Do aquatic macrophytes co-occur randomly? An analysis of null models in a tropical floodplain

One of the main issues in community ecology is the detection of structure and the identification of its related causes. In this study, co-occurrence null models were used to identify possible spatio-temporal patterns in the assemblage of aquatic macrophytes in the Upper Parana River floodplain. The samples were obtained through the Long Term Ecological Research (LTER) Program at two different grains: (1) a coarser spatial grain in January and August 2001 (entire floodplain lagoons); (2) and a finer spatial grain in November 2006 (1 m(2) quadrats). The study was conducted in 36 lagoons, both connected and disconnected to the main river channel, located in the sub-basins of the Baia, Ivinheima and Parana rivers. Two null models of species co-occurrence, the C-Score and Checkerboard indices, were used to test the null hypothesis of random structure of the aquatic macrophyte assemblages. The null models showed that the aquatic macrophyte assemblages were spatially structured in the distinct spatial grains. However, despite this general pattern, macrophyte assemblages are organized differently depending on the degree of connectivity, seasonal period and, at a finer grain, depth. Species co-occurrences were random in the disconnected lagoons during flood periods, in deep zones of the lagoons of the Baia River and in the shallow littoral zone in the lagoons of the Parana River. Analysis of the patterns of co-occurrence indicated that competition and/or habitat preferences are probably important influences on the nonrandom structure of assemblages. However, we suppose that at least three important factors (disturbances by water level fluctuation, dispersion and facilitation) counteract potential effects of competition in specific situations, leading macrophyte assemblages to assume random structure.