Tropical forests sequester approximately half of the world’s forest carbon and provide essential ecosystem services. However, they are increasingly threatened by land-use conversion and forest degradation driven by expanding human activities. Quantifying the recovery potential of degraded tropical forests is important for estimating future ecosystem service provision and for developing forest management strategies. In lowland timber production forests in Borneo, aboveground biomass (AGB) has been reported to recover within 40–60 years after selective logging. In contrast, some stands do not show such rapid recovery. In stands that experienced intense logging, forest floors and even canopies are covered by ferns and vines, and secondary succession is inhibited by this vegetation, potentially leading to arrested succession. Large-scale assessments of recovery potential across landscapes, including such arrested stands, remain limited.
    In this study, we analyzed 10-year time-series data from 30 permanent 0.12-ha plots spanning a broad degradation gradient in and around the Deramakot Forest Reserve (ca. 1,000 km2) in Sabah, Malaysia. We constructed a hierarchical Bayesian mixed-effects model with AGB increment as the response variable and initial AGB, fern and vine coverage, and elapsed time since the start of observation as predictors. We extrapolated the model to the entire Deramakot area to estimate potential biomass recovery from 2024 to 2034, using 2024 maps of AGB and fern and vine coverage.
    Preliminary results suggest that higher fern and vine coverage is associated with lower AGB accumulation rates. Approximately 30% of the landscape is covered by forests with ferns and vines, and the projected mean biomass accumulation rate across the forest reserve was approximately one-third of the values reported in prior studies. These findings suggest that evaluating ecosystem service provision without accounting for arrested succession may overestimate recovery potential in human-disturbed tropical forests.