Consequences of frugivore diversity for seed dispersal, seedling establishment and the spatial pattern of seedlings and trees

Many plants depend on frugivorous animals for the dispersal of their seeds. However, it is only poorly known whether regional differences in frugivore diversity have consequences for seed dispersal, seedling establishment, and the spatial distribution of seedlings and trees. This comparative study of seed dispersal investigated the consequences of regional differences in frugivore diversity for two tree species of the genus Commiphora. C. harveyi was studied in South Africa where avian frugivore diversity is high, C. guillaumini was studied in Madagascar where the avian frugivore community is depauperate. At both study sites, the percentages of handled and dispersed seeds in Commiphora trees were quantified by fruit traps, and visitation rates, seed handling rates and dispersal rates were quantified for each animal species for two consecutive years. Seedlings were mapped and the spatial distribution of trees quantified. At both study sites, fruits were mainly eaten by birds. The total percentage of dispersed seeds in South Africa was significantly higher than in Madagascar (70.8% vs. 7.9%) because there was a lack of effective dispersers that swallowed seeds in Madagascar. Seed dispersal benefit, i.e. the increase in the probability of becoming established as a seedling away from parent trees due to dispersal was much higher in Madagascar (80 times higher probability) compared to South Africa (6 times higher). Corresponding with the different dispersal percentages, seedlings in South Africa were found at relatively large distances from the nearest Commiphora tree (median distance=21.0 m), whereas in Madagascar seedlings were found mostly under and close to the nearest Commiphora tree (median distance=0.9 m). Finally, Commiphora trees in the Malagasy study site were clumped, but were more randomly distributed in the South African study site. These results suggest that regional differences in frugivore diversity and behaviour strongly affect seed dispersal of trees, seedling establishment and the spatial distribution of seedlings and trees.     

Fruit tracking,frugivore satiation,and their consequences for seed dispersal

Vertebrate frugivore communities are easily satiated by abundant fruit supplies and, contrary to abiotic dispersal agents, typically disperse only part of the available seed pool. This frugivore satiation is likely to be a widespread phenomenon and should be an influential predictor of plants’ ability to disperse their offspring to suitable establishment sites; yet it has never been systematically quantified. Here I investigate patterns of fruit abundance, frugivore activity and frugivore satiation, and their consequences for seed dispersal in the fleshy-fruited tree Frangula alnus. Based on constant-effort seed trapping conducted over 3 years, I assess densities of total and frugivore-consumed seedfall across two spatial (within/between populations) and two temporal (within/between ripening seasons) scales. Furthermore, I examine relationships between fruit abundance and the amount of seeds that are actually dispersed away from fruiting trees. Frugivore activity tightly matched fruit abundance, although some differences existed between scales. This marked fruit tracking did not prevent a significant frugivore satiation, however, and only 53% of the available fruit crops were actually consumed. The extent of satiation varied most at the within-population level, likely due to the territorial behaviour of important frugivore species. In contrast, levels of satiation remained remarkably invariable through time, suggesting that frugivores behave as opportunists and closely adjust the composition of their diet to the available food supply. Overall, greater fruit abundance resulted in a higher proportion of seeds falling beneath fruiting trees, but it also helped increase the (absolute) number of seeds dispersed. This study shows that frugivore satiation can be an important phenomenon even when frugivores tightly track fruit abundance. Its negative effects on recruitment may be attenuated, however, if greater fruit crops help increase population-wide frugivore activity and the amount of seeds being dispersed to suitable establishment sites.     

Effects of site and fruit size on the composition of avian frugivore assemblages in a fragmented Afrotropical forest

Factors influencing the interaction between fruiting trees and their frugivorous seed dispersers in fragmented Afrotropical landscapes are poorly known. With the use of Mantel statistics we analysed assemblages of frugivorous birds on 58 individual trees belonging to 11 species growing in seven Kenyan cloud forest fragments. Overall, frugivores showed little specialization on particular trees. Fruit size explained a substantial amount of the variation in frugivore assemblages among different tree species at the same site. In addition, frugivore assemblages on conspecific trees were significantly more similar when the trees occurred at the same site. This location effect was attributable to the different sites and forest fragments (of different sizes and disturbance levels) varying in the densities and composition of their avian frugivores, vegetation composition and tree fruiting phenologies. It was consolidated further by the low mobility of most of these avian frugivores, particularly their reluctance to cross between forest fragments. Habitat disturbance and fragmentation may therefore have affected fruit selection, with implications for both seed dispersal and regeneration.     

Comparing the Dispersal of Large-seeded Tree Species by Frugivore Assemblagesin Tropical Montane Forest in Africa

We examined frugivore visitation and seed dispersal of five large-seeded (≥ 5 mm) tree species in tropical montane forest based on their occurrence in frugivorous primate diets: Ekebergia capensis, Olea capensis, Parinari excelsa, Prunus africana , and Syzygium guineense. A total of 21 frugivores in five assemblages ( i.e. , chimpanzees, cercopithecines, large-bodied birds, small-bodied birds, and squirrels) were observed over the study period (August 2006 and October–April 2007). We observed seed dispersal in four of five tree species studied; no dispersal was observed for P. excelsa . Frugivore assemblages did not visit tree species equally. Primates spent the most time in trees and had the largest group size. Large-bodied birds (LB) and chimpanzees dispersed the highest number of seeds per minute. LB and cercopithecines potentially dispersed the greatest number of seeds for E. capensis , and chimpanzees for S. guineense . Our analyses indicated that the mean fruiting duration of the focal tree, time in the tree, and number of species present are important predictor variables for seed dispersal by small- and large-bodied birds, and cercopithecines. The number of fruiting trees in the immediate vicinity of the focal tree further predicted seed dispersal for small-bodied birds (SB). Large-bodied birdseed dispersal also was predicted by time in tree by SB, and the number of individuals for SB and cercopithecines. Cercopithecines (CS) were further explained by the time in tree and number of species (SB & LB), and number of individuals for CS. Our study highlights the complexity of describing the relative importance of a frugivore assemblage to the dispersal of a tree species seeds.     

Different contributions of birds and mammals to seed dispersal of a fleshy-fruited tree

Birds and mammals are important seed dispersers of fleshy-fruited plants. Although their behaviors are different, they frequently consume the same species. Thus to understand the dispersal of fleshy-fruited plants, the contribution of birds and mammals to seed dispersal has to be evaluated. Besides, within birds or mammals, some species may functionally different from others. In this study, we examined seed dispersal of the fleshy-fruited tree Swida controversa focusing on the difference between two frugivore groups (birds and mammals), and differences between species within groups. Collected seeds and S. controversa trees were identified using simple sequence repeat (SSR) genotyping, thus enabling to determine the distance between mother tree and dispersed seeds. The avian species were identified by DNA barcoding of feces, whereas the mammalian species were identified by the shape and smell of feces. Most seeds that fell near or under the maternal trees were dispersed by birds, resulting in short seed dispersal distances (average, 13 m). DNA barcoding detected five taxa of avian dispersers. No differences were detected in seed dispersal distance by different avian taxa (i.e., the distance between dispersed seeds and their maternal trees within the research plot); however the rate of seed immigration from outside the research plot by some avian taxa varied significantly. The seed dispersal distance by mammals was significantly further (127 m; min > 50 m) than that by birds. Additionally, immigrated seeds accounted for approximately two-thirds of mammal-dispersed seeds, indicating that these seeds were from outside the research plot, and that mammals significantly contributed to the long-distance seed dispersal of S. controversa. No differences in seed dispersal distance were detected between different mammalian taxa. Overall, this study revealed that birds and mammals show clearly different seed dispersal patterns, and thus, they play different roles in the regeneration of S. controversa.     

Frugivore Behavioural Details Matter for Seed Dispersal: A Multi-Species Model for Cantabrian Thrushes and Trees

Animal movement and behaviour is fundamental for ecosystem functioning. The process of seed dispersal by frugivorous animals is a showcase for this paradigm since their behaviour shapes the spatial patterns of the earliest stage of plant regeneration. However, we still lack a general understanding of how intrinsic (frugivore and plant species traits) and extrinsic (landscape features) factors interact to determine how seeds of a given species are more likely to be deposited in some places more than in others. We develop a multi-species mechanistic model of seed dispersal based on frugivore behavioural responses to landscape heterogeneity. The model was fitted to data from three-years of spatially-explicit field observations on the behaviour of six frugivorous thrushes and the fruiting patterns of three fleshy-fruited trees in a secondary forest of the Cantabrian range (N Spain). With such model we explore how seed rain patterns arise from the interaction between animal behaviour and landscape heterogeneity. We show that different species of thrushes respond differently to landscape heterogeneity even though they belong to the same genus, and that provide complementary seed dispersal functions. Simulated seed rain patterns are only realistic when at least some landscape heterogeneity (forest cover and fruit abundance) is taken into account. The common and simple approach of re-sampling movement data to quantify seed dispersal produces biases in both the distance and the habitat at which seeds arrive. Movement behaviour not only affects dispersal distance and seed rain patterns but also can affect frugivore diet composition even if there is no built-in preference for fruiting species. In summary, the fate of seeds produced by a given plant species is strongly affected by both the composition of the frugivore assemblage and the landscape-scale context of the plant location, including the presence of fruits from other plants (from the same or different species).     

Seed Dispersal by Avian Frugivores: Non‐random Heterogeneity at Fine Scales

Seed dispersal studies have primarily examined dispersal as a function of distance from the parent tree and/or heterogeneity in dispersal due to animal use of nesting, roosting and sleeping sites. However, non‐random heterogeneity in seed dispersal is also likely to result from the post‐foraging behavior and movement of frugivores which prefer certain trees. To characterize variation in seed rain at fine scales, we studied the dispersal curve of Prunus ceylanica, a primarily bird‐dispersed species. We compared seed rain at conspecifics, heterospecific fruiting trees with similar frugivore assemblages, emergent trees, and the landscape surrounding these trees. Seed rain of P. ceylanica was found to peak globally under the canopy of conspecifics but to peak locally under the canopy and immediate neighborhood of heterospecific fruiting trees. Our results demonstrate that seed rain is highly clumped even at fine spatial scales. A large proportion of seeds are dispersed in specific, localized regions. This variation can have important implications for plant population dynamics and might significantly alter the impact of post‐dispersal processes. Seed dispersal models may need to incorporate this heterogeneity to explain manifestations of spatially explicit dynamics like mixed species ‘orchards’.     

Early fate of Myristica hypargyraea seeds dispersed by Ducula pacifica in Tonga,Western Polynesia

Abstract Although pigeons from the genus Ducula are considered among the best avian dispersers of large seeds in Asia and the Pacific, little has been documented on their role. The early fate of dispersed and undispersed seeds of the large‐seeded tree Myristica hypargyraea A. Gray was studied in order to understand the advantage of seed dispersal by the Pacific Pigeon, Ducula pacifica Gmelin in Tonga. Frequency of visits by frugivores to fruiting trees and dispersal distance of seeds were measured. Pre‐dispersal vertebrate seed predation was assessed, then post‐dispersal predation was measured over 160 days. Overall, pre‐dispersal seed predation by parrots was low but variable among trees sampled. Most seeds (54.7%) in the study area were estimated to be dispersed by D. pacifica; 79.7% of those ingested were expelled directly beneath conspecific fruiting crowns, 20% were dispersed locally and < 0.3% were dispersed more than 300 m into a different forest type. Flying foxes (Pteropus tonganus Quoy and Gaimard) dispersed very few seeds (0.7%) and all were dropped below fruiting crowns. Between 4% and 39% of dispersed and undispersed seeds were still viable, or had established seedlings after 160 days. Most seeds had been removed or killed by rats, and seed survival was highest for locally dispersed seeds (approx. 20 m from source trees and within the M. hypargyraea forest). Although D. pacifica was the only frugivore observed to disperse seeds into this higher zone of survival, overall they did not confer a great advantage to seed survival since significant numbers of seeds/seedlings also persisted under fruiting crowns (27% under crowns compared with 39% locally dispersed). Nevertheless, D. pacifica was the only vector by which seeds were regularly moved within the M. hypargyraea forest and over longer distances, and hence, D. pacifica still plays a significant role in the regeneration of M. hypargyraea.     

Spatial–temporal distribution of ornithochorous seeds from an Elaeagnus umbellata community dominating a riparian habitat

The dispersal efficiency and potential distribution of ornithochorous seeds of Elaeagnus umbellata in a riparian habitat were evaluated to clarify this species' establishment site in relation to the disturbance regime of the floodplain. Fruit removal by avian frugivores was monitored using fruit bags, and the spatial distribution of excreted seeds was quantified by seed traps set randomly on a gravel bar as an isolated seed source in the Yoshino River throughout an autumn fruiting season. Although more than 45% of the fruits remained on the twigs in the fruit bags, almost all fruits on the control twigs without fruit bags were exploited by the beginning of January. The fruit removal rate and seed dispersal distance were positively correlated with an increase in wintering bird species and their abundance. Intact bird‐dispersed seeds of E. umbellata were trapped within a 400‐m range and damaged seeds were limited to traps set within 50 m from the seed source. Frugivore behavior, such as feces excretion on rocks near water drinking sites and perching on surrounding woodland, greatly influenced the spatial and temporal dispersal pattern of the seed rain. In the present study, the avian frugivores showed upstream seed dispersal; thus, in years with stochastic autumnal floods, secondary dispersal via hydrochory downstream may be facilitated. The intensive seed dispersal in E. umbellata indicates that the present distribution of parent trees in the restricted elevation range of the gravel bars is the result of survival through disturbance, rather than seed dispersal limitation.     

Can remnant frugivore species effectively disperse tree seeds in secondary tropical rain forests?

Seed dispersal by frugivores in tropical rain forests is important for maintaining viable tree populations. Over the years, vertebrate assemblages in tropical forests have been altered by anthropogenic disturbances, leading to concerns about the ability of remnant vertebrates to substitute for the lost or declining vertebrate populations. We compared vertebrate composition and frugivore visitation rates as an indirect measure of rate of seed dispersal in three tropical rain forests in Uganda, namely Mabira, Budongo and Kibale Forests. Mabira is highly disturbed, Kibale is little and Budongo is intermediate. The aim was to determine whether vertebrate assemblages in differentially disturbed forests had comparable abilities to disperse seeds and whether tree species were equally vulnerable to loss of seed dispersers. Assemblages of forest generalist species were similar in all forests, but specialists were less abundant in the heavily disturbed forest. Remnant frugivores in the heavily disturbed forest were mainly small-bodied species that spat seeds beneath fruiting trees compared to large-bodied species observed in the less disturbed forests that ingested and carried away the seeds. We postulate that the quantity of seeds dispersed in heavily disturbed forests is much reduced due to low visitation rates of frugivores and the absence of large frugivores that consume large quantities of fruit. The quality of seed dispersal is affected as well by the distance over which seeds are moved. Assessment of vulnerability of trees shows no evidence for disperser substitution for trees producing large fruits. Fruit trees with low nutritional contents and digestibility were least visited in frugivore-impoverished forests. The loss of large specialist frugivores is likely to affect recruitment of many trees, especially of species that cannot establish beneath adult conspecifics.     

Plants,position and pollination: Planting arrangement and pollination limitation in a revegetated eucalypt woodland

The spatial arrangement of plants is a key determinant of pollination services in natural ecosystems. Despite this, plant arrangements are rarely considered an important characteristic of revegetated communities and this may be limiting successful pollination dynamics in these re‐created systems. We assessed if aggregated South Australian Blue Gum (Eucalyptus leucoxylon, F. Muell) had greater pollination success than more dispersed trees by measuring fruit set, seed production and germination in a revegetated eucalypt woodland. The proportion of buds that developed into fruit (fruit set) was similar between aggregated (five flowering Blue Gum within 30 m) and dispersed (no flowering Blue Gum within 30 m) trees (<14%). Aggregated and dispersed trees produced a similar number of seeds per fruit in 2015, when fewer trees flowered and those that flowered produced fewer flowers (ca 8 seeds/fruit). In 2016, when flowering was more abundant, aggregated trees produced more seeds per fruit than dispersed trees (12 seeds/fruit compared to 4 seeds/fruit). Despite differences in seed number, there was no difference in the quality of seed produced, indicating outcrossed pollen was received by both aggregated and dispersed trees. Although outcrossing was likely, this did not prevent pollination limitation, with the addition of outcrossed pollen increasing fruit and seed set for trees in both arrangements. Consequently, we propose that planting individuals of the same species (conspecifics) in clumps, rather than intermixing species, can increase seed set in revegetated eucalypts, particularly in good flowering years. However, aggregated plantings will be most effective if the needs and foraging behaviour of pollinators are also considered, to attract and sustain pollinators in revegetated systems, thereby minimizing the risk of pollination limitation.     

Forest cover,hunting pressure,and fruit availability influence seed dispersal in a forest‐savanna mosaic in the Congo Basin

Forest fragmentation, reduced forest cover, and hunting pressure are the main threats affecting animal‐mediated seed dispersal. However, their combined effects on seed dispersal rates have been simultaneously investigated only rarely, and never in Africa. We aimed to disentangle the effects of forest cover, hunting pressure, frugivore abundance, and fruit availability at the local and landscape scales on the seed dispersal rates of Staudtia kamerunensis (Myristicaceae). To estimate the percentages of seed dispersal failure (undispersed seeds), we quantitated fruit remains below fruiting trees distributed across five contrasting sites in a semi‐natural forest‐savanna mosaic in the Democratic Republic of Congo. We used statistical analyses accounting for spatial autocorrelation and found that forest cover in the surrounding landscape, hunting level, the associated abundance of dispersers, and fruit availability all had significant effects on the percentage of seed dispersal failure. The combination of high fruit availability and reduced abundance of seed dispersers could accelerate seed disperser satiation, causing the seed dispersal system to be saturated. Our study highlights how two major factors associated with anthropogenic activities, forest cover and hunting, affect seed dispersal by animals. These findings could have far‐reaching implications for our understanding of tree‐frugivore interactions and the conservation of tropical communities.     

Bird Assemblage and Visitation Pattern at Fruiting Elmerrillia tsiampaca (Magnoliaceae) Trees in Papua New Guinea

Most tropical trees produce fleshy fruits that attract frugivores that disperse their seeds. Early demography and distribution for these tree species depend on the effects of frugivores and their behavior. Anthropogenic changes that affect frugivore communities could ultimately result in changes in tree distribution and population demography. We studied the frugivore assemblage at 38 fruiting Elmerrillia tsiampaca, a rain forest canopy tree species in Papua New Guinea. Elmerrillia tsiampaca is an important resource for frugivorous birds at our study site because it produces abundant lipid-rich fruits at a time of low fruit availability. We classified avian frugivores into functional disperser groups and quantified visitation rates and behavior at trees during 56 canopy and 35 ground observation periods. We tested predictions derived from other studies of plant–frugivore interactions with this little-studied frugivore assemblage in an undisturbed rain forest. Elmerrillia tsiampaca fruits were consumed by 26 bird species, but most seeds were removed by eight species. The most important visitors (Columbidae, Paradisaeidae and Rhyticeros plicatus) were of a larger size than predicted based on diaspore size. Columbidae efficiently exploited the structurally protected fruit, which was inconsistent with other studies in New Guinea where structurally protected fruits were predominantly consumed by Paradisaeidae. Birds vulnerable to predation foraged for short time periods, consistent with the hypothesis that predator avoidance enhances seed dispersal. We identified seven functional disperser groups, indicating there is little redundancy in disperser groups among the regular and frequent visitors to this tropical rain forest tree species.     

Impact of habitat structure and fruit abundance on avian seed dispersal and fruit predation

So far, it is poorly understood how differential responses of avian seed dispersers and fruit predators to changes in habitat structure and fruit abundance along land-use gradients may translate into consequences for the seed dispersal of associated plants. We selected a gradient of habitat modification (forest, semi-natural, and rural habitat) characterized by decreasing tree cover and a high variation in local fruit availability. Along this gradient we quantified fruit removal by avian seed dispersers and fruit predators from 18 Sorbus aucuparia trees. We analyzed the relative importance of tree cover and fruit abundance in explaining species richness, abundance and fruit removal rates of both guilds from S. aucuparia trees. Species richness and abundance of seed dispersers decreased with decreasing tree cover, whereas fruit removal by seed dispersers decreased with decreasing fruit abundance independent of tree cover. Both variables had no effect on species richness, abundance and fruit removal by fruit predators. Consequently, seed dispersers dominated relative fruit removal in fruit-rich sites but the dispersal/predation ratio shifted in favor of predation in fruit-poor habitat patches. Our study demonstrates that variation in local habitat structure and fruit abundance can cause guild-specific responses. Such responses may result in a shift in fruit removal regimes and might affect the dispersal ability of dependent fruiting plants. Future studies should aim at possible consequences for plant recruitment and guild-specific responses of frugivores to disturbance gradients on the level of entire plant–frugivore associations.     

Avian frugivory and seed dispersal of a large fruited tree in an Indian moist deciduous forest

Seed dispersal systems in degraded areas can be compromised following the decline of large-bodied frugivore populations responsible for their dispersal. In this context we examined the seed dispersal ecology of a large fruited deciduous tree (Dillenia pentagyna) along a forest degradation gradient in India. We examined the effect of structural components of vegetation and frugivore foraging behavior on D. pentagyna seed dispersal. Depauperate mammalian community and declined large avian frugivores e.g. hornbills in our study site make this system a specialized one and currently dependent on only two large bodied avian frugivores. Seed dispersal followed an overall leptokurtic pattern and the seed dispersal kernels were best explained by an inverse power function. Seed dispersal kernels in dense forest indicated longer dispersal distances than moderately dense forest and degraded forest. In degraded areas, no dispersal away from the crown was recorded for D. pentagyna and it occurred at low density. Canopy foliage abundance of the surrounding vegetation of the focal trees was best explained by quantity of seed dispersal by large avian frugivores. The number of avian frugivore species those are effective disperser of D. pentagyna decreased along the degradation gradient. Avian frugivore behavior in terms of visitation and seed swallowed is a determining factor that controls quantity of seed dispersal. Our study underscores deleterious impact of forest degradation on avian disperser community which in turn would affect regeneration capacity of degraded forest.     

Intraspecific variation in fruit–frugivore interactions: effects of fruiting neighborhood and consequences for seed dispersal

The extent of specialization/generalization continuum in fruit–frugivore interactions at the individual level remains poorly explored. Here, we investigated the interactions between the Neotropical treelet Miconia irwinii (Melastomataceae) and its avian seed dispersers in Brazilian campo rupestre. We built an individual-based network to derive plant degree of interaction specialization regarding disperser species. Then, we explored how intraspecific variation in interaction niche breadth relates to fruit availability on individual plants in varying densities of fruiting conspecific neighbors, and how these factors affect the quantity of viable seeds dispersed. We predicted broader interaction niche breadths for individuals with larger fruit crops in denser fruiting neighborhoods. The downscaled network included nine bird species and 15 plants, which varied nearly five-fold in their degree of interaction specialization. We found positive effects of crop size on visitation and fruit removal rates, but not on degree of interaction specialization. Conversely, we found that an increase in the density of conspecific fruiting neighbors both increased visitation rate and reduced plant degree of interaction specialization. We suggest that tracking fruit-rich patches by avian frugivore species is the main driver of density-dependent intraspecific variation in plants’ interaction niche breadth. Our study shed some light on the overlooked fitness consequences of intraspecific variation in interaction niches by showing that individuals along the specialization/generalization continuum may have their seed dispersed with similar effectiveness. Our study exemplifies how individual-based networks linking plants to frugivore species that differ in their seed dispersal effectiveness can advance our understanding of intraspecific variation in the outcomes of fruit–frugivore interactions.

     

Frugivory on Persea lingue in temperate Chilean forests: interactions between fruit availability and habitat fragmentation across multiple spatial scales

Habitat degradation and fragmentation are expected to reduce seed dispersal rates by reducing fruit availability as well as the movement and abundance of frugivores. These deleterious impacts may also interact with each other at different spatial scales, leading to nonlinear effects of fruit abundance on seed dispersal. In this study we assessed whether the degradation and fragmentation of southern Chilean forests had the potential to restrict seed dispersal the lingue (Persea lingue) tree, a fleshy-fruited tree species. Of five frugivore bird species, the austral thrush (Turdus falcklandii) and the fire-eyed diucon (Xolmis pyrope) were the only legitimate seed dispersers as well as being the most abundant species visiting lingue trees. The results showed little or no direct effect of habitat fragmentation on seed dispersal estimates, possibly because the assemblage of frugivore birds was comprised habitat-generalist species. Instead, the number of fruits removed per focal tree exhibited an enhanced response to crop size, but only in the more connected fragments. In the fruit-richer fragment networks, there was an increased fragment-size effect on the proportion of fruits removed in comparison to fruit-poor networks in which the fragment size effect was spurious. We suggest that such nonlinear effects are widespread in fragmented forest regions, resulting from the link between the spatial scales over which frugivores sample resources and the spatial heterogeneity in fruiting resources caused by habitat fragmentation and degradation.     

Dispersal modes affect tropical forest assembly across trophic levels

We examined assemblages of trees and two major groups of vertebrate seed dispersers, birds and primates, in Ugandan protected areas to evaluate the roles of dispersal limitation and species sorting in community assembly. We conducted partial Mantel tests to investigate relationships between community similarity, environmental distance and geographic distance. Results showed that environmental factors, specifically temperature and rainfall, significantly and more strongly structured tree assemblages than geographic distance. Analysis of tree dispersal modes revealed wind‐dispersed tree guilds were significantly dispersal limited but trees dispersed by animals were not. For assemblages of vertebrate seed dispersers, dispersal limitation significantly and more strongly structured assemblages of primates than species sorting whereas environmental factors significantly and more strongly structured assemblages of birds than dispersal limitation. We therefore examined whether trees dispersed by primates were more dispersal limited than trees dispersed by birds. We found consistent trends that primate fruit trees were more dispersal limited than bird fruit trees using three definitions of dispersal syndromes based on fruit color. Our results suggest that the dispersal abilities of primary consumers may affect the distribution of primary producers at large spatial scales.     

Standard yet unusual mechanisms of long-distance dispersal: seed dispersal of Corymbia torelliana by bees

Mellitochory, seed dispersal by bees, has been implicated in long-distance dispersal of the tropical rain forest tree, Corymbia torelliana (Myrtaceae). We examined natural and introduced populations of C. torelliana for 4 years to determine the species of bees that disperse seeds, and the extent and distance of seed dispersal. The mechanism of seed dispersal by bees was also investigated, including fruit traits that promote dispersal, foraging behaviour of bees at fruits, and the fate of seeds. The fruit structure of C. torelliana , with seed presented in a resin reward, is a unique trait that promotes seed dispersal by bees and often results in long-distance dispersal. We discovered that a guild of four species of stingless bees, Trigona carbonaria, T. clypearis, T. sapiens , and T. hockingsi, dispersed seeds of C. torelliana in its natural range. More than half of the nests found within 250 m of fruiting trees had evidence of seed transport. Seeds were transported minimum distances of 20–220 m by bees. Approximately 88% of seeds were dispersed by gravity but almost all fruits retained one or two seeds embedded in resin for bee dispersal. Bee foraging for resin peaked immediately after fruit opening and corresponded to a peak of seed dispersal at the hive. There were strong correlations between numbers of seeds brought in and taken out of each hive by bees ( r =  0.753–0.992, P  < 0.05), and germination rates were 95 ± 5%. These results showed that bee-transported seeds were effectively dispersed outside of the hive soon after release from fruits. Seed dispersal by bees is a non-standard dispersal mechanism for C. torelliana, as most seeds are dispersed by gravity before bees can enter fruits. However, many C. torelliana seeds are dispersed by bees, since seeds are retained in almost all fruits, and all of these are dispersed by bees.     

Introduction of mammalian seed predators and the loss of an endemic flightless bird impair seed dispersal of the New Zealand tree Elaeocarpus dentatus

Understanding the mutualistic services provided by species is critical when considering both the consequences of their loss or the benefits of their reintroduction. Like many other Pacific islands, New Zealand seed dispersal networks have been changed by both significant losses of large frugivorous birds and the introduction of invasive mammals. These changes are particularly concerning when important dispersers remain unidentified. We tested the impact of frugivore declines and invasive seed predators on seed dispersal for an endemic tree, hinau Elaeocarpus dentatus, by comparing seed dispersal and predation rates on the mainland of New Zealand with offshore sanctuary islands with higher bird and lower mammal numbers. We used cameras and seed traps to measure predation and dispersal from the ground and canopy, respectively. We found that canopy fruit handling rates (an index of dispersal quantity) were poor even on island sanctuaries (only 14% of seeds captured below parent trees on islands had passed through a bird), which suggests that hinau may be adapted for ground‐based dispersal by flightless birds. Ground‐based dispersal of hinau was low on the New Zealand mainland compared to sanctuary islands (4% of seeds dispersed on the mainland vs. 76% dispersed on islands), due to low frugivore numbers. A flightless endemic rail (Gallirallus australis) conducted the majority of ground‐based fruit removal on islands. Despite being threatened, this rail is controversial in restoration projects because of its predatory impacts on native fauna. Our study demonstrates the importance of testing which species perform important mutualistic services, rather than simply relying on logical assumptions.