Search

Loss and fragmentation of natural habitats are key contributors to the decline of populations and impoverishment of biological communities. The response to these disturbances can vary substantially among taxa and depends on spatial metrics of habitat fragments and the surrounding landscape. Herein we test how fragment area, shape, isolation, and matrix quality affect reptile richness, abundance, and occurrence in Brazilian Atlantic Forest fragments, a biodiversity hotspot with a poorly studied reptile fauna. We used 23 forest fragments, ranging from 2 to 30 hectares, surrounded by different matrix types, including sugarcane crop fields, cattle ranching, subsistence farmlands and rural communities. Species richness, total reptile abundance, population abundance, and occurrence probability of many species decreased with fragment area. Model selection suggested that fragment area is the main predictor of both richness and abundance, but matrix quality as well as fragment shape are also important predictors. For population abundance and occurrence probability, fragment area and proximity were the most important predictors followed by fragment shape and matrix quality, but the strength and even the sign of predictors varied substantially among species. We highlight that the value of small fragments should not be neglected for the conservation of Atlantic Forest reptiles. Resumo A perda e a fragmentacao de habitats naturais sao agentes-chave no declinio de populacoes e no empobrecimento de comunidades biologicas. A reposta a esses disturbios pode variar substancialmente entre taxons e depende de metricas espaciais dos fragmentos e da paisagem circundante. Nos testamos como a area, a forma, o isolamento e a qualidade da matriz de fragmentos da Mata Atlântica brasileira afetam a riqueza, a abundância e a ocorrencia de repteis. Nos usamos 23 fragmentos florestais, entre 2 e 30 hectares, rodeados por matrizes comuns no cenario atual da Mata Atlântica, um dos hotspots de biodiversidade. Ao contrario das expectativas teoricas, a riqueza, a abundância total, a abundância populacional e a probabilidade de ocorrencia de muitas especies diminuiu com o aumento da area dos fragmentos. A selecao de modelos sugere que a area dos fragmentos e o principal preditor tanto da riqueza quanto da abundância das especies, mas que a qualidade da matriz, assim como a forma dos fragmentos, sao tambem metricas importantes. Para a abundância populacional e a probabilidade de ocorrencia das especies os preditores mais importantes foram a area e o isolamento, seguidos pela forma e qualidade da matriz dos fragmentos, mas a forca e ate mesmo o sinal do efeito mudou substancialmente entre especies. Nos destacamos que, considerando o alto grau de disturbio da paisagem, o valor de pequenos fragmentos nao deve ser negligenciado em estrategias de conservacao de repteis da Mata Atlântica.

Aim Habitat split isolates forest fragments where terrestrial adults of many amphibian species live from aquatic-breeding sites where their larvae develop, causing population declines and shifts in community structure. In contrast to most fragmentation theoretical frameworks which point to the relevance of fragment area and isolation, a recent theoretical model predicts that split distance, defined as the shortest distance between a forest fragment edge and the nearby stream, should affect negatively the abundance and occurrence probability of populations in forest fragments and, consequently, their total abundance and species richness. The model further predicts that the quality of the matrix between the two habitats, called split matrix, will modulate such responses. Here, we provide the first test of these theoretical predictions. Location Atlantic Forest, Brazil. Methods Split distance was estimated for all fragments of the Brazilian Atlantic Forest, a highly disturbed biodiversity hotspot. Then, we selected 23 forest fragments, within a 1166-km2 landscape, in a design that insured orthogonality between split distance (0–2000 m) and fragment size (1.7–30 ha). Split matrix quality (SMQ) and proximity, a measure of the availability of forest around the focal fragment, were estimated and controlled for in the analyses. Using multiple regression analyses with control for spatial autocorrelation, and AIC multimodel inference, we evaluated the relative importance of split distance, SMQ, fragment size and proximity to determine amphibian community and population attributes. Results Most Atlantic Forest fragments were affected by habitat split. Consistent with the theoretical predictions, species richness, abundance and the occurrence probability of many species declined with split distance and increased with SMQ, even when controlling for fragment size and isolation. Main conclusions Split distance can be used as a surrogate for conservation value of forest fragments for aquatic-breeding amphibians in landscape planning and restoration enterprises.