Meu passado me condena! A autocorrelação temporal do fitoplâncton em um sistema lacustre neotropical

Abstract

Plankton metacommunity structure is known to respond to both local factors (i.e., biotic and abiotic) and patch connectivity. Many studies are seeking to quantify the influence of local and regional factors in determining the abundance of coexisting populations through the adequacy of metacommunity models such as species-sorting and neutral to real data. Most studies tested hypotheses in this regard from a spatial perspective, where community structures in neighboring patches were evaluated, as well as important environmental variables, and the importance of patches proximity and environmental similarity in predicting community structure is quantified by applying variance partition methods. However, as aquatic environments are not stable over time, the trade-off between species-sorting and neutral dynamics can also be tested from a temporal perspective. Therefore, this study aimed to quantify the influence of environmental and stochastic limitations of the phytoplankton population on a set of unconnected natural lakes, where the community was sampled monthly for eight years. Our hypothesis is that although its high turnover rate, phytoplankton populations have an expressive temporal autocorrelation, that may be interpreted as neutral dynamics. This interpretation comes from the fact that a population can persist in the environment, even if this environment may not be suitable for the species needs. Thus, it is expected that the autocorrelation can be as important as, or even more important than the environmental effect to explain community structure. To test our hypothesis, we performed the total variance partition separately for each of the 21 most abundant phytoplankton species, assessing the importance of the temporal (i.e. abundance in the last month), spatial (i.e. neighboring lakes), and environmental dimensions. Our hypothesis was corroborated, with the temporal dimension having the greatest importance over all other dimensions. We suggest that these organisms are “dispersing temporally”, with strong source-sink relationships over time, outperforming environmental filters.