Andrew C. Kindig

Responses of blue-green and green algae to streptomycin in unialgal and paired culture☆

Abstract Streptomycin sulfate prevented growth of six blue-green algae at concentrations (0.09 to 0.86 mg/l) substantially lower than needed to prevent growth of 7 of 8 green algae tested. Chlorella vulgaris, Scenedesmus obliquus and Ulothrix sp. grew in active streptomycin concentrations less than 21 mg/l, while Chlamydomonas reinhardtii growth was prevented at concentrations of 0.66 mg/l. Algal growth in sublethal concentrations of streptomycin was slowed or delayed, and the maximum density attained by several species was decreased. These data suggest that attempts to select ‘representative’ species for toxicity testing may have mixed success: any blue-green alga would be a suitable representative of Cyanophyta with respect to streptomycin, but no single green alga would adequately represent the wide range of response to streptomycin seen among the Chlorophyta species tested. To test for interactions between competition and chemical inhibition, paired cultures of S. obliquus with Anabaena cylindrica, Ankistrodesmus sp., C. vulgaris and Selenastrum capricornutum were observed with and without 6.6 mg/l streptomycin. The outcome of competition was consistent with predictions based on single-species assays, where major differences in chemical sensitivity were present, although significant interactions were observed. S. obliquus populations when treated with streptomycin were competitively superior to A. cylindrica and S. capricornutum. Significant interactions between competition and streptomycin treatement were also observed for paired cultures of C. vulgaris and S. obliquus; neither population fit the model of additive effects of chemical inhibition and competition.

Stepwise iterative calibration of a multi-species phytoplankton-zooplankton simulation model using laboratory data

Abstract A multi-species phytoplankton-zooplankton simulation model was rigorously calibrated using laboratory data. The model is a system of ordinary differential equations describing the dynamics of eight functional groups of phytoplankton, seven groups of zooplankton, and nitrogen and phosphorus forms. Processes represented in the model include nutrient uptake, photosynthesis, respiration, excretion, egestion, mortality, and nutrient recycling. A stepwise iterative calibration procedure was used with data from laboratory experiments involving various combinations of biota mixes and two initial nutrient concentrations. Calibration proceeded in four steps: (1) calibration of five of the eight phytoplankton groups to single and paired species experiments performed under both initial nutrient concentrations, (2) calibration of the remaining phytoplankton groups to single-species growth experiments and to an ‘ungrazed’ experiment (all phytoplankton without zooplankton), and (3 and 4) calibration of the zooplankton and nutrient recycling portions of the model to data from a full microcosm experiment (all phytoplankton and zooplankton together). Two constraints were imposed during calibration: (1) at each step, values of parameters determined during previous steps were not allowed to be varied from their calibrated values, and (2) a single set of parameter values had to be found for each functional group that provided satisfactory fits to all experiments involving that functional group. Based on both graphical comparisons and statistical goodness-of-fit measures, satisfactory model fits to data were obtained. The importance of representing intracellular nutrient pools for phytoplankton and size structure for zooplankton are demonstrated, and the benefits and drawbacks of using data from controlled laboratory and field systems for model development and calibration are discussed.

Modeling the effect of algal biomass on multispecies aquatic microcosms response to copper toxicity

Abstract A model was developed to simulate the effects of copper on an aquatic microcosm consisting of 10 phytoplankton species and 5 zooplankton species grown in a defined medium. The copper toxicity equations in the model were based on the results of single species toxicity tests at different initial cell densities. Model output was evaluated by comparing predictions with results of replicated [3] 35 day microcosm experiments having copper added at 3 different times (days 7, 14 or 21) at 500 ppb concentration. These experiments offered a wide range of plankton and nutrient conditions at the time of copper addition. The baseline for these experiments was set by a control experiment, with no copper addition and a model, MICMOD, developed to simulate microcosm behavior. Two toxicity models were tested; the ‘neutral’ model defined copper toxicity to plankton as depending only on the dissolved copper concentration in the medium, while the ‘biomass’ model also decreased toxicity in proportion to algal biomass. Graphical and statistical comparisons showed the toxicity model including algal biomass to better fit the data. Both the experimental results and the model suggest that the effects of copper may be strongly influenced by the density and species composition of the biota and the related differences in water chemistry at the time of copper addition. The apparent importance of algal biomass to copper toxicity may be due to (1) changing copper availability either through direct absorption or adsorption, (2) production of chelates by the algae which complex copper in less toxic forms or (3) changing the pH which affects copper ionization and particularly the concentration of the cupric ion (Cu2+) form, which is highly toxic.

Modeling the direct and indirect effects of streptomycin in aquatic microcosms

This paper compares the predictions of a multispecies, multi-trophic level aquatic ecosystem model with data from control and streptomycin treated laboratory aquatic microcosms. Preliminary calibration of streptomycin effect is based on single species bioassay experiments, while further calibration, maintaining the same relative species sensitivity, is done on the full microcosm to obtain improved model fits. Other model parameters were calibrated based on values in the literature and on other microcosm experiments. Statistical correlation and t-tests were used, along with graphs, to compare model results with data and overall agreement between the model and data was acceptable. The major discrepancy between the model and data was in the occurrence in the experiment of a very large Scenedesmus bloom whose magnitude was not replicated by the model. This was linked to differences in timing of the post-treatment drop in Daphnia and a larger than expected ammonia release. Experimental work was begun, primarly to investigate the possible source of the ammonia release, directed toward testing hypotheses proposed to explain the source of the model-data discrepancy. The model was shown to be useful in suggesting possible hypotheses to explain model-data discrepancies, to show how different processes dominate in the control and streptomycin treated microcosm and to improve our understanding of the effect of streptomycin on multispecies aquatic systems.

Potential Use of Microcosms to Assess Survival, Efficacy, and Environmental Safety of Genetically Engineered Microorganisms

“Standardized Aquatic Microcosms” for assessing the ecological effects of new chemicals are currently being interlaboratory tested, and could be adapted for use with genetically engineered organisms. They have the potential to evaluate the ability of new microorganisms to survive under conditions of competition and predation, to degrade test substrates in the presence of naturally abundant substrates, to determine pathogenicity to several species of algae and animals, and to disrupt ecological cycles. The 63-day microcosm protocol was developed and tested against a variety of chemicals.