Are landscape attributes a useful shortcut for classifying vegetation in the tropics? A case study of La Amistad International Park

Effective vegetation classification schemes identify the processes determining species assemblages and support the management of protected areas. They can also provide a framework for ecological research. In the tropics, elevation‐based classifications dominate over alternatives such as river catchments. Given the existence of floristic data for many localities, we ask how useful floristic data are for developing classification schemes in species‐rich tropical landscapes and whether floristic data provide support for classification by river catchment. We analyzed the distribution of vascular plant species within 141 plots across an elevation gradient of 130 to 3200 m asl within La Amistad National Park. We tested the hypothesis that river catchment, combined with elevation, explains much of the variation in species composition. We found that annual mean temperature, elevation, and river catchment variables best explained the variation within local species communities. However, only plots in high‐elevation oak forest and Páramo were distinct from those in low‐ and mid‐elevation zones. Beta diversity did not significantly differ in plots grouped by elevation zones, except for low‐elevation forest, although it did differ between river catchments. None of the analyses identified discrete vegetation assemblages within mid‐elevation (700–2600 m asl) plots. Our analysis supports the hypothesis that river catchment can be an alternative means for classifying tropical forest assemblages in conservation settings.     

Scale matters: fire–vegetation feedbacks are needed to explain tropical tree cover at the local scale

At a broad (regional to global) spatial scale, tropical vegetation is controlled by climate; at the local scale, it is believed to be determined by interactions between disturbance, vegetation and local conditions (soil and topography) through feedback processes. It has recently been suggested that strong fire–vegetation feedback processes may not be needed to explain tree‐cover patterns in tropical ecosystems and that climate–fire determinism is an alternative possibility. This conclusion was based on the fact that it is possible to reproduce observed patterns in tropical regions (e.g. a trimodal frequency distribution of tree cover) using a simple model that does not explicitly incorporate fire–vegetation feedback processes. We argue that these two mechanisms (feedbacks versus fire–climate control) operate at different spatial and temporal scales; it is not possible to evaluate the role of a process acting at fine scales (e.g. fire–vegetation feedbacks) using a model designed to reproduce regional‐scale pattern (scale mismatch). While the distributions of forest and savannas are partially determined by climate, many studies are providing evidence that the most parsimonious explanation for their environmental overlaps is the existence of feedback processes. Climate is unlikely to be an alternative to feedback processes; rather, climate and fire–vegetation feedbacks are complementary processes at different spatial and temporal scales.     

The effects of individual tree species on species diversity in a tropical dry forest change throughout ontogeny

Understanding how diversity is maintained in species‐rich communities, such as tropical forests, remains a challenge in ecology. Recent work suggests that the controversy between competing theories could be better resolved by considering the spatial scale at which different processes rule community assembly. Here we use individual species–area relationships (ISAR) to evaluate the spatial organization of tree diversity around individuals of different species in a completely‐mapped tropical dry forest in south Ecuador. We test two hypotheses. First, stressful environmental conditions promote facilitative interactions that will generate spatial signals of accumulation of diversity around individual trees – contrary to what has been reported in humid tropical forests. Second, spatial signals will shift through ontogeny. As, as larger, older trees generate new microsite conditions that affect the recruitment of younger, smaller trees. We compute ISAR functions for adult trees, for young trees and a new crossed‐ISAR function measuring the accumulation of diversity of young trees around the old trees. We compare observed ISARs to the expectations of inhomogeneous Poisson (i.e. null) models controlling for the effects of environmental variation and habitat association on tree distribution. Although the prevalent response among adult trees was not different from null expectations, which means that the organization of diversity in this size class could be explained by environmental heterogeneity alone, most species accumulated more diversity than expected over short spatial scales in the small stem size class. Only two species accumulated significant diversity in the crossed‐ISARs. Our study indicates the role of facilitation in the organization of plant diversity in this dry forest, but that facilitation is limited to some key species acting on early life stages and accumulating diversity around them. Our results demonstrate the benefit of considering different life‐stages and crossed analyses to disentangle the processes affecting community assembly in tropical dry forests.     

Monitoring shifts in plant diversity in response to climate change: a method for landscapes

Improved sampling designs are needed to detect, monitor, and predict plant migrations and plant diversity changes caused by climate change and other human activities. We propose a methodology based on multi-scale vegetation plots established across forest ecotones which provide baseline data on patterns of plant diversity, invasions of exotic plant species, and plant migrations at landscape scales in Rocky Mountain National Park, Colorado, USA. We established forty two 1000-m² plots in relatively homogeneous forest types and the ecotones between them on 14 vegetation transects. We found that 64% of the variance in understory species distributions at landscape scales were described generally by gradients of elevation and under-canopy solar radiation. Superimposed on broad-scale climatic gradients are small-scale gradients characterized by patches of light, pockets of fertile soil, and zones of high soil moisture. Eighteen of the 42 plots contained at least one exotic species; monitoring exotic plant invasions provides a means to assess changes in native plant diversity and plant migrations. Plant species showed weak affinities to overstory vegetation types, with 43% of the plant species found in three or more vegetation types. Replicate transects along several environmental gradients may provide the means to monitor plant diversity and species migrations at landscape scales because: (1) ecotones may play crucial roles in expanding the geophysiological ranges of many plant species; (2) low affinities of understory species to overstory forest types may predispose vegetation types to be resilient to rapid environmental change; and (3) ecotones may help buffer plant species from extirpation and extinction.     

The RAINFOR database: monitoring forest biomass and dynamics

Problem: Data from over 100 permanent sample plots which have been studied for 10–20 years need a suitable system for storage which allows simple data manipulation and retrieval for analysis. Methods: A relational database linking tree records, taxonomic nomenclature and corresponding environmental data has been built in MS Access as part of the RAINFOR project. Conclusion: The database allows flexible and long‐term use of a large amount of data: more than 100 tree plots across Amazonia, incorporating over 80 000 records of individual trees and over 300 000 total records of tree diameter measurements from successive censuses. The database is designed to enable linkages to existing soil, floristic or plant‐trait databases. This database will be a useful tool for exploring the impact of environmental factors on forest structure and dynamics at local to continental scales, and long term changes in forest ecology. As an early example of its potential, we explore the impact of different methodological assumptions on estimates of tropical forest biomass and carbon storage.     

Spatial pattern of diversity in a tropical rain forest in Malaysia

The diversity of trees (species richness, abundance and Shannon diversity) in a tropical rain forest of Malaysia has been studied from the point of view of its spatial organization in order to formulate hypotheses about the origin of the observed spatial patterns. The question that motivated this study is whether tropical forests communities are in a state of equilibrium or non-equilibrium. Three aspects have been examined: (1) changes in diversity were studied with respect to sampling area and sampling designs. A minimum area of 5–10 ha is recommended by the species–area curves, while 2–5 ha seem appropriate based on the Shannon diversity–area curves. Different sampling designs significantly affect the species–area curves. The power function, which can be derived under the equilibrium assumption, is not appropriate to fit the observed diversity–area curves. (2) The spatial features of diversity variables were then studied. Variograms showed that there are dominant short-range effects (around 150 m), obvious anisotropic distribution, and high random variation in the diversity data. (3) Partitioning the variation of the diversity measures into environmental (topographic) and spatial components indicated that the spatial organisation of that community was mostly unpredictable. There may be many processes controlling the formation of the spatial patterns in the tropical rain forest. Unidentified causes, affecting mainly the small-scale processes (<20 m), seem responsible for the large amount of undetermined variation in the diversity data sets. The study suggests that the Pasoh forest of Malaysia may not be in a state of equilibrium.     

Reconciling the contribution of environmental and stochastic structuring of tropical forest diversity through the lens of imaging spectroscopy

Both niche and stochastic dispersal processes structure the extraordinary diversity of tropical plants, but determining their relative contributions has proven challenging. We address this question using airborne imaging spectroscopy to estimate canopy β‐diversity for an extensive region of a Bornean rainforest and challenge these data with models incorporating niches and dispersal. We show that remotely sensed and field‐derived estimates of pairwise dissimilarity in community composition are closely matched, proving the applicability of imaging spectroscopy to provide β‐diversity data for entire landscapes of over 1000 ha containing contrasting forest types. Our model reproduces the empirical data well and shows that the ecological processes maintaining tropical forest diversity are scale dependent. Patterns of β‐diversity are shaped by stochastic dispersal processes acting locally whilst environmental processes act over a wider range of scales.     

Reproductive phenology of coastal plain Atlantic forest vegetation: comparisons from seashore to foothills

The diversity of tropical forest plant phenology has called the attention of researchers for a long time. We continue investigating the factors that drive phenological diversity on a wide scale, but we are unaware of the variation of plant reproductive phenology at a fine spatial scale despite the high spatial variation in species composition and abundance in tropical rainforests. We addressed fine scale variability by investigating the reproductive phenology of three contiguous vegetations across the Atlantic rainforest coastal plain in Southeastern Brazil. We asked whether the vegetations differed in composition and abundance of species, the microenvironmental conditions and the reproductive phenology, and how their phenology is related to regional and local microenvironmental factors. The study was conducted from September 2007 to August 2009 at three contiguous sites: (1) seashore dominated by scrub vegetation, (2) intermediary covered by restinga forest and (3) foothills covered by restinga pre-montane transitional forest. We conducted the microenvironmental, plant and phenological survey within 30 transects of 25 m × 4 m (10 per site). We detected significant differences in floristic, microenvironment and reproductive phenology among the three vegetations. The microenvironment determines the spatial diversity observed in the structure and composition of the flora, which in turn determines the distinctive flowering and fruiting peaks of each vegetation (phenological diversity). There was an exchange of species providing flowers and fruits across the vegetation complex. We conclude that plant reproductive patterns as described in most phenological studies (without concern about the microenvironmental variation) may conceal the fine scale temporal phenological diversity of highly diverse tropical vegetation. This phenological diversity should be taken into account when generating sensor-derived phenologies and when trying to understand tropical vegetation responses to environmental changes.     

Beta diversity and oligarchic dominance in the tropical forests of Southern Costa Rica

Recent studies have reported a consistent pattern of strong dominance of a small subset of tree species in neotropical forests. These species have been called “hyperdominant” at large geographical scales and “oligarchs” at regional‐landscape scales when being abundant and frequent. Forest community assembly is shaped by environmental factors and stochastic processes, but so far the contribution of oligarchic species to the variation of community composition (i.e., beta diversity) remains poorly known. To that end, we established 20.1‐ha plots, that is, five sites with four forest types (ridge, slope and ravine primary forest, and secondary forest) per site, in humid lowland tropical forests of southwestern Costa Rica to (a) investigate how community composition responds to differences in topography, successional stage, and distance among plots for different groups of species (all, oligarch, common and rare/very rare species) and (b) identify oligarch species characterizing changes in community composition among forest types. From a total of 485 species of trees, lianas and palms recorded in this study only 27 species (i.e., 6%) were nominated as oligarch species. Oligarch species accounted for 37% of all recorded individuals and were present in at least half of the plots. Plant community composition significantly differed among forest types, thus contributing to beta diversity at the landscape scale. Oligarch species was the component best explained by geographical and topographic variables, allowing a confident characterization of the beta diversity among tropical lowland forest stands. Abstract in Spanish is available with online material.     

Determinants of lepidopteran community composition and species diversity in eastern deciduous forests: roles of season, eco-region and patch size

Ecologists have long recognized that factors operating at both local and regional scales influence whether a given species occurs in an ecological community. The relative roles of variables manifested at local and regional scales on community structure, however, remain an unexplored issue for many faunas. To address this question, we compared the community composition and species diversity of forest Lepidoptera between (i) large forest tracts in historically glaciated and unglaciated regions of the eastern deciduous forest in North America, and (ii) large and small forest patches within a highly fragmented forest landscape. Specifically, we used seasonally stratified sampling to test whether regional and local differences in moth communities were related to variation in stand structure and floristic composition. At the local scale, we tested three alternative hypotheses describing the effects of patch size on moth species richness: species impoverishment, species replacement, or species supplementation. Cluster analysis revealed significant compositional differences in moth communities sampled between (i) early and late seasons, (ii) glaciated and unglaciated forest eco‐regions, and (iii) large and small forest patches. Canonical correspondence analysis suggested that floristic variation at regional scales had a greater role in determining moth community composition than local vegetation or patch‐size effects. Species richness was higher in the glaciated North Central Tillplain, and was attributable to a more diverse herbaceous feeding moth assemblage. Finally, we found evidence that both species impoverishment and species replacement processes structure the moth fauna of small woodlots; the richness of moths with larvae that feed on woody plants decreased with patch area, but herbaceous feeding species increased in diversity in smaller patches. Thus, our results suggest that local and regional differences in moth community structure are mediated by differences in host‐plant resources attributable to regional biogeographic history and local differences in patch size. Because community composition appeared to be more sensitive to environmental variation than species richness, we suggest that monitoring lepidopteran species diversity in forests will not detect significant changes in species composition due to environmental change.     

Vegetation responses to past climatic variation

Vegetation responses to climatic change can be studied retrospectively by utilizing the Quaternary fossil record. There has been controversy over the extent to which major changes in vegetation patterns at the continental scale lag behind the climatic changes that drive them, and to what extent vegetation can ever be said to be in equilibrium with climate. The equilibrium question has no single answer. The predominant mode of vegetation response to climatic change depends on the space and time frame and resolution of the data set in which the response is observed.Vegetation (as observed on particular space and time scales) can be in dynamic equilibrium with climate if its response time is sufficiently fast in relation to the rate of climatic change to which it is observed to be responding. Several processes can be involved in the response: successional, migrational, edaphic, and evolutionary. Successional response times can be deduced from forest succession models. The other processes are less well understood and different ideas exist concerning their rates. According to one hypothesis, migrational lags caused delays of thousands of years in the postglacial dynamics of forest composition. The alternative hypothesis explains these changes as dynamic equilibrium responses to changes in climatic seasonality and climatic anomaly patterns. Neither hypothesis need be universally true; gradient analysis and forest succession models are among the techniques that can be used in inferential tests of these hypotheses for particular space-time regions.Dynamic equilibrium may often be a reasonable approximation for the responses of the broadest continental-scale forest patterns to orbitally induced climatic changes. But as spatial and temporal frames of observation are diminished and resolution increased, biotic processes must eventually come to dominate. At sufficiently fine scales the main observable phenomena are successional responses to natural disturbance events. The late-Quaternary record of vegetation change allows a choice of observation scales and thus provides a continuum of possibilities for study, ranging from long-term dynamic bioclimatology to more conventional vegetation dynamics.I thank Margaret Davis, Honor Prentice, Jim Ritchie, Al Solomon, Geoff Spaulding and Tom Webb for their reviews of earlier drafts. Research supported by a US Department of Energy, Carbon Dioxide Research Division, grant to Brown University and a Swedish Natural Science Research Council grant to the project SlsSimulation of Natural Forest Dynamics.I thank Margaret Davis, Honor Prentice, Jim Ritchie, Al Solomon, Geoff Spaulding and Tom Webb for their reviews of earlier drafts. Research supported by a US Department of Energy, Carbon Dioxide Research Division, grant to Brown University and a Swedish Natural Science Research Council grant to the project SlsSimulation of Natural Forest Dynamics.     

Sixteen hundred years of increasing tree cover prior to modern deforestation in Southern Amazon and Central Brazilian savannas

Tropical ecosystems are under increasing pressure from land‐use change and deforestation. Changes in tropical forest cover are expected to affect carbon and water cycling with important implications for climatic stability at global scales. A major roadblock for predicting how tropical deforestation affects climate is the lack of baseline conditions (i.e., prior to human disturbance) of forest–savanna dynamics. To address this limitation, we developed a long‐term analysis of forest and savanna distribution across the Amazon–Cerrado transition of central Brazil. We used soil organic carbon isotope ratios as a proxy for changes in woody vegetation cover over time in response to fluctuations in precipitation inferred from speleothem oxygen and strontium stable isotope records. Based on stable isotope signatures and radiocarbon activity of organic matter in soil profiles, we quantified the magnitude and direction of changes in forest and savanna ecosystem cover. Using changes in tree cover measured in 83 different locations for forests and savannas, we developed interpolation maps to assess the coherence of regional changes in vegetation. Our analysis reveals a broad pattern of woody vegetation expansion into savannas and densification within forests and savannas for at least the past ~1,600 years. The rates of vegetation change varied significantly among sampling locations possibly due to variation in local environmental factors that constrain primary productivity. The few instances in which tree cover declined (7.7% of all sampled profiles) were associated with savannas under dry conditions. Our results suggest a regional increase in moisture and expansion of woody vegetation prior to modern deforestation, which could help inform conservation and management efforts for climate change mitigation. We discuss the possible mechanisms driving forest expansion and densification of savannas directly (i.e., increasing precipitation) and indirectly (e.g., decreasing disturbance) and suggest future research directions that have the potential to improve climate and ecosystem models.     

Landscape patterns and environmental controls of deciduousness in forests of central Panama

Aim Dry season deciduousness affects intra‐ and inter‐annual patterns of carbon, water and energy balance in seasonal tropical forests. Because it is affected by rainfall, temperature and solar radiation, deciduousness may be an indicator of the response of vegetation to climate change. Better understanding of how spatial patterns of deciduousness are affected by climate and other environmental gradients will improve the ability to predict responses to climate change. This study develops remote sensing methods for quantifying tropical forest deciduousness and examines the relationship between deciduousness and environmental factors in semi‐deciduous tropical forest. Location Central Panama. Methods I applied spectral mixture analysis (SMA) and the normalized difference vegetation index (NDVI) to Landsat images to predict deciduousness which was ground‐truthed with field observations of the percentage of overstorey deciduous trees. Using predicted deciduousness from SMA, patterns of deciduousness at three spatial scales were analysed. I determined how deciduousness varied spatially with rainfall and geological substrate. Results Both SMA and NDVI had strong correlations (r > 0.9) with field observations of deciduousness. On a landscape scale, deciduousness increased as rainfall decreased, but geological substrate altered this relationship. On some geological substrates, deciduousness was much greater than expected for a given rainfall total or showed a slight but significant increase with rainfall. At an intermediate spatial scale, there were highly deciduous patches from 3 to 250 ha in size embedded in non‐deciduous forest, which may have resulted from topography, soil variation or past land use. Main conclusions Dry season deciduousness can be accurately quantified using satellite images indicating that remote sensing can be a valuable tool for detecting change and understanding ecosystem processes in tropical forests from landscape to regional scales.     

Pathways, mechanisms and predictability of vegetation change during tropical dry forest succession

The development of forest succession theory has been based on studies in temperate and tropical wet forests. As rates and pathways of succession vary with the environment, advances in successional theory and study approaches are challenged by controversies derived from such variation and by the scarcity of studies in other ecosystems. During five years, we studied development pathways and dynamics in a chronosequence spanning from very early to late successional stages (ca. 1–60 years) in a tropical dry forest of Mexico. We (1) contrasted dynamic pathways of change in structure, diversity, and species composition with static, chronosequence-based trends, (2) examined how structure and successional dynamics of guilds of trees shape community change, and (3) assessed the predictability of succession in this system. Forest diversity and structure increased with time but tree density stabilized early in succession. Dynamic pathways matched chronosequence trends. Succession consisted of two tree-dominated phases characterized by the development and dynamics of a pioneer and a mature forest species guild, respectively. Pioneer species dominated early recruitment (until ca. 10 years after abandonment), and declined before slower growing mature-forest species became dominant or reached maximum development rates (after 40–45 years). Pioneers promoted their replacement early in succession, while mature-forest species recruited and grew constantly throughout the process, with their lowest mortality coinciding with the peak of pioneer abundance. In contrast to prevailing stochastic views, we observed an orderly, community driven series of changes in this dry forest secondary succession. Chronosequences thus represent a valuable approach for revealing system-specific successional pathways, formulating hypotheses on causes and mechanisms and, in combination with repeated sampling, evaluating the effects of vegetation dynamics in pathway variation.     

Core and Satellite Species in Degraded Habitats: an Analysis Using Malagasy Tree Communities

Core-satellite theory predicts that, via the “rescue effect”, widespread, abundant species should have reduced risk of local extinctions. We test this hypothesis in southeastern Malagasy littoral forest using data on distribution and abundance of trees and woody understory vegetation in tropical forest fragments along a disturbance gradient. We partition the mortality risk into two kinds of extinction factors, separately operating at demographic (local) and landscape (regional) scales, contrary to core-satellite predictions, for both trees and woody understory vegetation, that the relative number of core (abundant) species declined significantly with increasing disturbance. In the least-degraded forest fragments there was a strong mode of core species, while in the moderately- and severely-degraded fragments the species distributions were essentially log-normal, lacking a substantial core mode. While the rescue effect mitigates one kind of extinction risk, namely local environmental and demographic stochasticity, it may not counterbalance widespread pervasive sources of mortality. The amount of internal forest fragmentation appears to have a much greater effect on species richness and diversity than either fragment size or shape.     

Deforestation and Reforestation of Latin America and the Caribbean (2001–2010)

Forest cover change directly affects biodiversity, the global carbon budget, and ecosystem function. Within Latin American and the Caribbean region (LAC), many studies have documented extensive deforestation, but there are also many local studies reporting forest recovery. These contrasting dynamics have been largely attributed to demographic and socio‐economic change. For example, local population change due to migration can stimulate forest recovery, while the increasing global demand for food can drive agriculture expansion. However, as no analysis has simultaneously evaluated deforestation and reforestation from the municipal to continental scale, we lack a comprehensive assessment of the spatial distribution of these processes. We overcame this limitation by producing wall‐to‐wall, annual maps of change in woody vegetation and other land‐cover classes between 2001 and 2010 for each of the 16,050 municipalities in LAC, and we used nonparametric Random Forest regression analyses to determine which environmental or population variables best explained the variation in woody vegetation change. Woody vegetation change was dominated by deforestation (?541,835 km²), particularly in the moist forest, dry forest, and savannas/shrublands biomes in South America. Extensive areas also recovered woody vegetation (+362,430 km²), particularly in regions too dry or too steep for modern agriculture. Deforestation in moist forests tended to occur in lowland areas with low population density, but woody cover change was not related to municipality‐scale population change. These results emphasize the importance of quantitating deforestation and reforestation at multiple spatial scales and linking these changes with global drivers such as the global demand for food.     

Fragment shape and tree species composition in tropical forests: a landscape level investigation

Fragmentation of tropical forest alters community composition as a result of changes in forest shape. This paper uses 22 hypotheses to test the effect of fragment shape on tree species composition in Ghana, West Africa, within biological categories of regeneration guild, rarity, phenology and dispersal. For both regenerating and mature trees, relationships between species composition and the shape of forest fragments were complex; almost half were significant but many failed to support the established hypotheses. Irregular shaped fragments had high proportions of regenerating, light‐demanding pioneers and mature, animal‐dispersed species. Species common to Ghana formed the foundation of communities in fragments of all shapes. Investigation at the landscape level indicated broad patterns of species change. Rigorous hypothesis testing is needed, following extensive demographic work on the ground, before population dynamics within tropical forest fragments can be comprehended fully and applied to conservation management.     

Site fertility drives temporal turnover of vegetation at high latitudes

Experimental evidence shows that site fertility is a key modulator underlying plant community changes under climate change. Communities on fertile sites, with species having fast dynamics, have been found to react more strongly to climate change than communities on infertile sites with slow dynamics. However, it is still unclear whether this generally applies to high‐latitude plant communities in natural environments at broad spatial scales. We tested a hypothesis that vegetation of fertile sites experiences greater changes over several decades and thus would be more responsive under contemporary climate change compared to infertile sites that are expected to show more resistance. We resurveyed understorey communities (vascular plants, bryophytes, and lichens) of four infertile and four fertile forest sites along a latitudinal bioclimatic gradient. Sites had remained outside direct human disturbance. We analyzed the magnitude of temporal community turnover, changes in the abundances of plant morphological groups and strategy classes, and changes in species diversity. In agreement with our hypothesis, temporal turnover of communities was consistently greater on fertile sites compared to infertile sites. However, our results suggest that the larger turnover of fertile communities is not primarily related to the direct effects of climatic warming. Furthermore, community changes in both fertile and infertile sites showed remarkable variation in terms of shares of plant functional groups and strategy classes and measures of species diversity. This further emphasizes the essential role of baseline environmental conditions and nonclimatic drivers underlying vegetation changes. Our results show that site fertility is a key determinant of the overall rate of high‐latitude vegetation changes but the composition of plant communities in different ecological contexts is variously impacted by nonclimatic drivers over time.     

Large-scale phylogenetic analyses reveal the causes of high tropical amphibian diversity

Many groups show higher species richness in tropical regions but the underlying causes remain unclear. Despite many competing hypotheses to explain latitudinal diversity gradients, only three processes can directly change species richness across regions: speciation, extinction and dispersal. These processes can be addressed most powerfully using large-scale phylogenetic approaches, but most previous studies have focused on small groups and recent time scales, or did not separate speciation and extinction rates. We investigate the origins of high tropical diversity in amphibians, applying new phylogenetic comparative methods to a tree of 2871 species. Our results show that high tropical diversity is explained by higher speciation in the tropics, higher extinction in temperate regions and limited dispersal out of the tropics compared with colonization of the tropics from temperate regions. These patterns are strongly associated with climate-related variables such as temperature, precipitation and ecosystem energy. Results from models of diversity dependence in speciation rate suggest that temperate clades may have lower carrying capacities and may be more saturated (closer to carrying capacity) than tropical clades. Furthermore, we estimate strikingly low tropical extinction rates over geological time scales, in stark contrast to the dramatic losses of diversity occurring in tropical regions presently.