Woody plants diversity,floristic composition and land use history in the Amazonian rain forests of Madidi National Park,Bolivia

A floristic inventory of woody plants was carried out to analyse the relationships between floristic similarity and geographical distance, and to compare the effect of land use history on the floristic composition between sites. Three lowland and two submontane sites were studied in Madidi, Bolivia. In one site, there is evidence of an Inca ruin. A total of 877 species and 12,822 individuals of woody plants with a diameter at breast height ≥2.5 cm were recorded in 44 0.1–ha plots. Fisher’s Alpha index values were slightly higher for the lowlands than for the submontane. Floristic similarity was higher within sites than between sites as measured by both Sørensen and Steinhaus indexes. The fact that the 30 most important species per site (totalling 94 species) accounted for 61.7% of total individuals, support the hypothesis that Amazonian plant communities are dominated by a limited set of species, genera and families. On the other hand, 18 out of the 94 species were reported in a single site, suggesting that some species are patchy in distribution and may be environmentally determined. Both the oligarchy and environmental-determinism hypotheses can be complementary in order to understand floristic patterns of this region. The Ruins submontane site is floristically the most distinct, and past human disturbance is likely to be the main reason. Since species diversity (ranging from 53 to 122 species per plot) and density (ranging from 157 to 503 per plot) are highly variable in Madidi, to characterize the diversity of a site, it is necessary to quantify an average of 10 0.1-ha plots in a relatively small geographical area.     

Functional diversity of herbaceous species under different fire frequencies in Brazilian savannas

Fire plays an important role in determining the structure of savannas, so that frequent fires tend to favor the herbaceous species in savannas. Functional diversity is a measure of the range of the species functional traits and is thought to be related to many functions and properties like community stability, nutrient cycling, and productivity, for example. We examined if functional diversity of herbaceous species was different along a fire gradient, and if observed functional diversity differed from a random assembly of species traits. We sampled three sites with different fire frequencies. According to the intermediate disturbance hypothesis, we expected the functional diversity to be higher in the intermediate fire frequency site. Subsequently, we excluded rare species from the analyses – species with less than ten individuals in each site – to verify their contribution to the observed patterns of functional diversity. We expected to find no difference with the exclusion of rare species. Functional diversity values, either considering all species or excluding the rare ones, were not different among sites, suggesting that (1) frequent fires may not affect overall functioning of the herbaceous community in savannas and (2) rare herbaceous species seem to present lower functional complementarity. According to the neutral theory, most of the herbaceous species may have evolved stochastically in niches more similar than competitive exclusion could allow. Also, functional diversity was not different from what one may expect by chance in any of the three sites. We also calculated the floristic similarity among sites. When we considered qualitative data, we found floristic differences between the sites of high and intermediate as well as between high and low fire frequency sites, but there was no difference between the intermediate and low fire frequency situation. Floristic similarities were different among all the three sites when the abundances of species were considered. Consequently, different fire conditions may change species composition without affecting the functional diversity of the herbaceous layer of savannas, corroborating the idea that savannas are more stable in functional than in floristic terms.     

Vegetation structure: Fine scale relationships with soil in a cerrado site

There is a lack of understanding on factors influencing the occurrence of high species heterogeneity at fine scale in the Brazilian cerrado. Soil is a major determinant of vegetation in the Brazilian cerrado and an important candidate to influence species distribution at fine scales, since soil features vary at very small distances, whereas many environmental variables are relatively homogeneous at such scale. We tested plant-soil relationships at fine scales in a cerrado site. We placed 100 contiguous 25 m² plots, where we identified all woody individuals and measured several soil features. We did partial redundancy analysis, controlling for spatial autocorrelation, to test for relationships between soil features and floristic composition. We also did multiple regressions or spatial autoregressive models to test for relationships between soil features and: (1) the abundance of the five most common species, (2) total abundance, (3) richness, (4) evenness, and (5) diversity. We found weak relationships between soil and floristic composition, richness, and total abundance, which, coupled with also weak relationship found in another study with plant available water, indicate there is no major environmental variable influencing vegetation at fine scales, but several of them interacting. Organic matter was positively related with the abundance of Myrsine umbellata and was negatively related to evenness. Although a causal relationship cannot be inferred with certainty, the dominance of Myrsine umbellata seems to be related to a positive feedback with soil.     

Effects of the geomorphometric characteristics of the local terrain on floristic composition in the central Brazilian Amazon

The objective of this study was to identify the effects of local geomorphometry on the abundance, richness and floristic composition of tree species in the central Brazilian Amazon. Forty‐six 0.25‐ha plots in different phyto‐ecologic sites were sampled, and their trees were inventoried. Geomorphometric data (elevation, slope, aspect, plan and profile curvatures) were derived from Shuttle Radar Topography Mission data. A detrended correspondence analysis (DCA) was used to examine the floristic distribution patterns among plots. In addition, geomorphometric variables were submitted to multiple regression analysis to identify the variables influencing floristic composition (represented by the first DCA component), abundance and species richness. Correlation analyses between the number of individuals from each species and the first DCA component were performed to evaluate the contribution of each species. Analysis of the results could not confirm an effect of geomorphometry alone on species richness and abundance, although floristic composition was significantly influenced by profile curvature and elevation. Despite the relatively low variation in altitude at the study site, species were found to be sensitive to terrain peculiarities such as elevation and profile curvature, which can constrain particular ecologic niches and contribute to the spatial distribution patterns of species.     

Long-term Effects of <Emphasis Type="Italic">Lupinus lepidus</Emphasis> on Vegetation Dynamics at Mount St. Helens

The nitrogen-fixing legume Lupinus lepidus is the most abundant herb on new volcanic surfaces at Mount St. Helen. We compared vegetation structure in 30 Lupinus colonies in three age classes (old, mature, or young based on known years of their establishment) to adjacent sites that were sparsely populated by Lupinus. Our goals were to determine if the age of colonies affected either species composition or vegetation structure and if colonies altered vegetation structure of a site. Species richness increased with site age and colonies had significantly greater richness. Mature colonies had lower cover and frequency of other species, while old colonies had significantly higher cover of other species, particularly mosses. Frequencies in colonies were higher than in adjacent sites. Diversity measures were highest in old colonies and least in mature ones. ANOVA showed that density effects of Lupinus were stronger than age effects on richness, cover, frequency, dominance, and diversity. The floristic similarity between colonies and adjacent sites declined with colony age, and sites without Lupinus became heterogeneous with age. Detrended correspondence analysis demonstrated floristic change as colonies aged and that the differences between colonies and sparse sites increased with age. Regressions of Lupinus frequency with measures of structure showed weak second order relationships that suggested that low densities of Lupinus inhibited other species, while higher densities promoted abundance and diversity. Together these results suggest that Lupinus is a key element that alters floristic successional trajectories and accelerates succession. However, its effects are complicated by the wide ranges of Lupinus density, the sequence of invasion, and by herbivores.     

Diversity,floristic and structural patterns of cerrado vegetation in Central Brazil

The cerrado has been identified as one of the richest and most threatened biomes of the world, but few phytogeographical studies have been undertaken in the region. A total of 70 land systems based on climate, landscape and soils have been identified in the region, but it remains to be seen if the distribution and structure of the plant communities support these divisions. The aim of this work was to compare the floristic and structural similarity of cerrado sensu stricto within and between three physiographic units, named Pratinha, Veadeiros and São Francisco, which contain six land systems in central Brazil and cover 10 degrees of latitude and five degrees of longitude. The woody vegetation of 15 selected sites of the cerrado sensu stricto physiognomy was surveyed under a standardized methodology. The number of species per site varied from 55 to 97, with most sites having around 60 to 70 species, and Shannon´s diversity indices ranged from 3.44 to 3.73, with most sites around 3.5 suggesting high alpha diversity. Sørensen´s floristic similarity index was high, with all Figures above 0.5 between the sites in the same land system in each physiographic unit but low between sites in different land systems in the Veadeiros. Czekanowski similarity indices were lower than Sørensen’s in the comparisons due to a high structural differentiation between the sites. There is a large overlap in species occurrence in the sites but the size of their populations is very different at each site. Therefore, the high beta diversity is mostly due to differences in abundance of species between sites. The sites were separated by physiographic units, considering the first three divisions of TWINSPAN classification. The first axis of DCA ordination showed a gradient going from the cerrado on deep soils in Pratinha, through to those on sandy soils in São Francisco and ending on the shallower soils of the Veadeiros. Land systems conformed well with the floristic and structural variations of the vegetation, indicating their potential use in designing a network of conservation areas in the cerrado region and as a basis for decision-making on management.     

Genotypic diversity of a dominant C4 grass across a long-term fire frequency gradient

Aims and Methods Diversity-disturbance research has focused on community diversity, but disturbance frequency could impact diversity within species as well, with important consequences for community diversity and ecosystem function. We examined patterns of genetic diversity of a dominant grass species, Andropogon gerardii, in native North American tallgrass prairie sites located in eastern Kansas that have been subjected to a gradient of fire frequency treatments (burned every 1, 2, 4 or 20 years) since the 1970s. In addition, we were able to assess the relationships between genetic diversity of A. gerardii, species diversity and productivity across this range of fire frequencies.Important findings We found no significant relationships between genetic diversity of A. gerardii at the local scale (1 m 2 plot level) and disturbance frequency (burned 2 to 32 times over a 38-year period). However, at the site level (i.e. across all plots sampled within a site, ~100 m 2) there were differences in genotype richness and composition, as well as genomic dissimilarity among individuals of A. gerardii. Genotype richness was greatest for the site burned at an intermediate (4-year) frequency and lowest for the infrequently (20-year) burned site. In addition, genotypes found in the frequently burned sites were more similar from each other than expected by random chance than those found in the infrequently burned sites. Genotype composition of A. gerardii was not significantly different between the frequently burned sites (annual vs. 2 year) but did differ between frequently burned and infrequently burned sites (1 and 2 year vs. 4 and 20 year, etc.). Together, these results suggest site-level ecological sorting of genotypes in intact prairie across a broad gradient of disturbance frequencies, likely driven by alterations in environmental conditions. Frequent fire promotes the abundance of dominant grass species, reduces plant community diversity and impacts ecosystem processes such as productivity. Our study suggests that genetic diversity within dominant grass species also may be affected by disturbance frequency, which could have important implications for how species are able to respond to disturbance.     

Genetic diversity of Lotus corniculatus in relation to habitat type,species composition and species diversity

The genetic variability of Lotus corniculatus, a common and important fodder legume, was studied in relation to habitat type and to species diversity at a local level. The study was conducted in Cholomontas mountain, northern Greece, at altitudes of 760–870 m. Genetic material was selected from four forested, at the edge of forest, and four open grassland sites, and was studied with the aid of ISSR molecular markers. The plant cover at each study site was measured and the floristic composition was estimated. The percentage of graminoides increased in grassland sites with high grazing intensity, while the abundance of legumes, including L. corniculatus, forbs and woody species, increased in forested sites with low grazing intensity. Gene diversity H_E within the studied populations of L. corniculatus ranged from 0.167 to 0.213 and Shannon index (I) from 0.269 to 0.340. Genetic differentiation was detected between habitats as well as among the populations in each habitat, although it was low, 3% and 7% respectively. However, genetic differentiation was significant within the populations (90%). Genetic diversity of L. corniculatus was not correlated with species diversity (richness and evenness) either within or between populations, while was negatively correlated with L. corniculatus abundance.     

Improving the assessment of species compositional dissimilarity in a priori ecological classifications: evaluating map scale,sampling intensity and improvement in a hierarchical classification

Question: Can species compositional dissimilarity analyses be used to assess and improve the representation of biodiversity patterns in a priori ecological classifications? Location: The case study examined the northern‐half of the South‐east Queensland Bioregion, eastern Australia. Methods: Site‐based floristic presence–absence data were used to construct species dissimilarity matrices (Kulczynski metric) for three levels of Queensland's bioregional hierarchy – subregions (1:500 000 scale), land zones (1:250 000 scale) and regional ecosystems (1:100 000 scale). Within‐ and between‐class dissimilarities were compiled for each level to elucidate species compositional patterns. Randomized subsampling was used to determine the minimum site sampling intensity for each hierarchy level, and the effects of lumping and splitting illustrated for several classes. Results: Consistent dissimilarity estimates were obtained with five or more sites per regional ecosystem, 10 or more sites per land zone, and more than 15 sites per subregion. On average, subregions represented 4% dissimilarity in floristic composition, land zones approximately 10%, and regional ecosystems over 19%. Splitting classes with a low dissimilarity increased dissimilarity levels closer to average, while merging ecologically similar classes with high dissimilarities reduced dissimilarity levels closer to average levels. Conclusions: This approach demonstrates a robust and repeatable means of analysing species compositional dissimilarity, determining site sampling requirements for classifications and guiding decisions about ‘lumping’ or ‘splitting’ of classes. This will allow more informed decisions on selecting and improving classifications and map scales in an ecologically and statistically robust manner.     

Soil Bacterial and Fungal Community Structure Across a Range of Unimproved and Semi-Improved Upland Grasslands

Changes in soil microbial community structure due to improvement are often attributed to concurrent shifts in floristic community composition. The bacterial and fungal communities of unimproved and semi-improved (as determined by floristic classification) grassland soils were studied at five upland sites on similar geological substrata using both broad-scale (microbial activity and fungal biomass) and molecular [terminal restriction fragment length polymorphism (TRFLP), automated ribosomal intergenic spacer analysis (ARISA)] approaches. It was hypothesized that microbial community structure would be similar in soils from the same grassland type, and that grassland vegetation classifications could thus be used as predictors of microbial community structure. Microbial community measurements varied widely according to both site and grassland type, and trends in the effect of grassland improvement differed between sites. These results were consistent with those from similar studies, and indicated that floristic community composition was not a stable predictor of microbial community structure across sites. This may indicate a lack of correlation between grassland plant composition and soil microbial community structure, or that differences in soil chemistry between sites had larger impacts on soil microbial populations than plant-related effects.     

Preliminary assessment of plant community structure and arbuscular mycorrhizas in rangeland habitats of Cholistan desert, Pakistan

Plant species dominance and arbuscular mycorrhizas (AMs) were evaluated in four different habitats of Cholistan desert, Pakistan. The sites were selected on the basis of variations in topography, floristic composition and biotic interference. The phytosociological data revealed obvious differences in the floristic composition, plant species diversity, AM colonization and arbuscular mycorrhizal fungi (AMF) spore densities in the rhizospheres. The vegetation pattern and composition was governed by topography and edaphic characteristics of a particular site. Grasses were dominant and main components of vegetation at all the study sites and were invariably mycorrhizal. Some plant species were mycorrhizal at one site but non-mycorrhizal at the other. The AMF survey reported here offers an important starting point from which to analyse AMF community structure in different phytosociological habitats and land uses of Cholistan desert.     

The Role of Urban Structures in the Distribution of Wasteland Flora in the Greater Paris Area,France

Wastelands are likely to host a significant part of urban floristic diversity but have received limited attention because they are not considered interesting green zones. Here, we explore the potential role of wastelands in maintaining urban biodiversity to help define effective urban management plans. We quantified floristic diversity in 98 wasteland sites of Hauts-de-Seine, one of the most densely populated areas in France, and characterized the environmental parameters and spatial distribution of sites to identify some of the factors that influence plant species composition and to explore the impact of urban environment on the floristic interest of wastelands. Their floristic richness represented 58% of the total richness observed in the whole study area. Site richness depended on site area (the largest sites were the richest) and site age, with a maximum in sites of intermediate age (4–13 years). In the largest sites only (>2,500 m²), the floristic distance among sites was positively correlated with geographic distance, which suggests that migration of species among large sites partly controls local floristic composition. In contrast, the environmental distance among sites was not correlated with floristic distance. Finally, we showed that the presence of collective and individual dwellings within 200 m of a wasteland decreased its floristic rarity, whereas the presence of rivers or ponds increased it. We derive several recommendations to optimize the management of wastelands with respect to conservation of urban biodiversity.     

Allometric relationships of field populations of two clonal species with contrasting life histories, Cladium jamaicense and Typha domingensis

Allometric analysis was used to examine morphological relationships in field populations of two clonal plants, Cladium jamaicense and Typha domingensis, in a Florida Everglades wetland. We found that allometric relationships of individuals sampled from field populations could be adequately derived and applied to analyzing both leaf and ramet growth responses to site differences along a nutrient gradient. Overall, the allometric relationships showed a significant departure from isometry which indicates that the relationships were size-dependent. Leaf-level morphological relationships were significantly different between species and between sites along the nutrient gradient. These differences, however, were not expressed on the ramet-level. Neither species expressed a plastic allocation response to site differences along the nutrient gradient. Biomass allocation between above- and below-ground for both species indicated significant size-dependent relationships with decreasing relative allocation below-ground with increasing size. Models for predicting total plant biomass (above- and below-ground) for both C. jamaicense and T. domingensis were developed based on two non-destructive measurements that are easily obtainable in the field. The models followed the equation log (biomass) = α + β₁ × log (height) + β₂ × log (basal area), where α was species specific while β₁ and β₂ were similar for both species but significantly different according to site along the nutrient gradient. Analysis of this model showed that plant height had a relatively greater influence on biomass than basal area at all sites. This difference was greatest at the un-enriched area where plants tend to be short and thick and the least at the moderately enriched site where the relative influence of both parameters was similar.     

Patterns of diversity and habitat relationships in terrestrial mollusc communities of the Pukeamaru Ecological District, northeastern New Zealand

Aim The New Zealand terrestrial mollusc fauna is among the most speciose in the world, with often remarkably high richness at lowland forest sites. We sought to elucidate general explanations for patterns of richness in terrestrial mollusc communities by analysis of species coexistence and habitat relationships within a New Zealand district fauna. Location Pukeamaru Ecological District, eastern North Island, New Zealand. Methods We sampled molluscs using qualitative methods at twenty-three sites and quantitatively by frame sampling of scrubland-forest floor litter at sixteen of these sites and analysed patterns of species richness and turnover in relation to regional species pools and local habitat attributes. We then tested for nonrandom assemblage of taxa along diversity and habitat gradients. Results Ninety-four indigenous mollusc species were recorded from a district fauna estimated at 102 indigenous species: only two species were endemic. From the presumptive geological history of the district, the low endemism, and Brooks parsimony and indicator species analyses of faunal relationships, the communities were indicated to have resulted by accumulation of colonists from other New Zealand districts since the Miocene. Richness ranged from two or three indigenous species in dune habitats to fifty-nine species in a floristically rich forest. Beta diversity was high and site occupancy per species was low, indicating communities structured by successive replacement of ecological equivalents. Sites differing in vegetation had characteristic species assemblages, indicating a degree of habitat specialization. Canonical correspondence analysis indicated that canopy tree species, canopy height, floristic diversity, altitude, litter mass, and litter pH were important determinants of species assemblage in scrubland and forest. Richness was strongly associated with site floristic diversity and, for litter-dwelling species, the pH of litter substrate. High richness occurred at those sites supporting molluscs in high abundance. Shell-shape distributions were essentially Cainian unimodal, with communities dominated by snail species with subglobose to discoidal shells. Mean and variance of shell size increased with mollusc species richness and floristic diversity at sites, indicating dominance of communities by small-shelled species at early successional or floristically poor sites, and increased richness resulting from addition of larger snails into vacant niches. Shifts in shell form were associated with sympatry in several congeneric taxa. Main conclusions The underdispersion of shell shape, relative to faunas elsewhere in the world, indicates that community structure in New Zealand land snail faunas has been constrained by limited phylogenetic diversity and/or by convergence upon successful adaptations. The remarkably high richness that characterizes these communities indicates special conditions allow coexistence of numerous species. The relationship between floristic diversity at sites and the richness, diversity, and shell-size distributions of the molluscs suggests assemblages structured around niche partitioning among competing species. While there is an element of congruence between vegetation and mollusc pattern, this study indicates that assembly rules will be defined, and spatial pattern predicted, only through a better understanding of the linkage between regional species pool, organism traits, environment, and local community assemblage.     

Spatial distribution and floristic composition of trees and lianas in different forest types of an Amazonian rainforest

A quantitative inventory of trees and lianas was conducted (1) to compare floristic composition, diversity and stem density variation between three different forest types (tierra firme, floodplain and swamp), and (2) to analyse the relationships between floristic similarity and forest structure in two regions ~60 km apart in Yasuní National Park, Amazonian Ecuador. A total of 1,087 species with a diameter at breast height ≥ 2.5 cm were recorded in 25 0.1-ha plots. Tierra firme was the habitat with the highest number of species and stem density for trees and lianas, followed by floodplain and swamp in both regions. Two hypotheses that have been independently proposed to describe plant distribution in tropical rain forests, together explain species spatial distribution in this study. The fact that the 30 most important species per forest type (totalling 119 species) accounted for 48.2% of total individuals supports the oligarchy hypothesis. Likewise, 28 out of these 119 species are reported as restricted to a single forest type, which supports the environmental-determinism hypothesis. In general, both canopy and understorey trees and lianas showed rather similar floristic patterns across different forest types and regions.     

Scorpion diversity in two different habitats in the Arid Chaco,Argentina

Scorpions are one of the most important taxa of predators in terms of density, biomass, and diversity in various areas of the world. In this study, we compared population- and community-level data between a mature and a secondary forest in the Chancaní Reserve (Córdoba, Argentina). Scorpions were collected using pitfall traps (54 nights per site), and their nocturnal activity was observed by means of UV light (26 nights per site) over 7 months. Seven species of scorpions (1964 individuals) were observed in the study area (Bothriuridae and Buthidae). Brachistosternus ferrugineus composed >74 % of all individuals and was numerically dominant in most months. It was the most common species sampled with UV light method in all months (85.73 % in mature and 81.80 % in secondary forest). Timogenes elegans was the most common species sampled with the pitfall traps method in secondary forest (48.58 %). General sex ratio (males:females) for B. ferrugineus was 1:1.24 and for T. elegans was 1:0.53. The Shannon index was not significantly different between sites. Species richness was similar, and the Jaccard index was Cs = 0.86, indicating that both sites share 86 % of the species. Tityus confluens was the only species not shared between sites. Our results indicate that species composition in regenerating forest resembles that of primary forest after c. 15 years, but the relative abundances of these species differ.     

Effects of fire regime on plant species richness and composition differ among forest,woodland and heath vegetation

Question

Do the effects of fire regimes on plant species richness and composition differ among floristically similar vegetation types?

Location

Booderee National Park, south‐eastern Australia.

Methods

We completed floristic surveys of 87 sites in Sydney Coastal dry sclerophyll vegetation, where fire history records have been maintained for over 55 years. We tested for associations between different aspects of the recent fire history and plant species richness and composition, and whether these relationships were consistent among structurally defined forest, woodland and heath vegetation types.

Results

The relationship between fire regime variables and plant species richness and composition differed among vegetation types, despite the three vegetation types having similar species pools. Fire frequency was positively related to species richness in woodland, negatively related to species richness in heath, and unrelated to species richness in forest. These different relationships were explained by differences in the associations between fire history and species traits among vegetation types. The negative relationship between fire frequency and species richness in heath vegetation was underpinned by reduced occurrence of resprouting species at high fire frequency sites (more than four fires in 55 years). However, in forest and woodland vegetation, resprouting species were not negatively associated with fire frequency.

Conclusions

We hypothesize that differing relationships among vegetation types were underpinned by differences in fire behaviour, and/or biotic and abiotic conditions, leading to differences in plant species mortality and post‐fire recovery among vegetation types. Our findings suggest that even when there is a high proportion of shared species between vegetation types, fires can have very different effects on vegetation communities, depending on the structural vegetation type. Both research and management of fire regimes may therefore benefit from considering vegetation types as separate management units.     

Rendiconti Delle Sedute E Brevi Comunicazioni

Abstract

Totally, 26 xerohalophytic species belonging to the associations Cakilo-Xanthietum italici and Echinophoro-Elymetum farcti were registered at the Velika pla?a beach (Montenegro); they were distributed following the spatial succession common in this type of habitat. Three transects of different terrain morphology (positioned perpendicularly to the shoreline) were surveyed to establish if there were any differences in the floristic composition and species distribution due to the configuration discrepancy. Moreover, each transect was divided into zones: The first zone from 0 to 30 m inland, the second zone from 30 m to 60 m inland, and the third zone from 60 m to the end of beach (75–115 m, depending on transect length). The floristic composition was relatively uniform in the different transects. However, the number of individuals per species differed significantly among these transects. An analysis of the qualitative composition of transects/zones showed high differences between zones within transects I and II, whereas species composition of zones within transect III was mostly similar. When analyzing the quantitative composition of species, a positive correlation between the distance from the shoreline and the number of individuals per species in transects I and II was registered for most analyzed species (Spearman correlation, p < 0.001). This indicates that terrain configuration affected the ability of present species to become established and propagate.     

Army ants in four forests: geographic variation in raid rates and species composition

1. The New World army ants are top predators in the litter of tropical forest, but no comprehensive studies exist on variation in assemblage-wide activity and species composition. We used standardized protocols to estimate foraging raid rates and species composition of army ant communities in four Neotropical forests. The study sites spanned approximately 10 degrees latitude, with two sites each in Central and South America. 2. We recorded a total of 22 species of army ants. The four sites varied in observed and estimated species richness. Species overlap was highest between the Central American sites, and lowest between the South American sites. 3. Raid activity varied significantly among sites. Raid activity per kilometre of trail walks was over four times higher at the most active site (Sta. Maria, Venezuela) than at the least active site (Barro Colorado Island, Panama). Furthermore, each site showed a different diel pattern of activity. For example, raid activity was higher during daylight hours in Costa Rica, and higher at night in Venezuela. Raid activity relationships with ambient temperature also varied significantly among sites. 4. The overall rate of army ant raids passing through 1 m(2) plots was 0.73 raids per day, but varied among sites, from 0 raids per day (Panama) to 1.2 raids per day (Venezuela). 5. Primarily subterranean species were significantly more abundant in Venezuela, and above-ground foragers that form large swarm fronts were least abundant in Panama. The site heterogeneity in species abundance and diel activity patterns has implications for army ant symbionts, including ant-following birds, and for the animals hunted by these top predators.