Human activities alter disturbance regimes, influencing forest structure and ecosystem carbon. Identifying and quantifying natural and anthropogenic disturbances at fine spatial scales are critical to assessing the role of forests in climate change mitigation. This study investigated disturbance history and carbon storage in human-impacted forests and open ecosystems at elevations of 2000-4200 m in northwestern Yunnan, China. We established 50 sampling plots along the four due orientations of a mountain peak. Using tree rings, fire scars, satellite imagery, official records, and interviews, we reconstructed historical disturbances and identified fires, logging events, landslides, and icy precipitation since the 1950s. We analyzed the impact of disturbance history and topography on ecosystem carbon storage, including pools in soil (0-30 cm), woody debris, and non-woody and woody plants. Disturbances since the 1950s were largely driven by anthropogenic activities over time, along with climate and topography. Fires and logging were common near settlements in Pinus yunnanensis forests at lower elevations, while landslides primarily occurred in steep areas of Abies georgei forests and logged areas within broadleaf mixed forests. Icy precipitation was more frequent above 3500 m on the south and west slopes in A. georgei forests. Non-forest areas at higher elevations had a mean ecosystem carbon (including soil carbon down to 30 cm) density of 146 Mg C ha-1, while forest areas averaged 270 Mg C ha-1. Fire negatively impacted soil, woody plants, and overall ecosystem carbon, whereas logging impacts were limited to woody plants and overall ecosystem carbon. Carbon storage in woody plants and total ecosystem carbon followed a hump-shaped pattern with elevation, peaking near 3200 m. Our study links disturbance histories to spatial variation in carbon pools. This study helps improve carbon management and conserve biodiversity in human-modified forests and presents a multi-source approach that could be used in other human-impacted forests.
Cheng et al. (Mon,) studied this question.