Abiotic and biotic variable comparison among lakes
Lake trout from L260 had the greatest average total length and weight, followed by lake trout from L223. Although there were significant differences in average length and weight among lake trout populations, condition factor did not significantly differ suggesting a similar degree of health among populations. In terms of zooplankton populations, L260 and L223 contained the most abundant and species rich populations. As zooplankton represent a significant source of food for lake trout, the greatest abundance of zooplankton in L260 and L223 may explain the trend of larger individuals residing in those lakes. Moreover, L373 and L260 contained the most diverse, evenly-distributed zooplankton communities as suggested by Shannon diversity and equitability indices. Diversity and equitability were lower in L223 likely because the zooplankton community here was dominated by rotifer species (Keratella cochlearis , Kellicottia bostoniensis ,Polyarthra remata , Polyarthra vulgaris , Kellicottia longspina ) (Figure S5). High rotifer dominance in lakes has shown to increase bacterial abundance through top-down control of flagellates (Fermani et al., 2013) and, thus, the observed pathogen responses in L223 lake trout may potentially be explained by greater rotifer abundance and, consequently, bacterial abundance. However, more research is needed to establish a causative link among rotifer and bacterial abundances at IISD-ELA.
Average water quality parameters were input into PCA to elucidate similarities and differences among the four lakes based on their water quality characteristics. L260 and L223 were most similar in terms of water quality parameters, whereas L373 and L224 deviated from all other lakes (Figure 6). L260 and L223 had high scores on PC1, which received highest positive loadings from suspended nitrogen, suspended carbon, conductivity, and chlorophyll a. As nitrogen and carbon are important for zooplankton productivity (Brett, Arhonditsis, Chandra, & Kainz, 2012; Loick-Wilde et al., 2016) and chlorophyll-a is an indicator of productivity in aquatic environments, PC1 represents lakes with high productivity. The high positive PC1 scores for L223 and L260 suggest they had higher productivity relative to the other lakes, which is further supported by the fact that L260 and L223 had the greatest zooplankton, chlorophyll-a, and total dissolved nitrogen levels (Table 1). PC1 scores were weakest for L373 and were strongly negative for L224, suggesting L373 and L224 were less productive than L223 and L260. Lake 224 scored strongly on PC2 which had strong positive loadings from pH, alkalinity, and dissolved oxygen (Figure 6). These PCA results further demonstrate that Lake 224 was lower productivity as high pH levels limit phytoplankton growth (C. Y. Chen & Durbin, 1994). Furthermore, the positive loading of dissolved oxygen and negative loading of chlorophyll-a on PC2 is indicative of oligotrophic conditions.
Overall, results of the PCA suggest that L260 and L223 were most similar in terms of water quality characteristics likely driven by their higher productivity (Figure 6). Moreover, L224 and L373 varied from all other lakes, with variability likely being driven by more alkaline conditions in L373 and higher total dissolved phosphorus content in L224 (Figure 6). Unlike L260 and L223, water quality parameters and PCA results suggest that L373 and L224 were lower productivity lakes. Hierarchical clustering analysis confirmed that L223 and L260 were most similar based on water quality whereas L224 and L373 clustered apart from L223 and L260 (Figure S7). The divergence of L224 and L373 from the other lakes is likely primarily a function of water residence time which is driven by volume and catchment area. Thus, the smaller catchment area:volume ratio of L224 and L373 is likely a driver of the lower productivity of these lakes.