Thursday, August 5, 2010

PS 74-81: Habitat split, species life-history, and the matrix tolerance model

C. Guilherme Becker, Cornell University, Carlos Roberto Fonseca, Universidade Federal do Rio Grande do Norte, Celio F. B. Haddad, Universidade Estadual Paulista Juliode Mesquita Filho, and Paulo I. Prado, Universidade de Sao Paulo.

Background/Question/Methods

Species with different life-history traits are expected to respond to habitat change in distinct ways like tolerating more or less disturbed habitats. According to the matrix-tolerance model the vulnerability of species in fragments of natural vegetation is assumed to be directly related to their ability to use the disturbed matrix. Species that are not found in heavily degraded habitats are predicted to be less vulnerable, and vice-versa. However, species breeding habits and the degree of habitat split in the landscape may dictate when a given species will be forced to stay or abandon remnants of natural vegetation. Thus, we ask if breeding habits can influence species vulnerability. Here we (i) investigate how amphibians with distinct breeding habits are affected by habitat split and (ii) discuss how species life-history and landscape configuration influence the assumptions of the matrix-tolerance model. We compared capture rates of forest-associated amphibians with aquatic larvae to species with terrestrial development in forest fragments with streams and without streams, including the intervening grass-field matrix.

Results/Conclusions Species with aquatic larvae were often captured moving towards open habitats, and they presented lower population abundances in fragments disconnected from streams. They also embark on breeding migrations across long distances to find suitable water bodies, and are often forced to leave vegetation remnants. Terrestrial development species, which - can complete their life-cycle even if in small dry fragments clearly avoided open habitats, and were equally recorded at wet and dry fragments. In contrast to the matrix tolerance model, we argue that species that are not found in the matrix do not necessarily present higher susceptibility to decline in the case of habitat fragmentation. Species that can avoid disturbed habitats certainly experience long term genetic erosion due to lower demographic contributions of immigrants and via stochastic population collapses. Nevertheless, the impacts of induced migrations throughout inhospitable habitats may result in immediate declines on populations of forest-associated amphibians with aquatic larvae, which is the group that exhibits higher levels of population declines worldwide.