Are Assemblages of Aquatic-Breeding Anurans (Amphibia) Niches Structured or Neutral? Muryllo Melo 1,3 , Fernanda Fava 1 , Hugo B. Pinto 2 , and Fausto Nomura 1 1 Laborat orio de Herpetologia e Comportamento Animal, Departamento de Ecologia, Instituto de Ci^ encias Biol ogicas, Universidade Federal de Goi as, CP 131, CEP 74000-970, Goi^ ania, GO, Brazil 2 RAN – Centro Nacional de Pesquisa e Conservac ß~ ao de R epteis e Anf  ıbios, Setor Leste Universit ario, Rua 229, n° 95, CEP: 74.605.090, Goi^ ania, GO, Brazil ABSTRACT Local niche-based processes and dispersal are important determinants of assemblage composition and species diversity. However, there is no consensus about the relative importance of niche and spatial processes to explain the distribution of anuran species in tropical sys- tems. In our study, we analyzed the niche and neutral effects on anuran assemblages and found that biotic interactions were a predictor of assemblage structure. The Eltonian concept of niche was the best predictor for the structure of aquatic-breeding anuran assemblages, as species tended to co-occur more often than would be expected by chance. We suggest that the lack of environmental effect could be explained by differences in the pattern of movement between arboreal and non-arboreal anurans. Once there is a reduction in the num- ber of arboreal anurans in open areas, the importance of habitat heterogeneity to explain assemblage composition should decrease. The lack of correlation between the spatial component in our model and species composition is evidence that spatial processes, such as migration, did not play a major role in structuring local assemblages. Anurans are generally assumed as having poor dispersal ability, yet this assumption is not true for all anuran species. We suggest that future studies should include key behavioral traits, such as site fidelity and homing behavior, as these traits can represent the dispersal abilities of anurans and dispersal ability seems to be important when we try to predict patterns of anuran distribution. Key words : Cerrado; community structure; co-occurrence; Eltonian niche; Grinnellian niche; null models; redundancy analysis. MUCH OF COMMUNITY ECOLOGY IN THE LAST FIVE DECADES has focused on the analysis and interpretation of the relationship between species’ niche and communities to understand which mechanisms drive species distribution and structure ecological communities (Hutchinson 1959, Gotelli & McCabe 2002, Kneitel & Chase 2004). This effort led to the idea that communities result from short-term ecological interactions and/or long-term evolutionary processes that operate at different timescales, affect- ing the number and identity of co-occurring species (Ricklefs 2004). Thus, we could expect researchers to describe how an organism is affected by the distribution of resources, predators, or competitors (Schoener 1974) and how these organisms can alter these same biotic and abiotic factors (Schoener 1974, Lomo- lino & Brown 1998). Grinnell (1917) described the niche concept as the local abi- otic requirements for species persistence; this was later expanded to characterize the environmental conditions found throughout the range of the species (Leibold 1995). At the local scale, this idea can be related to the effect of habitat heterogeneity, which predicts that more complex environments could hold a higher number of species (Pianka 1966, Ricklefs 2004). A biotic compo- nent was introduced later by Elton (1927), in which communities are organized by biotic processes emphasising the role of the interactions among species (Woodward 1983, McCarthy 1997). Finally, it has also been suggested that spatially structured factors influencing the movements of organisms could also drive impor- tant ecological processes such as immigration and dispersal (Ros- indell et al. 2011). This was one foundation of the neutral theory of biodiversity and biogeography (Hubbell 2001), whereby the observed community composition would not be dependent on species’ traits or environmental tolerance, but by the chance an individual arrives at a site suitable for colonization (Hubbell 2001). Many recent studies have evaluated how anuran assemblages are structured, but results to date have been inconsistent (e.g., Parris 2004, Ernst & R€ odel 2008, Both et al. 2009, Nomura et al. 2012). One suite of studies has concluded that anuran assem- blages are structured mainly by local niche processes, with biotic interactions such as predation (e.g., Morin 1983, 1987) and com- petition (e.g., Gascon 1995) playing a major role. More recently, the importance of biotic processes has been reduced and abiotic factors, such as suitability of ponds (Zimmerman & Bierregaard 1986), hydroperiod (Baber et al. 2004), and local heterogeneity (Ernst & R€ odel 2006, Bastazini et al. 2007) have emerged as the most important predictors of anuran species distribution. In con- trast, other studies suggest an absence of biotic or abiotic effects on the structure of anuran assemblages (Gotelli & McCabe 2002, Santos et al. 2007) and considered that spatial effects such as the distribution of environmental variables (Ernst & R€ odel 2008, Received 12 June 2013; revision accepted 8 April 2014. 3 Corresponding author; e-mail: muryllomelo@gmail.com 608 ª 2014 The Association for Tropical Biology and Conservation BIOTROPICA 46(5): 608–614 2014 10.1111/btp.12130