A reconstruction of the recent fire regimes of Majete Wildlife Reserve, Malawi, using remote sensing

• Published in: Fire ecology 2021-01, Vol.17 (1), Article 4
• Main Authors: Nieman, Willem A., van Wilgen, Brian W., Leslie, Alison J.
• Format: Article
• Language: English
• Subjects: Annual rainfall; Biomedical and Life Sciences; Burning; Dry season; Ecology; Ecosystem management; Fires; Forestry; Image reconstruction; Infrared radiometers; Intervals; Life Sciences; MODIS; Original Research; Policies; Protected areas; Rainfall; Remote sensing; Remote sensors; Savannahs; Seasons; Sensors; Spectroradiometers; Vegetation; Wildlife; Wildlife conservation; Wildlife refuges
• ISSN: 1933-9747; 1933-9747
• DOI: 10.1186/s42408-020-00090-0

Background Fire is an important process that shapes the structure and functioning of African savanna ecosystems, and managers of savanna protected areas use fire to achieve ecosystem goals. Developing appropriate fire management policies should be based on an understanding of the determinants, features, and effects of prevailing fire regimes, but this information is rarely available. In this study, we report on the use of remote sensing to develop a spatially explicit dataset on past fire regimes in Majete Wildlife Reserve, Malawi, between 2001 and 2019. Moderate Resolution Imaging Spectroradiometer (MODIS) images were used to evaluate the recent fire regime for two distinct vegetation types in Majete Wildlife Reserve, namely savanna and miombo. Additionally, a comparison was made between MODIS and Visible Infrared Imager Radiometer Suite (VIIRS) images by separately evaluating selected aspects of the fire regime between 2012 and 2019. Results Mean fire return intervals were four and six years for miombo and savanna vegetation, respectively, but the distribution of fire return intervals was skewed, with a large proportion of the area burning annually or biennially, and a smaller proportion experiencing much longer fire return intervals. Variation in inter-annual rainfall also resulted in longer fire return intervals during cycles of below-average rainfall. Fires were concentrated in the hot-dry season despite a management intent to restrict burning to the cool-dry season. Mean fire intensities were generally low, but many individual fires had intensities of 14 to 18 times higher than the mean, especially in the hot-dry season. The VIIRS sensors detected many fires that were overlooked by the MODIS sensors, as images were collected at a finer scale. Conclusions Remote sensing has provided a useful basis for reconstructing the recent fire regime of Majete Wildlife Reserve, and has highlighted a current mismatch between intended fire management goals and actual trends. Managers should re-evaluate fire policies based on our findings, setting clearly defined targets for the different vegetation types and introducing flexibility to accommodate natural variation in rainfall cycles. Local evidence of the links between fires and ecological outcomes will require further research to improve fire planning.