Tropical ecosystems support a diversity of species and ecological processes that are unparalleled anywhere else on Earth. Despite their tremendous social and scientific importance, tropical ecosystems are rapidly disappearing. To usher tropical ecosystems and the human communities dependent upon them through the environmental transformations of the 21st century, tropical biologists must provide critical knowledge in three areas: 1) the structure and function of tropical ecosystems; 2) the nature and magnitude of anthropogenic effects on tropical ecosystems; and 3) the socio-economic drivers of these anthropogenic effects. To develop effective strategies for conservation, restoration, and sustainable management of tropical ecosystems, scientific perspectives must be integrated with social necessities. A new set of principles built on a framework for pursuing relevant tropical biological research will facilitate interdisciplinary approaches, integrate biological knowledge with the social sciences, and link science with policy. We propose four broad recommendations for immediate action in tropical biology and conservation that are fundamental to all biological and social disciplines in the tropics: 1) assemble and disseminate information on life's diversity in the tropics; 2) enhance tropical field stations and build a worldwide network to link them with tropical field biologists at their field sites; 3) bring the field of tropical biology to the tropics by strengthening institutions in tropical countries through novel partnerships between tropical and temperate zone institutions and scientists; and 4) create concrete mechanisms to increase interactions between tropical biologists, social scientists, and policy makers.

Litter quality parameters such as nitrogen and lignin content correlate with decomposition rates at coarse scales, but fine-scale mechanisms driving litter decomposition have proven more difficult to generalize. One potentially important driver of decomposition is the activity of extracellular enzymes that catalyze the degradation of complex compounds present in litter. To address the importance of this mechanism, we collected 15 Hawaiian plant litter types and decomposed them in fertilized and control plots for up to two years. We measured litter nutrient content and carbon chemistry prior to decomposition, as well as extracellular enzyme activities, mass loss, and litter nutrient content over time. We found that water-soluble carbon content, cellobiohydrolase activities, and polyphenol oxidase activities were significantly correlated with mass loss. Enzyme activities and decomposition rate constants both varied significantly by litter type, and fertilization increased mass loss rates in five litter types. Some litter types that decayed faster under fertilization also showed time-dependent increases in carbon-degrading enzyme activities, but others decayed faster independent of enzyme changes. These results suggest that extracellular enzyme activities partially determine litter decomposition rates, but high soluble carbon content may circumvent the requirement for enzyme-catalyzed decomposition.

Interest in tropical secondary forests has grown as large areas of agriculture have been abandoned in recent decades; yet, there are few long-term studies of post-agriculture vegetation recovery in the tropics. In this study, we compared the vegetation structure and floristic composition of old-growth and 40-year-old secondary riparian forests in the Cordillera Central, Dominican Republic. Canopy height and stem density of woody plants were similar between forest types, but basal area of trees was 27 percent lower in secondary forests. Introduced tree species comprised 20 percent of the basal area and dominated the understory of secondary forests. Life-form diversity was higher in old-growth forests as arborescent ferns, the palm species, and epiphytic bromeliads, orchids, and bryophytes were much more abundant. The number of species of epiphytic orchids and bromeliads, ground ferns, and herbaceous plants was also significantly higher in old-growth forests. The species density of woody plants and vines, however, was comparable between forest types, and vine abundance was significantly higher in secondary forests. The high importance of introduced tree species and the delayed recovery of several plant life-forms have important implications for the conservation of plant diversity in secondary forests in the tropics. The robust regeneration of woody structure despite the long land tenure (ca 60 yr) by farmers is probably due to the nutrient-rich alluvial soils and low-intensity agriculture. This study revealed the potential for the rapid recovery of woody plant diversity and structure in fertile secondary forests adjacent to mature forest seed sources and the more delayed recovery of nonwoody plant diversity and abundance.

We measured changes in insect herbivore intensity and guild structure in two tropical forests of Costa Rica over a three-year period. We estimated herbivory levels using discrete leaf litter samples to account for factors that influence the types of leaves most likely to be preserved in the fossil record. We found that leaf area removed by insect herbivores did not vary significantly between years within the two sampled forests, but differed across years between forests. The lack of variation in herbivore intensity within forests suggests that estimates of herbivore damage from fossil assemblages may reflect levels of herbivore intensity typically experienced by a given forest, at least over a short ecological timescale. The differences in herbivore intensity between similarly categorized forests suggest that simple one-to-one comparisons between fossil and modern sites can be problematic. The distribution of functional feeding groups also differed on an interannual basis both within and between forests. As guild structure can be highly variable, caution should be taken when comparing modern herbivore guild structures with those from fossil assemblages. Studies of herbivory in fossil assemblages have yielded much lower levels of damage than those found in this study as well as those measured from other modern forests. Leaf fossil preservation, environmental differences, and changes in plant–insect associations through time may account for these differences.

From an ecological standpoint, the symbiosis between leaf-cutting ants and their fungus can be described as an herbivore feeding on plant material harvested by the ants, metabolized by the fungus, and finally removed by the ants as refuse. Here, we report on a nutrient balance (soluble and insoluble carbon and nitrogen compounds) between harvested and refuse material collected from six Acromyrmex lundi nests from a semiarid woodland in Córdoba, Argentina. The nutrient balance was calculated using cellulose as an indicator. Harvested material had low protein (10%) and high fiber content (57%). Refuse material was rich in fiber (53%), total nitrogen (2%), and ash content (28%). Balance was negative for soluble carbon and phenols. Soluble fibers, lignin, and true protein remained unchanged. Surprisingly, crude protein and all soluble nitrogen fractions had a positive balance, suggesting additional nitrogen sources besides plant material. Phenol utilization by the leaf-cutting ant fungus suggests that phenols are not harmful to the fungus. Inability to metabolize insoluble carbon compounds and a net nitrogen gain in the refuse material were peculiar and interesting features of the ant–fungus system.

We examined feeding by the mangrove tree crab Aratus pisonii in Tampa Bay, Florida, in relation to the percent dry weight of carbohydrate, protein, phenolics, condensed tannins, ash, carbon, nitrogen, carbon:nitrogen ratio, water content, and sclerophylly for leaves of the red mangrove Rhizophora mangle. Comparisons of leaf chemistry were made among leaves that experienced variable levels of crab damage. Because R. mangle is the crab's preferred food source based on damage patterns in the field, comparisons of R. mangle leaf chemistry were made in relation to that of the black mangrove Avicennia germinans and the white mangrove Laguncularia racemosa. We observed a negative relationship between level of leaf damage and percent dry weight of nitrogen, carbohydrates, condensed tannins, and sclerophylly. In contrast, a positive relationship was found between leaf damage and the carbon:nitrogen ratio. The chemical constituents that provided the best explanation for differences in damage among the three mangrove species include condensed tannins, nitrogen, carbon:nitrogen ratio, carbohydrates, phenolics, water content, and ash. The results from this study suggest that chemistry only partially explains food preference by A. pisonii. It appears that A. pisonii feeding behavior and preference may be influenced by a more complex series of factors and interactions, which may include reproduction by, predation on, and interspecific competition with A. pisonii.

Ants of the genus Oecophylla are predators of other insects and are able to protect a variety of terrestrial plants against pest insects; however, observations on the ecology of these ants in mangrove forests are lacking. General observations on the ecology of Oecophylla smaragdina were carried out in a Thai mangrove forest to determine if these ants can protect their host plants in less favorable mangrove habitats. Leaf herbivory and the density of O. smaragdina ants were measured on Rhizophora mucronata trees at two sites. The results showed a negative correlation between ant density and herbivory. At both sites, the mean percent damaged leaf area was more than four times higher on trees without ants compared to “ant-trees.” A significant negative correlation was found between tree mean percent leaf damage and the density of ants on the tree. Furthermore, on trees with ants, there was less herbivory on leaves close to ant nests compared to other leaves on the tree. Most damage was caused by chrysomelid beetles (62%) and sesarmid crabs (25%) and both types of herbivory were significantly reduced on ant-trees.

The Neotropical beetle Cephaloleia fenestrata is a specialist herbivore on Pleiostachya pruinosa (Marantaceae) in upland and flood zone habitat of lowland tropical wet forest in Costa Rica. Cephaloleia fenestrata spends its entire life cycle on P. pruinosa, feeding primarily in rolled young leaves as adults, feeding in the concavity of leaf petioles as larvae, and laying eggs and pupating on the leaf petioles. Egg development time is about average for Chrysomelidae, approximately ten days. Cephaloleia fenestrata larvae, like other rolled leaf hispines, have an extremely long development period (94 d) consisting of only two instars (compared to a range of three to six in other Chrysomelidae). The 30-day pupal stage is also long, but within the range of that observed in other Chrysomelidae. Adults live an average of six weeks. The generation time for C. fenestrata is thus approximately six months. Male C. fenestrata lived longer and had a higher capture probability than females. There was no difference between males and females in the movement probabilities between upland and flood zone habitat. Rolled P. pruinosa leaves are both ephemeral (available to the beetles for approximately two days) and somewhat rare (1 in ca 15 ramets has a rolled leaf). Adults must move from leaf to leaf ca 36 times during their average adult life span. Still, C. fenestrata adults demonstrate high site fidelity.

The “deflection hypothesis” asserts that conspicuous marginal patches on insect wings function to deflect predator attacks toward such patches and away from more vital body parts. As a result of selection from predator attacks, these marks are predicted to increase the probability of escape by tearing relatively easily. To test if a conspicuous marginal patch is weak relative to a homologous wing area without such a patch, hindwing tear weight was compared among three Pierella species (Satyrinae) differing in the presence of a conspicuous patch in the hindwing tornus. The species with a conspicuous white hindwing patch (P. astyoche) had significantly lower tear weights than the two species lacking the patch (P. lamia and P. lena). Forewing length did not explain variation in wing-tear weight, but wing-tear weight was positively related to insect age in a manner consistent with the deflection hypothesis. Older individuals of P. lamia and P. lena had higher tear weight, whereas this relationship was absent in P. astyoche. These results represent the first direct evidence that deflection marks on butterfly wings are relatively weak and should have an increased tendency to tear when handled by a predator.

Detrimental effects of introduced honeybees (Apis mellifera L.) on native plant pollination have been predicted based on their observed deterrence or expulsion of native pollinators or their depletion of floral resources. Here, a case is reported in which floral visitation by the introduced honeybee affects male fitness and probably fruit and seed production of a tropical tree without affecting the resource sought by the native pollinator. Clusia arrudae is a dioecious species and is pollinated by individuals of Eufriesea nigrohirta that visit its flowers to collect resin. Male flowers, however, are also visited by individuals of A. mellifera, which remove ca 99 percent of their pollen grains. When E. nigrohirta leaves flowers previously visited by A. mellifera, they carry on their bodies less than 0.1 percent of the pollen grains carried by bees leaving flowers not visited by the honeybee. This may explain why the frequency of A. mellifera at male flowers is negatively correlated with the number of seeds produced by female flowers. This novel situation may affect other plants offering alternative rewards for pollinators, such as resins, oils, and aromatic compounds.

We describe the association between a new species of flower mite (tribe Melicharini) and Neotropical flower-visiting bats. Analogous to the well-known hummingbird–mite interactions, these mites appear to use bat visitors of the lowland tropical palm Calyptrogyne ghiesbreghtiana as a means of traveling to new inflorescences. Phenology data show tight coupling between the occurrence of mites on bats and flowering phenology of C. ghiesbreghtiana, suggesting that the palm is the most important host for the mites.

We studied the population dynamics of the lesser long-nosed bat Leptonycteris curasoae in a cave in central Mexico for a two-year time. The population had substantial seasonal changes in size and composition during this time. Colony size was largest from February to July (22,000–27,000 adults) and contained equal numbers of males and females. In June and July, males had enlarged testes and presumably mated with females. In August, the population size began to decline and was mostly composed of pregnant females. Between September and December, the colony was composed exclusively of pregnant and lactating females and their young, and population size decreased to about one-third of the observed maximum size. In January, adult males returned to the roost and the population size increased. We also observed an increase in body mass and fat accumulation in both sexes, apparently related to reproductive activity. This is the first report of the continuous presence of a substantial female population of L. curasoae throughout the year in a single roost in Mexican tropics, indicating that some populations of L. curasoae in central Mexico complete their life cycle without having to migrate.

Tropical forests are among the most heterogeneous environments on earth, and food resources for many animals are patchy both in time and space. In Africa's equatorial forest, permanent trails created and maintained by forest elephants are conspicuous features. Trails may be several meters wide and continue for tens of kilometers. Speculation on which resources determine the distribution of trails has identified fruit, browse, and mineral deposits as candidates. In this study, the relationships between these habitat variables and elephant trails were investigated. The size of individual trails and the density of the trail system increased dramatically with proximity to mineral deposits. Fruit tree basal area decreased with perpendicular distance from trails, while that of non-fruit trees did not. Fruit tree abundance and basal area were significantly higher on trail intersections than random sites and increased with intersection size. No relationship was found between monocotyledon browse abundance and elephant trail system characteristics. Clumped resources, which are at least partially reliable, provide a high nutritional payback, and are not rapidly depleted and can thus be visited repeatedly, appear to influence permanent trail formation by forest elephants. Permanent trails may allow naive individuals or those with imperfect knowledge to locate and acquire important resources.

Canopy invertebrates may reflect changes in tree structure and microhabitat that are brought about by human activities. We used the canopy fogging method to collect ants from tree crowns in primary forest, secondary forest, and pasture in a Neotropical cloud forest landscape. The total number of species collected was similar in primary forest (21) and pasture (20) habitats, but lower in secondary forest (9). Lower diversity in secondary forest was caused by lower species density (no. of species per sample). Rarefaction curves based on number of species occurrences suggest similar community species richness among the three habitats. This study has implications for conservation of tropical montane habitats in two ways. First, arboreal ant species density is reduced if secondary forest replaces primary forest, which increases the chance of extinction among rare species. Second, pasture trees may serve as repositories of primary forest ant communities due to similar tree structure.

We show that tropical forest tree saplings with greater belowground carbon allocation have more breakage scars along their stems. We suggest the existence of alternative carbon allocation strategies in relation to physical damage in the forest understorey. “Tolerators” allocate more belowground, have enhanced resprouting ability and slower aboveground growth, whereas “escapers” allocate less belowground, are not well prepared for recovering from damage, but grow fast enough to escape from the damage-susceptible size class.

The tropical mountain cloud forest bromeliads studied thus far have been reported as self compatible; however, whether or not autogamy and apomixis occur is not known. We characterized the breeding system of Werauchia sintenisii using a series of treatments to detect self-compatibility, autogamy, and apomixes. The natural fruit set of this population was 97 percent. Pollen origin (self vs. cross) did not affect fruit and seed set. Werauhia sintenisii is autogamous but not apomictic. All pollinator-excluded flowers produced fruits, but flowers that were emasculated and caged did not produce any fruit. Factors affecting the evolution of autogamy in this system are discussed.

In tropical forests, rodents exert a prominent role as post-dispersal seed predators, while other vertebrates apparently have a minor effect on seed predation. In some forest fragments in southeast Brazil, however, the density of rodents is unusually low, whereas terrestrial granivorous birds are abundant. In this study, we used seeds of a second-growth tree (Croton priscus, Euphorbiaceae) to investigate experimentally the relative importance of rodents and birds as seed predators in a forest fragment. We also estimated the density of terrestrial granivorous birds and trapped small mammals. Results indicate that, as a consequence of decreased rodent population and increased bird abundance, the latter are at least as important as the former as the main post-dispersal seed predators of C. priscus. We interpret such apparent imbalance in the interaction between C. priscus and their seed predators as a possible consequence of forest fragmentation.

We quantified predation pressure on first instar nymphs of a stick insect in predator-exclusion experiments in the forest of Barro Colorado Island, Panamá. After considering intrinsic mortality (19%) and potential emigration (negligible), we estimated that 54 percent of the nymphs died due to predation in a two-week period. Predation on nymphs was highest at night and may explain the low abundance of Metriophasma diocles in the understory.

We analyzed Ficus perforata seed removal by canopy ants under four treatments: syconia seeds; howler monkey-defecated washed seeds; syconia seeds with howler dung; and defecated seeds with howler dung. Ants removed more defecated washed seeds in the wet season, while during the dry season, removal was greater for syconia seeds and defecated washed seeds. Our results suggest that ants tended to remove more seeds when they were not embedded in feces; however, even seeds set on howler dung were still attractive to ants.