Patterns in phytoplankton taxonomic composition across temperate lakes of differing nutrient status

We describe the relationships between summer average total phosphorus (TP) and the biomass of six major phytoplankton taxonomic groups from 91 north temperate lakes. Both regression and a locally weighted smoothing (LOWESS) analyses show that all groups increase with TP but over different nutrient ranges. At nutrient extremes, i.e. very low and high levels of TP, the few taxonomic groups that dominate total biomass are generally dissimilar. In oligotrophic and eutrophic lakes, most group biomass curves show corresponding increases or decreases, but different rates of change. The curves converge at intermediate TP levels (-10-30 p,g liter-l) where there is increased equitability among algal groups. In highly eutrophic lakes, the slope of the total biomass curve decreases, as do all the relative abundances of all groups except blue-greens and diatoms. We conclude that the curvilinearity of the TP-total biomass curve is not attributable to a single taxonomic group, because all groups show some nonlinearity in relation to TP We suggest that morphological diversity, differential herbivory, and, in particular, mixing regime may explain some of the-observed patterns.There is a well-established, positive relationship between nutrient loading and productivity in lakes (e.g. Schindler 1978), which is most readily apparent as increasing phytoplankton biomass with eutrophication (Nicholls and Dillon 1978). Growth of phytoplankton may be limited by one of several nutrients (e.g. phosphorus, nitrogen, carbon, silica) in a manner that has been related to cellular stochiometric ratios (e.g. Hecky and Kilham 1988), but in the majority of temperate zone lakes, annual productivity is most frequently phosphorus limited (e.g. Smith 1982). Numerous studies have shown that phytoplankton taxonomic composition and species diversity change with increasing nutrient levels (e.g. LaZerte and Watson 198 1; Smith 1990) and that these changes are related to differences among taxa in nutrient uptake, storage, growth, and loss rates (Kalff and Knoechel 1978; Reynolds 1984). In temperatezone lakes, oligotrophic systems support minimal phytoplankton biomass with low species diversity and are generally dominated by nanoflagellates belonging to the Chrysophyceae and Cryptophyta or by cyanobacterial or chlorophyte picoplankton (e.g. Pick and Caron 1987; Willen et al. 1990). Moderate enrichment results in higher and more seasonally variable levels of phytoplankton biomass, along with increased taxonomic diversity (Sommer et al. 1986). In these mesotrophic systems, most algal taxonomic groups are represented over the growing season, especially Bacillariophyceae (diatoms), Chlorophyta (green algae), Cryptophyta, and Dinophyta, as well as Cyanobacteria and Chrysophyceae (e.g. Rosen 1981; Eloranta 1986). Eutrophic and hypereutrophic lakes sustain very high average algal biomass often I Present address: Iowa State University, Department of Animal Ecology, 124 Science II, Ames, Iowa 50011-3221. Acknowledgments This study was supported by Natural Sciences and Engineering Research Council of Canada operating grants to E.M. and J.A.D. dominated by very few taxa, usually Cyanobacteria, diatoms, or in some water-bodies, chlorococcales or dinoflagellates (e.g. Paerl 1988b; Padisak and Dokulil 1994; Jensen et al. 1994). Surprisingly, there are few quantitative comparisons of changes in the average biomass of major phytoplankton taxa with nutrient levels among lakes, although some studies have measured changes in the average proportions (%), or dominance, of some algal divisions (Smith 1990; Duarte et al. 1992; Chow-Frazer et al. 1994). Shifts in phytoplankton taxonomic structure with enrichment have been documented by extensive single sample surveys of lakes (e.g. Rosen 1981; Eloranta 1986; Willen et al. 1990) and have been qualitatively described in excellent reviews of general successional dynamics (e.g. Reynolds 1984; Hecky and Kilham 1988) and of individual taxonomic groups (e.g. Paerl 1988a; Sandgren 1988; Willen 1991). These shifts have been successfully predicted for individual species on a local scale (i.e. within lakes) using resource-competition models (e.g. Tilman 1982; Sommer 1993) and more complex mechanistic models (e.g. Smith et al. 1987; Varis 1993). Furthermore, several studies have empirically described the increase of cyanobacterial biomass with total phosphorus in north temperate (Smith 1986; McQueen and Lean 1987) and subtropical lakes (Canfield et al. 1989); however, similar quantitative relationships for other major groups have not, to our knowledge, been derived. In previous work, we and other authors have shown that algal biomass measured as either chlorophyll a (Chl a) or algal biovolume varies nonlinearly with total phosphorus (TP) (McCauley et al. 1989; Prairie et al. 1989; Watson et al. 1992), and we suggested that this curvilinear&y reflects nonuniform growth and loss rates of one or more components of the total phytoplankton community. We tested the idea that this nonlinearity can be explained by differences in the responses to enrichment of two functional size fractions with high and low susceptibility to zooplankton grazing