On the role of natural water level fluctuation in structuring littoral benthic macroinvertebrate community composition in lakes

We used traditional hydrologic endpoints and extracted novel water-level fluctuation (WLF) characteristics through principal components analysis (PCA), to determine the relationship between WLF and rocky littoral benthic macroinvertebrate (BMI) richness in 16 boreal lakes. Yearly WLF amplitude (maximum minus minimum water levels) ranged from 35.9 cm to 157.5 cm. Using PCA we derived a new variable D80-D210 (31 March minus 01 August) as a surrogate for change in mean water level and potential habitat squeeze (loss). Analyses of BMI richness with several physicochemical variables, including water quality (8), habitat variability (4), lake and basin morphology (6), land classification (28), water temperature (8), hydrology (9), and PCA axes (10) resulted in only three significant relationships. We found a classic species-area relationship, as BMI richness increases with increasing lake area (r(2) = 0.38(linear), r(2) = 0.69(unimodal)). Similarly, as littoral slope increases macroinvertebrate richness decreases (r(2) = 0.32(linear)). Most importantly, lower water levels, quantified using D80-D210, have higher macroinvertebrate richness (r(2) = 0.38(linear)). Together these results suggest that a habitat squeeze in littoral areas is the direct result of lower mean water levels and that relatively small changes in natural WLF can be associated with changes in BMI communities.