Search

AimMutualistic interactions between plants and animals are fundamental for the maintenance of natural communities and the ecosystem services they provide. However, particularly in human-dominated island ecosystems, introduced species may alter mutualistic interactions. Based on an extensive dataset of plant–frugivore interactions, we mapped and analysed a meta-network across the Caribbean archipelago. Specifically, we searched for subcommunity structure (modularity) and identified the types of species facilitating the integration of introduced species in the Caribbean meta-network.LocationCaribbean archipelago (Lucayan archipelago, Greater Antilles, Lesser Antilles).MethodsWe reviewed published scientific literature, unpublished theses and other nonpeer-reviewed sources to compile an extensive dataset of plant–frugivore interactions. We visualized spatial patterns and conducted a modularity analysis of the cross-island meta-network. We also examined which species were most likely to interact with introduced species: (1) endemic, nonendemic native or introduced species, and (2) generalized or specialized species.ResultsWe reported 3060 records of interactions between 486 plant and 178 frugivore species. The Caribbean meta-network was organized in 13 modules, driven by a combination of functional or taxonomic (modules dominated by certain groups of frugivores) and biogeographical (island-specific modules) mechanisms. Few introduced species or interaction pairs were shared across islands, suggesting little homogenization of the plant–frugivore meta-network at the regional scale. However, we found evidence of “invader complexes,” as introduced frugivores were more likely to interact with introduced plants than expected at random. Moreover, we found generalist species more likely to interact with introduced species than were specialized species.Main conclusionsThese results demonstrate that generalist species and “invader complexes” may facilitate the incorporation of introduced species into plant–frugivore communities. Despite the influx of introduced species, the meta-network was structured into modules related to biogeographical and functional or taxonomic affinities. These findings reveal how introduced species become an integral part of mutualistic systems on tropical islands.1 INTRODUCTIONMutualistic interactions between plants and animals, such as pollinators and frugivores, are critically important for maintaining the functionality of natural communities (Jordano, 1987;Ollerton et al., 2011; Rech et al., 2016). While most flowering plants are dependent on animals for pollination and seed-set (Ollerton et al., 2011; Rech et al., 2016), animal frugivores may ingest or otherwise manipulate and consequently disperse millions of seeds annually (Bueno et al., 2013). Frugivory is thereby crucial for the maintenance of plant diversity (Harms et al., 2000), as it allows plants to populate new sites, maintains gene flow between distinct populations and decreases density-dependent mortality in proximity of the parent individuals (Rogers et al., 2021). In some tropical systems, approximately 90% of all woody plants depend on frugivores for seed dispersal (Almeida-Neto et al., 2008; Howe & Smallwood, 1982). In addition to providing direct dispersal to specific, favourable sites for the plant (Wenny & Levey, 1998), frugivores can enhance the probability of successful germination, for example through the passage of seeds in the intestinal system (e.g. Traveset et al., 2001). The most important frugivore groups are birds, mammals and reptiles with birds and reptiles being particularly important in tropical island ecosystems (Kaiser-Bunbury et al., 2010; Valido & Olesen, 2007).Globally, co-evolved plant–frugivore communities are suffering from an array of drivers associated with global change, such as the introduction of species into new environments, where they become integrated into local communities through species interactions (Gallardo et al., 2016; Vilà et al., 2011). Species communities are thus being altered, which in turn may have consequences for biotic interactions and ecosystem functions, such as seed dispersal (Aslan et al., 2013; Lugo et al., 2012; Traveset & Richardson, 2006; Vizentin-Bugoni et al., 2021). Island ecosystems are particularly vulnerable to the disruption of native plant–frugivore interactions as island mutualists have evolved in isolation, and frequently developed specific traits, such as altered dispersal, or loss of defence traits in plants (Burns, 2019). Furthermore, as islands harbour many endemic species found nowhere else on Earth (Kier et al., 2009; Paulay, 1994), and have experienced disproportionally high extinction rates and numerous extant island species are threatened with extinctions (Blackburn et al., 2004; Fernández-Palacios et al., 2021; Groombridge, 1992), it is especially important to understand how introduced species integrate into island communities (Wood et al., 2017).Introduced species may integrate into existing communities and establish themselves in different ways. For instance, the concept of “invader complexes” suggests that introduced species facilitate the establishment of other introduced species, resulting in groups of introduced species interacting strongly with each other and less with the remaining community (D'Antonio & Dudley, 1993). Alternatively, endemic species that have become superabundant and highly generalized species due to ecological release and density compensation may readily include new arrivals into their interactions and thereby facilitate the establishment of introduced species on islands (Olesen et al., 2002). Furthermore, a growing number of studies show that species with many mutualistic partners (i.e. generalized species irrespectively of being nonendemic native or endemic) are more likely to incorporate new partners into their networks (Bascompte & Stouffer, 2009; Maruyama et al., 2016). In network theory, this is called “preferential attachment” (Newman, 2001), and thus, most generalized species would be expected to interact with introduced species.In addition to understanding which species are responsible for incorporating introduced species into native communities, we have little quantitative understanding of how introduced species affect the structure of native interaction networks and how this varies biogeographically (Fricke & Svenning, 2020). As for other mutualistic networks, plants and frugivores form complex interaction networks with reccurring structural properties (Bascompte & Jordano, 2007). One such property of interaction networks is modularity, which describes how interacting species aggregate into modules consisting of species that interact strongly within the respective module but much less with species of other modules (Thébault, 2013). The modular structure of mutualistic networks may reflect “co-evolutionary units” (Olesen et al., 2007) determined by an array of factors, such as phenological overlap, morphological traits, taxonomic relatedness and biogeography (Araujo et al., 2018; Dalsgaard et al., 2013; Donatti et al., 2011; Martín González et al., 2018; Maruyama et al., 2014; Schleuning et al., 2014). However, it is poorly understood whether introduced species influence the modular structure of mutualistic systems.Here, we present an extensive dataset on plant–frugivore interactions compiled from published and unpublished resources across the islands of the Caribbean archipelago: Lucayan archipelago, Greater Antilles and Lesser Antilles. We use the data to (i) explore the distribution of frugivory records across the Caribbean islands; (ii) assess island connectivity through shared species and interactions; (iii) evaluate the modular structure of the regional plant–frugivore meta-network and (iv) determine whether generalized vs. specialized species and introduced vs. endemic species are more likely to integrate introduced plants and frugivores into native plant–frugivore communities in island systems.2 METHODS2.1 Data collection and study regionAll our data were collected on the Caribbean islands, that is the Lucayan archipelago (The Bahamas and Turks and Caicos), the Greater Antilles (Cuba, Cayman Islands, Jamaica, Hispaniola and Puerto Rico) and the Lesser Antilles (a series of islands from the US and British Virgin Islands in the north to Grenada in the south). We did not include plant–frugivore interactions from islands such as Trinidad and Tobago, Curaçao, and Bonaire just north of South America, as these are continental islands with biotas with strong affinities to the South American mainland (Carstensen et al., 2012; Ricklefs & Bermingham, 2008). The low-lying sedimentary islands of the Lucayan Archipelago are part of the North American platform (Iturralde-Vinent & MacPhee, 1999; Trejo-Torres & Ackerman, 2001), and some of the islands have been interconnected in the Pleistocene (Murphy et al., 2004; Trejo-Torres & Ackerman, 2001). The mostly large and mountainous islands of the Greater Antilles are old with different geological origins (Graham, 2003; Iturralde-Vinent & MacPhee, 1999). The Greater Antilles emerged as fragments in the Eocene about 49 Ma; the geological history of the region has been highly dynamic with some parts connected in the past (Buskirk, 1985; Graham, 2003; Iturralde-Vinent & MacPhee, 1999; Ricklefs & Bermingham, 2008). The current biota of the Greater Antilles was only in small parts formed by vicariance, with dispersal facilitated by the Aves Ridge about 32–35 Ma (Iturralde-Vinent & MacPhee, 1999) or a more likely overwater dispersal at least for the avifauna (Buskirk, 1985; Graham, 2003; Ricklefs & Bermingham, 2008). The Lesser Antilles form a volcanic arc where the North and South American plates subduct under the Caribbean plate and likely originated at least 20 Ma (Ricklefs & Bermingham, 2008). To the east of the volcanic arc are some younger and low-lying islands such as Antigua and Barbuda, which consist of uplifted marine sediments (Ricklefs & Bermingham, 2008; Ricklefs & Lovette, 1999). Some islands were interconnected during the last glacial maximum, but most Lesser Antilles islands have never been interconnected (Ricklefs & Bermingham, 2004, 2008). The isolation of the Caribbean islands from the mainland differs greatly (Carstensen et al., 2012). Bimini in the Bahamas, for instance, is only approx. 87 km from the North American continent and Grenada in the Lesser Antilles is only 137 km from the continental landmass of South America. By contrast, islands such as Grand Turk (993 km) and South Caicos (999 km) are much more isolated from the mainland. On average, the isolation from any continental landmass in the Caribbean is over 500 km (Mean: 593 km ± 248 km SD; see details in Supporting Information Table S1). The distances between single islands are much smaller, for example the distance between Martinique and Dominica and Martinique and Saint Lucia is approx. 40 km. An island size threshold of 10,000 km2 has previously been suggested to be important for islands to be considered sources for colonization (Weigelt & Kreft, 2013), and on average, the islands of the Caribbean are approx. 304 km (±174 km SD) from the nearest island that exceeds 10,000 km2 (Table S1). Given the geological history and isolation of the Caribbean, the biota is characterized by being depauperate with high levels of endemism.To collect data on interactions between plants and frugivores in the Caribbean, we screened the Web of Science (WoS) and Google Scholar search engines. We used the combination of the following search terms: (“frugivory” OR “seed dispersal” OR “seed removal” OR “mutualism”) AND (“Caribbean” OR “Lesser Antilles” OR “Greater Antilles” OR “West Indies” OR “Bahamas” OR “Turks and Caicos” OR “Cayman Islands” OR “Jamaica” OR “Cuba” OR “Hispaniola” OR “Haiti” OR “Dominican Republic” OR “República Dominicana” OR “Puerto Rico” OR “Mona” OR “Virgin Islands” OR “Saint Martin” OR “Anguilla” OR “St. Kitts and Nevis” OR “Antigua” OR “Barbuda” OR “Montserrat” OR “Guadeloupe” OR “Dominica” OR “Martinique” OR “St. Lucia” OR “St. Vincent” OR “Grenadines” OR “Barbados” OR “Grenada”). To also include the grey literature, we contacted local ornithologists and ecologists working in the Caribbean region. This approach allowed us to obtain non-English publications, such as theses and dissertations not available online. We screened each of the studies manually, discarding studies where no appropriate data were presented (e.g. mutualistic interactions in marine environments). Interactions were only included when the respective authors presented original evidence for interaction events, that is evidence of fruits and/or seeds being ingested by frugivores. Thus, we discarded records where interactions between species were speculative (e.g. observation of frugivores on fruiting plant species without any evidence of fruit ingestion).We standardized the species names of plants and frugivores using the R-package taxize (Chamberlain & Szocs, 2013; Global Names Resolver, 2021) and data from the Integrated Taxonomic Information System (ITIS, 2021). We also retrieved information about species taxonomies (i.e. class, order and family) from ITIS. Finally, we compiled information about the native status of species and classified them into nonendemic native (species native to the Caribbean, but also naturally occurring elsewhere), endemic (only occurring within the Caribbean) and introduced (not naturally occurring within Caribbean) species (see details in Supporting Information Text S1). Of the original records, 95 plant (approx. 16% of all reported plants) and one frugivore record were not identified to species level (e.g. only genus name reported) and were thus excluded from data analyses. The final data used in statistical analyses consisted of interactions between 486 plant and 178 frugivore species.2.2 Data analysis2.2.1 Cross-island patterns of shared species and interactionsWe summarized patterns of shared species and interaction pairs across the Caribbean by calculating the proportion of shared species and interaction pairs across all islands. We calculated this proportion as the number of species/interaction pairs found on any two islands, divided by the total number of species/interaction pairs found on the given islands (Fricke & Svenning, 2020). We summarized these patterns separately for all reported records, for endemic, nonendemic native and for introduced plant and frugivore species and interaction pairs, respectively.2.2.2 Modularity of the Caribbean plant–frugivore meta-networkTo detect a modular structure of the meta-network, that is the network of plant–frugivore interactions across all islands, we employed Beckett's DIRT-LPA algorithm in the computeModules function of the R-package “bipartite” (Dormann et al., 2008, 2009). We ran 10 independent runs of the algorithm on the binary meta-network containing interactions between all identified species and identified the run with the single best division into modules, that is the highest degree of modularity Q. For the run with the highest Q value, we recorded the Q value, the number of modules as well as the respective plant and frugivore species in each module (Schleuning et al., 2014) and the islands on which they were recorded. To test whether the identified modular structure of the meta-network differed from random, we compared our results to 100 null models. To this end, we used an algorithm proposed by Patefield (1981) to randomize the interactions between species, using fixed marginal totals to produce networks with randomly associated species without constraining the degree of specialization (Blüthgen et al., 2008; Schleuning et al., 2014). For each of the null models, we applied the same approach as with the original matrix, that is we identified the single best configuration from 10 independent runs (Schleuning et al., 2014). We then tested whether modularity of the original matrix was significantly different from the best 100 null models by looking at the proportion of null modularity values that were greater than the empirical one, that is if .05; Figure S2c) and the frugivore perspective (slope = 0.003, p > .05; Figure S2d).4 DISCUSSIONHere, we present a comprehensive review of published plant–frugivore interactions across the Caribbean archipelago, including the Lucayan archipelago, the Greater and Lesser Antilles. All islands shared species and unique interaction pairs with neighbouring islands and archipelagos, thereby forming a cohesive meta-network. We show that the meta-network of plant–frugivore interactions across the Caribbean was structured into modules, with at least some modules determined by a combination of functional or taxonomic (i.e. certain groups of frugivores) and biogeographical (i.e. island-specific modules) mechanisms. While relatively few species in the dataset were introduced to the Caribbean (16% plant and 8% frugivore species), we found support for the “invader complexes” theory, whereby introduced species facilitate the establishment of other introduced species (D'Antonio & Dudley, 1993; Olesen et al., 2002). Moreover, we found that generalized species were more likely to incorporate introduced species into their interactions, giving support for the “preferential attachment” theory (Newman, 2001). Below, we first discuss the available data on frugivory in the Caribbean, whereafter we discuss how species and interactions are shared across islands. We end by discussing the drivers of modularity and the integration of introduced species into plant–frugivore communities across the Caribbean.4.1 Data on frugivores and their plants in the Caribbean archipelagoAcross all islands, the vast majority of reported frugivores were birds (79%), followed by reptiles (13%) and mammals (8%), of which in turn the majority were bats (71%). These data thus reflect patterns that are typical for oceanic islands, as there is generally a lack of nonvolant, large-bodied, frugivorous mammals which may be ecologically replaced by birds and reptiles (Kaiser-Bunbury et al., 2010). The low number of mammal species in the dataset could also reflect past mammal extinctions particularly on the islands of the Greater Antilles (Turvey et al., 2021), potentially leaving some plants without their main seed dispersers.A large proportion of the plant species (28%) and the majority of frugivore species in the dataset (65%) were classified as endemic to the Caribbean. High degrees of endemism in local species communities are characteristic of island ecosystems (Kier et al., 2009; Paulay, 1994). In a review of plant–frugivore interactions on the Galapagos archipelago, Heleno et al. (2011) found similarly high proportions of endemic frugivores in the species pool (71%), underlining the importance of endemic frugivores for island communities. By contrast, only a few species in the dataset were classified as introduced to the Caribbean (16% plants and 8% frugivores), which was lower than other studies on island ecosystems. Notably on Hawai'i, the proportion of introduced seed disperser species ranged from 50% to 100% for plants and from 60% to 100% for birds (Vizentin-Bugoni et al., 2019). On the Galapagos, the proportion of introduced plants and frugivores was 28% and 23%, respectively (Heleno et al., 2011). However, on the Galapagos, all introduced frugivore species were mammals, whereas in our data, the vast majority of introduced species were birds (86%) and only two species (14%) were mammals (a primate: Chlorocebus pygerythrus and a carnivore: Herpestes javonicus).4.2 Cross-island patterns of shared species and interactionsWhen examining the role of different groups of plants and frugivores in connecting islands and archipelagos, we found that nonendemic native species and interaction pairs were shared most widely across islands (Figure 2), which is expected, as these species are widespread species occurring throughout the Caribbean and the Neotropical mainland. They are thus supposedly good dispersers, and their ranges often occur across multiple islands and cross-borders of archipelagos (Dalsgaard et al., 2014). Although endemic frugivores made up more than 60% of the frugivore species, generally they overlapped much less between islands compared to nonendemic natives, which only accounted for less than 30% of the frugivores in the data (Table 2). This pattern is not surprising, since the distributional ranges of endemic species are per definition confined within limited geographical areas (Kricher, 2011), many species being single-island endemics or occurring on few islands within each of the archipelagoes, that is the Lucayan archipelago, the Greater and Lesser Antilles (Dalsgaard et al., 2014). In the Caribbean, for instance, there is a high number of single-island endemic frugivorous birds, such as various species of parrots like the Saint Vincent Parrot Amazona guildingii (Birds Caribbean, 2021). Introduced plant species were shared widely across the Caribbean (Figure 2), which was expected, as most were agricultural and widely cultivated plants, reflecting that the Caribbean is historically heavily impacted by humans (Kemp et al., 2020; Walters & Hansen, 2013). By contrast, introduced frugivores were reported on few islands only (Table 2), and these islands shared mostly low proportions of introduced frugivores (Table S4); introduced interaction pairs were almost not shared between islands. Globally, a recent study showed how introduced species caused an increase in the proportion of regions sharing species and interactions (Fricke & Svenning, 2020), demonstrating that species introductions led to increasing similarity and homogenization in plant–frugivore communities across the world (Fricke & Svenning, 2020). In the Caribbean, however, given our data, especially nonendemic natives played a bigger role in interconnecting islands.4.3 Modularity of the Caribbean plant–frugivore meta-networkThe Caribbean plant–frugivore meta-network was organized in modules, as are most mutualistic plant–animal interaction networks, both local networks (e.g. Dalsgaard et al., 2013; Dupont & Olesen, 2009; Mello et al., 2011a, 2011b; Olesen et al., 2007) and meta-networks (Araujo et al., 2018; Emer et al., 2018; Martín González et al., 2018). The separation of the meta-network into modules was at least partly driven by functional or taxonomic (i.e. modules dominated by certain species groups) and biogeographical (i.e. island-specific modules) mechanisms. For instance, one module consisted of small- to medium-sized bird species recorded in Jamaica (100% birds; 88% of frugivores recorded in Jamaica; module nine in Figure 4). Another module consisted mostly of various bat species (63% bats) recorded in Cuba (88% of frugivores recorded in Cuba; module one in Figure 4), whereas another module consisted almost exclusively of rock iguanas (Cychlura spp.) found in the Bahamas only (88% Iguanas; 88% of frugivores were recorded on the Bahamas only; module 13 in Figure 4). These modules associated with specific functional/taxonomic groups or specific islands were thus positioned in the periphery of the Caribbean meta-network (Figure 3). The separation into modules according to biogeographical affinities, such as single islands, was expected given that interactions between plants and frugivores are inherently spatial as species must be in the same place to interact (Morales & Vázquez, 2008) and many species are restricted to specific islands. Spatial patterns that correspond to insularity in the broad sense have previously been shown to partially explain the modular structure of mutualistic plant–animal networks in landscape matrices, where species are restricted to different types of patchily distributed habitats (Maruyama et al., 2014). Patterns of modularity have also previously been suggested to be explained by behavioural or functional traits of species (Dicks et al., 2002; Donatti et al., 2011; Maruyama et al., 2014). In plant–frugivore interactions, although plants typically aim to attract functionally diverse seed dispersers (Plein et al., 2013), there is evidence of functional matching between interaction partners, especially with birds (Vollstädt et al., 2017). Morphologically different frugivore species tend to forage on morphologically distinct sets of plant species (Dehling et al., 2016; Gautier-Hion et al., 1985; Lomáscolo et al., 2010; Mello et al., 2011b), which might be reflected in the modules composed primarily of specific frugivore groups with characteristic morphological and functional traits. Bats, for instance, consume different types of fruits than birds and may show a clear separation in their dietary composition (Gorchov et al., 1995). The patterns of modularity we detected were therefore in line with expectations of functional/taxonomic and biogeographical mechanisms as drivers of modularity. However, there were also modules consisting of a mix of species from various islands. One module consisted of about 50% of large parrot species (Amazona spp.), but the frugivores were recorded in the entire Caribbean (module six in Figure 4) and, notably, the module in the centre of the Caribbean meta-network consisted of various types of frugivores occurring throughout the Caribbean, thereby interconnecting islands and archipelagos in the Caribbean meta-network (module seven; Figure 3 and Table S5).4.4 Interactions with introduced speciesRegarding how introduced species were integrated into the meta-network, we found that nonendemic native and endemic plants and frugivores interacted significantly less with introduced frugivore species than expected at random (Figure 4a,b). Among Caribbean frugivores and their fruiting plants, there is therefore no support for the idea that endemic super-generalists are the main facilitators of introduced species, as suggested for pollination networks on tropical islands (Olesen et al., 2002). On the contrary, introduced frugivores were recorded interacting with introduced plants significantly more often than expected at random (Figure 4c). This pattern suggests that introduced frugivores “prefer” to feed on introduced plants, which in turn suggests the presence of “invader complexes,” that is introduced species interacting more among themselves than expected at random, thus facilitating their establishment (D'Antonio & Dudley, 1993). Such facilitation processes between introduced species can lead to “invasional meltdowns,” as large groups of introduced species may have increasingly negative impacts on native communities (Jeschke et al., 2012; Simberloff & von Holle, 1999). Other island ecosystems have been found to be even more dominated by introduced frugivores, notably Hawai'i is almost exclusively dominated by introduced frugivores, as most of the endemic species have gone extinct (Vizentin-Bugoni et al., 2019; Vizentin-Bugoni et al., 2021). These findings from various archipelagos are concerning, regarding the potential impact of introduced species on native ecosystems. Such findings are particularly worrying when considering that on other island ecosystems, introduced species were also more often involved in seed-dispersal interactions (rather than seed/pulp predation) than native species (Heleno et al., 2011; Vizentin-Bugoni et al., 2019, 2021). For many of the interaction records, our data do not distinguish between seed-dispersal interactions or seed/pulp predation events; thus, it is not possible to estimate the effect of introduced species on local native and endemic plant communities in the Caribbean. Nevertheless, in Hawai'i, it was shown that introduced frugivores do not sufficiently replace the species roles of lost seed dispersers, since they preferentially disperse seeds of introduced rather than native plants (Vizentin-Bugoni et al., 2019). This raises the question why introduced plant species seem so attractive. One reason could be that introduced plants may have specific traits, such as longer fruiting duration, which increase the probability of encounters and are therefore more likely to be consumed by frugivores (Heleno et al., 2011; Sperry et al., 2021). In the Caribbean meta-network, many of the observations were from agricultural areas, where agricultural plants such as Mangifera indica (Mango) are often abundant with large crops, and although they are not dispersed by any native frugivore, they do overall attract many frugivores. Fruiting plant and thus resource abundance is in turn linked to increased fruit consumption, because frugivores often track available fruits in the landscape (Quitián et al., 2019), and consequently, the patterns we find may be partially driven by the high abundance of introduced agricultural plants and their crop sizes in human-dominated environments. Such patterns may be more pronounced on densely populated islands than on islands with few people and relatively more protected areas.In addition to “invader complexes,” we found that generalist species, that is species with many interaction partners, were more likely to interact with introduced species, which was consistent for both plants and frugivores (Figure S2a,b). These results are in line with previous findings, underlining the importance of highly generalized species for the establishment of introduced species, especially on islands (Maruyama et al., 2016). This gives support for the “preferential attachment” hypothesis (Newman, 2001), that is that species with wide ecological niches include and facilitate the establishment of new species, such as introduced species on islands. Our finding that generalized species do not have a higher proportion of interactions with introduced partners in their total set of interactions than specialized species (Figure S2c,d) reflects the overall low numbers of introduced species in the Caribbean data. Since only few of the potentially available interaction partners are introduced species, generalized species with many interaction partners would also be expected to have a decreasing proportion of their interactions with introduced species. Thus, although generalized species are likely to incorporate introduced species into their niche (Figure S2a,b), they do not have a specific preference for introduced species (Figure S2c,d).5 CONCLUSIONSBased on a comprehensive review of accessible data on plant–frugivore interactions, we showed that the Caribbean meta-network is structured into modules and demonstrate how introduced species are integrated into native communities in the Caribbean archipelago. These results provide valuable insight into plant–frugivore interactions in insular biodiversity hotspots, showing how insular plant–frugivore systems are susceptible to invasion. Future studies are needed to demonstrate the importance of introduced species as seed dispersers compared with seed/pulp predators (Nogales et al., 2017). Specifically, research quantifying the relative importance of different frugivore groups as seed dispersers and their respective effectiveness is lacking for most plant–frugivore interactions in the Caribbean. This would provide valuable information and could help with the conservation of endemic plants in the Caribbean archipelago. Moreover, we also in general lack information on frugivory in the Caribbean. Kim et al. (2022) reported 4336 species of plants with animal-dispersal syndromes in the Caribbean archipelago, and our dataset represents only 11% of those species with some regional variation (Table S6). For instance, whereas Puerto Rican plants were covered relatively well (31% of the species), plants in Hispaniola (approx. 7%), Jamaica (approx. 8%) and the Lesser Antilles (approx. 9%) were less well represented. There may also be taxonomical differences in sampling completeness. Palms (Arecaceae) are highly diverse in the Caribbean representing 135 species (Roncal et al., 2008), and our dataset had only 23 palm species (17%). Several endemic and highly threatened fleshy-fruited plants do not have any information on the main seed dispersers (e.g. Catesbea spinosa, Brunfelsia portoricensis, Diospyros spp. and many cactus species). We also have limited and incomplete information on the fruit diet of several endemic frugivores (e.g. pigeons, thrashers and thrushes) that could play an important role for seed dispersal of Caribbean plants. There is therefore an urgent need to increment more scientific information on plant–frugivore interactions in the Caribbean, one of the world's insular biodiversity hotspots.ACKNOWLEDGEMENTSThis work is dedicated to our colleague and coauthor Michel Faife-Cabrera who passed away due to Covid-19 complications. M. Galetti thanks CNPq and University of Miami for financial support. M.G.R. Vollstädt, C. N. Kaiser-Bunbury, B. I. Simmons, F. Gonçalves, and B. Dalsgaard thank the Independent Research Fund Denmark (grant no. 0135-00333B). Funding for A. Strong's work was provided by an NSF grant to T. W. Sherry (Tulane University) and R. T. Holmes (Dartmouth College), the Chicago Zoological Society, Sigma Xi Grants-in-Aid-of Research, the World Nature Association, and The Louisiana Educational Quality Support Fund. A. Strong's work benefitted from collaborations with M. Johnson, T. Sherry, A. Sutton and the late R. Sutton. Funding for S. Schubert's work was provided by Rufford Foundation Small Grants 1, 2, & Booster, in addition to the Old Dominion University Paul W. Kirk Jr Student Research Award, a British Ornithologists' Union Student Research Award, and the BirdsCaribbean David S. Lee Fund. J. Salazar's work was funded by FONDOCyT. (Ministerio de Educaciòn Superior, Ciencia y Tecnología), Project 1B4-9. F. L. Tarazona-Tubens is supported by McKight Fellowship. B. I. Simmons was supported by a Royal Commission for the Exhibition of 1851 Research Fellowship. We also thank all researchers who have worked intensively in the Caribbean.CONFLICT OF INTERESTThe authors declare that there is no conflict of interest to report.Open ResearchSupporting InformationREFERENCESBIOSKETCHMaximilian Vollstädt is an ecologist, and a researcher at the University of Copenhagen, Denmark. His research interests include mutualistic interactions in plant-pollinator and plant-seed disperser communities in tropical island ecosystems.Author contributions: M.G.R.V., M.G., B.D. Conceptualization, Methodology; M.G.R.V., B.I.S. Formal analysis, Visualization; M.G.R.V., M.G., C.N.K.B., B.D. Original draft preparation, Writing; All Authors: Writing, Review & Editing.Volume28, Issue11November 2022Pages 2361-2374FiguresReferencesRelatedInformationRecommendedNatural mixing of species: novel plant–animal communities on Caribbean IslandsA. E. Lugo, T. A. Carlo, J. M. WunderleAnimal ConservationFrugivore distributions are associated with plant dispersal syndrome diversity in the Caribbean archipelagosSeokmin Kim, Lilian Sales, Daiane Carreira, Mauro GalettiDiversity and DistributionsInteractions between resource availability and enemy release in plant invasionDana M. BlumenthalEcology LettersHuman disturbances affect the topology of food websFrederico Mestre, Alejandro Rozenfeld, Miguel B. AraújoEcology LettersA general framework for species-abundance distributions: Linking traits and dispersal to explain commonness and rarityThomas Koffel, Kaito Umemura, Elena Litchman, Christopher A. Klausmeier

Aim There has been a wide interest in the effect of biotic interactions on species' occurrences and abundances at large spatial scales, coupled with a vast development of the statistical methods to study them. Still, evidence for whether the effects of within-trophic-level biotic interactions (e.g. competition and heterospecific attraction) are discernible beyond local scales remains inconsistent. Here, we present a novel hypothesis-testing framework based on joint dynamic species distribution models and functional trait similarity to dissect between environmental filtering and biotic interactions. Location France and Finland. Taxon Birds. Methods We estimated species-to-species associations within a trophic level, independent of the main environmental variables (mean temperature and total precipitation) for common species at large spatial scale with joint dynamic species distribution (a multivariate spatiotemporal delta model) models. We created hypotheses based on species' functionality (morphological and/or diet dissimilarity) and habitat preferences about the sign and strength of the pairwise spatiotemporal associations to estimate the extent to which they result from biotic interactions (competition, heterospecific attraction) and/or environmental filtering. Results Spatiotemporal associations were mostly positive (80%), followed by random (15%), and only 5% were negative. Where detected, negative spatiotemporal associations in different communities were due to a few species. The relationship between spatiotemporal association and functional dissimilarity among species was negative, which fulfils the predictions of both environmental filtering and heterospecific attraction. Main conclusions We showed that processes leading to species aggregation (mixture between environmental filtering and heterospecific attraction) seem to dominate assembly rules, and we did not find evidence for competition. Altogether, our hypothesis-testing framework based on joint dynamic species distribution models and functional trait similarity is beneficial in ecological interpretation of species-to-species associations from data covering several decades and biogeographical regions. peerReviewed

Agriculture and the exploitation of natural resources have transformed tropical mountain ecosystems across the world, and the consequences of these transformations for biodiversity and ecosystem functioning are largely unknown1-3. Conclusions that are derived from studies in non-mountainous areas are not suitable for predicting the effects of land-use changes on tropical mountains because the climatic environment rapidly changes with elevation, which may mitigate or amplify the effects of land use4,5. It is of key importance to understand how the interplay of climate and land use constrains biodiversity and ecosystem functions to determine the consequences of global change for mountain ecosystems. Here we show that the interacting effects of climate and land use reshape elevational trends in biodiversity and ecosystem functions on Africa's largest mountain, Mount Kilimanjaro (Tanzania). We find that increasing land-use intensity causes larger losses of plant and animal species richness in the arid lowlands than in humid submontane and montane zones. Increases in land-use intensity are associated with significant changes in the composition of plant, animal and microorganism communities; stronger modifications of plant and animal communities occur in arid and humid ecosystems, respectively. Temperature, precipitation and land use jointly modulate soil properties, nutrient turnover, greenhouse gas emissions, plant biomass and productivity, as well as animal interactions. Our data suggest that the response of ecosystem functions to land-use intensity depends strongly on climate; more-severe changes in ecosystem functioning occur in the arid lowlands and the cold montane zone. Interactions between climate and land use explained-on average-54% of the variation in species richness, species composition and ecosystem functions, whereas only 30% of variation was related to single drivers. Our study reveals that climate can modulate the effects of land use on biodiversity and ecosystem functioning, and points to a lowered resistance of ecosystems in climatically challenging environments to ongoing land-use changes in tropical mountainous regions.

Many experiments have shown that biodiversity enhances ecosystem functioning. However, we have little understanding of how environmental heterogeneity shapes the effect of diversity on ecosystem functioning and to what extent this diversity effect is mediated by variation in species richness or species turnover. This knowledge is crucial to scaling up the results of experiments from local to regional scales. Here we quantify the diversity effect and its components-that is, the contributions of variation in species richness and species turnover-for 22 ecosystem functions of microorganisms, plants and animals across 13 major ecosystem types on Mt Kilimanjaro, Tanzania. Environmental heterogeneity across ecosystem types on average increased the diversity effect from explaining 49% to 72% of the variation in ecosystem functions. In contrast to our expectation, the diversity effect was more strongly mediated by variation in species richness than by species turnover. Our findings reveal that environmental heterogeneity strengthens the relationship between biodiversity and ecosystem functioning and that species richness is a stronger driver of ecosystem functioning than species turnover. Based on a broad range of taxa and ecosystem functions in a non-experimental system, these results are in line with predictions from biodiversity experiments and emphasize that conserving biodiversity is essential for maintaining ecosystem functioning.

Recently, network approaches have gained increasing popularity in studies of species interactions. These analyses provide important information about structural and functional organisation, as well as on the dynamics of species interactions. Common model systems for network studies include seed dispersal, pollination, and also parasite interactions. Bat flies (Diptera: Streblidae, Nycteribiidae) are obligate blood-sucking ectoparasites of bats. Resource partitioning allows multiple fly species to co-occur on a single host individual, making them an ideal model system for network analyses. Between 2013 and 2018 in Central Panama, 6528 bats from 53 species were examined for the presence of bat flies. Thereof, we collected a total of 6077 bat flies belonging to 52 species. The resulting interaction network showed a significantly higher specificity (H2' = 0.97) and modularity (Q = 0.78) than expected by chance, indicating a very high host specificity of the bat flies. To investigate parasite interactions in the context of host size, host abundance and roosting preferences, we pooled parasite identifications on genus level. The majority of our identified modules were associated with bats using persistent roosting structures. Neither host size nor host abundance appeared to affect module structure. Further, module structure appeared not to be host-phylogeny driven, instead modules were often composed of species known to share roosting structures. Their high host-specificity could put bat flies at risk of extinction in changing environments.

While patterns in species diversity have been well studied across large‐scale environmental gradients, little is known about how species’ interaction networks change in response to abiotic and biotic factors across such gradients. Here we studied seed‐dispersal networks on 50 study plots distributed over ten different habitat types on the southern slopes of Mt Kilimanjaro, Tanzania, to disentangle the effects of climate, habitat structure, fruit diversity and fruit availability on different measures of interaction diversity. We used direct observations to record the interactions of frugivorous birds and mammals with fleshy‐fruited plants and recorded climatic conditions, habitat structure, fruit diversity and availability. We found that Shannon interaction diversity (H) increased with fruit diversity and availability, whereas interaction evenness (EH) and network specialization (H₂) responded differently to changes in fruit availability depending on habitat structure. The direction of the effects of fruit availability on EH and H₂ differed between open habitats at the mountain base and structurally complex habitats in the forest belt. Our findings illustrate that interaction networks react differently to changes in environmental conditions in different ecosystems. Hence, our findings demonstrate that future projections of network structure and associated ecosystem functions need to account for habitat differences among ecosystems.

Aim Climate change and an increase in human disturbance are major drivers of global biodiversity loss. Yet it is not clear to what extent their effects on animal communities are direct or indirectly mediated by changes in biotic factors, such as plant diversity. Here, we disentangle the direct and indirect effects of climate, human disturbance, vegetation structure and plant functional traits on the functional diversity of avian frugivore communities across a large environmental gradient. Location Mount Kilimanjaro. Time period Sampling between November 2013 and October 2015. Major taxa studied Fleshy-fruited plants, frugivorous birds. Methods We sampled plant and bird communities along an elevational and a human disturbance gradient and measured corresponding morphological traits of plants and birds to calculate indices of functional identity and functional diversity of plant and bird communities. We used structural equation models to disentangle direct and indirect effects of all variables on functional identity and diversity of frugivorous bird communities. Results Both functional identity and diversity of frugivorous bird communities were consistently related to the functional identity and diversity of plant communities. Climate had almost exclusively indirect effects on functional identity and diversity of bird communities mediated through effects on plant functional identity and diversity. In contrast, human disturbance also had direct negative effects on bird diversity. Main conclusions We show that plant functional identity and diversity are the most important drivers of functional identity and diversity of frugivorous birds. Although effects of climate on bird communities are almost exclusively mediated indirectly through plant communities, human disturbance resulted in a direct reduction of bird diversity. The high degree of trait matching between interdependent trophic levels over a large environmental gradient demonstrates the importance of biotic drivers for animal communities and shows that biodiversity models need to consider such bottom-up effects in future conditions.

Agriculture and the exploitation of natural resources have transformed tropical mountain ecosystems across the world, and the consequences of these transformations for biodiversity and ecosystem functioning are largely unknown

Species’ functional traits set the blueprint for pair-wise interactions in ecological networks. Yet, it is unknown to what extent the functional diversity of plant and animal communities controls network assembly along environmental gradients in real-world ecosystems. Here we address this question with a unique dataset of mutualistic bird–fruit, bird–flower and insect–flower interaction networks and associated functional traits of 200 plant and 282 animal species sampled along broad climate and land-use gradients on Mt. Kilimanjaro. We show that plant functional diversity is mainly limited by precipitation, while animal functional diversity is primarily limited by temperature. Furthermore, shifts in plant and animal functional diversity along the elevational gradient control the niche breadth and partitioning of the respective other trophic level. These findings reveal that climatic constraints on the functional diversity of either plants or animals determine the relative importance of bottom-up and top-down control in plant–animal interaction networks., Differential responses of plant and animal functional diversity to climatic variation could affect trait matching in mutualistic interactions. Here, Albrecht et al. show that network structure varies across an elevational gradient owing to bottom-up and top-down effects of functional diversity.

[Aim] Biogeographical comparisons of interaction networks help to elucidate differences in ecological communities and ecosystem functioning at large scales. Neotropical ecosystems have higher diversity and a different composition of frugivores and fleshy-fruited plants compared with Afrotropical systems, but a lack of intercontinental comparisons limits understanding of (a) whether plant–frugivore networks are structured in a similar manner, and (b) whether the same species traits define the roles of animals across continents., [Location] Afrotropics and Neotropics., [Time period] 1977-2015., [Taxa] Fleshy-fruited plants and frugivorous vertebrates., [Methods] We compiled a dataset comprising 17 Afrotropical and 48 Neotropical weighted seed-dispersal networks quantifying frugivory interactions between 1,091 fleshy-fruited plant and 665 animal species, comprising in total 8,251 interaction links between plants and animals. In addition, we compiled information on the body mass of animals and their degree of frugivory. We compared four standard network-level metrics related to interaction diversity and specialization, accounting for differences related to sampling effort and network location. Furthermore, we tested whether animal traits (body mass, degree of frugivory) differed between continents, whether these traits were related to the network roles of species and whether these relationships varied between continents., [Results] We found significant structural differences in networks between continents. Overall, Neotropical networks were less nested and more specialized than Afrotropical networks. At the species level, a higher body mass and degree of frugivory were associated with an increasing diversity of plant partners. Specialization of frugivores increased with the degree of frugivory, but only in the Neotropics., [Main conclusions] Our findings show that Afrotropical networks have a greater overlap in plant partners among vertebrate frugivores than the more diverse networks in the Neotropics that are characterized by a greater niche partitioning. Hence, the loss of frugivore species could have stronger impacts on ecosystem functioning in the more specialized Neotropical communities compared with the more generalized Afrotropical communities., We thank Beth A. Kaplin and Norbert J. Cordeiro for their guidance and support for P.J.D., who received a travel grant by The Center for Tropical Studies and Conservation (CTEC). L.C. and I.G. were supported by the Robert Bosch Foundation. D.M.D. (DE 2754/1‐1), F.S. (HE 3041/20‐1), M.Q., V.S., E.L.N. (Research Unit 823‐825), and K.B.G., M.S. and M.G.R.V. (FOR 1246) thank the German Research Foundation (DFG) for funding. F.A.F.J. acknowledges funding by a CAPES scholarship, N.F. and D.G.S. by the Robert Bosch Foundation, M.G., C.E., A.P. and M.A.P. by Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP 2010/52315‐7; 2015/15172‐7; 2016/18355‐8) and Conselho Nacional de Desenvolvimento Científico (CNPq), M.C.M. by Doctoral Fellowships from COLCIENCIAS and Rufford, M.S.S. by Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET) and FONCyT (PICT2013‐2759 and PICT2016‐0608), P.G.B. by CONICET (PIP 2014‐592) and FONCyT (PICT 2013‐1280), R.A.R. by a Doctoral Fellowship from CONICET, R.H. and S.T. (IF/00441/2013) and M.C. (SFRH/BD/96050/2013) by Fundação para a Ciência e Tecnologia, Portugal, and A.T. (CGL2013‐44386‐P) and D.G. (CGL2015‐68963‐C2‐2‐R) by the Spanish government. T.