Water level fluctuations and the ecosystem functioning of lakes

Abstract

Hydrological regimes are key drivers of productivity and structure in freshwater ecosystems but are increasingly impacted by human activity. Using 17 published food web models of 13 African lakes as a case study, we explored relationships between seasonal and interannual water level fluctuations and 15 attributes related to ecosystem function. We interpreted our results in the context of Odum's ecosystem maturity hypothesis, as systems with higher magnitude fluctuations may be kept at an earlier maturity stage than those that are relatively stable. The data we compiled indicate that long-term changes in the hydrological regimes of African lakes have already taken place. We used Least Absolute Shrinkage and Selection Operator (LASSO) regression to examine relationships between ecosystem attributes and seven physical characteristics. Of these characteristics, interannual water level fluctuation magnitude was the most frequently retained predictor in the regression models. Our results indicate that interannual water level fluctuations are positively correlated with primary and overall production, but negatively correlated with fish diversity, transfer efficiency, and food chain length. These trends are opposite those expected with increasing ecosystem maturity. Interestingly, we found seasonal water level fluctuations to be positively correlated with biomass. An increase in standing biomass is generally associated with more mature ecosystems. However, we found that less production and biomass occurred at high trophic levels in highly fluctuating compared to relatively stable systems. This synthesis provides evidence that water level fluctuations are a key process influencing ecosystem structure and function in lakes.