Landscape resistance index aiming at functional forest connectivity.

ABSTRACT

Resistance models may quantify the ability of the landscape to impede species' movement and represent suitable habitats. Moreover, the performance of resistance models parameterized by land-use/land cover attributes evidence that the integrity of the environments subject to urban sprawl is poorly understood. In this sense, the study assumed we could identify the forest functional connectivity in a landscape considering the disparity in the landscape mosaic. In this context, we sought to develop a landscape resistance index through structural equation modeling (SEM), supported by the criteria of heat emission, biomass, and anthropogenic barriers, obtained by remote sensing, called observed variables. The landscape studied in the Green Belt Biosphere Reserve of São Paulo has significant remnants of the Atlantic Forest, a biodiversity hotspot. However, our results indicated criteria variability in the landscape modeled through the SEM, obtaining a significant adjustment of the landscape resistance index, with comparative fit index (CFI) of 1.00 and root mean square error of approximation (RMSEA) of 0.00. The index reflects the resistance levels of the land use/land cover, expressed by the class interval, ranging from 0% (1.73) to 100% (493.88), with the highest values associated with the anthropized uses and forest isolation. Thus, our index based on environmental attributes reflects the structure of functional forest connectivity and offers a framework to design forest corridors across landscapes.