Intra-annual variation of phytoplankton community responses to factorial N, P, and CO2 enrichment in a temperate mesotrophic lake

Across primary producer communities in different lakes, nitrogen (N) and phosphorus (P) can exhibit many different patterns of limitation. Here, we look at the intra-annual variability of these patterns in a single lake. Furthermore, we investigate whether a third resource, carbon dioxide (CO2) can have significant effects on phytoplankton biomass and community composition. We performed five in situ lacustrine mesocosm experiments at different times of the year. In each experiment, we had a factorial design with two levels of N, P and CO2 enrichment (no enrichment or double lake concentrations for N and P and atmospheric (400 ppm) and c. 1,000 ppm for CO2) resulting in a total of eight treatments. Mesocosms of c. 1,600 L were suspended in a temperate, mesotrophic lake (Lac Hertel, Canada). Each experiment lasted 2 weeks and chlorophyll a biomass, coarse chemotaxonomic community composition (measured using fluorometry), and several environmental variables were recorded at a minimum of four time points. We found that the limiting, synergistic, and community composition effects of N and P varied between experiments. TN:TP ratios explained, in part, some of this variability, along with insolation and water temperature. Despite relatively high levels of CO2 in the control mesocosms, we found a constant synergistic effect of CO2 with N. In combination with the synergistic effect of P with N found in some experiments, this provides support for CO2 as one of the multiple limiting resources in nutrient-enriched systems. This finding could have implications for eutrophic lakes exposed to increasing concentrations of CO2. We also found that the effects of CO2 on community composition varied intra-annually. Thus, we conclude that generalised predictions about the effect of CO2 on community composition at a coarse chemotaxonomic scale are unlikely to hold, but predictions specific to season and system are likely to be reliable.