Forest elephants in a rainforest fragment: preliminary findings from a wildlife conservation project in southern Ghana

1. What they did

The Kakum and Assin-Attandanso Forest Reserves in southern Ghana hold a small population of African forest elephants, living in a patch of logged rainforest surrounded by farms. During a three-month survey of wildlife in the reserves (July to September 1990), the authors gathered information on how many elephants there were, how they lived in groups, what they ate, and the damage they did to farmers’ crops.

Because the forest was so dense, they never saw an elephant. Everything was based on spoor:

• Surveys. A helicopter flight showed the state of the forest cover. Most work was on foot along existing paths (boundary lines, new hiking trails, hunters’ paths, old logging roads and game trails), plus about 20 km of motorable logging tracks.
• Footprints. Clear, fresh forefoot prints were measured across two perpendicular dimensions, sketched and photographed where possible, and any peculiarities in foot or toenail shape and footpad ridges were noted. These were used to tell elephants apart and judge their age and size class.
• Dung. Fresh, intact dung boli were measured for length and diameter, and their texture, consistency and contents noted, again to judge age and size class and to tell individuals apart.
• Crop raids. Reports of crop raiding were followed up as soon as possible; spoor and crop damage were recorded at the farm, and the elephants were tracked to find the routes they used to leave and enter the reserve.

Group size, age and composition were estimated only from fresh spoor clearly left in a single event, not from old or scattered spoor. Several dung piles of similar size at one site counted as one elephant if other signs (heavy browsing, flattened vegetation where an elephant had lain down, or churned ground) showed it had stayed half an hour or more.

2. What they found

Population

• Spoor showed at least 25 to 30 different elephants in focal survey areas covering about 15% of the reserves.
• A conservative estimate was 100 elephants and a liberal one up to 200; the authors’ working estimate was 100 to 150, about 0.3 to 0.4 elephants per km² over 347 km². Densities reported elsewhere in West Africa were 0.33 per km² in Bia National Park, Ghana, and 0.23 per km² in Tai National Park, Ivory Coast.
• Elephants avoided the western “panhandle” of Kakum, the western half of Assin-Attandanso and some higher ground, probably because of human disturbance and patchy food.
• The population appears completely cut off from other elephant populations. Elephants no longer live in the neighbouring Pra Suhien Reserve.

Groups

• Groups were small: 2 to 6 elephants, typically 3 or 4 (n = 9). Pairs were adults only, and spoor of lone elephants was common.
• Spoor of herds with small calves (under about 8 to 10 years) was found only well inside the reserves. Groups with young calves rarely if ever left the reserves to raid farms.

Trails and travel

• There were many established elephant trails. These often followed old skid trails and logging tracks, and elephants had begun to use new hiking trails too.
• Raiding elephants often walked along the cleared 1 to 2 m wide reserve boundary lines until they reached farm plots.

Feeding

• Clear signs of browsing were often missing even where elephants were very active. No serious bark stripping or damage to mature trees was found.
• Browsing was recorded most often on the umbrella tree and the raffia palm. Umbrella trees of 10 to 30 cm diameter had been pushed over and stripped of leaves. Raffia swamps were used for drinking and mud bathing.
• Pieces of the hard shell of Tieghemella heckelii fruit were common in dung throughout the study.

Crop raiding

• Elephants caused severe, local damage to maize, cassava, banana and plantain, cocoyam and papaya, and trampling often did as much damage as feeding.
• They rarely went more than 100 to 200 m beyond the forest edge, even under cover of cocoa or dense bush fallow.
• They did not damage oil palm, orange or cocoa. They avoided a large oil palm plantation and sometimes walked through cocoa groves without feeding on them, unlike elephants elsewhere in Ghana.
• Overall, birds and other mammals probably caused more crop loss than elephants, but elephants caused the greatest losses to individual farmers.
• Reports of raids dropped sharply from mid-August to mid-September, at the second fruiting peak of Tieghemella heckelii.

Poaching

Elephants were sometimes shot for ivory until about 1970, and some were killed in crop-raiding control in the past (now stopped). No elephants were known to have been poached recently; local poaching was limited to monkeys, rodents and small antelopes.

3. What they concluded

• The authors suggest raiding dropped in late August and September because elephants moved into the interior to feed on fruiting Tieghemella heckelii. Logging may have reduced these trees in the south and west of Kakum, or there may be a natural gradient; data to test this were lacking.
• Selective logging now takes about fifty tree species in Ghana, compared with one fifty years ago. Many are important fruit trees for elephants, such as Tieghemella heckelii, and the loss of elephants as seed dispersers may in turn affect regrowth of valuable timber trees.
• Forest elephants favour plants in forest gaps, because the food is within reach and these plants have fewer chemical and physical defences.
• West African rainforest elephants are browsers and fruit eaters with almost no grass in their diet. Because their food is scattered, the authors predict they will live at lower densities and in smaller groups than elephants in grassland and savanna mosaics, and never form large herds (more than 25). Their seasonal movements are most likely tied to fruit, not water or grass.
• The authors consider their spoor method practical for these conditions, while noting its considerable potential for subjective error.

4. How tracking was or was not used

The whole study rests on tracks and signs, since no elephant was seen. The authors used fresh footprints and dung to count individuals, estimate group size and age classes, and recognise individual elephants from footprint peculiarities and dung characteristics. They used signs such as browsing damage, flattened vegetation and disturbed ground to judge how long an elephant had stayed at a site, and judged spoor freshness so that only fresh, single-event spoor was used for group estimates. After crop raids they trailed the elephants to find the routes they used to leave and enter the reserve. Elephant trails, feeding signs on plants and crops, and fruit remains in dung gave information on movement and diet.

5. Tracker or observer certification

Not reported. The spoor data were collected by the authors; no tracking experience or evaluation is described, although the population and group size estimates depend entirely on spoor interpretation. Game Protection Guards and Game Scouts were being trained in the techniques, which the authors found were readily understood and learned by staff with limited formal education.

6. Other technologies, skills or innovations

• A helicopter survey to assess forest cover (aerial counts of elephants were impossible under the dense canopy).
• Photographs and sketches of footprints.

7. Implications for conservation (from the paper)

• Ghana’s forest elephant populations are small, isolated and confined to a few parks and reserves. Managing national parks and neighbouring forest reserves as large, joined conservation units will be needed for them to survive.
• Forest management plans should recognise that timber trees provide food and habitat for wildlife. Removing mature fruiting trees such as Tieghemella heckelii may seriously reduce the seasonal food elephants need.
• Elephant poaching was not a problem at the time, but could become a threat.