The benthic macroinvertebrate community of Lake Ontario was assessed through a lakewide survey in 2008. Diporeia was very rare throughout the lake at all depths in 2008, and only four of 52 locations had densities >100 m−2, all of them at depths >90 m. The maximum density of Diporeia found at any location was at 257 m−2 in 2008, which can be compared to maximum densities of 13,280 m−2 observed in 1994. Lakewide Diporeia abundance declined with an additional order of magnitude from an average of 342 m−2 in 2003 to 21 m−2 in 2008. The Quagga Mussel (D. rostriformis bugensis) dominated the benthic macroinvertebrate community in 2008, comprising over 70% of the density and 98% of the biomass. No Zebra Mussels were identified in the 2008 samples. Quagga Mussels, Oligochaetes and Chironomids were most abundant between 31 and 90 m. Sphaeriids were rare at all depths, but were more abundant at sites deeper than 90 m. Between 2003 and 2008, lakewide Dreissena abundance declined by 43% primarily due to significant declines in the 10–30 m depth region (from 6500 m−2 to 900 m−2). Dreissena did not decline significantly in the 30–90 m or over 90 m depth regions. The 2008 survey revealed a continued decline in Diporeia and Sphaeriid Clams, a replacement of Zebra Mussels by Quagga Mussels, and a decline in Quagga Mussels at depths shallower than 30 m. Oligochaetes and chironomids showed no significant changes since the 1990s.
The continuous introduction of non-native species to the Great Lakes is changing the ecosystem, making ecological monitoring increasingly important. This may be particularly true for the benthic community. The benthos in Lake Ontario's deeper waters was historically dominated by the burrowing Amphipod Diporeia spp., which together with Fingernail Clams (Sphaeriidae), Worms (Oligochaeta) and Midge larvae (Chironomidae), were the main components of the cold-stenotherm macrobenthic community, occupying most of the deeper waters of the Great Lakes (Cook and Johnson, 1974). These organisms are primarily detritivores, dependent upon surface productions (particularly diatom blooms) sinking to the benthic profundal zones.
Since the early 1990s, the non-native Dreissenid Mussels (Dreissena polymorpha and D. rostriformis bugensis, henceforth collectively referred to as Dreissena) have increased dramatically, to the extent that these invasive species became the dominant benthic macroinvertebrates in Lake Ontario by 1995 (Watkins et al., 2007). These Mussels are efficient filter feeders that often form dense, carpet-like beds that cover the bottom substrate. They have the capacity to filter large volumes of water, with an estimated clearance rate of 20 to over 30% of the water column per day for the 30–50 m depth range in Lake Michigan in 2008 (Vanderploeg et al.,2010). These Mussels have also been identified as a major energy sink in Great Lakes ecosystems, shunting nutrients and minerals toward the benthos (Hecky et al.,2004; Nalepa et al., 2009b; Vanderploeg et al.,2010). Bioenergetics models for western Lake Erie proposed by Nalepa et al. (2009b) postulated that much of the energy sequestered by Dreissena will not be utilized by upper trophic levels.
The proliferation of Dreissenid Mussels is closely correlated with the decline of Diporeia in the Great Lakes, although the mechanism involved is not known (Nalepa et al., 2006, 2009a; Watkins et al., 2007). From 1994 to 2003, the populations of Diporeia in Lake Ontario declined by an order of magnitude, from a lakewide average of over 3000 m−2 to less than 400 m−2 (Lozano et al., 2001; Dermott and Geminiuc, 2003; Watkins et al., 2007). Diporeia declined first in shallow water, then in 30–90 m depth region, and finally at depths >90 m, mirroring the range expansions of Quagga Mussels into deeper water (Watkins et al., 2007). By 2003, Diporeia had nearly disappeared at depths shallower than 90 m.
These dramatic changes in the Lake Ontario benthic macroinvertebrate community through 2003 were documented by Watkins et al. (2007). This article extends this time series by analyzing the benthic data from the Lake Ontario Lower Trophic Level Assessment year of 2008 (LOLA, 2008). We were particularly interested in whether the decline in Diporeia and concomitant increase in Quagga Mussels continued, and whether the large changes associated with the proliferations of Quagga Mussels has also affected Fingernail Clams, Oligochaetes and Chironomids. The specific objectives of this research are 1) to assess the status of these five major taxonomic groups in Lake Ontario in 2008 and 2) to investigate any changes in population densities and biomass of these key organisms over time with an emphasis on changes since 2003. Time trends were investigated by comparing results of the LOLA 2008/2009 benthic survey to those of earlier EPA, EMAP, and LOLA studies (Lozano et al., 2001; Dermott and Geminiuc, 2003; Watkins et al., 2007).
The 2008/2009 data were collected using Environment Canada's long-term monitoring plan for Lake Ontario. Benthos samples were collected at sites located along multiple north-south transects of the binational Lake Ontario Lower Aquatic Food Web Study (LOLA) (Figure 1, Table 1). The 2008/2009 LOLA collection was taken over the course of three separate cruises due to time constraints and inclement weather. 44 sites were sampled while on the CCGS Limnos during the week of 18 to 21 August 2008. An additional 3 sites were visited in September 2008 on the CCGS Limnos, and 4 sites were visited in September 2009 on the R/V Lake Guardian.
At each site, three samples were taken with a Ponar grab (area = 0.048 m2). These samples were pooled into a single sample and elutriated using a nylon sleeve with a 500-μm opening (Lozano et al., 2001). The remaining residue was preserved in 5% formalin containing rose Bengal stain. In the laboratory, all organisms were removed from samples under a dissecting microscope and identified. Dreissenids were identified to species, while Oligochaetes, Chironomids, Amphipods and Sphaeriids were identified at least to genus. Counts of Oligochaetes were based on identified heads and excluded fragments.
Samples were classified by lake basin (West: longitude >78.6°; Central: longitude between 77.6° and 78.6°; East: longitude <77.5°) and depth (<30 m, 30–90 m and >90 m). Two-way ANOVAs were employed to test for differences in population densities of all taxonomic groups among lake basins and depth regions, as well as examining differences in abundance between the 2003 and 2008/2009 sampling periods. Significant differences were subsequently evaluated by post-hoc multiple range tests (Tukey's HSD). In all statistical analyses, statements of significance are at an error level of α = 0.05. To account for the heteroscedasticity of benthic abundance estimates, a natural logarithmic transformation (loge(#/m2+1)) was used for all statistical tests. Distributional maps for abundance of Diporeia and Dreissena were created with Surfer 9.0 (Golden Software) for the years between 1994 and 2008. Sample collection and processing was similar to previous years. Although all benthic macroinvertebrates found in the samples were counted in 2008/2009, only six stations were counted for Oligochaetes, Chironomids, and Sphaeriids in 2003. Therefore, we excluded data for these groups from 2003, as 6 stations are not sufficient for reasonable comparisons among years. Dreissena were not counted in 1994.
To estimate dry weight (henceforth referred to as “biomass”), lengths of Mussels collected in 2008 were measured using the Zebra Scan IDL program (Nalepa et al., 2008). This program uses a scanned image to measure each organism. Subsamples were taken in cases where there were over 200 Dreissenids collected at a given site. Using these length measurements, an average shell-free dry weight per Mussel was calculated using the methods in Nalepa et al. (2008). Lengths of 50 Diporeia, 161 Oligochaetes, and 111 Chironomids from selected sites were measured using Image-Pro. All Oligochaetes and Chironomids scanned were from the 2008 collection. Diporeia measured were collected in 1994, as there were too few present in the 2008 samples to produce a useful estimate. An average dry weight of Diporeia, Chironomids, and Oligochaetes were calculated based on length to dry weight conversions in Nalepa and Quigley (1980). The average weights of Dreissena, Diporeia, Oligochaetes, and chironomids were 3.34, 0.74, 0.09 and 0.63 mg dry weight, respectively. For sphaeriids, we used the value 0.26 mg dry weight, as all sphaeriids collected in 2008/2009 were Pisidium spp. Nalepa and Quigley (1980) estimated the average weight of Pisidium to be 0.26 mg.
Biomass for each depth interval was multiplied by the proportion of the area within a specific basin and depth interval. Weights from each depth interval were summed to arrive at a lake-wide weighted average. Average lakewide biomasses were calculated for of Diporeia, Dreissena, Oligochaetes, Sphaeriids and Chironomids for each year between 1994 and 2008. Biomass was not calculated for Dreissena in 1994 since the Mussels were not saved for that year. In 2003, dry weight of Oligochaetes, Sphaeriids and Chironomids were estimated from the six of the stations sampled that year (station 6, 9, 41, 63, 71 and 81, locations in Table 1).
The Quagga Mussel was the most abundant species found in the 2008/2009 survey, comprising 70% of all organisms sampled lakewide and 98% of the biomass. All Dreissenid Mussels identified were Quagga Mussels. This species was significantly more abundant at depths between 30–90 m, irrespective of basin, with similar average densities of less than 1000 m−2 in the shallower and deeper regions (Figure 1, Tables 2 and 4). There were no significant differences observed in population density between the three basins (Table 3). Between 2003 and 2008, there was a highly significant 86% decrease in the average density of Mussels in the 10–30 m depth interval, with population densities showing no significant change in the 30–90 m and >90 m zones (Table 4). Dreissena was already the dominant macroinvertebrate by biomass in water less than 90 m by 1997, and Quagga Mussels also exceeded Diporeia biomass in deeper water in 2003. By 2008, Quagga Mussels constituted over 90% of the biomass in all depth regions (Table 5, Figure 2). The time trend in lakewide biomass of the combined Dreissena group is not signficant since 1997 (Figure 2, R2 = 0.35, N = 5, P = 0.291).
Very few Diporeia were present in the 2008/2009 samples, with average lakewide densities of 0, 5 and 45 organisms m−2 for 0–30 m, 30–90 m, and >90 m depth intervals, respectively (Table 2). Diporeia was only found at one station shallower than 90 m and at only 7 out of 19 deep stations, mostly in the western part of the lake (Table 1). Densities over 100 m−2 were only found at 4 stations (Station 14, 19, 52 and 55; Table 1). Average Diporeia densities at depths >90 m have declined by an order of magnitude since 2003 (Table 4, Figure 3). Between 1994 and 2008, the average lakewide biomass of Diporeia fell from 3.1 g m−2 to 0.01 g m−2, representing a 99.9% decrease (Figure 2, Table 5). The decline in Diporeia biomass over time is significant (R2 = 0.75, N = 6, P = 0.025).
Oligochaetes were the second most abundant taxonomic group lakewide, accounting for 24% of all benthic macroinvertebrates collected during the 2008/2009 sampling period. Twenty-three taxonomic groups were identified, of which Lumbriculidae, Tubificidae, Potomothrix vejdovskyi and Stylodrilus heringianus were the most common (see the Appendix, available in the online supplementary information). Population densities averaged 1032 m−2 and ranged from 135 to 2,824 m−2. They were also the most ubiquitous taxonomic group, found at 50 of the 51 sampling locations. Densities of greater than 1,000 m−2 were found at 33% of the sites, with the maximum density of 4,419 m2 found in the western basin at a depth of 11 m. Densities were >500 m−2 at 29 of the 31 sites with depths <90 m (Table 1, Figure 1). Oligochaetes were also significantly more abundant in the western basin (Table 5). Overall, lakewide Oligochaete population densities have not changed significantly from 1994 to 2008 (R2 = 0.32, N = 5, P = 0.32).
Chironomids were significantly more abundant in the 10–30 m and 30–90 m depth intervals, exceeding 200 m−2 at 11 of 32 locations at these depths (Table 1, Figure 1). Eighteen taxa were identified, with Heterotrissocladius subpilosus and Micropsectra sp. were most common (Appendix). Population densities were highest at shallow locations located near Toronto, ON; Hamilton, ON; and Cobourg, ON. Chironomids were significantly more abundant in the western basin, with an average density of 308 m−2. Mean densities in the central and eastern basins were comparable at 145 and 99 organisms m−2, respectively (Table 1). There is no significant time trend in chironomid biomass (Figure 2, Table 5; R2 = 0.52 N = 5, P = 0.17).
Sphaeriids (all in the genus Pisidium) were relatively scarce in the 2008/2009 samples, and were only present in densities greater than 100 m−2 at 7 of the 51 sites (Table 1). They were marginally more abundant within the >90 m depth interval (Table 3). There were no significant differences in population densities amongst the three basins. Sphaeriid biomass has continued to decline since 1994 (Table 5, Figure 2) and this decline is marginally significant (R2 = 0.76, N = 5, P = 0.056).
The restructuring of Lake Ontario's benthic macroinvertebrate community continued throughout the first decade of the twenty-first century. The benthic ecosystem of Lake Ontario is now dominated by the Quagga Mussel, an ecosystem engineer that has profound effects on water clarity and bottom structure (Higgins and VanderZanden, 2010; Nalepa and Schlosser, 2013). Although D. polymorpha was the dominant Mussel species in the early 1990s, Quagga Mussels gradually replaced Zebra Mussels and became the dominant species by 2003 (Watkins et al., 2007). This process appears to be complete, as we did not identify any Zebra Mussels in the 2008/2009 samples. However, Zebra Mussels do persist in the lake on rocky, nearshore substrates that cannot be sampled with Ponar grabs (Karatayev et al., 2013). Possible reasons for the greater success of Quagga Mussels include a higher assimilation efficiency (Baldwin et al., 2002), a lower respiration rate (Roe and MacIsaac, 1997), higher filtration rates during the warmer months (Diggins, 2001) and a faster growth rate in the presence of predators (Naddafi and Rudstam, 2014a). D. r. bugensis also possess longer siphons and byssal threads than D. polymorpha, which enables them to thrive in soft sediments found in the deeper waters of Lake Ontario (Nalepa et al., 2009a; Dermott et al., 1998).
Population densities of Dreissena declined significantly from 2003 to 2008 in the 10–30 m depth interval, which contrasts with the large population increases observed in previous sampling years. Lakewide densities declined by 45%, although the decline was not significant in water deeper than 30 m. A population decline of an invasive species after an initial population explosion immediately after establishment is a common observation in invasion biology. For Quagga Mussels, this could be caused by increased predation by round Goby (Naddafi and Rudstam, 2014b). Round Goby populations have increased dramatically in Lake Ontario between 2003 and 2008, and are primarily found in water shallower than 30 m during the summer and fall (Weidel et al., 2013).
The severe decline in population densities of Diporeia continues to be a major concern. In 2008/2009, they were absent from almost all stations shallower than 90 m, and the lakewide average was as low as 16 m−2. There were only four sampling locations at which they were present in somewhat higher numbers (>100 m−2), and all four of those sites were >90 m deep (Table 1). Even in deeper waters, Diporeia densities have declined substantially, from 625 to 45 m−2 from 2003 to 2008. If this trend continues, this highly important component of the benthos and a critical link between the benthic and pelagic communities (Dermott, 2001) could be extirpated from Lake Ontario in the near future.
The cause of the declining numbers of Diporeia has yet to be determined. The strong correlation in both space and time of the increase in Quagga Mussels and decrease in Diporeia is strongly suggestive of a negative effect of Quagga Mussels on Diporeia (Nalepa et al., 2006). Competition with Quagga Mussels over diminishing food sources, such as has been proposed with the decline of the spring diatom bloom in Lake Michigan (Kerfoot et al., 2010; Vanderploeg et al., 2010), is a possible mechanism. In addition to monopolizing food resources, Dreissena also affects water chemistry by assimilating large amounts of calcium to construct their shells. Barbiero et al. (2006) found that calcium concentrations in Lake Ontario's offshore waters decreased by 4–5 mg l−1, and that this coincides with a three-fold reduction in turbidity values due to a decrease in whiting events. Whiting events transport organic material to the bottom as calcium particles form on an organic nucleus (Hodell and Shelske, 1998), so a decline in whiting events could be a contributing factor to the decline in Diporeia (Watkins et al., 2013). Toxins associated with Mussel pseudofeces are another possible direct link between the Quagga Mussel increase and the Diporeia decline (Nalepa et al., 2009a, 2010; Dermott et al., 2005a).
Although the decline of Diporeia appears to be connected with the expansion of Quagga Mussels, there are alternative hypotheses. Diporeia began to disappear prior to the appearance of Quagga Mussels in the deeper areas of both Lake Ontario (Dermott, 2001; Watkins et al., 2007, 2013) and Lake Michigan (Nalepa et al., 2009a). Barbiero et al. (2011) reported declining numbers of Diporeia in Lake Huron despite much lower abundances of Mussels there, and Mussels coexist with Diporeia in several smaller lakes (Dermott et al., 2005b; Watkins et al., 2012), suggesting that Dreissenids may not be directly implicated in the Diporeia decline. Interestingly, Diporeia from declining populations do not show decreased condition (Nalepa et al., 2009a), and stressors applied to the animals give the same metabolic response regardless of the presence of Quagga Mussels (Maity et al. 2013). Furthermore, the decline in the spring bloom observed in Lake Michigan (Mida et al., 2010; Vanderploeg et al., 2010; Kerfoot et al., 2010) did not occur in Lake Ontario, even though Diporeia declined in both lakes (Watkins et al., 2013). Overall productivity of the lake as measured by total phosphorus content in the surface water has declined since the 1960s (Dove, 2009), although no further decline in total phosphorus has occurred in Lake Ontario since 1995 (Holeck et al., 2015), and algal species composition in 2008 indicates more eutrophic conditions (Munawar et al., 2015). Additionally, there is a productive deep chlorophyll layer in the lake, and this production has probably increased over time (Watkins et al., 2015). Other possibilities include viral diseases or some other unknown pathogen (Hewson et al., 2013). As long as the exact mechanism remains unknown, the coupling of the decline in Diporeia and the increase in Quagga Mussel remains a correlation.
Interestingly, Oligochaete and Chironomid densities have not declined significantly in response to the proliferation of Dreissena since 1994. The Mussel beds provide good-quality habitat for these organisms because they are spatially heterogeneous and complex, providing them with cover from predators (Mayer et al., 2001) and supply these detritivores with a food source in the form of pseudofeces and associated bacteria (Ricciardi, 1994). Experiments conducted by Ricciardi et al. in the St. Lawrence River indicated a strong preference for Dreissena beds amongst several different benthic taxa, particularly deposit feeders (Ricciardi et al., 1997). Long-term studies in southwestern Lake Ontario indicated a strong correlation between Dreissena densities and non-dreissenid macroinvertebrate abundances and diversity on both cobble and reef substrates (Haynes et al., 2005). This phenomenon has also been documented in Lake Simcoe, Ontario (Ozersky et al., 2011), Oneida Lake, NY (Mayer et al., 2002) and Lake Erie (DeVanna et al., 2011). In addition, Oligochaetes are ubiquitous in Great Lakes benthic communities and are more tolerant of habitat degradation than other taxa, and therefore may be less affected by ecological and environmental changes (Krieger, 1984). It is worth noting that the decline in food resources suggested as a cause for the decline in Diporeia does not result in similar declines in Oligocheates, Chironomids and other deposit feeders between 1994 and 2008, raising additional questions about decline in food resources as a cause for the decline of Diporeia in Lake Ontario (Watkins et al., 2007, 2012).
We detected a marginally significant decline in Fingernail Clams (Sphaeriids) in Lake Ontario. As of 2008, these bivalves comprised a very small part of the overall benthic biomass. Sphaeriids are filter feeders and require calcium to build their shells, and thus compete for both food and calcium with the burgeoning Quagga Mussel population. This suggests a plausible mechanism for the coupling of the increase in Quagga Mussels and the decline in sphaeriids. This decline was not detectable in the data available up to 1999 (Watkins et al., 2007).
To conclude, the results of this study illustrate the restructuring of the benthic community of Lake Ontario. Mussels now completely dominate the benthic biomass with associated ecological consequences (Karatayev et al., 2002; Higgins and VanderZanden, 2010; Mayer et al., 2014). This increase may also be the cause for the decline in both Diporeia and Fingernail Clams. The complete replacement of Zebra Mussels by Quagga Mussels and the decline of Dreissena in the 10–30 meter depth interval are also important trends that warrant continued long-term monitoring.
The authors would like to thank the captains and crews of the Canadian Coast Guard Ship Limnos and the US EPA R/V Lake Guardian and the Research Support Branch of Environment Canada in Burlington for assisting in sample collection. Melissa Clouse and Kerrin Mabrey assisted in processing samples and identifying and measuring organisms. Identifications for Chironomids and Oligochates in 2008 were made by EcoAnalysts, Inc. Fred Luckey, Alice Dove, and Glenn Warren provided input and advice.
Financial assistance was provided by an Interagency Agreement between the US EPA and DOC/NOAA (DW13942146-01). This project is part of a larger program that is investigating the lower trophic level of Lake Ontario (LOLA) in support of the US-Canada Lake Ontario Lakewide Management Plan LaMP. LaMP partners include federal, state, and provincial government agencies charged with environmental quality and natural resource management responsibilities for the lake.
Supplemental data for this article can be accessed on the publisher's website.