Early ecological outcomes of assisted natural regeneration are strongly site-dependent: Evidence from the Philippines

Assisted natural regeneration (ANR) offers a promising strategy to restore native vegetation after deforestation or severe degradation. However, limited studies have evaluated its ecological impacts, especially across growth stages and incorporating broader ecological attributes. This study assessed ANR outcomes in two regions of the Philippines and examined the roles of remnant vegetation, topography, and soil on native vegetation recovery. We established 50 plots of 0.05 ha across two study sites (Bohol – 20 and Zambales – 30), equally divided between ANR (with weeding and firebreaks implemented in 2017) and adjacent natural regeneration areas of similar history but without ANR interventions (UNR). We sampled woody vegetation across three size classes (i.e., seedlings, saplings, and trees) to compare ecological metrics between treatments after controlling for environmental covariates, and used fixed-effects models to evaluate environmental and landscape drivers of recovery. Within 7 years of recovery, ANR generally promoted more recruitment of endemic, native, zoochorous, and secondary species, whereas UNR plots recruited more exotic and pioneer species. However, effects on species diversity, composition, forest structure, and biomass were strongly site-dependent: ANR effects were strong and positive across all size classes in Bohol but weak, inconsistent, or absent in Zambales. Multivariate analyses further revealed that remnant tree density was the strongest environmental/landscape predictor of forest recovery in Bohol, whereas distance to forest fragments was the strongest predictor in Zambales. Soil variables and aspect affected specific ecological metrics in both sites. While trees and saplings responded predictably to different environmental factors, seedlings exhibited more stochastic relationships in both sites. Our findings show that low-cost ANR interventions, such as firebreaks and grass removal, can enhance native tree species diversity, biomass, and structure of degraded forests, but their effectiveness varies with biotic legacies, abiotic conditions, and landscape contexts. It may take longer for positive outcomes of ANR to emerge in harsher environments. Restoration planning and monitoring should consider these factors to ensure ANR effectiveness and promote long-term forest resilience.