Search

Species contribute differently to the structure and stability of networks of interacting species. However, species contributions to the importance of other species is usually neglected, thus limiting our understanding of species dynamics beyond the general network structure. We combined knowledge on diversity effects on focal species and network ecology to analyze the influence of plants and herbivores on big-leaf mahogany's (Swietenia macrophylla) importance in the plant–herbivore network. To this end, we built interaction networks using information from a large-scale tree diversity experiment and performed computer simulations of species removal with redistribution of interactions. We compared the importance of big-leaf mahogany in the observed networks to simulated networks where we removed: (a) tree species with similar beetle assemblages to mahogany, (b) the most interconnected (core) tree species and (c) the core beetle species. Removal of the core and similar tree species increased the importance of mahogany, whereas eliminating core beetle species decreased it. Interestingly, the effect of core tree species on mahogany’s importance was mediated by core beetles (R2 = 0.46). Neither tree nor core beetle species’ effects were contingent on tree species richness or abundance. These results indicate that highly connected tree and herbivore species jointly determine the role of mahogany in the plant–herbivore network. Likewise, these results provide insight into the effect of tree species composition and highly interactive herbivores in shaping the role of species in the herbivore network. CONACyT grant awarded to VPT and LAR (128856).

Ant-plant interactions are widespread and have profound effects on plant fitness, food webs, and species diversity. Whereas most experimental work has tested for ant effects on herbivores, few studies have investigated ant effects on plant pathogens, and even less dual effects on herbivores and pathogens. In this work, we tested whether ants protect wild cotton (Gossypium hirsutum L.) plants against insect herbivores and fungal pathogens by means of an ant exclusion field experiment. In addition, we measured extrafloral nectar (EFN) traits (volume and concentration) to assess their role in ant attraction, and controlled for cotton direct chemical defenses (phenolics) in testing for ant-mediated defense. Ant exclusion did not have a detectable effect on either insect herbivory or pathogen infection levels on wild cotton leaves, and, accordingly, did not impact plant growth (height) over the growing season. We found that EFN traits were associated with ant recruitment to wild cotton plants, with both nectar amount and concentration being positively correlated with ant abundance on control plants. The above tests accounted for leaf phenolics, a type of direct defense, indicating no independent effect of indirect defense mediated by ants on wild cotton during the study. These results provide limited evidence of biotic defense by ants against insect and pathogens on wild cotton at the study site, but call for further assessments of spatiotemporal variation in wild cotton-ant interactions., This study was fnancially supported by CONACYT Grant (CB-250925) to LAR and VPT.

Animal pollination is responsible for the majority of the human food supply. Understanding pollination dynamics in agricultural systems is thus essential to help maintain this ecosystem service in the face of human disturbances. Surprisingly, our understanding of plant–pollinator interactions in widely distributed smallholder agricultural systems is still limited. Knowledge of pollination dynamics in these agricultural systems is necessary to fully assess how human disturbances may affect pollination services worldwide. In this study, we describe the structure of a plant–floral visitor network in a smallholder Cucurbitaceae agricultural system. We further identify the main floral visitors of these crops and tested their importance by simulating how their extinction affected network structure and robustness. The observed network was highly connected and generalized but it was neither nested nor compartmentalized. Our results suggest that the structure of agricultural plant–pollinator networks could be inherently different from those in natural communities. These differences in network structure may reflect differences in spatial distribution of floral resources between agricultural and natural systems. We identified Augochlora nigrocyanea and Peponapis limitaris as the two most frequent floral visitors. However, removal of these species did not affect network structure or its robustness, suggesting high levels of interaction rewiring. To our knowledge, this is one of the first studies to describe the structure of a plant–floral visitor network in diverse agricultural systems in the tropics. We emphasize the need for more studies that evaluate network structure in agricultural systems if we want to fully elucidate the impact of human disturbances on this key ecosystem service.