Effects of plant diversity and primary production on decomposition in tropical streams.
biodiversity, decomposition, nutrient cycling, aquatic hyphomycetes, priming effect, tropical streams.
Tropical streams are ecosystems with a high richness in their riparian forests, represented mostly by a high number of rare species. However, these ecosystems have suffered a considerable and rapid loss of species diversity due to anthropic disturbances. Rare plant species are more likely to be lost due to the smaller size of their populations, and so it is often these species that drive such diversity changes in tropical streams. However, the dominant plant species are usually the only ones included in experiments evaluating the consequences of the loss of diversity, and therefore the effect of the loss of rare plant species on these ecosystems is still poorly understood. Importantly, the presence of these species in plant litter mixtures has the potential to accelerate the litter decomposition, as their unique functional traits can lead to increased functional diversity of litter mixtures, which typically accelerates this process. Another factor that has not yet been investigated in tropical streams that can accelerate the litter decomposition is the priming effect – the acceleration of the recalcitrant litter decomposition in the presence of the labile carbon of algae. Studies of the priming effect in aquatic ecosystems have focused on laboratory experiments or mesocosms, and therefore evidence of a priming effect in natural systems is rare. Thus, the objective of this thesis was to evaluate the importance of rare plant species and the priming effect for the litter decomposition in tropical streams. To investigate the importance of rare plant species for the litter decomposition in tropical streams, we carried out field experiments in a stream of the Cerrado biome simulating scenarios of loss of rare plant species of different functional types (i.e., with functional traits similar or dissimilar to those of dominant species): an experiment taking in consideration the proportion of plant litter species entering the stream, and another experiment where we used the same biomass for all species in litter mixtures. The experiments were designed considering the traits of communities of low diversity in natural ecosystems, commonly characterized by the persistence of dominant plant species. In chapter 1 we demonstrated that the loss of rare plant species reduced the functional diversity of litter mixtures and consequently reduced decomposition and two other important related processes – nitrogen (N) loss and fungal biomass production –, via the resource dissimilarity effect. Furthermore, we observed that when only similar rare species were present in litter mixtures, the processes were more reduced than in the scenario with only dissimilar rare species (except for N loss). In chapter 2 we showed that the loss of rare plant species reduced the biomass of fungi, considered the main decomposer organisms in tropical streams, but did not affect other fungal parameters evaluated (i.e., aquatic hyphomycetes sporulation rate, diversity and taxonomic composition). Furthermore, we observed that the reduction in fungal biomass was greater in litter mixtures with proportions of natural species than in mixtures with the same biomass for all species, indicating that evenness modulated the effect of rare species loss on fungal biomass. To investigate the importance of the priming effect for the litter decomposition in tropical streams, in chapter 3 we carried out a field experiment using artificial shading in streams of the Cerrado biome. Although we found only evidence of reduced litter decomposition with increasing algal labile carbon (negative priming effect), we demonstrate that the priming effect plays an important role in the litter decomposition mediated by decomposer microorganisms and shredder invertebrates in these ecosystems. Our findings highlight the relevance of rare plant species (especially functionally dissimilar rare species), of maintenance the natural proportion of species in riparian forests, and of the priming effect for the functioning of tropical streams. Therefore, this thesis has the potential to contribute to practical questions of the restoration of riparian zones (mainly the Cerrado), since it indicates that the species and aspects mentioned above should be considered in future restoration projects.