Selection on fruit traits is mediated by the interplay between frugivorous birds, fruit flies, parasitoid wasps and seed‐dispersing ants

Abstract

Every organism on Earth must cope with a multitude of species interactions both directly and indirectly throughout its life cycle. However, how selection from multiple species occupying different trophic levels affects diffuse mutualisms has received little attention. As a result, how a given species amalgamates the combined effects of selection from multiple mutualists and antagonists to enhance its own fitness remains little understood. We investigated how multispecies interactions (frugivorous birds, ants, fruit flies and parasitoid wasps) generate selection on fruit traits in a seed dispersal mutualism. We used structural equation models to assess whether seed dispersers (frugivorous birds and ants) exerted phenotypic selection on fruit and seed traits in the spiny hackberry (Celtis ehrenbergiana), a fleshy-fruited tree, and how these selection regimes were influenced by fruit fly infestation and wasp parasitoidism levels. Birds exerted negative correlational selection on the combination of fruit crop size and mean seed weight, favouring either large crops with small seeds or small crops with large seeds. Parasitoids selected plants with higher fruit fly infestation levels, and fruit flies exerted positive directional selection on fruit size, which was positively correlated with seed weight. Therefore, higher parasitoidism indirectly correlated with higher plant fitness through increased bird fruit removal. In addition, ants exerted negative directional selection on mean seed weight. Our results show that strong selection on phenotypic traits may still arise in perceived diffuse species interactions. Overall, we emphasize the need to consider diverse direct and indirect partners to achieve a better understanding of the mechanisms driving phenotypic trait evolution in multispecies interactions.

The relationship between host plant traits and biodiversity across three sympatric seed‐feeding tri‐trophic systems in a tropical region of Brazil

“…The sampling effort was sufficient to estimate all herbivore species, but rarefaction curves suggested that further sampling may be required for a complete assessment of parasitoid species associated with S. multijuga and I. vera (Figure 2). Further, we established the relationship between resources and insects by using path analysis, which is a powerful tool to assess direct and indirect connections in complex, tri‐trophic systems (Beduschi et al, 2015; Nascimento et al, 2015; Palacio et al, 2020). We found that the number of seeds per fruit was more important for both the abundance and richness of herbivores and parasitoids than the seed and fruit weight, which had a negative or nonsignificant relationship.…”

On the adequacy of fruit removal as a proxy for fitness in studies of bird‐mediated phenotypic selection

“…Third, both components are inherently linked, as fruits enclose seeds and fruits must be pecked or swallowed before seeds can be dispersed, so the contribution of fruit removal to plant fitness is expected to be significant (Nakashizuka et al, 1995; Izhaki, 2002; Snell et al, 2019). This is most evident in the case of species with a fixed number of seeds per fruit, for which the number of fruits removed often scales linearly with the number of seeds dispersed (e.g., Sobral et al, 2013; Palacio et al, 2020). By contrast, seed removal is expected to be more decoupled from fruit removal in multi‐seeded species with a variable number of seeds per fruit (e.g., many species produce fruits that develop acropetally, where seed number per fruit changes drastically from bottom to top; Diggle, 1995) as the same number of fruits removed in two plants may lead to different numbers of seeds removed.…”

A conceptual framework linking pollen limitation and seed dispersal effectiveness

“…Additionally, variation in the disperser traits can also affect their responses to dispersal units' traits, since low and high-quality dispersers can differ in their foraging preferences (e.g. Russo 2003;Leal et al 2014b;Palacio et al 2020). Therefore, by influencing the dispersal units' traits, PL can indirectly affect the assemblage of dispersers interacting with the dispersal units and, consequently, the quality of seed dispersal received by pollen-limited plants.…”