Fig 1.
Map of Kenya showing the study area and the spatial distribution of anthrax occurence data (red circles) recorded between 2011 and 2017.
Areas 1–4 arbitrarily represent important regions for describing the predicted distribution of anthrax: 1) western highlands; 2) Lake Victoria basin; 3) southwestern region; 4) central highlands.
Table 1.
Variables fitted in BRT algorithm for niche modeling.
Fig 2.
Panel A: Predicted geographic distribution of anthrax in southern Kenya based on the mean prediction of an ensemble of 100 boosted regression tree experiments.
Panel B shows the upper 97.5% and C the lower 2.5% confidence intervals. Regions 1–4 are used as reference areas for the discussion: 1) western highlands; 2) Lake Victoria basin; 3) southwestern region; 4) central highlands.
Fig 3.
Location of National Parks and Game Reserves overlaid on the predicted distribution of anthrax in southern Kenya based on boosted regression tree experiments.
Fig 4.
Variable relative influence for final variable set used to model the distribution of anthrax in southern Kenya using boosted regression tree experiments.
Error bars represent variability across an ensemble of 100 BRT experiments.
Fig 5.
Partial dependency plots (PDP) showing marginal effects on the mean prediction probability of potential anthrax distribution by each variable across the 100 BRT experiments.