Abstract
Chlorophyll a is an important pigment used by algae to absorb solar energy for photosynthesis. Because chlorophyll a is used by all algae (including cyanobacteria) for photosynthesis, it is often measured as an index of algal abundance. Although chlorophyll a is an imperfect representation of algal abundance, other methods for quantifying algal abundance are time consuming, expensive, and still imperfect. Chlorophyll a can be quantified using a probe that can be deployed for weeks or months at a time, generating high frequency chlorophyll a data alongside other water quality data collected with deployed sensors. These other water quality metrics include measurements of several variables that could represent hypothesized drivers of variation in algae abundance. Here, we used water quality and chlorophyll a data compiled for the purpose of predicting harmful algal blooms in the Illinois River to identify the strength of relationships between chlorophyll a concentration in the water column and other water quality variables that may have a mechanistic link to algal abundance. We used statistical models and causal modeling to evaluate the relationships between environmental data and chlorophyll a concentration. We found the highest concentrations of chlorophyll a occurred when discharge was low and temperature was high. Relationships were weak to moderate in all modeling approaches that related environmental data to chlorophyll a concentration, even when accounting for optimized lag times. The direction and magnitude of statistical associations between chlorophyll a and environmental data also varied by site. From a causal modeling perspective, the available data may poorly represent the hypothesized mechanisms, or we may be missing causal drivers of variation in chlorophyll a concentration.