Long-term recoveries of forest ecosystem components after shallow landslides were asynchronized among slope positions

Abstract

Landslides are a common disturbance in mountainous areas of the world. Transporting and accumulating landslide debris, i.e., disturbance legacies, such as coarse woody debris (CWD), vegetation patches, and surface soils, generate a heterogeneous environment along slopes (zones), which are suggested to affect forest recovery. However, the long-term changes in forest ecosystems after landslides remain unknown, particularly zone-dependent change patterns. We aimed to reveal the differences in the changes in live trees, understory vegetation, CWD, and soils among zones by surveying forests with landslide ages (years since the landslide) ranging from 3 to 74 years and reference stands. The increase in live tree aboveground biomass occurred at a faster rate at the lower part of the slopes where the disturbance legacies were rich and surface soils were stabilized due to the smaller slope angle. Chronological patterns of understory vegetation amounts were determined by the differences in disturbance legacy richness and the timing of subsequent canopy closure. The amounts and decay-class diversity of CWD were determined by the differences in legacy richness and mortality through stand development. These zone-dependent chronological changes influenced litter production and determined the recovery rates of surface soil carbon stocks and nitrogen contents. The increase in the dominance rates of forest herbaceous species was faster in the lower part of the slopes due to the faster surface soil development and canopy closure. Our results illustrate that long-term forest ecosystem succession and recovery after landslides occurs more rapidly at the lower parts than at the upper parts of slopes. HighlightsO_LIRich landslide legacies and low slope angles promoted vegetation and soil recovery C_LIO_LILegacy richness and timing of canopy closure determined changes in the understory C_LIO_LILandslide mortality and stand development mortality determined changes in dead wood C_LI