Degradation and deforestation in the Conchos river basin (Mexico): Predictive modeling through logistic regression (1985-2016)

Abstract

In this work, the pine, oak and mixed forest areas, with the highest probability of being degraded-deforested, were identified in the Conchos river basin (Chihuahua-Durango, Mexico). Degraded-deforested of pine, oak and mixed forest areas were calculated from 1985-2016 through map algebra (Inventory of land uses and vegetation series I and VI). Subsequently, the logistic regression technique was applied considering eight physical and socioeconomic variables possibly associated with the loss degradation of the forest. A pseudo R2 value of 0.204 was obtained, which indicates high predictive power of the model. An area of 733.43 km2 was identified at risk of being deforested  degraded in the future, representing 6.7% of the forest area currently existing in the basin. The variables with the greatest specific weight in deforestation forest degradation were, in descending order: Marginalization, mining concessions and land tenure. The results are helpful to prevent the loss degradation of pine-oak forests in the basin, and facilitate its environmental management.