Statistical Optimization of Microcystis Growth and Microcystin Production in Natural Phytoplankton Community using Microcosm Bioassays

Abstract

Eutrophication of freshwater lakes and reservoirs causing toxic cyanobacterial blooms has become a global health concern. Commonly used approaches of statistical modelling have not fully captured the complex response of cyanobacterial biomass and mi- crocystin concentration in response to stochastic variation of ambient conditions. This study applied statistically-based experimental design to screen out critical environmental factors, and to investigate the interactive effects of total nitrogen (TN), total phosphorus (TP), temperature (T) and light intensity (L) on Microcystis (MA) growth and microcystin (MC) production in the natural phytoplankton com- munity. A batch of central composite designed nutrient enrichment bioassays (CCD-NEBs) were conducted in an ambient-controlled microcosm based on pneumatic annular flume system. Second-order polynomial CCD regression models predicting MA density and MC concentration were established and validated. MC concentration was found positively related to MA dominance. The interactive effects of TN-TP and TN-L on both MA density and MC concentration were determined to be more significant than other factors. Ranges of the optimum TN/TP ratio for MA growth and MC production under various environmental conditions were defined as 9.5 ~ 10.2 and 9.9 ~ 10.4, respectively. The ranges of nutrient thresholds for MA density of 2.00 × 104 cells mL-1 and MC concentration of 1.0 μg L-1 under different temperature and light conditions were evaluated, respectively. This study revealed that nutrient thresholds for MA blooming events and cyanotoxins safety guideline were fluctuant under lake-dependant conditions of temperature and light intensity, and the optimum TN/TP mass ratios should be alerted for the eutrophication management of blooming formation and microcystin safety.