Diversity gradients of Neotropical freshwater fish: evidence of multiple underlying factors in human‐modified systems. (1st April 2016)

• Record Type: Journal Article
• Authors: Bailly, Dayani; Cassemiro, Fernanda A. S.; Winemiller, Kirk O.; Diniz‐Filho, José Alexandre F.; Agostinho, Angelo A.
• Is Part Of: Journal of biogeography. Volume 43:Number 8(2016:Aug.)
• Page Start: 1679
• Page End: 1689
• Publication Date: 2016-04-01
• Subjects: abundance -- anthropogenic impacts -- diversity hypotheses -- latitudinal diversity gradient -- metabolic theory of ecology -- reservoir age -- reservoirs -- spatial autocorrelation -- species richness; Biogeography -- Periodicals; 578.09
• DOI: 10.1111/jbi.12749
• Languages: English
• ISSNs: 0305-0270

Abstract

Abstract: Aim: Contemporary patterns of species diversity are the result of a hierarchy of natural processes and modern anthropogenic influences. However, studies of these patterns in human‐modified systems from a macroecological perspective are lacking. Considering that fish assemblages in reservoirs reflect both long‐term evolutionary responses of species and shorter term responses to anthropogenic stressors, we employed a multi‐hypothesis approach using different magnitudes of predictors to analyse the processes that drive fish diversity in reservoirs at a broad spatial scale. Location: Brazil. Methods: We derived species richness from an extensive database of fish inhabiting Neotropical reservoirs, and using multiple regression analysis, we tested seven hypotheses that link species richness to continuous variables associated with regional, local and population components. We analysed the spatial structure using Moran's I autocorrelation coefficients and used spatial eigenvector mapping to explicitly account for the spatial component when testing the hypotheses by multiple regressions. Partial regressions were performed to map the relative contributions of the different components in explaining species richness. Results: Predictors related to six hypotheses were retained in the best‐fit models, and our data supported the species–energy, metabolic, species–area, species–distance from the source and time of habitat alteration hypotheses. However, the predictions of the population abundance hypothesis (PAH) were not supported by the data. The shared effects of the different components explained the greatest proportion of the variation in species richness, indicating that distinct mechanisms related to these alternative hypotheses interact or are not spatially independent. Main conclusions: Our findings indicate that the latitudinal diversity gradient typically displayed by freshwater fish is maintained in human‐modified habitats, such as reservoirs, and that multiple mechanisms drive fish diversity in reservoirs over large spatial scales. The lack of support for the PAH implies that mechanisms structuring diversity patterns can be influenced by anthropogenic stressors.