The reality of environmental dynamics in the Anthropocene is that true reference conditions are often lacking, and many systems are experiencing continuing impacts. For example, many watersheds have experienced multiple types of land-use changes over time, but we have a limited understanding of how subsequent alterations to already impacted ecosystems may affect communities. Working in an agricultural region of Brazil, we test the hypothesis that the degree change to already impacted streams and their watersheds is positively correlated with change in taxonomic and functional structure of fish assemblages. Forty streams were sampled 10 years apart (2003-2013), and instream habitat and watershed land-use were quantified in both periods. We quantified 15 functional traits associated with habitat use and foraging for 54 species and six functional groups were defined using clustering. PCoA was used to quantify environmental distances for each stream between periods using instream, watershed and all variables combined. Jaccard and Bray-Curtis indices were used to compare taxonomic and functional assemblage structure between years. Taxonomic (richness, diversity, evenness) and functional (FRic, FEve, FDiv) changes were calculated by subtracting values in 2013 from values in 2003. To test our hypothesis we used linear regressions between environmental distances and assemblages change.
Results/Conclusions
Contrary to our expectations, changes in taxonomic and functional attributes were not correlated with further degree of environmental change (e.g. replacement of pasture by sugarcane). One exception was a significant but weak (P=0.02, R2adj=0.11) correlation between degree of watershed change and Bray-Curtis functional similarity. Specifically, watershed change correlated with alteration of the distribution of abundances among functional groups, but not the richness or composition of those groups. The lack of significant relationships between changes in almost all taxonomic/functional attributes and environmental change may be due to the degree of biotic homogenization already present in these streams as a consequence of the long and intense process of deforestation to agricultural use. In this context, the degree of environmental change required to induce further changes in assemblages may be much higher than initial stages of alteration. However, assemblages were dynamic in time, but simply not in response to changes to instream habitat and watershed land-use that have been shown to be important in structuring fish assemblages in many systems worldwide. This suggests that assemblage dynamics in response to continuing change in already impacted systems may not follow the same patterns identified in earlier stages of alteration.