Indications of Offshore Spawning by Southern Gulf of Maine Winter Flounder

Abstract

This is the first study to document quantitatively that Gulf of Maine (GOM) Winter Flounder Pseudopleuronectes americanus spawn offshore. Three sites (southern Jeffreys Ledge, Bigear [an area southwest of Tillies Bank], and a portion of Stellwagen Bank National Marine Sanctuary) were sampled during spring 2016 to determine whether Winter Flounder are spawning in non-coastal, deep waters in the southern GOM. In total, 1,384 Winter Flounder were caught by trawl, measured, sexed, and assessed for reproductive stage during the peak spawning season (March–May). These fish showed clear signs that spawning was occurring either at or very near to all three sites surveyed. In all sites, a shift from prespawn to postspawn females occurred. Running ripe females were caught at both Jeffreys Ledge and Stellwagen Bank on multiple occasions, and ripe and recently spawned females were caught at Bigear. Given that these sites are over 15 km from the coast, it is not energetically likely that fish in such advanced reproductive condition are migrating to shore to spawn. Essential fish habitat (EFH) for GOM Winter Flounder is categorized as waters with depths of 5 m or less. The present study clearly shows that this depth range is incomplete and thus inaccurate for southern GOM Winter Flounder. Although it is irrefutable that GOM estuaries and shallow bays are important as nursery grounds for juvenile Winter Flounder and as postspawn feeding grounds for adults, the current GOM EFH designation warrants reclassification.

Received May 5, 2017; accepted August 2, 2017

The Winter Flounder Pseudopleuronectes americanus is a commercially and recreationally important flatfish that ranges from Georgia, USA, to Labrador, Canada (Scott and Scott 1988), but is most common from New Jersey to Nova Scotia (Perlmutter 1947). In U.S. waters, Winter Flounder are managed as two coastal stocks (Gulf of Maine [GOM] and southern New England/Mid-Atlantic [SNE/MA]) and one offshore stock (Georges Bank; Pereira et al. 1999). Although the biology and ecology of Winter Flounder have been well studied (see reviews by Able and Fahay 1998; Pereira et al. 1999; Klein-MacPhee 2002), unresolved questions remain concerning their spawning movements. Specifically, GOM Winter Flounder may exhibit more plasticity in spawning patterns and locations than previously thought.

The spawning season for Winter Flounder typically lasts for 2–3 months in the late winter and spring, depending on location (DeCelles and Cadrin 2011). In the SNE/MA stock, reproductively isolated adult populations generally undergo inshore migrations in early spring into specific estuaries or coastal embayments, where spawning occurs (Lobell 1939; Perlmutter 1947; Saila 1961; Howe and Coates 1975; Phelan 1992) and where the young of the year remain for their first 2 years before moving offshore (Pereira et al. 1999). Although the occurrence of ripe females on the continental shelf of the New York Bight indicates that some SNE/MA populations may remain in deeper waters to spawn (Wuenschel et al. 2009), it is the estuaries and inshore areas that are recognized as essential fish habitat (EFH) for Winter Flounder (Pereira et al. 1999).

For GOM Winter Flounder, EFH is classified the same as for the SNE/MA stock, with the assumption that spawning patterns are similar, and until recently this was undisputed. Relatively few studies have examined adult Winter Flounder movements in the GOM through tagging, and most of these did not examine spawning behavior (Perlmutter 1947; McCracken 1963; Howe and Coates 1975). It is accepted that GOM Winter Flounder populations undergo shorter migrations than SNE/MA populations, which would favor discrete spawning populations, and water temperature plays an important role in governing seasonal movements (McCracken 1963; Howe and Coates 1975; Van Guelpen and Davis 1979; Hanson and Courtenay 1996). However, recently two studies questioned whether GOM Winter Flounder actually spawn in estuaries (DeCelles and Cadrin 2010; Fairchild et al. 2013). DeCelles and Cadrin (2010) used acoustic telemetry to monitor 72 adult fish tagged during the spawning season in the Plymouth Bay and Estuary, Massachusetts, during 2007–2009. Two groups of Winter Flounder were identified based on location (bay or estuary) during the spawning season (March–May). A small group migrated into the estuary, but the majority (76%) of tagged Winter Flounder remained in the coastal bay waters (DeCelles and Cadrin 2010). Fairchild et al. (2013) examined the spawning movements of adult Winter Flounder in Ipswich Bay, a 66-m-deep offshore area in the southwestern GOM. A total of 357 Winter Flounder collected by bottom trawl in the offshore study area were tagged with conventional tags in March–June 2009, examined to quantify reproductive status, and released. Fairchild et al. (2013) found that in Ipswich Bay, the Winter Flounder spawning season begins in early March and continues through May and into early June. Peak spawning occurs in late April and early May. Using telemetry, fine-scale movements of 40 ripe adult Winter Flounder tagged with acoustic transmitters were tracked during April–August 2009 (Fairchild et al. 2013). Only six fish (16%) were detected as entering estuaries between the end of April and August, indicating that the majority of the tagged fish did not spawn in estuaries but remained in deeper, coastal waters. Forty-two percent (n = 16) of the fish were never found in waters shallower than 44 m; many of these fish were never detected as moving inshore either. By mid-May, adults that were present in estuaries had already spawned and were feeding.

These telemetry studies show that Winter Flounder in the southern GOM are not estuarine dependent but employ several different reproductive strategies. The Ipswich Bay study (Fairchild et al. 2013) provided detailed depth ranges occupied by a population of spawning Winter Flounder and clearly identified that—contrary to established paradigms—some GOM Winter Flounder populations spawn in deep, offshore areas. However, these offshore areas have not been well documented, although there is some evidence of their locations. First, commercial fishermen have reported (C. Bouchard, FV Stormy Weather; J. Ford, FV Lisa Ann III; and D. Goethel, FV Ellen Diane, personal communications) and University of New Hampshire (UNH) scientists have noted (E. A. Fairchild, unpublished data; H. Howell, personal communication) that spawning Winter Flounder are found in offshore locations. From 1995 to 2012, ripe and running ripe Winter Flounder were caught each March in the southern part of Jeffreys Ledge for UNH research projects (E. A. Fairchild, unpublished data). Contrary to what is reported in the scientific literature, the commercial fishermen suspected that GOM Winter Flounder are spawning in deeper waters (30–50 km) offshore, where the fishermen categorized the substrates as including gravel, pea-stone, sand, and mud. Second, the GOM Industry-Based Survey (Hoffman et al. 2012) indicated that the largest aggregations of adult Winter Flounder caught during the prespawning months are not found near the mouths of estuaries but rather in locations further offshore (Fairchild et al. 2013).

Based on these observations and prior studies (DeCelles and Cadrin 2010; Fairchild et al. 2013), it is clear that GOM Winter Flounder are not limited to spawning in estuarine and near-coastal areas but also spawn offshore. However, neither documentation of offshore locations nor quantitative assessment of Winter Flounder in those locations has been performed. Therefore, the purpose of this study was to sample the adult population of Winter Flounder in three offshore study sites during the spawning season to quantify how the relative abundance and reproductive status of these fish changed over the spawning season in each of these areas.

STUDY AREAS

In collaboration with commercial fishermen, I identified three areas in the southwestern GOM where adult Winter Flounder are located during the spawning season (Figure 1): Bigear, the southwest corner of Stellwagen Bank, and southeastern Jeffreys Ledge.

FIGURE 1. Relative abundance of mature (>28-cm TL) Winter Flounder captured in seventy-one 30-min tows from March 14 to May 15, 2016, in three offshore areas within the southern Gulf of Maine: southeastern Jeffreys Ledge (blue), Bigear (green), and Stellwagen Bank (red). Circle sizes are proportional to the Winter Flounder catch.

Bigear

Bigear is a small, approximately 1-km² area southwest of Tillies Bank, approximately 22 km from Cape Ann, Massachusetts, with its northwest corner at 42.30°N, 70.24°W and its southeast corner at 42.26.5°N, 70.17°W. Depth ranges from 70 to 80 m, and the substrate is muddy. According to industry logbooks, high abundances of Winter Flounder have been caught here from late February through March (J. Ford, FV Lisa Ann III; and C. Bouchard, FV Stormy Weather, personal communications). Due to the GOM groundfish closures to protect spawning populations of Atlantic Cod Gadus morhua, no fishing has occurred in this area during April–May since 1999 (U.S. Office of the Federal Register 2017), but when the area reopens in June, Winter Flounder are no longer present. This is one of the areas where fishermen repeatedly have collected prespawn, spawning, and spawned-out Winter Flounder, which have been used for broodstock in UNH aquaculture studies.

Southwest Corner of Stellwagen Bank

The southwest corner of Stellwagen Bank consists of an approximately 40-km², hard-sand area ranging in depth from 27 to 40 m within the Stellwagen Bank National Marine Sanctuary in Massachusetts Bay. The area is located about 22 km from Cape Cod and 44 km from mainland Massachusetts. According to the fishermen, their catches are dominated by Winter Flounder in late May through June (end of the spawning season). Fishermen suspect that Winter Flounder are present in the area during the months prior, but they have not had access to fishing there during earlier months.

Southeastern Jeffreys Ledge

The study area on the eastern edge of southern Jeffreys Ledge is a narrow, approximately 40-km², sandy-bottom stretch along the 35-fathom (64-m) contour, situated about 15 km from Cape Ann, Massachusetts. Like Bigear, high abundances of Winter Flounder have been caught here for many years starting in late February and extending through March (C. Bouchard, FV Stormy Weather; and D. Goethel, FV Ellen Diane, personal communications). This area also is closed to groundfishing in April–May (again similar to Bigear), and when the area reopens in June (U.S. Office of the Federal Register 2017), Winter Flounder are no longer present. This too is one of the areas where fishermen repeatedly have collected prespawn, spawning, and postspawn Winter Flounder that are used for broodstock in UNH aquaculture studies.

METHODS

To determine whether and when Winter Flounder spawning occurs in the three offshore areas, trawl surveys were conducted over a 2-month period during the Winter Flounder spawning season. Sampling of the three sites generally was conducted biweekly from March 17 to May 19, 2016, but some temporal variation occurred due to logistics (i.e., two vessels for three sites, weather, and staffing; Table 1). Both Jeffreys Ledge and Bigear were sampled five times, while Stellwagen Bank was sampled four times during the study. On each sampling trip, short (30-min) tows using standard legal groundfish gear (15.2-cm body, 16.5-cm cod-end mesh size) were deployed from two contracted commercial fishing vessels. Except for one six-tow Stellwagen Bank trip (on April 6), five tows occurred on each sampling trip, resulting in a total of 71 tows. Because of the limited towable ground in these areas, surveys did not follow a stratified random sampling design but instead were directed based on fishers’ knowledge and the available towable area at specified depths. Tows were offset as much as possible on each trip.

TABLE 1. Sampling site characteristics and Winter Flounder catch composition, including the number of females assigned to various reproductive stages; the total number of females, males, and fish of unknown sex; and the total number of individuals that were sacrificed for examination of gonads.

						Reproductive stage of females							Unknown sex
Site	Date	Average depth range (m)	Bottom temperature range (°C)	Bottom type^a	Total Winter Flounder	Developing	Ripe	Running ripe	Spent	Recovering	Total females	Males	Immature	Mature	Total sacrificed
Jeffreys Ledge	Mar 20	58–70		Sand	83	31	25	2	15	0	73	8	0	2	6
	Apr 6	55–69		Sand	74	26	17	0	16	1	62^c	2	5	5	2
	Apr 21	53–68		Sand	71	22	17	1	21	0	61	6	0	4	1
	May 7	55–68		Sand	23	0	10	0	6	3	19	2	1	0	5
	May 19	55–64		Sand	6	0	0	0	2	1	3	2	0	1	2
Bigear	Mar 17	91–183	6.3–6.4	Mud/clay	69	38	22	0	1	0	62^b	3	0	4	11
	Mar 30	121–144	5.5–6.5	Mud	160	116	21	0	8	1	146	11	1	2	9
	Apr 13	113–128	6.1–6.4	Mud	152	106	25	0	5	0	136	7	2	7	4
	Apr 25	110–139	6.4–6.6	Mud	40	16	12	0	7	0	35	2	0	3	1
	May 4	91–119	5.7–6.2	Mud	30	3	0	0	17	3	23	2	1	4	2
Stellwagen Bank	Apr 6	32–83		Sand	176	0	132	1	23	0	156	13	0	7	0
	Apr 22	30–41	5.4–6.4	Sand	166	0	33	0	97	0	130	9	1	26	0
	May 3	33–82	5.9–6.4	Sand	147	5	8	0	55	34	103^b	6	10	28	2
	May 10	27–37	6.0–6.9	Sand	187	3	0	1	24	6	36^c	1	6	144	2
All					1,384	366	322	5	297	49	1,045	74	27	237	47

^a As classified by vessel captains.

^b The reproductive status of one female was not evaluated.

^c The reproductive status of two females was not evaluated.

For each tow, the following information was recorded: starting and ending times and positions, tow speed, depth range, estimated mean depth, bottom temperature (available only for two sites), bottom substrate type, and catch composition (species and amount). All nonmarketable bycatch was released alive immediately after each tow; all marketable bycatch was landed and sold by the captains. All captured Winter Flounder were placed in flowing seawater tanks onboard the vessels and were measured (TL), tagged with conventional external tags (either 2-cm disc tags or 5-cm T-bar tags), and examined to determine sex and reproductive condition if possible; the majority of Winter Flounder were then released alive.

Sex and reproductive staging criteria

For external sexing of Winter Flounder, males were identified only if they extruded milt (i.e., were spermiating). Females were classified as developing (showing a small, distinctive bulge in the abdominal cavity posterior to the genital pore), ripe (exhibiting a large, elongate, distinctive bulge in the abdominal cavity posterior to the genital pore), running ripe (extruding eggs), spent (showing a conspicuous concave depression in the gonadal area), or recovering (resting) per the methodology of Fairchild et al. (2013) and McBride et al. (2013).

If a fish was less than 28 cm TL and its sex was not identifiable externally, it was classified as an immature individual of unknown sex based on Winter Flounder maturity guidelines from the nearby area, Ipswich Bay. This size was based on the average of the two mean lengths at 50% maturity for males (27 cm) and females (29.7 cm) from past surveys (1982–2007) of the inshore western GOM stock (O’Brien et al. 1993; Fairchild et al. 2013; P. Nitschke, National Marine Fisheries Service [NMFS], Northeast Fisheries Science Center, unpublished data). Immature Winter Flounder were tagged and released alive.

Winter Flounder larger than 28 cm TL and for which sex was not identifiable externally were classified as mature individuals of unknown sex. Up to six fish from each tow were sacrificed, and whole gonads were photographed and inspected macroscopically to identify sex and reproductive stage. Females in this group also were examined for evidence of skipped spawning, characterized by opaque, thick ovarian walls (McBride et al. 2013).

Statistical analysis

Sampling trips were grouped biweekly, starting with March 13–26 as week 2 and ending with May 8–21 as week 10. Nonparametric paired Wilcoxon tests (JMP version 12.2) were used to examine temporal changes in CPUE within each study area. All differences were considered significant at P-values less than 0.05. Spatial comparisons (i.e., between sites) were not possible due to differences in fishing vessels and gear deployment techniques.

RESULTS

Overall, 1,384 Winter Flounder were caught via trawling from March 14 to May 15, 2016 (Table 1; Figure 1). The CPUE of Winter Flounder varied temporally (Figure 2). At Jeffreys Ledge, mean Winter Flounder CPUE, reported hereafter as the mean (±SD) number of fish caught per 30-min tow, ranged from 1.2 ± 1.3 to 16.6 ± 8.4. Winter flounder catch at Jeffreys Ledge was similar in mid-March through mid-April (mean CPUE = 14.2 ± 4.5 to 16.6 ± 8.4; P ≥ 0.672), but then decreased significantly (P = 0.01) from 14.2 ± 4.5 in mid-April to 4.6 ± 0.9 in early May and 1.2 ± 1.3 in late May (Figure 2). At Stellwagen Bank, mean CPUE of Winter Flounder remained constant (P ≥ 0.52) across all four sampling trips, ranging from 29.3 ± 9.0 to 37.4 ± 23.4. Winter Flounder catches at Bigear exhibited a bell-shaped curve in that catches were significantly lower (P ≤ 0.04) at the beginning (13.8 ± 4.9) and end (7.0 ± 6.1) of the study period but reached a peak in late March (32.08 ± 8.74) to early April (30.4 ± 10.6).

FIGURE 2. Mean CPUE (+SD) of Winter Flounder in three offshore areas within the southern Gulf of Maine during the 2016 spawning season. Tow duration was 30 min. Within a site, columns labeled with different letters are significantly different (P < 0.05). Corresponding survey data appear in Table 1.

With the exception of the last sampling trip at Stellwagen Bank (May 10), Winter Flounder catches were dominated by mature females, regardless of the date or site (Figure 3). Generally, as time progressed during the spawning season, the proportion of Winter Flounder for which sex was externally unidentifiable (i.e., unknown) increased at all locations; the percentage of mature, unknown-sex individuals ranged from 0% to 17% among fish caught at Jeffreys Ledge, from 1% to 13% at Bigear, and from 4% to 77% at Stellwagen Bank (Table 1). Male Winter Flounder were caught on every sampling trip at all sites but were captured in very low quantities on average; CPUE per tow was 1.0 ± 1.3 for males compared to 14.7 ± 11.5 for females (Table 1).

FIGURE 3. Sex composition (green = unknown; blue = male; purple = female) of Winter Flounder caught biweekly from March 14 to May 15, 2016, in three offshore areas within the southern Gulf of Maine: Jeffreys Ledge (left panels), Bigear (middle panels), and Stellwagen Bank (right panels). Top panels show raw data, while bottom panels show data adjusted for classification accuracy. The unknown category includes both immature and mature fish. Numbers within gray bars represent the total number of Winter Flounder that were assessed each week.

The proportion of prespawn to postspawn females changed over time, with more prespawn females caught during the first half of the study and more postspawn females caught during the second half (Figure 4; Table 1). Furthermore, during May, a growing percentage of females caught at each site was classified as recovering (Jeffreys Ledge: 18%; Bigear: 13%; Stellwagen Bank: 29%; Table 1). Only five running ripe females were caught: three at Jeffreys Ledge (weeks 2 and 6) and two at Stellwagen Bank (weeks 4 and 10; Figure 4; Table 1).

FIGURE 4. Reproductive stages of female Winter Flounder (adjusted for classification accuracy) caught from March 14 to May 15, 2016, in three offshore areas within the southern Gulf of Maine: Jeffreys Ledge (top panel), Bigear (middle panel), and Stellwagen Bank (bottom panel). Catch data are grouped into biweekly periods. Numbers above biweekly data represent the total number of female Winter Flounder that were assessed each week.

Nineteen percent (n = 264) of all Winter Flounder caught were not externally identifiable to sex, with most deemed mature (>28 cm TL; n = 237). To verify classification (sex and reproductive stage) accuracy and to identify those individuals with the “mature, unknown sex,” designation, 47 fish were sacrificed for gonad inspection. These fish included non-spermiating males, fish that had already spawned and physically recovered, and all other reproductive stages except running ripe. Ninety-six percent (96%; n = 25) of fish that were not in the mature, unknown-sex category were identified to sex correctly. There was only one case of a fish being misidentified as a female, and no fish were misidentified as males. Of the fish that were classified as mature individuals of unknown sex (47%; n = 22), the majority was female (82%; n = 18). The observed reproductive stage of all sacrificed females (n = 42) was compared to the predicted reproductive stage (Table 2). Our predictive capabilities were very robust (100% accuracy, but sample sizes were small) for all stages except developing females, in which 20% were misclassified as either ripe or recovering. Adjusting for classification errors in sex composition (Figure 3) and reproductive staging (Table 2) did not substantially alter our interpretation of Winter Flounder population dynamics in these areas. From these adjustments was verification that Winter Flounder deemed “mature, unknown sex” in May samples from Stellwagen Bank were predominantly resting females. Skipped spawning in females was not observed in this study.

TABLE 2. Classification accuracy of sexing and reproductive condition staging for 42 female Winter Flounder larger than 28 cm TL.

	Observed reproductive stage
Predicted reproductive stage	Immature	Developing	Ripe	Running ripe	Recovering	Spent	Misclassified (%)
Immature	0	0	0	0	0	0	0
Developing	0	12	0	0	0	0	0
Ripe	0	1	4	0	0	0	20
Running ripe	0	0	0	0	0	0	0
Recovering	0	2	0	0	5	0	29
Spent	0	0	0	0	0	7	0
Unknown	1	6	1	0	1	2

As of September 2017, 1.9% (26 of 1,337 fish) of all tagged and released Winter Flounder had been recaptured and reported by fishermen, with complete recapture information provided for 23 fish. In the months after the 2016 spawning season (June–January 2016), 10 fish were recaptured. The majority (7 of 10 fish), which included individuals from each of the three study sites, had moved substantially closer to shore. Winter Flounder from Bigear (n = 4) and Stellwagen Bank (n = 2) were recaptured during the following spawning season (spring 2017) near their original tagging locations.

DISCUSSION

Winter Flounder caught during March–May 2016 showed clear signs that spawning was occurring either at or very near to all three sites surveyed. Running ripe females were caught at both Jeffreys Ledge and Stellwagen Bank on multiple occasions; at Bigear, ripe and recently spawned females were caught. In addition, more prespawn females were caught earlier in the sampling period, and more postspawn females were caught later. For a female Winter Flounder from Jeffreys Ledge, Bigear, or Stellwagen Bank to spawn by the shore, she would have to swim approximately 15, 22, or 22 km, respectively (as measured from the midpoints of the areas sampled to the closest shore point). It is unlikely that fish in such an advanced reproductive condition are migrating those distances to spawn. Although the Winter Flounder spawning season lasts several months and although at the population level, females with ovulated oocytes are observed over 2 months (Press et al. 2014), individual movements of running ripe fish are unknown. However, at least in the laboratory, individuals will release the vast majority of their eggs relatively quickly—within 1–2 min (E. A. Fairchild, personal observation). Given that female Winter Flounder cease feeding during the spawning season, active swimming must be energetically costly. He (2003) observed that the in situ adult Winter Flounder swimming speed associated with feeding was 0.95 body lengths (BL)/s at 4.4°C, and speeds decreased with colder temperatures. Joaquim et al. (2004) measured the maximum cardiac output levels of Winter Flounder in a flume tank at 0.65 BL/s (temperature = 4°C) and 0.73 BL/s (10°C); observed that the fish were less stressed when swimming slower (<0.4 BL/s), regardless of temperature; and suggested that cardiac outputs have been overestimated in flatfishes. In this study, bottom temperature ranged from 5.4°C to 6.9°C, and the mean size of captured Winter Flounder was 36 cm TL (Table 1). Using the parameters from He (2003) and Joaquim et al. (2004), it would take 19.4–48.6 h of sustained swimming for a fish to reach shore from Bigear and 28.5–71.3 h for a fish to reach shore from Jeffreys Ledge or Stellwagen Bank. These estimates are based on a straight-line distance with no resting periods, and they do not consider currents, tides, or diel activity patterns, thus making the estimates physiologically improbable. However, these times and the energetic costs to the fish could be reduced if, like many other flatfishes (e.g., European Flounder Platichthys flesus: Wirjoatmodjo and Pitcher 1984; Plaice Pleuronectes platessa: Cushing 1990; Summer Flounder Paralichthys dentatus: Sackett et al. 2007), Winter Flounder utilize selective tidal transport to facilitate long-distance movements. As of now, this has not been described for Winter Flounder.

In the present study, 93.4% of all Winter Flounder that were externally identifiable to sex were females (1,045 females of 1,119 fish). Similar highly female-skewed sex ratios have been observed for many populations of both GOM and SNE/MA Winter Flounder (Table 3). Because age-5 and older female Winter Flounder are larger and faster growing than males (Klein-MacPhee 2002), one could naturally surmise that the observed sex ratio in this study was an artifact of trawl gear selecting for the larger females. However, this trend also was observed in the spring bottom trawl surveys conducted by the NMFS Northeast Fisheries Science Center, which uses trawls with a 1.3- or 2.5-cm-mesh belly and cod end (ESB 2017). During 2000–2016, in areas and months overlapping this study, 63.4% of all Winter Flounder that were caught and sexed in the spring bottom trawl survey were females (1.74:1 ratio; total n = 424 fish; Table 3; ESB 2017). Rather than gear selectivity, skewed sex ratios may reflect a higher mortality rate among male Winter Flounder (Witherell and Burnett 1993), although the causes for and implications of this difference are poorly understood. In this study, the presence of so many female Winter Flounder during the spring lends further credence to the idea that the sampled sites and the immediately surrounding areas are Winter Flounder spawning grounds.

TABLE 3. Incidence of female-skewed populations of Winter Flounder. Calculations are based on external observations of mature fish (i.e., nonspawning fish are not included), except for Bejda and Phelan (1998), Wuenschel et al. (2009), and the National Marine Fisheries Service spring trawl survey data (ESB 2017). Table is modified from Fairchild (2012).

U.S. location	Female : male ratio	Months	Year	Reference
Gulf of Maine
Ipswich Bay	4.6	Feb–Jun	2007	Fairchild et al. (2013)
Stellwagen Bank/Jeffreys Ledge/Bigear	14.1	Mar–May	2016	Present study
Stellwagen Bank	5.5	Mar	2012	E. A. Fairchild, unpublished data
Stellwagen Bank	1.7	Apr–Jun	2000–2016	ESB (2017)
Southern New England/Mid-Atlantic
South of Cape Cod^a	2.3	Dec–Apr	1960–1965	Howe and Coates (1975)
South and east of Cape Cod	1.3	May	1983–1991	Witherell and Burnett (1993)
Nantucket Sound	4.1	Mar–Apr	2012	E. A. Fairchild, unpublished data
Rhode Island Sound, Rhode Island	2.3	Nov–Apr	1956–1958	Saila (1961)
Narragansett Bay, Rhode Island	1.5	Feb–Mar	1958	Saila (1962)
Niantic River, Connecticut	2.5	Feb–Apr	1977–2015	DNC (2016)
New York Harbor	1.7	Jan–Mar	2002–2010	Wilber et al. (2013)
Raritan Bay and Sandy Hook, New Jersey	2.7	Mar–Sep	1989	Bejda and Phelan (1998)
Navesink Estuary, New Jersey	1.5	Feb–Apr	1997	Stoner et al. (1999)
New York Bight, New Jersey	1.8	Jan–Oct	2007	Wuenschel et al. (2009)

^aIncludes Cape Cod to Buzzards Bay and south to Nantucket Sound and Shoals and Vineyard Sound.

Although running ripe females were caught at Jeffreys Ledge, the relative abundance of Winter Flounder there decreased toward May, indicating that although fish may be spawning on southern Jeffreys Ledge, they do not remain there after spawning and are moving elsewhere. At Bigear, where catches were predominantly prespawn, developing females, Winter Flounder abundance dropped significantly by early May, suggesting that the fish move through Bigear on the way to a nearby spawning area. Industry partners believe that Winter Flounder are utilizing an adjacent, structurally complex area for spawning, but this cannot be validated with trawl gear because tows cannot be conducted in that area. At Stellwagen Bank, the relative abundance of Winter Flounder remained constant and high, and all reproductive stages were caught; it appears that Winter Flounder spawn and remain there. In addition, it seems that the spawning season occurs earlier and lasts longer at Stellwagen Bank than at the other, more northerly, sites. By early April, developing females were not caught at Stellwagen Bank, whereas to the north and east, they were the most frequently caught female reproductive stage (Figure 4). By mid-May, a small but growing percentage of females caught on Stellwagen Bank were developing, whereas this group was in decline at Jeffreys Ledge and Bigear, indicating that on Stellwagen Bank, the spawning season continued over a longer period (Table 1; Figure 4).

From the few fish (n = 10) that were recaptured in the months (June–January) after the 2016 spawning season, there is some evidence that Winter Flounder migrate inshore after spawning, presumably to a feeding ground. Seven of 10 recaptured Winter Flounder, including individuals from each of the three study sites, moved substantially closer to shore during June–October. If recapture locations indicate general movement patterns, then all fish moved west–southwest in Massachusetts Bay rather than toward the closest shores—Cape Ann or the tip of Cape Cod. The remaining three recaptured fish originated from Stellwagen Bank and were still there when recaptured in August. Such inshore movements also were observed for the Winter Flounder population in nearby Ipswich Bay; after spawning in deeper coastal waters, adults moved inshore, presumably to feed (Fairchild et al. 2013). Gut content analyses of inshore adults during months and at locations where acoustically tagged Winter Flounder were present revealed that postspawn fish were feeding on soft-shell clam Mya arenaria siphons and on American Sand Lance Ammodytes americanus (Fairchild et al. 2013).

Based on advances in tagging technology, divergent movement patterns of Winter Flounder populations during the spawning season are apparent, with some GOM Winter Flounder populations clearly neither estuarine dependent nor necessarily coastally dependent for spawning (DeCelles and Cadrin 2010; Fairchild et al. 2013). Furthermore, although highly variable depending on the year and hatching date, modeling of Winter Flounder larval drift over a 5-year period, with early, mid-, and late-spawning groups, showed that Winter Flounder spawned and hatched from Stellwagen Bank, southern Jeffreys Ledge, and Ipswich Bay would contribute to estuarine juvenile populations (DeCelles et al. 2015). Unlike the conventional belief that Winter Flounder larvae found within 25 km of shore are a result of outflowing from natal estuaries (Smith et al. 1975), the inverse may be more apt for southern GOM Winter Flounder populations; these larvae originate from offshore spawning grounds, and a proportion of the larvae is advected into estuaries.

The SNE/MA Winter Flounder exhibit spawning site fidelity, returning to the same spawning areas each year. Although no studies have focused directly on this phenomenon, mature Winter Flounder tagged during their spawning season have been recaptured in subsequent years in the same locations (Phelan 1992; DNC 2016). Movements of tagged Winter Flounder (206 recaptures of 7,346 tagged fish; 3%) in the inner New York Bight showed separate populations with homing movements to distinct inshore locations (Phelan 1992). In the Niantic River, Connecticut, a total of 41,069 Winter Flounder larger than 20 cm TL have been tagged and released each spring since 1983 (DNC 2016). As of 2015, the recapture rate has been 5.4% (n = 2,208 fish; DNC 2016). In the GOM, 4 (1%) of 395 mature Winter Flounder that were originally tagged in Ipswich Bay during the spawning season were recaptured at the original release location in the following year, potentially suggesting that Winter Flounder also return to the same spawning areas each year north of Cape Cod (Fairchild et al. 2013). The current study follows suit; during the subsequent spawning season (spring 2017), 6 of the 1,337 tagged and released Winter Flounder (4 tagged at Bigear and 2 tagged at Stellwagen Bank) were recaptured close to their original capture locations. These returns possibly are indicative of spawning site fidelity, but to conclusively prove this pattern, a dedicated study is needed in which a significantly greater return rate is documented. We can look to studies of other flatfishes for guidance, specifically studies of European Plaice. Similar to Winter Flounder, Plaice migrate between discrete feeding and offshore spawning locations, returning to the same sites by the same routes each year (Hunter et al. 2003; Solmundsson et al. 2005). In Icelandic waters, these patterns were documented by using conventional tags on relatively few fish (~700), with return data reported by the commercial industry; recapture rates were very high (74%; Solmundsson et al. 2005). Similarly, in the North Sea, 46 (39%) of 117 female Plaice that were fitted with data storage tags were recaptured by the industry (Hunter et al. 2003). Unlike the northern European fleet, the U.S. Northeast groundfish fleet has declined dramatically. Since 2000, there has been a 77% reduction in the number of gill-net and bottom trawl vessels in the greater GOM (Maine, New Hampshire, and Massachusetts) that land Winter Flounder (GARFO 2017). Even if more Winter Flounder can be tagged, the expected return rate would not significantly increase due to the lack of possible industry effort. In a study of Yellowtail Flounder Limanda ferruginea during 2003–2006, only an 8% return rate was achieved, despite the tagging of 45,000 fish (Wood and Cadrin 2013). To ensure a higher return rate on tagged Winter Flounder and thus verify spawning site fidelity, a dedicated effort to recapture fish in the ensuing years and the use of data storage tags with high monetary incentives for return are recommended.

In summary, this was the first study to document quantitatively that GOM Winter Flounder spawn offshore. In addition, one of the offshore sites studied, Stellwagen Bank, appears to support a robust Winter Flounder population that remains after spawning rather than moving inshore to feed. This is not entirely surprising, as the habitat on Stellwagen Bank where sampling occurred is sandy, much shallower than either southern Jeffreys Ledge or Bigear, and situated furthest from the coast. In many ways, Stellwagen Bank is more typical of the habitats that are considered traditional inshore Winter Flounder spawning areas.

Identifying and studying where and when Winter Flounder spawn are important to ensure proper management of the resource by protecting both spawning fish populations and their essential habitat. Although this study has clearly shown that Winter Flounder in the southern GOM do spawn offshore, finer-scale movements around the spawning grounds remain less clear. Depending on these patterns and whether there is solid evidence that Winter Flounder show spawning site fidelity, managers may want to consider time or area closures to protect spawning populations if the GOM stock biomass declines. The EFH for spawning Winter Flounder is classified based on where the demersal eggs occur, and this is identical for both the GOM and SNE/MA stocks: in waters with depths of 5 m or less (Pereira et al. 1999). Clearly, this depth range is incomplete for southern GOM Winter Flounder. Harder to discern is how much deeper this designation should extend into offshore waters and whether the depth range should be truncated if Winter Flounder are not spawning in estuaries (anymore). In the northeastern U.S. waters, Winter Flounder EFH designations carry great economic import; time-of-year restrictions prohibit activities such as dredging (for navigation and marine infrastructure development) during certain months in waters of 5-m or lesser depths to protect spawning fish and egg masses. Although it is irrefutable that estuaries and shallow bays are important nursery grounds for juvenile Winter Flounder and serve as important postspawn feeding grounds for adults, the current GOM EFH designation does not include the full depth range where offshore-spawning fish occur. Future studies should focus on definitively demonstrating spawning site fidelity, mapping finer-scale movements of spawning fish in the offshore areas (including whether selective tidal transport is employed), and determining current estuarine use by adults. Similar studies in other areas of the GOM would help to determine whether these observed patterns persist throughout the entire range of the stock or are only prevalent in the south.

ACKNOWLEDGMENTS

Special thanks are due to industry collaborators, Captains Jim Ford, David Goethel, and Norman Pike, without whose expertise this study would not have been possible. Numerous students assisted in this research, including E. Berghahn, K. Butler, E. Groover, S. Mace, M. Owings, J. Rackovan, T. Rodenbaugh, S. Sykes, and J. Taylor. Scientific support was provided by Hunt Howell and Rich McBride. I am grateful to S. Dwyer (Dominion Nuclear Connecticut), P. Kostovick (National Oceanic and Atmospheric Administration [NOAA]), and K. McGrath (NOAA) for providing supporting data and to the two anonymous reviewers who provided constructive criticism. Funding for this research was provided by the New England Fisheries Management Council through an award to the Northeast Consortium.