Analysis of phytoplankton responses to water chemistry dynamics in a moderately eutrophic north Texas reservoir

Canonical correspondence analysis (CCA) was used to explore relationships between measured environmental variables and in situ phytoplankton communities in a moderately eutrophic North Texas Reservoir. Soluble-reactive phosphate-phosphorus (SRP), nitrate-nitrogen (NO₃), nitrite-nitrogen (NO₂), pH, total suspended solids, and turbidity each explained significant (p $\le$ 0.05) directions of variation in the distribution of phytoplankton taxa. A series of partial, constrained CCA analyses, run with the first axis constrained to each of the environmental variables in turn, showed that concentrations of soluble-reactive phosphate-phosphorus (SRP) and nitrate-nitrogen (NO ₃-N) best explained variation in phytoplankton communities. Weighted averaging regression techniques were then used to develop transfer functions to infer (SRP) and (NO₃-N) concentrations from phytoplankton assemblages. Transfer function models showed relatively good correlations between inferred and measured environmental variables in the study lake. Application of the models to phytoplankton assemblages from six other regional water bodies, however, showed only moderate correlations between inferred and measured environmental variables.