Frugivorous fish play a prominent role in seed dispersal and reproductive dynamics of plant communities in riparian and floodplain habitats of tropical regions worldwide. In Neotropical wetlands, many plant species have fleshy fruits and synchronize their fruiting with the flood season, when fruit-eating fish forage in forest and savannahs for periods of up to 7 months. We conducted a comprehensive analysis to examine the evolutionary origin of fish-fruit interactions, describe fruit traits associated with seed dispersal and seed predation, and assess the influence of fish size on the effectiveness of seed dispersal by fish (ichthyochory). To date, 62 studies have documented 566 species of fruits and seeds from 82 plant families in the diets of 69 Neotropical fish species. Fish interactions with flowering plants are likely to be as old as 70 million years in the Neotropics, pre-dating most modern bird-fruit and mammal-fruit interactions, and contributing to long-distance seed dispersal and possibly the radiation of early angiosperms. Ichthyochory occurs across the angiosperm phylogeny, and is more frequent among advanced eudicots. Numerous fish species are capable of dispersing small seeds, but only a limited number of species can disperse large seeds. The size of dispersed seeds and the probability of seed dispersal both increase with fish size. Large-bodied species are the most effective seed dispersal agents and remain the primary target of fishing activities in the Neotropics. Thus, conservation efforts should focus on these species to ensure continuity of plant recruitment dynamics and maintenance of plant diversity in riparian and floodplain ecosystems.

Competition plays a central role in the maintenance of biodiversity. A backbone of classic niche theory is that local coexistence of competitors is favoured by the contraction or divergence of species' niches. However, this effect should depend on the diversity of resources available in the local environment, particularly when resources vary in multiple ecological dimensions. Here, we investigated how available resource breadth (i.e. prey diversity) and competition together shape multidimensional niche variation (between and within individuals) and interspecific niche overlap in 42 populations of congeneric tropical frog species. We modelled realized niches in two key trophic dimensions (prey size and carbon stable isotopes) and sampled available food resources to quantify two-dimensional resource breadth. We found a 14-fold variation in multidimensional population niche width across populations, most of which was accounted for by within-individual diet variation. This striking variation was predicted by an interaction whereby individual niche breadth increased with resource breadth and decreased with the number of congeneric competitors. These ecological gradients also interact to influence the degree of niche overlap between species, which surprisingly decreased with population total niche width, providing novel insights on how similar species can coexist in local communities. Together, our results emphasize that patterns of exploitation of resources in multiple dimensions are driven by both competitive interactions and extrinsic factors such as local resource breadth.

Although neglected by classic niche theory, individual variation is now recognized as a prevalent phenomenon in nature with evolutionary and ecological relevance. Recent theory suggests that differences in individual variation across competitors can affect species coexistence and community patterns. However, the degree of individual variation is flexible across wild populations and we still know little about the ecological drivers of this variation across populations of single species and, especially, across coexisting species. Here, we aimed to (a) elucidate the major drivers of individual niche variation in natural communities and (b) to determine how consistent this variation is across coexisting species and communities. We analysed natural patterns of individual-level niche variation in four species of coexisting generalist frogs across a wide range of tropical communities. Specifically, we used gut contents and stable isotopes (δ C and δ N) from frog species and their prey to quantify individual niche specialization. Then, we combined data on local community structure, availability of prey, phylogenetic relationships and predator-prey size models to test how this variation is related to four ecological factors which are predicted to be key drivers of individual specialization: intraspecific competition, interspecific competition, ecological opportunity (i.e., diversity of resources) and predation. We found that the degree of individual trophic specialization varied by up to ninefold across populations within the same species. This sizable variation in trophic specialization across populations was at least partially explained by gradients of density of competitors (both conspecifics and heterospecifics) and intraguild predation. However, the specific relationships between individual specialization and these ecological gradients were strongly species-specific. As consequences, the identity of the species with more individual variation changed among sites and there was typically no spatial correlation in the degree of individual specialization across coexisting species. Our results show that individual niche specialization within and across species can be strongly context-dependent and that hierarchies of individual variation among coexisting species are not necessarily consistent across communities. Recent theory suggests that this pattern could lead to concurrent changes in competitive interactions across sites and thereby could play a key role in species coexistence at the landscape level. Our results suggest that individual variation across and within coexisting species has the potential to affect not only species coexistence at local communities, but also regional diversity patterns.