Living in groups may have several advantages such as protection from predators or help in raising young ones, but living in groups also has one major cost – competition for food. Fluid grouping, wherein group size and composition change over short time periods – a few hours or days – in response to varying resource availability is one possible solution to competition for food within groups.
Large groups may split into small groups in response to limited resources, and small groups may fuse together to form large groups when resources are plentiful. Societies that have nonrandom groupings of individuals but show temporal changes in group sizes and compositions have been termed fission-fusion societies and include chimpanzees, lions, spotted hyaenas, dolphins, and humans.
That fission-fusion dynamic is a response to changing resources had been suggested in spider monkeys, chimpanzees, brown hyaenas, spotted hyaenas, and African savannah elephants. However, there had been very little previous work on the social organization of Asian elephants, which occupy the same ecological niche as African elephants but diverged from their African cousins over 7 million years ago.
As elephants inhabit ecologically diverse habitats, from savannahs to tropical rain forests to deserts, they are useful organisms to study in order to understand the effect of ecology on social structure. African savannah elephants (Loxodonta africana) were known to show a multi-tiered, fission-fusion society, where clans are the largest stable social units and include bond groups that include family groups.
We had found recently (Nandini et al. 2018) that clans – communities within which females associate with one another and across which they do not usually have amicable associations – were the most inclusive units of social organization in the Asian elephant population that we studied. However, there was greater fluidity in social structure, and clans did not comprise strict bond groups that comprised strict family groups, unlike the case of African savannah elephants.
Individual females within clans in our study population could form a group with almost any other female in her clan, although not entirely randomly, through fission-fusion dynamics. A “group” here is a set of associating elephants (that feed and move in a coordinated manner, usually within 50-100 m of one another) and is usually a subset of the clan based on how much food is available in a particular site for the elephants to feed upon.
The clan itself can be detected using long-term data on associations and network statistics. While a clan may include over 30 adult females, the average group comprises only 2-3 adult females (and their offspring). The difference in social structure between the Asian elephant population we studied and the African savannah elephant population was possibly due to the average group size in the latter being larger than the average group size in the former because of more abundant food resources in the African savannah.
In this paper (Nandini et al. 2017), we examined whether changing resource availability, in the form of seasonality, affected group size and social structure of our study population, as would be expected in a fission-fusion society. The population of Asian elephants that we study is in the Nagarahole and Bandipur National Parks (Kabini population), in southern India. We collected field data from March 2009 to July 2014, spanning 878 field days, as part of the long-term Kabini Elephant Project. The two parks are separated by the River Kabini. A dam on the river results in parts of the forest on both sides being submerged during the wet season, and, being exposed, forming large grassy areas where elephants congregate during the dry season. Elephant ranging is affected seasonally, with lower density around the reservoir during the wet season compared to the dry season.
Because of locally abundant resources (water and grass), we expected to find larger group sizes and more and stronger associations between females during the dry season than during the wet season, when food would be more dispersed. Thus, fission-fusion dynamics might then help regroup females in response to food resource availability.
Individual Asian elephants can be identified using various natural physical characteristics just the way humans can be identified. Females were identified based on a combination of physical characteristics such as ear folds and length, ear depigmentation, nicks, tears, and holes in the ears, the shape of the back, tail length, tail brush shape, and the presence of tushes, warts, and so on. We collected data on who was seen with whom at different points in time. In order to compare group sizes between the two seasons, one can simply list out the sizes of all the groups seen in each season, irrespective of which clan the females of the group belong to, or separate them out by clan and then look for differences across seasons.
When we did the first, dry season group sizes were slightly but significantly larger than the wet season group sizes, and the level of associations was slightly higher in the wet season than in the dry season. However, pooling all the data irrespective of clans can bias the results towards clans that are commonly sighted. Therefore, we analyzed the data by clan and found that there was no seasonal difference in group sizes and no consistent difference across clans in the level of association amongst females or other social network properties. Since most previous studies on Asian elephants were not based on identified individuals, their findings of seasonal differences in group sizes may need to be re-examined.
Interestingly, we found that group sizes formed by the females were not different on average whether the clans the females belonged to were large or small. Females formed small groups of 2 to 3 females on average even when their clan sizes were large. This, therefore, indicated that there was some constraint on group size, possibly from food limiting the number of animals that can feed at a site. This constraint on group size resulted in reduced associations within larger clans, since all the females would not be able to meet one another as much as females in smaller clans could do while maintaining a small group size. There were some preferred associations within clans, but there was also a turnover in who associated with whom so that different females within the clan could be met. Thus, we found that small groups of varying composition were being formed instead of small groups of fixed composition.
Unlike in several other species, we found that fission-fusion dynamics in the Kabini elephant population was not a means to change group size in response to varying resources, but primarily enabled females to meet their clan mates while maintaining small group sizes. This suggests a benefit to extended associations within the clan and it would be interesting to find out what those benefits are. Allomothering is a possibility that comes to mind and is being examined. Benefits may also include an exchange of social and ecological knowledge.
Since we found that group size and social structure were not significantly affected by seasonality, although elephant movement patterns and habitat usage were, it would also be interesting to examine the actual abundance and dispersion of elephant food plants in the area to find out if they change seasonally.
These findings are described in the article entitled Seasonal variation in female Asian elephant social structure in Nagarahole-Bandipur, southern India, recently published in the journal Animal Behaviour. This work was conducted by S. Nandini, P. Keerthipriya, and T.N.C. Vidya from the Jawaharlal Nehru Centre for Advanced Scientific Research (JNCASR).
References:
- Nandini S, Keerthipriya P, Vidya TNC. 2018. Group size differences may mask underlying similarities in social structure: a comparison of female elephant societies. Behavioral Ecology 29(1): 145-159.
- Nandini S, Keerthipriya P, Vidya TNC. 2017. Seasonal variation in female Asian elephant social structure in Nagarahole-Bandipur, southern India. Animal Behaviour 134: 135-145. DOI 10.1016/j.anbehav.2017.10.012.