Plataforma de modelagem baseada em agentes (ABM) em R: simulação de abundância e revisão de métodos de controle de Sus scrofa

Abstract

The wild boar (Sus scrofa ) is considered one of the most impactful invasive species in the world, causing both environmental and economic damage. Control strategies for this animal have proven ineffective, often based on methods lacking adequate population monitoring and scientific basis. This work aimed to present an agent-based computational platform (ABM) developed in R, which performs fauna monitoring simulations via camera traps. The species Sus scrofa was chosen as a case study in a region of southwestern São Paulo state. 40,000 agent trails, representing herds of wild boars, were simulated, and the performance of Random Encounter Models (REM) in estimating population density was evaluated. The results revealed that the use of exclusion windows, a common practice in camera trapping studies, tends to overestimate population density and reduce agreement with real values, contradicting bibliographic recommendations. Furthermore, landscape structure was a determining factor in the accuracy of the estimates. In parallel, a review of control and monitoring methods highlighted the difficulty of controlling the species in Brazil. It is concluded that the MAB platform represents a promising tool for optimizing monitoring studies and that the effective management of Sus scrofa requires a transition to evidencebased strategies, integrating precision monitoring, predictive modeling, and planned control.