Abstract
In this study, we examine the spatio-temporal patterns of citizen-reported human-bear conflict (HBC) from 2002 to 2022 and use the Forest-Based and Boosted Classification (FBBC) technique to assess the significance of several factors in the occurrence of HBC. Our analysis reveals a significant increase in HBC incidents over the study period, with the fewest conflicts in 2002 (217) and the most in 2022 (4455). These were concentrated in northwestern Connecticut, particularly eastern Litchfield County and western Hartford County. The results of geostatistical analysis, including measures of dispersion and emerging hot spot analysis indicated a southward trend in HBC on both annual and monthly scales. The validation results of the FBBC highlighted the relevance of forest fragmentation, intermediate housing density, proximity to water bodies, and snowfall in predicting HBC. Each variable demonstrated nearly equal importance (20%) in predicting HBC occurrences from 2010 to 2022, though land cover showed no significant predictive power. These findings elucidate the spatio-temporal dynamics of HBC and offer valuable insights for wildlife managers to prioritize conflict mitigation strategies effectively. The results of this study identify locations prone to HBC. Moreover, FBBC results show that this technique can be used to predict future HBC based on projected changes in these variables due to climate change and expansion of the human-wildlife interface. Our analysis can aid in the development of targeted, evidence-driven, and ethical management interventions in Connecticut.