The impacts of disturbances on mountain ecosystem services: Insights from BEAST and Bayesian network

Abstract

Disturbances, such as wildfire, insect outbreak, wind throw and building construction, can significantly alter ecological processes and impact the supply of mountain ecosystem services (ESs). The comprehensive impacts of vegetation-related disturbances on ESs are poorly known, as most of existing studies have been focused on the impacts of single disturbance. Using the Eastern Helan Mountain as a case study, the vegetation-related disturbances during 1989–2020 were detected and mapped through the BEAST algorithm. Then two critical regulating ESs of carbon sequestration and sand fixation were quantified using CASA model and RWEQ model, respectively. Bayesian Network (BN) was further applied to explore the impacts of disturbances on ESs and identify the “win-win” solutions of ESs in terms of disturbance. The results suggested that BEAST could effectively identify vegetation-related disturbances at various spatio-temporal scales, which was difficult to be captured by normal land use/cover products. The established BN model could simulate the spatio-temporal pattern of sand fixation and carbon sequestration with an average classification error of 13.36% and 9.83%, respectively. Forest conservation and return pasture to forest policy can reduce disturbance frequency and intensity, and improve these two ESs. Both ES generally decreased with disturbance intensity, but firstly decreased and then increased with disturbance frequency. Thus, slight disturbance is beneficial for these ESs, especially in grassland at medium elevation. The “win-win” solutions of these two ESs can be achieved under frequent disturbances with the lowest intensity. Despite the low impacts of disturbances on ESs than other nodes in the BN, their impacts should also be taken into account into ecosystem management due to their inevitability. This BN-based approach turns out to be a promising decision support tool for mountain ecosystem managers to integrate small-scale disturbances into management and provide key information for post-disturbance recovery.