Disappearance and Return of Plethodontid Salamanders to Clearcut Plots in the Southern Blue Ridge Mountains

The disappearance and return of plethodontid salamanders on clearcuts has been monitored since 1979 in the southern Blue Ridge Mountains at three sites near Highlands, North Carolina. Salamander abundance on 225 m² plots located on clearcuts and in nearby forest was determined by nightly non-destructive searches. Abundances on clearcut and forest plots at a given site were compared for each year in which sampling occurred. Numbers of salamanders on clearcut plots decreased to 30–50% of forested plots in the first year after logging and were almost zero by the second year. Decreases in standing crop and moisture content of leaf litter seem responsible for salamander disappearance. Salamanders returned to clearcuts 4–6 years after cutting, and their numbers increased rapidly. Linear regressions estimate that salamander numbers on clearcut plots will equal or exceed numbers on forested plots by 20–24 years after cutting. The pattern of salamander return to clearcuts appears closely correlated with the timing of litter layer reformation. All sex and age classes of the most common species, Plethodon jordani, disappear from clearcuts at equal rates, whereas the earliest colonizers are predominantly large adults. Plethodon oconaluftee, a desiccation-resistant species, exists on regenerating clearcuts in disproportionately large numbers. Large adults of all species, including Plethodon oconaluftee, may be better able to withstand the drier, sparse litter cover of young, regenerating stands. Adults might move to clearcuts to avoid competition from smaller and immature salamanders restricted to mature forests with abundant, moist litter.     

Biotic Impoverishment and Homogenization in Unfragmented Forest Understory Communities

Abstract:  Ecological change is often hard to document because of a lack of reliable baseline data. Several recent then-versus-now surveys of temperate forest and grassland communities demonstrate losses of local plant species, but most are based on data from a single site. We resurveyed understory communities in 62 upland forest stands in northern Wisconsin (U.S.A.) for which quantitative baseline data exist from 50 years ago. These stands are within a largely unfragmented region but vary in species composition and successional stage. We collected data on changes in (1) total and native species richness, (2) the ratio of exotic to native species, (3) the relative abundance of habitat generalists, and (4) community similarity among sites. We also compared how these rates of change varied over time. Over the past 50 years, native species density declined an average of 18.5% at the 20-m² scale, whereas the ratio of exotic species to native species increased at 80% of all sites. Habitat generalists increased, and habitat specialists declined, accounting in part for an 8.7% rise in average similarity in species composition among sites. Most of these changes cannot be related to succession, habitat loss, or invasion by exotic species. Areas without deer hunting showed the greatest declines in native species density, with parks and research natural areas faring no better than unprotected stands. Animal-pollinated and animal-dispersed species also declined, particularly at unhunted sites. These results demonstrate the power of quantitative multistand data for assessing ecological change and identify overabundant deer as a key driver of community change. Because maintaining forest habitats alone fails to preserve plant diversity at local scales, local biotic simplification seems likely to continue in the region unless active efforts are taken to protect diversity.     

Relation of Terrestrial-Breeding Amphibian Abundance to Tree-Stand Age

There is a lack of research on the effects of logging on Canadian amphibians. We compared the abundance of terrestrial salamanders in old-growth forests with that in young and mature post-harvest stands. We also measured habitat features of amphibians and contrasted these with old-growth and harvested stands to assess the effects of forest harvest. Quadrat searches demonstrated that clearcut harvesting reduces terrestrial amphibian populations by up to 70% in coastal old-growth forests. We suggest that this reduction results from a decrease in availability of moist microhabitats. Salamander densities within 10 meters of streams in managed stands were similar to those near and away from streams in old growth. We recommend that forest managers recognize the needs of terrestrial amphibians and help maintain amphibian populations by preserving cool, moist habitats. This can be accomplished within cutblocks by (1) maintaining an even distribution of logs and snags as stable, moist microhabitats; (2) retaining some understory as sources of shade, and (3) preserving streamside buffers. Managers must also ensure some level of landscape connectivity to enable climate-sensitive amphibians to disperse and recolonize marginal habitats.     

The Value of Primary, Secondary, and Plantation Forests for a Neotropical Herpetofauna

Abstract:  Plantation forests and second-growth forests are becoming dominant components of many tropical forest landscapes. Yet there is little information available concerning the consequences of different forestry options for biodiversity conservation in the tropics. We sampled the leaf-litter herpetofauna of primary, secondary, and Eucalyptus plantation forests in the Jari River area of northeastern Brazilian Amazonia. We used four complementary sampling techniques, combined samples from 2 consecutive years, and collected 1739 leaf-litter amphibians (23 species) and 1937 lizards (30 species). We analyzed the data for differences among forest types regarding patterns of alpha and beta diversity, species-abundance distributions, and community structure. Primary rainforest harbored significantly more species, but supported a similar abundance of amphibians and lizards compared with adjacent areas of second-growth forest or plantations. Plantation forests were dominated by wide-ranging habitat generalists. Secondary forest faunas contained a number of species characteristic of primary forest habitat. Amphibian communities in secondary forests and Eucalyptus plantations formed a nested subset of primary forest species, whereas the species composition of the lizard community in plantations was distinct, and was dominated by open-area species. Although plantation forests are relatively impoverished, naturally regenerating forests can help mitigate some negative effects of deforestation for herpetofauna. Nevertheless, secondary forest does not provide a substitute for primary forest, and in the absence of further evidence from older successional stands, we caution against the optimistic claim that natural forest regeneration in abandoned lands will provide refuge for the many species that are currently threatened by deforestation .     

Use of Multispecies Occupancy Models to Evaluate the Response of Bird Communities to Forest Degradation Associated with Logging

Forest degradation is arguably the greatest threat to biodiversity, ecosystem services, and rural livelihoods. Therefore, increasing understanding of how organisms respond to degradation is essential for management and conservation planning. We were motivated by the need for rapid and practical analytical tools to assess the influence of management and degradation on biodiversity and system state in areas subject to rapid environmental change. We compared bird community composition and size in managed (ejido, i.e., communally owned lands) and unmanaged (national park) forests in the Sierra Tarahumara region, Mexico, using multispecies occupancy models and data from a 2‐year breeding bird survey. Unmanaged sites had on average higher species occupancy and richness than managed sites. Most species were present in low numbers as indicated by lower values of detection and occupancy associated with logging‐induced degradation. Less than 10% of species had occupancy probabilities >0.5, and degradation had no positive effects on occupancy. The estimated metacommunity size of 125 exceeded previous estimates for the region, and sites with mature trees and uneven‐aged forest stand characteristics contained the highest species richness. Higher estimation uncertainty and decreases in richness and occupancy for all species, including habitat generalists, were associated with degraded young, even‐aged stands. Our findings show that multispecies occupancy methods provide tractable measures of biodiversity and system state and valuable decision support for landholders and managers. These techniques can be used to rapidly address gaps in biodiversity information, threats to biodiversity, and vulnerabilities of species of interest on a landscape level, even in degraded or fast‐changing environments. Moreover, such tools may be particularly relevant in the assessment of species richness and distribution in a wide array of habitats. Uso de Modelos de Ocupación para Múltiples Especies para Evaluar la Respuesta de las Comunidades de Aves a la Degradación de Bosques Asociada con la Tala     

Effects of Fragmentation of Araucarian Vine Forest on Small Mammal Communities

Abstract: Habitat loss and fragmentation are major threats to the survival of forest-dependent fauna. We examined the abundance of small mammal species in forests, corridors, remnants of araucarian vine forest, and Araucaria cunninghamii plantations and pastures. None of the forest mammal species persisted following conversion of forest to pasture. Plantations supported lowered abundances of a subset of forest species that were mainly habitat generalists with respect to their occurrence in different floristic types of undisturbed native forest. Within plantations, an increased subcanopy cover was associated with a more forest-like small mammal assemblage. Species' responses to habitat fragmentation varied. The floristic habitat generalists were largely tolerant of habitat fragmentation, their abundance being similar in forests, corridors, and remnants, and were capable of persisting in remnants a few hectares in area. Floristic habitat specialists were vulnerable to habitat fragmentation and thus were abundant in continuous forest, were less abundant in corridors, and were generally absent from remnants. Species that avoid the corridor matrix and are therefore constrained to the corridor may be disadvantaged by the linearity of the habitat, consistent with the predictions of central-place foraging theory. Although small remnants and corridors provide habitat for some species, those that are more specialized in their use of undisturbed habitat types require the retention or reestablishment of large intact areas.     

Effects of Timber Harvesting on Southern Appalachian Salamanders

We compared the species richness and abundance of salamanders on six recent clearcuts (< 10 years old) with that of salamanders on 34 mature forest stands (>50 years old) in southern Appalachian forests in western North Carolina, U.S.A. Catches of salamanders from plots in mature forest stands were about five times higher than those on recent clearcuts. Almost all species and major taxonomic groups of salamanders were adversely affected by timber removal. Mean number of species collected per plot was about twice as great in mature forest stands as in clearcuts. Analyses of stand age versus salamander catch for 47 plots indicate that 50–70 years are required for populations to return to predisturbance levels following cutting. We conservatively estimate that clearcutting in U.S. national forests in western North Carolina results in a loss of nearly 14 millton salamanders annually. It also is chronically reducing regional populations by more than a quarter of a billion salamanders (9%) below that which could be sustained if mature forests were not cut.     

Effects of Selective Logging on Bat Communities in the Southeastern Amazon

Abstract:  Although extensive areas of tropical forest are selectively logged each year, the responses of bat communities to this form of disturbance have rarely been examined. Our objectives were to (1) compare bat abundance, species composition, and feeding guild structure between unlogged and low-intensity selectively logged (1–4 logged stems/ha) sampling grids in the southeastern Amazon and (2) examine correlations between logging-induced changes in bat communities and forest structure. We captured bats in understory and canopy mist nets set in five 1-ha study grids in both logged and unlogged forest. We captured 996 individuals, representing 5 families, 32 genera, and 49 species. Abundances of nectarivorous and frugivorous taxa (Glossophaginae, Lonchophyllinae, Stenodermatinae, and Carolliinae) were higher at logged sites, where canopy openness and understory foliage density were greatest. In contrast, insectivorous and omnivorous species (Emballonuridae, Mormoopidae, Phyllostominae, and Vespertilionidae) were more abundant in unlogged sites, where canopy foliage density and variability in the understory stratum were greatest. Multivariate analyses indicated that understory bat species composition differed strongly between logged and unlogged sites but provided little evidence of logging effects for the canopy fauna. Different responses among feeding guilds and taxonomic groups appeared to be related to foraging and echolocation strategies and to changes in canopy cover and understory foliage densities. Our results suggest that even low-intensity logging modifies habitat structure, leading to changes in bat species composition.     

Forests Too Deer: Edge Effects in Northern Wisconsin

Abstract: Browsing by white-tailed deer (Odocoileus virginianus) can profoundly affect the abundance and population structure of several woody and herbaceous plant species. Enclosure studies and population surveys reveal that past and current deer densities as low as 4 deer/km² may prevent regeneration of the once common woody species, Canada yew (Taxus canadensis), eastern hemlock (Tsuja canadensis), and white cedar Puja occidentalis), as well as several herbaceous species. Prior to European settlement, forests in northern Wisconsin contained relatively sparse deer populations (<4/km²), but extensive timber cutting in the late nineteenth century boosted deer populations. Continued habitat fragmentation resulting from scattered timber harvests and the creation of "wildlife openings" to improve deer forage maintain these high densities throughout much of the Northeast.
Because deer wander widely, the effects of high deer densities penetrate deeply into remaining stands of old and mature forest, greatly modifying their composition Thus, abundant early successional and "edge" habitat, and the high deer densities they engender, represent significant external threats to these plant communities. We hypothesize that establishing large (200–400 km²) continuous areas of maturing forest, especially in conjunction with increased hunting, could reduce local deer densities and so provide a simple and inexpensive method for retaining species sensitive to the deleterious effects of browsing.     

Wildfire, Fuel Reduction, and Herpetofaunas across Diverse Landscape Mosaics in Northwestern Forests

Abstract:  The herpetofauna (amphibians and reptiles) of northwestern forests (U.S.A.) is diverse, and many species are locally abundant. Most forest amphibians west of the Cascade Mountain crest are associated with cool, cascading streams or coarse woody material on the forest floor, which are characteristics of mature forests. Extensive loss and fragmentation of habitat resulted from logging across approximately 50% of old-growth forests in northern California and approximately 80% of stands in Oregon and Washington. There is a complex landscape mosaic and overlap of northern and southern biotic elements in the Klamath-Siskiyou Region along the Oregon and California border, creating a biodiversity hotspot. The region experiences many low-severity fires annually, punctuated by periodic major fires, including the Biscuit fire, the largest in North America in 2002. In the fire's northern portion, severe fire occurred on >50% of stands of young, managed trees but on only about 25–33% of old-growth stands. This suggests that the legacy of timber harvest may produce fire-prone stands. Calls for prescribed fire and thinning to reduce fuel loads will remove large amounts of coarse woody material from forests, which reduces cover for amphibians and alters nutrient inputs to streams. Our preliminary evidence suggests no negative effects of wildfire on terrestrial amphibians, but stream amphibians decrease following wildfire. Most reptiles are adapted to open terrain, so fire usually improves their habitat. Today, the challenge is to maintain biodiversity in western forests in the face of intense political pressures designed to "prevent" catastrophic fires. We need a dedicated research effort to understanding how fire affects biota and to proactively investigate outcomes of fuel-reduction management on wildlife in western forests.     

Logging Concessions Can Extend the Conservation Estate for Central African Tropical Forests

Abstract:  The management of tropical forest in timber concessions has been proposed as a solution to prevent further biodiversity loss. The effectiveness of this strategy will likely depend on species-specific, population-level responses to logging. We conducted a survey (749 line transects over 3450 km) in logging concessions (1.2 million ha) in the northern Republic of Congo to examine the impact of logging on large mammal populations, including endangered species such as the elephant ( Loxodonta africana ), gorilla ( Gorilla gorilla ), chimpanzee ( Pan troglodytes ), and bongo ( Tragelaphus eurycerus ). When we estimated species abundance without consideration of transect characteristics, species abundances in logged and unlogged forests were not different for most species. When we modeled the data with a hurdle model approach, however, analyzing species presence and conditional abundance separately with generalized additive models and then combining them to calculate the mean species abundance, species abundance varied strongly depending on transect characteristics. The mean species abundance was often related to the distance to unlogged forest, which suggests that intact forest serves as source habitat for several species. The mean species abundance responded nonlinearly to logging history, changing over 30 years as the forest recovered from logging. Finally the distance away from roads, natural forest clearings, and villages also determined the abundance of mammals. Our results suggest that logged forest can extend the conservation estate for many of Central Africa's most threatened species if managed appropriately. In addition to limiting hunting, logging concessions must be large, contain patches of unlogged forest, and include forest with different logging histories.     

Ontogeny of provisioning in two populations of wild golden lion tamarins (Leontopithecus rosalia)

Provisioning weaned young is an important part of the cooperative rearing system of marmosets and tamarins (family: Callitrichidae). Juvenile callitrichids receive a substantial proportion of their diet from all adult group members, whereas juveniles of most other primate species only receive food from the mother infrequently via scrounging. We conducted a longitudinal study of provisioning to 21 young wild golden lion tamarins through their first year of life in two Brazilian reserves. Juveniles were predicted to experience higher provisioning rates at the site with mature forest (the presumed ancestral habitat) and in territories that contained a higher proportion of preferred habitat within the mature forest reserve. Key measures of provisioning did not differ substantially between habitat types. The rates at which juveniles begged for food were very similar across habitats, while the influence of habitat on the rates at which juveniles received food was small and varied across age groups. The most pronounced differences between the reserves were: adult resistance to food-transfer attempts was less frequent, food-offering calls before food transfer were more frequent, and the proportion of prey provisioned was higher in the reserve in which earlier successional-stage forest predominated. Within the mature forest reserve, begging success of young juveniles covaried positively with preferred habitat. Rather than overt caretaker–offspring conflict, an information–feedback loop between juveniles and adults appears to drive provisioning changes as juveniles mature. Direct measures of resource abundance are needed to help clarify the mechanisms by which forest successional stage influences provisioning.     

Life History as a Predictor of Salamander Recovery Rate from Timber Harvest in Southern Appalachian Forests,U.S.A

Forest management often represents a balance between social, economic, and ecological objectives. In the eastern United States, numerous studies have established that terrestrial salamander populations initially decline in abundance following timber harvest, yet the large‐scale and long‐term consequences are relatively unknown. We used count data from terrestrial survey points to examine the relation between salamander abundance and historic timber harvest while accounting for imperfect detection of individuals. Overall, stream‐ and terrestrial‐breeding salamanders appeared to differ by magnitude of population decline, rate of population recovery, and extent of recolonization from surrounding forest. Specifically, estimated abundance of both species groups was positively associated with stand age and recovery rates were predicted to increase over time for red‐legged salamanders (Plethodon shermani) and decrease in stream‐breeding species. Abundance of stream‐breeding salamanders was predicted to reach a peak by 100 years after timber harvest, and the population growth rate of red‐legged salamanders was predicted to undergo a significant increase 100 years after harvest. Estimated abundance of stream‐breeding salamanders in young forest stands was also negatively associated with the distance to adjacent forest, a result that suggests immigration has a role in the recovery of these species. Our results indicate that salamander abundance in young forest stands may be only modestly lower than in more mature forest but that full recovery from timber harvest may take a substantial amount of time and that species life history may affect patterns of recovery. Historia de Vida como un Vaticinador de la Tasa de Recuperación de una Salamandra a la Colecta de Madera en los Bosques del Sur de los Apalaches, E.U.A     

Plethodontid Salamander Response to Silvicultural Practices in Missouri Ozark Forests

Abstract: There is little information on the effects of tree harvest on salamander populations in the midwestern United States. We present data on plethodontid salamander densities in replicated stands of three forest age classes in the southeastern Ozarks of Missouri. Forest age classes consisted of regeneration-cut sites <5 years old, second-growth sites 70–80 years old, and old-growth sites> 120 years old. Salamander abundance on 21, 144-m² plots was determined by area- and time-constrained searches. We also compared age-class habitat characteristics, including downed woody debris, canopy cover, ground area cover, herbaceous vegetation, and woody vegetation. Salamander density was lowest in newly regenerated forests and highest in forests> 120 years old. Comparisons of recently regenerated forests with mature forests> 70 years old indicated that terrestrial salamanders were reduced to very low numbers when mature forests had been intensively harvested. This reduction may result from a decrease in microhabitat availability. Forest age-class comparisons further indicated that salamander abundance slowly increased over time after forests had regenerated. Management decisions that take into account plethodontid salamander abundance and their response to forest structural diversity are important components in sustaining ecosystem integrity while maximizing economic yield.     

A novel statistical method for classifying habitat generalists and specialists

We develop a novel statistical approach for classifying generalists and specialists in two distinct habitats. Using a multinomial model based on estimated species relative abundance in two habitats, our method minimizes bias due to differences in sampling intensities between two habitat types as well as bias due to insufficient sampling within each habitat. The method permits a robust statistical classification of habitat specialists and generalists, without excluding rare species a priori. Based on a user-defined specialization threshold, the model classifies species into one of four groups: (1) generalist; (2) habitat A specialist; (3) habitat B specialist; and (4) too rare to classify with confidence. We illustrate our multinomial classification method using two contrasting data sets: (1) bird abundance in woodland and heath habitats in southeastern Australia and (2) tree abundance in second-growth (SG) and old-growth (OG) rain forests in the Caribbean lowlands of northeastern Costa Rica. We evaluate the multinomial model in detail for the tree data set. Our results for birds were highly concordant with a previous nonstatistical classification, but our method classified a higher fraction (57.7%) of bird species with statistical confidence. Based on a conservative specialization threshold and adjustment for multiple comparisons, 64.4% of tree species in the full sample were too rare to classify with confidence. Among the species classified, OG specialists constituted the largest class (40.6%), followed by generalist tree species (36.7%) and SG specialists (22.7%). The multinomial model was more sensitive than indicator value analysis or abundance-based phi coefficient indices in detecting habitat specialists and also detects generalists statistically. Classification of specialists and generalists based on rarefied subsamples was highly consistent with classification based on the full sample, even for sampling percentages as low as 20%. Major advantages of the new method are (1) its ability to distinguish habitat generalists (species with no significant habitat affinity) from species that are simply too rare to classify and (2) applicability to a single representative sample or a single pooled set of representative samples from each of two habitat types. The method as currently developed can be applied to no more than two habitats at a time.     

Habitat-mediated variation in predation risk by the American marten

The probability of prey encounter, attack, capture, and kill are often hypothesized to depend on habitat structure, but field evidence in terrestrial systems is rare. We tested whether predation efficiency by the American marten (Martes americana) and fear of predation by their primary prey, the red-backed vole (Clethrionomys gapperi), differed between 20- to 50-year-old regenerating forest stands and older uncut stands. Our results showed that the frequency of prey encounter, prey attack, and prey kill were higher in old uncut forests, despite the fact that small-mammal density was similar to that in younger logged forests. These differences in predation efficiency were linked to higher abundance of coarse woody debris, which seems to offer sensory cues to martens, thereby increasing the odds of hunting success. Red-backed voles in regenerating forest stands exhibited increased wariness compared to voles living in old uncut forest, suggestive of a behavioral response to habitat-mediated variation in predation risk.     

Spatial occurrence of a habitat-tracking saproxylic beetle inhabiting a managed forest landscape.

Because of the dynamic nature of many managed habitats, proper evaluation of conservation efforts calls for models that take into account both spatial and temporal habitat dynamics. We develop a metapopulation model for successional-type systems, in which habitat quality changes over time in a predictable fashion. The occupancy and recruitment of the predatory saproxylic (dependent on dead wood) beetle Harminius undulatus was studied in a managed boreal forest landscape, covering 24,449 ha, in central Sweden. In a first step, we analyzed the beetle's occupancy pattern in relation to stand characteristics, and the amounts of present and past habitat in the surrounding landscape. Managed forest is suitable habitat when > or =60 years old, and immediately after cutting, but not between the ages of 10 and 60 years. The observed occupancy of H. undulatus was positively correlated with the stand's age as habitat. We used a metapopulation model to predict the current probability of occurrence in each forest stand, given the spatiotemporal distribution of suitable forest stands during the last 50 years. Metapopulation parameters were estimated by matching predicted spatial distributions with observed spatial distributions. The model predicted observed spatial distributions better than a similar model that assumed constant habitat quality of each forest stand. Thus, metapopulation models for successional-type systems, such as dead wood dependent organisms in managed forest landscapes, should include habitat dynamics. An estimated 82% of the landscape-wide recruitment took place in managed stands, which covered 87% of the forest area, in comparison with 18% in unmanaged stands, which covered 13% of the forest area. Among the managed stand types, > or =60-year-old stands and 3-7-year-old clear-cuttings contributed to 79% of the total recruitment while 8-59-year-old stands only contributed 3%. The results suggest the following guidelines to improve conditions for H. undulatus and other species with similar habitat requirements: (1) the proportion of the landscape constituted by younger stands should not be allowed to grow too large, (2) the rotation period of managed stands should not be allowed to be too short, and (3) dead wood should be retained and created at final cutting.     

Managing Boreal Forest Landscapes for Flying Squirrels

Abstract: Flying squirrel (Pteromys volans) populations have declined severely during the past few decades, and the species has become a focal species in forest management and the conservation debate in Finland. We compared landscape structure around known flying squirrel home ranges with randomly chosen forest sites to determine which landscape patterns characterize the areas occupied by the species in northern Finland. We sought to identify the key characteristics of the landscape that support the remaining flying squirrel populations. We analyzed landscape structure within circular areas with 1- and 3-km radii around 63 forest sites occupied by flying squirrels, and around 96 random sites. We applied stepwise analysis of the landscape structure where landscapes were built up step-by-step by adding patch types in order of their suitability for the flying squirrel. The land-use and forest-resource data for the analysis were derived from multisource national forest inventory and imported to a geographical information system. Landscape patch types were divided into three suitability categories: breeding habitat (mixed spruce-deciduous forests); dispersal habitat ( pine and young forests); and unsuitable habitat ( young sapling stands, open habitats, water). Flying squirrel landscapes contained more suitable breeding habitat patches and were better connected by dispersal habitats than random landscapes. Our results suggest that for the persistence of the flying squirrel, forest managers should 1) maintain a deciduous mixture, particularly in spruce-dominated forests; 2) maintain physical connectivity between optimal breeding habitats; and 3) impose coarse-grained structures on northeastern Finnish landscapes at current levels of habitat availability.     

Understanding Trait‐Dependent Community Disassembly: Dung Beetles,Density Functions,and Forest Fragmentation

Abstract: Anthropogenic disturbances such as fragmentation are rapidly altering biodiversity, yet a lack of attention to species traits and abundance patterns has made the results of most studies difficult to generalize. We determined traits of extinction‐prone species and present a novel strategy for classifying species according to their population‐level response to a gradient of disturbance intensity. We examined the effects of forest fragmentation on dung beetle communities in an archipelago of 33 islands recently created by flooding in Venezuela. Species richness, density, and biomass all declined sharply with decreasing island area and increasing island isolation. Species richness was highly nested, indicating that local extinctions occurred nonrandomly. The most sensitive dung beetle species appeared to require at least 85 ha of forest, more than many large vertebrates. Extinction‐prone species were either large‐bodied, forest specialists, or uncommon. These explanatory variables were unrelated, suggesting at least 3 underlying causes of extirpation. Large species showed high wing loading (body mass/wing area) and a distinct flight strategy that may increase their area requirements. Although forest specificity made most species sensitive to fragmentation, a few persistent habitat generalists dispersed across the matrix. Density functions classified species into 4 response groups on the basis of their change in density with decreasing species richness. Sensitive and persistent species both declined with increasing fragmentation intensity, but persistent species occurred on more islands, which may be due to their higher baseline densities. Compensatory species increased in abundance following the initial loss of sensitive species, but rapidly declined with increasing fragmentation. Supertramp species (widespread habitat generalists) may be poor competitors but strong dispersers; their abundance peaked following the decline of the other 3 groups. Nevertheless, even the least sensitive species were extirpated or rare on the smallest and most isolated islands.     

Sustainable Use of the Tropical Rain Forest: Evidence from the Avifauna in a Shifting-Cultivation Habitat Mosaic in the Colombian Amazon

We carried out a one-year study of the avifauna along a regeneration gradient created by indigenous shifting cultivation in the Colombian Amazon rain forest. Regenerating one-hectare crop fields, from 1 to 17 years after abandonment, and undisturbed understory sites were compared in a sample effort of 4323 net hours. We made 878 captures of 103 species belonging to 11 trophic groups. Differences in species richness and number of individuals captured between regenerating areas and primary forest were low. A comparison of similarity indices among the sites for each habitat type showed no consistent differences, suggesting similar levels of local heterogeneity. Similarity indices calculated between sites of different-aged habitats, however, showed that young secondary growth at less than 10 years of abandonment was least similar to other sites, whereas old secondary growth between 13 to 17 years of abandonment, and forest understories, had the most similar avifauna. Two trophic groups were absent from the samples of the forest interior, and the abundance of the other nine varied among sites of regeneration. We conclude that slash and burn agriculture represents a high-intensity, small-magnitude disturbance to the forest avifauna. Large changes occur when the crop field is established and last until about 10 years after abandonment. Our data suggest that abandoned crop fields older than this age mimic the natural patch dynamics of the forest, making this type of agriculture an example of a potentially sustainable use of tropical forests, at least from the standpoint of its impact on bird communities.