Validating movement corridors for African elephants predicted from resistance-based landscape connectivity models

Abstract

CONTEXT: Resistance-based connectivity models are widely used conservation tools for spatial prioritization and corridor planning, but there are no generally accepted methods and recommendations for validating whether these models accurately predict actual movement routes. Hence, despite growing interest and recognition of the importance of protecting landscape connectivity, the practical utility of predictions derived from connectivity models remains unclear. OBJECTIVES: The difficulties in validations are mainly related to the unavailability of independent data and lack of appropriate, easily applied statistical frameworks. Here, we present a case study where two independently collected datasets were used to validate resistance-based landscape connectivity models and movement corridors identified by these models. METHODS: We used annual aerial counts to evaluate the connectivity model, and a field survey to assess the performance of predicted corridors. We applied these two independent datasets to validate a previously developed connectivity model for the African elephant (Loxodonta africana) in the Borderland region between Kenya and Tanzania. RESULTS: The results of this study confirm that the resistance-based connectivity model is a valid approach for predicting movement corridors for the African elephant. We show that high connectivity values are a strong predictor of the presence of large numbers of the elephants across the years. The probability of