Diet of the sable antelope, part 3: Hippotragus niger variani, with special mention of geophagy


...continued from

Please see:

My reference (xeroxed) is:
Richard D Estes and Runhild K Estes (1970) Preliminary report on the giant sable (Hippotragus niger variani). Third progress report. National Geographic Society, hippotragine antelope study.

Fieldwork was conducted in September 1969-March 1970.

The special interest of this report is that it dates back to the period before the Angolan Civil War ( During this war, which lasted more than a quarter of a century, all scientific contact with Hippotragus niger variani was lost.

After the war, H. n. variani was found to have survived, albeit barely so. Intensive efforts have been made in the last two decades to secure the remaining populations. However,

  • there are currently only two observations of this subspecies in iNaturalist (,
  • as far as I know the diet has not been studied beyond what Estes found in this report, and
  • to this day there has never been a study of the diet of H. n. variani in the dry season.

Also of special interest is the mention of geophagy ( at the bases of large mounds of macrotermitines (

The following excerpts, relevant to diet, are verbatim except for the clarifications and links I have inserted (in square brackets) and interspersed. References cited within these excerpts are:


Pages 5-6:

"There are three main vegetation types [in Luando Nature Reserve,], determined primarily by topography and drainage:

  • Brachystegia/Julbernardia woodland, on elevations and other well-drained sites. The greater part of the reserve consists of this woodland. Julbernardia paniculata, Brachystegia boehmi, Isoberlinia angolensis, B. spiciformis, and B. floribunda are the dominant trees, while a variety of other pinnately compound species of the Caesalpinioideae are common. Which species dominate varies according to position in the soil catena. For instance, B. floribunda tends to dominate on gravelly and rocky soils in the upper end of the catena, while Isoberlinia angolensis, B. ?longifolia together with three species of Uapaca [] are commonest in the ecotone between edaphic grassland and woodland. Large termite mounds, many supporting sizeable trees, are prominent features of the woodland.
  • Anhara grassland, equivalent to the dambos [] of Zambia, caused by seasonal waterlogging, interrupts and interdigitates with the woodland. The dominant grasses are Loudetia spp., though small sedges (especially Kyllinga spp.) are more abundant in many places than are grasses. The anharas are studded with grey conical and umbrella-shaped termite mounds [ and]. These provide slight elevations upon which trees and shrubs are able to gain a footing in an otherwise inhospitable environment.
  • Floodplain grassland bordering the Luando, part of the Cuanza [] and their major tributaries. Grasses grow taller and more luxuriantly here than on the anharas."


Kyllinga is now part of Cyperus (

Pages 11-12:

"The giant sable of the Quimbango region [] frequented woodland during the rainy season. The main reason for preferring woodland over anhara was long-ago deduced by Blaine (1922: 321): 'The undergrowth is light, consisting of little low seedlings of bush a few feet high, and a fine, soft, sparsely-growing grass, which is the principal food of the sable...The soil is a sandy loam enriched with leaf-mould, giving place on the dambos to the usual sun-baked knobbly grey clay, where a hard, coarse grass grows which the sable never seems to eat.' This statement is as true today as it was 50 years ago. The Loudetias which dominate the meadows are all wiry stems and no leaf - making them exceptionally good for thatching but useless for grazing (except in their youth, when sable eat a certain amount). In 72 observations of Herds A and C in which the habitat was specified, it was found, however, on the edge of an anhara or tree grassland 47 times (65 percent). Here again, the main reason is that their favourite grasses grow more luxuriantly here, where the ground stays moister than in the woodland without becoming waterlogged like the anharas. Thus the giant sable, like its counterparts in Kenya and Rhodesia, may be thought of as being largely an edge species...But during the rains, good grasses also grow luxuriantly on and around the termite mounds which are so abundant throughout the woodland. The way sable move from mound to mound is an outstanding characteristic of their feeding behaviour. Investigations of termite mounds have shown that they possess a higher content of humus, nitrogen, and colloidal matter than the surrounding soil, as well as a higher maximum water retaining capacity (Murray, 1938 in Glover [sic], 1964). Large termitaria also occur in the anharas, forming islands upon which grow many of the same grasses found in the woodland. These too are visited by sable, even when the surroundings are muddy or actually underwater, for they are not reluctant to venture on to such ground in order to reach an objective. Not only did they readily cross anharas during the rains, but Herd A regularly went out onto Congolo and Cibila anharas both to drink and to visit a number of different salt licks (characteristically found at the bases of termite mounds; samples have been collected for chemical analysis). On at least one occasion, the herd spent some time wandering about a flooded area while feeding upon the grasses growing on slight elevations - most of these being defunct termite mounds of a small type confined to the anharas. On the other hand, sable are only transient in such places; as a rule they frequent firm ground. In fact, after a heavy rainfall in the study area, it was predictable that Herd A would be found in the highest and driest woodland near the Camana picada [motorable track] until the water had drained off and sunk into the low-lying part. As the water table receded, the sable gradually moved back down the catena [,to%20accumulate%20near%20the%20bottom.], showing a preference for the grasses growing on soil that was still moist but firm enough so that their hooves did not sink in deeply. The apparent danger - of injury or increased vulnerability to predators - may deter sable from frequenting very soft ground, as an adult's foot may sink in as much as 45 cm (by actual measurement in a patch of loamy soil). The tendency to favour one part of the catena over another was also discernible on a seasonal basis. The highest ground, with gravelly soils and generally short grasses (except on the termite mounds) was most frequented during the heavy rains of February and early March. From mid-March to mid-April the rains failed. Within a fortnight, the grasses in the high woodland began to turn yellow, while the leaves of some herbs and trees turned yellow or red and began to fall. By that time, the herd had shifted its ground to the lowest woodland bordering the Cibila anhara where the water had been ankle-deep after the last big storm. Here, the general level of the grass was higher (almost 50 cm) with stands up to three metres tall on some termite mounds, and the pasture was as lush as ever. Then came a final week of rain during which the herd moved back to the high woodland, where the grasses quickly resumed growing. The existence of a catena in Herd A's home range is obviously of major significance in its distribution, for it enables the herd to find preferred habitat in one part or another under any given set of conditions."

Pages 13-14:

"The giant sable prefers the same types and selects many of the same species that members of other sable populations were seen to select, especially in Tanzania's Rungwa Game Reserve [], where habitat conditions are very similar...Thus such tufted perennials as Brachiaria, Digitaria, Panicum and Setaria spp. are universal favourites. In short, they prefer the best available pasture grasses; furthermore, they select each species at its tenderest and most nutritious stage of growth. A sable typically bites off the outer 15-30 cm of the plant and seldom crops closer than 20-30 cm. Larger, coarser grasses such as Tristachya superba and Hyparrhenia spp. are also taken as long as they remain tender, but the lower 50 cm or so of these plants may be left, whereas such feathery grasses as Themeda triandra and Hyparrhenia filipendula may be cropped to 10 cm or less. From superficial examination, it often appeared that a herd had specialised almost entirely on one or two different species for a period of a week or two. But whenever grazed plants were actually collected and compared, it turned out that a number of very similar-looking species had been selected. Usually ten or more species accounted for 90 percent of the plants in the sample; seldom did one species make up more than 25-30 percent of the total. The grazing calendar of the sable is closely tied to the phenology of the grasses...Each grass reaches maturity at a particular time, with the result that there is a regular succession of (apparently) dominant grasses, as one common species after another comes into flower. Thus two Digitaria spp. dominated the woodland in November and December, respectively, succeeded by a Tristachya sp. in February and an Andropogon sp. in March. On the anharas, which lagged behind the woodland pastures for some time, an Eragrostis sp. was the first to flower in November, followed by a different Tristachya sp. and finally by Loudetia spp. beginning the end of January. The last of all to mature were the very tall species such as Pennisetum purpureum and Hyparrhenia spp., of which at least 10 occur. Interestingly enough, they started to grow most vigorously after the rains ceased in mid-March, and appear to be largely confined to well-drained deep soils. They are most abundant along the roadside, where they form a narrow strip on either side, in abandoned fields, where tall Hyparrhenia form almost pure stands, and on termite mounds, especially those in the lower parts of the catena. Lately the sable have been seen to feed heavily in some of the flowering Hyparrhenias, but the very tall and coarse species that grow in the plantations and along the roadside are apparently only palatable in their early stages. Sable are also browsers. Considering the general lack of reliable information about this animal, it is remarkable that two of its favourite browse plants have been faithfully recorded by almost every person who has ever observed and written about it. Thus Statham (1922) reported that...Diplorhynchus condylocarpon...and...Dolichos sp. were of major importance in the sable's diet. Undoubtedly it was the local people who called attention to these plants, just as they pointed them out to us, along with a number of other plants that sable browse. In practically every case, observations have borne out their information, although so far we have not seen [Dolichos] eaten. In the case if Diplorhynchus, we have seen sable browsing this very common small tree in almost every area visited to date, and giant sable have been taking it in quantity since the beginning of the study. A low leguminous shrub, Mucuna stans, was equally heavily browsed by Herd B in November and December. For instance, of 75 plants examined in a place where the herd had been feeding, 45 had been browsed. It was one of the dominant shrubs in this part of the woodland, whereas Diplorhynchus was far commoner in the home range of Herd A, which fed on Mucuna stans only occasionally. It is worth noting that both species, the latter in particular, are abundant in abandoned cultivation. Of the dozen-odd other herbs and woody plants which the sable have been seen to browse in Luando, only three were heavily utilised. Early in the season, two succulent Commelina spp. were taken while in flower. Beginning in January, Julbernardia paniculata, the dominant tree, was browsed with increasing frequency. Lately, herds in both the study area and near Mulundo have been feeding heavily on a Blepharis sp., a herb with blue flowers and spiny bracts; last July sable in Victoria Falls National Park were observed eating the dried heads of B. bainesii. The actual percentage if the sable's diet that is made up of browse is unknown and of course variable. It seems unlikely, though, that browse ever accounts for more than 20 percent of the diet."

Tristachya superba:

Diplorhynchus condylocarpon:

Mucuna stans:

The following are spp. of Dolichos in the miombo biome:

Julbernardia paniculata:

Posted on December 07, 2023 03:06 PM by milewski milewski



Which spp. of Cyperus would grow as small plants (formerly Kyllinga) in seasonally wet drainage lines in the miombo biome in central Angola?

Posted by milewski 8 months ago

Add a Comment

Sign In or Sign Up to add comments