Regeneration and changes in diversity in wildfires from satellite images (Landsta TM and ETM+)

Abstract

The role played by direct regeneration in early post-fire years is well known in mediterranean plant communities. However, the effect of wildfires on medium and long-term vegetation dynamics needs a better understanding. Spatial and temporal post-fire vegetation patterns are fundamental to evaluate forest succession processes, forest structure and species diversity. Focussing on that need, this work presents an analysis of regeneration patterns and diversity dynamics after a fire which occurred in 1993, affecting around 7000 Ha in southern Spain. Field data were recorded from permanent plots in 2000 and 2005. Remotely sensed data were obtained from Landsat ETM+ images of the years 1992, 1993, 2000 and 2004. Both sets of data, fieldwork and remotely sensed data, were analyzed and modulated separately through statistical and quantitative analysis, and then related on the basis of ecological relations. Based on those results, we have developed a stochastic model to predict medium-term forest dynamics. Forests in this area followed the auto-succession model in the early years. However, because of the history of disturbance in the area, it seems unlikely that forests will return to their pre-fire state. Change on NDVI was highly correlated with the changes recorded in the regrowth of the main species, which conditioned the composition of the plant species, and, therefore, diversity indexes such as Shannon (H) and Simpson (D). As a general conclusion, it can be affirmed that NDVI was highlighted as a good understanding of succession stage changes, showing sensitivity to post-fire plant cover changes and diversity dynamics, which therefore recommends its application in the management and restoration of wildfire-affected areas.