Leveraging water quality monitoring data to guide algaecide use in aquaculture ponds

Aquaculture has increased in importance to feed a growing world population given that some commercial fishing efforts may threaten natural fish populations. Although hypereutrophic conditions are common in aquaculture systems, the resulting blooms of phytoplankton can be harmful to fish because of toxin production and lead to tainted products through the release of off-flavor compounds. Commercial farmers routinely use applications of various algaecides to control algal blooms; however, such use carries possible adverse effects to non-target organisms including fish, zooplankton, and beneficial algae, including chlorophytes. This study aims to determine the chemical or physical water quality parameters that are associated with toxic algal blooms by leveraging three years of water quality monitoring data collected across 21 commercial catfish aquaculture ponds in western Alabama, USA. The response variables were used to indicate eutrophic and harmful conditions, including two algal pigments (chlorophyll- a (all phytoplankton) and phycocyanin (cyanobacteria)) and the hepatotoxin, microcystin, produced by some cyanobacteria. Initially, response trends over time were evaluated. Next, linear mixed effects models guided by AIC and all subsets model selection were used to determine which model (e.g., chemical or physical) best explained each response variable. Long-term monitoring of aquaculture ponds showed clear seasonal trends reinforced by the model selection. The chemical parameter top model explained phycocyanin and microcystin, which included total and dissolved forms of nitrogen and phosphorus, while chlorophyll- a was best explained by physical parameters, like light and dissolved oxygen. The model highlights important parameters that negatively influence water quality in these hypereutrophic systems, which will help guide algaecide treatment periods to manage toxigenic and off-flavor producing cyanobacteria.