cod 40 41 42 152 observed from 1987 through 2015 during the Canadian Department of Fisheries and Oceans 43 44 153 (DFO) annual trawl survey during February and March were examined. Morrison (1990) 45 46 47 154 described the maturity staging guidelines that are used on the DFO trawl survey. For both 48 49 155 surveys, the abundance of spawning cod (defined here as maturity stages; ripe, ripe and 50 51 52 156 running, or spent) observed at each survey station was plotted in ten-year periods to identify 53 54 55 157 areas where spawning cod were sampled, and to investigate potential changes in the 56 57 158 distribution of spawning activity over time. For both the NEFSC spring trawl survey and the DFO 58 59 60 8

http://mc.manuscriptcentral.com/icesjms Page 9 of 54 Manuscripts submitted to ICES Journal of Marine Science

1 2 3 159 trawl survey, we performed a G* hotspot analysis (Getis and Ord, 1992) to identify survey tow 4 5 6 160 locations where the number of spawning cod were significantly greater than average (p < 0.05) 7 8 161 and were surrounded by other tows with above average catches of spawning cod, following the 9 10 11 162 methods outlined by the Closed Area Technical Team (NEFMC, 2013). For both surveys, the 12 13 163 number of hotspots within a 100km 2 grid was summarized to identify spatial clusters of 14 15 16 164 spawning activity. 17 18 For Review Only 19 165 Ichthyoplankton sampling datasets from the U.S. and Canada were investigated to describe the 20 21 22 166 temporal and spatial distribution of cod eggs on Georges Bank. We obtained data on the 23 24 167 distribution and abundance of cod eggs observed during the United States Global Ocean 25 26 27 168 Ecosystems Dynamics (GLOBEC) program, which sampled on Georges Bank from February 28 29 169 through July in 1995, and from January through June from 1996 through 1999 (Sibunka et al., 30 31 32 170 2006). Cod eggs were classified to three stages, and stage-one eggs represented the period 33 34 171 from “spawned to just before blastopore closure” (Sibunka et al., 2006). The monthly 35 36 37 172 distribution and abundance of stage-one cod eggs that were observed from 1995 through 1999 38 39 173 were plotted. We chose to examine the distribution of stage-one eggs, because these eggs had 40 41 42 174 been adrift for the shortest period of time and would presumably be closer to the spawning 43 44 175 grounds than later stage eggs. We also reviewed published reports from the Marine Resource 45 46 47 176 Monitoring Assessment and Prediction (MARMAP) program, which collected monthly bongo 48 49 177 net samples of fish eggs and larvae in the Northwest Atlantic from 1977 through 1987 (Page et 50 51 52 178 al., 1998; Berrien and Sibunka, 1999), and Hanke et al. (2000), which summarized the results of 53 54 55 179 three ichthyoplankton sampling programs completed by the Canadian Department of Fisheries 56 57 180 and Oceans on Georges Bank and the Scotian Shelf from 1975 through 1997. 58 59 60 9

http://mc.manuscriptcentral.com/icesjms Manuscripts submitted to ICES Journal of Marine Science Page 10 of 54

1 2 3 181 Data collected by the Canadian Fisheries Observer Program between 1983 and 2015 were used 4 5 6 182 to investigate the timing and location of cod spawning on eastern Georges Bank. Canadian 7 8 183 fishery observers sampled the maturity stages of cod using the same classification scheme 9 10 11 184 employed on the DFO trawl survey. Cod were often subsampled for maturity when the catches 12 13 185 were large. The spatial and temporal distribution of observations reflect patterns of fishing 14 15 16 186 effort on eastern Georges Bank by the trawl fleet, which is constrained by factors such as the 17 18 187 For Review Only 19 Canadian jurisdiction and seasonal fishery regulations, which typically prohibit trawling on 20 21 188 eastern Georges Bank from February through May to protect spawning cod (Wang et al., 2015). 22 23 24 189 2.2 Interviews With Fishermen 25 26 27 190 We used a semi-structured interview approach to gather FEK regarding their understanding of 28 29 191 cod spawning dynamics on Georges Bank, with a focus on the spatial and temporal distribution 30 31 32 192 of spawning activity. The semi-structured interview format allowed us to collect information 33 34 193 that could be compared and standardized across interviews. In this flexible format, the 35 36 37 194 interview was guided by a common set of questions, but the fisherman had latitude to discuss 38 39 195 their observations and perspectives. Two or more scientists were present for almost every 40 41 42 196 interview, with one scientist leading the discussion, and the other scientist(s) recording answers 43 44 197 and marking spatial information by hand on paper nautical charts of Georges Bank. An audio 45 46 47 198 recording was also captured for each interview. Current and retired captains (n = 40) with 48 49 199 experience fishing for cod on Georges Bank were interviewed. We started by interviewing 50 51 52 200 experienced captains known from previous collaborative research projects, and used a snowball 53 54 55 201 sampling process (Babbie, 1989) to generate additional interviews by asking captains to refer 56 57 202 other fishermen who also targeted cod on Georges Bank. One interviewer is fluent in 58 59 60 10

http://mc.manuscriptcentral.com/icesjms Page 11 of 54 Manuscripts submitted to ICES Journal of Marine Science

1 2 3 203 Portuguese, which enabled us to interview Portuguese speaking fishermen in New Bedford, 4 5 6 204 MA, USA, where many of the captains in the trawl fishery are of Portuguese descent. 7 8 9 205 The interviews were done at times and locations that were convenient for fishermen, and were 10 11 12 206 typically conducted aboard their vessels while in port, in cafes, their homes, or in our office. 13 14 207 Participation was voluntary, and fishermen were not required to share any information they did 15 16 17 208 not want to. Participating fishermen were compensated for their time with a $50USD gift card, 18 For Review Only 19 209 although most fishermen were not aware of the compensation until after the interview had 20 21 22 210 been completed. Each fisherman was required to read and sign an interview consent form, 23 24 211 which had the approval of the University of Massachusetts Dartmouth Office of Institutional 25 26 27 212 Compliance and Ethics. The duration of interviews ranged from forty five minutes to three 28 29 213 hours, depending upon each fisherman’s knowledge and their willingness to provide 30 31 32 214 information. We interviewed fishermen from throughout New England and Nova Scotia, in 33 34 215 order to gain the perspectives of a diverse group of fishermen with experience fishing across all 35 36 37 216 regions of Georges Bank. A conscious effort was made to interview fishermen without the 38 39 217 presence of other fishermen, so that participants could answer survey questions freely and to 40 41 42 218 the best of their knowledge without influence from others. However, in one instance, three 43 44 219 fishermen were interviewed together on Cape Cod due to unavoidable logistical reasons. 45 46 47 48 220 Each interview began with a series of demographic questions designed to gauge the 49 50 221 fishermen’s experience fishing for groundfish, including specifically targeting cod on Georges 51 52 53 222 Bank (Table 1). Next, fishermen were asked to recall times and locations where they had 54 55 223 observed spawning cod. Fishermen were asked to identify and delineate these locations on 56 57 58 59 60 11

http://mc.manuscriptcentral.com/icesjms Manuscripts submitted to ICES Journal of Marine Science Page 12 of 54

1 2 3 224 NOAA nautical chart 13200 (http://www.charts.noaa.gov/OnLineViewer/13200.shtml) which 4 5 6 225 encompasses the area from southern New England to eastern Georges Bank. Many fishermen 7 8 226 also provided information from their own charts, logbooks, and plotters, and used this 9 10 11 227 information to precisely identify the locations and time of year where they captured spawning 12 13 228 cod. For each spawning ground that fishermen identified, we asked them a series of questions 14 15 16 229 to classify the timing of spawning at that location and to collect information on the biological 17 18 230 For Review Only 19 characteristics and habitat associated with that spawning ground (Table 1). Fishermen were 20 21 231 asked to specify the criteria they used to determine that area was a cod spawning ground. 22 23 24 232 Fishermen were also shown pictures of male and female cod at different maturity stages, and 25 26 233 were asked to identify which maturity stages they observed at that location. Having the photos 27 28 29 234 of the gonads of spawning cod helped to distinguish cod feeding grounds from spawning 30 31 235 grounds. The cod maturity stage guide was obtained from the NOAA NEFSC (Richard McBride, 32 33 34 236 pers. comm.) and is the same guide used during trawl surveys (Burnett et al., 1989). Finally, as 35 36 237 time allowed, we asked the fishermen questions related to the biology and stock structure of 37 38 39 238 cod on Georges Bank (Table 1). 40 41 42 239 After each interview, the spawning grounds identified by the fishermen were digitally mapped 43 44 240 in ArcGIS. By relying on the nautical charts and the GIS mapping, we were able to standardize 45 46 47 241 the spatial information provided by the fishermen. A unique shapefile was created for each 48 49 242 spawning ground, and a layer file containing all of the spawning grounds identified by each 50 51 52 243 fisherman was created. The spawning grounds were also grouped in space and time for 53 54 55 244 visualization purposes. For example, layer files were created which contained all of the 56 57 245 spawning grounds that were identified in each month. Spawning grounds were classified 58 59 60 12

http://mc.manuscriptcentral.com/icesjms Page 13 of 54 Manuscripts submitted to ICES Journal of Marine Science

1 2 3 246 according to different regions of Georges Bank (Figure 2), and layer files were created to display 4 5 6 247 all of the spawning grounds identified in each region. We used geoprocessing tools in ArcGIS to 7 8 248 quantify the amount of overlap between spawning grounds (i.e., shapefiles) and to quantify the 9 10 11 249 number of fishermen that identified spawning activity at a given location. We identified 12 13 250 “consensus spawning grounds” which were locations where spawning activity was reported by 14 15 16 251 three or more fishermen. 17 18 For Review Only 19 252 3. Results 20 21 22 253 3.1 Analysis of Scientific Data 23 24 254 Goode (1884) and Rich (1929) described the “Winter Fishing Grounds”, an area on eastern 25 26 27 255 Georges Bank between the parallels of 41°30’N and 42°N and 66°38’W and 67°30’W with rocky 28 29 256 bottom and depths ranging from 55 to 73m. The Winter Fishing Ground was reported to serve 30 31 32 257 as a major cod spawning ground in February, March, and April, when dense aggregations of 33 34 258 spawning cod would predictably form every year (Goode, 1884). Bigelow and Schroeder (1953) 35 36 37 259 described major cod spawning activity that occurred on Georges Bank in February, March and 38 39 260 April, and defined the grounds as occurring in depths of approximately 64m from about 40 41 42 261 41 o21’N to 41 o31’N and from 66 o50’W to 67 oW. 43 44 45 262 Nantucket Shoals has long been known to serve as a cod spawning and feeding ground (Bigelow 46 47 48 263 and Schroeder, 1953), and Goode (1884) described a number of important cod fishing grounds 49 50 264 on Nantucket Shoals and east of Cape Cod, including Outer Crab Ledge, Pollock Rip, Great Rip, 51 52 53 265 Fishing, Rip, and Phelps Bank. Smith (1902) collected mature cod from Nantucket Shoals, held 54 55 266 them in a laboratory, and observed that spawning occurred from the middle of November until 56 57 58 59 60 13

http://mc.manuscriptcentral.com/icesjms Manuscripts submitted to ICES Journal of Marine Science Page 14 of 54

1 2 3 267 the end of January. Based on tag returns, Schroeder (1930) concluded that cod spawn on 4 5 6 268 Nantucket Shoals from November through April, with peak spawning activity occurring in 7 8 269 December and January. Schroeder (1930) also mapped cod spawning grounds on Nantucket 9 10 11 270 Shoals that appear to correspond to areas that are currently known as the Fingers, Old Man 12 13 271 Shoal, Old South Shoal, and Davis Bank. Tagging studies documented two migratory 14 15 16 272 behavioural patterns of cod on Nantucket Shoals, with a group of fish that would remain 17 18 273 For Review Only 19 resident on Nantucket Shoals throughout the year, and a second group of cod that would 20 21 274 migrate in the fall to spawning grounds at several locations in southern New England and the 22 23 24 275 mid-Atlantic (Smith, 1902; Schroder, 1930; Wise, 1963). 25 26 27 276 From 1970 through 2014, 22,402 cod were sampled for maturity on the NEFSC spring trawl 28 29 277 survey, of which 415 (1.9%) were ripe, 160 (0.7%) were ripe and running, and 2,837 (12.7%) 30 31 32 278 were spent. These observations suggest that most of the cod spawning activity had ended by 33 34 279 the time the survey sampled on Georges Bank (i.e., 70.9% of tows in April). Nearly all cod 35 36 37 280 spawning hotspots (p < 0.05) were observed on the northern and eastern portions of Georges 38 39 281 Bank from 41 o40’N to 42 o12’N and between 66 oW and 68 oW, with the greatest concentrations 40 41 42 282 of hotspots located on the Northeast Peak (Figure 3). 43 44 45 283 A total of 16,234 cod were sampled for maturity during the NEFSC fall survey, 305 (1.9%) cod 46 47 48 284 were ripe, nine (0.06%) were ripe and running, and 221 (1.4%) were spent. In addition, 3,031 49 50 285 cod (18.7%) were noted to be developing, suggesting that the fall survey, which primarily 51 52 53 286 samples Georges Bank in October (80.1% of tows), occurs before most cod spawning activity. 54 55 56 57 58 59 60 14

http://mc.manuscriptcentral.com/icesjms Page 15 of 54 Manuscripts submitted to ICES Journal of Marine Science

1 2 3 287 Of the 16,668 cod that were sampled for maturity on the DFO trawl survey, 2,359 (14.2%) were 4 5 6 288 ripe, 2,412 (14.5%) were ripe and running, and 2,467 (14.8%) were spent. The greatest 7 8 289 concentrations of spawning cod were observed in Canadian waters on eastern Georges Bank, 9 10 11 290 primarily in depths between 50 and 100m. Spawning cod were occasionally sampled across the 12 13 291 Southwest Part of Georges Bank in depths <100m, but were not observed in any tows 14 15 16 292 completed in waters >100m on the Southwest Part. The majority of cod spawning hotspots (p < 17 18 293 For Review Only o 19 0.05) were observed in Canadian waters on eastern Georges Bank, between 66 5’W and 20 21 294 66 o50’W and 41 o35’N and 42 o5’N, although some spawning hotspots were also present in 22 23 24 295 Closed Area II (Figure 4). 25 26 27 296 While cod eggs have been collected across a wide range of depths on Georges Bank, the highest 28 29 297 concentrations of eggs were typically observed in depths < 100m (Berrien and Sibunka, 1999; 30 31 32 298 Hanke et al., 2000). Cod eggs collected during the MARMAP program indicated that 60% of cod 33 34 299 spawning occurs on Georges Bank between February 23rd and April 6 th , with 90% of the 35 36 37 300 spawning activity occurring between mid-November and mid-May (Page et al., 1998). Cod eggs 38 39 301 were observed on both eastern and western Georges Bank, with the highest concentration of 40 41 42 302 eggs observed on the Northeast Peak (Page et al., 1998; Berrien and Sibunka, 1999). Hanke et 43 44 303 al. (2000) reported that cod eggs were not collected on Georges Bank in July and August, and 45 46 47 304 that cod spawning on Georges Bank began in September and October, and continued with 48 49 305 increasing intensity in November and December. Hanke et al. (2000) noted that cod eggs were 50 51 52 306 abundant on Georges Bank in March and April, and that the spawning season ended in May or 53 54 55 307 June. 56 57 58 59 60 15

http://mc.manuscriptcentral.com/icesjms Manuscripts submitted to ICES Journal of Marine Science Page 16 of 54

1 2 3 308 During the GLOBEC ichthyoplankton sampling program, stage one cod eggs were present in the 4 5 6 309 majority of tows completed from January through March, which suggests that cod spawning 7 8 310 activity is widespread across Georges Bank during these months, and the greatest 9 10 11 311 concentration of stage one cod eggs were observed in February and March (Table 2). In 12 13 312 January, stage one eggs were rarely observed on the Southwest Part and on the Northern Edge, 14 15 16 313 but were abundant on the Northeast Peak, Eastern Georges, and Closed Area II. In February 17 18 314 For Review Only 19 stage one cod eggs were observed to be abundant on the Northeast Peak, Eastern Georges, and 20 21 315 Closed Area II, with lower concentrations on Georges Shoals, the Southwest Part, and the 22 23 24 316 Northern Edge. The distribution of stage one eggs was similar in March, although they were 25 26 317 less commonly observed on the Northern Edge and Georges Shoals. Stage one cod eggs were 27 28 29 318 observed in 35% of the samples collected in April, although the concentrations were 30 31 319 substantially reduced from prior months. Stage one cod eggs were rarely observed in May and 32 33 34 320 June, and no cod eggs were present in the 38 samples collected in July of 1995. 35 36 37 321 Of the 18,804 cod that were sampled by the Canadian Fisheries Observer Program, 7,713 (41%) 38 39 322 were in spawning condition, including 1,780 (9.5%) that were ripe and running. Spawning 40 41 42 323 activity appeared to increase from November to January, before peaking in February and 43 44 324 March, which was when the highest proportion of spawning fish were observed (Table 3). A 45 46 47 325 clear understanding of monthly trends is confounded to some extent by the paucity of fishery 48 49 326 observations in March and April, but samples collected on the DFO bottom trawl survey 50 51 52 327 indicate that cod are still actively spawning on eastern Georges Bank in early spring. The lowest 53 54 55 328 proportions of ripe and ripe and running cod were observed from May through October. In 56 57 58 59 60 16

http://mc.manuscriptcentral.com/icesjms Page 17 of 54 Manuscripts submitted to ICES Journal of Marine Science

1 2 3 329 May and June, a relatively large proportion of cod were spent or recovering, indicating that 4 5 6 330 most spawning activity had ended. 7 8 9 331 3.2 Interviews with Fishermen 10 11 12 332 Of the 52 fishermen we contacted for the project, 40 agreed to complete an interview. The 13 14 333 fishermen we interviewed were captains, and 39 of the 40 had been captains for the majority of 15 16 17 334 their careers. These fishermen collectively had 1,566 total years of fishing experience, including 18 For Review Only 19 335 1,373 years fishing specifically for cod on Georges Bank. On average, each fishermen had 34 20 21 22 336 years of experience targeting Georges Bank cod (range = 12 to 52 years). Eight of the 40 23 24 337 fishermen used multiple gear types to target Georges Bank cod. The majority of fishermen (n = 25 26 27 338 35) used an otter trawl, while others used gillnets (n = 8), longline (n = 7), and rod and reel (n = 28 29 339 2). The captains we interviewed fished from New Bedford, MA, USA (n = 21), Chatham, MA, 30 31 32 340 USA (n = 7), Hyannis, MA, USA (n = 1), Gloucester, MA, USA (n = 1), Nantucket, MA, USA (n = 1), 33 34 341 Boston, MA, USA (n = 1), Montauk, NY, USA (n = 1), Portland, ME, USA (n = 1), Pubnico, Nova 35 36 37 342 Scotia, Canada (n = 3), Yarmouth, Nova Scotia, Canada (n = 2), and Lunenburg, Nova Scotia, 38 39 343 Canada (n = 1) (Figure 1). 40 41 42 43 344 The captains were attentive to the reproductive condition of the cod they had caught, and 44 45 345 could frequently recall in great detail the spawning condition of the fish in their catches. U.S. 46 47 48 346 fishermen dressed their cod catch at sea, and often described seeing eggs and milt spilling out 49 50 347 of the fish as an indication the cod were spawning. Some U.S. fishermen also described eating 51 52 53 348 the roe of the female cod, and indicated that they preferred to eat the roe of developing 54 55 56 57 58 59 60 17

http://mc.manuscriptcentral.com/icesjms Manuscripts submitted to ICES Journal of Marine Science Page 18 of 54

1 2 3 349 females. Some fishermen also reported that they could use their sounders to differentiate 4 5 6 350 between feeding and spawning aggregations of cod. 7 8 9 351 The fishermen identified 210 cod spawning grounds on Georges Bank and Nantucket Shoals, 10 11 12 352 but the same spawning grounds were often identified independently by multiple fishermen 13 14 353 (Figure 5). On average, each fisherman identified five cod spawning grounds (range = 2 to 25 15 16 17 354 spawning grounds). In addition, fishermen distinguished 47 feeding grounds that had high 18 For Review Only 19 355 catch rates of cod during certain times of the year. The region around “Nantucket Shoals and 20 21 22 356 the Channel” was most commonly identified as having cod spawning grounds (Table 4). 23 24 357 Georges Shoal and the Northern Edge were also identified as important regions for cod 25 26 27 358 spawning, while relatively few spawning grounds were reported on the Southwest and Eastern 28 29 359 Parts on Georges Bank. 30 31 32 33 360 Twenty six consensus spawning grounds (areas independently identified by three or more 34 35 361 fishermen) were identified during the interviews (Figure 6). These spawning grounds were 36 37 38 362 widespread throughout the Georges Bank stock area. Many of the spawning grounds are 39 40 363 discrete, and associated with specific bathymetric features such as channels between shoals, 41 42 43 364 edge habitats adjacent to shoals, complex rocky bottom, or areas with steep bathymetric 44 45 365 contours. These consensus spawning grounds were familiar to most of the fishermen that we 46 47 48 366 interviewed, and had common names that fishermen used to identify each location. Nine of 49 50 367 these spawning grounds were located on Nantucket Shoals and two were located in relatively 51 52 53 368 shallow water just east of Cape Cod. Consensus spawning grounds were also identified on 54 55 369 Georges Shoals, the Northern Edge of Georges Bank, and in Closed Area II. Only two consensus 56 57 58 59 60 18

http://mc.manuscriptcentral.com/icesjms Page 19 of 54 Manuscripts submitted to ICES Journal of Marine Science

1 2 3 370 spawning grounds were identified in Canadian waters on the Northeast Peak, and none were 4 5 6 371 identified on the Southwest Part of Georges Bank. 7 8 9 372 The captains had detailed knowledge about the timing of cod spawning, and the seasonal 10 11 12 373 availability of cod on the fishing grounds. Fishermen reported few observations of spawning 13 14 374 cod between July and September, and observed that spawning activity increases from October 15 16 17 375 through December (Figure 7). Over half of the participating fishermen reported active 18 For Review Only 19 376 spawning grounds in January, February, and March. Fishermen noted that cod spawning 20 21 22 377 declined from relatively high levels in April to lesser amounts in May and June. 23 24 25 378 Fishermen reported that the timing of spawning varied between regions of Georges Bank. 26 27 28 379 Although at least one fisherman described spawning activity on Nantucket Shoals and the 29 30 380 Channel in each month, most fishermen identified the spawning period in this region from 31 32 33 381 October through April, with peak spawning in November and December, which is earlier than 34 35 382 regions further east (Figure 8). On Georges Shoal, the spawning season was described as lasting 36 37 38 383 from October through June, with the majority of spawning taking place between December and 39 40 384 May. Peak spawning in the relatively deep waters of the Northern Edge was reported in April 41 42 43 385 and May, which is later than peak spawning reported in other regions. 44 45 46 386 Fishermen typically identified abiotic or biotic habitat characteristics associated with each 47 48 49 387 spawning ground. Cod spawning grounds were most commonly characterized as occurring in 50 51 388 areas with sandy substrates, and fishermen often described sand “lumps” as the preferred 52 53 54 389 spawning habitat (Table 5). Rocky, hard bottom habitats and areas with gravel substrate were 55 56 390 also identified as important for cod spawning activity. The fishermen observed that cod 57 58 59 60 19

http://mc.manuscriptcentral.com/icesjms Manuscripts submitted to ICES Journal of Marine Science Page 20 of 54

1 2 3 391 spawning grounds were often in areas with complex bathymetric features such as ridges, 4 5 6 392 valleys, and deep holes. Fishing in these complex habitats is often difficult and dangerous, 7 8 393 particularly with mobile gear, and requires a priori knowledge that the fishermen acquire 9 10 11 394 through direct experience and information sharing among captains. The persistence of mobile 12 13 395 gear fishermen to pursue cod in these dangerous habitats reflects the importance of these 14 15 16 396 areas as locations where cod would densely aggregate to spawn. The fishermen also reported 17 18 397 For Review Only 19 that cod spawning grounds were frequently in areas with high concentrations of shellfish, 20 21 398 including surf clams (Spisula solidissima ), quahogs (Mercenaria spp. ), and mussels (Mytilus 22 23 24 399 edulis ). Cod spawning aggregations were also often associated with areas that held high 25 26 400 concentrations of forage fish, such as herring (Clupea harengus ), mackerel (Scomber scombrus ), 27 28 29 401 or sand lance (Ammodytes spp .). 30 31 32 402 Fishermen reported that cod spawning occurs across a wide range of depths on Georges Bank 33 34 403 and Nantucket Shoals (Figure 9). Although the bulk of spawning activity was reported to occur 35 36 37 404 in depths ranging from 20 to 91 meters, fishermen also reported that cod also spawn in shoal 38 39 405 habitats (< 20m) and in deep water (> 150m) off the Northern Edge of Georges Bank. 40 41 42 43 406 Of the 210 cod spawning grounds that were identified in this study, the fishermen reported 28 44 45 407 locations that they thought were no longer active and 60 spawning grounds that they thought 46 47 48 408 were still active. However, in the majority of cases (122 of 210 spawning grounds), fishermen 49 50 409 were unsure if the spawning ground they identified was still active. 51 52 53 54 410 Many fishermen participating in this study expressed concern that the current management 55 56 411 units for cod in U.S. waters do not match the biological population structure of the resource. 57 58 59 60 20

http://mc.manuscriptcentral.com/icesjms Page 21 of 54 Manuscripts submitted to ICES Journal of Marine Science

1 2 3 412 Fifteen of the sixteen fishermen who had experience fishing on both eastern and western 4 5 6 413 Georges Bank remarked that cod on eastern Georges Bank were distinct from those on 7 8 414 Nantucket Shoals and western Georges Bank, while one fisherman felt that cod move widely 9 10 11 415 throughout Georges Bank in search of food and optimal temperatures. In particular, the 15 12 13 416 fishermen noted that cod on eastern Georges Bank often attained larger sizes than those taken 14 15 16 417 inshore. Fishermen also consistently described differences in the colour and shape of cod 17 18 418 For Review Only 19 between eastern and western Georges Bank, and observed that cod from eastern Georges Bank 20 21 419 were firmer and higher quality than cod from Nantucket Shoals. Many fishermen identified a 22 23 24 420 longitudinal divide that they believe separate the two groups of cod on eastern and western 25 26 421 Georges Bank, and most suggested a boundary line of either 68 o W or 69 oW. Furthermore, ten 27 28 29 422 fishermen perceived that there is connectivity between cod in the Gulf of Maine and Nantucket 30 31 423 Shoals, and observed that cod make seasonal migrations from the western Gulf of Maine to 32 33 34 424 Nantucket Shoals in search of food. 35 36 37 425 4. Discussion 38 39 426 Fishermen’s reported timing and location of cod spawning activity generally agreed with the 40 41 42 427 scientific literature, but FEK improved the resolution of information available to investigate cod 43 44 428 spawning dynamics on Georges Bank. Many of the spawning grounds identified by fishermen 45 46 47 429 were documented on broad spatial scales from prior scientific research (e.g., ICES, 2005), but 48 49 430 the reports from fishermen suggest that spawning activity can occur on much finer spatial 50 51 52 431 scales, and that spawning grounds are often associated with specific habitat features. Cod 53 54 55 432 spawning dynamics were not well represented by only examining data from seasonal trawl 56 57 58 59 60 21

http://mc.manuscriptcentral.com/icesjms Manuscripts submitted to ICES Journal of Marine Science Page 22 of 54

1 2 3 433 surveys, which do not routinely sample many of the consensus spawning grounds, particularly 4 5 6 434 those on Nantucket Shoals or Georges Shoals. 7 8 9 435 Across all of Georges Bank and Nantucket Shoals, the timing of spawning reported by the 10 11 12 436 fishermen (Figure 7) was in close agreement with results from the MARMAP and GLOBEC 13 14 437 ichthyoplankton sampling programs (Berrien and Sibunka, 1999; Table 2). Developing cod were 15 16 17 438 commonly observed during the NEFSC fall trawl survey, which primarily samples Georges Bank 18 For Review Only 19 439 in October, while ripe and ripe and running cod were relatively rare. Similarly, the fishermen 20 21 22 440 reported relatively little cod spawning activity on Georges Bank in October, and observed that 23 24 441 cod spawning activity increases in November and December. Fishermen observed peak 25 26 27 442 spawning activity from January through March, which matches closely with data collected by 28 29 443 the DFO trawl survey and the Canadian Fisheries Observer program (Table 3). Fishermen 30 31 32 444 reported that cod spawning activity continues into April and May, and this observation is 33 34 445 corroborated by the GLOBEC program, which sampled stage one cod eggs during these months 35 36 37 446 (Table 2). The NEFSC spring trawl survey also observed cod in spawning condition during April 38 39 447 and May, although a higher proportion of the cod were spent. 40 41 42 43 448 Although fishermen reported that the spawning season on Nantucket Shoals is protracted, they 44 45 449 identified November and December as the peak spawning season (Figure 8). This timing aligns 46 47 48 450 well with the scientific information. Sampling conducted during the MARMAP program 49 50 451 observed a relatively high density of cod eggs in the waters around Nantucket Shoals in 51 52 53 452 November and December, which may indicate spawning activity in the region (Berrien and 54 55 453 Sibunka, 1999). Schroeder (1930) documented spawning on Nantucket Shoals from November 56 57 58 59 60 22

http://mc.manuscriptcentral.com/icesjms Page 23 of 54 Manuscripts submitted to ICES Journal of Marine Science

1 2 3 454 through April, with peak spawning in December and January, while Smith (1902) reported that 4 5 6 455 spawning lasted from November until January. 7 8 9 456 The four spawning grounds on Nantucket Shoals described by Schroeder (1930) were identified 10 11 12 457 by three or more fishermen during the interviews, as were a number of other spawning 13 14 458 grounds in the region. The NEFSC trawl survey occasionally samples the consensus spawning 15 16 17 459 grounds that were identified on eastern Nantucket Shoals (“Asia Rip”, ”Middle Rip” and “Fishing 18 For Review Only 19 460 Rip”), and small catches of spawning cod have been observed near these spawning grounds 20 21 22 461 during both the NEFSC fall and spring surveys. However, consensus spawning grounds on 23 24 462 western Nantucket Shoals (e.g., “The Lagoon”, “The Fingers”) are not routinely sampled during 25 26 27 463 the NEFSC trawl survey, possibly due to the complex, shallow bathymetry and strong tidal 28 29 464 currents which make it dangerous to trawl in this area. 30 31 32 33 465 Fishermen reported widespread spawning activity on Georges Shoals from December through 34 35 466 April, and identified several spawning grounds in this region (Figures 5 and 8). Interestingly, 36 37 38 467 Georges Shoals had not been indicated as a cod spawning ground in prior scientific reports. 39 40 468 Cod eggs were sampled on Georges Shoals during the MARMAP program, primarily from 41 42 43 469 November through April, although the egg stages were not reported (Berrien and Sibunka, 44 45 470 1999). However, stage one cod eggs were not commonly observed on Georges Shoals during 46 47 48 471 the GLOBEC program. The consensus spawning grounds identified on Georges Shoals are not 49 50 472 routinely sampled during the NEFSC trawl survey, potentially due to their small size and the 51 52 53 473 complex bathymetry in this region. However, relatively small catches of spawning cod were 54 55 56 57 58 59 60 23

http://mc.manuscriptcentral.com/icesjms Manuscripts submitted to ICES Journal of Marine Science Page 24 of 54

1 2 3 474 observed in close proximity to the “Northeast of Georges Shoal” and “Northeast Rip” consensus 4 5 6 475 spawning grounds during NEFSC spring trawl survey tows. 7 8 9 476 Goode (1884) and Rich (1929) described large cod spawning aggregations on the “Winter 10 11 12 477 Fishing Grounds” that occurred predictably every year in February, March, and April. A number 13 14 478 of the consensus spawning grounds identified in the present study are within the “Winter 15 16 17 479 Fishing Grounds” including “The Leg”, “Billy Doyle’s Hole”, “The Clover”, “East of the Leg”, 18 For Review Only 19 480 “Northeast Rip”, and “Closed Area II” (Figures 5 and 6). The timing of spawning reported by 20 21 22 481 Goode (1884) and Rich (1929) matches closely with the contemporary reports we received from 23 24 482 the fishermen who indicated that most of the cod spawning activity occurs at “The Leg” and 25 26 27 483 “Billy Doyle’s Hole” from January through March. Spawning cod were routinely caught in this 28 29 484 region during the NEFSC spring trawl survey, and many of the cod spawning hotspots from the 30 31 32 485 NEFSC spring survey were located in the “Winter Fishing Grounds”. Spawning cod were also 33 34 486 observed in the “Winter Fishing Ground” during DFO trawl surveys, and stage one cod eggs 35 36 37 487 were sampled in this area from January through April during the GLOBEC program. 38 39 40 488 The fishermen reported cod spawning activity on the Northern Edge from December to June, 41 42 43 489 with a peak in April and May. Cod spawning on the Northern Edge was reported to occur across 44 45 490 a wide range of depths (45-200m), and on a variety of substrates including mud, sand, gravel 46 47 48 491 and rocks. Spawning cod were rarely observed on the Northern Edge during the NEFSC fall 49 50 492 survey, which matches the fishermen’s reports that cod do not spawn on the Northern Edge in 51 52 53 493 October and November. During the NEFSC spring survey, spawning cod were commonly 54 55 494 observed on the Northern Edge in depths ranging from 40-140m, and were occasionally 56 57 58 59 60 24

http://mc.manuscriptcentral.com/icesjms Page 25 of 54 Manuscripts submitted to ICES Journal of Marine Science

1 2 3 495 sampled in waters up to approximately 200m, which is well aligned with fishermen’s reports in 4 5 6 496 this region. During the GLOBEC program, stage one cod eggs were sampled on the Northern 7 8 497 Edge from January through May, with the greatest concentrations observed in February. 9 10 11 12 498 Fishermen reported that the cod spawning season on the Northeast Peak was protracted, 13 14 499 extending from January through August, and identified two consensus spawning grounds in this 15 16 17 500 region. The greatest concentrations of spawning cod observed during both the NEFSC spring 18 For Review Only 19 501 and DFO trawl surveys were on the Northeast Peak of Georges Bank, and the majority of cod 20 21 22 502 spawning hotspots identified by the DFO trawl survey were on the Northeast Peak. Spawning 23 24 503 cod were routinely sampled in this area from January through March by the Canadian Fisheries 25 26 27 504 Observer Program. Cod eggs were also abundant on the Northeast Peak during the MARMAP 28 29 505 and GLOBEC sampling programs, with peak spawning noted in February and March (Berrien and 30 31 32 506 Sibunka, 1999; Sibunka et al., 2006). The Northeast Peak has long been known to serve as a cod 33 34 507 spawning ground (Rich, 1929), and spawning cod sampled on the Northeast Peak were 35 36 37 508 genetically distinct from those sampled on Nantucket Shoals (Ruzzante et al., 1998; Lage et al., 38 39 509 2004; Wirgin et al., 2007; Kovach et al, 2010). Although there is ample scientific evidence of 40 41 42 510 cod spawning on eastern Georges Bank, relatively few fishermen identified spawning grounds 43 44 511 in this region, and no consensus spawning grounds were identified on eastern Georges Bank. 45 46 47 48 512 Few fishermen identified cod spawning activity on the Southwest Part, and no consensus 49 50 513 spawning grounds were observed in this region. Relatively small catches of spawning cod were 51 52 53 514 observed on the Southwest Part during both the NEFSC spring and DFO trawl surveys. Cod eggs 54 55 515 were routinely sampled on the Southwest Part during the MARMAP and GLOBEC 56 57 58 59 60 25

http://mc.manuscriptcentral.com/icesjms Manuscripts submitted to ICES Journal of Marine Science Page 26 of 54

1 2 3 516 ichthyoplankton surveys (Berrien and Sibunka, 1999; Sibunka et al, 2006), although it is unclear 4 5 6 517 if these eggs were spawned on the Southwest Part or advected from other spawning locations. 7 8 9 518 One objective of this study was to use FEK to examine long-term shifts in the distribution of cod 10 11 12 519 spawning activity on Georges Bank and Nantucket Shoals. However, meeting this objective 13 14 520 proved to be difficult. Many of the fishermen we interviewed were retired, and nearly all of the 15 16 17 521 fishermen that are still active reported that they no longer target cod because of the low 18 For Review Only 19 522 quotas in recent years. The majority of fishermen noted that they no longer fish in areas where 20 21 22 523 cod are abundant in order to target stocks with higher allocations and to avoid exceeding their 23 24 524 cod quota. Closed areas that were implemented to reduce fishing mortality (e.g., Closed Areas 25 26 27 525 I and II) or protect habitat (e.g., Nantucket Shoals habitat closure) have prohibited fishing in 28 29 526 many of the areas where the U.S. fleet used to target spawning cod, and U.S. fishermen have 30 31 32 527 been banned from fishing on eastern Georges Bank since the Hague Line was established in 33 34 528 1984. On the Canadian portion of eastern Georges Bank, the trawl fleet is required to use 35 36 37 529 selective fishing gear (e.g., haddock separator trawl) which is designed to reduce cod bycatch, 38 39 530 and fishing is typically prohibited from early February through May to protect spawning cod 40 41 42 531 (Wang et al., 2015). These regulations, and resulting shifts in fishing behaviour, make it difficult 43 44 532 to assess the current productivity of cod spawning on Georges Bank. These regulations could 45 46 47 533 also help to explain why so few fishermen were able to identify cod spawning grounds on 48 49 534 eastern Georges Bank, despite the ample scientific evidence which indicates that major 50 51 52 535 spawning activity occurs in this region. 53 54 55 56 57 58 59 60 26

http://mc.manuscriptcentral.com/icesjms Page 27 of 54 Manuscripts submitted to ICES Journal of Marine Science

1 2 3 536 FEK can serve as an important source of information for stock identification, because fishermen 4 5 6 537 can identify the spawning locations, movement patterns, and morphological characteristics of 7 8 538 their target species (Neis, 1998; Murray et al., 2008b). Fifteen fishermen believed that cod on 9 10 11 539 eastern Georges Bank are separate from those on western Georges Bank and Nantucket Shoals, 12 13 540 based on their observations of disparate migratory patterns and persistent morphological 14 15 16 541 differences between the two groups. Their perceptions are consistent with the available 17 18 542 For Review Only 19 scientific information (Zemeckis et al., 2014c) and have important implications for the 20 21 543 assessment and management of the resource. Fishermen’s observations of distinct groups of 22 23 o 24 544 cod on eastern and western Georges Bank, with a suggested longitudinal boundary near 68 W, 25 26 545 is also consistent with the boundary that Wise (1963) suggested from tagging observations. 27 28 29 546 Tallack (2011) and Loehrke (2013) also observed that few cod moved between Georges Bank 30 31 547 and southern New England, and Schroeder (1930) reported that cod tagged on Nantucket 32 33 o 34 548 Shoals were rarely recaptured on Georges Bank east of 68 W. Additionally, genetic differences 35 36 549 have been observed between cod on eastern Georges Bank and those on Nantucket Shoals 37 38 39 550 (Lage et al., 2004; Wirgin et al., 2007; Kovach et al., 2010). 40 41 42 551 Fishermen also described connectivity between cod on Nantucket Shoals, the Great South 43 44 552 Channel, and the western Gulf of Maine. These observations agree with a genetic study 45 46 47 553 completed by Kovach et al. (2010), which identified a “southern complex” of winter spawning 48 49 554 cod in the Gulf of Maine and locations south of Cape Cod, Massachusetts. Based on 50 51 52 555 conventional tag returns, Tallack (2011) observed considerable movement of cod between 53 54 55 556 Cape Cod and the inshore Gulf of Maine, although some sub-legal cod were also reported to 56 57 557 migrate from Cape Cod to Georges Bank. We suggest that FEK should be routinely collected as 58 59 60 27

http://mc.manuscriptcentral.com/icesjms Manuscripts submitted to ICES Journal of Marine Science Page 28 of 54

1 2 3 558 part of interdisciplinary stock identification studies along with traditional stock identification 4 5 6 559 techniques such as genetics, analysis of life history traits, and applied marks. 7 8 9 560 The FEK collected in this study can serve as a valuable guide for future research efforts to better 10 11 12 561 understand the stock dynamics and productivity of cod. For example, a directed trawl survey 13 14 562 could be initiated to sample at the spawning locations that were identified by fishermen and 15 16 17 563 assess which spawning grounds remain active. In addition, the location and timing of cod 18 For Review Only 19 564 spawning grounds that were identified by fishermen could be used to inform the initial 20 21 22 565 conditions for Individual Based Modeling experiments designed to examine the transport and 23 24 566 survival of cod eggs and larvae. Finally, these putative spawning areas could be used to guide 25 26 27 567 the design of future tagging studies to investigate stock identity and connectivity of spawning 28 29 568 cod. 30 31 32 33 569 The high participation rates, and the quality and reliability of the information provided by the 34 35 570 fishermen in this study were likely the result of several factors. First, through prior cooperative 36 37 38 571 research projects we had a pre-existing relationship with many of these fishermen, enabling us 39 40 572 to gain their trust. Secondly, we conducted the interviews at times and locations that were 41 42 43 573 convenient to the fishermen, which didn’t cause them to lose fishing opportunities, and 44 45 574 showed them that we valued their perspectives. Many fishermen were thankful that we 46 47 48 575 solicited their knowledge and expertise, and the appreciated the ability to contribute to 49 50 576 scientific research. One interviewer is fluent in Portuguese, which allowed us to interview eight 51 52 53 577 fishermen that have extensive experience fishing for Georges Bank cod, but often do not 54 55 578 participate in fishery management because of a language barrier. Many of the fishermen that 56 57 58 59 60 28

http://mc.manuscriptcentral.com/icesjms Page 29 of 54 Manuscripts submitted to ICES Journal of Marine Science

1 2 3 579 we interviewed were retired, and may have been more willing to share their information 4 5 6 580 because they would not be affected by any resultant management measures. Finally, given the 7 8 581 depleted status of Georges Bank cod, some fishermen were motivated to participate out of a 9 10 11 582 desire to share knowledge that could potentially help contribute to cod rebuilding. 12 13 14 583 In conclusion, FEK has improved our understanding of cod spawning activity on Georges Bank. 15 16 17 584 There was substantial agreement between the available scientific information and the 18 For Review Only 19 585 information reported by the fishermen about the spatial and temporal distribution of cod 20 21 22 586 spawning on Georges Bank, which suggests that the reports provided by the fishermen were 23 24 587 reliable. Although the major spawning grounds reported by fishermen were also identified in 25 26 27 588 the scientific literature (e.g., the Northeast Peak and Nantucket Shoals), the fishermen we 28 29 589 interviewed had extensive knowledge related to the timing and location of cod spawning, and 30 31 32 590 identified some cod spawning grounds that were not previously described in scientific reports. 33 34 591 Multiple fishermen who were interviewed independently identified several fine-scale 35 36 2 37 592 geographic regions (often < 50km ) where cod spawning occurs, and they had detailed 38 39 593 knowledge about the habitat characteristics of these fine-scale locations. This level of spatial 40 41 42 594 resolution could not typically be obtained using traditional scientific data collection approaches. 43 44 595 The information gathered in this study can serve as a valuable guide for future research, and 45 46 47 596 can help inform future management actions. 48 49 50 597 Acknowledgements 51 52 53 598 We are indebted to the fishermen who generously shared their time and knowledge with us. 54 55 599 We are especially grateful to Gilbert Donaldson, who helped us contact fishermen in Nova 56 57 58 59 60 29

http://mc.manuscriptcentral.com/icesjms Manuscripts submitted to ICES Journal of Marine Science Page 30 of 54

1 2 3 600 Scotia and schedule interviews. Data requests were fulfilled by Sean Lucey, Rich McBride, and 4 5 6 601 Jon Hare at the NOAA Northeast Fisheries Science Center, and Yanjun Wang and Irene 7 8 602 Andruschenko at the Canadian Department of Fisheries and Oceans. We also thank Crista Bank 9 10 11 603 for her assistance with the interviews. David Pierce provided valuable feedback on an earlier 12 13 604 version of this manuscript. We also thank Ted Ames who helped us formulate the interview 14 15 16 605 questions, Rob Stephenson and Greg Skomal for their helpful feedback on the project, and Mike 17 18 606 For Review Only 19 Armstrong, Micah Dean, and Bill Hoffman for their assistance. 20 21 22 607 Funding 23 24 608 Funding for this research was provided by the New England Fishery Management Council 25 26 27 609 through a grant administered by the Northeast Consortium. 28 29 30 610 References 31 32 33 611 Ames, T. 1997. Cod and haddock spawning grounds in the Gulf of Maine from Grand Manan 34 35 612 Channel to Ipswich Bay. In The Implications of Localized Fisheries Stocks, pp. 55-64. Ed. by I. H. 36 37 38 613 von Herbing, and M. Tupper. Natural Resource, Agriculture, and Engineering Service, Ithaca, 39 40 614 New York. 200 pp. 41 42 43 615 44 45 616 Ames, E. P. 2004. Atlantic cod stock structure in the Gulf of Maine. Fisheries, 29(1), 10-28. 46 47 48 617 49 50 618 Anuchiracheeva, S., Demaine, H., Shivakoti, G. P., and Ruddle, K. 2003. Systematizing local 51 52 53 619 knowledge using GIS: fisheries management in Bang Saphan Bay, Thailand. Ocean and Coastal 54 55 620 Management, 46: 1049-1068. 56 57 58 621 59 60 30

http://mc.manuscriptcentral.com/icesjms Page 31 of 54 Manuscripts submitted to ICES Journal of Marine Science

1 2 3 622 Armstrong, M. P., Dean, M. J., Hoffman, W. S., Zemeckis, D. R., Nies, T. A., Pierce, D.E., Diodati, 4 5 6 623 P. J., et al . 2013. The application of small scale fishery closures to protect Atlantic cod spawning 7 8 624 aggregations in the inshore Gulf of Maine. Fisheries Research, 141: 62-69. 9 10 11 625 12 13 626 Babbie, E. 1989. The Practice of Social Research, 5th Edition. Wadsworth, Belmont, California. 14 15 16 627 17 18 628 For Review Only 19 Bergmann, M., Hinz, H., Blyth, R. E., Kaiser, M. J., Rogers, S. I., and Armstrong, M. 2004. Using 20 21 629 knowledge from fishers and fisheries scientists to identify possible groundfish ‘Essential Fish 22 23 24 630 Habitats’. Fisheries Research, 66: 373-379. 25 26 631 27 28 29 632 Berrien, P. L., and Sibunka, P. 1999. Distribution patterns of fish eggs in the U.S. northeast 30 31 633 continental shelf ecosystem, 1977-1987. NOAA Technical Memorandum NEFSC 145. 316 pp. 32 33 34 634 35 36 635 Bigelow H. B., and W. C. Schroeder. 1953 Fishes of the Gulf of Maine. Fishery Bulletin US 53: 1- 37 38 39 636 577. 40 41 637 42 43 44 638 Burnett, J., O’Brien, L., Mayo, R. K., Darde, J. A. and Bohan, M. 1989. Finfish maturity sampling 45 46 639 and classification schemes used during Northeast Fisheries Center bottom trawl surveys, 1963- 47 48 49 640 89. NOAA Technical Memorandum NMFS-F/NEC-76. 14 pp. 50 51 641 52 53 54 55 56 57 58 59 60 31

http://mc.manuscriptcentral.com/icesjms Manuscripts submitted to ICES Journal of Marine Science Page 32 of 54

1 2 3 642 Dean, M. J., Hoffman, W. S., and Armstrong, M. P. 2012. Disruption of an Atlantic cod spawning 4 5 6 643 aggregation resulting from the opening of a directed gill-net fishery. North American Journal of 7 8 644 Fisheries Management, 32(1): 124-134. 9 10 11 645 12 13 646 Dean, M. J., Hoffman, W. S., Zemeckis, D. R., and Armstrong, M. P. 2014. Fine-scale diel and 14 15 16 647 gender-based patterns in behaviour of Atlantic cod ( Gadus morhua ) on a spawning ground in 17 18 648 For Review Only 19 the western Gulf of Maine. ICES Journal of Marine Science, 71(6): 1474-1489. 20 21 649 22 23 24 650 Getis, A., and Ord, J. K. 1992. The analysis of spatial association by use of distance statistics. 25 26 651 Geographical Analysis, 24(3): 189-206. 27 28 29 652 30 31 653 Goode, G. B. 1884. The fisheries and fishing industry of the United States. United States 32 33 34 654 Commission of Fish and Fisheries. 35 36 655 37 38 39 656 Grabowski, T. B., Thorsteinsson, V., McAdam, B., and Marteinsdóttir, G. 2011. Evidence of 40 41 657 segregated spawning in a single marine fish stock: Sympatric divergence of ecotypes in 42 43 44 658 Icelandic cod? PLoS ONE, 6(3): 1-9. 45 46 659 47 48 49 660 Gurshin, C. W. D., Howell, W. H., and Jech, J. M. 2013. Synoptic acoustic and trawl surveys of 50 51 661 spring-spawning Atlantic cod in the Gulf of Maine cod spawning protection area. Fisheries 52 53 54 662 Research, 141: 44-61. 55 56 663 57 58 59 60 32

http://mc.manuscriptcentral.com/icesjms Page 33 of 54 Manuscripts submitted to ICES Journal of Marine Science

1 2 3 664 Hanke, A. R., F. H. Page, and J. Neilson. 2000. Distribution of Atlantic cod (Gadus morhua) eggs 4 5 6 665 and larvae on the Scotian Shelf. Canadian Technical Report of Fisheries and Aquatic Sciences 7 8 666 2308. 140 pp. 9 10 11 667 12 13 668 Hedeholm, R. B., Jacobsen, R. B., and Nielsen, E. E. 2016. Learning from ‘apparent consensus’ in 14 15 16 669 TAC disputes: exploring knowledge overlaps in LEK and genetic categorization of Atlantic cod. 17 18 670 For Review Only 19 Marine Policy, 69: 114-120. 20 21 671 22 23 24 672 Hind, E. J. 2015. A review of the past, the present, and the future of fishers’ knowledge 25 26 673 research: a challenge to established fisheries science. ICES Journal of Marine Science, 72(2), 27 28 29 674 341-358. 30 31 675 32 33 34 676 ICES, 2005. Spawning and life history information for north Atlantic cod stocks. ICES 35 36 677 Cooperative Research Report No. 274. 162 pp. 37 38 39 678 40 41 679 Johannes, R. E., Freeman, M. M. R., and Hamilton, R. J. 2000. Ignore fishers’ knowledge and 42 43 44 680 miss the boat. Fish and Fisheries, 1: 257-271. 45 46 681 47 48 49 682 Kovach, A. I., Breton, T. S., Berlinsky, D. L., Maceda, L., and Wirgin, I. 2010. Fine-scale spatial 50 51 683 and temporal genetic structure of Atlantic cod off the Atlantic coast of the USA. Marine 52 53 54 684 Ecology Progress Series, 410: 177-195. 55 56 685 57 58 59 60 33

http://mc.manuscriptcentral.com/icesjms Manuscripts submitted to ICES Journal of Marine Science Page 34 of 54

1 2 3 686 Lage, C., Kuhn, K., and Kornfield, I. 2004. Genetic differentiation among Atlantic cod ( Gadus 4 5 6 687 morhua ) from Browns Bank, Georges Bank, and Nantucket Shoals. Fishery Bulletin, 102: 289- 7 8 688 297. 9 10 11 689 12 13 690 Loehrke, J. L. 2013. Movement patterns of Atlantic cod ( Gadus morhua ) spawning groups off 14 15 16 691 New England. MS Thesis, University of Massachusetts, Dartmouth, MA, USA. 17 18 692 For Review Only 19 20 21 693 Macdonald, P., Angus, C. H., Cleasby, I. R., and Marshall, C. T. 2014. Fishers’ knowledge as an 22 23 24 694 indicator of spatial and temporal trends in abundance of commercial fish species: megrim 25 26 695 (Lepidorhombus whiffiagonis ) in the northern North Sea. Marine Policy, 45, 228-239. 27 28 29 696 30 31 697 Maurstad, A. 2002. Fishing in murky waters: ethics and politics of research on fisher knowledge. 32 33 34 698 Marine Policy, 26: 159-166. 35 36 699 37 38 39 700 Morgan, M. J., DeBlois, E. M., and Rose, G. A. 1997. An observation of the reaction of Atlantic 40 41 701 cod ( Gadus morhua ) in a spawning shoal to bottom trawling. Canadian Journal of Fisheries and 42 43 44 702 Aquatic Sciences, 54 (Suppl. 1): 217-223. 45 46 703 47 48 49 704 Morrison, C. 1990. Histology of the Atlantic cod, Gadus morhua : an atlas. Part Three. 50 51 705 Reproductive Tract. Canadian Special Publications of Fisheries and Aquatic Sciences 100. 183 52 53 54 706 pp. 55 56 707 57 58 59 60 34

http://mc.manuscriptcentral.com/icesjms Page 35 of 54 Manuscripts submitted to ICES Journal of Marine Science

1 2 3 708 Murray, G., Neis, B., Schneider, D. C., Ings, D., Gosse, K., Whalen, J., and Palmer, C. T. 2008a. 4 5 6 709 Opening the black box: methods, procedures, and challenges in the historical reconstruction of 7 8 710 marine social-ecological systems. In Making and Moving Knowledge: Interdisciplinary and 9 10 11 711 Community-Based Research in a World on the Edge, pp. 100-120. Ed. by J. S. Lutz, and B. Neis. 12 13 712 McGill-Queen’s University Press, Montreal. 360 pp. 14 15 16 713 17 18 714 For Review Only 19 Murray, G., Neis, B., Palmer, C. T., and Schneider, D. C. 2008b. Mapping cod: fisheries science, 20 21 715 fish harvesters’ ecological knowledge and cod migrations in the northern Gulf of St. Lawrence. 22 23 24 716 Human Ecology, 36: 581-598. 25 26 717 27 28 29 718 Neis. B. 1998. Fishers’ ecological knowledge and the identification and management of localized 30 31 719 populations of Atlantic cod ( Gadus morhua ) in northeast Newfoundland. In The Implications of 32 33 34 720 Localized Fisheries Stocks, pp. 177-188. Ed. by I. H. von Herbing, and M. Tupper. Natural 35 36 721 Resource, Agriculture, and Engineering Service, Ithaca, New York. 200 pp. 37 38 39 722 40 41 723 Neis, B., Felt, L. F., Haedrich, R. L., and Schneider, D. C. 1999a. An interdisciplinary method for 42 43 44 724 collecting and integrating fishers’ ecological knowledge into resource management. In Fishing 45 46 725 Places, Fishing People: Traditions and Issues in Small-Scale Canadian Fisheries, pp. 217-238. Ed. 47 48 49 726 by D. Newell and R. E. Ommer. University of Toronto Press, Toronto. 412 pp. 50 51 727 52 53 54 55 56 57 58 59 60 35

http://mc.manuscriptcentral.com/icesjms Manuscripts submitted to ICES Journal of Marine Science Page 36 of 54

1 2 3 728 Neis, B., Schneider, D. C., Felt, L., Haedrich, R. L., Fischer, J., and Hutchings, J. A. 1999b. Fisheries 4 5 6 729 assessment: what can be learned from interviewing resource users? Canadian Journal of 7 8 730 Fisheries and Aquatic Sciences, 56: 1949-1963. 9 10 11 731 12 13 732 New England Fishery Management Council (NEFMC). 2009. Northeast Multispecies Fisheries 14 15 16 733 Management Plan - Amendment 16, 905 pp. 17 18 734 For Review Only 19 20 21 735 New England Fishery Management Council (NEFMC). 2013. Omnibus essential fish habitat 22 23 24 736 amendment 2 draft environmental impact statement. Appendix E – synopsis of Closed Area 25 26 737 Technical Team analysis of juvenile groundfish habitats and groundfish spawning areas, 67 pp. 27 28 29 738 30 31 739 Northeast Fisheries Science Center (NEFSC). 2013. 55th Northeast Regional Stock Assessment 32 33 34 740 Workshop (55th SAW) Assessment Report. US Department of Commerce, Northeast Fisheries 35 36 741 Science Center Reference Document 13-11, 845 pp. 37 38 39 742 40 41 743 Northeast Fisheries Science Center (NEFSC). 2015. Operational Assessment of 20 Northeast 42 43 44 744 Groundfish Stocks, Updated Through 2014. US Department of Commerce, Northeast Fisheries 45 46 745 Science Center Reference Document 15-24, 251 pp. 47 48 49 746 50 51 747 Page, F. H., Losier, R., and Berrien, P. 1998. Spawning time of haddock and cod on Georges Bank 52 53 54 748 as indicated by MARMAP ichthyoplankton data. Canadian Stock Assessment Secretariat 55 56 749 Research Document 97/130. Fisheries and Oceans Canada. 26 pp. 57 58 59 60 36

http://mc.manuscriptcentral.com/icesjms Page 37 of 54 Manuscripts submitted to ICES Journal of Marine Science

1 2 3 750 4 5 6 751 Pederson, J., and Hall-Arber, M. 1999. Fish habitat: a focus on New England fishermen’s 7 8 752 perspectives. American Fisheries Society Symposium, 22: 188-211. 9 10 11 753 12 13 754 Rich, W. H. 1929. Fishing grounds of the Gulf of Maine. U.S. Commissioner of Fisheries, 14 15 16 755 Washington, D.C. 17 18 756 For Review Only 19 20 21 757 Rose, G. A. 2007. Cod: The ecological history of the North Atlantic fisheries. Breakwater Books 22 23 24 758 Ltd., St. John’s, NF. 592 pp. 25 26 759 27 28 29 760 Rosenberg, A. A., Bolster, W. J., Alexander, K. E., Leavenworth, W. B., Cooper, A. B., and 30 31 761 McKenzie, M. G. 2005. The history of ocean resources: modeling cod biomass using historical 32 33 34 762 records. Frontiers in Ecology and the Environment, 3(2): 84-90. 35 36 763 37 38 39 764 Rothschild, B. J., Keiley, E. F., and Jiao, Y. 2014. Failure to eliminate over-fishing and attain 40 41 765 optimum yield in the New England groundfish fishery. ICES Journal of Marine Science, 71: 226- 42 43 44 766 233. 45 46 767 47 48 49 768 Ruzzante, D. E., Taggart, C. T., and Cook, D. 1998. A nuclear DNA basis for shelf- and bank-scale 50 51 769 population structure in northwest Atlantic cod ( Gadus morhua ): Labrador to Georges Bank. 52 53 54 770 Molecular Ecology, 7: 1663-1680. 55 56 771 57 58 59 60 37

http://mc.manuscriptcentral.com/icesjms Manuscripts submitted to ICES Journal of Marine Science Page 38 of 54

1 2 3 772 Scholz, A., Bonzon, K., Fujita, R., Benjamin, N., Woodling, N., Black, P., and Steinback, C. 2004. 4 5 6 773 Participatory socioeconomic analysis: drawing on fishermen’s knowledge for marine protected 7 8 774 area planning in California. Marine Policy, 28: 335-349. 9 10 11 775 12 13 776 Schroeder, W. C. 1930 Migrations and other phases in the life history of the cod off southern 14 15 16 777 New England. Bulletin of the Bureau of Fisheries, 46. 136 pp. 17 18 778 For Review Only 19 20 21 779 Serchuk, F. M., and Wigley, S. E. 1992. Assessment and management of the Georges Bank cod 22 23 24 780 fishery: An historical review and evaluation. Journal of Northwest Atlantic Fishery Science, 13: 25 26 781 25-52. 27 28 29 782 30 31 783 Sibunka, J. D., Johnson, D. L., and Berrien, P. L. 2006. Distribution and abundance of fish eggs 32 33 34 784 collected during the GLOBEC broad-scale Georges Bank surveys, 1995-1999. NOAA Technical 35 36 785 Memorandum NEFSC-NE-199. 72 pp. 37 38 39 786 40 41 787 Siceloff, L., and Howell, W. H. 2013. Fine-scale temporal and spatial distributions of Atlantic cod 42 43 44 788 (Gadus morhua ) on a western Gulf of Maine spawning ground. Fisheries Research, 141: 31-43. 45 46 789 47 48 49 790 Silvano, R. A. M., MacCord, P. F. L., Lima, R. V., and Begossi, A. 2006. When does this fish 50 51 791 spawn? Fishermen’s local knowledge of migration and reproduction in Brazilian coastal fishes. 52 53 54 792 Environmental Biology of Fish, 76: 371-386. 55 56 793 57 58 59 60 38

http://mc.manuscriptcentral.com/icesjms Page 39 of 54 Manuscripts submitted to ICES Journal of Marine Science

1 2 3 794 Smedbol, R. K., and Stephenson, R. 2001. The importance of managing within-species diversity 4 5 6 795 in cod and herring fisheries of the north-western Atlantic. Journal of Fish Biology, 59:109-128. 7 8 796 9 10 11 797 Smith, H. M. 1902. Notes on the tagging of four thousand adult cod at Woods Hole, 12 13 798 Massachusetts. U.S. Commission of Fish and Fisheries, Part XXVII. Report of the Commissioner 14 15 16 799 for the year ending June 30, 1901. pp. 193-208. 17 18 800 For Review Only 19 20 21 801 Stephenson, R. L., Paul, S., Pastoors, M. A., Kraan, M., Holm, P., Wiber, M., Mackinson, S., 22 23 24 802 Dankel, D. J., Brooks, K., and Benson, A. 2016. Integrating fishers’ knowledge research in science 25 26 803 and management. ICES Journal of Marine Science, doi:10.193/icesjms/fsw025. 27 28 29 804 30 31 805 Svedäng, H., Stål, J. Sterner, T., and Cardinale, M. 2010. Consequences of subpopulation 32 33 34 806 structure on fisheries management: cod ( Gadus morhua ) in the Kattegat and Oresund (North 35 36 807 Sea). Reviews in Fisheries Science, 18(2): 139-150. 37 38 39 808 40 41 809 Tallack, S. M. L. 2011. Stock identification applications of conventional tagging data for Atlantic 42 43 nd 44 810 cod in the Gulf of Maine. In Proceedings from the 2 International Symposium on Advances in 45 46 811 Fish Tagging and Marking Techniques, pp. 1-15. Ed. By J. McKenzie, B. Parsons, A. C. Seitz, R. K. 47 48 49 812 Kopf, M. Mesa, and Q. Phelps. American Fisheries Society, 76. 560 pp. 50 51 813 52 53 54 814 Wang, Y., O’Brien, L., Andruschenko, I., and Clark, K. J. 2015. Assessment of eastern Georges 55 56 815 Bank Atlantic cod for 2015. TRAC Reference Document 2015/03. 90pp. 57 58 59 60 39

http://mc.manuscriptcentral.com/icesjms Manuscripts submitted to ICES Journal of Marine Science Page 40 of 54

1 2 3 816 4 5 6 817 Wirgin, I., Kovach, A. I., Maceda, L., Roy, N. K., Waldman, J., and Berlinsky, D. L. 2007. Stock 7 8 818 identification of Atlantic cod in U.S. waters using microsatellite and single nucleotide 9 10 11 819 polymorphism DNA analyses. Transactions of the American Fisheries Society, 136: 375-391. 12 13 820 14 15 16 821 Wise, J. P. 1963. Cod groups in the New England Area. Fishery Bulletin, 63: 189-203. 17 18 822 For Review Only 19 20 21 823 Yates, K. L. (2014). View from the wheelhouse: perceptions on marine management from the 22 23 24 824 fishing community and suggestions for improvement. Marine Policy, 48: 39-50. 25 26 825 27 28 29 826 Zemeckis, D. R., Hoffman, W. S., Dean, M. J., Armstrong, M. P., and Cadrin, S. X. 2014a. 30 31 827 Spawning site fidelity by Atlantic cod ( Gadus morhua ) in the Gulf of Maine: implications for 32 33 34 828 population structure and rebuilding. ICES Journal of Marine Science, 71(6): 1356-1365. 35 36 829 37 38 39 830 Zemeckis, D. R., Dean, M. J., and Cadrin, S. X. 2014b. Spawning dynamics and associated 40 41 831 management implications for Atlantic cod ( Gadus morhua ). North American Journal of Fisheries 42 43 44 832 Management, 34(2): 424-442. 45 46 833 47 48 49 834 Zemeckis, D. R., Martins, D., Kerr, L. A., and Cadrin, S. X. 2014c. Stock identification of Atlantic 50 51 835 cod (Gadus morhua) in US waters: an interdisciplinary approach. ICES Journal of Marine 52 53 54 836 Science, 71(6): 1490-1506. 55 56 57 58 59 60 40

http://mc.manuscriptcentral.com/icesjms Page 41 of 54 Manuscripts submitted to ICES Journal of Marine Science

1 2 3 Part 1: Demograhic Questions 4 1) What is your year of birth? 5 2) How many total years of experience do you have commercial fishing? 6 2) How many years of experience do you have commercial fishing for groundfish on Georges Bank and Nantucket Shoals? 7 3) How many years of experience do you have fishing specifically for cod on Georges Bank and Nantucket Shoals? 8 4) Of the total years you have spent fishing for groundfish on Georges Bank and Nantucket Shoals, how many years 9 were you a deckhand, a mate, and a captain? 10 5) What gear types did you use when fishing for cod? 11 12 Part 2: Questions Specific to Each Spawning Ground 13 1) What was the name of the spawning ground? 14 2) What months did spawning occur here? 15 3) How were you able to determine this area was a cod spawning ground? 16 4) What maturity stages did you see at this spawning ground? 17 5) Were there more males or females in the catch? 18 6) What depths did you find spawningFor cod at Reviewthis location? Only 19 7) How would you describe the magnitude of cod spawning at this site? 20 8) What was the predictability or consistency of this spawning ground from year to year? 21 9) What size were the majority of cod you encountered at this spawning ground? 22 10) How would you describe the habitat at this spawning ground? 23 11) Is this spawning ground still active? 24 11a) If the spawning ground is still active, what is the magnitude of spawning now, compared to past levels? 25 11b) If the spawning ground is no longer active, what year did it cease to be active, and why do you think 26 spawning stopped at this location? 27 Part 3: Secondary Questions that Were Asked As Time Allowed 28 1) Did cod feed during the spawning season? Were cod also feeding at the spawning ground? 29 2) Was cod spawning activity related to time of day? 30 3) Is there connectivity between the spawning sites you identified within the larger spawning ground? 31 4) Did you notice a change in the size of the cod over the course of the year? 32 5) Was there anything unique about the shape of the cod at any of the spawning grounds? 33 6) Was there anything unique about the colour of the cod at any of the spawning grounds? 34 7) Do you think there is a boundary between Georges Bank cod and inshore cod? 35 36 37 38 Table 1. Questions that we asked fishermen during the interviews. 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 http://mc.manuscriptcentral.com/icesjms Manuscripts submitted to ICES Journal of Marine Science Page 42 of 54

1 2 3 Month # of Tows # of Tows With # of Tows With Mean 4 Cod Eggs (all Stage One Cod Concentration 5 6 stages) Eggs Stage One Cod Eggs 7 (number/10m^2) 8 9 January 229 148 (65%) 115 (50%) 49.1 10 February 338 260 (77%) 212 (63%) 173.1 11 March 342 223 (65%) 171 (50%) 157.9 12 13 April 334 179 (54%) 118 (35%) 17.9 14 May 374 55 (15%) 34 (9%) 11.9 15 June 200 7 (4%) 3 (2%) 2.2 16 July 38 0 (0%) 0 (0%) 0 17

18 For Review Only 19 20 Table 2. Monthly sampling intensity and concentration of cod eggs observed during the GLOBEC 21 ichthyoplankton sampling program from 1995 through 1999. The data were provided by Jon 22 Hare at the NOAA Northeast Fisheries Science Center, and are described in Sibunka et al. 23 24 (2006). 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 http://mc.manuscriptcentral.com/icesjms Page 43 of 54 Manuscripts submitted to ICES Journal of Marine Science

1 2 3 Number of Number of Ripe 4 Developing Developing Month Tows Cod Immature Ripe and Spent Recovering Resting 5 1 2 Sampled Sampled Running 6 7 Jan. 852 9537 3% 14% 29% 38% 9% 3% 3% 2% 8 Feb. 240 3983 2% 8% 20% 31% 20% 7% 9% 3% 9 Mar. 6 28 0% 4% 4% 39% 50% 4% 0% 0% 10 Apr. 0 0 11 12 May 18 362 2% 36% 2% 1% 0% 17% 25% 16% 13 Jun. 47 496 10% 22% 9% 3% 2% 23% 22% 8% 14 Jul. 85 1089 11% 20% 7% 2% 0% 4% 15% 40% 15 Aug. 132 1466 8% 20% 6% 1% 0% 5% 11% 48% 16 17 Sep. 86 721 8% 35% 13% 5% 1% 0% 5% 34% 18 Oct. 30For 366 Review 3% 32% 17%Only 3% 0% 4% 10% 30% 19 Nov. 29 372 5% 33% 30% 8% 5% 2% 4% 12% 20 Dec. 41 384 3% 20% 34% 22% 5% 2% 3% 12% 21 22 23 24 Table 3. Cod maturity stages observed each month by the Canadian Fisheries Observer 25 Program on eastern Georges Bank from 1983 through 2015. The data were provided by Irene 26 Andruschenko at the Canadian Department of Fisheries and Oceans. 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 http://mc.manuscriptcentral.com/icesjms Manuscripts submitted to ICES Journal of Marine Science Page 44 of 54

1 2 3 Region Number of Number of Number of Number of 4 Fishermen that Spawning Fishermen that Feeding 5 6 Identified Spawning Grounds Identified Grounds 7 Grounds Identified Feeding Grounds Identified 8 Nantucket Shoals and Channel 27 84 8 13 9 10 Closed Area I 7 7 2 2 11 Northern Edge 18 28 1 1 12 Georges Shoals 20 41 7 8 13 Southwest Part 8 9 2 2 14 15 Closed Area II 12 21 1 2 16 Northeast Peak 12 13 5 15 17 Eastern Georges 4 7 2 4 18 For Review Only 19 20 21 Table 4. Number of cod spawning and feeding grounds that were identified in each region of 22 Georges Bank during interviews with fishermen. 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 http://mc.manuscriptcentral.com/icesjms Page 45 of 54 Manuscripts submitted to ICES Journal of Marine Science

1 2 3 Abiotic Habitat Attributes Biotic Habitat Attributes 4 5 Sand or sand lumps 105 Shellfish 50 6 Rocks and "hard" bottom 67 Herring 9 7 Gravel 36 Sand lance 8 8 Pebble and cobble 11 Macroalgae 5 9 10 Mud 9 Mackerel 3 11 Strong tides and currents 9 Sea stars 3 12 Shipwrecks 7Worms 2 13 "Broken" bottom 3 Squid 1 14 15 Silver hake 1 16 Sponges 1 17 Small haddock 1 18 For ReviewCrabs Only 1 19 20 21 22 23 Table 5. Fishermen reported habitat characteristics associated with cod spawning grounds. 24 The table shows the frequency with which each habitat characteristic was reported. 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 http://mc.manuscriptcentral.com/icesjms Manuscripts submitted to ICES Journal of Marine Science Page 46 of 54

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 For Review Only 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 Figure 1. Management boundaries for cod in U.S. waters, including the Gulf of Maine and Georges Bank 34 stocks. Cod on eastern Georges Bank (outlined in bold) are managed as a transboundary resource by the 35 United States and Canada. The black numbered circles depict the ports where we interviewed fishermen: 1 = Montauk, NY, 2 = New Bedford, MA, 3 = Hyannis, MA, 4 = Chatham, MA, 5 = Nantucket, MA, 6 = Boston, 36 MA, 7 = Gloucester, MA, 8 = Portland, ME, 9 = Yarmouth, Nova Scotia, 10 = Pubnico, Nova Scotia, and 11 37 = Lunenburg, Nova Scotia. 38 39 215x166mm (600 x 600 DPI) 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 http://mc.manuscriptcentral.com/icesjms Page 47 of 54 Manuscripts submitted to ICES Journal of Marine Science

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 For Review Only 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 Figure 2. Eight geographic regions that were used to classify the location and timing of cod spawning on 34 Georges Bank. 35 215x166mm (600 x 600 DPI) 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 http://mc.manuscriptcentral.com/icesjms Manuscripts submitted to ICES Journal of Marine Science Page 48 of 54

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 For Review Only 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 Figure 3. Distribution of cod spawning hotspots observed during the NEFSC spring trawl survey from 1970 34 through 2014. The number of hotspots within a 100km2 grid were summarized to identify spatial clusters of 35 spawning activity. Cod fishing has been prohibited in the Closed Areas since 1994, with some exceptions. The Exclusive Economic Zone marks the boundary between the U.S. and Canadian portions of 36 Georges Bank. 37 38 215x166mm (300 x 300 DPI) 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 http://mc.manuscriptcentral.com/icesjms Page 49 of 54 Manuscripts submitted to ICES Journal of Marine Science

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 For Review Only 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 Figure 4. Distribution of cod spawning hotspots observed during the DFO trawl survey from 1987 through 34 2015. The number of hotspots within a 100km2 grid were summarized to identify spatial clusters of 35 spawning activity.

36 215x166mm (300 x 300 DPI) 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 http://mc.manuscriptcentral.com/icesjms Manuscripts submitted to ICES Journal of Marine Science Page 50 of 54

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 For Review Only 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 Figure 5. Cod spawning grounds that were identified by fishermen. Each polygon represents a spawning 34 ground that was identified by a single fisherman. The shading is used to identify areas where there is 35 overlap in the spawning locations reported by multiple fishermen. The rectangle outlined in black depicts the “Winter Fishing Grounds” that were described by Goode (1884) and Rich (1929). The hashed rectangle 36 represents the cod spawning grounds that were reported by Bigelow and Schroeder (1953). 37 38 215x166mm (300 x 300 DPI) 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 http://mc.manuscriptcentral.com/icesjms Page 51 of 54 Manuscripts submitted to ICES Journal of Marine Science

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 For Review Only 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 Figure 6. Consensus cod spawning grounds that were identified by at least three fishermen during the 34 interviews. 35 215x166mm (300 x 300 DPI) 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 http://mc.manuscriptcentral.com/icesjms Manuscripts submitted to ICES Journal of Marine Science Page 52 of 54

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 For Review Only 19 20 21 22 23 24 25 26 27 28 29 30 Figure 7. The number of fishermen who reported spawning activity in each month, and the number of 31 spawning sites identified in each month throughout the entire study area. 32 33 87x60mm (300 x 300 DPI) 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 http://mc.manuscriptcentral.com/icesjms Page 53 of 54 Manuscripts submitted to ICES Journal of Marine Science

1 2 3 4 5 6 7 8 9 10 11 12 13 14 Figure 8. Proportion of fishermen who reported spawning activity each month in the different geographic 15 regions of Georges Bank. 16 17 47x9mm (300 x 300 DPI) 18 For Review Only 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 http://mc.manuscriptcentral.com/icesjms Manuscripts submitted to ICES Journal of Marine Science Page 54 of 54

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 For Review Only 19 20 21 22 23 24 25 26 27 Figure 9. Depth range (m) of cod spawning grounds reported by fishermen. 28 29 76x45mm (300 x 300 DPI) 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 http://mc.manuscriptcentral.com/icesjms