Open Access Te Herenga Waka-Victoria University of Wellington
2 files

Resource-use and recursion by a mega-herbivore Elephas maximus borneensis

posted on 2024-06-10, 21:40 authored by English, Megan

This thesis examines the ecology of elephants (Elephas maximus borneensis) that inhabit the Lower Kinabatangan region of Sabah, Malaysia. My research focused on improving our understanding of their habitat use and food-plant preference over varying spatial and temporal scales, and tested recursion hypotheses. Recursion, the reuse of sites or plants over time, has rarely been explored in wild animals. Studies of recursion promote understanding of species ecology as they explore temporal variation in resource-use. Recursion by herbivores may be a foraging strategy for optimising resource-use by returning to sites to coincide with plant recovery. A review of the recursion literature revealed that previous studies had not considered recursion that leads from foraging theory; this informed the research and the design of the chapters on recursion at two spatial scales – site and plant. The review also demonstrated the need to integrate the large amount of research on recursion-like processes with the new research topic of recursion. Such processes include site reuse associated with spatial memory, resource recovery and foraging site-fidelity. The scarcity of studies of these topics in large, wild herbivores was also evident.  I chose to investigate recursion ecology in the Bornean elephant because this provided an opportunity to test hypotheses for repeated resource-use, and to improve our understanding of resource ecology for a mega-herbivore. I expected recursion to occur less frequently in elephants compared with smaller herbivores. Mega-herbivores have a reduced requirement for high quality food. They should also consume more resources per visit, resulting in more time needed for resources to recover and therefore less frequent recursions. However, elephants have a more highly developed spatio-temporal memory than other herbivores and therefore may have greater ability to return more often to profitable foraging sites and plants.  To investigate elephant habitat use I first characterised the habitat types in the Lower Kinabatangan Wildlife Sanctuary (LKWS) based on their floristic composition and physical characteristics. Elephant habitat-use was measured using indirect observations of feed sign along ten 1 km long and 4 m wide strip-transects randomly placed throughout the LKWS. Elephants exhibited a strong preference for open grass areas along forest margins and avoided swamps and recently logged or cultivated habitat. Two of the most common habitats, lowland mixed dipterocarp and semi-inundated forest, were neither selected nor avoided. The habitat types identified and their use by elephants underpinned the subsequent chapters on recursion and elephant food-plant preferences.  I investigated recursion to foraging sites using GPS collars to track the two main elephant herds in the LKWS in order to examine their behaviour and sample sites. Recursion was common, occurring at 48 of 87 foraging sites, within 48 hours and between 151-250 days. This indicated that elephant foraging strategies involve site sampling and timing of returns to coincide with some periodicity in site quality. I also found that recursion occurred if the site had previously been occupied for longer periods than sites receiving no recursions. The number of days that had passed between the first visit and recursion was also positively correlated with time spent at the recursion site. Sites that received most recursions were within the elephants’ preferred habitat; that is, open grass areas along forest margins. These findings indicate that recursion occurs for the repeated exploitation of higher quality foraging sites and is perhaps timed for the recovery of their food plants.  The hypothesis that plant recovery rates influence recursion periodicity had not been tested previously in wild populations. The growth of new shoots on plants from 30 species that were previously selected for feeding by elephants were measured each month for 9 months, or until they were re-browsed by elephants, to learn if plant recovery rates influence recursion time. Recursion to grasses was found to coincide with full recovery but the elephants prematurely browsed other plant types. This suggests that elephant foraging strategies influence vegetation community structuring and may maintain or enhance grass patches. My results from Chapters 3-5 demonstrate recursion at two spatial scales, site and plant, and indicate that the elephants forage optimally in their preferred habitat.  It was also necessary to understand the proportion of grass and browse in the elephant diet and what influences the selection of browse species. Food-plant use and availability analyses found that contrary to what was expected of a generalist herbivore, plants were not selected by elephants in proportion to their availability. The finding that grasses form a major and highly preferred part of Bornean elephant diet was unexpected because they had been regarded as a forest-feeding species. One hundred and eighty-two plants were eaten and 185 plants from 18 species were measured for species availability along 12 transects. I identified a preference for grass rather than browse species despite grasses being less abundant, confirming the importance of grass and grass patches – two spatial scales – to elephants in the LKWS. Previous work has implicated the size and vigour of plants as important in herbivore food-plant selection. I found that elephant browse preferences were not influenced by plant vigour and plant size. Bornean elephant foraging strategies are therefore primarily focussed on the optimum harvesting of grass in discrete patches within the forest and the riparian zone.  To manage this population of elephants it is important to understand their relationship with their food resources. My study has contributed in this regard in ways that are of general importance to understanding elephants and other mega-herbivores, as well as in ways that are of specific importance to the conservation of Bornean elephants in the LKWS. Firstly, in order to meet the elephant population’s dietary needs, access to open grass areas must be maintained. Forest rehabilitation projects should incorporate these areas as part of their landscape management strategy. For example, corridors should be designed and placed to assist the movement of elephants amongst preferred grassy foraging sites.  Secondly, this study provides scientific support for effective ways to identify profitable foraging areas for wild herbivores across a variety of landscapes. Studies of recursion and the characters of sites and plants that receive repeated visits provide more direct and rapid ways of identifying what is important to herbivore populations.  Moreover, if the elephant, the largest land mammal and herbivore, is so elaborately and concurrently recursive at different spatial and temporal scales, then recursion is likely to be widespread amongst other herbivores. I therefore recommend further examination of recursion behaviour across a wider range of wild herbivores. Advances in our understanding of herbivore ecology and our ability to conserve and manage wildlife habitat will require a direct refocus on repeated resource-use and will depend on redesigning studies to consider the scales at which recursion occurs.


Copyright Date


Date of Award



Te Herenga Waka—Victoria University of Wellington

Rights License

Author Retains Copyright

Degree Discipline

Ecology and Biodiversity

Degree Grantor

Te Herenga Waka—Victoria University of Wellington

Degree Level


Degree Name

Doctor of Philosophy

ANZSRC Type Of Activity code


Victoria University of Wellington Item Type

Awarded Doctoral Thesis



Victoria University of Wellington School

School of Biological Sciences


Linklater, Wayne; Gillespie, Graeme; Goossens, Benoit