A large dataset of animal-plant interactions was collected across 11 East Caribbean islands.
Specifically, the project focused on birds as providers of two crucial ecosystem functions: pollination and seed dispersal. Interactions were sampled by mist-netting birds (>1500 individuals) and collecting pollen samples from their beaks and feathers, and by collecting a fecal sample from each individual, which allowed to screen for seeds. Additionally, functional traits were measured for each bird individual (beak measurements, body weight), as well as for all flowers and fruits encountered in the sample area.
From this data local interaction networks were constructed for the bird-pollination and bird-seed dispersal ecosystem functions respectively. Additionally, a regional meta-network was constructed for bird-mediated seed dispersal, by pooling all interactions across all islands.
Several significant scientific findings have been contributed so far and others are in production.
The project contributed the first empirical data set on bird-mediated seed dispersal across multiple islands of the Lesser Antilles archipelago significantly. This dataset will be made available (open access) shortly and will hopefully contribute to many scientific questions.
The data on bird-mediated pollination, was used to compare hummingbird communities in the continental Americas with hummingbirds on the Caribbean islands.
This comprehensive analysis of a large, unprecedented dataset of plant–hummingbird interactions and associated functional traits across large spatial gradients indicates the presence of an ornithophilous pollination syndrome in plants, but shows that in the Caribbean islands, hummingbird-visited plants display low proportions of traits typically associated with bird-mediated pollination. Caribbean hummingbirds probably expand their feeding niche to include plants with traits that they tend to avoid in the mainland because the lower diversity and abundance of insect pollinators makes it easier to fill this niche. Interestingly, against theoretical expectations, there was a higher functional diversity in the plant–hummingbird interactions of the Caribbean islands. This contradicts theoretical predictions but could be a response to competition over comparatively scarce resources in the environmentally unstable oceanic islands. Results published in Functional Ecology:
https://doi.org/10.1111/1365-2435.70068(s’ouvre dans une nouvelle fenêtre).
The data on bird-mediated pollination further allowed to show that on the island of Dominica in the Lesser Antilles, following Hurricane Maria in 2017 which resulted in a devastating population crash of the Purple-throated Carib (Eulampis jugularis), resulted in clear pollinntor-shifts in a tightly coadapted plant-hummingbird pollination system. Throughout the Lesser Antilles, the Purple-throated Carib competitively excludes other hummingbirds and small birds from access to the flowers of two Heliconia plants. But following their population crash on Dominica, this system has shifted and Heliconias are very frequently visited by other birds as well. These results are important, because hurricanes will increase in frequency, and this study thus suggests that tightly coadapted plant-animal interactions may be reshuffled in the future with potential negative impacts on entire ecosystems. Results published in New Phytologist:
https://doi.org/10.1111/nph.19938(s’ouvre dans une nouvelle fenêtre).
Further, the data allowed to provide genomic evidence for a coadapted mutualism and long-term interdependence between the purple-throated carib and its two Heliconia nectar-food plants on Dominica. The project demonstrates that long-term dependency of specialized pollination systems creates an evolutionary signature on the genomes of interacting partners and highlights the strength of combining comparative genomics with ecology to better understand the strength and evolutionary outcomes of mutualistic systems. Based on the results it would be interesting to expand the analysis to other specialized mutualistic systems, both within pollination but also other types of mutualisms, such as large frugivorous animals and their fruit plants which may also be highly interdependent over long-time frames. Furthermore, these types of analyses could be extended to more generalized mutualistic systems which constitute the bulk of interactions in nature to examine whether long-term dependency can be demonstrated for plants and their animal pollinators and seed-dispersers which are putatively less tightly associated. This study is under review in Current Biology.
Very recently, the data on bird-mediated seed dispersal and associated functional traits was used to to assess the extinction threat for large-seeded plants across the Caribbean islands.
This analysis found that large-fruited plant species are effectively “orphaned” due to the absence of animal dispersers, resulting in elevated rates of endangerment across islands. Threat status of plant species was better explained by trait matching than other pressures such as human use, suggesting that extinction or reduced abundance of large-bodied frugivores is compromising their plant mutualists’ ability to persist. These results argue strongly for the need to restore populations of large-bodied frugivores across the Caribbean, either through population enhancement of currently endangered species or rewilding with close relatives of extinct species. Results published in PNAS:
https://doi.org/10.1073/pnas.2535192123(s’ouvre dans une nouvelle fenêtre).
The data further contributed to a study, where we found that non-native Rattus sp. were responsible for nearly 80% of all frugivorous activity detected by camera traps and that increased non-native mammalian frugivory activity was associated with significant declines in avian frugivory rates. Non-native mammalian frugivory activity , in turn, was positively associated with road proximity and neighboring island proximity, illustrating the potential for both human activity and island geography to influence non-native mammal frugivory interactions. These results suggest that non-native mammals are dominating frugivory dynamics in understory vegetation in the eastern Caribbean, with consequences for native frugivory interactions. Given the generally seed predatory behavior of Rattus sp., we argue that plant dispersal could be negatively affected on islands with abundant non-native mammal populations. Thus, we emphasize the need for controlling non-native mammal populations and highlight the diverse inter-trophic effects that non-native mammals can have on insular tropical ecosystems. This study is under review in Ecology and Evolution.
Another main achievement from this project is currently being prepared for submission to a peer-reviewed journal. Based on the regional meta-network of bird-mediated seed dispersal which was obtained by pooling all data across all islands, we show that islands and island groups strongly correspond to modularity of the network. This study thus suggests that different biogeographical histories may act in distinct islands and island groups of the same archipelago, with potential implications for the design of local conservation strategies. Further, the data suggests that birds in the oceanic islands have larger niche breadths than those in the continental islands, i.e. they interacted with a higher proportion of the locally available plants. However, resource niche overlap did not vary between birds of different island groups, indicating that different species may build functionally similar interaction networks regardless of biogeographical setting. In sum, this work is an important contribution, shedding light on a crucial ecosystem function across the islands of an archipelago, where empirical data is largely missing.
