Impacts of Fires on the Normalized Difference Vegetation Index (NDVI) : Case of the Pendjari (…)

ABSTRACT

The Pendjari Biosphere Reserve (PBR), located in the West African savanna, plays a key role in maintaining regional biodiversity. However, the reserve faces a significant challenge due to the recurrent occurrence of early fires, which may influence vegetation dynamics and ecosystem functioning. This study aimed to identify trends and changes in vegetation cover and to assess their implications for the management and conservation of natural resources in the PBR. To achieve this, NDVI data (MOD13Q1) and burned area data (MCD64A1) derived from MODIS satellites were collected over a 22-year period. Using ArcGIS, NDVI pixels were matched with burned area data to generate vegetation indices associated with fire events. The spatio-temporal dynamics of vegetation were mapped, and the rate of vegetation change before and after fires was quantified using Ranson's equation. In addition, Autoregressive Integrated Moving Average (ARIMA) models were applied to forecast future vegetation dynamics under current fire regimes. The results revealed a progressive shift from dense to more open vegetation formations. A t-test indicated a highly significant difference (p < 0.001) in NDVI values between areas subjected to repeated early fires and those affected by late fires. Under both fire scenarios, ARIMA models project a continued decline in NDVI values. It is important to note that these trends are indicative rather than conclusive, as key drivers such as rainfall, grazing intensity and fire severity were not explicitly incorporated into the analysis. Overall, the findings highlight that both fire timing and frequency are critical drivers of vegetation dynamics in the PBR. Adaptive fire management strategies, including controlled early burning and reduced fire recurrence, are essential to sustain vegetation resilience and ecosystem stability in this fire-prone savanna landscape.