In the tropics, limestone caves in karstic areas are known for their unique biodiversity. However, many caves remain unstudied and little is known about underlying gradients that determine diversity and biomass in aquatic microhabitats. Here, we sampled zooplankton and benthos in a set of 12 aquatic caves, locally called closed cenotes in Yucatán, Mexico. Our aim was to explain diversity patterns and differences in biomass with particular attention for correlations between bat colony characteristics and other biota. Compared with caves that support photosynthesis, diversity was low with an average of four planktonic and two benthic species in these dark caves. Undetectable phosphorus concentrations in the water suggest this nutrient is limiting. Several associations hint at a potential link between bat abundance and functional guild composition, water quality and aquatic biota. As such, more bats were linked to higher nitrate concentrations. Yet this was not translated to higher invertebrate biomass, probably since phosphorus is limiting. Overall, the trends found in this survey suggest that bats could be important as fertilizers of the caves although mechanistic links that mediate the flux of nutrients need to be confirmed experimentally.

Differential tree seedling recruitment across forest-savanna ecotones is poorly understood, but hypothesized to be influenced by vegetation cover and associated factors. In a 3-y-long field transplant experiment in the forest-savanna ecotone of Ghana, we assessed performance and root allocation of 864 seedlings for two forest (Khaya ivorensis and Terminalia superba) and two savanna (Khaya senegalensis and Terminalia macroptera) species in savanna woodland, closed-woodland and forest. Herbaceous vegetation biomass was significantly higher in savanna woodland (1.0 ± 0.4 kg m−2 vs 0.2 ± 0.1 kg m−2 in forest) and hence expected fire intensities, while some soil properties were improved in forest. Regardless, seedling survival declined significantly in the first-year dry-season for all species with huge declines for the forest species (50% vs 6% for Khaya and 16% vs 2% for Terminalia) by year 2. After 3 y, only savanna species survived in savanna woodland. However, best performance for savanna Khaya was in forest, but in savanna woodland for savanna Terminalia which also had the highest biomass fraction (0.8 ± 0.1 g g−1 vs 0.6 ± 0.1 g g−1 and 0.4 ± 0.1 g g−1) and starch concentration (27% ± 10% vs 15% ± 7% and 10% ± 4%) in roots relative to savanna and forest Khaya respectively. Our results demonstrate that tree cover variation has species-specific effects on tree seedling recruitment which is related to root storage functions.

Megaherbivores play a central role in the evolution and functioning of ecosystems. In tropical forests elephant species are some of the few remaining megaherbivores. Through elephant foraging, nutrients that would be locked in leaves and stems, taking months or years to decay, are quickly liberated for use. In 10 experimental sites in Kibale National Park, Uganda, we set up 10 pairs of plots (4 × 4 m), each pair involved one treatment, elephant dung addition, and one control. After 1 y, we quantified growth (height and leaf number) and survival of young light-demanding (12) and shade-tolerant (19) plant species (439 stems in total). In general, the addition of elephant dung did not increase seedling growth, and it only increased the number of leaves in shade-tolerant plants with a large initial number of leaves. Researchers have speculated that the loss of elephants would shift the composition of African forests to slow-growing tree species. However, this is not supported by our finding that shows some slow-growing shade-tolerant plants grew more new leaves with additional nutrient input from elephant dung, a condition that would occur if elephant numbers increase.

Tree establishment in tropical dry forests is constrained by drought-related seedling mortality during early stages of recruitment. Predicted increases in the duration of growing-season droughts in the future pose a significant threat to these ecosystems that could significantly alter their vegetation structure and composition. Here, we examined drought tolerance in seedlings of seven common dry-forest tree species from the Indian subcontinent. We conducted a dry-down experiment on 3-wk-old seedlings, and asked whether the key plant functional traits, specific leaf area (SLA), leaf dry matter content (LDMC), seed size and stem specific density (SSD) were good predictors of seedling growth under well-watered conditions, and survival during drought. Seedlings displayed substantial drought tolerance with most seedlings surviving for more than 2 wk under protracted drought. Seed size in combination with SLA predicted seedling growth under well-watered conditions and seed size predicted survival under drought. In contrast to our expectations, seedlings with lower SSD survived for longer without water. Our results suggest that dry-forest species will be differentially affected by the predicted increases in the duration of growing-season droughts, and detrimental effects will be more severe for species with smaller seeds.

Understanding why ecological communities contain the species they do is a long-standing question in ecology. Two common mechanisms that affect the species found within communities are dispersal limitation and environmental filtering. Correctly identifying the relative influences of these mechanisms has important consequences for our understanding of community assembly. Here variable selection was used to identify the environmental variables that best predict tropical forest primate community similarity in four biogeographic regions: the Neotropics, Afrotropics, Madagascar and the island of Borneo in South-East Asia. The environmental variables included net primary productivity and altitude, as well as multiple temperature, precipitation and topsoil variables. Using the best environmental variables in each region, Mantel and partial Mantel tests were used to reanalyse data from a previously published study. The proportion of variance explained increased for each region. Despite increases, much of the variation remained unexplained for all regions (R2: Africa = 0.45, South America = 0.16, Madagascar = 0.28, Borneo = 0.10), likely due to different evolutionary and biogeographic histories within each region. Nonetheless, substantial variation among regions in the environmental variables that best predicted primate community similarity were documented. For example, none of the 14 environmental variables was included for all four regions, yet each variable was included for at least one region. Contrary to prior results, environmental filtering was an important assembly mechanism for primate communities in tropical forests worldwide. Geographic distance more strongly predicted African and South American communities whereas environmental distance more strongly predicted Malagasy and Bornean communities. These results suggest that dispersal limitation structures primate communities more strongly than environmental filtering in Africa and in South America whereas environmental filtering structures primate communities more strongly than dispersal limitation in Madagascar and Borneo. For communities defined by genera, environmental distance more strongly predicted primate communities than geographic distance in all four regions, which suggests that environmental filtering is a more influential assembly mechanism at the genus level. Therefore, a more nuanced consideration of environmental variables affects conclusions about the influences of environmental filtering and dispersal limitation on primate community structure.

Staminodes are sterile stamens that produce no pollen, exhibit diverse structures and perform various functions. Flowers of Phanera yunnanensis possess three fertile stamens with large anthers and long filaments, and seven staminodes with tiny anthers and short filaments. To investigate the adaptive significance of staminodes in this species, we studied effects of staminode removal on pollen removal and deposition, flower visitation rate and fruit set in Xishuangbanna, south-western China. Four species of nectar-foraging pollinators visited flowers, mostly Amegilla zonata and Apis cerana (2.80 ± 0.15 and 1.76 ± 0.41 visits h−1 per flower, respectively). Staminode removal did not affect fruit set, but increased visitation by A. cerana by 2.6-fold, reduced visitation by A. zonata by 68% and increased the pollen removal rate for both pollinators (all effects were significant). Staminode removal significantly reduced pollen deposition rate for A. zonata, but not for A. cerana. These results suggest that the staminodes of P. yunnanensis filter which insects act as pollinators and affect pollen removal and deposition rates. By reducing pollen removal rates, staminodes may implement a pollen-dispensing schedule that spreads pollen dispersal from individual flowers over multiple pollinators. By altering pollen deposition rates, staminodes may influence reproductive fitness in other ways.

The rain forests of South-East Asia represent a biodiversity hotspot of terrestrial leeches, but we have only fragmentary and often anecdotal information on this component of tropical communities. To address the foraging tactics of terrestrial leeches, we studied the vertical distribution of Haemadipsa picta on foliage in a Bornean tropical rain forest. We investigated the links between leech body length and the above-ground height of their ambush positions under natural conditions and in a choice experiment performed under semi-natural conditions. We studied 167 leeches, which varied considerably with respect to body length (4–29 mm). On average, the leeches ambushed at lower heights under natural conditions than under experimental conditions (47.7 cm vs. 67.5 cm), though the heights of ambush positions overlapped considerably. Leeches that chose higher ambush positions under natural conditions consistently chose higher ambush positions in the experiment (Pearson r = 0.29). Under both natural and experimental conditions, leech body length was positively correlated with the height of ambush positions (Pearson r = 0.48); on average, a 1-mm increase in body length corresponded to a 2–3 cm higher ambush position. Our findings suggest that H. picta individuals actively choose hunting locations according to their above-ground height, shifting their ambush sites over ontogeny to higher foliage. We hypothesize that growing leeches might target different prey following ontogenetic shifts in the foraging optima, ultimately decreasing intraspecific competition.

We investigated whether changes in nutrient availability affected N, P, S and polyphenol concentrations in different leaf-development stages of three brevideciduous and three evergreen dominant woody species in a nutrient-limited savanna woodland in Central Brazil. Treatments included eight years of annual fertilization with 100 kg ha−1 of N, P, N plus P and control, each replicated in four randomized 15 × 15-m plots. All species increased S concentrations (minimum 28%) in young and mature leaves in fertilized plots. Dalbergia miscolobium decreased total phenol concentrations with P (−34.3%, −23.7%) and NP fertilization (−28.2%, −17.1%). Blepharocalyx salicifolius increased total phenol (27.6%, 18.8%) and tannin (46.3%; 43.5%) in P fertilized and increased total phenol (33.9%) and tannin (27.8%, 43.5%) in NP fertilized plots. Total phenol concentration decreased with leaf age in Ouratea hexasperma, Styrax ferrugineus and Blepharocalyx salicifolius, which also decreased tannin concentration with leaf age. For all treatments, brevideciduous species had higher N, P, total phenols and tannin concentrations and lower S concentration than evergreens. These differences between phenological groups suggest that tropical ecosystems responses to environmental changes are more complex than anticipated by global vegetation models, with consequences for predictions in ecosystem functions and resilience.

Cascading effects of high trophic levels onto lower trophic levels have been documented in many ecosystems. Some studies also show evidence of extended trophic cascades, in which guilds dependent on lower trophic levels, but uninvolved in the trophic cascade themselves, are affected by the trophic cascade due to their dependence on lower trophic levels. Top-down effects of large mammals on plants could lead to a variety of extended trophic cascades on the many guilds dependent on plants, such as pollinators. In this study, floral-visitor and floral abundances and assemblages were quantified within a series of 1-ha manipulations of large-mammalian herbivore density in an African savanna. Top-down effects of large mammals on the composition of flowers available for floral visitors are first shown, using regressions of herbivore activity on metrics of floral and floral-visitor assemblages. An extended trophic cascade is also shown: the floral assemblage further altered the assemblage of floral visitors, according to a variety of approaches, including a structural equation modelling approach (model with an extended trophic cascade was supported over a model without, AICc weight = 0.984). Our study provides support for extended trophic cascades affecting floral visitors, suggesting that trophic cascades can have impacts throughout entire communities.

Ground-dwelling ants are active foragers that may extend their foraging area into the vegetation, although the factors affecting their diversity in the suspended litter of understorey plants remain overlooked. To evaluate the influence of the distance between strata, litter biomass and plant size on the ant fauna, the litter ant assemblage of the suspended stratum was compared with the ground immediately below the understorey treelet Erythrochiton brasiliensis (Rutaceae) in an Atlantic Forest, south-eastern Brazil. We collected 1364 ants from 26 ant species. The suspended litter ant assemblage represented a subset of the ground-dwelling ants present in soil litter. The beta diversity results primarily from the high ant species turnover among individual suspended-litter samples, and among ground-litter samples, while species turnover among suspended-ground pairs is lower. Additionally, plant height was not important in determining the species turnover between strata. However, plant height positively correlated with ant species richness, probably because of the increased number of microhabitats. These results suggest that suspended litter in the forest understorey can provide the conditions for ground-dwelling ants to forage and nest, functioning as a vertical extension of resources and microhabitat conditions present in the ground litter.

Litter breakdown is an important ecological process at the bottom of food webs in streams. Previous studies have been based only on a temporal interval of a single season, thus ignoring seasonal variation in litter input and community structure. We investigated organic matter input in a Brazilian savanna stream and the influence of its associated hyphomycetes on the invertebrate community. Organic matter input was sampled monthly and the leaves submitted to decomposition experiments. There were lower breakdown rates and higher invertebrate species richness and abundance during the dry season, which reached their maximum in July due to low stream discharge. Invertebrate composition was best explained by hyphomycetes (mainly by Flagellospora curvula and Anguillospora filiformis). Hyphomycetes have the capacity to degrade complex compounds of litter and to rapidly absorb nutrients by growing branched filaments, thus making the leaves more favourable for consumption by invertebrates. Shredder abundance was negatively related to litter richness, indicating possible species-specific relationships. We observed a sequential process with increased leaf litter input promoting an increase in hyphomycetes biomass, which in turn favoured invertebrate density.

Caliochory, or seed dispersal by birds as nest material, has been reported for several species, but its effectiveness remains unclear in most cases. Darwin’s finches are traditionally regarded as seed predators, but the observation of two nests challenges this assumption by demonstrating that they can act as seed dispersers via caliochory. Darwin’s finches incorporate cotton-like materials into their nests, including seeds of Darwin’s cotton (Gossypium darwinii), a shrub endemic to the Galápagos (Ecuador). Bird nests typically break down after intense rainfall, so the seeds incorporated into nests might benefit from suitable conditions for germination. By simulating the germination conditions experienced over a 72-h period by cotton seeds in a naturally fallen nest, this study qualitatively confirms the long-term viability of at least a small fraction of the seeds at the surface of the nest. Darwin’s finches might therefore provide seed-dispersal services to Darwin’s cotton and possibly, other native and exotic plants of the Galápagos commonly incorporated into nests. However, larger confirmatory studies are needed.

Addressing what affects predation among caterpillars, we conducted an experiment in a Bornean rain forest on 212 clay models of Tortricidae caterpillars (the herbivore) and 53 trees of Kopsia pauciflora (the host), located either in the open or under closed canopies. We predicted that the frequency of predatory attacks towards caterpillars increases (1) in canopy gaps and (2) on leaves damaged by herbivory, but (3) decreases among caterpillars that wrap their body in leaves. Each plant with caterpillar models was consecutively allocated to one of four treatments: caterpillars artificially rolled in leaves vs caterpillars on unrolled leaves, and caterpillars on artificially damaged vs undamaged leaves. Each time, caterpillar models were placed on the plants for 48 h, and then replaced with new models that were subjected to a new treatment. On average, our caterpillar models had a 21% chance of being attacked per 24 h. More attacks were performed by insects (81.6%) than birds. The attack frequency did not depend on the canopy cover or on leaf damage, but decreased among models rolled in leaves. This pattern was mainly attributable to insect attacks, which were also more frequent than bird attacks. Overall, the Tortricidae caterpillars seem to suffer comparable predation rates in the open and under closed canopies irrespective of the herbivory damage of leaves, but their leaf-rolling behaviour might reduce predation.

Leaf-cutting ants are dominant herbivores in Neotropical rain forests, and their colony densities increase in disturbed habitats such as forest edges. However, while it is well-established that leaf-cutting ants profit from changes to the food-plant community, the phylogenetic dimension of this ant–plant interaction remains poorly understood in fragmented forests. We studied diet composition of Atta cephalotes in the edge and interior of Atlantic forest in north-east Brazil (8°30′S, 35°50′W). We applied phylogenetic signal analysis to investigate the diet across plant lineages and performed phylogenetic generalized linear models to analyse the diet in both habitats. We found a phylogenetic signal in diet and in leaf mechanical resistance, which means that A. cephalotes selects closely related food plants with less resistant leaves. Most preferred species belong to Malpighiales, Rubiaceae and Melastomataceae. We also found that irrespective of phylogeny, ants select food plants with less resistant leaves, both in edge and interior. However, ants choose more abundant plants only in edges. High abundance of optimal diet facilitates foraging in forest edges and explains why colony densities increase in disturbed habitats. Finally, by favouring or disfavouring specific clades, leaf-cutting ants contribute to changes in the phylogenetic structure of tropical rain forests, e.g. phylogenetic impoverishment.

Herbaceous plants are often under-studied in tropical forests, despite their high density and diversity, and little is known about the factors that influence their distribution at microscales. In a 25-ha plot in lowland Amazonian rain forest in Yasuní National Park, Ecuador, we censused six species of Heliconia (Heliconiaceae) in a stratified random manner across three topographic habitat types. We observed distribution patterns consistent with habitat filtering. Overall, more individuals occurred in the valley (N = 979) and slope (N = 847) compared with the ridge (N = 571) habitat. At the species level, Heliconia stricta (N = 1135), H. spathocircinata (N = 309) and H. ortotricha (N = 36) all had higher abundance in the valley and slope than ridge. Further, H. vellerigera (N = 20) was completely absent from the ridge. Conversely, H. velutina (N = 903) was most common in the drier ridge habitat. The two most common species (H. stricta and H. velutina) had a reciprocal or negative co-occurrence pattern and occurred preferentially in valley versus ridge habitats. These results suggest that taxa within this family have different adaptations to the wetter valley versus the drier ridge and that habitat partitioning contributes to coexistence.

Studies integrating variation in species composition among sites are useful in understanding the impacts of land-use changes on the spatial distribution of biodiversity. However, the failure to recognize the distinction between beta diversity components, dissimilarity due to species replacement (i.e. turnover) and dissimilarity due to species loss from site to site (i.e. nestedness), can lead to inappropriate use of some indices. Here, we evaluated how the spatial distribution of anuran beta diversity components, turnover and nestedness, is associated with local and landscape descriptors in a tropical agricultural landscape with a recent history of agriculture expansion in south-eastern Brazil. Overall, 27 anuran species were found in the region with average ± SD species richness in each pool of 9.5 ± 3.5 species, ranging from 4 to 15 species. We observed that species turnover was the major component for anuran dissimilarity among pools, indicating that anuran species occurring in species-poor pools are not subsets of anuran species occurring in species-rich pools. Local variables and geographic distance were not important descriptors explaining the variation of anuran beta diversity. In contrast, the distance of the pools to the nearest forest fragment explained 16% of the variance in total beta diversity, 5% of the nestedness component and 2% of spatial turnover. Our results show that pools distributed across farmland landscapes are harbouring different anuran species composition, and together, these pools are contributing to the regional diversity of anurans in this region which is considered one of the most deforested and fragmented within Brazil.