We reviewed responses associated with the invasion of dreissenid mussels by two eastern Lake Ontario fish populations and the fisheries they support. Resurging lake whitefish and walleye populations declined following dreissenid mussel invasion in the early 1990s. Impacts on whitefish were associated with the loss of a key diet item, Diporeia, and its replacement with diet items of lower energy value. Impacts featured a die-off, dispersal, declines in juvenile and adult condition and growth rates, delayed age-at-maturity, and several years of reproductive failure. Impacts on walleye were consistent with dreissenid driven ecosystem change, particularly, clearer water. The key response by the walleye population was a downward shift in recruitment levels. This shift appears to be due to a change in the stock-recruitment relationship caused by decreased survival during early life (i.e. egg to 4-months), and suggests that the carrying capacity for these early life stages has diminished. Currently, whitefish reproduction has resumed and walleye reproduction appears stabilized at a lower level. Recent (i.e. 2003 and 2005) whitefish year-classes were relatively large but the fish are growing slowly and annual survival rate is not yet known. The whitefish commercial harvest continues to decline in synchrony with the declining adult whitefish population. The walleye recreational fishery (i.e. effort and harvest) has stabilized at a smaller size consistent with lower walleye year-class strength.
Lake whitefish (Coregonus clupeaformis; referred to as whitefish) and walleye (Sander vitreus) are important elements of the fish community and fisheries of eastern Lake Ontario. From a historical perspective, both species are native to Lake Ontario and, although having shown large fluctuations in abundance, have always been among the most abundant of the large-bodied species in the fish community. A low point for the two species occurred in the 1970s when these species, having declined in the 1960s, persisted only at remnant levels. Factors involved with the declines in the 1960s and subsequent resurgences in the 1980s have been documented and included cultural eutrophication, over-exploitation, introduction of non-native fish species, and climate (Christie, 1972; Hurley and Christie, 1977; Hurley, 1986a; Christie et al., 1987). More recently, whitefish and walleye populations and fisheries have responded rather dramatically to ecosystem changes that followed the invasion of dreissenid mussels (zebra, Dreissena polymorpha and quagga, D. bugensis) to Lake Ontario.
Whitefish and walleye are both highly migratory in Lake Ontario. Whitefish inhabit the open-waters of eastern Lake Ontario, especially the Kingston Basin (Fig. 1), often at depths of 20–35 m, for much of the year with mature fish migrating during fall to major spawning areas in the Bay of Quinte and the southern shore of Prince Edward County (Fig. 1) in Lake Ontario (Hoyle, 2005). After hatching in April, the young whitefish inhabit the lower reaches of the Bay of Quinte and nearshore areas of eastern Lake Ontario. They are commonly caught in routine bottom trawls, about 20 m deep, during August (commonly 70–110 mm fork length) near Conway and Timber Is (Fig. 1). Walleye spawn during April, primarily along the shoreline and in the major rivers of the Bay of Quinte. Soon after spawning, adult walleye migrate during May to eastern Lake Ontario where they reside during the summer. Walleye fry hatch during May and are caught in Bay of Quinte bottom trawls at 4–20 m deep during August (commonly 100–150 mm fork length). Juvenile walleye inhabit the Bay of Quinte, at all depths, year-round to about age-4 or age-5 years (Payne, 1963; Hurley, 1986b).
Whitefish feed on benthic macro-invertebrates. Historically, they relied on the deepwater amphipod (Diporeia spp.) for much of their food energy in eastern Lake Ontario and elsewhere in the Great Lakes (Dermott et al., 2005; Mohr and Nalepa, 2005). Walleye are piscivorous and the most abundant top predator in the fish community of the Bay of Quinte and eastern Lake Ontario. Adult walleye feed, nearly exclusively, on alewife in eastern Lake Ontario. In the Bay of Quinte, the walleye diet is more diverse with alewife, gizzard shad, yellow perch, white perch all contributing significantly at times (Hurley, 1986b; Bowlby et al., 1991; Morrison et al., Lake Ontario Management Unit, Ontario Ministry of Natural Resources, Picton, Ontario, in press).
Dreissenid mussels first became abundant in eastern Lake Ontario in 1993 (Dermott, 2001; Hoyle et al., 2003) and measurable impacts on water quality were observed that year (Johannsson et al., 1998). Larger impacts on water quality were apparent by 1994 (e.g. rapid and dramatic reductions in chlorophyll and phytoplankton, Nichols, 2001). Benthic macroinvertebrate populations were also greatly impacted. Diporeia rapidly disappeared over the 1993–1995 time-period from eastern Lake Ontario (Dermott, 2001; Hoyle et al., 2003), and whitefish were forced to include items with lower energy content in their diet, especially dreissenids (Hoyle et al., 2003; Lumb et al., 2007). In the Bay of Quinte, dreissenid mussels proliferated and impacted water quality about one year later than eastern Lake Ontario; in 1994 and 1995, respectively (Bailey et al., 1999; Dermott, 2001; Hoyle et al., 2003). Increased water clarity resulted in an expansion of the aquatic vegetation distribution in the Bay of Quinte (Leisti et al., 2006) and eventually expanded the distributions and increased the abundances of fish species preferring this habitat, especially bluegill sunfish, black crappie and largemouth bass, during the late 1990s.
Following the resurgence of whitefish and walleye populations in the 1980s, commercial fishing for the two species was re-established and closely regulated. Whitefish harvest quotas were established and gradually increased as the population increased. Most whitefish harvest was taken by 4½ inch gillnets although a portion of the harvest was taken in impoundment gear. Harvest was largely restricted to seasons and locations to best avoid incidental catch of lake trout and walleye in gillnets—primarily during fall in the whitefish migration and spawning areas of eastern Lake Ontario and the Bay of Quinte (Hoyle, 2005). Walleye harvest was primarily allocated to the recreational fishery but small quotas were established for commercial impoundment gear and to cover walleye incidental catch in the whitefish gillnet fishery. No commercial walleye harvest was permitted in the upper Bay of Quinte where the recreational walleye fishery largely occurred (i.e. west of Glenora; see Fig. 1; Bowlby et al., 1991). The recreational walleye fishery developed in the 1950s, following an increase in walleye abundance in the 1940s (Christie, 1973). The fishery collapsed in the 1960s but increased dramatically as the walleye population recovered in the 1980s. Some winter ice-fishing occurred but most of the recreational walleye harvest occurred during the open-water season in the Bay of Quinte, west of Glenora. An aboriginal fishery (spear fishing for spawning walleye and gillnetting primarily during spring and fall) also occurs in the Bay of Quinte but long-term monitoring data do not exist.
In this paper, we review the synchronous responses of whitefish and walleye populations to ecosystem change after dreissenid mussel invasion. We draw upon long-term annual index fishing, commercial harvest reporting, and recreational fishery survey data. The status of these fish populations and their supported fisheries are updated and contrasted a decade after dreissenid mussel invasion impacted the eastern Lake Ontario ecosystem.
We updated long-standing abundance indices from index gillnetting and bottom trawling operations based out of the Glenora Fisheries Station on the Bay of Quinte near Picton, Ontario, and conducted on the Bay of Quinte and eastern Lake Ontario. Updated results were presented graphically and interpreted with respect to the timing of dreissenid mussel impacts on the local ecosystem; pre and post 1994 in eastern Lake Ontario, and pre and post 1995 in the Bay of Quinte. The detailed sampling methodologies of these long-term and intensive programs are described elsewhere (e.g. Hurley, 1986a; Christie, 1987; Casselman et al., 1999; Casselman et al., 2002; Casselman and Scott, 2003; Morrison et al., Lake Ontario Management Unit, Ontario Ministry of Natural Resources, Picton, Ontario, in press) and briefly summarized here.
Gillnets were set in the Bay of Quinte and the Kingston Basin of Lake Ontario (Fig. 1), and captured age-1 and older walleye and whitefish (age-0 fish were not vulnerable to capture). Gillnets were comprised of multiple mesh sizes from 38.1–152.4 mm (12.5 mm increments) stretched mesh and catch-per-gillnet (CUE) values were standardized to represent 1000 m of each gillnet. Up to nine fixed index fishing areas (EB01, EB02, EB03, EB04, EB05, EB06, Melville Shoal, Grape Is, and Flatt Pt) were sampled in the Kingston Basin; three of the areas (Melville Shoal, Grape Is, and Flatt Pt) were depth-stratified (5–10 m, 10–15 m, 15–20 m, 20–25 m, and 25–30 m). Two fixed index fishing sites were used in the Bay of Quinte (Big Bay and Hay Bay). Least squares means were used to account for years when all areas/depths were not sampled. Gillnets were set for approximately 24 hr. All netting occurred during June–September.
Multifilament gillnets were replaced with monofilament in 1991 (Bay of Quinte) or 1992 (Kingston Basin). Gear comparisons revealed no significant differences in whitefish CUE between the two gear types; hence no correction factor was applied. For walleye, monofilament gillnets caught twice as many fish as multifilament; therefore all multifilament CUEs were multiplied by a factor of two.
Kingston Basin index netting sites (Fig. 1), at depths greater than 20 m, were used to index the abundance of whitefish age-1 and older. Bay of Quinte index netting sites (Fig. 1), at depths less than 15 m were used to index the abundance of juvenile walleye (age-1 to approximately age-4 or age-5) and Kingston Basin index netting sites, at depths less than 15 m (Fig. 1) were used to index the abundance of adult walleye. Walleye age distribution in gillnets was determined using otoliths, sectioned through the origin, for fish caught in 2004 and 2005.
Bottom trawls were used to sample age-0 whitefish and walleye. Catches were standardized (CUE) as the total catch in a single 6 min trawl (¼ mile in distance) using a ¾ Yankee (Lake Ontario) or ¾ Western (Bay of Quinte) bottom trawl. All trawling occurred during August-September when age-0 whitefish and walleye were about four months of age. Two fixed sites (Timber Is. in the Kingston Basin and Conway in the lower Bay of Quinte; Fig. 1) were used to index the abundance of whitefish while up to six sites (Trenton, Belleville, Big Bay, Deseronto, Hay Bay and Conway) were used to measure walleye abundance. No trawling was done in 1989. Least squares means were used to account for years when all sites were not sampled.
Commercial fish harvest reporting is mandatory. Whitefish and walleye harvest statistics were summarized from Baldwin et al. (1979) for the years 1957–1977 and from records maintained by the Ontario Ministry of Natural Resources (OMNR) at the Glenora Fisheries Station (Picton, Ontario, Canada, K0K 2T0, unpublished data) for the years 1978–2005.
Bay of Quinte walleye recreational fishery
Assessment of the Bay of Quinte recreational fishery (May-November open-water fishery) was completed only sporadically prior to 1979 (1957–1960, 1974, 1976). Beginning in 1979, the open-water fishery was surveyed annually using a roving stratified sampling methodology (boat counts and angler interviews) to estimate angler effort and harvest-per-unit effort (HUE). All results were expanded to represent the open-water fishing season from opening weekend in early May to November 30.
Whitefish were nearly absent from gillnet catches in the 1970s. CUE increased exponentially during the 1980s and early 1990s, peaking in 1993 (Fig. 2). Thereafter, CUE declined, equally dramatically, and by 2005 reached its lowest level since 1981. The timing of this decline coincides with the time that Dreissena colonized the area and impacted water quality in eastern Lake Ontario (Fig. 2).
Walleye CUE in Bay of Quinte gillnets declined in the late 1960s, was very low for several years in the 1970s, increased dramatically in the 1980s, was very high until the early 1990s, declined through to about 2000, and finally stabilized at moderate levels through to 2005 (Fig. 3). When gillnetting began in 1978 in the Kingston Basin of eastern Lake Ontario, walleye were at very low abundance. The increase in Kingston Basin walleye CUE through the 1980s and early 1990 lagged behind the increase in the Bay of Quinte by several years. Walleye CUE in the Kingston Basin peaked in 1990, declined through to about 2000, and finally stabilized at moderate levels through to 2005. The walleye CUE decline in both the Bay of Quinte and the Kingston Basin of eastern Lake Ontario appears to have begun 3 or 4 years prior to the time that Dreissena had colonized these areas and impacted water quality (Fig. 3). Juveniles dominate (95%) the gillnet CUE for walleye in the Bay of Quinte, whereas older walleye make-up the bulk (92%) of the gillnet CUE for walleye of the Kingston Basin (Fig. 4).
Age-0 whitefish CUE was very low during the 1970s but several large year-classes were observed during the 1980s and early to mid-1990s, especially the 1987, 1991, 1992, 1994 and 1995 year-classes. Relatively small numbers of age-0 whitefish were caught in 1996 and 1997 and very few age-0 fish were caught from 1998–2002 (Fig. 5). Age-0 CUE increased again after 2002 with relatively large numbers of fish caught in 2003 and 2005. Both the Conway (lower Bay of Quinte) and Timber Is. (Kingston Basin) sampling locations produced large numbers of age-0 fish during the 1980s and early to mid-1990s although CUE at Conway tended to be higher. After 2002, CUE of age-0 fish tended to be higher at Timber Is. (Fig. 5).
When annual bottom trawling began in 1972, no age-0 walleye were caught until 1977 when a few were taken. However, large numbers of age-0 walleye were caught in 1978 and, in the years following, a general trend of increasing numbers of age-0 fish were caught until 1995. After 1995, age-0 CUE was lower and appears to have stabilized at a level about one-third that of the 1977–1995 time-period. The lower CUE of age-0 walleye after 1995 corresponds with the time that Dreissena had colonized these areas and impacted water quality (Fig. 6).
Commercial harvest of whitefish was high in the late 1950s and early 1960s (average 169,000 kg annually from 1957–1963), declined rapidly to extremely low levels during the 1970s and early 1980s, then increased steadily to a peak of nearly 300,000 kg in 1996 (Fig 7). Whitefish harvest declined rapidly after 1996 and in 2005 was the lowest observed (less than 22,000 kg) since 1985 (Fig. 7).
Commercial harvest of walleye was highest in the 1950s (averaged nearly 56,000 kg annually from 1957–1962), declined to negligible levels during the 1970s, increased in 1979 and 1980, and was closed to fishing in 1981 (Fig. 7). The fishery was re-opened in 1989 and harvest increased gradually to just less than 19,000 kg in 1997 then declined to less than 5,000 kg after 2001.
The walleye recreational fishery harvest was low and declining in the late 1950s and early 1960s at which time surveys were discontinued. Harvest remained low in 1974 and 1976 but when annual surveys resumed in 1979, harvest had increased to nearly 10,000 kg (15,000 fish). Walleye harvest increased sharply to nearly 90,000 kg (192,000 fish) in 1980. Subsequently, high harvest levels were maintained through the early 1990s. Harvest declined steadily after 1991 until about 2000 after which harvest levels remained relatively consistent at a lower level. The harvest levels from 1998-2005 averaged about one-quarter those of the 1980–1997 time-period (Fig. 7 and Fig. 8).
Recreational fishing effort for walleye was relatively low (less than 130,000 angler hours) in the late 1950s, 1976 and in 1979 when annual surveys resumed. Fishing effort increased sharply in 1980 then increased steadily until 1996 (over 600,000 angler hours), declined to a low point in 2002, and finally increased slightly and stabilized at about 200,000 angler hours annually from 2003–2005. The trend in walleye harvest per unit effort was closely correlated with walleye abundance in the Bay of Quinte gillnets (compare Fig. 8 with Fig. 3; r = .851, p < .001).
When dreissenid mussels invaded eastern Lake Ontario in the early 1990s whitefish and walleye populations were about ten-fifteen years into dramatic resurgences after having existed only at remnant levels in the 1970s (Bowlby et al., 1991; Casselman et al., 1996). At that time, dreissenid mussel invasion was imminent, and the major anticipated impact was increased water clarity due to the mussels' filter feeding activities. In the Bay of Quinte, where aquatic vegetation distribution was limited by turbid water conditions, it was hypothesized that aquatic vegetation would increase. These ecosystem changes were considered to be less favourable to walleye and more favourable to walleye predators and competitors (Bowlby et al., 1991). Negative impacts on the whitefish in eastern Lake Ontario were not anticipated (Casselman et al., 1996). A decade, or so, following dreissenid mussel invasion and proliferation, populations of the two fish species are now much reduced from peak levels of the early to mid-1990s.
Impacts on the whitefish population were undoubtedly and directly related to the astonishing loss of Diporeia, their primary food supply, which coincided with the invasion of dreissenids. Immediate impacts on whitefish included a die-off, dispersal, and a change in diet (Hoyle et al., 2003, Hoyle, 2005). The post dreissenid whitefish diet has a decreased energy content compared to pre dreissenid conditions (Lumb et al., 2007). Whitefish lost body condition and growth rate declined. The whitefish population experienced near reproductive failure for five consecutive years, 1998–2002, before moderate levels of reproductive success resumed in 2003. Post dreissenid whitefish have a very different life history strategy than that of the previous generation. Fish are now much slower growing and later maturing on their replacement diet (Hoyle, 2005). Life history traits that should accompany slower growth would also include lower natural mortality rates and increased longevity but this remains to be seen. Therefore, the future for whitefish in Lake Ontario is uncertain.
Achievement of whitefish population biomass or fishery yield levels observed during the late 1980s and early to mid-1990s seems highly unlikely given the current ecosystem in eastern Lake Ontario. Given much reduced growth, whitefish would have to be more abundant than at the peak abundance in the early 1990s; survival rate would have to be very high. To date, the relatively strong 2003 and 2005 year-classes that showed up during routine bottom trawling (Fig. 4) have not recruited strongly to index gillnets (Fig. 2). We anticipated that slower whitefish growth would result in delayed recruitment to our Kingston Basin gillnets but the 2003 year-class was expected to recruit more strongly by 2005 than was the case.
The whitefish commercial fishery has declined markedly. Quota reductions may have played a role but the fishery was unable to harvest its quota even at the reduced levels. Lower whitefish abundance and relatively poor body condition of the fish (i.e. low fat content) must be major factors contributing to the harvest decline. A much slower growing and later maturing fish results in fish that are largely inaccessible to the 114 mm (4.5 inch) mesh gillnet fishery for at least two to three additional years compared to the past. More fish would have to be harvested to achieve the same harvest weight.
Impacts on walleye, though anticipated, were more subtle and complex than those on whitefish. Walleye gillnet CUE actually began to decline prior to dreissenid mussel invasion, suggesting that the resurging population had “over-shot” its carrying capacity. Mills et al. (2003) suggested that walleye distribution generally “moved” down the Bay of Quinte to the lower Bay and eastern Lake Ontario. This change in distribution occurred as walleye density increased and as alewife, the walleye's preferred prey, were depleted during the late 1980s in the upper Bay (Ridgeway et al., 1990) and during the early 1990s in the lower Bay (Casselman and Scott, 2003). Further distribution changes may have occurred due to increasing water clarity following dreissenid mussel invasion. Chu et al. (2004) found that a decline in suitable walleye habitat (increased water clarity and temperature), since the invasion of dreissenid mussels, coincided with walleye abundance decline in the upper Bay of Quinte. However, the pattern was not consistent through time and Chu et al. suggested that other factors must have affected these walleye. Nonetheless, clearer water conditions post-dreissenid mussel likely changed walleye behaviour and distribution patterns in the Bay of Quinte.
There is no compelling indication that changes in walleye distribution patterns have significantly changed the overall annual walleye migration pattern, first described by Payne (1963), nor caused decreased walleye survival. Immature walleye still reside in the Bay of Quinte and mature walleye, having spent the summer months in eastern Lake Ontario (see age distributions in summer gillnets, Fig. 4), return to the Bay of Quinte to spawn. The relationship between juvenile walleye abundance in the Bay and adult walleye abundance in the Kingston Basin (Fig. 9) gives no indication that mortality rate has changed for these life history stages. Old fish still figure prominently in the walleye population (Fig. 4) suggesting that mortality is not excessive. However, clearly, recruitment levels have been reduced post dreissenid mussel (Fig. 5 and Fig. 9). It may be that less adult walleye are returning to the Bay of Quinte to spawn but even a small parental stock size can produce a very strong year-class (e.g. 1978 year-class; Fig. 6). The post dreissenid walleye population has stabilized at a level lower than that pre dreissenid mussel with juvenile and adult abundance levels consistent with lower early life history recruitment.
Reduced recruitment levels since dreissenid mussel invasion may be due to decreased early life history survival (i.e. at a life stage prior to recruitment as age-0 fish in our August bottom trawls). It appears that there has been a decrease in the number of recruits for a given parental stock size (Fig. 9). This suggests a reduced carrying capacity for eggs/fry in the Bay of Quinte and may indicate that survival has declined for these life history stages. Potential hypotheses for increased mortality include a decline in suitable spawning area, increased levels of predation and/or competition, or reduced food supply/availability. The relative foraging efficiencies of walleye fry and their competitors/predators in clearer water conditions may also be involved. Hoxmeier et al. (2006), in a study of 15 Illinois reservoirs, found that prey availability was an important factor for larval (6 mm total length) and small (46 mm total length) walleye survival, and that juvenile centrachid density had a negative effect on larval walleye survival, presumably caused by predation. Quist et al. (2003) demonstrated that recruitment of walleye can be reduced by centrarchid predation on larval walleye, but they also found that water clarity had no direct effect on recruitment. More detailed analysis of stock-recruitment relationships, predator/competitor abundance levels, and prey availability pre and post dreissenid mussel may reveal more information about the cause of reduced recruitment levels in the Bay of Quinte. For example, potential walleye fry predators that increased in abundance following dreissenid mussel invasion included yellow perch, black crappie, bluegill, and largemouth bass (Ontario Ministry of Natural Resources, Glenora Fisheries Station, Picton, Ontario, unpublished data).
In summary, lake whitefish and walleye populations showed dramatic, negative responses to dreissenid mussel invasion in eastern Lake Ontario. The most important finding was that these responses led to the populations having new equilibriums—driven by lower growth rate for lake whitefish and decreased early life history survival for walleye. We suggest that given the current ecosystem status, recovery of lake whitefish and walleye populations to pre dreissenid mussel levels will be difficult.
We gratefully acknowledge past and on-going contributions made by operations staff at the Glenora Fisheries Station and the thoughtful suggestions offered by two anonymous reviewers.