Correlation is not causation: a case study of fisheries, trophic state and acidity in Florida (USA) lakes

Environmental pollution studies are often based on observational data and correlative statistics that indicate relationships between pollutants and ecological or human health attributes. Such relationships should be considered preliminary information from which hypotheses can be developed. Often, however, they are used as the basis for inferring cause and effect. That approach can lead to erroneous decisions about pollution control, resource management, or human health standards. This is illustrated using a case study involving fisheries, trophic state attributes (nutrients and algal biomass), and lake water acidity. Fish biomass in Florida lakes is positively correlated with nutrient concentrations and algal biomass, perhaps indicating that fish are controlled by resource availability. However, fish biomass is inversely correlated with hydrogen ion concentration, perhaps indicating that fish decline in acid lakes due to physiological stress or changes in food web structure. Both explanations are ecologically sound, but they point to drastically different resource management strategies: lake fertilization versus neutralization. This issue and others in the environmental pollution arena cannot be fully resolved until experimental research and ecological process studies are coupled with the collection of standard observational data. ©