An improved understanding of past interactions between terrestrial vegetation and various forcings, such as climate change, human impact, and paleofire, is crucial for assessing impacts of future global change on terrestrial ecosystems. This study seeks to find the key factor or factors that have driven Holocene vegetation change along the East Asian monsoon margin. Several high-resolution pollen records are reviewed and new phytolith-based paleovegetation reconstructions and physical geochemical datasets are presented from a peatland in northeastern China. Using 108 modern topsoil samples as a training set, canopy cover and vegetation composition are estimated for the period since 5100 cal. yr BP. Variation partitioning analysis (VPA) is used to determine the relative importance of climate change, human impacts, and paleofire disturbance. The generalized dynamic vegetation model LPJ-GUESS is forced with climate anomaly output from an atmospheric general circulation model to simulate vegetation dynamics during the mid-Holocene and the pre-industrial era. The proxy-based estimates are compared to modelling output. Results indicate that regional tree cover varied from 10% to 40% during the past five millennia. The single-core, phytolith-based reconstructions are generally consistent with stacked tree pollen z-scores calculated from different records along the East Asian monsoon margin, implying that mid-Holocene tree cover decrease was persistent and almost synchronous over extensive areas. VPA demonstrates that long-term monsoon marginal vegetation successions were mainly caused by climate effects. Numerical modelling suggests that since the mid-Holocene the retreat of forests along the monsoon margin was primarily associated with precipitation deficits. Our investigation highlights that the precipitation associated with the East Asian monsoon system has exerted a stronger influence than the westerlies on the monsoon margin climate and vegetation change. With ongoing global change, close attention to variations in precipitation patterns and amounts should be especially helpful in efforts aimed at ecological monitoring, change prediction, and restoration.
Read full abstract