Ecosystem functions and densities of contributing functional groups respond in a different way to chemical stress.

Understanding whether and to what extent ecosystem functions respond to chemicals is a major challenge in environmental toxicology. The available data gathered by ecosystem-level experiments (micro- and mesocosms) often describe the responses of taxa densities to stress. However, whether these responses are proportional to the responses of associated ecosystem functions to stress is unclear. By combining a carbon budget modeling technique with data from a standardized microcosm experiment with a known community composition, we quantified three ecosystem functions (net primary production [NPP], net mesozooplankton production [NZP], and net bacterial production [NBP]) at three Cu concentrations, with a control. Changes of these ecosystem functions with increasing chemical concentrations were not always proportional to the Cu effects on the densities of the contributing functional groups. For example, Cu treatments decreased mesozooplankton density by 100-fold and increased phytoplankton density 10- to 100-fold while increasing NZP and leaving NPP unaltered. However, in contrast, Cu affected microzooplankton and the associated function (NBP) in a comparable way. We illustrate that differences in the response of phytoplankton/mesozooplankton densities and the associated ecosystem functions to stress occur because functional rates (e.g., photosynthesis rates/ingestion rates) vary among Cu treatments and in time. These variations could be explained by food web ecology but not by direct Cu effects, indicating that ecology may be a useful basis for understanding environmental effects of stressors.