Environment shapes the spatial organization of tree diversity in fragmented forests across a human-modified landscape.

Altered stand structure and tree allometry reduce carbon storage in evergreen forest fragments in India’s Western Ghats

Summary Fragmentation is ubiquitous across tropical forests and drives marked shifts in tree community composition by differentially affecting species' dispersal, establishment and survival. Such compositional shifts can potentially alter ecosystem‐level properties such as above‐ground carbon storage, but our understanding of the factors linking compositional shifts to carbon storage is limited. We compared tree communities of contiguous and fragmented tropical rain forests in the Western Ghats (India) and assessed the ability of various plant functional traits associated with seed dispersal, establishment and survival processes to predict species' responses to fragmentation. Further, we assessed relationships between functional traits that predict tree community turnover and those that govern carbon storage to examine how fragmentation effects on species' composition can alter the ability of tree communities to store carbon. Seed size, as indexed by seed length, was the best predictor of species' responses, with larger‐seeded species declining in fragments. Across species, seed length was positively correlated with maximum attainable height, which decreased by 10% on average at the community level in fragments. Such shifts towards smaller‐seeded communities could decrease forest stature and reduce above‐ground carbon stocks by 8%. Synthesis and applications. Our study highlights a previously undescribed mechanism by which fragmentation‐driven declines of large‐seeded tree species can reduce above‐ground carbon stocks by promoting shorter‐statured forests. These results imply that strict protection alone might be insufficient and that a multipronged conservation strategy would be required to sustain carbon stocks in tropical forest fragments. Such interventions will need to combine restoration programmes for large‐seeded tree species in fragments with broader‐scale efforts to maintain hospitable and well‐connected landscapes for their seed dispersers.

Seed predation and reduced predation risk with distance from conspecific trees are important influences on tree regeneration in tropical forests. Shifts in animal communities, such as an increase in rodents and other small mammals due to forest fragmentation, could alter patterns of seed predation and affect tree regeneration and community dynamics in forest fragments. We performed a field experiment on four native rainforest tree species in the Western Ghats, India, to test whether fragmentation increases seed predation by mammals and alters the distance dependence of seed predation. We monitored seed predation within open and mammal exclosure plots, near and far from the canopies of conspecific trees, in contiguous and fragmented forests. Seed predation of Cullenia exarillata, Ormosia travancorica, and Syzygium rubicundum was markedly higher in forest fragments, and more so within open plots than exclosures, while the predominantly insect‐predated Acronychia pedunculata experienced similar predation in contiguous forests and fragments. Seed predation of C. exarillata and S. rubicundum was unrelated to distance from conspecific trees in open plots in both contiguous forests and fragments, in contrast to exclosures that showed marked near versus far differences in seed predation. Our findings suggest that by increasing overall seed predation risk and imposing similar seed predation risk near and far from adults variably across the tree species, small mammals could alter processes that shape tree diversity and species composition in fragmented tropical rainforests.

Predator–primate interactions are understudied, yet predators have been shown to influence primate behavior, population dynamics, and spatial distribution. An understanding of these interactions is important for the successful management and conservation of these species. Novel approaches are needed to understand better the spatial relationships between predators and primates across changing landscapes. We combined photographic surveys of predators and humans with line-transect sampling of lemurs across contiguous and fragmented forests in Madagascar to 1) compare relative activity; 2) estimate probability of occupancy and detection; 3) estimate predator–primate and local people–primate co-occurrence; and 4) assess variables influencing these parameters across contiguous and fragmented forests. In fragmented (compared to contiguous) forest sites endemic predator and lemur activity were lower whereas introduced predator and local people activity were higher. Our two-species interaction occupancy models revealed a higher number of interactions among species across contiguous forest where predator and lemur occupancy were highest. Mouse lemurs show evidence of “avoidance” (SIF 1.0) with feral cats and local people in contiguous forest. Feral cats demonstrated the highest number of interactions with lemurs, despite their distribution being limited to only contiguous forest. Distance to forest edge and distance to nearby villages were important in predicting predator occupancy and detection. These results highlight the growing threat to endemic predators and lemurs as habitat loss and fragmentation increase throughout Madagascar. We demonstrate the effectiveness of a novel combination of techniques to investigate how predator species impact primate species across a gradient of forest fragmentation.

We evaluated if male age and body size, density of conspecifics, and arthropod biomass contributed to variation in territory size and overlap of Ovenbirds, Seiurus aurocapillus (L., 1766), in a fragmented and contiguous boreal forest. Territory size and overlap were determined by radio-tracking territorial male Ovenbirds in fragmented (n = 22) and contiguous forest (n = 13) from late May to the end of June 1999 and 2000. Variation in male territory size was most strongly associated with individual characteristics, specifically body size and age. Furthermore, we found strong support for an effect of density of conspecifics on territory overlap, suggesting that the exclusivity of territories and perhaps levels of territoriality were greater for males in contiguous forest than for those in fragments. Our findings (i.e., mean territory size was similar between landscapes and territory overlap was greater in fragments than in contiguous forest) suggest that fragments either have larger areas of unsuitable habitat or are less saturated with Ovenbirds. Furthermore, because resources were not distributed equally among individuals, our results were consistent with the ideal dominance model of habitat selection. Finally, previous studies based on acoustical surveys have likely underestimated space-use requirements in forest passerines.

We examined whether forest fragmentation by agriculture influenced the abundance of North American red squirrels (Tamiasciurus hudsonicus) in the southern boreal mixedwood forest of Saskatchewan. During the summers of 1995 and 1997, we determined the relative abundance of red squirrels in 43 forest fragments ranging in size from 0.2 to 82 ha and in 15 sites in nearby contiguous forest. Relative abundance was determined using acoustic point count surveys, in which all squirrel vocalizations were recorded. Within the fragmented agricultural landscape, we found that red squirrel abundance increased slightly with forest-fragment size, whereas the abundance of red squirrels was negatively correlated with the amount of forest cover within 1 km of a site. We also compared squirrel abundance in forest fragments with that in contiguous forest and found that red squirrels were significantly more abundant in forest fragments than in contiguous forest. We speculate that changes in dispersal patterns in fragmented forest habitats, higher squirrel survival in forest fragments, and differences in diet between contiguous and fragmented forest habitats contributed to this pattern. Our results suggest that forest fragmentation may not be particularly detrimental to habitat generalists like the red squirrel. However, an increased abundance of red squirrels in forest fragments may result in negative effects on other species, such as forest songbirds.

We evaluated the effects of forest fragmentation caused by agriculture on arthropod prey biomass and vegetation structure found in territories of male Ovenbirds (Seiurus aurocapillus) breeding in the southern boreal mixed woods of Saskatchewan, Canada. The objective of this study was to determine if previously documented differences in pairing success of male Ovenbirds in contiguous forests and forest fragments in our study area were associated with differences in arthropod prey biomass and vegetation structure between contiguous and fragmented forests. A secondary objective was to examine the correspondence between vegetation and arthropods to evaluate whether vegetation cues could be useful for birds attempting to predict future arthropod biomass during territory selection. Our results indicate that both vegetation structure and arthropod prey composition in Ovenbird territories differed between fragmented and contiguous forests, whereas total arthropod biomass did not. Furthermore, the correspondence of vegetation with arthropod prey composition and total prey biomass was weak, a result that questions the use of vegetation structure by male Ovenbirds for predicting future prey availability during territory selection. Overall, the current extent of forest fragmentation in our study area is not likely reducing pairing success of territorial male Ovenbirds by lowering the biomass of arthropod prey. However, landscape differences in vegetation structure could influence pairing success of male Ovenbirds in forest fragments by reducing suitable microhabitats for nesting or by increasing habitat suitability for nest predators and Brown-headed Cowbird (Molothrus ater) parasites.

-Declines of many forest-dwelling Neotropical migrants have been attributed, in part, to fragmentation of forest habitat on the breeding grounds in North America. During 1990-1991, we determined reproductive success of Wood Thrushes (Hylocichla mustelina) nesting within contiguous forest habitat (>10,000 ha) and in nine forest fragments ranging in size from 9.2 to 126.5 ha in Berks County, Pennsylvania. We located 171 Wood Thrush nests. Nesting success differed significantly among forest size categories, with 86% of the nests successful in contiguous forest, 72% successful in large fragments (>100 ha), and 43% successful in small fragments (<80 ha). The variable that best predicted nest survival was forest area (R2 = 0.86). Rates of predation differed significantly among forest size categories, and predation was the primary cause of nesting failure. We found 56% of the nests in small fragments were lost to predators as compared to 22% in large fragments and 10% within the contiguous forest. Visitation by mammalian predators to scent posts was significantly different between small and large forest sites (41 vs. 14%, respectively), and relative abundance of avian nest predators was significantly higher in small forest fragments than in the large forest sites (x = 1.04 vs. 0.41 birds per census point). Brown-headed Cowbirds (Molothrus ater) parasitized 9% of the nests. Rates of brood parasitism did not differ significantly among forest size categories and had little influence on nesting success. Our results suggest that reproductive success of Wood Thrushes nesting within contiguous forest is high and that severe reproductive dysfunction as a result of high rates of nest predation is an important consequence of forest fragmentation. Received 6 July 1993, accepted 21 November 1993. MANY FOREST-DWELLING NEOTROPICAL migrant songbirds have undergone population declines in eastern deciduous forests of North America (Robbins et al. 1989a, Askins et al. 1990). Most declining forest species are found less commonly than expected in small areas of forest than in large (area sensitive), and some of these songbirds have decreased in abundance or disappeared completely from small forest patches (Ambuel and Temple 1983, Blake and Karr 1984, Askins et al. 1987). Because forest fragmentation results in smaller forest patches or fragments separated from each other by nonforest habitat, these declines have been attributed, in part, to fragmentation of forest habitat in North America (Robbins 1979, Askins et al. 1990). Numerous hypotheses have been proposed to explain why forest-dwelling Neotropical migrants decrease in abundance or disappear from small forest patches (Askins et al. 1990). As a forest becomes fragmented, the amount of edge habitat increases and the amount of interior de3Present address: Illinois Natural History Survey, 607 E. Peabody Drive, Champaign, Illinois, 61820, USA. creases. Mammalian nest predators such as eastern chipmunks (Tamias striatus) and raccoons (Procyon lotor; Bider 1968, Forsyth and Smith 1973) tend to be more abundant along the forest edge than in the forest interior. Whitcomb et al. (1981) suggested that avian nest predators, such as Blue Jays (Cyanocitta cristata) and American Crows (Corvus brachyrhynchos), show a similar pattern. The Brown-headed Cowbird (Molothus ater), a brood parasite, also is more abundant along edges (Brittingham and Temple 1983). In addition, rates of both nest depredation and brood parasitism are higher near forest edges than within the forest interior (Gates and Gysel 1978, Chasko and Gates 1982, Brittingham and Temple 1983, Temple and Cary 1988). Consequently, one hypothesis is that forest birds decline in number in small forest patches as a result of poor reproductive success due to high rates of predation on eggs and nestlings, and to brood parasitism (Brittingham and Temple 1983, Wilcove 1985, Askins et al. 1990). Experiments with artificial nests indicate that nest predation is more frequent in small forest patches than within extensive areas of forest, thus supporting this hypothesis (Wilcove 1985,

Variations in tree community composition and structure in a fragment of tropical semideciduous forest in southeastern Brazil related to different human disturbance histories

In Neotropical rain forests, fresh mammal dung, especially that of howler monkeys, constitutes an important resource used by dung beetles as food and for oviposition and further feeding by their larvae. Tropical rain forest destruction, fragmentation, and subsequent isolation causing reductions in numbers of and the disappearance of howler monkeys may result in decreasing numbers of dung beetles, but this has not been documented. In this study, we present information on the presence of howlers and dung beetles in 38 isolated forest fragments and 15 agricultural habitats. Howler monkeys were censused by visual means, while dung beetles were sampled with traps baited with a mixture of howler, cow, horse, and dog dung. Results indicated that loss of area and isolation of forest fragments result in significant decrements in howlers and dung beetles. However, dung beetle abundance was found to be closely related to the presence of howler monkeys at the sites and habitats investigated. Scenarios of land management designed to reduce isolation among forest fragments may help sustain populations of howler monkeys and dung beetles, which may have positive consequences for rain forest regeneration.

In Neotropical rain forests, fresh mammal dung, especially that of howler monkeys, constitutes an important resource used by dung beetles as food and for oviposition and further feeding by their larvae. Tropical rain forest destruction, fragmentation, and subsequent isolation causing reductions in numbers of and the disappearance of howler moneys may result in decreasing numbers of dung beetles, but this has not been documented. In this study, we present information on the presence of howlers and dung beetles in 38 isolated forest fragments and 15 agricultural habitats. Howler monkeys were censused by visual means, while dung beetles were sampled with traps baited with a mixture of howler, cow, horse, and dog dung. Results indicated that loss of area and isolation of forest fragments result in significant decrements in howlers and dung beetles. However, dung beetle abundance was found to be closely related to the presence of howler monkeys at the sites and habitats investigated. Scenarios of land management designed to reduce isolation among forest fragments may help sustain populations of howler monkeys and dung beetles, which may have positive consequences for rain forest regeneration. Am. J. Primatol. 48:253–262, 1999. © 1999 Wiley-Liss, Inc.

Soil chronosequences provide an opportunity to examine the influence of long-term pedogenesis on the biomass and composition of associated tree communities. We assessed variation in the species composition of trees, saplings, and seedlings, and the basal area of adult trees, in lowland temperate rain forest along the Haast chronosequence on the west coast of the South Island of New Zealand. The sequence consists of Holocene dune ridges formed following periodic earthquake disturbance and is characterized by rapid podzol development, including a marked decline in phosphorus concentrations, accumulation of a thick organic horizon, and formation of a cemented iron pan. Tree basal area increased for the first few hundred years and then declined in parallel with the decline in total soil phosphorus, consistent with the concept of forest retrogression. There were also marked changes in the composition of the tree community, from dominance by conifers on young soils to a mixed conifer–angiosperm forest on old soils. Although a variety of factors could account for these changes, partial Mantel tests revealed strong correlations between tree community composition and soil nutrients. The relationships differed among life history stages, however, because the adult tree community composition was correlated strongly with nutrients in the mineral soil, whereas the seedling community composition was correlated with nutrients in the organic horizon, presumably reflecting differences in rooting depth. The changes in the tree community at Haast are consistent with disturbance-related succession in conifer–angiosperm forests in the region, but the opposite of patterns along the nearby Franz Josef post-glacial chronosequence, where conifers are most abundant on old soils. The Haast chronosequence is therefore an important additional example of forest retrogression linked to long-term soil phosphorus depletion, and provides evidence for the role of soil nutrients in determining the distribution of tree species during long-term succession in lowland temperate rain forests in New Zealand.

This study, one of the first to document the effects of forest fragmentation on insects in the tropics, showed that dung and carrion beetle communities in 1-ha and 10-ha forest fragments differed from those in contiguous forest, even though the fragments had been isolated by <350 m for an ecologically short time (2-6 yr). During 288 pitfall trap days at the Minimum Critical Size of Ecosystems study site 80 km north of Manaus, Brazil, I captured 55 species in 15 genera. Trap days were divided equally between three 1-ha, three 10-ha, three clear-cut, and three contiguous forest areas. Pitfall samples from clear-cut areas separating forest fragments from intact forest in- dicated that beetles rarely moved from intact forest into fragments. The apparent barrier imposed by clearcuts diminished with the invasion of second growth. Except for the four species in the genus Glaphyrocanthon, all species were found more frequently in forested areas than in clearcuts. Glaphyrocanthon constituted 97% of the 717 individuals captured in clearcuts and was never captured in contiguous forest or 10-ha fragments. Forest fragments had fewer species, sparser populations, and smaller beetles than com- parable intact forest areas. The changes in dung and carrion beetle communities help explain the low rates at which dung decomposed in 1-ha fragments. Thus, forest fragmentation not only changes the dung and carrion beetle fauna; its effects ramify through other related community and ecosystem processes, as well.

A majority of the research on forest fragmentation is primarily focused on animal groups rather than on tree communities because of the complex structural and functional behavior of the latter. In this study, we show that forest fragmentation provokes surprisingly rapid and profound alterations in tropical tree community. We examine forest fragments in the tropical region using high-resolution satellite imagery taken between 1973 and 2004 in the Southern Western Ghats (India) in relation to landscape patterns and phytosociological datasets. We have distinguished fragmentation in six categories--interior, perforated, edge, transitional, patch, and undetermined--around each forested pixel. Furthermore, we have characterized each of the fragment class in the evergreen and semi-evergreen forest in terms of its species composition and richness, its species similarity and abundance, and its regeneration status. Different landscape metrics have been used to infer patterns of land-use changes. Contiguous patches of >1,000 ha covered 90% of evergreen forest in 1973 with less porosity and minimal plantation and anthropogenic pressures; whereas in 2004, the area had 67% forest coverage and a high level of porosity, possibly due to Ochlandra spread and increased plantations which resulted in the loss of such contiguous patches. Results highlight the importance of landscape metrics in monitoring land-cover change over time. Our main conclusion was to develop an approach, which combines information regarding land cover, degree of fragmentation, and phytosociological inputs, to conserve and prioritize tropical ecosystems.

Traits determine species response to climate conditions and the match between phenotypes and climate mediates spatial variation in species composition.These trait-climate linkages can be disrupted in human-modified landscapes. Human land use creates forest fragments where dispersal limitation or edge effects exclude species that may otherwise suit a given macroclimate. Furthermore, stressful macroclimate can limit viable trait combinations such that only a subset of values of any given trait occurs with respect to another trait, resulting in stronger trait covariance. Because forest loss can compound climatic stress, trait covariance from benign to harsher climates is expected to be stronger in fragments compared to contiguous forests. In a wet tropical forest landscape in the Western Ghats Biodiversity Hotspot of peninsular India, I compared fragments with adjacent contiguous forests for signatures of trait-mediated assembly of tree communities. Using four key plant traits-seed size, specific leaf area (SLA), wood density, and maximum height-I evaluated trait-abundance associations and trait covariance across climate, soil, and elevation gradients. In the contiguous forest, smaller-seeded, shorter, thinner-leaved species became more abundant from low to high elevations. In fragments, species with higher SLA were more abundant at sites with more seasonal climates and lower precipitation, and larger seeded species were less abundant at warmer sites. However, traits only weakly predicted abundances in both habitats. Moreover, only contiguous forests exhibited significant compositional change via traits, driven by trait syndromes varying along a composite gradient defined by elevation, water deficit, and soil C:N ratio. Site-level trait covariance revealed that warmer, wetter conditions in fragments favored taller species for given seed size, as compared to similar conditions in contiguous forests. Overall, trait syndromes and trait covariance, rather than single traits, determined the phenotypes best suited to macroclimate conditions and should inform management or restoration goals in fragments.