The influence of local precipitation and temperature on long-term growth dynamics in two species of seasonally dry tropical forest trees were investigated. Growth records were extracted from tree rings in Guanacaste province, Costa Rica. These chronologies provide a long-term (c. 85-y) record of tree growth for two species with contrasting phenologies. Annual growth, in both species, was dependent on annual and/or monthly variation in local precipitation but less so on temperature. For each species, however, patterns of growth reflected unique degrees of sensitivity to monthly rainfall and rainfall during previous years. It is hypothesized that such differences were due to the rooting depth of these species. A review of the literature also indicated similar diverse cambial growth responses by tropical trees to variation in annual and monthly climate. Lastly, it was shown that variation in longer term fluctuations in the Pacific and Atlantic oceans, as measured by the El Niño Southern Oscillation (ENSO) and North Atlantic Oscillation (NAO), significantly influenced local precipitation in Guanacaste only during the wettest portion of the wet season. Such temporal sensitivity may have differentially influenced the longer-term growth of some tropical tree species but not others. Together, these results support the hypothesis that tropical tree species respond individualistically to variation in local and regional climate and that some tropical assemblages may in fact be structured by species-specific differences in soil water-use.

Plant establishment was studied in a swamp forest in the lowland Atlantic rainforest in southeastern Brazil. A topographic gradient allowed a clear distinction between a periodically flooded and a permanently flooded site in the study area. Both sites were mosaics of patches: the former showed soil patches of varying sizes subject to distinct flooding durations (from 0 to 300 d per year) and in the latter, the tank bromeliads which densely occupied the understorey, trap litter and provided ‘suspended soil’ patches. In the periodically flooded site, the occurrence of regenerants (< 1.0 m tall) of the actively regenerating tree populations of Calophyllum brasiliense, Symphonia globulifera and Tovomitopsis paniculata (all Clusiaceae), originating from seed or vegetative organs, was recorded for 1 m2 quadrats placed in flooded (n = 82) and unflooded (n = 103) patches. In the permanently flooded site, dominated by Tabebuia cassinoides (Bignoniaceae), the presence of germinated seedlings was counted for 400 rosettes of the terrestrial tank bromeliad Nidularium procerum. The results showed that unflooded patches, including tank bromeliads, favoured seedling establishment. However, T. paniculata and T. cassinoides colonize patches where flooding lasts longer or is permanent mostly by reproducing vegetatively, while the seeds of C. brasiliense, which showed no form of vegetative reproduction, and S. globulifera are successful in colonizing flooded patches. Germination tests and field observations indicated that, for C. brasiliense, pre- and post-dispersal seed dormancy, a two-phase seed dispersal (water and bats) and seedling tolerance to flooding accounted for this pattern.

A comparative study of the damage caused by Hurricane Hugo on the three main types of forest in the lesser Antilles was carried out in the mangrove forest, the semi-evergreen forest and the rainforests of Guadeloupe. Hurricane impact on these forests showed at first that high canopy trees generally prevented understorey trees from heavy direct damage (shield effect). In the species-poor, structurally homogeneous stands, in the mangrove forest, species' susceptibility was the major determinant in both intensity and type of damage. Damage was heavy in places and showed a certain spatial uniformity. In the rainforest, the high structural complexity of the vegetation generally induced a great variability in the spatial distribution of damage even at a small scale. Due to the presence of tall trees, indirect damage was locally very heavy. Clumps of large trees formed resistant structures which diminished lethal damage below the main canopy (cluster effect). In such a highly structured, species-rich forest, species' susceptibility was concealed at the local scale by the prevailing influence of vertical structure. Because of its relatively rich flora but low stature, the semi-deciduous forest exhibited intermediate patterns of damage. It was concluded that floristic composition and forest structure help to explain, and to predict, hurricane damage to forest cover.

An exhaustive floristic inventory was conducted in a 400-ha block of tropical semideciduous forest in the Chiquitanía region of Santa Cruz, Bolivia. A total of 501 species were collected using both quantitative and traditional plant collecting methods. Trees represented the most diverse life form (124 spp), followed by herbs (101 spp), lianas (85), shrubs (66), herbaceous climbers epiphytes (15) and parasites (two). Floristic diversity was greatest for the forest floor community, when compared to either the understorey or canopy tree communities. Quantitative data were obtained for 336 species in 100 plots which contained a series of nested subplots to sample smaller-stature plants. Stem density for trees (dbh≥5 cm) was 914 trees ha−1, with a total density for all life forms estimated to be 135,000 plants ha−1; the total basal area for the forest was estimated to be 27.6 m2 ha−1. Species richness when measured by standard methods was found to be one of the highest for a dry forest region reported for the Neotropics with a mean of 70.8 spp 0.1 ha−1 (dbh ≥ 2.5 cm) and 50 spp ha−1 (dbh ≥ 10 cm). Comparison of life forms and vertical strata showed that the flora on the forest floor was more diverse than all other vertical strata combined. Habitat heterogeneity was studied using ordination procedures based on floristic data and to identify the characteristic species of three plant communities: granite outcrops, valley forest and upland forest. The most abundant tree species in the study area were Acosmium cardenasii, Neea hermaphrodita, Aspidosperma tomentosa and Galipea trifoliata, while the species with the greatest basal area were Anadenanthera colubrina, Acosmium cardenasii, Caesalpinia floribunda, Aspidosperma tomentosa, Piptadenia viridiflora, Chorisia speciosa, Tabebuia impetiginosa, Centrolobium microchaete, and Machaerium scleroxylon. Most canopy and understorey tree species had a population structure characterized by numerous juveniles and relatively few large trees, while emergent species tended to have a size-class distribution with relatively few juvenile individuals. Lianas and canopy trees were predominantly anemochorous, while understorey trees and shrubs were predominantly zoochorous; herbaceous species were largely autochorous, with only fern species relying on wind dispersal (sporochory). A comparison of structural attributes with other dry forest areas in the Neotropics, demonstrates the heterogeneous nature of vegetation types that are commonly assigned to this ecosystem. The Chiquitano dry forest shares many floristic elements with the semideciduous forests of the Andean piedmont of northwestern Argentina, the Misiones region of eastern Paraguay and northeastern Argentina, as well as the Caatinga region of northeastern Brazil.

This study investigated the nutritional ecology of forest elephants in Kibale National Park, Uganda relative to crop-raiding behaviour, and examined nutritional differences between crops and food consumed by wild elephants. An index of dietary nutrient concentration was determined by quantifying the species and parts of plants consumed along feeding trails, collecting food items, and analysing foods for energy, fibre, protein, minerals and secondary compounds. Frequency of crop raiding was quantified over 13 mo. Energy and protein concentration was within suggested levels, but concentrations of several minerals, particularly sodium, were low relative to requirements based on captive elephants and values reported for other wild populations. The very low sodium concentrations of Kibale elephant diets and low availability of alternative sodium sources, such as soil or water, suggest that sodium drive is very likely in this population. Crops consumed by Kibale elephants had higher Na concentrations and lower concentrations of fibre and secondary compounds than wild diets. The known attraction of elephants to mineral sources throughout their range and the low mineral concentration of leaves, fruits, bark, and stems consumed by forest elephant in this study suggest that mineral nutrition is likely to be an important factor driving elephant behaviour and patterns of habitat use.

Arbuscular mycorrhizal (AM) fungi colonization and response were studied in seedlings of 80 native woody species belonging to different successional groups from the Tibagi River Basin, Paraná State, south Brazil. This study includes data from 43 native woody species already published. The results with 80 species did not differ from the results of the 43 species. The experiment was carried out in a greenhouse in plastic bags filled with a mix of subsoil (85%) and sand (15%), inoculated or not with spores of native AM fungi obtained from rhizosphere soil of different native tree species in an area with natural vegetation dominated by woody pioneer species. The successional groups were represented by 16 pioneer, 20 early secondary, 29 late-secondary and 15 climax species. The AM response and colonization in the greenhouse were 5.9 and 4.2 times greater in the early successional species than in the late-successional species, respectively. Seedlings of 49 woody species were collected in the interior under the canopy of the tropical forest of the Mata dos Godoy State Park and in a cleared area dominated by woody pioneer species. The percentage of AM colonization in the field was 54.9, 40.4, 7.2 and 3.1 for the pioneer, early secondary, late-secondary and climax species, respectively. The response to AM inoculation was strongly and directly related to AM colonization in the greenhouse and field and inversely related to seed weight. The AM colonization in the greenhouse was strongly and directly related to AM colonization in field. The late-successional species showed lower AM colonization and response than early successional species. The accentuated mycotrophism of the early successional species may be involved in their establishment, growth, survival and early forest structuring on low-fertility soils.

Phenological patterns of herbaceous and woody plants well represented in the Sudano-Sahelian savannas of northern Cameroon were studied in relation to water availability in three contrasted soil types. Mean annual precipitation in this region is 800 mm, entirely restricted to the period May-June to September-October, but redistribution of rainwater by varying soil substrates leads to considerable heterogeneity in available water resources for plants. Results are discussed in terms of probable competition for water resources among co-occurring plants and the adaptations shown by observed patterns to prevailing environmental conditions. Water stress is seen to be a limiting factor to plant growth but does not constitute the sole trigger for the phenological phases observed, nor is it the principal factor responsible for the preponderance of annual species in the herbaceous stratum.

Fifteen species of Lepidoptera occurred in large numbers in spring and early summer after the passage of Hurricane Hugo over the north-east of Puerto Rico. Spodoptera eridania (Noctuidae) was the most common of the larvae and fed on 56 plant species belonging to 31 families. All the Lepidoptera fed on early successional vegetation. Some of the plants represent new host plants for these species. The outbreaks appeared to be based on the flush of new foliage that developed in the Luquillo Mountains after the passage of the hurricane. The end of the S. eridania outbreak was concurrent with the consumption of its preferred host plants and to an apparent increment in parasitism by ichneumonids (Hymenoptera). Parasitism by tachinids (Diptera) may have contributed to the reduction in abundance of other Lepidoptera species that were temporarily very abundant. Natural enemies of S. eridania were recorded for the first time in Puerto Rico.

Many woody species in humid tropical forests synchronously flush entire canopies of young red leaves. Numerous unsuccessful attempts have been made to explain the adaptive value of this visually striking phenomenon. In the humid tropics, fungal attack is a potentially important source of mortality for expanding young leaves. We propose that the anthocyanins responsible for the red coloration of young leaves may play a protective role against invasions by leaf-attacking fungal pathogens.

Fungus-growing leaf cutting ants (Atta columbica Guerin) were used in choice tests because they are known to select against leaves or chemicals containing fungicidal properties. In feeding trials with leaf discs from 20 common species, ant preference decreased significantly with increasing anthocyanin content. In feeding trials with pure anthocyanin (3,3',4',5,7-pentahydroxyflavylium chloride) presented on oat flakes, ants again showed a significant dosage dependent preference. This suggests that even low concentrations of anthocyanins may be harmful to the fungal colonies of ants. Additional work on the effects of anthocyanin on leafattacking fungi is encouraged.

To examine the influence of seasonality on the abundance and composition of species in the community, we studied the seed rain and the seed bank in a seasonal, semi-deciduous forest in south-eastern Brazil. Soil samples (depth 3 cm, 0.25 m2) were collected within a 1-ha plot in the dry seasons of 1996 and 1997 and the rainy season of 1997 (84 samples per collection). Thirty-five seed traps (0.25 m2 each) were placed randomly in the forest from March 1997 to February 1998. The greatest density of seeds in the soil occurred in samples collected during the rainy season, the period which coincides with the main fruiting period in this forest. The Sørensen similarity index values for the seed bank composition among the three periods were high (> 0.50). The fallout of propagules was strongly seasonal, with more than half of the annual total number of seeds being caught in the two months around the end of the dry season and beginning of the rainy season. The mean density of seeds in the soil bank was nearly 86% lower than the seed rain density. There was no clear evidence of seasonal effects on species density and richness in this forest.

It has been suggested that vascular epiphyte composition and abundance may be influenced by forest host composition. We studied the epiphyte species distribution among host species of a tropical dry forest in Mexico. All the epiphyte stands supported by woody plants (dbh > 3 cm) of ten forest plots (0.1 ha each) were counted. We measured the dbh of all the hosts, noted their bark characteristics (texture and peeling behaviour), and measured the bark thickness of the most abundant host species. Epiphyte distribution was biased toward a high concentration of epiphytes in three host species and a lower abundance of epiphytes on five host species. This was consistent among epiphyte taxa and host species, allowing us to classify hosts as preferred (with more epiphyte stands than expected by chance) and limiting species (with fewer epiphyte stands than expected by chance), at a community level. Host quality did not relate to mean phorophyte size (measured as basal area) or to bark characteristics (peeling behaviour, thickness and texture) between species. For some epiphyte taxa, the observed distribution indicated that the forest contained preferred and limiting host species mainly. Our data suggest that the epiphyte species in the forest studied could be host limited. We concluded that neither host size nor obvious bark characteristics can be used to predict epiphyte distribution and that further research is necessary.

Tree characteristics were recorded from 2829 standing trees in 24 0.25-ha terra firme forest plots in central Amazonia, 3 y after a surface fire had swept through the study area. Sixteen of the plots were within forest that burnt for the first time at the end of the 1997-98 El Niño (ENSO) event, and the remaining eight plots were within unburnt primary forest. In order to investigate the morphological correlates of tree mortality, we measured tree diameter at breast height (dbh) and bark thickness, and recorded burn height, bark roughness and the presence of latex, resin and buttress roots. Leaf litter depth was also recorded at the base of all trees in the unburnt forest. Using logistic regression models, tree mortality was best explained by the burn height, although dbh and the presence of buttresses were also important. Buttressed trees were associated with deeper leaf litter accumulation at their bases and higher char heights than trees without buttresses. Moreover, trees surviving the fire had significantly thicker bark than living trees in unburnt forest plots, indicating that thin-barked trees are more prone to selective mortality induced by heat stress. Latex did not appear to have had any significant effects on mortality, though resins were less abundant amongst the live trees in the burnt forest than in the unburnt controls. Levels of fire-mediated tree mortality in this study are compared with those in other Amazonian forest regions in light of historical factors affecting tree resistance to fires.

Predators of herbivorous insects play important roles in tropical ecosystems as herbivory may affect structure and diversity of plant populations. Although insectivorous bats are particularly abundant and diverse in the tropics, their impact on herbivorous insects is little understood. To assess prey consumption, we observed the gleaning bat Micronycteris microtis (Phyllostomidae) continuously for 3 mo including 16 full nights at a nightly feeding roost on Barro Colorado Island in Panama using infrared videotaping combined with collection of prey remains. Individual bats consumed about 61–84% of their body mass in arthropods per night. Diet analysis revealed a high percentage of herbivorous insects, constituting more than half (51%) of all prey and over 70% of prey biomass. Dominant prey were caterpillars (33% of prey biomass), and other herbivores including crickets, katydids, scarab beetles and phasmids. Furthermore, a novel feeding behaviour was observed as M. microtis selectively discarded parts of intestines of phytophagous insects before consumption, probably to avoid intake of plant material either for ballast reduction and/or for protection from secondary plant compounds. Combined with estimated feeding rates of insects in sympatric bat species, our data suggest that gleaning bats are important predators of herbivorous insects and might be under-estimated reducers of herbivory in the tropics.

Long-term data on flower and fruit production of the forest community in two lowland tropical rain forests in north-western Sumatra are presented. The proportion of years with mast fruiting was found to be similar to that found elsewhere in Malesia. However, masting at the two sites, 70 km apart, did not coincide, and showed no correlation with the El Niño–Southern Oscillation phenomenon (ENSO). Comparisons with other sites in Malesia suggest a general waning of ENSO's impact toward western Malesia. Spatial variation at various scales in the timing of masting events was noted in Sumatra and elsewhere. This suggests that additional factors to ENSO play a role in determining forest-level mast fruiting, and we hypothesize that frugivorous animals have the opportunity to track mast fruiting. It is hypothesized that asynchrony between nearby areas in masting increases toward the western edge of Malesia.

Large mammals often play important roles in determining the structure and composition of plant communities. This study focused on the extent to which wild pigs (Sus scrofa) influence the dynamics of tree seedlings and saplings in a lowland rain forest at Pasoh Forest Reserve in West Malaysia. Native wild pigs are common in the study area and may significantly influence growth and survivorship of woody plants in the understorey through several activities namely, nest building, soil rooting and seed predation. To test experimentally the impact of pigs on the plant community, eight 49-m2 exclosures were constructed. After 2 y, the number of recruits inside exclosures was three times greater than in unfenced control plots. Stem density was highly correlated with species richness, which also increased significantly inside exclosures. Height growth of plants was greater in the exclosures by 52.5% for trees between 1 and 7 m tall. Trees less than 1 m tall, however, exhibited no differences in growth. Mortality of plants also did not differ between treatments. The observed differences between exclosure and control plots can be attributed to soil-rooting and seed predation, suggesting that these two behaviours of wild pigs are important to plant dynamics in the understorey.

The distribution of the nutrients N, P and K in soil and vegetation and their mobility through litterfall and decomposition in mature and successional stands of a subtropical dry forest were studied in Guánica, Puerto Rico. Soils of the Guánica forest have high total amounts of N (9100 kg/ha), P (1820 kg/ha), and K (7460 kg/ha). However, high extractable Ca (>4000 mg/g) and pH (> 7–8) may explain why only 1.3 and 25% of the total P and K, respectively, were extractable. Total ecosystem storage of N, P and K was 10,300, 1900 and 7700 kg/ha, respectively, of which vegetation stored only 10, 2 and 3%, respectively. Litterfall returned 26, 18 and 180% per year of the N, P and K stored in the ground litter compartment. Trees retranslocated about 30 and 65% of the N and P required to satisfy aboveground net primary production and immobilized P in dead roots. Slow leaf decomposition (7.3 yr for 95% decomposition) released K faster than mass, P as fast as mass, and ash and N slower than mass. The use efficiency of P by litterfall was high compared with other tropical forests, while that of N and K was similar to other tropical and temperate forests. Cutting and regrowth of vegetation resulted in differences in the nutrient concentration in litterfall and nutrient use efficiency of successional vegetation.

We investigated the association between lianas and Bertholletia excelsa (Brazil nut), a long-lived, emergent tree of significant ecological and economic importance in Amazonia. Our objectives were: (1) to determine the relationship between crown liana load and liana number, basal area, and origin in relation to the B. excelsa host; and (2) to determine the relationship between liana load and B. excelsa fruit and nut production, diameter growth, and crown form, position and area. One hundred and forty trees (≥50 cm dbh) were selected with representatives of 10 diameter classes and four liana load categories. To quantify fruit and nut production, fruit counts and nut fresh weights per tree were measured in 2002 and 2003, and annual diameter growth was quantified using dendrometer bands. Trees with lianas produced significantly fewer fruits and had reduced nut fresh weights than liana-free trees. Trees with the most extensive liana loads (>75% crown coverage) were 10.2 times more likely to have crown forms categorized as less than half-crowns or few branches than trees with reduced liana loads. No statistically significant relationship was found between liana load and tree diameter growth. Results suggest that liana cutting might increase B. excelsa fecundity and commercial nut yields.

Relating patterns of species abundance to the processes that determine them at local and regional spatial scales is one of the central aims of ecological research. To examine the relationship between distribution and growth in tropical rain forest, seasonal and annual changes in tree diameter of two tree species with contrasting distribution patterns, Celtis mildbraedii and Strombosia glaucescens, were monitored across topographic gradients in moist semi-deciduous and moist evergreen forest in Ghana over 2 years, 1997-9. Concurrent measurements were made of soil water availability, and during 1997/8, of rainfall, nutrient availability and photosynthetically active radiation (PAR). The two sites received similar, low, total rainfall during 1997/8, but soil water availability in the dry season at the evergreen site exceeded that at the semi-deciduous forest site. Soil water availability was higher during 1998/9 at both sites. During 1997/8, PAR was similar at the two sites. The evergreen site had soils of lower pH, available P, Mg, Ca and K concentrations than the semi-deciduous forest site. Water availability and the sand content of the soil increased and concentrations of total N and C decreased, from summit to valley positions at both sites. Celtis mildbraedii had significantly faster growth rates than S. glaucescens, and growth of both species was greater at the semi-deciduous than the evergreen forest site during the wet year (1998/9) but not during the dry year (1997/8). Celtis mildbraedii in semi-deciduous forest grew less in valley than summit and slope positions. We conclude that in the absence of a severe dry season, growth is higher in semi-deciduous than evergreen forest, and is related to the higher soil fertility in more seasonal forest. The patterns are consistent with the hypothesis that concentrations of available P and/or cations in the soil limit growth in evergreen forest, while concentrations of N in valley soils limit growth of C. mildbraedii in semi-deciduous forest. There was no evidence that variation in PAR influenced growth rates in these forests. A reduction in growth rate on less fertile soils may be a factor determining the distribution limit of dry forest species, such as C. mildbraedii, in evergreen forest.

Changes in structure of a Costa Rican upper montane Quercus forest were studied in twelve 0.1 ha plots along a chronosequence including 8- to 20-year-old Early Secondary Forest (ESF), 25- to 32-year-old Late Secondary Forest (LSF) and mature Primary Forest (PF). In 1.2 ha 2854 stems ≥ 3.0 cm dbh belonging to 42 tree species were recorded. Tree species richness per 0.1 ha ranged from 15 to 30. Diversity indices did not change significantly during succession. Stem density did not differ significantly among successional phases, while basal area was significantly higher in PF than in ESF and LSF. Maximum canopy height and basal area increased linearly during the first three decades of recovery. Height and dbh showed a significant, logarithmic regression for all forest phases. A period of 84 y was estimated as the theoretically minimum time needed for structural recovery. The maximum canopy height and basal area recovered two to five times slower in upper montane than in lower montane or lowland Neotropical forests.

The germination of 43 tree species native to the lowland forests of Malaysia was monitored on forest soil in trays placed in closed-canopy forest, an artificial forest gap, and a large clearing. Germination varied significantly among habitats, with only seven species germinating well in all three sites. Seed germination of most species demonstrated clear patterns of shade tolerance or intolerance identical to those long recognized for tree seedlings. Favoured locations for germination were similar to known distributional patterns of adult individuals in mature and secondary forests. Most forest canopy species germinated in the gap, but germination in the large clearing was dramatically reduced or nil; the regeneration of these species in large clearings will be severely retarded even when seeds are present. Canopy-induced inhibition of germination was apparent in several pioneer species; seeds moved from the forest into the gap or clearing germinated rapidly. Results indicate that natural treefall gaps do not inhibit the germination of most species, but do allow pioneer species to germinate, and therefore regenerate. Although germination occurred in all three locations, the level of germination for particular species varied greatly among habitats. We conclude that germination requirements play a very important role in controlling the species composition of regeneration in forest understorey, gap, and large clearing habitats.