| Sumario: | Studying interactions among species and their multiple forms and outcomes has become a central question of ecological research. Mutualisms have changed the way ecologists understand the functioning of ecosystems. Generalised mutualistic networks, especially among free-living species, have been observed to include highly heterogeneous assemblages, in which most species establish few interactions while a minority of species establish a wide range of interactions. These highly connected species are referred to as super-generalists. They are characterised for connecting otherwise isolated groups of interacting species or modules and for providing stronger cohesion to ecological networks. Their characteristics make them fundamental for local biodiversity and the structure and stability of interactions. In this PhD thesis we aim to better comprehend the super-generalist strategy and the mutualistic interactions at the individual level. For this purpose, we use as a model organism Pistacia lentiscus L. (Anacardiaceae), a woody shrub species abundant in the Mediterranean Basin, together with the coterie of frugivorous animals that consume its fruits and disperse its seeds. In Chapter 1 we review sampling methods used to record frugivory interactions, assessing their strengths, caveats, and convenience in different contexts. We discuss different approaches for combining data gathered using different methodologies. In Chapter 2, we explore the effectiveness of the frugivory/seed-dispersal mutualism between individual plants of P. lentiscus and its avian frugivore assemblage, assessing the level of reciprocity in reward exchange and dependence asymmetry among partners. In Chapter 3 we investigate the delayed outcomes of animals’ fruit consumption in early seedling recruitment of P. lentiscus plants. Lastly, in Chapter 4 we analyse individual-based networks of different plant species and regions of the world, we compare their topology and structure with that of species-based networks, and explore the specialisation and interaction profile of individual plants within populations. We highlight the significant advancements brought by emerging methods like camera traps and molecular tools, which allow recording interactions across large spatiotemporal scales and even tracking the movement of individual seeds by animals. We also show how data integration improves network completeness and representativity, and compare different data merging approaches for adjacency matrices. We document how the effectiveness of interactions established between P. lentiscus individual plants and their frugivores is mainly determined by fruit consumption frequency (the component with highest variation), leading to highly reciprocal exchange of services but still highly asymmetric dependence among partners. Despite being mostly determined by the number of fruits consumed, the dispersal service provided by birds is decoupled from microhabitat suitability: frugivores deposit a smaller number of seeds in the most suitable microhabitats. These results underscore the role of different frugivores in spatial recruitment of heterogeneous landscapes. Finally, we demonstrate that the structure of individualbased networks is very similar to that of species-based networks. Plant individuals presented similar interaction profiles irrespective of the species or regions their population belonged to. Within populations, plants present low to medium levels of specialisation, and few frugivore species contribute most interactions in all studied networks. Overall, these results suggest that super-generalist species may evolve when they combine sets of traits that make them accessible and attractive to a diversified assemblage of frugivores. Numerical effects, such as abundant fruit crops, facilitate plenty of frugivory interactions, characterised in most cases by high reciprocity. In exchange for a fair nutritious reward, plants secure their recruitment thanks to a vast amount of seeds dispersed. Finally, our results indicate that the variation encountered in the way plant individuals structure their interactions is highly consistent across populations of generalised mutualisms worldwide. Collectively, the four chapters in this thesis contribute to better understanding the origin and maintenance of super-generalist species within complex ecological networks, by focusing on the biological scale at which interactions actually occur in nature, that is, the individual scale.
|
|---|
