Forest Elephant Movements and Habitat Use in a Tropical Forest-Savanna Mosaic in Gabon

Poaching of forest elephants for their ivory has decimated their populations in Central Africa, with a decline of over 62% since 2002. The densely forested country of Gabon in the Congo Basin provides critical habitat for approximately half of the world’s remaining forest elephants. With forest elephants under intense poaching pressure in most of their range, there is a need for information on their habitat use, movements, and ecology to understand how they modify their environment and to maximize the effectiveness of conservation efforts.

Movement ecology can provide insights into forest elephant resource requirements and temporal, seasonal, or spatial patterns regarding when or where an animal is at greatest risk. In 2015, the Gabon Parks Agency (ANPN) in collaboration with Duke University initiated a forest elephant monitoring program using global positioning system (GPS) collars to 1) investigate the spatial distribution and environmental drivers of elephant movements and 2) inform proactive law enforcement strategies by predicting where elephants will be found.

The Wonga Wongué Presidential Reserve (WW), on the western coast of Gabon, consists of a forest-savanna matrix, providing an opportunity to evaluate how elephants use different habitat types. In 2015 and 2016, the Gabon Parks Agency collared 17 forest elephants in the reserve. Here I used the location data from these elephants to ask:

1. Do forest elephant movements and home ranges differ temporally, seasonally, or by elephant sex?
2. Do forest elephants differentially use forest and savanna habitats?
3. What environmental variables (ecological and anthropogenic) most strongly determine habitat use by elephants?

I characterized elephant movements and measured home ranges, employing three different methods of quantifying and visualizing home range areas. I assembled environmental variables of interest (e.g. land cover type, vegetation, and Euclidean distance to streams, villages, and secondary roads) at each elephant location, and analyzed ecological patterns of habitat use across elephant sex, season, and time of day. I developed a prediction surface for the likelihood of elephant occupancy by modelling elephant habitat use during the dry and wet seasons using mixed effects logistic regression.

Forest elephants travelled up to 3000 km annually and exhibited average home range areas of 713 km2, with males having significantly larger home ranges than females. Both female and male elephants remained largely in and around the central savanna of WW, with a few males travelling outside the park borders and travelling up to 110 km to return to the central savanna. Forest elephants demonstrated both temporal and seasonal movement patterns. Temporally, they moved between forest and savanna at dawn and dusk, spending more time in forests during the day when temperatures are highest and entering the savanna as the sun goes down. During the short wet season when grass recruits, forest elephants spent proportionally more time in the savanna. Elephants also travelled at greater speeds during the short wet season when, in addition to greater grass availability, the abundance and diversity of fruiting tree species is greatest, suggesting that availability of fruit influences movements.

The most significant determinant of elephant movements was vegetation; elephants use areas with higher vegetation density, which afford cover and browse. When villages were nearby, elephants tended to spend time nearer to them, perhaps for access to agricultural crops. During the wet season, elephant presence was more likely at greater distances from perennial water sources, indicating decreased water limitation during this season.

The interaction between forest elephants and the large central savanna highlights the important role of the savanna for food provision and social gatherings. Park management should continue to manage the savannas of WW to maintain openness and grass regeneration. Unlike other sites, the interior of WW is well protected, and ecological factors including food and water availability more strongly affect elephant movements than anthropogenic features such as roads and villages. Even so, there is a high likelihood of human-elephant conflict along the border of the reserve where villages are located. To conserve elephants, stakeholder engagement and law enforcement should focus on these areas of potential conflict. Future GPS-tracking efforts should focus on the park boundaries and multiple-use areas between protected areas to assess the anthropogenic impacts on forest elephant movements and the capacity of the protected area network in Gabon to protect and maintain forest elephant populations.