The spatial and seasonal distribution of walleye were surveyed with gillnets in the Bay of Quinte and Lake Ontario during 1992–2008. Walleye movements were determined with tagging in the Bay of Quinte and Lake Ontario surrounding Prince Edward County during fall 1998–2003, and with recaptures from angling and other fisheries in Lake Ontario and the St. Lawrence River during 1998–2003. Immature walleye (age <4 yr) were abundant in the upper Bay of Quinte from April to November, moved small distances down the bay during summer, and were less frequently observed in Lake Ontario. Mature walleye (age >4 yr) were found in the upper Bay of Quinte during spring, and farther down the bay during summer. Older mature walleye (age-7+) were observed in eastern Lake Ontario during summer. The tagging data were consistent in showing that older fish moved down the Bay of Quinte toward Lake Ontario during summer. During fall mature walleye moved back up the Bay of Quinte either from Lake Ontario or the lower bay. Immature and mature walleye moved farther up the Bay of Quinte during fall to spring, again with older walleye tending to move longer distances. Walleye migration between the Bay of Quinte and Lake Ontario during spring and fall was consistent with avoiding warm temperature in the upper bay, and foraging on alewife in the lower bay and Lake Ontario during summer and young-of the-year fishes such as gizzard shad during fall. The distribution of walleye between the Bay of Quinte and Lake Ontario did not change after dreissenid colonization.
Walleye (Sander vitreus) is the most abundant piscivorous fish in the Bay of Quinte and the nearshore areas of the eastern half of Lake Ontario. These walleye are largely one genetic stock (Wilson and Mathers, 2003), and migrate between the Bay of Quinte and Lake Ontario (Payne, 1963; Hurley, 1986b; Bowlby et al., 2010). Walleye spawn during April along the shoreline and in the major rivers of the Bay of Quinte. Soon after spawning, adult walleye migrate to Lake Ontario where they reside during summer. Juvenile walleye inhabit the Bay of Quinte year-round. The walleye population in Lake Ontario has experienced declines in abundance (Hurley and Chrsitie, 1977; Hurley, 1986a; Bowlby et al., 1991; Bowlby et al., 2010). Most recently, during the 1990s, walleye abundance declined sharply, following establishment of Dreissenid Mussels in the Bay of Quinte (Hoyle et al., 2008; Bowlby et al., 2010). Dreissenids became abundant in eastern Lake Ontario in 1993 (Dermott, 2001). Dreissenids proliferated and impacted water clarity in the Bay of Quinte by 1995, (Bailey et al., 1999; Dermott, 2001). Increased water clarity resulted in an expansion of aquatic vegetation in the Bay of Quinte (Leisti et al., 2006), leading to increased distributions and abundances of fish species preferring this habitat, especially bluegill (Lepomis macrochirus), black crappie (Pomoxis nigromaculatus), and largemouth bass (Micropterus salmoides) during the late 1990s (Hoyle et al., in press). Chu et al. (2004) modeled walleye habitat in the Bay of Quinte based on water clarity and temperature during 1972–2001. They suggested that increased water clarity in 1995 after dreissenid invasion reduced the quality of habitat for walleye in the upper Bay of Quinte. Walleye were thought to have responded to increased water clarity by moving to the lower Bay of Quinte, eastern Lake Ontario, and the St. Lawrence River (Casselman et al., 1999; Casselman and Scott, 2003).
Hurley (1986b) determined that when temperature exceeded 23.5°C in the Bay of Quinte, growth ceased for age-5 walleye, and for age-1 to age-3 above 25°C. He suggested that the Bay of Quinte was too warm for growth of walleye during summer causing them to migrate to Lake Ontario. Kershner et al. (1999) and Wang et al. (2007) have proposed the same explanation for migration of walleye out of the warmer western basin of Lake Erie during summer, but also suggested that walleye may migrate in response to abundance of soft-rayed prey fish.
In the Bay of Quinte, the walleye diet was predominantly alewife (Alosa pseudoharengus) until the invasion of dreissenids (Hurley, 1986b; Bowlby et al., 1991; Bowlby et al., 2010). Since dreissenids have become established, the diet has become more diverse with alewife, gizzard shad (Dorosoma cepedianum), yellow perch (Perca flavescens), white perch, and Round Goby (Neogobius melanostomus) contributing significantly at times (Bowlby et al., 2010). The Round Goby invaded eastern Lake Ontario in 1999 (Dietrich et al., 2006), and was abundant in Quinte surveys (Hoyle et al., in press) and walleye diets by 2003.
The objectives of this paper are threefold. Firstly, we use gillnet and tagging data to investigate if recent distribution patterns of walleye in the Bay of Quinte and Lake Ontario during 1992–2008 are consistent with the historical trends outlined by Payne (1963), Bowlby et al. (1991), and more recently Bowlby et al. (2010). Secondly, we investigate the contention that walleye moved to Lake Ontario in response to increased water clarity in 1995 (Casselman et al., 1999; Casselman and Scott, 2003) related to the invasion of dreissenids. Thirdly, we consider hypotheses for walleye movements, related to temperature preference (Hurley, 1986b), and to prey availability (Kershner et al., 1999; Wang et al., 2007), specifically the spawning migration of alewife into the Bay of Quinte (Ridgway et al., 1990), the production of young-of-the-year fishes (YOY), and the invasion of Round Gobies about 2000 (Hoyle et al., in press).
The spatial and seasonal distributions of walleye, alewife, and gizzard shad were surveyed with gangs of multiple-mesh gill nets at fixed sites in the Bay of Quinte and in the eastern half Lake Ontario (Figure 1) from 1992 to 2008. We limited our analyses to these years to avoid previous changes in fishing gear and sampling locations (Casselman et al., 1999; Casselman et al., 2002; Casselman and Scott, 2003; Bowlby et al., 2010). The survey was divided into three regions (Bay of Quinte, eastern Lake Ontario, and western Lake Ontario). Three sites were in the Bay of Quinte (upper Bay - UB, middle Bay - MB, and lower Bay - LB). Five sites were in eastern Lake Ontario (Grape Island - GI, Melville Shoal - MS, Flatt Point – FP, Main Duck Island - MD, and Rocky Point - RP). MD was sampled only during 1992–1994 and was discontinued due to similarity in catch with RP. In this analysis MD was grouped with RP. Three sites were in western Lake Ontario (Middle Ground - MG, Brighton - BR, and Wellington - WE). During April to November, 1997–1999, samples were collected in the Bay of Quinte during spring (April 28–June 4) summer (June 23–August 30), and fall (September 29–November 5 - excluding 1999). Sampling was conducted at all sites during summer, except in 1993 the only sampling in the Bay of Quinte was at UB. A standard gang of monofilament gill nets comprised ten 15.2-m panels with stretched-mesh sizes ranging from 38.1–152.4 mm in 12.5-mm increments with a 1-m gap, arranged in order. During 1992–1994 MG was sampled with only 38.1–76.2-mm meshes, which significantly reduced the catch of older walleye. Sites in Lake Ontario (except MG) and at LB were stratified into five 5-m depth zones ranging from 7.5 (median depth) to 27.5 m. Walleye were rarely caught (2%) in the deepest zones (22.5–27.7 m), which subsequently were excluded from analyses. MG and UB contained only the 7.5-m zone and MB contained both 7.5-m and 12.5-m zones. A total of 7,505 walleye were captured in 1,680 net sets. Walleye were aged with scales from 1992–1995 and otoliths from 1996 to 2008. Most (73%) of the walleye were aged directly from these structures and the remaining with age-length keys. We had previously determined that otolith ages from Lake Ontario Walleye were accurate for all age groups, and scale ages were accurate up to age-6, and so we grouped older ages (7+) for all years.
Walleye movements were determined with tagging throughout the Bay of Quinte and Lake Ontario surrounding Prince Edward County (Figure 1) during fall 1998–2000 and 2002–2003, and with recaptures from angling and other fisheries in Lake Ontario and the St. Lawrence River during 1998–2000, 2002–2003. We captured walleye with trap nets and tagged 9,163 with individually numbered Floy tags. Tagged fish were aged with age-length keys based on otoliths developed from gill net samples (above) in 1998, and from samples collected while tagging in later years. We grouped older ages (12+) for all years, as age-length keys could not effectively discriminate older ages due to asymptotic growth in length of walleye. Fishers reported the tag number, location and date of recapture. The movement distance from mark location to recapture locations was determined with the ruler tool on Google Earth (Google Inc., 2009), along the closest path through Lake Ontario. The direction of movement was determined relative to the flow of the Bay of Quinte. Movement towards the Trent River or “up” the bay was defined as positive, and movement “down” the bay was defined as negative. Results were summarized by season: spring (Mar-May), summer (June–August), fall (September–November), and winter (December–February).
Water temperature was recorded at the location and depth of each gill net sample. The daily mean water temperature of the Bay of Quinte was obtained from the Belleville water treatment plant, as described by Casselman et al. (1999), for 1943–2009 to estimate the mean daily temperature for April to November across years. Hourly water temperature of Lake Ontario was obtained from buoy C45135 at Prince Edward Point, 43.79N 76.87W for 1989–2009 (www.meds-sdmm.dfo-mpo.gc.ca) to estimate the mean daily temperature for April to November. Outliers, apparently related to air exposure of the thermometer, were removed after analysis of the residuals from a fifth order polynomial fit to the temperature-date data.
During summer walleye in Lake Ontario were most abundant at MS (1.6 per net) in eastern Lake Ontario followed by UB (1.2 per net) and LB (1.0 per net) in the upper and lower Bay of Quinte (Figure 2). The catch was less at all other sites (<0.5 per net). The fewest walleye were observed at FP (<0.1 per net), also in eastern Lake Ontario. Sites in western Lake Ontario had the next lowest catch (<0.2 per net). Walleye catch decreased with depth and temperature (Figure 3). This pattern was similar across all regions and 60% or more walleye were caught at 7.5 m, where the mean temperature was warmer (17.5–23.1°C). Temperature was more variable in eastern Lake Ontario than other regions, and the mean temperatures at 7.5 m at FP, GI, MS, and RP were 15.0, 16.8, 18.8, and 19.6°C, respectively, and may have contributed to the low catch of walleye at FP and high catch at MS (Figure 2). Most of the walleye catch in eastern Lake Ontario was comprised of age 7+ (68%), and the catch from age 1 to 6 increased from 2 to 8% (Figure 4). In contrast, age-7+ walleye were 11% of the catch in the Bay of Quinte and the catch of the younger age classes peaked with a mode of 26% at age 3, more typical of a stable age distribution. In western Lake Ontario the age distribution peaked at age 7+ (32%) with a secondary mode at age 2 (24%), showing characteristics of both eastern Lake Ontario and the Bay of Quinte (Figure 4).
Declines in walleye catch were observed during the 1990s in the Bay of Quinte, and similar declines were observed in eastern and western Lake Ontario (Figure 5). In the Bay of Quinte the mean catch of walleye was highest (1.7 per net) during 1992–1996, declined to 0.74 per net during 1997–2000, and declined again to 0.26 per net during 2001–2008. Also, in western Lake Ontario the mean catch of walleye was highest (0.23 per net) during 1992–1996, and declined to 0.06 per net during 1997 to 2008. In eastern Lake Ontario the decline in walleye catch appears to lag behind the Bay of Quinte by 2–3 years (Figure 5), and the mean catch was 0.72 per net during 1992–1999, and 0.36 per net during 2000–2008. The decline of walleye within the Bay of Quinte was more complex across reaches of the bay (Figure 5). The mean catch of walleye increased in the middle bay from 0.17 per net in 1994 to 0.95 per net in 1999, and then declined to a mean catch of 0.30 per net during 2000-2008. In contrast, the mean catch of walleye in the lower bay was highest (2.6 per net) during 1992-1996, and declined to 0.73 per net during 1997 to 2008. The mean catch of walleye in the upper bay declined sharply from 5.1 per net in 1992 to 3.1 per net in 1993, then gradually declined over the next 6 years during which the mean catch was 1.1 per net during 2000–2008.
The catch of age-1 and age-2 walleye increased from spring to fall at all sites in the Bay of Quinte (Figure 6) consistent with increasing catchability in gill nets as they grow (Hamley and Regier, 1973). The catch of younger walleye in the Bay of Quinte peaked at age-3 when they appeared to be fully recruited to the gill nets (Figure 4). Accordingly, interpretation of seasonal changes in the distribution of younger walleye may be more difficult. Catch of age-3 and age-4 walleye declined through the summer and fall in the upper bay, but increased in the middle and lower bay (Figure 6) suggesting a movement down the bay in late summer. Age-5 and age-6 walleye had a similar pattern of increasing catch from spring to fall, and their catch during fall was generally more than double the previous season. During spring high catches of the oldest walleye (age-7+) were observed only in the upper bay, but their pattern of catch was similar to age-5 and age-6 walleye through the remaining seasons at all Quinte sites.
During 1998–2008, 668 walleye tags with date and location of recapture were returned by fishers. Tagged walleye spent an average of 1.8 years in Lake Ontario before being recaptured. Most returns came from the upper Bay of Quinte (380) and middle bay (173). Fewer returns came from the lower bay (41), western Lake Ontario (44), and eastern Lake Ontario and the St. Lawrence River (30). Walleye movements from tag to recapture locations generally increased with age (Figure 7) from an average of 14 km for age-2 walleye to a peak of 43 km for age-11. The direction of movement changed with age and season (Figure 7). During winter age-6, age-9, age-11, and age-12+ were significantly (P < 0.05) farther up the Bay of Quinte relative to tagging location, and other ages had no significant movement. During March-May all ages except 2, 5, and 10 were significantly farther up the Bay of Quinte relative to tagging location. During summer age-11 and age-12+ walleye were significantly farther down the Bay of Quinte relative to tagging location, age-4 walleye were significantly farther up the bay, and other ages had no significant movement. During fall walleye movements were not significantly different than zero for all ages. We observed a progression of recaptures of walleye tagged in western Lake Ontario from the lower bay to upper bay from fall to winter, suggesting that older walleye in western Lake Ontario migrate into the Bay of Quinte through eastern Lake Ontario, rather than via the Murray Canal.
Seasonal gillnet catches during 1997–1999 (Figure 8) showed alewife migrated into the lower Bay of Quinte during summer (76 per net). Alewife migrated into the bay at the beginning of summer as catches in June of 167 per net declined by August to 2 per net. Few alewife were observed in the middle bay (<5 per net) and upper bay (<2.0 per net). In eastern Lake Ontario during summer the mean catch of alewife (86 per net) was similar to the lower bay. More alewife were caught at GI (120 per net) which was closest to the lower bay; fewest were caught at MS (24 per net), and at FP and RP, 96 and 105 alewife per net were caught, respectively. In western Lake Ontario the alewife catch (5.9 per net) was slightly higher than in the Bay of Quinte, and substantially lower than eastern Lake Ontario. YOY gizzard shad were caught only at the upper and middle Bay of Quinte during summer and fall (Figure 8).
The mean daily temperature of the upper Bay of Quinte rises rapidly from 3.0°C at the beginning of April to peak at 24.0°C for a period from July 19 to August 3, and drops rapidly during fall to 2.6°C on November 30 (Figure 9). Lake Ontario warms more slowly than the Bay of Quinte. The mean daily temperature of Lake Ontario is 1.8°C on April 1, and by June 1 is 10.5°C, which is 7.3°C lower than the Bay of Quinte. Lake Ontario temperature peaks at 22.5°C on August 14, almost 4 weeks behind the Bay of Quinte. Lake Ontario cools more slowly than the Bay of Quinte and is 7.2°C on November 30. We have defined the preferred temperature zone for walleye as 20.6–22.7°C, based on the 95% CI of the mean (21.6°C) of six preferred temperature studies ranging from 20–23.2°C (Wismer and Christie, 1987). The mean daily temperature in the Bay of Quinte reached preferred temperatures on June 15 for 19 days until the temperature rose higher, and then dropped into the preferred zone on August 21 for 21 days. Lake Ontario reached the preferred temperature zone on July 22 for 48 days (Figure 9). The mean daily temperature in the upper Bay of Quinte reached or exceeded 23.5°C on July 14 for 30 days (Figure 9) when growth of age-5 walleye ceased (Hurley, 1986b). Moreover, the temperature of the Bay of Quinte reached or exceeded 25°C at least one day in 76% of the years from 1943 to 2008, and growth of age-1 to age-3 ceased (Hurley, 1986b).
Complementary gillnet and tagging data indicated walleye distribution and movement in the Bay of Quinte and Lake Ontario were consistent with previous observations of walleye migratory behaviour in Lake Ontario (Payne, 1963; Hurley, 1986b; Bowlby et al., 2010) and Lake Erie (Kershner et al., 1999; Wang et al., 2007). Gillnet data indicated that mature walleye are found in the Bay of Quinte during spring and that older walleye (age-7+) are found in eastern Lake Ontario during summer. The tagging data were consistent with this pattern showing that older fish moved farther down the Bay of Quinte toward Lake Ontario by summer. During fall mature walleye moved back up the Bay of Quinte either from Lake Ontario or from farther down the bay. Immature and mature walleye moved farther up the Bay of Quinte from fall to spring, again with older walleye tending to move longer distances. The Bay of Quinte, like the western basin of Lake Erie, provides the major spawning and nursery habitats for walleye; older fish migrate to cooler habitats during summer and return in fall.
Walleye spawn in the Bay of Quinte and its tributaries during April (Payne, 1963) when water temperatures correspond with the normal walleye spawning temperature of 7–9°C (Scott and Crossman, 1973). However, spawning in Lake Ontario appears to be limited because few immature (age ⩽4) walleye are observed in eastern and western Lake Ontario. Immature walleye are abundant in the upper bay from April to November. Increasing catch with age in eastern Lake Ontario suggests immigration from the Bay of Quinte rather than local production. Lake Ontario summer water temperatures are optimal for walleye; therefore the lack of natural reproduction seems surprising. In spring, Lake Ontario water temperatures lag about one month behind that of the Bay of Quinte. Therefore, walleye eggs and fry in the Bay of Quinte have an additional month to develop than in the lake. It is possible that this additional growth is needed to become large enough to ensure survival the following winter, but this hypothesis needs to be investigated.
Bioenergetics modeling of Quinte Walleye has shown a growth disadvantage for walleye that remain in the upper bay during summer when temperature exceeds 23.5°C for older walleye and 25°C for younger walleye (Hurley, 1986b). Walleye might benefit from migration away from the warmer waters of the upper Bay of Quinte during summer to avoid temperatures that would cause growth cessation. Both immature and mature walleye tended to move farther down the Bay of Quinte or to Lake Ontario toward cooler waters during summer. In the middle and lower Bay of Quinte, and Lake Ontario the depth is sufficient for temperature stratification, allowing walleye a choice of temperatures during summer (Figure 3). Where water temperatures were below the preferred temperature of 21.6°C, walleye tended to choose the depth with the warmest temperature. Many mature walleye migrated down the Bay of Quinte and into Lake Ontario well before the temperature reached critically warm levels in July. During May and June walleye migrating from the Bay of Quinte to Lake Ontario would experience a drop in temperature of about 5–10°C away from their preferred temperature. As well, walleye migrating from Lake Ontario into the Bay of Quinte during October or November would experience a drop in temperature 2–5°C away from their preferred temperature. Accordingly, temperature is likely a driver for summer migration, but it may not determine the exact timing of migration in or out of the Bay of Quinte.
Walleye migration is also likely related to their need to find available prey fish. Walleye migrating down the Bay of Quinte meet alewives, their primary prey, migrating up the bay from Lake Ontario in May and June. Walleye predation is a major factor in alewife mortality in the Bay of Quinte, and as alewives migrate up the bay their population is decimated (Ridgway et al., 1990). Few alewives are available for those walleye that delay migration. Prior to June, alewives are only available to walleye that migrate to Lake Ontario. Alewives are concentrated in shallow water during the spawning migration, and may be easy prey for walleye. However, later in the year the distribution of alewife may be more widespread after spawning and less available as forage for walleye. Walleye may return to the Bay of Quinte to prey upon the summer production of YOY of other major prey fishes, such as gizzard shad, yellow perch, and white perch. YOY production of these species in the in the Bay of Quinte is substantial (Hoyle et al., in press). These fish must grow large enough to become prey for walleye. Ridgway et al. (1990) found that alewife in walleye stomachs averaged 138 mm fork length in the Bay of Quinte. During late summer and fall, YOY of these species grow large enough for walleye to eat (e.g. gizzard shad: Figure 8). Madenjian et al. (1996) found that recruitment of walleye in Lake Erie was correlated with the availability of YOY gizzard shad as prey for the parental stock. YOY gizzard shad may be an important prey species influencing the fall migration of mature walleye up the Bay of Quinte.
Quinte Walleye appear to employ three size-dependent strategies associated with high temperatures in the upper bay during summer. The first strategy is to remain in the upper bay and wait out the warm period. The younger age groups which continue to grow at higher temperatures than older walleye use this strategy. As well, with a smaller body size younger walleye may consume smaller YOY prey fish earlier in the year compared with older walleye. Year to year variability in temperature may result in the upper bay being suitable for older walleye in cool years. The effect of warmer summers due to climate change (Casselman, 2002) may reduce the viability of this strategy. In a second strategy, some Quinte Walleye may avoid high summer temperatures by gradually moving to the middle and lower bay, thereby maintaining optimum temperatures for growth. These fish may take advantage of the alewife spawning run for food if they migrate early and far enough down the bay. The third strategy is to migrate earlier to Lake Ontario into colder water to take advantage of concentrated spawning alewife for foraging. This strategy is employed almost exclusively by mature walleye. The concentrated availability of alewife may offset the lower temperatures of Lake Ontario. Moreover, mature walleye may have high energetic needs in their post-spawning state, and their energetic needs may be more immediate than for immature walleye. If post-spawning walleye are in a critical state, the advantage of immediate food for increased survival could offset lower temperatures and lower prey availability later in the summer in Lake Ontario.
The distribution of walleye did not change through time among the regions: Bay of Quinte, eastern and western Lake Ontario. Rather, abundances declined in concert across all regions. Accordingly, we see no support for the contention of Casselman et al. (1999) and Casselman and Scott (2003) that walleye moved to Lake Ontario in response to increased water clarity from dreissenids. Within the Bay of Quinte the distribution of walleye shifted after the onset of dreissenid mussel invasion in the middle and lower bay during 1997–1999. Walleye increased in abundance the middle bay along with a corresponding decline in the lower bay (Figure 5). A decline in abundance in the upper bay after dreissenid invasion was consistent with the predictions of a decline in thermal-optical habitat for walleye (Chu et al., 2004). As well, Chu et al. predicted an increase in thermal-optical habitat for walleye in the middle bay in 1999 which corresponded with one of the years we saw an increase in walleye abundance. In the lower bay, however, Chu et al. (2004) predicted increasing thermal-optical habitat after dreissenid invasion but walleye abundance declined. The relationship between water clarity and shifts in walleye distribution in the Bay of Quinte remain unclear. An alternate explanation for the distributional shift of walleye in the Bay of Quinte may be related to prey abundance. Yellow perch increased in the diet of walleye in the Bay of Quinte after dreissenid invasion, particularly in 1998 and 1999 (Bowlby et al., 2010). As well, catch of small yellow perch in trawls increased in concert with walleye abundance in the middle bay (Hoyle et al., in press), and walleye may have been attracted to an area with high concentration of prey. Dreissenids were likely responsible for habitat changes in the Bay of Quinte resulting in the increase in yellow perch abundance (Hoyle et al., in press).
Walleye distribution appeared to be unaffected by colonization of Round Goby. Gobies went from low to high abundance throughout most of the Bay of Quinte over the period from 2002–2003 (OMNR, 2010), and likely did not create enough spatial contrast to affect walleye distribution.
Walleye migration between the Bay of Quinte and Lake Ontario during spring and fall was consistent with avoiding warm temperature in the upper bay, and foraging on alewife in the lower bay and Lake Ontario during summer and young-of the-year fishes such as gizzard shad during fall. The distribution of walleye between the Bay of Quinte and Lake Ontario did not change after dreissenid colonization.
We gratefully acknowledge contributions made by staff at the Glenora Fisheries Station and the thoughtful suggestions of Bruce Morrison and two anonymous reviewers.