Determining trophic state in experimental lakes

Carpenter et al. (1998) investigated how nutrients (i.e., phosphorus [P] input), dissolved organic carbon (DOC), and grazing are related to phytoplankton biomass and production in four manipulated lakes intensely sampled during five summers. Multivariate probability distributions of algal biomass and production were constructed for varying data on P input rates, DOC, and grazer length, where DOC was used as a surrogate for humic and fulvic acids (also called colored DOC by Carpenter et al.). It is important to study the joined effects of these three predictor variables to improve predictability of lake trophic state (Nürnberg and Shaw 1999). However, as happens frequently in science, specialization has progressed so far that related disciplines use different expressions and methods, hindering the exchange of ideas and comparison of results. In this context I do not mean to diminish the value of the study by Carpenter et al. (1998) on the effect of trophic interactions and biological components on nutrient–phytoplankton relationships. My purpose is to caution against overgeneralization of these results and to suggest the usage of more appropriate measures. In particular, Carpenter et al. (1998) used several concepts that may be misleading: (1) areal instead of volumetric measures of algal biomass and productivity in correlations with volumetric measures of DOC concentrations, (2) nutrient load instead of lake concentration to assess the effects of P on lake trophic state, and (3) transfer of results from small, potentially meromictic lakes that experienced various types of manipulations to trophic state interactions in natural lakes.