Project Update: September 2007
Project Web Page: www.elephantlisteningproject.org

In June 2006, the Elephant Listening Project began the groundwork for research in Gabon with a reconnaissance trip by Karyn Rode, funded by the Cleveland Metroparks Zoo. It was during this initial trip that key contacts were established, potential research sites identified, and some of the more difficult logistical issues determined. This initial grant was instrumental in attracting additional funding for full implementation of the proposed research. Based on experience and information gained by Karyn, we were able to make modifications to some of the equipment, and were prepared for many of the logistical challenges that we encountered during implementation. In June and July 2007, in collaboration with the Wildlife Conservation Society (WCS-Gabon), we initiated an ambitious research program. Research activities are focused in two areas of interest, supported by a research grant from the U.S. Fish and Wildlife Service. Considerable logistical assistance has come from WCS-Gabon, and 'Projet Protection des Gorilles' (in Plateau Batéké N.P.). This report summarizes activities in each of the targeted areas of interest.

Use of Forest Clearings (Bias) by Forest Elephants
The 'Bai' study aims to better understand the factors that affect forest elephant use of bais and to quantify the diurnal and seasonal pattern of use by elephants. At a sample of bais, we will use acoustic recordings to quantify the activity of elephants as well as any active hunting. In addition, we are measuring rainfall and analyzing the mineral content of water and soil resources at each location. Other external factors that might affect bai use include the intensity of hunting pressure in the vicinity (both current and recent historical), frequency of visits by tourists, and potentially the proximity of logging operations. Data on each of these factors will be obtained either directly (detected on the ARUs or from records of visitation at established ecotourism destinations, e.g. Langoué Bai) or indirectly (interviews, signs of activity around bais detected when servicing the ARUs or by other researchers).

Autonomous Recording Units (ARUs) were placed at six different bais in July 2007. We attempted to assemble a sample that would give reasonable variation on the factors of interest, and this appears to have been achieved. Each of the study sites is described below.

• Ivindo National Park, established in 2002, encompasses 3000 km2 of tropical forest in central Gabon. Ivindo has no roads and no resident human population. Human density in surrounding areas is low and the greatest threats within and surrounding the park consist of poaching and habitat destruction, both of which are related to new logging concessions and associated road networks. Langoué Bai is an approximately 400 m x 900 m clearing in the southern portion of the park: the bai has been monitored by the Wildlife Conservation Society (WCS) since 2001. Ecotourism facilities are being expanded at Langoué, but currently include a functioning camp approximately 4 km from the bai and a pair of observation platforms.
• Plateau Batéké National Park, established in 2002, encompasses 2050 km2 of savannah and gallery forest in southeastern Gabon. Batéké has very few roads, little tourist activity, and no resident human population. Although illegal hunting has recently been brought under control through successful anti-poaching efforts, there is a history of extreme hunting pressure throughout the park, including in the vicinity of Jobo Bai, near the international boundary with Republic of Congo. Some hunting pressure continues, particularly in these most remote corners of the National Park. Jobo Bai is approximately 100m x 100m and is receiving increased attention from the WCS because of its proximity to the new trans-boundary park system established between Gabon and Congo, and the two or three bais situated in Congo, all within 20 km of Jobo Bai.
• CEB (Compagnie Equatorialle de Bois) is a French logging operation with huge holdings north and east of Lastourville, in central Gabon. They are a leader in the implementation of 'sustainable' harvesting methods and in seeking 'green' certification for their wood products. Several bais have been identified within their logging concession, and the company has voluntarily created a buffer zone around each to encourage continued use by forest elephants. Two of the bais included in this monitoring study are 25 km from Langoué Bai in Ivindo N. P. A third bai is more than 105 km from any protected area and will provide an interesting contrast to the bais close to, and within, national parks.
• Djidji Bai is a small forest clearing to the east of Ivindo N. P., within the logging concession of the Rougier Logging Company. Like CEB, this company has also expressed interest in further exploring practices that would reduce the impact of logging activities on wildlife and habitats.

Effects of Human Activity and Environmental Variables - In order to develop a protocol for acoustically monitoring elephant populations from activity at bais, information is needed on the human and environmental variables that affect bai use. Human and environmental factors quantified at all bais will include: 1) gunshot occurrence, 2) rainfall, 3) surface area of the bai, 4) number of waterholes or soil excavation sites, 5) mineral availability in bai water (e.g. sodium, calcium, etc.), 6) hours of patrols/monitoring at each bai and 7) proximity to villages and roads. Bais with heavy hunting pressure are thought to alter elephant use patterns by increasing night-time use ((Maisels, 2002); S. Blake pers. comm.), and this study will also examine whether consistent eco-tourism activity alters elephant activity in a similar manner. In areas where long-term monitoring has occurred and poaching has increased significantly, elephants became increasingly nocturnal (e.g., Odzala National Park; (Maisels, 1996).

Autonomous Recording Units at the bais will record for 60-90 days on a set of batteries. Although only a single ARU was placed at each bai, each unit will record from an area of 1.5-3.5 km2, depending on the signal/noise ratio. Two bais in the CEB forestry concession had fast-flowing streams running through them, decreasing the signal/noise ratio and thus reducing the distance at which we can detect elephant vocalizations. Nonetheless, a detection area of 1.5 km2 will still easily cover all elephant activity in the vicinity of the bai. Between October 2007 and early January 2008 the acoustic data will be recovered from ARUs at each bai and the batteries replaced for a further three months of recording. Funding has been requested from the International Elephant Foundation to extend this study for an additional three month period of recording, which would provide data from all seasons of the year.

Call Rates vs. Elephant Numbers- Validation at Low Density
Detailed behavior observations are planned at Langoué Bai beginning in October 2007. One objective of this study is to validate the algorithm relating calling rates to elephant numbers derived from work at Dzanga-Sanga Bai in the Central African Republic (Thompson et al., submitted). Dzanga Bai has very high densities of forest elephants relative to densities observed at other bais in the Congo Basin, and certainly higher than the size of groups moving through the forest. Observations of elephant groups in the forest suggest an average group size of 3-4 animals (White et al., 1993; Morgan & Lee, 2007). The variance in call rate at low densities has not been well characterized but is critical to using acoustic monitoring methods to estimate elephant abundance in the forest at large. Visitation by elephants at Langoué Bai frequently involves only a single family group, but sometimes involves 20-30 animals (4-6 family units). This will allow us to characterize the relationship between calling rate and elephant numbers at low densities.

In October, 2007, an array of 6 ARUs will be established at Langoué Bai, each ARU equipped with a GPS receiver to provide highly accurate time information on the acoustic recordings. This will allow us to localize the source of each recorded vocalization. Observations will be made for 6-8 weeks, including regular scan samples of the number, sex, and age class of elephants in the clearing, as well as video records of behaviors. Observations will be concentrated during daylight hours, but nighttime observations will be made when possible on clear moonlit nights. We will test the use of night-vision equipment to extend the days on which we can make nocturnal observations.

LITERATURE CITED:

MAISELS, F. (1996). Synthesis of information concerning the Parc National d'Odzala, Congo. -- In. Projet ecofac-Composante Congo.

-. (2002). Long term monitoring, research, and protection of elephants in the tri-national region (Nouabale-Ndoki Nation Park - Republic of Congo; Dzanga Sangha Reserve - Central African Republic; Lobeke Reserve - Cameroon): Final Report September 200-November 2001. -- In. U.S. Fish and Wildlife Service.

MORGAN, B. J. & LEE, P. C. (2007). Forest elephant group composition, frugivory and coastal use in the Réserve de Faune du Petit Loango, Gabon. -- African Journal of Ecology 'OnlineEarly',???-???

THOMPSON, M., PAYNE, K. B., SCHWAGER, S. J. & TURKALO, A. K. (submitted). Acoustic estimation of wildlife abundance part I: Methodology for African forest elephants. --.

WHITE, L. J. T., TUTIN, C. E. G. & FERNANDEZ, M. (1993). Group Composition and Diet of Forest Elephants, Loxodonta-Africana-Cyclotis Matschie 1900, in the Lope Reserve, Gabon. -- African Journal of Ecology 31, 181-199.