Landscape resistance index aiming at functional forest connectivity

• Published in: Environmental monitoring and assessment Vol. 195; no. 10; p. 1224
• Main Authors: Vanderley-Silva, Ivan, Valente, Roberta Averna
• Format: Journal Article
• Language: English
• Published: Cham Springer International Publishing 01.10.2023; Springer Nature B.V
• Subjects: Anthropogenic factors; Approximation; Atmospheric Protection/Air Quality Control/Air Pollution; Biodiversity; Biodiversity hot spots; biomass; Biosphere; Brazil; class; conservation areas; Criteria; Earth and Environmental Science; Ecology; Ecotoxicology; Environment; Environmental Management; Environmental monitoring; equations; Forests; Green belts; Heat; Land cover; Land use; Landscape; Landscape design; landscapes; Mathematical models; Modelling; Monitoring/Environmental Analysis; Multivariate statistical analysis; Remote sensing; species; Structural equation modeling; Urban areas; Urban sprawl; Urbanization; Vegetation; Brazil; Landscape structure analysis; Land use and land cover; Structural equation model; Environmental criteria
• ISSN: 0167-6369; 1573-2959; 1573-2959
• DOI: 10.1007/s10661-023-11749-x

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.