Current Status of Mathematical Models for Population Dynamics of Prymnesium parvum in a Texas Reservoir1

Grover, James P., Jason W. Baker, Daniel L. Roelke, and Bryan W. Brooks, 2010. Current Status of Mathematical Models for Population Dynamics of Prymnesium parvum in a Texas Reservoir. Journal of the American Water Resources Association (JAWRA) 46(1):92-107. DOI: 10.1111/j.1752-1688.2009.00393.x Abstract: Blooms of the harmful alga Prymnesium parvum have apparently increased in frequency in inland waters of the United States, especially in western Texas. A suite of mathematical models was developed based on a chemostat (or continuously stirred tank reactor) framework, and calibrated with data from Lake Granbury, Texas. Inputs included data on flows, salinity, irradiance, temperature, zooplankton grazing, and nutrients. Parameterization incorporated recent laboratory studies relating the specific growth rate of P. parvum to such factors. Models differed in the number of algal populations competing with P. parvum, and whether competition occurred only by consumption of shared nutrients, or additionally through production of an allelopathic chemical by one of the populations, parameterized as cyanobacteria. Uncalibrated models did not reproduce the observed seasonal dynamics of P. parvum in Lake Granbury, which displayed a maximum population in late February during a prolonged bloom in cooler weather, and reduced abundance in summer. Sensitivity analyses suggested two modifications leading to predictions that better resembled observations. The first modification greatly reduces the optimal temperature for growth of P. parvum, an approach that disagrees with laboratory experiments indicating a strong potential for growth at temperatures above 20°C. The second modification increases the growth rate of P. parvum at all temperatures, in models including cyanobacterial allelopathy. Despite these adjustments, calibrated models did not faithfully simulate all features of the seasonal dynamics of P. parvum.