Estimating the probability of exceeding elevated pH values critical to fish populations in a hypereutrophic lake

Abstract

Current eutrophication models typically are used to predict seasonal mean conditions. However, the risk of summer fish kills in hypereutrophic lakes is likely to be more closely dependent on periodic extreme events, such as potentially lethal peaks in pH driven by algal photosynthesis. In hypereutrophic Upper Klamath and Agency lakes, Oregon, peak summertime pH values frequently exceed critical levels that can reduce fish growth and survival (pH > 9.50, a likely sublethal tolerance limit for two resident endangered fish species). We developed two empirical models, one parametric and one nonparametric, that predict the likelihood of exceeding user-defined critical values of pH from concentrations of chlorophyll a in these lakes. Separate models were derived to incorporate seasonal dynamics and differences between the two lakes, and the behavior of these models was tested under four different critical pH scenarios. Both parametric and nonparametric models performed similarly, suggesting that management efforts to reduce chlorophyll a in these lakes from 200 to 100 µg·L-1 should decrease the probability of exceeding pH 9.5 by 45%. We suggest that this general approach potentially can be applied to the management of fish populations in other hypereutrophic lakes as well.