Hippos of the Black Volta River




Don't be scared. I'm Nicky Green and I'll protect you from the hippos.

3. Estimating the Hippopotamus Population in Bui National Park
Daniel Bennett, Dept. Zoology, University of Aberdeen, Scotland.

The hippo (Hippopotamus amphibius) is considered widespread and secure by the IUCN (Eltringham in Oliver 1993). However populations in West Africa have been in decline for at least a hundred years and only 7,000 animals are thought to be left in the entire subcontinent, compared with 150,000 animals in Eastern and Southern Africa. In Ghana the hippo is classed as endangered and is afforded complete protection under national law. It was once widespread throughout the Volta system but the only surviving animals are found in the Black Volta River from Bui National Park upstream to the border with Burkina Faso. Within this area two populations have been identified, one within Bui National Park and another around Wichan, near Wa in the north of the country. Neither population is well defined and only in the former area is the hippos' habitat afforded protection.

The aim of this study was to make an estimate of the hippopotamus population within Bui National Park and to investigate interactions between hippos and people in the area. The work was conducted by members of the Aberdeen University Black Volta '97 Expedition between June and July 1997.

Proposed methodology assumed;

1. That hippos would be in the water during the day and out of the water during the night.

2. That individual animals could be identified and subsequently recognised by taking photographs of the creases around the eye with a 600mm lens, having been enticed to lift their heads above the water using the mirror technique described by Kardstad & Hudson (1984) and Norton (1988).

3. That it would be possible to survey the entire length of the Black Volta River within Bui National Park during the study.

In practice, none of these assumptions were valid. Heavy rains just prior to our visit caused the level of the river to rise and much of the riverine forest was flooded. As a result the river was twice as wide as had been expected and access to the riverbank was impossible except in a few places. Cloud cover prevented the use of the mirror technique, and the 600mm lens used to photograph individual hippos did not provide sufficient magnification to reveal eye creases. Subsequent investigations revealed that the movement patterns of individual hippos could not be predicted from the time of day, with animals found in the water at all hours of the night, and on the riverbank at most times of the day. Finally, the flooded condition of the forest and the swollen tributaries made walking along the river course very slow and precluded studies in central and northern sectors of the river. On the recommendation of Ghana Wildlife Department the aims of the hippo survey were altered to make an estimate of hippo numbers in the southern portion of Bui National Park, from Batoo village upstream as far as Bope, which includes a large part of the area that will be affected by the Bui Dam (Map 1). Traditional local canoes were used for the survey, each containing two local fishermen and two expedition members. Areas of the river were surveyed daily (except Sundays) throughout July. Estimates were made by recording numbers of heads visible every 15 seconds for at least five minutes, and using the maximum value obtained. Usually fishermen’s’ estimates of total numbers were higher than those of researchers, but the lower value was used. Sexes were assigned on an arbitrary basis with the largest animals classified as males and smaller adults as females. Babies were easy to identify by their very small size. Juveniles were animals other than babies whose size suggested they had not reached sexual maturity. In addition some groups were seen from the riverbank.

Tables 3.1 and 3.2 give locations, numbers and age classes of hippos seen during the study.
TABLE 3.1. Number and location of hippo groups seen from the river

In total 31 hippo sightings were made 23 from the river and eight from the riverbank. Hippos were most commonly seen in groups of 1-9 animals (mean 5, standard deviation 2.7, median 5, interquartile range 3-7.8). These groups included adult males, adult females, juveniles and babies. All groups of four or more hippos included babies. Aggregations were most common around the outlets of three tributary steams (Lubia, Coco and Japoli), but we could find no evidence that hippos ever enter these tributaries. In the case of the Coco and Lubia streams, dense vegetation make it unlikely that they are frequented by hippos. Outlets of tributaries therefore provide good vantage points from which to estimate hippo numbers during riverbank counts.
TABLE 3.2. Number and location of hippo groups seen from the riverbank.

Because few, if any, individuals could be recognised it was not possible to calculate the total number of animals present in this portion of the river. The minimum number present is derived from riverbank sightings on 24 July, when 15 individuals were seen at the same time. However, this is likely to be an underestimate of total numbers, because the hippos seen on this occasion probably represent two groups from the northern sectors of the study area. Another group of at least nine animals occupies the rocky sectors A & B and occasionally stray towards Batoo Village.

Observations suggest the following age structure for the hippo groups in this part of the Black Volta:

TABLE 3.3 Population structure of hippos

A male seen on 19, 21, 28 & 29th July was the only individual we could identify with certainty, because of his large size and dark colouration. This animal was associated with at least one group (Group 3) most of the time, but left the group towards evening, presumably to forage on his own.

Length of river sections used for survey (estimated from the map) averaged 2.3km  (standard deviation 0.47, range 1.8-3.0). The mean number of hippos seen per section of the river (Table 3.3) suggests a  mean density of  2.11 (+/-0.17) hippos per km of river. Extrapolating this for the whole park suggests a total population of  140-164 animals. This number is likely to be an underestimate for three main reasons.

1.  Length of the river in the wet season is underestimated by the map.
2.  "Maximum counts" were based on the largest number of heads seen at any one time. Because large sections of the river were inaccessible to canoes, hippo groups could have been missed. Fishermen’s’ estimates of hippo group sizes were usually higher than ours by one or two animals. They estimated that the group seen on 24th July included over 20 animals. They did not agree with any of our attempts to identify individuals (except the male mentioned above) and did not agree with our conclusion that the river sector contained three groups of animals. According to fishermen, the hippos are very mobile and will travel large distances along the river in a few hours. The hippos seen in the study site could therefore have come from anywhere.
3.  The density of hippopotamus probably increase further into the park. Casual observations further upstream (see below) suggest that there many hippos around Abru Bhunu. The many thickly forested islands in this area of the park could only be surveyed in dry season conditions.

AREA     A B C D E
LENGTH (KM) 1.8 3.0 2.0 2.3 2.5
MEAN NO. HIPPOS 4 6.5 4 5 4.7
STANDARD DEVIATION 4.24 1.52 1.41 2.62 4.03
HIPPOS PER KM 2.29 2.17 2.0 2.22 1.88
TABLE 3.4. Distribution of hippos in study area.

FIGURE 3.1. Group size for hippos seen from the river


Because of the lack of individual identification, little can be said about daily movement patterns of hippo groups. Local fishermen claimed that areas of rapids (even very fast white water such as Fagoo) present no obstacle to hippos, who often simply walk through these areas. The main obstacle to hippo movement is dense vegetation, which may limit their movements away from the river to areas of more open forest and grassland.

Of nine foraging zones identified as "hippo lawns" eight were on the western riverbank (Map 6)  Area of these zones was estimated by eye. The steep topography of the Northern bank must be at least partly responsible for this trend. It is presumed that steep hillsides are not suitable for hippo foraging, regardless of their vegetation types.

Observations on hippos in late afternoon and early evening suggest that large males leave aggregations before dusk and presumably forage away from the rest of the group. Based on the numbers of hippos tracks leading to and from the river, most animals in one group must leave the water by the same route, although thereafter they may disperse to forage.  Following tracks made by foraging hippos was relatively easy, but because animals tend to use previously established tracks it was not possible to estimate mean nightly distance travelled. Some tracks indicated that a single large individual (presumably the mother) foraged with a very small individual at least twice over a ten day period in the vicinity of the Coco Stream.

Attempts at observing hippo foraging behaviour were not successful. Nocturnal censusing of the type described by Ogen-Odoi and Dilworth (1987) is inappropriate for this area because hippos are not accustomed to vehicles (or engines of any type). Two tree houses were built in an area that appeared to be a major foraging ground for hippos. The buildings were constructed over six days and thereafter were manned nightly. No hippos were seen from the treehouses, although on some nights they could be heard nearby,  in or at the edge of, the river. It is likely that the disturbance caused by treehouse construction deterred the hippos from foraging in this area.

Because it was not possible to deduce how many animals had been feeding in any location, measurements of the total feeding area (i.e. the area of "hippolawns") are not meaningful. Attempts to measure footprint size and therefore assign spoor to individual animals were not successful, probably because of differential shrinkage of footprints depending on the moisture content of the mud.

Two excursions on foot further upstream (to Abru Bhunu) suggest that hippopotamus density increases further into the park This is based on the higher incidence of tracks seen and the fact that wherever the riverbank was accessible hippos could be seen or heard. Although it was not possible to make estimates of numbers in these areas, desnities appeared to be much higher than in surveyed areas of the river.

During riverbank surveys it was noted that hippos tended to remain in one place for at least several hours at a time. They are much easier to see when they are resting than when they are travelling through the water. In the latter case their presence is usually revealed only as a distinctive zigzagged wake.


Above & below: The hippos of Bui photographed with a 600mm lens on 800ASA film.


Interactions between people and hippopotamus
Hippopotamus in Bui National Park have been responsible for at least one human death in recent years. The incident involved a local fisherman who was killed at night when a hippopotamus overturned his canoe. Although his friends and family were adamant that he was catching fish, wildlife staff claim that he was trying to catch the hippo. Conversely, a number of hippopotamus have been poached in the park. Because few, if any, of the local people have access to the high velocity weapons needed to kill hippos safely from a distance, hippo hunting remains a very dangerous occupation. We found no evidence that pitfall traps are used to catch hippos in this area. Local people freely admit that hippo is their favourite meat, but insist that the only ones they have tasted had strayed outside the park boundary and were killed because they were a menace to local people and their crops. The low population density of both people and hippopotamus in this area make conflict between them relatively uncommon compared with Eastern Africa (e.g. Mkanda 1994). However, because of the very poor nature of the soils in this region, formation of  buffer zones between the river bank and agricultural lands to reduce human/hippo interactions do not appear to be a very desirable option; the only good farmland is within, or adjacent to, the floodplain of the river.  In general, hippos rarely stray outwith the southern boundary of the park and this greatly reduces the number of potentially antagonistic encounters between them. A major cause for future concern is that engineering activity within the park or pressure from an increasing hippo population will cause animals to spill over the southern border of the park, where they will undoubtedly cause problems.

Assuming that hippo densities throughout the park are at least as high as those recorded by us in the southern sector, the hippo population at Bui appears secure, in the short term at least, and, based on the high incidence of juvenile animals, numbers may be increasing. Thus Faber's (1996) estimate of 18 individuals within the park (and 27 throughout the Ghanaian portion of the Black Volta) is a gross underestimate, and highlights the dangers of using cursory observations as a basis for estimating animal populations, even when the animals are as large and conspicuous as the hippopotamus. However, compared to hippopotamus populations surveyed in eastern Africa, the numbers of hippos in the Black Volta is very small (e.g. Norton 1988; Bhima 1996; Karstad & Hudson 1984). Large increases in hippo density may have the effect of displacing individuals so that more animals cross the southern park boundary and thus increase the number of encounters between humans and hippos, which is clearly undesirable.

The hippopotamus population at Bui numbers at least 140 animals and is probably in the range of 250-350. There are grounds for supposing that the population is currently growing. Hippos appear unable to survive outside the park because they are not tolerated by local people.

The high incidence of hippo observations, both on the riverbank and by canoe, indicate that the Bui hippo population has the potential to be an important tourist attraction. However, the dilapidated state of the park's infrastructure, its remote location and lack of even basic facilities make it unlikely that the park could cope with large influxes of tourists without substantial investment. We believe that the park could attract significant numbers of more adventurous tourists, and that they could be catered for with very rudimentary facilities, most of which already exist. The main priority of such a plan must be the safety of tourists and their guides. For riverbank and treehouse observations this can be accomplished simply by ensuring that each group is accompanied by a Wildlife Ranger trained in the use of high-velocity firearms. For canoe trips along the Black Volta life jackets should be regarded as essential. The facility that is most obviously lacking at Bui is a vehicle that can transport paying visitors on demand. If the number of tourists rises in the way we anticipate, such a vehicle would pay for itself within 18 months.

One other hippopotamus population still exists in Ghana, at Wichan, near Wa. Population is estimated to be around 100 animals (P.Choribe, pers. comm.). This population is at present in an unprotected area, and the local districts assemblies are keen that the Wildlife Department takes some action to reduce conflict between hippos and people.

Possible effects of the Bui Hydro Electric Project
Whether hippos will continue to inhabit this area after the construction of Bui Dam probably depends largely on whether suitable foraging areas will exist or can be created close to the edges of Bui Lake. The fact that all foraging areas identified by us were in riverine forest habitats suggests that it may be necessary to create suitable foraging areas prior to flooding in order to maintain a hippo population at Bui. However O’Connor & Campbell (1986) note that hippos feed closer to the riverbank during the wet season in Zimbabwe, and that they favour alluvial areas. It may be that the dry season feeding grounds of the Bui hippos are much larger than those used in June and July.

Our limited experience supports that of other workers who stress the unpredictability of hippo behaviour. On several occasions lone hippos were encountered by boats in areas where they were not expected (by local fishermen). In no cases were the hippos aggressive towards canoes, but care was taken to stay as far away from the animals as possible. In at least one case a hippo was taken unawares by a canoe while it rested on the riverbank during the day. The animal immediately sought refuge in the water and it was sheer luck that the encounter passed without further incident.

Hippopotamus are used to the activity of traditional canoes. Despite the protected nature of the park, fishermen are regular visitors and the hippos take very little notice of them. The opportunity to used larger, motor-powered boats for the survey was available, but declined by the expedition leader, on the grounds that the animals were not accustomed to large boats, or to motors, and would therefore be more likely to respond in a defensive or aggressive manner. We also have doubts about the suitability of the large Ghana Wildlife Department boat at Bui for any use on the Black Volta River. Even at the height of the wet season many areas of the river are very shallow, and the boat is too big to pass easily through rapids. The outboard motor supplied with the boat is for marine use and much too powerful for a river of this nature.

Recommendations for future work.

Further work should concentrate on;

  • A census of the hippo population throughout Bui National Park, conducted during the dry season.
  • The diet of the hippo and the availability of suitable food plant around the banks of  the proposed Bui Lake.
  • The development of basic tourist facilities and the establishment of viewing areas suitable for use by tourists in Bui National Park.
  • A survey and assessment of the conservation potential of the hippo population at Wichan

Bhima, R. 1996. Census of hippopotamus (Hippopotamus amphibius L.) in the Upper Shire River, Malawi. African Journal of Ecology 34:83-85.

Fabre,K. 1996. Unpublished report of a hippopotamus survey along the Black Volta River in Ghana,

Karstad,E.L. & R.J.Hudson. 1984. Census of the Mara River hippopotamus (Hippopotamus amphibius), southwest Kenya, 1980-1982. African Journal of Ecology 22:143-147.

Karstad,E.L. & R.J.Hudson. 1986. Social organisation and communication of riverine hippopotami in southwestern Kenya. Mammalia 50(2):153-158.

Mkanda,F.X. 1994. Conflicts between hippoptamus (Hippopotamus amphibius L.) and man in Malawi. African Journal of Ecology. 32: 75-79.

Ngog Nje, J. 1988. Contribution a l’etude de la structure de la population des hippopotames (Hippopotamus amphibius L.) au Parc National de la Benoue (Cameroun).
Mammalia 15(2): 149-158.

Norton,P.M. 1988. Hippopotamus numbers in the Luangwa Valley, Zambia, in 1981. African Journal of Ecology 26: 337-339.

O’Connor,T.G. & B.M.Campbell. 1986. Hippopotamus habitat relationships on the Lundi River,  Gonarezhou National Park, Zimbabwe. African Journal of Ecology 24: 7-26

Ogen-Odoi, A.A. & T.G. Dilworth. 1987. Effects of burning and hippopotamus grazing on savanna hare habitat utilisation. African Journal of Ecology 25: 47-50.

Oliver,W.L.R. 1993. Pigs, peccaries and hippos. Status Survey and Conservation Plan. IUCN.

The hippos of the Black Volta photographed with a 500mm lens on 800ASA film.

For full details of this project, and others conducted at Bui National Park, see our Final Report.

The Black Volta Project was sponsored by


Supported by Barclays PLC / Royal Geographical Society