Appendices included. "Project number 16010 DQB." "EPA 16010-DQB-10-71." "October 1971." Cover title. "SERL report no. 71-6." Includes bibliographical references (pages 169-177).

Batch and continuous flow (chemostat) assays were investigated as part of a joint industry-government sponsored, multilaboratory effort to develop a standardized algal assay procedure for nutrient level assessment. Assays were conducted with Selenastrum capricornutum as a standard assay organism. Batch culture assays were found to have a lower level of precision than chemostat assays in the assessment of growth response as a function of nutrient concentration. The biomass parameter, maximum cell concentration, X, of the batch assay generally responded to the nutrient concentration of the sample; however, the chemostat biomass parameter, steady state cell concentration, Xb1s, always was found to be proportional to the nutrient concentration of the samples. The results of spiking tests with batch assays generally were inconclusive with respect to identification of the growth rate limiting nutrient whereas the results of spiking tests with chemostats indicated clearly the growth rate limiting nutrient. It is recommended that batch type algal assays be used only for crude screening or routine monitoring purposes and that the chemostat should be used for the quantitative assessment of the algal growth supporting properties of waters as well as for the development of kinetic descriptions for nuisance algae and the rate limiting nutrients of concern. A kinetic description of Selenastrum capricornutum indicated a low half saturation constant, Ks, (the concentration of nutrient supporting one-half the maximum growth rate) of about 5 [mu]g P/ for phosphate phosphorus and a yield coefficient, Y, that varied as a function of growth rate. A theoretical model was proposed and evaluated which describes the varying yield coefficient (the result of "excess" uptake) as a function of the growth rate (mean cell residence time). The function was verified experimentally at a very high statistical confidence level. The significance of these findings and their application to the practical problem of eutrophication assessment is presented.