Abstract
The distributions of chrysophyte stomatocysts in the sediments of 221 lakes from four study areas located in British Columbia, Canada (two datasets), New York, U.S.A., and south central Siberia were examined. The combined lake set spanned a wide range of chemical and morphometric variables. Approximately 400 cyst morphotypes were identified, from which we selected sixty‐three common types for further analysis. Canonical ordination (detrended canonical correspondence analysis (DCCA) and constrained DCCA) showed that four environmental variables (pH, conductivity, depth and total phosphorus (TP) ) were significant (Monte Carlo permutation test, P =0.01 for all four variables) and independent determinants of the cyst distributions. Weighted averaging (WA) inference models were developed for each variable. The best models, in terms of their r 2 , were developed for pH and conductivity ( r 2 =0.78 and 0.73, respectively). The models for depth and TP were less robust ( r 2 =0.47 and 0.46, respectively). We discuss the relative importance of conductivity and salinity, lakewater acidity and metal concentrations, lake morphometry and trophic status as determinants of chrysophyte population distributions, and possible causal mechanisms for each variable. We identify assemblages of cysts that appear to be associated with six different groups of lakes: (1) very acidic (pH 8.5) lakes, including sub‐ and hyposaline lakes. Finally, we suggest further possible avenues for research that will improve the usefulness of chrysophyte stomatocysts as paleolimnological indicators.