Isabela Galarda Varassin

Plant–animal interactions result in many ecosystem services, such as pollination. In the western hemisphere, and especially the Neotropics, hummingbirds are important pollinators for thousands of plant species, yet there is much to be learned about the spatial and temporal variation in mechanisms influencing these interactions. We collected an extensive dataset from 39 sites across elevation gradients spanning 4.7–3522 m above sea level in three regions: the Pacific slope of the Talamanca Mountain Range in central‐southern Costa Rica, the Chocó‐Andean region of northwestern Pichincha province in Ecuador, and the Atlantic Forest in southeastern Brazil. Because these regions have very different biogeographic histories, comparing them allows us to identify general patterns of interactions. Data on flower abundance and plant–hummingbird interactions were gathered monthly for 1.5–2 years along 1.5‐km transects situated along an elevation gradient in each country during 2017 and 2024. Trait data for flowers were also measured. This effort resulted in 95,824 pairwise interactions between 715 plants and 79 hummingbird species, and also 873 interactions with another 12 bird species in Ecuador. These data are being used to develop a predictive framework on species interactions necessary to understand and conserve biodiversity and ecosystem services. These data also provide detailed information on plants important for hummingbird conservation which have, for example, been used to guide restoration projects. The data are available for reuse under CC BY 4.0 licensing. Please cite the data release on Zenodo and this data paper when reusing the dataset.

Understanding how species roles in mutualistic networks relate to geographical distributions is key to predicting ecosystem stability across heterogeneous landscapes. We investigated plant–pollinator networks in the Southern Grasslands of Brazil, integrating data from the Atlantic Forest high-altitude grasslands and the Pampa. Using published studies, grey literature, and biodiversity databases, we built a regional meta-network with 8001 interactions among 422 plant and 338 bee species. Network structure was assessed using degree centrality and modularity, and we tested whether centrality relates to species geographical range size. The network showed strong modular organization, with nine well-defined modules and significantly higher modularity than null expectations. Interaction patterns were highly asymmetric, with few highly connected species and many rare interactors. Bees exhibited higher centrality than plants, particularly eusocial and generalist tribes, such as Meliponini, Apini, and Augochlorini. Despite broad variation in distribution area, geographical range size did not predict centrality for either plants or pollinators. These results indicate that species roles are driven primarily by taxonomic identity rather than spatial distribution. Conservation strategies should prioritize key functional groups, especially highly connected bees, which disproportionately sustain network structure.

Plant-pollinator interactions are essential for the reproduction of most flowering plants and the maintenance of biodiversity. With each new assessment of threat status, new plants and pollinators are included in the list of threatened species. However, the ecological interactions established by threatened plants and pollinators are largely unknown. To support conservation planning for threatened plants and pollinators, we surveyed interactions for the plant and pollinator species listed in the official threatened species lists for Brazil. We found data on plant-pollinator interactions for less than 30% of the species listed as threatened. From what is known, abundant species with generalist characteristics play a central role in the interaction meta-network of threatened species. This information can be used to promote better actions for their conservation. Nevertheless, more efforts are required to focus on pollination interactions of threatened animals and plants. We highlight and discuss the relevance of exotic species such as Apis mellifera and Lantana camara in our set of mapped interactions. We conclude that the data synthesized here can support actions for the long-term maintenance of threatened species and the functioning of ecosystems.

Ecological communities result from complex species interactions, often summarized in interaction networks. The structure of these networks is described by metrics that provide insight into community assembly, ecosystem functioning and coevolutionary processes. Despite advances in measuring and mapping network structure, the mechanisms underlying its formation remain less explored. Network metrics may vary across communities owing to changes in species diversity and environmental conditions. However, network metrics may remain invariant if mechanisms influencing interactions (linkage rules) are independent of species composition and environmental conditions and instead influenced by traits. We investigated whether changes in taxonomic, phylogenetic and functional diversity along elevation gradients influence network modularity, nestedness, connectance and specialization across 32 sites in Brazil, Costa Rica and Ecuador. Despite elevation’s impact on diversity, we found that it had no effect on network structure, which remained consistent across elevations. Instead, trait-based mechanisms, specifically the matching between hummingbird bills and flower corollas, emerged as a consistent driver of network structure. Species showing strong trait matching contributed more to modularity and specialization, but less to nestedness and connectance than expected by chance. These results suggest that trait matching influences the invariant structure of plant–hummingbird networks, persisting despite shifts in diversity across biogeographical regions and elevational gradients.

Plant species pollinated by animals differ in the pool of visitors they attract. Some attract functionally similar animals, while others engage with a broader set of functionally distinct visitors; we term this species‐level variation functional generalism. Here, we explore the influence of plant traits and context‐dependent factors on functional generalism, and the potential effects of this overlooked aspect of mutualistic interactions on species' reproductive outcomes. We quantified functional generalism in 211 hummingbird‐visited plant species across 32 sites in three regions of Costa Rica, Ecuador and Brazil. We related functional generalism to plant traits (corolla length, opening, curvature) and context‐dependent factors, including the abundance and flowering duration of the focal plant species, as well as the functional diversity of co‐flowering plant species and of hummingbird pools during its flowering period. Finally, we assessed how functional generalism influenced plant reproductive outcomes (visitation rate and proportion of nectar‐robbing visits). Species with long, narrow corollas had the highest functional generalism, attracting more functionally diverse hummingbirds. Functional generalism increased in functionally diverse plant communities, highlighting the influence of neighbours within the same trophic level. Abundant species with short flowering periods had greater functional generalism than less abundant, continuously flowering species. Functional generalism had a hump‐shaped relationship with visitation rate, peaking at moderate levels but declining afterward. Moreover, the proportion of nectar‐robbing visits increased with functional generalism. These findings reveal the joint influence of plant traits and community context on functional generalism, which in turn affects plants' reproductive success. Read the free Plain Language Summary for this article on the Journal blog.

A myriad of factors influence species interactions, and determining their relative importance is a major challenge in community ecology. Here, we explored the multiple factors influencing hummingbird visitation rates to flowers by considering pollinator and floral traits alongside the broader ecological context (i.e. phenology and floral abundance). Our data were collected from 32 systematically surveyed sites within native vegetation across three distinct Neotropical bioregions, each characterized by unique species composition and evolutionary histories. In total, we sampled 6423 flowering individuals, representing 404 plant species and 68 hummingbird species. At the community level, we found that hummingbirds primarily visit flowers with long, tubular corollas, aligning with predictions from pollination syndromes, and that visitation is also higher to flowers with straight corollas. At the species level, visitation frequency was associated with specific characteristics of the hummingbird species, such as its bill length and foraging behavior, where species often forage on flowers with matching morphological traits. Further, our findings extended beyond the expected trait preferences, revealing that visitation rates are also the outcome of the ecological context; particularly, the abundance of conspecific, but not heterospecific, flowers. The effect of conspecific abundance varied across bioregions, suggesting that from the plant perspective, intraspecific competition and facilitation could prevail in different regions. Additionally, we found that plant species whose floral morphology was more distinct from the other co‐flowering species tended to attract more visits, but only if they were abundant enough to stand out within the floral community. These findings highlight how plant and pollinator traits, modulated by the ecological context, jointly shape pollination dynamics.

Tropical mountains often harbour high species richness. Yet the mechanisms behind such high richness remain poorly understood. One prominent hypothesis for high species richness is niche partitioning, which reduces competition and promotes coexistence. Here, we evaluate niche structure and specialisation of plant species based on the floral traits related to pollination interactions across an elevational gradient in the northern Andes. Niche structure can vary among sites, either expanding or contracting, or becoming more or less packed. We sampled eleven communities of hummingbird‐visited plants along an elevation gradient in the Ecuadorian Andes and measured a series of functional traits associated with hummingbird pollination. We used the traits of all co‐occurring species to calculate community weighted means, functional richness, and mean nearest neighbour distances and evaluated how they varied across elevation. Additionally, we measured specialisation based on plant–hummingbird interaction records to assess if packing is associated with narrower resource use or greater niche overlap. Species and functional richness were constant along the elevation gradient; however, niche packing was stronger at mid‐elevation. We found changes in network specialisation, where the least specialised communities were those with higher niche packing. These results suggest that traits related to pollination and plant reproduction help to explain species co‐occurrence and specialisation.

The presence in ecological communities of unfeasible species interactions, termed forbidden links, due to physiological or morphological exploitation barriers has been long debated, but little direct evidence has been found. Forbidden links are likely to make ecological communities less robust to species extinctions, stressing the need to assess their prevalence. Here, we used a dataset of plant–hummingbird interactions, coupled with a Bayesian hierarchical model, to assess the importance of exploitation barriers in determining species interactions. We found evidence for exploitation barriers between flowers and hummingbirds across the 32 studied communities; however, the proportion of forbidden links changed drastically among communities because of changes in trait distributions. The higher the proportion of forbidden links, the more they decreased network robustness because of constraints on interaction rewiring. Our results suggest that exploitation barriers are not rare in plant–hummingbird communities and have the potential to limit the rescue of species experiencing partner extinction.