Modeling spatial distribution of bark beetle susceptibility using the maximum entropy approach

Abstract

Bark beetles cause significant damage to forests, which are valuable natural resources. The
creation of susceptibility maps for bark beetles is a significant stage in the management and
reduction of bark beetle-related harm. The present investigation involved the development of
a susceptibility map for bark beetles, utilizing the Maximum Entropy (MaxEnt) model, a
machine learning technique, and incorporating 19 different bioclimatic climate variables. The
model's accuracy was evaluated through receiver operating characteristic (ROC) analysis, and
the area under the curve (AUC) was computed to be 0.705. The MaxEnt model indicated that
the annual mean temperature (BIO 1) had the greatest impact on the susceptibility of bark
beetles. Categorization of bark beetles' susceptibility was delineated into four different
categories, namely low, moderate, high, and extreme. Based on the results, approximately 58%
of the study area included areas that exhibit vulnerability to bark beetle infestation. The
accuracy of the bark beetle susceptibility map, which was developed based on these results,
was found to be high and consistent with the observed bark beetle damage.