Mapping Terrestrial Conservation: Showcasing Model Assessment for Map Performance Evaluation

Abstract

 Habitat preservation, conservation, and protection of threatened species require careful site selection procedures. The presence of several site-selection models encourages model assessments to determine their usefulness and functionality for conservation purposes. Map model assessments are not conventionally conducted between site selection models to understand the nature of their decision-making strategies and prioritization. Therefore, the objective of this study was to determine the relationship between a recently published site selection model, the Safe Zone with MaxEnt, representing the Species Distribution Model, and the Forest Landscape Integrity Index, representing an intact forest habitat. As expected, there was considerable variation among the three models due to their different algorithms and themes of concern; nevertheless, these had some level of agreement in selecting potential areas for conservation. In addition, this study presented a convenient method for map data collection and the analysis required for a comprehensive understanding of the model agreements and relationships with other models. The nature of the decision making of the Safe Zone model and its potential utility for selecting Strict Protection Zones was presented. This study highlighted the intersection of geospatial analysis, conservation, and model evaluation. Finally, study emphasized the importance of informed decision-making in environmental planning and management through rigorous map assessment. The limitations of this study include the varied pixel resolutions of global and local maps, temporal predictions, and the fact that the models to which the SZ was compared are limited to models and maps available in the area of interest.

Keywords: model evaluation; map comparison; category agreements; map model relationships; safe zone for conservation