The importance of landscape parameters in determining African Elephants (Loxodonta africana) directional movement decisions

Background/Question/Methods

Understanding how movement decisions of individual animals are related to their surrounding landscape can promote the conservation of species and their habitat. Moreover, the effect of the animal’s internal state on its movement pattern has not been closely examined. In southern Africa, land use changes and fencing of reserves restrict the movement of African elephants (Loxodonta africana), causing vegetation degradation and leading to increased human-wildlife conflict. Here, we examine what is the relative importance of detailed landscape parameters in determining elephants’ movement direction and speed. We fitted fifteen elephants with GPS-GSM collars in Etosha National Park, Namibia. Using extensive field work and satellite imagery we created detailed vegetation classification and continuous maps of vegetation density, cover, productivity, and temporal variability. Distances to roads, fences, and water sources were calculated. We inferred elephants’ internal state (hunger, thirst) based on the biophysical characteristics of their previous locations. We used a step selection function to compare the relative importance of these landscape parameters in determining elephants’ directional movement, and examined the effect of individuals’ condition and the season on this relationship. We further used logistic regression to understand the relationship between elephants’ movement speed and the habitat quality, as perceived by the individual.

Results/Conclusions

Our results show that elephants prefer to walk into areas with higher annual maximum and annual average productivity, but lower temporal variance of productivity. Medium cover of Mopani or Acacia trees was the preferred habitat type. Elephants chose to move towards areas with lower annual rainfall, perhaps since these areas contained shorter trees and higher grass cover. Further, we found that elephants stay closer to water resources; and, surprisingly, closer to fences and roads. The sex of the individual and the season affected how elephants responded to the landscape. Interestingly, these results persisted at different temporal resolutions but were sensitive to the spatial scale of the underlying landscape information. Analysis of movement speed yielded similar conclusion about the elephants’ habitat preference, while providing further insight into why particular areas might be preferred. Elephant moved slower in areas with high productivity of vegetation, but moved faster through areas with higher temporal variability of vegetation productivity, which reinforces their preference of habitats with more predictable vegetation quality. Our methodology for combining detailed landscape maps, previous location attributes and both movement direction and speed can be used to improve the understanding of how wildlife respond to natural and anthropogenic landscape features.