Forest elephants (Loxodonta cyclones) are a distinct species from African savanna elephants (L. africana), are smaller, and highly frugivorous. They typically live in rainforest habitats throughout West and Central Africa. They are thought to make up one-third of the entire African elephant population, but are much less studied because they inhabit dense forest environments, making them difficult to see. Forest habitats make it difficult to monitor and manage populations, which is a concern because populations are currently under severe threat from poaching. These factors make forest elephants a priority for scientific merit and conservation value.
My research goals were to identify forest elephant social structure and ranging behaviors using satellite telemetry, non-invasive genetic techniques, and network models. Studies in savanna elephants have shown that a higher-level, fission-fusion social structure exists, where core family groups of elephants can join, separate from, and re-join larger groups. Groups that associated with one another are related in populations undisturbed from poaching. I was interested in determining the relatedness patterns of groups in forest elephants to detect if higher-level social structure exists by investigating the association patterns over time and relationships across the larger landscape. Observations on forest elephants suggest that they have smaller groups, but the forest environment may prevent individuals from being detected and therefore cause a downward bias in group sizes. Therefore, it is uncertain if this reflects the true relationships in forest elephant society, or they are more similar to African savanna elephants, but just more difficult to detect.
Space Use Patterns and Home Ranges
With GPS satellite telemetry, I analyzed the movement patterns of six forest elephants in Loango National Park, Gabon. Surprisingly, their home ranges were very small (11-105 km squared) even though they were tracked for close to two years. Their home ranges were not only small, but also adjacent to one another with minimal overlap. This differs from many spatial patterns found in savanna elephants, and possibly as a result of difference in sociality, which inspired further questions in my dissertation research. These results are published in Biotropica.
Spatial Genetic Patterns
Using non-invasive genetics (from elephant dung) and spatial autocorrelation, I also examined the relationship between spatial proximity and genetic relatedness between adult females (males are solitary, and calves/subadults are dependent on mothers) in Lopé National Park, Gabon. If females are associating with other related females, than genetic structure should exist when females are in closer proximity to one another. Although I did find evidence of genetic structure, structure was not strong and genetic diversity as represented by mitochondrial haplotypes within the area was high. However, when I investigated the relationship between individuals as inferred from their grouping patterns from dung, I found that individuals were mostly associating with related individuals and of those of the same mitochondrial haplotype. Given the high diversity of haplotypes in the area and that they were not spatially segregated within the area, this suggests that associations are not random or based on resources, but individuals are choosing to be with those of the same matriline. These results are in review.
Social Network Analyses
Finally, I identified individual adult females and their associates when found both solitarily and in groups to track which females they were grouping with over time. Network models reveal the patterns of all of the females in the population to visualize and quantify their social patterns. When possible, I also collected dung from individuals in order to examine the genetic relatedness patterns between individuals in relation to their social patterns. I found that forest elephants do associate in larger groups than are represented through observational patterns alone and groups appear to be based on matrilines, however the associations that take place are not nearly as complex than those that occur in both African savanna elephants and Asian elephants. Also, not all individuals were found to be in groups. Therefore forest elephants, do seem to differ in sociality and have the least stable groups and the least number of associates. Results are in review.
Given their different ecosystems, it is interesting to compare forest to savanna elephants to aid in understanding how ecological factors favor or prevent the formation of higher-level social structure.