Estrutura de comunidade de fitoplâncton no rio intermitente Cruxati, Ceará, Brasil

Abstract

The term Metacommunity can be understood as a set of local communities linked by the dispersal of multiple potentially interactive species. This theoretical framework helps studies that seek to understand how local and regional communities are structured. Among its approaches, there is the mechanistic approach, which synthesizes classical theories of communities in four descriptor models of the processes that act in this structuring: neutral (neutral), patch dynamics, mass effects and species allocation (species sorting). Based on the theoretical framework of metacommunities, our objectives were (i) to understand the role of environmental and spatial variables in the structuring of the phytoplankton metacommunity of the Cruxati River (ii) to identify which of the descriptor models act in the structuring of these communities. We hypothesized that the structure (composition and richness) of phytoplankton species varies according to the temporal dynamics of intermittent rivers. We observed that the species richness was higher in the Pre-Dry period, where there was a connection between the points, but the species composition did not change between the hydrological periods, since the PERMANOVA result showed that there was no significant difference in the composition of the metacommunity when compared the two connectivity scenarios between the water bodies (F = 1.1307; p = 0.3223), but showed a direct relationship with the salinity and conductivity variables, as both composition and richness decreased with increasing values of these parameters. When these variables were grouped with others, the set of environmental variables formed did not significantly explain the variation of species (p = 0.162). On the other hand, spatial variables proved to be effective in structuring this community, reaching to explain significantly (p = 0.001) 9% (R² = 0.09) of the metacommunity composition. Despite this significance, most of the explanation for our results was attributed to residuals, that is, to variables that were not analyzed. Thus, our metacommunity of study appears to be structured in a stochastic way, following the definitions of the neutral model, but we also observe other models acting, even if on a smaller scale, such as patch dynamics and species allocation. Phytoplankton proved to be a good bioindicator of models in metacommunities, due to the representative amount of existing species, as well as their variation in community structure already observed when related to environmental and spatial filters, but the number of predictors proved to be small for us to assess. definitively on the descriptor models. We recommend that more ecological variables and interactions, as well as a functional approach, be used in future studies.