< Back to Article

Forest Elephant Crisis in the Congo Basin

Figure 2

Results of Fitting a Logistic Regression Model to Elephant and Human Presence/Absence Data Pooled across MIKE Survey Sites

Distance to road (in kilometres) and site were used as explanatory variables. (A) shows the elephant data, and (B) shows the human data. The observations and regression lines are colour-coded by site and the dashed line shows the regression line without the inclusion of site as a covariate. The covariates distance to road and site are significant for both elephant and human probability of occurrence. The dissimilarity between sites is more pronounced when modelling the probability of elephant occurrence.

Figure 2