Modeling the risk of illegal forest activity and its distribution in the southeastern region of the Sierra Madre Mountain Range, Philippines

Abstract

Illegal activity within protected forests, such as illegal logging, slash-and-burn farming, and agricultural expansion, has brought many plant and animal species to the brink of extinction and threatens various conservation efforts. The Philippine government has introduced a number of actions to combat environmental degradation, including the use of mobile platforms such as the SMART-Lawin system to collect patrol data from the field, which represents a remark-able step towards data-driven conservation management. However, it remains difficult to control illegal forest activity within protected landscapes due to limited patrol and law enforcement resources. A better understanding of the spatial distribution of illegal activity is crucial to strengthening and efficiently implementing forest protection practices. In the present study, we predicted the spatial distribution of illegal activity and identified the associated environmental factors using maximum entropy modeling (MaxEnt). Patrol data collected using the SMART-Lawin system from the Baliuag Conservation Area for the period 2017-2019 were used to train and validate the MaxEnt models. We tuned the MaxEnt parameter setting using the ENMeval package in R to over-come sampling bias, avoid overfitting, and balance model complexity. The re-sulting MaxEnt models provided a clear understanding of the overall risk of illegal activity and its spatial distribution within the conservation area. This study demonstrated the potential utility of data-driven models developed from patrol observation records. The output of this research is beneficial for conservation managers who are required to allocate limited resources and make informed management decisions.