Restoration targets for nearshore fish populations in Hamilton Harbour and Toronto Harbour–Areas of Concern–were developed. Fish were captured in trap nets at Hamilton Harbour, Toronto Harbour, and nine unimpaired embayments of Lake Ontario. Embayments were sampled at least once and up to nine times during 2006–2013. Principal components analysis indicated the fish communities in unimpaired embayments clustered into two groups. One group had larger exposure distance (opening) to Lake Ontario and larger exposure index (surface area ÷ opening), and was termed “exposed.” The alternate group was “sheltered.” Fish communities in unimpaired embayments followed an aquatic macrophyte gradient of open-water species (e.g. White Bass, Gizzard Shad) at one end and species associated with vegetative cover (e.g. Bluegill, Pumpkinseed) at the other. A second fish community gradient was correlated with embayment opening and exposure index, and may be associated with habitat stability and exchanges with colder water from Lake Ontario. Hamilton Harbour is a sheltered embayment, but principal components analysis indicated the fish community was significantly different from all other embayments, consistent with its Area of Concern status. Northern Pike and Yellow Perch were the only target species in Hamilton Harbour close in abundance to unimpaired sheltered embayments. Common Carp, Brown Bullhead, Channel Catfish, and others were an order of magnitude higher than in unimpaired sheltered embayments. Toronto Harbour is an exposed embayment, and principal components analysis indicated the fish community was not significantly different from unimpaired embayments. Likewise, the abundance of most of the target species in Toronto Harbour was close to unimpaired embayments. Walleye, Smallmouth Bass, and Rock Bass were depressed and Gizzard Shad was elevated in Toronto Harbour, compared to unimpaired embayments.
Fish communities in Hamilton Harbour and Toronto Harbour were depressed by the early 20th Century due to dredging, contaminated sediments, eutrophication and other human activities (Whillans, 1979; Holmes and Whillans, 1984). As part of an effort to restore degraded hotspots in the Great Lakes, both of these harbours were listed as an Area of Concern (AOC) by the International Joint Commission in 1987. Subsequently, Remedial Action Plans (RAP) designated their fish populations as “impaired” (Metro Toronto and Region RAP Office, 1994; O'Connor, 2003). The fish community RAP target for delisting the Hamilton Harbour AOC is based on an Index of Biotic Integrity (IBI) (Minns et al., 1994). This target is for the nearshore fish community, based on attributes of the fish community related to species richness, trophic structure, and native species abundance. Water quality and fish habitat in Hamilton Harbour have been partially restored, but IBI and fish abundance data indicate the fish community remains impaired (Hiriart-Baer et al., 2016; Hamilton Harbour RAP Office, 2012; Boston et al., 2016). The Hamilton Harbour and Watershed Fisheries Management Plan (HHWFMP) recommended the creation of abundance targets for fish species important to ecosystem function and fisheries (Bowlby et al., 2010), but these targets have not been created. Nearshore species targets could support the IBI targets for the RAP, and contribute to efforts to address habitat needs for each species. Fish community or species targets for the Toronto RAP are under development, and this article can contribute to this effort.
Species targets may be empirically developed using historic fish abundances or with a comparative approach using fish abundances in similar habitats. Unfortunately, historic fish communities in Hamilton Harbour and Toronto Harbour were not quantified. Whillans (1979) described these communities as dominated by coolwater and warmwater fishes — such as Northern Pike (Esox lucius), Walleye (Sander vitreus), Largemouth Bass (Micropterus salmoides), Yellow Perch, and Sunfish—typical of fish communities in other Lake Ontario embayments (Hoyle et al., 2012). These coolwater and warmwater fishes are most abundant in warmer nearshore areas usually associated with embayments (Hoyle et al., 2007). Much of Lake Ontario is colder and dominated by coldwater fishes (Stewart and Bowlby, 2009). Lake Ontario has significant upwelling events, bringing cold hypolimnetic water close to shore during summer (Rao and Murthy, 2001). Wind-driven currents, fetch, and other physical processes may result in exchanges of cold water from the depths of Lake Ontario with warmer water along the shoreline and in embayments (Rao et al., 2009). Sheltered embayments may have less exchange with Lake Ontario resulting in better habitat for nearshore fishes than exposed embayments (Murphy et al., 2011). As a result, nearshore fish communities in Lake Ontario differ between exposed and sheltered habitat (Chu et al., 2014). Hamilton Harbour and Toronto Harbour differ in their exposure to the open waters of Lake Ontario, and this must be taken into account to predict restored nearshore fish population abundance.
The objective for this article is to develop abundance targets for restored nearshore fish populations in Hamilton Harbour and Toronto Harbour. We take a comparative fish community approach with unimpaired Lake Ontario embayments to predict the abundance of key species after habitat restoration. We test underlying assumptions that fish communities are similar among the unimpaired embayments, and different from those in Hamilton Harbour and Toronto Harbour. Jointly we consider the hypothesis that fish communities in Lake Ontario embayments differ based on their degree of exposure to Lake Ontario.
Fish were captured in 6-ft trap nets at Hamilton Harbour, Toronto Harbour, and nine unimpaired embayments of Lake Ontario (Figure 1). Although the Bay of Quinte is an AOC, its fish communities are “likely are not impaired” (Environment Canada and the Ontario Ministry of the Environment, 2011) and have been so since the mid-1990s (Brousseau et al., 2011). Trap netting was based on an Ontario Provincial standard method (nearshore community index netting) for sampling aquatic littoral area fishes of a size greater than 90 mm (Stirling, 1999). These nets miss smaller fishes including darters and most minnows (Hoyle et al., 2012). Embayments were sampled at least once and up to nine times on an annual basis during 2006–2013 between August and September (Table 1). At each embayment 12–36 sampling sites (Table 1) were randomly chosen 1.7–3.5 m deep. This number of sampling sites has resulted in a percent relative standard error generally <30% for the catch of the top 10 species at each embayment in a year (Ontario Ministry of Natural Resources [OMNR], 2014).
All netted fish were identified to species and counted. The mean catches of fish by species and year were calculated for each embayment, averaged across years, and then log-transformed. Redhorse species (Moxostoma sp.) catches were combined due to difficulties in identification, and their occupation of very similar ecological niches. Coldwater species were excluded as they were not considered part of the nearshore fish community (Stewart et al., 2013). Sporadic species that were observed at fewer than 7 embayments were excluded.
Two sets of principal components analyses (PCA) were performed using Statistica (StatSoft, Inc., 2007) with transformed mean catch. Twenty-one species were included in the PCAs (Table 2). PCA-I included all embayments to evaluate the differences in fish community between Hamilton Harbour and Toronto Harbour and unimpaired embayments. PCA-II included only fish communities from unimpaired embayments. Differences in fish communities between embayments were determined by examining the significance (p<0.05) within each correlation matrix.
Surface area, distance across opening exposed to Lake Ontario, and exposure index (surface area ÷ opening) were determined to indicate the exposure of embayments to Lake Ontario for water exchange (Table 1). Surface area and the length of the opening were measured from digital images. Correlations between the fish community PC-II factor coordinates and these indicators of exposure were tested.
Restoration targets for nearshore fish populations in Hamilton Harbour and Toronto Harbour were estimated as the mean trapnet catch in unimpaired embayments. The selection of target fishes was based on target species defined in the HHWFMP. Hyper-abundant species were defined based on catch more than 10 times the mean catch in unimpaired embayments.
PCA-I indicated that the nearshore fish community in Hamilton Harbour differed greatly from the fish communities in other Lake Ontario embayments, but the Toronto Harbour fish community was similar to some of the unimpaired communities (Figure 2). The fish community in Hamilton Harbour was not significantly correlated with fish communities in other Lake Ontario embayments, but the fish community in Toronto Harbour was significantly correlated with those in the North Channel, Lower Bay of Quinte, Prince Edward Bay, and Presqu'ile Bay (Table 3). Most of the variation (85%) in these fish communities was described by the first 3 principal components (PC). PC 1 and PC 2 were similar for PCA-I and PCA-II, and these are described for PCA-II below. PC-I 3 accounted for 10.1% of the variation in fish communities and was almost exclusively explained by the difference between Hamilton Harbour and the other embayments. Much of this difference was related to relatively higher abundance of Brown Bullhead, Channel Catfish, and White Perch, and lower abundance of Smallmouth Bass (Micropterus dolomieu) and White Sucker (Catostomus commersoni) in Hamilton Harbour compared with other embayments.
Fish communities in the unimpaired embayments (PCA-II) separated into two clusters (Figure 3) based on fish community correlations (Table 4). The first cluster included North Channel, Lower Bay of Quinte, and Prince Edward Bay; the second cluster included West Lake, Upper Bay of Quinte, East Lake, Middle Bay of Quinte and Wellers Bay. Presqu'ile Bay was not different from either cluster. PC-II 1 accounted for 68.2% of the variation in fish communities and described a fish community gradient with major contribution by Brown Bullhead (Ameiurus nebulosus) and Sunfishes at one end and White Bass (Morone chrysops), Channel Catfish, and Golden Shiner (Notemigonus crysoleucas) at the other (Figure 4, Table 5). PC-II 1 was significantly correlated with surface area of the embayments (r = 0.69, P = 0.039). PC-II 2 accounted for 14.6% of the variation in fish communities and described a fish community gradient with major contributions by Brown Bullhead, White Sucker, and Rock Bass (Ambloplites rubestris) at one end and Bluegill, White Perch, and Black Crappie (Pomoxis nigromaculatus) at the other (Figure 4, Table 5). PC-II 2 was significantly correlated with opening (r = 0.86, P = 0.003) and exposure index (r = 0.72, P = 0.029). These correlations were clearly dependent on fish community differences between the two clusters. The first cluster (North Channel, Lower Bay of Quinte, and Prince Edward Bay) of fish communities relates to “exposed” embayments, and the second cluster relates to “sheltered.” The fish community in Presqu'ile Bay could not be clearly excluded from exposed or sheltered clusters. With the analysis of opening it belonged with the sheltered cluster, but with the analysis of exposure index it belonged with the exposed cluster. Consequently, the Presqu'ile Bay fish community was excluded from further analysis.
Restoration targets for nearshore fish populations in Hamilton Harbour were based on sheltered embayments. Hamilton Harbour's opening (88 m) and exposure index (4.2) were within the range of sheltered embayments’ openings (21–1,033 m) and exposure index (1.4–14.1, Table 1). Catches of nearshore fishes in sheltered embayments were used to predict restored fish populations in Hamilton Harbour (Table 6). With the exception of Northern Pike and Yellow Perch, the catches of HHWFMP target species were extremely low in Hamilton Harbour, ranging 0.06 (Smallmouth Bass)–0.29 (Rock Bass) times lower than sheltered embayments. Northern Pike and Yellow Perch (Perca flavescens) catches in Hamilton Harbour were similar to the sheltered embayments (Table 6). Catches in Hamilton Harbour were dominated by Bluegill, Gizzard Shad (Dorosoma cepedianum), Common Carp, Brown Bullhead, Channel Catfish, and White Perch (Morone americana) (Table 1). Of these, only Bluegill (Lepomis macrochirus) catches are lower than those in sheltered embayments (Table 6), and the rest are hyper-abundant species, ranging from 10 (Gizzard Shad)–69 (Channel Catfish) times higher than in sheltered embayments.
Similarly, restoration targets for nearshore fish populations in Toronto Harbour were based on exposed embayments. Toronto Harbour's opening (1,960 m) was higher than sheltered embayments, but lower than exposed embayments (5,513–9,247 m, Table 1). Toronto Harbour's exposure index (137.1) was higher than exposed embayments (45.6–90.7, Table 1). Above, PCA-I indicated that the fish community in Toronto Harbour was not significantly different than unimpaired exposed embayments. Predictably, trapnet catches in Toronto Harbour were similar to these embayments (Table 6). Northern Pike, Pumpkinseed (Lepomis gibbosus), Largemouth Bass, and Yellow Perch were higher (1.29–2.09 times) than exposed embayments (Table 6), and Walleye, Smallmouth Bass, and Rock Bass were much lower (0.05–0.16 times). Among the hyper-abundant species (based on Hamilton Harbour designation) Channel Catfish (Ictalurus punctatus) in Toronto Harbour were not elevated (0.14 times) relative to exposed unimpaired embayments, but the remainder were elevated (Table 6). Gizzard Shad were extremely elevated (57.38 times), but species that were hyper-abundant in Hamilton Harbour were less elevated (2.24–7.70 times).
Both Hamilton Harbour and Toronto Harbour are situated in the Greater Toronto and Hamilton Area (Ontario Ministry of Infrastructure, 2012), a metropolis that includes eight other distinct cities along the western Lake Ontario waterfront between and beyond Toronto and Hamilton. Toronto and Hamilton have a long history of industrial and residential development with associated impacts on fish communities (Whillans, 1979). In comparison, the unimpaired embayments are in the eastern part of the lake, where urban development is lower and includes only three small cities along the Bay of Quinte: Quinte West (formerly Trenton), Belleville and Kingston. Nonetheless, like Hamilton Harbour and Toronto Harbour, the Bay of Quinte is an AOC, and during the 1960s and 1970s fish communities were impaired, largely as a result of hyper-eutrophication and low dissolved oxygen (Hurley and Christie, 1977). The impaired fish community in the Bay of Quinte of that time (Hurley, 1986) was similar to the current community in Hamilton Harbour. This suggests that the current east-west differences are not biogeographical, but rather are related to the presence of impaired habitat in Hamilton Harbour.
Although a nearshore fish community gradient (PC-II 1) was correlated with surface area of unimpaired embayments, this may be a surrogate for aquatic macrophytes. Fishes that prefer macrophytes, such as Bluegill, Pumpkinseed and Brown Bullhead (Scott and Crossman, 1973) were at one end of this gradient. Smaller lakes tend to be shallower, with less fetch, leading to greater relative abundance of aquatic macrophytes, compared with larger lakes (Chambers, 1987; Gasith and Hoyer, 1998). Aquatic macrophytes increase habitat complexity, resulting in a higher abundance and diversity of fishes (Thomaz and Cunha, 2010). In contrast, fishes that prefer open water, such as Gizzard Shad, Channel Catfish and White Perch (Scott and Crossman, 1973) were at the other end of the fish community gradient. Larger deeper embayments may lack sufficient light penetration and limit aquatic macrophytes. As well, wave action resulting from greater fetch may impact aquatic vegetation (Doolittle et al., 2010) by physical action or habitat modification through the erosion and sedimentation of fine material (Wilson and Keddy, 1986).
The fish species composition along the second fish community gradient (PC-II 2) was consistent with the correlations of this gradient with exposure to Lake Ontario. Studies of fish distribution with respect to harsh and sheltered environments are uncommon. In a rare study involving temperate nearshore fishes, Murphy et al. (2011) suggested that sheltered embayments were more suitable for Bluegill growth because they were warmer due to less intrusion of cold water from Lake Ontario. Moreover, they noted that the variability in temperature in some Lake Ontario embayments might impact the spawning and early life stages of Bluegill and similar species resulting in variable year classes. Exposed embayments may be more variable in temperature and water currents (Rao et al., 2009), like rivers. Considering a river as comparable variable habitat, Brown Bullhead, White Sucker, and Smallmouth Bass at the exposed embayment end of this gradient may occupy runs and riffles in rivers, but Bluegill, Black Crappie, Longnos Gar (Lepisosteus osseus) and White Perch at the sheltered embayment end of this gradient occupy pools and quiet water in rivers (Eakins, 2015).
The nearshore fish community in Hamilton Harbour was clearly impaired, and it differed significantly from all other fish communities. Moreover, the relative abundances of target species were an order of magnitude lower in comparison with unimpaired sheltered embayments. Hypereutrophication in Hamilton Harbour has created a suite of habitat impairments including low oxygen, high turbidity, and low aquatic vegetation (O'Connor, 2003) resulting in hyperabundance of species such as Common Carp, Brown Bullhead and Channel Catfish that are tolerant to these conditions. Hyperabundance of Common Carp (Cyprinus carpio) contributes to higher turbidity that in turn, retards restoration of the aquatic macrophytes and zooplankton, and subsequently much or the fish community in Hamilton Harbour wetlands (Lougheed et al., 2004). A barrier to selectively exclude Common Carp from Cootes Paradise, a 250 ha wetlands in Hamilton Harbour, has contributed to an overall decline in Common Carp abundance in Hamilton Harbour (Bowlby et al., 2010). Northern Pike appears to be doing well in Hamilton Harbour due to restoration of spawning and nursery habitat (Theÿsmeÿer et al., 2002) and as observed in this study, due to a lack of competitive piscivores. Most of the target species, including sunfishes (e.g. Bluegill and Pumpkinseed) and piscivores such as Walleye and Largemouth Bass, are depressed in Hamilton Harbour, consistent with habitat impairment. Restoration of Walleye in the Bay of Quinte started after improvements to dissolved oxygen following phosphorus control in waste water (Hurley, 1986). Walleye has been stocked in Hamilton Harbour recently, in large part to increase the abundance of piscivores and provide top-down influence on fish community structure (OMNR, 2014).
Impairment of the nearshore fish community in Toronto Harbour was equivocal. Significant correlations between fish communities in Toronto Harbour and unimpaired exposed embayments suggest no impairment. As well, the relative abundance of six of nine target species were close to or exceeded those in exposed embayments. However, the relative abundance of key species–Rock Bass, Smallmouth Bass and Walleye–in Toronto Harbour was lower than exposed embayments, and Gizzard Shad was hyper-abundant. These key species would be expected to structure fish communities in exposed embayments through community interactions. Smallmouth Bass is the most abundant piscivore, and Rock Bass is the most abundant secondary consumer in unimpaired exposed embayments. The hyper-abundance of Gizzard Shad may reflect the low abundance of piscivores. As well, high abundance of Common Carp may indicate habitat impairment, and a concern is for its potential to impair aquatic macrophytes (Lougheed, 2004). The influence of individual species in structuring the fish communities or indicating habitat impairment was not evaluated in the PCAs, as all species were weighted equally. Alternative analyses provide a weighted approach and complement this study. Hoyle and Yuille (2016) have developed an IBI for nearshore fish communities in Lake Ontario following Minns et al. (1994) but with the trapnet data as used here. The trapnet IBI indicates that Toronto Harbour and Hamilton Harbour are impaired compared with 11 other Lake Ontario embayments and St. Lawrence River locations.
In conclusion, nearshore fish communities of unimpaired Lake Ontario embayments reduce to two clusters that can be used to predict restored fish populations in Hamilton Harbour and Toronto Harbour. Differences in these fish communities are related to indices of exposure of the embayments to exchange of water from Lake Ontario.
We gratefully acknowledge contributions made by Lake Ontario Management Unit staff who have helped collect and process fish community data. Comments from Alex Bowlby, Mike Yuille, Colin Lake and two anonymous reviewers significantly improved this article.
The Province of Ontario supported this project with funds from the Canada Ontario Agreement program.