Human disturbance intensity reduces drought resistance without prolonging recovery in karst forests of southwest China

Read full text
Authors: Lu Wang, Yuemin Yue, Liang Shi, Weiye Wang, Jiawen Cui, Qian Li, Boyi Liang, Chongyang Xu, Hongyan Liu, Martin Brandt, Kelin Wang
Cited by: 0

Abstract

Role of human disturbance in shaping forest drought responses remains insufficiently understood, particularly in widely-distributed and ecologically-fragile karst landscapes. Here, we analyzed six major drought events over the past two decades in the karst forests of Southwest China, a region representative of tropical and subtropical karst ecosystems with intensive human activity. Using high-resolution gross primary productivity (GPP) and Standardized Precipitation Evapotranspiration Index (SPEI) datasets, we compared drought resistance and recovery time between low- and high-disturbance forests. High-disturbance forests experienced significantly greater maximum GPP losses during droughts, with differences reaching up to 13% under comparable drought intensities. Beyond the dominant influence of drought intensity, climatic conditions (temperature and precipitation) and soil factors (soil organic carbon and soil thickness) exerted stronger effects on drought-induced GPP loss in high-disturbance forests than in low-disturbance ones. Interestingly, despite suffering greater productivity loss, high-disturbance forests exhibited comparable, and occasionally even shorter, recovery times than low-disturbance forests, suggesting a possible trade-off between drought resistance and recovery capacity. Although recovery time was not consistently prolonged, the greater productivity losses and elevated climatic and edaphic sensitivity in high-disturbance forests indicate reduced resistance and heightened vulnerability to future climate extremes and progressive drying trends.

Fields of Study & Subjects

Plant Water Relations and Carbon Dynamics Soil erosion and sediment transport Karst Systems and Hydrogeology

Publication details

Journal / source
Figshare
Publication date
1 January 2026