Predictive models of the natural regeneration of pine in ecosystems with forest fires

Abstract

In ecosystems the effects of forest fires are variable depending on the severity of the fire. Consequently, the recovery that the
vegetation will have in these areas is different. However, the evaluation in the field of the ecosystems' response to this impact
implies a significant expenditure of resources, either due to the extent of the fire or the inaccessibility of the terrain. Due to this,
alternative strategies are sought for the evaluation and determination of priority areas for the implementation of management, such as the use of spectral indices derived from remote sensors. In this work, the correlation presented by different variables taken directly in the field and obtained by applying spectral indices to satellite images was evaluated to determine predictive models of the natural regeneration of pine after the occurrence of a forest fire. Being the variables of tree canopy diameter, improved vegetation index and slope those that were included in the predictive models, presenting a higher value of R² in the model in which both environmental variables and those taken from satellite images are taken together. Finally, the resulting model with the improved vegetation index was applied to a forest fire one year later and two years after the occurrence of the fire, obtaining as a result a decrease in regeneration individuals two years after the fire, however, it is notorious the tendency to find more regeneration in fire affected areas compared to non-burned areas.