Diversidade e composição funcional de comunidades fitoplanctônicas: determinantes ambientais e incertezas nos índices

Abstract

The functional diversity and functional composition can be used as a good strategy to predict the ecosystems dynamics as well as the factors affecting its stability. Environmental variables on the local scale and landscape are responsible for different structuring processes within communities, including the functional traits composition. In addition, these can also influence the functional diversity values. Currently, there are numerous functional indices available in the literature, and these are distinct as the methodology used in its calculation. Thus, we analyzed the influence of environmental variables at the local scale and landscape on the functional composition of phytoplankton and also investigate the uncertainties on the use of functional diversity indices based on dendrograms and their linkage methods over of sampling points distributed in an environmental gradient. We sampled 29 floodplain lakes in the middle Araguaia river and tributaries. These communities were characterized functionally by measuring 10 functional traits divided into three categories (i.e. morphological, physiological and behavioral). Then, these were used to calculate four functional diversity indices based on dendrograms (RGF - Richness Functional Groups; FD - Functional Diversity, MPD - Average Distance between Pairs of Species and MNTD - Average Distance from the Nearest Taxon) through seven different linkage methods. We found that the environmental variables at the local scale are the main determinants for the physiological and behavioral traits, and the landscape variables have no significant influence to the phytoplankton functional composition. On the other hand, the functional diversity indices proved to be different from each other, both the values obtained within the same community as well as by the environmental factors which are predicted. In addition, no exist significant differences between the linkage methods found and out the absence of a variation pattern along the sampling units. These results indicate that the physical, chemical and biological characteristics the habitats act to determine functional traits should occur. Accordingly, changes (especially anthropogenic) on local characteristics of the floodplain lakes must be avoided since this can affect the functional composition. Moreover, that better attention should be dedicated to the choice of the indices.