Most anurans undergo a major ontogenetic niche shift, as tadpoles they occupy aquatic habitats and then terrestrial habitats as adults. Thus, considering that habitats occupied by tadpoles are determined by the breeding choices of adults, we hypothesized that: 1) environmental filters have a stronger effect on adults than on tadpoles; 2) adult traits (e.g., presence of adhesive disc, reproductive behavior) are more associated with extrinsic characteristics of breeding habitats (e.g., canopy cover), and tadpole traits (e.g., body size, fin height) are more associated with intrinsic characteristics (e.g., aquatic vegetation); 3) phylogeny is a proxy for trait diversification in communities of both tadpoles and adults. We tested whether phylogenetic and trait diversification were different than expected by chance across breeding habitats along the ontogenetic gradient. We also investigated the relationship between phylogeny, morphological and life-history traits, environment, and space using an extension of the RLQ ordination.
Results/Conclusions
We found that phylogenetic diversity in adults and both phylogenetic and trait diversification in tadpoles were randomly distributed across breeding habitats. Additionally, trait diversification in adults was more clustered in breeding habitats than expected by chance, indicating that adults are regulated by environmental filters. Traits of adults, which were significantly correlated with the environment, exhibited a strong phylogenetic signal. Phylogeny was not a good proxy for trait diversification in tadpoles as it was for adults. The first axis of RLQ explained 90.7% of the total variation for adults and 72.9% for tadpoles. Among adult anurans, explosive breeders such as Mycrohylidae occurred in large ephemeral habitats with a high abundance of predators. Prolonged breeders such as Hylidae were associated with long-hydroperiod habitats and high canopy cover. Tadpoles with relatively small bodies were associated with lotic habitats, high canopy cover, and low aquatic vegetation. Conversely, tadpoles with larger fins, larger caudal muscles, and larger eyes were associated with lentic habitats and high aquatic vegetation. As expected, habitat filtering was stronger for adults than for tadpoles, which suggested that tadpoles should be more affected at the microhabitat scale. The amount of invertebrate predators, for instance, was correlated only with adults, which indicates that adults may detect safe aquatic conditions for tadpoles. We conclude that phylogenetic and trait diversification for adults and tadpoles does not seem to be affected in the same scale and by the same environmental characteristics.