Patterns and drivers of disturbance in tropical forest reserves of southern Ghana

Background

Human activities such as logging, mining, and agriculture, combined with climate change effects like drought, drive forest disturbances and increase wildfire risk. While tropical forests typically remain moist, wildfires have grown more frequent, particularly during El Niño-related droughts. This study focuses on the Upper Guinean Forest (UGF) region in West Africa—a globally significant biodiversity hotspot facing intense climatic and human pressures. This region experiences persistent droughts that are likely to intensify in the future along with mounting pressure from a rapidly growing population. Protected reserves in Ghana contain most of the remaining forests, but human encroachment increasingly threatens these areas.

Goals and Methods

This study examines the spatial patterns and drivers of fire and non-fire disturbances in Ghana's open- and closed-canopy tropical forest reserves from 2013 to 2020. The authors used Landsat imagery to classify forest cover and detect changes, incorporating VIIRS active fire data to distinguish between fire and non-fire disturbances. They applied boosted regression tree (BRT) models to analyze how climate variability (precipitation), human accessibility (population density, proximity to roads and non-forest areas), and landscape structure (fragmentation, percentage of closed forest, slope) influence the likelihood of each disturbance type.

Conclusions and Takeaways

The researchers determined that closed-canopy forests are more vulnerable to degradation, while open forests face higher risks of forest loss. Furthermore, climate variability, particularly during droughts, drives fire as a major disturbance agent. Human accessibility also plays a significant role, as forests near non-forest areas and regions with intermediate population density experience elevated disturbance risks. Notably, previously disturbed forests become more susceptible to future disturbances, creating a positive feedback loop that accelerates degradation and loss. The authors highlight the importance of using models based on climate variability, human accessibility, and landscape structure to identify high-risk areas, allowing practitioners to target conservation and restoration efforts more effectively.