Dynamics of the giant panda habitat suitability in response to changing anthropogenic disturbance in the Liangshan Mountains

Abstract Through decades of conservation efforts at the national and worldwide scales, the giant panda (Ailuropoda melanoleuca) has been downgraded from “endangered” to “vulnerable” on the IUCN Red List. The decision has not gone unquestioned, however. Although the population has increased, pandas still face serious human disturbances (e.g., livestock grazing and logging). Human disturbance is a crucial factor affecting the distribution of wildlife and their habitats, and evaluating the change of habitat quality in response to this disturbance is vital to wildlife conservation. However, the quantification of human activity variables that result in habitat changes has been rare. To fill this knowledge gap, we used human disturbance (i.e., livestock grazing, logging, farming, collection of bamboo shoots, collection of medicinal plants, road construction), environmental (i.e., forest cover, elevation, slope) and panda GPS locational data to evaluate panda habitat suitability and space use, the impact of human disturbance on panda habitat, and the distribution characteristics of human disturbances over time. Our results show that pandas' habitat suitability has improved, with substantial areas of stabilizing and increasing habitat across our Liangshan Mountains study area between the 3rd national survey of the giant panda (2000) to the 4th national survey of the giant panda (2012). The percent contribution of human disturbance variables to MaxEnt panda habitat suitability models decreased greatly between the 3rd and 4th survey (93.6% to 59.5%), which was especially driven by the decrease in logging's contribution (78% to 19%). The percent contribution of environmental variables consequently increased (6.4% to 40.5%), the largest growth occurring in the forest cover variable (2.7% to 14.2%). Our results also indicate that panda space use expanded between the national surveys. It is concerning that human disturbances are still widely distributed and the percent contribution of some disturbance variables to panda habitat suitability increased. For example, livestock grazing went from a 1.4% to 13.1% contribution between the 3rd and 4th national surveys, and indicates a need to strengthen management in this area. The methods used in this study could also be applied to the assessment of human disturbance effects on habitat quality for other species, and our results provide scientific support to relevant management departments for wildlife and biodiversity conservation in South-central China.