Fig. 6: Empirical validation of CAPTAIN. | Nature Sustainability

Fig. 6: Empirical validation of CAPTAIN.

From: Improving biodiversity protection through artificial intelligence

Fig. 6: Empirical validation of CAPTAIN.The alt text for this image may have been generated using AI.

a,b, Maps showing the ranking of priority areas for protection across Madagascar based on the distribution of endemic trees and on a limited budget allowing for up to 10% of protected area overall: protection units identified through Marxan optimizations (with BLM = 0.1, as in Carrasco et al.33; a) and equivalent CAPTAIN results based on a full initial monitoring policy (b). The colour scale represents the relative importance of each protected unit, measured as the frequency at which it was included in the optimized solution. c,d, Histograms of the fraction of species ranges included in the protected units calculated for each species with Marxan (c) and CAPTAIN (d). The red dashed lines indicate the 10% threshold set as the target of the policy. Orange bars show species whose protection did not meet the target.

Back to article page