Abstract
Strategies that focus on reducing nutrient loading to freshwater lakes have historically been successful in improving water quality by curbing large phytoplankton blooms. However, as waters warm, little is known about the resultant phytoplankton physiology and ensuing perturbations in the food web that may occur. Here, we designed a mesocosm experiment to investigate the impact of warming water on phytoplankton physiology and further validate a previously developed, coarse-grained model that predicts the key aspects of phytoplankton physiology, including elemental stoichiometry and macromolecular allocation, across varying temperatures. We found that higher temperatures double the maximum cellular density (cells L(-1)) of phytoplankton, suggesting that high temperature stimulates cell division over maximizing carbon storage. Also, the cells in warmer waters dedicate fewer resources to proteins and RNA production, leading to higher fractions of carbon allocated to storage. This work illustrates the potential impact warming waters may have on the ecosystem, as higher fractions of carbohydrates are often associated with less nutritious food for higher trophic levels.IMPORTANCEWe take a novel approach to investigating the impact of warming on phytoplankton physiology by utilizing mesocosms and a coarse-grained cellular model. Previous work in this field tends to use idealized laboratory experiments, mesocosms, or models alone. By synthesizing model and mesocosm results, we test the model's ability to capture physiology in semi-natural environments. We conducted this experiment under phosphorus limitation and saw high cell densities in the heated, treatment tanks. Thus, warming waters may negate some successful management practices that curb eutrophication. With increased temperatures, we also observed increased N:P values in both the experimental and model results, which may be due to the combined effects of a lack of P storage, fewer enzymes required, and a corresponding decrease in RNA production. Our model predictions closely aligned to mesocosm observations, suggesting the capability of our model to represent lower trophic organisms in ecosystem models.