Previous PageTable Of ContentsNext Page

Report on the state of knowledge on browse in East Africa in 1980*

H.F. Lamprey. D.J. Herlocker And C.R. Field

UNEP-MAB Integrated Project in Arid Lands (IPAL)


1. Introduction

2. Floristics

(a) Flora, check-list and keys

(b) Ecological surveys: vegetation maps

3. Woodland dynamics

(a) General

(b) Linear growth rates

(c) Productivity

(d)Tree and shrub population dynamics

(e) Ecological succession in woodland communities

Browse impact

The control of trees and shrubs in pasture

The diets of browsing animals 

Insect impact

Chemical composition and nutritive values of browse




1. Introduction

In this report, browse is taken to include only the woody plant species. The area covered comprises Kenya, Tanzania and Uganda. The plant species included in the report are those of the subhumid, semi-arid and arid wooded savanna ecosystems. The sources of information fall mainly into two categories:

  1. Information on woody plants which are browsed by wild and domestic herbivores (or are otherwise important to browsers) on the plant communities in which they occur.
  2. Information on the diets of browsing animals.

Although there is a considerable body of knowledge on the woody vegetation of East Africa, some of which constitutes food for browsing animals, existing data specifically on the characteristics of woody species in their role as browse plants is scarce. The bulk of the literature appears to be centred upon the diets of wild herbivorous mammals, some of it describing the chemical composition and nutritive value of the browse plants eaten by them, and including some data on woody plants observed to be browsed by livestock.

A few papers deal solely with the browsing behaviour of goats, one on the comparative effects of goats and cattle on wooded grassland and one on the crude protein content of the diet of zebu cattle (including the browse component) in wooded grasslands in upland Kenya. The above mentioned papers will be summarized in this report in greater detail than other material owing to their relevance to browsing.

An account of the browse plants of a region must be limited somewhat arbitrarily to the species known to be important to browsing animals, since it would be impractical to deal at length with the hundreds of woody species indigenous to East Africa which are likely to be browsed to some degree. Brenan and Greenway (1949) included 4,694 tree and shrub species in their check-list for Tanzania. Less than 150 browse species can be mentioned in this report.

2. Floristics

(a) Flora, check-list and keys

Brenan and Greenway (1949) prepared an annotated checklist of the trees and shrubs of Tanganyika, giving brief descriptive information on distribution, botanical characteristics, economic uses and, for some species, brief notes on phenology. Gillet and MacDonald (1970) published a numbered check-list of the trees, shrubs and important lianes indigenous to Kenya. Eggeling (1952) gave botanical descriptions and information on the economic value of the trees of Uganda. Dale and Greenway (1961) provided a compact botanical account of the trees and shrubs of Kenya, giving some information on their economic significance. Burtt (1939), revised by Welch (1957), produced a pocket key to the genera and most species of the trees, shrubs and climbing plants of Tanganyika, and a second volume, revised by Glover et al (1953), on the species of the more important genera. These keys, which are applicable to the subhumid and semi-arid woodlands of Tanzania, are also useful to a considerable degree in the same climatic zones in Zambia, Malawi, Uganda and Kenya, although the nomenclature is somewhat out of date.

The most important of the botanical references to woody plants are those in the Flora of Tropical East Africa (eds. Turrill and Milne-Redhead; Hubbard and Milne-Redhead), which is being published progressively. The taxa of woody plants already published in the Flora are as follows:

Sub-families: Caesalpinioidae and Mimosoideae of the Leguminosae. Families: Aizaceae, Alangiaceae, Annonaceae, Aquilifoliaceae, Araliaceae, Berberiadaceae, Brexiaceae, Buxaceae, Canellaceae, Capparidaceae, Caprifoloaceae, Caricaceae, Caryophyllaceae, Chenopodaceae, Connaraceae, Convolvulaceae, Cornaceae, Dilleniaceae, Elatinaceae, Cycadecaeae, Hamamelidaceae, Hypericaciae, Icacianaceae, Lecythidaceae, Linaceae, Loganiaceae, Maliphiaceae, Melianthaceae, Menispermaceae, Olacaceae, Opiliaceae, Phytolaceaceae, Pittosporaceae, Polygonaceae, Ranunculaceae, Rhamnaceae, Rhizophoraceae, Rosaceae, Salvadoraceae, Sapotaceae, Scytopotalaceae, Sonnerataceae, Tamaricaceae, Theaceae, Turneraceae, Ulmaceae.

The Flora of Tropical East Africa and the previously mentioned floristic publications contain no information specifically relevant to the subject of browsing.

Pratt and Gwynne (1977), in their appendix on important rangeland plants, included a section on trees and shrubs, briefly describing some sixty woody species common in East Africa. They mention the importance of some of them, either as browse plants or as species poisonous to livestock. Verdcourt and Trump (1969) described the common poisonous plants of East Africa, including several woody species. Ivens (1967), in a small book on East African weeds and their control, provided a brief section including nineteen common woody species.

(b) Ecological surveys: vegetation maps

Knowledge on the habitats of tree and shrub species is contained in several ecologically oriented publications, descriptions of plant communities and vegetation maps. The most comprehensive work is that of Lind and Morrison (1974) on the vegetation of East Africa, which provides descriptions of the main communities, with lists of their characteristic plant species and accounts of the environmental factors influencing their geographical distribution. One section of this book discusses the history of the vegetation of East Africa.

Pratt, Greenway and Gwynne (1966) provided a classification of the main vegetation types of the East African rangelands, based mainly upon physiognomic features, but they also characterised the main eco-climatic zones by their dominant tree genera. Greenway (1973) prepared a classification of all the important vegetation types of East Africa, giving lists of plant species to illustrate the composition of the main communities.

Several vegetation maps and map series have been published which cover the region or parts of it, including:

Langdale-Brown and Trapnell (1962): East Africa.
Edwards (1940): Kenya.
Gilman (1949): Tanganyika.
Langdale-Brown et al (1964): Uganda. 
Dale (1939): Coast Province, Kenya.
Dale (1940): Mt. Elgon; Kenya and Uganda.
Moomaw: (1960): The Coastal Region, Kenya. 
Wilson (1962): Karamoja District, Uganda.
Trapnell et al (1969): South-west Kenya.
Herlocker and Dirschl (1962): Ngorongoro Conservation Area, Tanzania
Herlocker (1979): South-western Marsabit District, Kenya.

Other descriptions and analyses of woodland plant communities which form parts of localised ecological surveys include the following:

Phillips (1929): Centra Province, Tanganyika. Scott (1934): Central Province, Tanganyika.

Burtt (1942): Tanganyika (several woodland communities).

Greenway (1933): Mpwapea District, Tanganyika. 
Pielou (1952): Rukwa Valley, Tanganyika.
Bogdan (1958): Kiboko, Kenya.
Langdale-Brown (1959a): Eastern Province, Uganda. 
Langdale-Brown (1959b): Buganda, Uganda.
Langdale-Brown (1960a): Northern Province, Uganda.
Langdale-Brown (1960b): Western Province, Uganda.
Greenway and Vesey Fitzgerald: (1969): Lake Manyara National Park, Tanzania.
Glover (1966): Narok District, Kenya.
Hemming (1972): South Turkana, Kenya.
Barkham and Rainy (1976): Samburu-Isiolo Game Reserve, Kenya.

The conservation status of woodlands (among the whole range of vegetation communities in East Africa) was discussed by Polhill (1968) for Tanzania, Osmaston (1968) for Uganda and Lucas (1968) for Kenya, and a regional synthesis was provided by Verdcourt (1968). Lamprey (1975) made an evaluation of the distribution of protected areas in relation to the needs of ecosystem conservation in eastern Africa, drawing attention to the vulnerability of certain forest and woodland communities to destruction or degradation.

3. Woodland dynamics

(a) General

Information on dynamic aspects of woodland ecology in East Africa is scarce and compares unfavourably with the extent of knowledge on the dynamics of grassland communities. It appears that this is partly due to the difficulties and the time involved in obtaining a series of measurements of trees and woodlands as compared with measurements on the herb layer.

(b) Linear growth rates

Growth-rate measurements have been made in Brachystegia woodland communities (Glover, 1939) and in acacia and Combretum woodlands (Herlocker, 1976) in Tanzania. In the former, the heights and trunk diameters of small samples of nine common tree species were correlated with numbers of growth rings (assumed to be annually produced). Glover obtained an approximate growth curve for the tree Julbernardia globiflora, indicated a relatively fast mean annual increase in trunk diameter of 4.2 mm for the first 50 years of growth, declining to approximately 2.0 mm thereafter, with old trees showing up to 120 or 130 rings and having trunk diameters of 3640 cm and heights of 1617 m. Similar data on small samples of six acacia species in the Serengeti region are given by Herlocker. Variability in growth rates due to site conditions and climate is very high but certain generalizations on characteristic growth rates for each species can be made: e.g. Acacia xanthophloea under favourable conditions shows annual diameter increments of 17-25 mm and annual height gains of 1.4 m in the sapling stages. A. tortilis under favourable conditions has an annual diameter increment of 10-14 mm and a height gain of 0.8 m in the young stages. This species normally reaches maturity at c. 6 m in height after about 20 years. In all species measured, growth-rates are relatively slow in the drier parts of their range and may be less than half of those in the wetter areas.

Similarly, a tree's position on the drainage catena will influence its growth rate. For the great majority of browse species no information on linear growth rates is available.

(c) Productivity

The type of linear increment measurements referred to above, as yet have not been related either to measurements of biomass nor to productivity in East African trees and shrubs, although work is now in progress in northern Kenya to obtain such correlations. Whilst it is desirable to obtain estimates of biomass and productivity of all parts of the plant (roots, stems, bark, branches, twigs, foliage, flowers and fruits) only the last four of these categories can normally be classed as browse (although the roots and bark of some trees are commonly eaten or chewed by elephants). It is evident that, with limited time and assistance, research effort should be concentrated upon the parts which are normally browsed by ungulates.

It appears that no browse production measurements have been published or reported for East Africa. However data have been obtained recently on browse production in the Serengeti National Park, Tanzania and in Marsabit District, Kenya which are due to be published in 1980 or 1981. In the former locality, R. Pellew (personal communication), carried out an ecological study of the giraffe population and its relationship with the Acacia food resource.

With the use of fenced enclosures, the main effort was devoted to the determination of seasonal standing crop biomass/ha of browse available to giraffe (leaf and new shoots only, and less than 5.75 m high), rates of net primary productivity/ha/annum of certain Acacia species, and rates of browse consumption/ha/annum by giraffe.

For the three Acacia species, A. tortilis, A. xanthophloea and A. hockii, data are available as follows:

  1. Height specific vertical growth-rates for browsed and unbrowsed trees.
  2. Bush area/ha increments for browsed and unbrowsed trees.
  3. Canopy volume/ha increments for browsed and un-browsed trees.
  4. Growth rates of individual tagged twigs (expressed as linear or weight increments) for browsed and unbrowsed trees-these were used as the basis for the productivity estimates.
  5. Standing crop biomass/ha of browse available to giraffe for four separate woodland vegetation types (riverine, drainage-line, open mid-slope and ridge-top Acacia regeneration thicket), expressed on a wet/dry season basis. For the ridge-top vegetation type, monthly variations in available biomass have also been determined. These biomass figures are all gross quantitative estimates only, with no qualitative or palatability assessment incorporated.
  6. Rates of primary productivity of browse available to giraffe (excluding lignified stems and branches) for the three acacia species, expressed as kg dry weight/ha/annum.
  7. Rates of browse offtake by giraffe of the three Acacia species, expressed as kg dry weight/ha/annum.
  8. Estimates of the longer-term stimulatory feed-back of browsing upon the productivity of the three Acacia species by comparison with unbrowsed shoots clipped to simulate giraffe browsing (these data are still being analysed).

Results to date suggest the following early conclusions:

a) The biomass of available browse represents a significant proportion of total habitat biomass, particularly in the areas of acacia regeneration thicket.

b) Browse productivity rates are high, reaching 6000 kg/ha/ annum (dry weight) in the riverine A. xanthophloea regeneration, a production which is similar to that of the neighbouring Serengeti long grass plains but double that of most savanna grasslands.

c) Acacia tortilis ridge-top thicket has a production of approx. 1500 kg/ha/annum (dry-weight).

d) Offtake by giraffe is high, reaching c. 80% per annum of A. xanthophloea production and c. 60% per annum of A. tortilis production below the height of 5.75 m.

Limits of tolerance of the trees to these high levels of offtake remains to be investigated.

In northern Kenya, D.J. Herlocker (IPAL) is engaged in a study of primary and secondary productivity patterns associated with sheep, goat and camel herds in northern Kenya in cooperation with H.J. Schwartz and C.R. Field. The following data are being collected.

  1. Systematic samples of vegetation composition and biomass through time: 
    a) fortnightly, for the herb and small dwarf-shrub layer; 
    b) less frequently for large dwarf-shrub layer (because of the difficulty of destructive sampling at such     frequent intervals); 
    c) shrubs and small trees only, sampled when a new site is used by the sheep and goat or camel herd or when a particular species is shown to be preferred by the animals;
    d) food preference surveys carried out every two weeks and samples taken for chemical. analysis of those plants which are preferred.
  2. In order to allow more frequent and/or extensive sampling of large dwarf-shrub and shrub species, correlations are being determined, through destructive sampling, of plant component biomass with a number of easily measurable attributes (crown diameter, plant height, stem diameter and capacitance meter, phytometer and 10-point frame readings).
  3. Correlations of plant and plant component biomass with measurable attributes are then combined with samples of population structure to obtain whole plant and plant component (including available forage) biomass from sample aerial photography of woodlands.
  4. Herd attributes, such as mortality, lambing rates, incidence of disease, growth rates, population structure and milkyield are monitored at the same time: 
    a) one half of the herd is a control and is under traditional management; 
    b)the other half is periodically treated for the elimination of intestinal worms.
  5. Several satellite herds under traditional management, are monitored at less frequent intervals for the more easily-monitored attributes.
  6. Forage intake quantity is estimated using the chromic acid method.

It is worthwhile mentioning that comparable studies of woodland standing crop and primary productivity are reported by Rutherford (1978) and Dayton (1978) from South Africa. These authors provide (and quote) foliage and fruit production figures from Colophospermum mopane, Burkea africana, Acacia albida, Acacia erioloba, Acacia tortilis, Piliostigma thonningii, Combretum apiculatum and Combretum zeyheri.

In his paper on browse production and utilization in the Tarangire National Park, Tanzania, Vesey Fitzgerald (1973) carried out a survey to estimate the offtake by wild browsing animals relative to the foliage production in Acacia and Combretum woodlands. He did not measure browse production or consumption in absolute terms. His results are discussed under the heading "browse impact" below.

(d)Tree and shrub population dynamics

The dynamic state of woodland communities has been recognised by many authors and several accounts exist of structural changes observed (or inferred from evidence) in woodland ecosystems in East Africa. The greater part of the extensive literature discusses not only the changes themselves but also their causes. Descriptions of the size-class structure of sample tree populations in the Serengeti National Park were given by Lamprey et al (1976), Herlocker (1976), Vesey-Fitzgerald (1970) and Croze (1974). In each case the tree populations studies were influenced by fire and elephants and the size-class structures were related to observable or predicted changes in tree numbers and total canopy cover. These studies are referred to at greater length under the subject heading "browse impact" below.

(e) Ecological succession in woodland communities

Vesey-Fitzgerald (1973) discussed the nature of successional change in woodland communities, illustrating his discussion from successions taking place in the Tarangire National Park, Tanzania. Herlocker (1976) described the apparent reduction of riverine forest and its replacement by wooden grassland in the northern Serengeti, and thinning out of woodland trees in the same region, resulting in increased areas of grassland. These and other successional changes taking place in the East African national parks are discussed further in the following section on "browse impact".

It is widely recognized that many rangeland plant communines in East Africa, which are not edaphic grasslands, are liable to "bush encroachment" (Pratt, 1964). Such bush encroachment commonly involves the increase in density of one or more indigenous woody species (possibly after an initial invasion of the grassland). In semi-arid grassland communities, the natural succession is normally towards woodland. "Where open grassland occurs (except in edaphic grasslands) it is usually because the natural succession has been diverted or arrested by extraneous factors, notably by fire (Phillips, 1930; West, 1965)" (Pratt and Gwynne,1977). Under these circumstances the encroaching "bush" is usually regarded by the cattle rancher or the pastoralist as a "weed", to be removed if possible. Woody species included in this category, are Acacia drepanolobium, Acacia hockii, Tarchonanthus camphoratus, Euclea divinorum, Combretum spp., Commiphora spp.. In relatively dry areas, grass species are commonly replaced by dwarf shrubs, e.g. Duosperma, Indigofera spp., Serococomopsis, Solanum incanum. If sufficient grass cover remains (after grazing or encroachment by woody species) to provide fuel for periodic burning, the woody species can normally be held in check, if not eradicated by grass fires. Heavy grazing may prevent grass fires and lead to progressive colonisation by woody species. Under these circumstances the woodland or shrubland which develops may become suitable habitat for the browsing ungulates.

Browse impact

The greater part of the published and reported data on the impact of browsing upon woody vegetation in East Africa deals with the effects of wildlife, especially elephants, in national parks. The volume of literature reflects the seriousness of the management problems caused by the impact of elephants upon national park tree populations over the last thirty years. There is general agreement that the problem of excessive elephant influence originates from the over-concentration of elephant populations within the parks and reserves which offer them protection as their former habitats are increasingly affected by human occupation and activity.

Buechner and Dawkins (1961), describing vegetation changes taking place in the Murchison Falls (now Kabalega) National Park in Uganda, said, "All types of woody vegetation are in the process of conversion to grassland under the combined influence of elephants and fire. Large trees are killed by fire damage to tissues exposed by the action of the animals in gouging, peeling and ripping the bark while foraging and rubbing on the boles of the trees". In the twenty sample plots (each of 0.1 ha) in Terminalia glancescens - Prosopis africana woodland which these authors analysed, bark removal and damage by elephants had occurred in a varying but generally large proportion of the trees.

In several other studies in East Africa estimates of rates of damage and killing of trees by elephants and the resultant changes in Acacia woodland size structure have been published (e.g. Lamprey et al, 1967; Field, 1971; Douglas-Hamilton, 1972; Groze, 1974; Harrington and Ross, 1974; Herlocker, 1976; Pellew, 1979). The last named author, reviewing previous data on the Serengeti Acacia woodlands confirmed previous estimates of mortality in mature Acacia xanthophloea and Acacia tortilis as c. 6% per annum over the six years 196874, and predicted that, at current elephant numbers, all existing mature trees in the Acacia tortilis and A. xanthophloea woodlands may be lost within 15 years or less. Meanwhile, a combination of giraffe and fire impact prevents the development of replacements. Fire burns small trees back to ground level; giraffe browsing tends to hold the trees at a height vulnerable to burning; those few trees which emerge from the giraffe-fire cycle are killed by elephants. Mature woodlands are being replaced by dense thickets which are now prevented from developing into woodlands by browsing and intermittent burning. Pellew provides a simulation model of A. tortilis woodland dynamics under the influence of fire and browsing which is programmed for 25 years. It incorporated a seedling input and assumes a constant level of Bruchid beetle predation on the seeds (see "insect impact" below).

Vessey-Fitzgerald (1973) estimated the relative browse impact (mainly the elephants) on ten common and widespread tree and shrub species in Acacia woodlands in the Tarangire National Park, Tanzania. The woody vegetation observed had a mean density of c. 600 plants/ha. Mean densities of mature trees were 21/ha; young trees, 225/ha and shrubs 350/ha. The overall relative frequencies of the three classes were: mature trees 4%; young trees 42% and shrubs 54%.

In Combretum-Commiphora woodland the mean density was measured as 1400 plants/ha: mature trees 24/ha; young trees 650/ha; shrub 720/ha. The relative frequencies were: mature trees 2%; young trees 46%; shrubs 52%.

This author observed that 46.5% of the trees and shrubs had been browsed to some degree, the percentage varying between species. A total of about half of the available browse material was utilized. Browse impact was assessed in three degrees of intensity. All of the ten plant species experienced light and intense browsing. At the latter degree of intensity the plants exhibited characteristic dense and pollarded growth forms with reduced height increment rates. At a severe intensity of browsing, which was experienced by about 4%of each seven plant species, the degree of damage was sufficient to reduce their chances of survival and to result in death within one or two years, particularly if exposed to grass fires.

Vesey-Fitzgerald concluded that the regeneration, recruitment and replacement of the commonest tree species was more than sufficient to compensate for the observed offtake. He did not quantify these rates in absolute terms.

The main browsing animals in the Tarangire area were elephants, giraffe and black rhinoceros, which were present in the area studied at mean densities which had been determined between 1958 and 1961 as 1.11/km2, 1.14/km2 and 0.15/km2 respectively (Lamprey, 1964). Together these animal populations amounted to a biomass density of c. 3395 kg/km2 (liveweight). As an approximation, given an estimated standing crop foliage biomass density of 1500 kg/ha dry weight, (for woodlands at a mean annual rainfall of 600 mm: (cf. Rutherford,1978) and a browse offtake of 50% of the foliage standing crop, c. 40 kg/ha (liveweight) of browsing animal biomass is supported by an offtake of c. 750 kg/ha dry weight of browse biomass, under the relatively natural conditions of a national park. (This does not take into account the relatively small amount of grass consumed by elephants).

Effects of browsing by livestock are discussed in a small number of East African publications. Staples et al (1942) described a study of the comparative effects of goats and cattle on a prepared area of wooded grassland at Mpwapwa, central Tanzania in semi-arid woodlands (mean annual rainfall c. 700mm), basing their observations on five similar fenced plots (32 32 m). The experiment, which ran from January 1938 to March 1942, provided qualitative data on vegetation changes which took place in the two plots browsed only by goats, two grazed only by cattle and the one which was unused and served as the control. The pasture had been cleared from "climax" bush cutting, burning the trash and planting the burnt patches with Cynodon plectostachyum two years earlier and had since been protected from heavy grazing. "The result was a good stand of grasses, mixed with herbs and regenerating, though inconspicuous, woody plants; a useful pasture tending to revert to bush".

Although the authors record that "either two oxen or about 14 goats were put on one or two days a week whenever there was enough food to keep them all reasonably contented", actual stocking rates were not recorded but amounted to 30 to 40 grazing days per annum and was the same for each plot. From this information it is apparent that the goat stocking rate was approximately 1.4 stock unit/ha/annum and cattle 2 stock units/ha/annum. (1 stock unit = 250 kg). "If one plot became too bare for use, grazing of all was discontinued until all were ready again". A plant species list was prepared for each plot at the beginning and the end of the experiment. The lists included a total of some 55 woody species as well as a similar number of grasses and other herbs.

After four years, inspection of the plots showed great modifications from their original states which the authors described in some detail. The goats browsed all the plants within reach, including the taller grasses. They did not browse any plants down to ground level and consequently did not destroy any. They did little damage to the bark of the well-grown trees, and young trees which succeeded in putting out branches beyond the reach of goats were able to continue their growth. The net result of stocking with goats under the conditions of the experiment was to preserve a good ground cover including many plant species. The cover was too low to support tsetse and sufficient to protect the soil. The authors concluded that it would be difficult to crowd goats, on the unfenced grassy bushland in such high concentration that they bare the ground sufficiently to initiate serious erosion. The authors further conclude:

"By contrast the effect of pasturing cattle on grassy bushland is invariably bad. By concentrating on the grasses they relieve the bushes from competition, and though the advantages (to the bushes) of this are more than offset by packing of the soil through trampling, the unrelieved tendency is towards the formation of open thicket with little ground cover; obviously a condition of continuous reduction of carrying capacity and accelerated erosion, until the appalling state of affairs seen around Dodoma, Kondoa Irangi and Singida (Central Tanzania) is reached."

"Wherever in bushland country there is some sort of ground cover of grasses and herbs which can be expected to extend if given a chance, such an area will not be damaged but may be much improved by even heavy goat browsing, whereas certainly it would be damaged by anything more than the lightest grazing by cattle. The current idea that the goat is a pernicious animal, of which the numbers must everywhere be reduced if the land is to be saved, is quite wrong when one is considering the economics of bushland utilization. In fact if one desires to obtain animal produce from this type of country without first clearing it and putting it down to grass, the only way the attempt can safely be made is by making use of goats".

From the experiment of Hornby et al (1948) it is evident that wooded grasslands used at moderate and regulated stocking rates by goats can be maintained in good condition. General, but apparently undocumented, experience on goats indicates that they become destructive to woody vegetation at densities which are high in relation to the productivity of the vegetation (as would be expected).

Field (1979), in an ecological study of sheep and goats observed their feeding under controlled stocking rates and recorded diets and quantities of forage taken. By weighing a small sample of sheep and goats before and after a day's feeding and allowing for measured intake of water and output of faeces and urine, one male sheep of 37.6 Kg body weight in dry matter, a second male sheep of 37.2 Kg consumed 2.1% and a male goat of 27.1 Kg consumed 2.0% of its body weight. These figures are similar to those of other investigators (e.g. Quartermain and Broadbent, 1974). This author observed that in the arid environment of northern Kenya, goats consume approximately equal quantities of browse and herb material (relatively more browse in the wet season), the diet of sheep also includes a high proportion of browse (approximately 26% in the dry season and 38% in the wet season). The corresponding figures for goats were 45% and 57%. In these experiments the highly palatable dwarf shrub Indigofera spinosa comprised between 22% and 40% of the diets of both sheep and goats except where the latter were at a light stocking rate, when it comprised 7% of the diet.

Field C.R. (1979) in a study of camel diets in northern Kenya, gives the daily dry matter consumption of a camel as 2.5% of body weight. Using population data collected by Unesco IPAL in the Mt. Kulal study area (22.500 km2, were estimated to have a daily dry matter intake of 14.7 kg/km2 (mean annual D.M. I = 5380 kg/km2). The combined productivity of the herb and dwarf shrub layer in April and May 1977, measured in 7 paddocks totalling 40 ha, was 62.729 ± 24.526 kg/km2 which was obtained with 133 mm rainfall. On this basis the annual dry matter production of the herb and dwarf shrub layers is estimated at 117.911 kg:km2. Extrapolating to the study area as a whole, the camel population would appear to consume 4.6% of this production. However, since there has been no calculation of the productivity of the trees which are also browsed by the camels, it is estimated that approximately 30% should be subtracted from the herb and dwarf shrub intake figures giving an intake by camels of approximately 3.2% of the annual production.

These observations require further qualification because camels are not distributed evenly over the region. Since ecological degradation is taking place mainly in the vicinity of human settlements and such areas appear to be the foci for the spread of desert conditions, the main interest lies in the intensity of use by livestock of such areas. The highest observed mean density of camels is 8.0/km2 but densities of 3 to 4/km2 are not uncommon.

At the highest observed camel density the offtake of the dwarf shrub and herb layers would be of the order of 16% of the production. However, since the sheep and goat populations (existing in the region at a biomass density similar to that of the camels) are thought to eat rather more than the camels near settlements and water holes, the total consumption of the herb and dwarf shrub layers is likely to exceed a level equal to 32% of the mean productivity of the region, and considerably higher than that of the degraded areas near the settlements.

The control of trees and shrubs in pasture

Bush encroachment control, as an aspect of range management in East Africa has been the subject of several publications. In this review on information on browse, it is sufficient to mention briefly some of the more important references.

In his paper on the reclamation of severely overgrazed pastures with bush encroachment in Kitui District, Kenya, Jordan (1975) placed emphasis on accelerating pasture recovery by complete closure to grazing and re-seeding with grasses. Bush control, largely involving killing scrub Acacia spp., was achieved by cutting down the trees and burning the trash which had been left scattered. Eighteen to twenty-four months later, the grass growth was sufficient to provide fuel for a burn at the end of the dry season which killed the regenerating trees and shrubs.

Pratt (1964, 1966a, 1976b); Thomas and Pratt (1967) and Pratt and Knight (1968) published a series of papers on bush control experiments in the control of several woody species by burning, by grazing management and by mechanical and chemical methods. The main species concerned were Acacia brevispica, Maytenus putterlickioides, Acalypha fruticosa, Tarchonanthus camphoratus and Duosperma eremophilum. In the case of each type of bush encroachment experimental results were given indicating appropriate control measures. The economics of bush control measures under current and potential land use were not discussed.

Harker (1959) studied the problem of controlling Acacia hockii, regarded as a weed in pastures in Uganda. A variety of control measures, including chemical, mechanical and browsing by goats, were discussed and the economics of chemical control methods assessed. Ivens (1967) summarized information on the control of woody weeds in East Africa.

The diets of browsing animals

Hofmann and Stewart (1972) provided an analysis of information on the dets of wild ruminants in East Africa and on their stomach structure, upon which they based their classification of the ruminant species as grazers or browsers. Browsers, described as selectors of juicy, concentrated herbage, were divided into (a) tree and shrub foliage eaters and (b) fruit and dicot (tree, shrub or forb) foliage eaters. Category (a) includes Boocerus, Giraffa, Litocranius, Strepsiceros, and Tragelaphus stresiceros. Category (b) included Cephalophus, Nesotragus, Oreotragus, Madoqua, Sylvicapra, Tragelephus scriptus. Intermediate feeders (a) preferring grasses included Aepyceros and Gazella thonsoii, (b) preferring forbs and shrub and tree foliage, included Gazella gramti, Raphicerus, Tanrotragus.

These categories are contrasted with (a) the roughage grazers (Alcephelus, Damaliscus, Oryx and Redcunca fulvo: (b) the fresh grass grazers dependent on water (Adenota Connochaetes, Kobus, Ourebia, Redunea reduca and Syncerus, (c) the dry region grazers: Oryx b. beisa. The descriptions and photographs of types of rumen and omasum structures found in these different feeding categories are given in Hofmann (1968). In his classification cattle are typical roughage grazers; sheep and goats have the stomach structure described as transitional between the extreme grazers and browsers, adapted for occasional, perhaps seasonal selective intake of some dry, fibrous forage. The camel, which is not included in this author's classification, is likely to fall into the specialised browsing group with the giraffe.

The feeding behaviour and diets of wild herbivores are described by a number of authors. Some give observed food preferences of a number of browsing species (e.g. Lamprey, 1963). Others provide detailed observations on single species: Leuthold (1970) on Gerenuk(Litocranius); Goddard (1968) on black rhinoceros (Diceros); Field (1971); Douglas-Hamilton (1972) on the elephant (Loxodonta); Field and Ross (1976) on elephant and giraffe; Napier Box and Sheldrick (1973) on elephant. Several of these authors give an analysis of the percentage contribution to the animals' diets provided by each of the important browse species; some relate food selection to food plant availability; some show seasonal preferences.

Information on the diet of goats in a semi-arid area in Kenya, is given by Edwards (1948). He classifies 67 plant species, the great majority woody (and including no grasses), into three categories of palatability to goats based upon the relative frequency of selection. Acacia tortilis, Grewia kakothamnos and Ochna stuhlamnnii were found to be "extremely palatable". Staples et al (1942) described goats as primarily browsers, taking little grass. Wilson (1957) observed that grasses made up 33.5% of all the plants eaten by the goats he was studying. Knight (1965) recorded a large proportion of grass in the diets of goats in Baring o District. In 1956 observations ("feeding stops") of 20 goats, in September 1962, 38.4% were for grasses, 33.6% for herbs and 28% for trees and shrubs. In December 1959 and February 1960 when 4 goats were observed, on each occasion 65.4% and 82.5% respectively of feeding observations were of grasses. This author also noted a marked individual diet preferences in certain goats.

Field (1979) observed the diets of goats and sheep in five different areas and in six vegetation types in northern Kenya making over 5000 records. Her results are summarized as follows:

% in diet of



Dwarf Shrubs,



Leaf Litter















Insect impact

The effects of insects upon browse plants appear to have received very little attention in East Africa. The tree locusts, Anacridium melanorhodon and A. wernerellum (Acridiidae), undergo localised outbreaks which, although small in comparison with those of the migratory locust, have a large impact on the foliage of several common Acacia species and are mentioned briefly in the Locust Handbook (1966) and a map shows their approximate distribution in Africa. No quantitative information on the browse impact of these insects appears to be available in East Africa.

Lamprey et al (1974) described the role of the seed beetle Bruchidius spadiceus in infesting the seeds of Acacia tortilis spirocarpa in Tanzania. A high proportion of the seeds observed were infested and it was found that, unless ingested by a browsing mammal they were killed by the insects. Passage through the gut of a mammal (for which the seeds of this and other Acacia species are evidently adapted) kills the larval insects and renders the seed viable, resulting in relatively high germination rates. Further studies of the insect fauna of Acacia tortilis are in progress in Kenya.

Chemical composition and nutritive values of browse

The results of chemical analysis of browse plants have been given by several authors in connection with studies of the diets of browsing ungulates, both wild and domestic. The most important publication is that of Dougal and Drysdale (1964) who provided a table giving information in a substantial list of East African indigenous fodder plants observed to have been eaten by browsing ungulates. The greater part of this list is given in Table 1 of this report, including all the woody species analysed and some grasses for comparison. Also included in this report are a series of histograms (Figures 15) from the same authors, giving the percentage distribution of ash, silica calcium, phosphorus, crude protein, crude fibre and nitrogen-free extract in grasses, legumes, leguminous browse, and non-leguminous browse collected in East Africa. This paper provides a discussion on the nutritive value of browse plants, pointing out that the legumes are the richest source of protein and that they are the least fibrous of all the vegetation.

Figures 1-5 . The percentage distribution of ash; silica ; calcium; phosphorus; crude protein crude fibre and nitrogen-free extractives in grasses ( G ) ; legumes (L ) ; leguminous browse (LB)and non-leguminous browse (NLB).

Figure 2. Silica

Figure 3

Figure 4

Figure 5

Douglas and Sheldrick (1964) recorded a day's diet of an elephant in Tsavo National Park, giving the species browsed, the parts of the plants selected, the number of occasions each species was browsed and the chemical composition of the parts browsed.

Field (1975) gave the chemical composition of fodder plants on the Galana Ranch in South East Kenya (semi-arid scrub-woodlands) and analysed the monthly crude protein levels of the stems and leaves of Grewia and Combretum spp.. He also compared the diets of cattle, buffalo, oryx and eland in terms of the grasses, herbs and browse and found the latter species only to be primarily a browser.

McKay and Fransden (1969) described the chemical and floristic component of the diet of zebu cattle in semi-arid wooded grasslands in Kenya. While these authors confirmed that the cattle mainly selected grasses, some browsing was observed, especially in overgrazed areas, where woody plants gave high levels of crude protein (often between 15%and 25% of dry weight. This paper provides graphs of variable crude protein percentage in Acacia nilotica and Acacia brevispica (leaves and pods separately); Rhus natolonsis, Tarchonanthus camphoratus and Grewia bicolor (leaves and stem tips); Dichorostachys cineria and Capparis fascicularis (leaves). Over a period of nearly two years, relating protein levels to rainfall, the high seasonal variability of protein content of browse plants (e.g. between 10% and 23% in Acacia nilotica foliage) shows that a single chemical analysis of a particular browse sample is likely to give an incomplete picture of the plant's nutritive value on a year-round basis.

Gwynne (1969) discussed the high nutritive value of Acacia seed pods and its significance for seed dispersal by ungulates in some Acacia species. He gave the results of chemical analysis of the dehiscent and indehiscent pods of four Acacia species and their seeds. The author concluded that the indehiscent pods (A. tortilis, A. nilotica, A. siebenaia and A. albida) are especially sought by browsing ungulates, not because of their nutritive content, which is similar to that of the dehiscent species, but because of their large size and conspicuousness, which makes them easy to pick up.

Taylor (1969), in a publication on water economy in eland and oryx in Kenya, described how the requirements of these two species were provided largely by the water contained in their browse. The weight of acacia foliage which an eland would normally ingest in a day, containing 58% by weight of water, would provide a total of about 5.3 litres of water per 100 kg of body weight, which is the amount needed by the eland for survival. This author showed that, under average conditions of temperature and humidity at night, the sub-shrub Duosperma, with a water content of as little as 1% in the daytime, absorbed atmospheric water at night amounting to a total water content of 42% of its weight. He suggested that, by eating mainly at night, oryx could take in food containing an average of 30% water and thus become independent of free water. Grant's gazelle which is known to live without drinking (even in the dry season) must obtain its total water requirements from its browse plants, of which Indogofera spinosa is one of the most important over large areas of semi-arid and arid rangeland in East Africa.

In livestock, as in wild ungulates, the intake of free water by drinking becomes minimal when there is abundant green browse available. Under these circumstances camels may not need to drink and other livestock species will drink relatively infrequently. Being denied the opportunity to browse at night, as wild ungulates do, livestock cannot obtain water absorbed by their forage plants and therefore are very largely dependent upon drinking during the dry season, a factor which may reduce their foraging range at that time of year.


Published information on the phenology of browse plants in East Africa is very scarce, although some is available for certain species in floristic works (e.g. Brenan and Greenway 1949). Unpublished notes on the phenology of trees in the Serengeti National Park have been made by Herlocker (1969-1972), which appear to fill a considerable gap in the existing knowledge of these trees. The species included are: Acacia xanthophloea, A. kirkii, A. seyal, Acacia tortilis, A. clavigera, A. sieberiana, A. hockii, A. mellifera, A. Senegal, A. gersardii, A. polycantha, A. nilotica, A. drepanolobium, Erthrina abyssinica, Erythrina burtii, Entada abyssinica, Lannea stuhlmannii, Commiphora trothae, Commiphora madagascarensis, Lonchocarpus eriocalyx, Commiphora africana, Combretum molle, Heeria retienlata, Grewia fallax, Albizia harvey, Albizia peterian, A. amara, Ormocarpum triehocarpum, Sclerocarea birrea, Kigelia aethiopum, Termindia mollis.

Table 2. Average chemical composition of Kenya browse and pasture herbage (per cent of dry matter)









10.85 (535)

11.12 (204)

  8.60 (64)

11.73 (170)

Crude protein

11.51 (518)

21.86 (225)

14.77 (64)

12.70 (170)

Crude fibre

30.34 (455)

21.90 (141)

29.54 (64)

29.27 (170)

Nitrogen-free extract

45.01 (455)

43.28 (141)

45.25 (64)

43.75 (170)

Nitrogen (N)





Silica (S102)

  4.95 (427)

  1.18 (186)

  0.56 (48)

  1.46 (132)

Silica-free ash

  5.90 (427)

  9.91 (186)

  8.04 (48)

10.27 (132)

Calcium (Ca)

  0.42 (518)

  1.17 (225)

  1.82 (64)

  1.82 (170)

Phosphorus (P)


  0.288 (225)

  0.186 (64)

  0.212 (170)

Sodium (Na)


  0.028 (15)

  0.066 (45)

  0.074 (111)

Potassium (K)

  2.54 (120)

  3.08 (15)

  1.19 (38)

  2.13 (110)

N/P ratio





Ca/P ratio





(S102)/Ash ratio






Existing knowledge on browse in East Africa is evidently deficient in most respects, particularly in view of the need for information which can be applied in range and livestock management. The foregoing summary on the present state of knowledge indicates no aspect of the subject on which there is adequate data, with the possible exception of the diets of certain wild browsing herbivores observed at specific localities. There is, therefore, a very wide field of research which could be undertaken on browse. Nevertheless, a limited number of priorities may be suggested as having the greatest urgency in relation to the immediate problems of range management and livestock production.

The outstanding gaps in present knowledge on browse concern productivity and methods for measuring it. The first requirement is for the establishment of reliable and economic techniques for assessing browse production at various scales from individual trees to large and diverse woodland areas. It seems inevitable that a great deal of destructive sampling of trees and shrubs, is required to provide initial reliable data for calibrating more rapid methods of biomass and production estimation.

For each type of browse community, dry weight measurements of the browse components must be obtained from adequate series of trees and shrubs. Such measurements should permit correlation regressions of browse biomass against more easily measurable values, such as canopy area (to be measured on aerial photographs), average crown diameter measured from samples on the ground and possibly phytometer readings.

A further problem which requires urgent solution is that of assessing the availability of browse forage to the browsing animal. What proportion of the total mass of foliage, shoots and fruits constitutes potential food and what proportion does not (by virtue of its inaccessibility or unpalatability)? To a great extent this question must be answered by observation of the feeding behaviour of browsing animals, taking account of seasonal changes in food requirements and preferences, and the constraints caused by drought conditions.

Tolerance levels to varying degrees of browse impact in each of the important browse species are virtually unknown. Future research should include studies of the viability of trees and shrubs under varying levels and regimes of browse offtake and under varying climatic conditions. The population dynamics of tree and shrub communities under browsing impact is a research subject of fundamental importance, especially where there is the possibility of ecological decline through overexploitation. Knowledge of tree population structures may be expected to lead to better assessment of the conservation status of browse plant communities and to practical estimates of sustainable browse offtakes and hence to livestock carrying capacities.

Existing data on the composition and nutritional values of browse species could provide a good starting point for a systematic programme of chemical analysis. It would provide for the chemical analysis of browse species from all eco-climatic zones and would also cover seasonal variation in composition and nutritional value.

The importance of browse as the main resource supporting the pastoral peoples of the arid zone appears to have been underestimated in the past. There can be little doubt that the rational development of range and livestock management in the dry savannas in the future will depend on a greatly increased fund of knowledge on the browse plants.


Barkham, J., Rainv, M (1976). "The vegetation of the Samburu-Isiolo Game Reserve". E. Afr. Wild. J. 14: 297329.

Bodgan, AN. (1968). "Some edaphic vegetation types at Kiboko, Kenya". J. Ecol. 46: 115126.

Brenan, J.P.M., Greenway, P.J. (1949). "Checklist of the forest trees and shrubs of the British Empire 5. Tanganyika Territory Part I". Oxford. Imperial Forest Institute.

Buechner, H.K., Dawkins, H.C. (1961). "Vegetation change induced by elephants and fire in Murchison Falls National Park, Uganda". Ecology. 42: 752766.

Burtt, B.D. (1939). "A field key to the savanna genera and species of trees, shrubs and climbing plants of Tanganyika Territory. Part I: genera and some species". (Revised, 1957, Welch J.R.). Dar-es-Salaam. Government Printer.

Burtt, B.D. (1942). "Some East African vegetation communities". J. Ecol. 30:65146.

Burtt, B.D. (1953, revised: Glover, P.E. and Jackson, C.H.N.) "A field key to the savanna genera and species of trees, shrubs and climbing plants of Tanganyika. Part II: The species of the more important genera with general index". Dar-es-Salaam. Government Printer.

Croze, H (1974). "The Seronera bull problem. II: the trees". E. Aft. Wildl. J. 12: 2948.

Dale, L.R. (1939). " The woody vegetation of the Coast Province of Kenya". I.F.I. Institute Paper 18.

Dale, I.R. (1940). "The forest types of Mt. Elgon". J.E. Afr Nat. Hist. Soc., 36: 7482.

Dale, I.R. and Greenways P.J. (1961). "Kenya trees and shrubs". Nairobi. Buchanan's Kenya Estates Ltd, in association with Hatchards, London.

Dayton, B.R. (1978). "Standing crops of dominant Combretum Species at three browsing levels in the Kruger National Park". Koedoe 21: 6776.

Dougal, H. W., Drysdale, V.M., Glover, P.E. (1964). "The chemical composition of Kenya browse and pasture herbage". E. Afr. Wildl. J. 2: 82125.

Dougal H.W., Sheldrick D.L.W. (1964). "The chemical composition of a day's diet of an elephant". E. Afr Wildl J. 2: 5159.

Douglas-Hamilton, M. (1952). "On the ecology and behaviour of the African elephant". Ph. D. thesis. Oxford.

Edwards, D.C. (1940). "A vegetation map of Kenya with particular reference to grassland types". J. Ecol. 28:377385.

Edwards, D.C. (1948). "Some notes on the food of goats in a semi-arid area". E. Afr. Agric. For. J. 13: 221223.

Eggeling, W.J. (1952). "Indigenous trees of the Uganda Protectorate, 2nd Edition revised by LR. Dale. Entebbe, Government Printer.

Field, A.C. (1979). "The impact of sheep-goats on the vegetation in the arid zone". UNEPMABIPAL Tech. Rep. E2.

Field, C.R. (1971). "Elephant ecology in the Queen Elizabeth National Park, Uganda". E. Afr. Wildl. J. 9: 99124.

Field, C.R. (1975). "Climate and the food habits of ungulates on Galatia Ranch". E. Afr. Wildl. J. 13: 203219.

Field, C.R. (1979). "The food habits of camels in northern Kenya". UNEP-MAB-IPAL. Tech. Rep. E. 16.

Field, C.R., Ross. LC. (1976). "The savannah ecology of Kedepo Valley National Park, Uganda II. Feeding ecology of elephant and giraffe". E. Afr. Wildl J. 14(1): 116.

Gillet, J.B. (1970), "A numbered check list of trees and shrubs and noteworthy lianes indigenous to Kenya". Nairobi, Government Printer.

Gillman. C. (1949). "A vegetation types map of Tanganyika Territory". Geog. Rev. 39: 737.

Glover, P.E. (1939). "A preliminary report on the comparative ages of some important East African trees in relation to their habitats". S. Afr. J. Sci. 36: 316327.

Glover, P.E. (1966). "An ecological survey of the Narok District of Kenya Masailand 19611965". Mimeo. S.L: S.N.

Goddard, J. (1968). "Food preferences of two black rhinoceros populations". E. Afr. Wildl. 6: 118.

Greenway, P.J. (1933). "The vegetation of Mpwapawa". J. Ecol. 21: 2843.

Greenway, P.J. (1962). "The vegetation of the Serengeti National Park". Typscript Rep. to Tanzania N.P.

Greenway, P.J. (1973). "A classification of the vegetation of East Africa". Kirkia. 9 (1): 168.

Greenway, P.J., Vesey-Fitzerald, D.F. (1969). "The vegetation of the Lake Manyara National Park". E. Ecol. 57: 127149.

Gwynne, M.D. (1969). 8 "The nutritive values of Acacia pods in relation to Acacia seed distribution by ungulates". E. Afr. Wildl. J. 7: 176178.

Harker K. W. (1959). "An Acacia weed of Uganda grasslands". Trop. agric. Din. 36: 4551.

Harrington G.N. and Ross, I.C. (1974). "The savanna ecology of Kidepo Valley National Park I: The effects of burning and browsing on the vegetation". E. Afr. Wild. J. 12: 93105.

Herlocker. D.J. (1969-1972). "Phenology of trees in the Serengeti National Park". Unpublished notes.

Herlocker. D.J. (1976). "Woody vegetation of the Serengiti National Park". Kleberg studies in Natural Resources, R.M. 5/KS1, Texas A & M University.

Herlocker, D.J. (1979). "The vegetation of South Western Marsabit District, Kenya". UNEPlMABlIPAL Tech. Rep. D.l.

Herlocker, D.J. and Dirshl, H.J. (1972). "Vegetation of Ngorongoro Conservation area, Tanzania, Can. Wildl. Serv. Rep. Ser. no. 19.

Hofmann RR and Scholtz, P. (1968). "Applied topographic anatomical studies of East African ruminants". E. Afr. Wild. J. 6: 107125.

Hofmann R.R. & Stewart, D.R.M. (1972). "Grazer on browser: a classification based on the stomach-structure and feeding habits of East African ruminants". Extrait de: Mammalia 36 (2): 226240.

Hornby, H.E. and van Rensburg, H.J. (1948). "The place of goats in Tanganyika farming systems I. In deciduous bushland formation". E. Afr. Agri. For. J. 14: 9498.

Ivens, G.W. (1967. "East African weeds and their control Nairobi, London O.U.P.

Jordan, S.M. (1957). "Reclamation and pasture management in the semi-arid areas of Kitui District, Kenya". E. Afr. Agric. For. J. 23: 8488.

Knight J. (1965). "Some observations on the feeding habits of goats in South Baringo District of Kenya". E. Afr. Agric. For. J. 30: 182188.

Lamprey, H.F. (1963). "Ecological separation of large mammal species in the Tarangire Game Reserve, Tanganyika". E. Agri. Wildl. J. I: 6393.

Lamprey, H.F. (1964). "Estimation of the large mammal densities, biomass and energy exchange in the Tarangire Game Reserve and the Steppe of Masai Tanganyika". E. Afr. Wildl. 2: 146.

Lamprey, H.F. (1975). "The distribution of protected areas in relation to the needs of biotic community conservation in Eastern Africa". IOCN Occasional Paper 16:

Lamprey. H.F., et al (1967). "L'invasion of the Serengeti Park by elephants". E. Aft. Wildl J. 5: 151166.

Lamprey, H.F., Halevy, G and Makacha, S. (1974). "Interactions between Acacia, bruchid seed beetles and large herbivores". E. Afr. Wild. J. 12: 8185.

Langdale-Brown, I. (1959a). "The vegetation of the Eastern Province of Uganda". Mem. res. div., Series 2 No. 1. Kampala, Dept. of Agric.

Langdale-Brown, I. (1959a). "The vegetation of Buganda". Mem. res. div. Series 2. No. 2. Kampala, Dept. of Agric.

Langdale-Brown, I. (1960a). "The vegetation of the West Nile, Acholi and Lango Districts of the Northern Province of Uganda". Mem. res. div. Series 2. No. 3. Kampala. Dept. of Agric.

Langdale-Brown, I (1960b). "The vegetation of the Western Province of Uganda". Mem. res. div. Series 2 No. 4 Kampala, Department of Agric.

Langdale-Brown, I. Osmaston, H.A. and Wilson J.G. (1964) The vegetation of Uganda. Government of Uganda.

Langdale-Brown, I. Trapnell C.G. (1962) "The natural vegetation of East Africa". In The natural resources of East Africa. Nairobi, E.W. Russell.

Leuthold, W. (1970). "Preliminary observations on food habits of gerenuk in Tsovo National Park, Kenya". E. Afr Widdl J. B: 7384.

Lind E.M. and Morrison M.E.S. (1974). East African vegetation. London, Longman.

Locust Handbook (1966). Anti-locust Research Centre, London.

Lucas, G.L. (1968) In conservation of Africa South of the Sahara: Kenya, Hedberg I. & Hedberg, O. Act. phyt. Suecica 54: 152159.

McKay, A.D. and Frandesen, P.E. (1969). "Chemical and floristic components of the diet of Zebu cattle (Bos indicus) in browse and grass range pastures in a semi-arid upland area of Kenya. I. Crude Protein". Trop. Agric. Trin.

Milne-Redhead, E. et al. (ongoing publication) Flora of Tropical East Africa. London.

Moomaw, J.C.A. (1960). "A study of the plant ecology of the Coast Region of Kenya". Nairobi: Government Printer.

Napier-Bax, P. Sheldrick, D.L.W. (1963). "Some preliminary observations on the food of elephant in the Tsavo Park (East) of Kenya". E. Afr. Wildl. J. 1: 4053.

Osmaston, H.A. (1968). "Conservation of vegetation south of the Sahara: Uganda". Acta phytogeogr Suec. 54: 148151.

Pellew R. (1979). "A simulation model of the Seronera Acacia tortilis wood, incorporating the impacts of fire, elephants and giraffe. Unpublished report.

Philipps, J.F.V. (1929). "Some important vegetation communities in the Central Province of Tanganyika Territory". S. Afr. J. Sce: 26: 332372.

Philipps, J.F.V. (1930). "Fire: its influence on biotic communities and physical factors in South & East Africa". S. Afr. J. Sci: 27: 352367.

Pielou, E.C. (1952). "Notes on the vegetation of the Rukwa Rift Valley, Tanganyika". J. Ecol. 40: 383392.

Polhill, R.M. (1968). "Conservation of vegetation South of the Sahara: Tanzania". Acta Phytogeog Suec 54: 166178.

Pratt, D.J. (1964). "Bush control studies in the drier areas of Kenya. Preliminary work with the Holt machines". Emp. J. Exp. Agric. 32: 1824.

Pratt, D.J. (1966a). "Bush control studies in the drier areas of Kenya, II, An Evaluation of the Holt TXa `Bush Breaker', on Tarchonanthus/Acacia thicket". J. Appl. Ecol. 3: 97-115.

Pratt., D.J. (1966b). "Bush control studies in the drier areas of Kenya III Control of Disoerma in semi-desert dwarf shrub grassland. J. Appl. Ecol. 3: 287291.

Pratt, D.J., Greenway, P.J., Gwynne M.D. (1966). "A classification of East African rangelands with an appendix on terminology" J. Appl. Ecol. 3: 36982.

Pratt, D.J., Gwynne, M.D. (1977). Rangeland management and ecology in East Africa. London, Hodder and Stoughton.

Pratt, D.J. and Knight, J. (1971). "Bush control studies in the drier areas of Kenya. V: Effect of controlled burning and grazing management on Terchonanthus/Acacia thicket". J. Appl. Ecol. 8: 217237.

Rutherford, M.C. (1978). "Primary production ecology in southern Africa". In Biogeography and ecology of Southern Africa, (ed. M.J.A. Werger). The Hague, W. Junk.

Scott, J.D. (1934). "The ecology of certain plant communities of the central province of Tanganyika Territory". J. Ecol. 22: 177229.

Staples, R.R., Hornby, H.E., Hornby R.M. (192). "A study of the comparative effects of goats and cattle on a mixed grassbush pasture". E. Afr. Agric. For. J. 8: 6270.

Taylor, C.R. (1969). "The eland and the oryx". Sci. Am. 220: 8885.

Thomas, D.B., Pratt, D.J. (1967). "Bush control studies in the drier areas of Kenya IV Effects of controlled burning on secondary thicket in upland Acacia woodlands". J. Appl. Ecol. 4: 325335.

Trapnell, C.G., et al (1969a) Kenya vegetation: sheet 3 (Mapsheet 1: 250,000). Tolworth, England. Directorate of Overseas Surveys.

Verdcourt; B. (1968). "Conservation of vegetation in Africa South of the Sahara: regional synthesis (Eastern Africa)". Act. Phytogeog. Suecica 54: 186192.

Verdcourt, B., Trump E.C. (1969). Common poisonous plants of East Africa. London, Collins.

Vesey-Fitzerald, D.F. (1973). "Browse production and utilization in the Tarangire National Park". E. Afr. Wildl. J. 11: 291306.

West. O. (1965). "Fire in vegetation and its use in pasture management, with special reference to tropical and subtropical Africa". Commonw. Bur. Pastures, Field crops. Farnham Royal, England.

Wilson, J.G. (1960). "The vegetation of Karamoja District, Northern Province of Uganda". Mem. res. Div Uganda Dep. Agric. Ser. 2-Veg. 5, 1-163. Entebbe, Department of Agriculture.

Wilson, R.N. (1957). "Studies of the browsing and reproductive behaviour of the East African dwarf goat". E. Afr Agric. For. J. 23: 138147.

Previous PageTop Of PageNext Page