Browse in Miombo woodland*

R. M. Lawton

Land Resources Development Centre, Overseas Development Administration, Foreign and Commonwealth Office, London

Introduction

Location and distribution

Environmental factors

Climate and phenology

Sociological factors

Accessibility of browse

An experiment in the utilization of browse

Comments on some common browse species

Differences between cattle and wildlife browsing

Suggestions for the utilization and management of browse

Conclusions

References

Introduction

In this review of the browse resource in miombo it is necessary to consider a number of ecological and sociological constraints, particularly the widespread presence of the tsetse fly (Glossina morsitans); the long dry season; and the lack of traditional skills in animal husbandry. It is not intended to discuss the important role played by the introduced pasture legumes, Stylosanthes guianensis, Macroptilium atropurpureum, Glycine wightii etc., in the utilization of miombo grazing. Nor is it intended to discuss the problems of bush encroachment, although some of these species may be of value as browse.

Location and distribution

Miombo is one of the most extensive woodland communities in Africa. It stretches from the south-western shores of Lake Victoria in northern Tanzania, through western, central and southern Tanzania to Mozambique; it covers the whole of Zambia, Malawi, and extends as far south as central Rhodesia; to the east it occurs in the Shaba (Katanga) province of Zaire and spreads across northern and central Angola. The distribution is shown on the vegetation map of Africa (Keay, 1959), which is at present under revision (White, in press).

Miombo forms a single storey, light canopy, deciduous woodland, usually growing to a height of 10-15 m, dominated by leguminous trees of the genera Brachystegia and Julbernardia. It may have a scattered shrub layer and has a ground cover of grasses and forbs that reach a height of about one metre. Brief descriptions of the woodlands are given for each country in Hedberg (1968), more detailed descriptions with an up-to-date comprehensive list of references are given by Werger and Coetzee (1978).

It is necessary to consider a number of associated woodland communities that occur within the broad complex of the miombo environment. In northern and central Tanzania, as the rainfall decreases in amount and duration, the miombo is replaced by Acacia, Combretum and Commiphora woodland associations. Sometimes there is a transitional zone where remnants of miombo are surrounded by the other associations.

In central Tanzania there is a mosaic of miombo and the distinctive Itigi thicket, in this case, soil differences determine the vegetative patterns.

In the Luangwa and Zambezi Valleys of Zambia, in the southern half of Rhodesia, and in western Angola, Colophospermum mopane woodland occurs on black cotton soils and on sandy soils with a hard pan near the surface. C. mopane is an important habitat for cattle and wildlife. The tree C. mopane is a valuable browse species.

In southern Zambia and Rhodesia, as the rainfall decreases in amount and duration, the miombo is replaced by open Acacia and Combretum woodlands. Dry Baikiaea plurijuga forests and associated thickets occur on Kalahari sands in the drier areas of south-western Zambia, south-eastern Angola and southern Rhodesia.

There are complex dynamic successional patterns within the miombo, e.g. the miombo-chipya-Marquesia evergreen forest complex of north-eastern Zambia (Lawton, 1978), and the miombo-chipya-Cryptosepalum evergreen forest complex of north-western Zambia (Cottrell and Loveridge, 1966). When Marquesia forest is cleared for cultivation it is replaced by open fire-hardy chipya vegetation which will then pass through a series of successional stages culminating in the restoration of the Marquesia forest. In addition there are edaphic associations related to topographical features, e.g. Brachystegia microphylla occurs on rocky hilltops often in association with B. bussei.

Environmental factors

Miombo and the associated woodland communities are the habitat of the tsetse fly, Glossina morsitans, the presence of which has a profound effect on the ecology and land-use potential of the woodland. The tsetse fly is the vector of the fatal disease trypanosomiasis, known in man as sleeping sickness and in cattle as nagana. The tsetse is a blood-sucking insect. Once it becomes infected with trypanosomes it will remain infected for the rest of its life, which may be up to 6 months, and during this period it spreads the disease to some of the animals on which it feeds. Wildlife carry the trypanosomes in their blood; although they are immune to the disease, they act as a reservoir of infection. Where there is a large wildlife population in miombo tsetse fly will be present, often in large numbers. The combination wildlife/miombo/tsetse fly is a natural ecosystem, it is unfavourable to the alternative ecosystem of human settlement/cultivation/pastoralism. If the miombo is cleared, wildlife and the tsetse fly will retreat due to the destruction of their habitat and to disturbance. The land can then be cultivated and used for grazing, provided the woodland is not allowed to regenerate and so create a favourable habitat for the re-invasion of the tsetse fly. Many examples of the dynamic relationships between the two ecosystems have been described by Ford (1971). The effects of the rinderpest outbreak in the 1980's that devasted wildlife and cattle populations, and other diseases like contagiuos bovine pleuropneumonia, on the distribution of the tsetse fly, have been discussed by Ford (1871). He also mentions historical factors, for example, in Rhodesia he observed, "that before the Ngoni invasion of the early ninetenth century, these fly-belts were much smaller than they had become by about 1870. Their expansion between 1830 and 1870 had been a response to depopulation brought about by famine and disease that followed the social disorganization produced by the Ngoni impact (Ford, 1971). Periodic droughts have also played a role in the ecology of the tsetse fly.

Fire has an effect on the ecology of miombo woodland. A fire early in the dry season favours woodland regeneration and so creates a habitat that is suitable for the tsetse fly, but a fire during the latter half of the dry season destroys the woodland and tends to create an open environment which is not favourable to the tsetse.

Climate and phenology

The miombo has a single rainy season lasting for five to six months, the rest of the year is dry. The rains usually start in late October or early November and finish in March or April. The average annual rainfall for the Lubumbashi region of Shaba province, Zaire, is given as c. 1 270 mm, with a range of from 716 mm to 1 551 mm (Malaisse, 1978), this range will cover the rainfall over most of the miombo. The dry season can be divided into two halves; from May to mid-August is the cool season, when temperatures may fall to below freezing at night and rise to about 21–28°C during the day; from mid-August until the onset of the rains in October or November is the hot season, when night temperatures do not fall below 10–15° and the day temperatures may reach 33–37°C. A heavy precipitation of dew may fall during the cool season.

All the Brachystegia species are deciduous, with the exception of B. taxifolia which is evergreen. The leaves become senescent at the end of the rains but may remain on the branches for much of the dry season, unless they are removed by strong winds, frost, or destroyed by fire. As soon as the temperature rises from about mid-August the Brachystegia and Julbernardia species, growing on deep soils, come into new leaf. The new flush is often a shade of pink or brown. Within about three weeks the leaves will turn green. Trees growing on shallow soils, particularly on steep escarpments, or in dry marginal miombo habitats, may not come into new leaf until the first storms have occurred, but generally over most of the miombo woodland, trees are in leaf at least two months before the start of the rains. It is during this time that browse is important.

Sociological factors

Some of the main tribal groups that live in the miombo have no tradition of animal husbandry. This may of course be partly due to the presence of the tsetse fly, but it may also be due to their origins, or to their method of cultivation. In central Tanzania the Nzamwezi live in miombo, they do not keep cattle, although the related Sukuma that live to the north are pastoralists as well as cultivators (Morgan 1972). At present the Sukuma are moving south into the miombo, where they graze agricultural fallow land following cultivation by the Nzamwezi.

The Bemba group that occupy north-eastern and central Zambia have no tradition of animal husbandry (Trapnell 1953), this may be due to their origin, because they migrated from the tropical forests of Zaire some two to three hundred years ago. Also the Bemba method of cultivation, known as chitimene, does not create grazing lands.

In southern and south-western Zambia the Tonga, I1a and Lozi are pastoralists and cultivators. They live in open Acacia Combretum woodlands and have access to extensive dry season floodplain grazing. The Tonga and I1a graze the Kafue Flats during the dry season and the open woodlands during the rains (Trapnell 1957). The Lozi graze the Zambezi floodplains during the dry season and the Baikiaea forest and thickets during the rains (Trapnell 1957).

Further south in Rhodesia the two main tribal groups, the Shona and the Ndebele, are pastoralists and cultivators. They wage an endless war against the tsetse fly and woodland encroachment.

It is clear that the main tribal groups in the miombo are traditionally non-pastoralists. Although the wildlife population and the accompanying tsetse fly have now disappeared from north-eastern Zambia, the Bemba have not yet become cattle owners. It may take many years before a non-pastoralist community acquires the desire and knowledge to become pastoralists.

Accessibility of browse

One of the first problems to consider is the accessibility of browse to cattle. The usual miombo canopy at a height of 10 m or more is out-of-reach, (Photo 1), although the natural browsing animals, particularly elephant and giraffe, can reach up to the canopy. Eland may use their horns to bring down some of the branches (Lightfoot and Posselt, 1977). The stems must therefore be pollarded at a height of about 1-2 metres (Photo 2). The pollarded branches soon grow out-of-reach (Photo 3) and will require cutting back every few years. The aim should be to develop a crown of small branches that will produce a large amount of foliage in the latter half of the dry season. Coppice regrowth from ground level (Photo 4) may be preferable, particularly if browsing induces a proliferation of shoots from the root stock.

Photo 1. A typical stand of Brachystegia-julbernardia or miombo woodland. North-eastern Zambia.. R.M. Lawton.

Photo 2. pollarded stem of Brachystegia floribunda. A suitable height for browsing. North-eastern Zambia. R.M. A Lawton.

Photo 3. Pollarded stems soon grow out-of-reach and need to be cut back frequently. North-eastern Zambia. R.M. Lawton.

Photo 4. Coppice re-growth in miombo, suitable for browse. Northeastern Zambia. R.M. Lawton.

From September to November the young leaves are high in crude protein and mineral content, and at this time of the year the browse is most palatable. Most of the common species are browsed by wildlife and cattle during this period. A list of published and known analytical data of the crude protein and mineral content of some of the common browse species is given in Annex 1.

Species	Crude protein content %	N %	P %	K %	Mg %	Ca %	Mn ppm	Carotene ug/g	Location	Reference	Remarks
Adansonia digitata	16.6	2.57	0.20	1.22	0.44	2.60	46	–	Central Tanzania	Lawton, 1979	Mature leaves
Aeschynomene fluitants	–	–	–	2.40	0.41	0.52	136	–	Kafue Flats, Zambia	Rees, 1978	No N or P analysis
Albizia adianthifolia	29.13	–	0.35	1.66	0.20	0.20	–	–	Copperbelt, Zambia	Lawton, 1968
A. amara	13.48	2.15	0.11	0.82	0.42	0.87	104	189	Central Tanzania	Lawton (in prep.)
	26.00	–	–	–	–	–	–	–	Kenya	5kerman, 1977
Afzelia quanzensis	18.63	2.98	0.30	2.28	0.29	0.43	29	186	Central Tanzania	Lawton, (in prep.)
Baphia bequaertii	22.38	–	0.31	1.40	0.17	0.17	–	–	Copperbelt, Zambia	Lawton, 1968
	23.50	–	–	–	–	–	–	–	Misamfu, Zambia	Rees, 1974
B. massaiensis	23.89	3.84	0.24	0.75	0.21	0.59	308	79	Central Tanzania	Lawton, (in prep.)
Brachystegia spiciformis	16.75	2.68	0.26	1.35	0.20	0.46	220	–	Central Tanzania	Lawton, 1979
	18.30	–	–	–	–	–	–	–	Misamfu, Zambia	Rees, 1974
B. boehmii	11.12	1.78	0.15	0.97	0.16	0.48	198	–	Central Tanzania	Lawton, 1979
Capparis tomentosa	19.94	3.19	0.12	5.06	1.44	1.89	100	162	Central Tanzania	Lawton (in prep.)	High K % checked
C. sepiaria	31.25	5.00	0.15	2.60	0.68	1.82	29	156	Central Tanzania	Lawton (in prep.)
Cassipourea mollis	19.50	3.12	0.22	1.38	0.38	0.68	985	21	Central Tanzania	Lawton (in prep.)
Colophospermum mopane	15.40	–	–	–	–	–	–	–	Rhodesia	Lightfoot and Posselt, 1977	Ranges from 9-20% during year
	13.1	–	0.27	2.04	0.21	0.52	–	–	Zambia	Lawton, 1968
Combretum apiculatum	13.2	–	–	–	–	–	–	–	Rhodesia	Lightfood and Posselt 1077	Mean CP %
C. fragrans	11.62	1.86	0.15	1.44	0.24	0.50	22	–	Central Tanzania	Lawton, 1979
C. obovatum	22.06	3.53	0.29	1.93	0.28	0.86	600	–	Central Tanzania	Lawton, 1979
C. purpureiflorum	22.68	3.63	0.20	1.50	0.22	0.76	710	–	Central Tanzania	Lawton, 1979
	22.37	3.58	0.33	1.68	0.21	0.55	500	133	Central Tanzania	Lawton (in prep.)
Dichrostachys cinerea	20.87	3.34	0.18	1.33	0.24	0.44	26	–	Central Tanzania	Lawton, 1979
	15.00	-	0.18	1.22	0.46	1.53	–	–	Zambia	Van Rensburg, 1968
										Skerman, 1977
Diplorhynchus condylocarpon	18.50	2.69	0.30	2.05	0.21	0.22	580	–	Central Tanzania	Lawton, 1979
	11.30	–	–	–	–	–	–	–	Zambia	Rees, 1974
Grewia spp.	22.50	3.60	0.22	1.98	0.34	2.37	339	227	Central Tanzania	Lawton (in prep.)
	17.10	–	–	–	–	–	–	–	Rhodesia	Lightfoot and Posselt, 1977	Mean - yearly range
											from 14.7-19.3
Julbernardia globiflora	16.37	2.62	0.25	1.31	0.20	0.54	80	–	Central Tanzania	Lawton, 1979
J. paniculata	12.38	–	0.24	1.50	0.25	0.70	–	–	Copperbelt, Zambia	Lawton, 1968
Justicia salvioides	44.06	7.05	0.33	3.30	0.40	0.80	342	196	Central Tanzania	Lawton (in prep.)	High N % checked
Markhamia obtusifolia	25.87	4.14	0.40	2.51	0.25	0.42	224	–	Central Tanzania	Lawton, 1979
M. acuminata	15.44	2.47	0.22	1.41	0.35	1.10	94	192	Central Tanzania	Lawton (in prep.)
Thylacium africanum	22.87	3.66	0.09	2.83	0.80	3.09	45	106	Central Tanzania	Lawton (in prep.)

An experiment in the utilization of browse

A series of investigations into the utilization of browse by cattle was carried out by Hood (1972) at the Misamfu Agricultural Research Station, near Kasama, north-eastern Zambia. As an initial investigation a herd was followed for two days each month from December to August and dates of the first record of browsing by species were noted (Annex 2). Browsing started in March when the grasses have become dry, unpalatable, and low in crude protein content. The twenty-nine species that were browsed make up the majority of the species in the miombo at Misamfu. This initial experiment clearly showed that cattle did utilize browse, so it was decided to plan a further experiment.

Species	Date of first record
Parinari curatellifolia Planch ex Benth	March 21
Baphia bequaertii De Wild	March 21
Vitex mombassaea yatke	March 28
Combretum molle R. Br. ex G. Don	March 30
Uapaca nitida mull. Arg.	March 30
Isoberlinia angolensis welw. ex Benth	March 30
Julbernardia paniculata Benth Troupin	March 30
Salacia rhodesiacaa Blakelock	March 30
Syzygium guineense (willd.) DC	March 30
Dichrostachys cinerea Wight & Arn	April 6
Piliostigma thonningii Schumach	April 6
Ficus ingensa (mig.) Mig	April 6
Entada abyssinica Steud ex A. Rich	April 6
Cussonia arborea Hochst ex A. Rich	April 6
Ochthocosmus lemaireanus De Wild & Th. Durand	April 29
Brachystegia spiciformis Benth	May 4
Rothmannia englerana (K Schum) Keay	May 4
Pterocarpus angolensis DC	May 5
Brachystegia utilis Burt Davey & Hutch	May 5
Albizia antunesiana Harms	May 11
Maytenus senegalensisa (Lam.) Exell	May 11
Combretum fragrans F. Hoffm.	May 11
Anisophyllea boehmira Engl.	May 18
Lannea discolor (Sond.) Engl.	May 18
Maprounea africana Mull. Arg.	June 7
Pavetta schumanniana F. Hoffm. ex K. Schum	June 7
Vitex fischerta Giirke	June 22
Protea sppa	July 14
Canthium sppa	Aug. 22

a Species only eaten by the bull.

Copied from Hood (1972) with corrections to botanical names.

The experiment consisted of four randomised blocks each containing six paddocks of 0.21 ha. Three treatments were duplicated in each block:

1. Untouched woodland.
2. All trees lopped at one metre above ground.
3. All trees removed except for a few left for shade.

The experiment ran for four years (1968–71), although the untouched woodland treatment was abandoned soon after the first year. A herd of four steers were placed in each paddock for two days and the paddock was then rested for 46 days. Each 48 day period was called a cycle and the cycles were continuous. The most palatable species, that is the species that were browsed each cycle, were: Baphia bequaertii, Dalbergia nitidula, Parinari curatellifolia.

Only a few plants of the Dalbergia and the Parinari were present in the paddocks, so the utilization of these two species has not been discussed. Some of species were browsed to a lesser extent each cycle, these were: Brachystegia spiciformis, Diplorhynchus condylocarpon, Julbernardia paniculata.

Others were only browsed during part of the dry season. Pericopsis angolensis (July-September), Brachystegia floribunda (April-September), B. utilis (March-September), Combretum molle (March-September), C. zeyheri (June-September), Isoberlinia angolensis (March-September), Vitex mombassae (April-September). September was the mean date of the cycle, but this cycle did extend over into October.

The most palatable species are browsed throughout the year, even during the rains, when the quality of grazing is good. One species, Baphia bequaertii, had a crude protein content of 28.1% in August and 23.5% in October (Rees, 1974), and on the copperbelt 22.38% was recorded in October (Lawton, 1968). Hood noted that B. bequaertii produced new leaves after each cycle of browsing, whereas other species were slow to produce new leaves. The branches of B. bequaertii became stunted and some of the trees died in the second year with the remainder succumbing in the third year. The cattle then turned their attention to Brachystegia spiciformis, the second most palatable species, and this was destroyed in the same way as the Baphia. The third species to come in for heavy browsing was the common Julbernardia paniculata, and by the end of the experiment this species was showing signs of loss of vigour and deaths were increasing.

Hood observed that in one season the new flush of J. paniculata was defoliated by caterpillars; such outbreaks are not infrequent in miombo. The details of tree species on which a number of miombo caterpillars feed has been given by Malaisse (1978). Many of these caterpillars are edible, and are a valuable source of protein for the local population. Defoliation by edible caterpillars in Colophospermum mopane woodland was noted by Rose-Innes (pers. comm.). It is concluded from Hood's experiment that the palatable miombo species will not survive if they are browsed frequently throughout the year.

Comments on some common browse species

Brachystegia spiciformis and Julbernardia globiflora are the two most widespread canopy tree species in the miombo. The young leaves of both species were browsed in Tanzania in November when their crude protein content was in excess of 16% (Lawton, 1979). Brachystegia spiciformis was killed by frequent browsing throughout the year (Hood, 1972).

With a crude protein content of over 22% in October and November, the small understorey tree, Baphia bequaertii, is probably the most palatable browse in the miombo. This species was killed by frequent browsing (Hood 1972). The shrub B. massaiensis is browsed in the Itigi thickets of central Tanzania, where a crude protein content of nearly 24% was recorded in leaves collected in December (Lawton, in preparation). From the Luangwa Valley, Zambia, it was noted that elephant browse this species in July (Gough, 1973).

Combretum fragrans is a common tree of the drainage lines, or mbugas, in central Tanzania. The young leaves were browsed in November when the crude protein content was about 11% (Lawton, 1979). In the Luanga Valley, Zambia, the old leaves were stripped off the trees by elephant in May and June (Gough 1973). Two very common Combretum climbers or scandent shrubs, C. obovatum and C. purpureiflorum, grow in the secondary miombo re-growth of central Tanzania. They were browsed in November and December when their crude protein content was over 22% (Lawton, 1979).

The multi-stemmed shrub Dichrostachys cinerea is a source of nutritious browse (20% crude protein, Tanzania, Lawton 1979), but it is inclined to form dense thickets and it has been suggested that it should be eradicated (Van Rensburg, 1968; Skerman, 1977). It is considered an indicator of overgrazed land (Pratt and Gwynne, 1977).

Diplorhynchus condylocarpon is widespread in miombo. It was browsed by cattle in November in central Tanzania (Lawton, 1979). Elephant browse it in August in the Selous Game Reserve and in central Tanzania south of Tabora.

Species of the genus Grewia are frequently a component of thickets and are often a source of browse. One species with a crude protein content of 22% was collected in Central Tanzania in December (Lawton, in peparation).

In the drier areas bordering miombo, particularly in Acacia woodland and in thickets, the family Capparidaceae contain a number of palatable species. The armed scrambling shrubs Capparis tomentosa, and C. sepiaria var. fischeri had a crude protein content of nearly 20% and over 31% respectively in December in central Tanzania (Lawton, in preparation). In the same area the small leaves of the shrub Maerua parvifolia (18% crude protein) were being stripped probably by small antelopes and goats. The dark green trifoliate leaves of Thylacium africanum (nearly 23% crude protein) were being browsed by cattle (Lawton in prep.).

Two species of Markhamia are browsed by cattle and wildlife; M. obtusifolia with a crude protein content of nearly 26% was collected in November in central Tanzania (Lawton, 1959). This species is also browsed by elephant and buffalo in the Luangwa Valley, Zambia. M. acuminata is a thicket species in central Tanzania, where it was browsed by cattle in December, when the crude protein content was 15% (Lawton, in preparation).

Quite a number of species in the Itigi thickets are a source of browse for wildlife and cattle that are passing through. The young leaves of one outstanding shrub, Justicia salvioides, had a crude protein content of 44% when collected in December. (The analysis was checked to make quite sure this figure was correct). This shrub is said to be a favourite browse plant of elephant, who also eat the green stems during the dry season, although these only had a crude protein content of 10% (Lawton, in preparation).

The tree Colophospermum mopane forms almost pure stands and provides a good supply of browse. In the Luangwa Valley, Zambia, elephant pollard the tree at a height of about one to two metres and so bring the browse within reach of other animals (Lawton and Gough, 1970). Elephant browse the old leaves in July and the new flush in November/December (Gough, 1973). In the dry areas of Rhodesia cattle partly subsist off C. mopane browse in the late dry season.

Only one exotic tree species, Leucaena leucocephala, has been planted for browse in the miombo (Bogdan, 1977; Skerman, 1977). The foliage has a crude protein content of about 28%, but if fed in excessive amounts, cattle will show symptoms of mimosine poisoning (Craufurd,1979). Leucaena may taint milk and meat (Lamb, 1978). It is however a prolific plant and can be cut and browsed frequently. Yields of up to 10 000 kg/ha dry matter per annum with a minimum protein content of 20% have been achieved (Craufurd, 1979).

More information on browse plants can be found in Skerman (1977), Bogdan (1977) and Pratt and Gwynne (1977).

Differences between cattle and wildlife browsing

If cattle are confined to small paddocks, as in Hood's experiment, the most palatable species will be destroyed by over-browsing, a similar situation may arise if wildlife were confined, but usually wildlife is free to range over a large area and so browsing pressure is light.

It was noted in the Luangwa Valley, Zambia, that elephant and other browsers, eat the senescent leaves of Colophospermum mopane and other species in July, that is just at about the time of natural leaf-fall, and again in November or December when the young leaves are fully expanded (Gough, 1973). As soon as fresh grass is available, the elephant graze, so the browse plants are able to put on new leaves and grow throughout the rains. Provided the dry grass is not destroyed by fire, elephant will graze during the dry season and supplement their diet with browse. Cattle may behave in a similar way if they were free to range over a sufficiently large area, but this is usually not the case. In a comparative study by Taylor and Walker (1978) it was concluded that cattle do not utilize the available forms of vegetation as fully as wild herbivores. It was noted that cattle only make limited use of browse species. The browsing patterns of wildlife would appear to be more effective and less destructive than those of cattle. Some wildlife browsing patterns have been described by Vesey-Fitzgerald (1973), Brown (1967), Guy (1976), Jarman (1971), Lamprey (1963), Rees (1978), and Rodgers (1976).

Suggestions for the utilization and management of browse

The most favourable time to utilize browse in miombo and in the neighbouring woodland associations is from September to November, when the young leaves have a high crude protein content and the grazing is at its poorest, or even non-existent. A number of grazing systems have been described by Pratt and Gwynne (1977), and it is suggested that the utilization of browse may be integrated into one or more of these systems, provided of course the various constraints, including the presence of tsetse fly, can be overcome.

It is suggested that during the rains from December to April the cattle should graze in a series of paddocks in which all the woody vegetation, apart from a few shade trees, had been removed. From May to August they should graze a similar series of paddocks in which pasture legumes, e.g. Stylosanthes guianesis, Macroptilium atropurpureum, etc., had been established. From September to November they would browse in a series of paddocks where most of the trees had been pollarded at a height of one metre, or cut at ground level to provide coppice regrowth. Throughout the system the paddocks would be grazed or browsed on a rotation.

An alternative method would be to harvest the foliage, dry and mill it, and feed it to the cattle during the latter half of the dry season. The leaves could probably be harvested in December and perhaps again during the rains. Harvesting will present problems, it may have to be done by hand, and during December most of the labour force will be growing food crops. The cost of drying large quantities of leaves may be expensive in equipment and power. The system may well be uneconomic, but the production of ground bush meal was tried experimentally in Rhodesia and was said to show promise (Bullock and Hunt, 1966).

Another alternative would be to grow some of the species with a high crude protein content, like Justicia salvioides, harvest the leaves and stall feed them to cattle.

Conclusions

Although miombo is an extensive vegetation community, it does not provide a favourable habitat for domestic livestock. The main constraints are, the presence of tsetse fly, the lack of traditional skills in animal husbandry, and a long dry season. Many of the canopy species are a source of browse, but they are often out-of-reach. When browse is made accessible by pollarding the stems, repeated browsing destroys the most palatable species. The main difference between natural browsers, i.e. elephant, giraffe, eland, kudu etc., and cattle, is that the natural browsers are not confined, whereas cattle are confined to either paddocks, or a range. The browsing pressure of wildlife therefore tends to be light, in comparison to that of cattle.

Browse is nutritious during the latter half of the dry season and it is suggested that it should be utilized at that time. It may be possible to incorporate it into a rotational grazing system. Alternatively leaves could be harvested, dried and ground into meal and stall-fed to cattle. Some of the most palatable species could probably be grown and harvested and fed to cattle.

References

Bodgan,A.V. (1977). Tropical pasture and fodder plants. London, Longman Group Ltd.

Brown, L.H. (1967). 'Problems of wildlife in the grazing areas of Africa'. Publications of the ITC-UNESCO Centre for integrated surveys. Delft, The Netherlands.

Bullock, E.J., Hunt, T.E. de la (1966). `Ground bush meal as a source of roughage for ruminants'. Zambia Commercial Farming. February pp. 16–17.

Cottrel, C.B., Loveridge, J.P. (1966). `Observations on the Cryptosepalum forest of the Mwinilunga District of Zambia'. Proc. Trans. Rhodesian Scient. Assoc., Vol 51, Salisbury, Rhodesia.

Craufurd, R.Q. (1979). 'Leucaena'. Farming in Zambia, Vol 12, No. 2. pp. 33–35.

Ford, J. (1971). The role of the Trypanosimiases in African ecology. A study of the tsetse fly problem. Oxford, Clarendon Press.

Gough, M. (1973). 'The Luangwa valley: an ecological study'. Oryx 12. pp. 223–226.

Guy, P.R. (1976). 'Diurnal activity patterns of elephant in the Sengwa Area, Rhodesia'. E. Afr. Wildl. J. Vol. 14 pp. 285–295.

Hedberg, I. and O. (Editors) (1968). 'Conservation of vegetation in Africa south of the Sahara'. Proc. Sym. 6th plenary meeting AETFAT Uppsala, 12–16 Sept. 1966. Act. Phytogeog. Suecica 54.

Hood, R.J. (1972). 'The development of a system of beef production for use in the Brachystegia woodlands of Northern Zambia'. Ph. D. thesis for University of Reading.

Jarman, P.J. (1971).'Diets of large mammals in the woodlands around Lake Kariba, Rhodesia'. Oecologia (Berl.) 8. 157–178 pp.

Keay, R.W.J. (1959). 'Vegetation map of Africa. South of the tropic of Cancer. (Explanatory notes)'. AETFAT/ UNESCO. London, OUP.

Lamb, R. (1978) 'A high speed tree for Africa'. West Africa (UK) No. 3197. pp. 2088–2087.

Lamprey, H.F. (1963). 'Ecological separation of the large mammal species in the Tarangire Game Reserve, Tanganyika'. E. Afric. Wildl. J. Vol. 11: pp. 63–92.

Lawton, R.M. (1968). 'The value of browse in the dry tropics' E. Afiic. Agric. For. J. Vol 33 No. 3 pp. 227–230.

Lawton, R.M. (1978).'A study of the dynamic ecology of Zambian vegetation'. J. Ecol. Vol 66 pp. 175–198.

Lawton, R.M. (1979). 'Woodland ecology reconnaisance survey'. Tabora Rural Development Programme. Tolworth UK. Land Resources Development Centre. Project Report 76. 24 p.

Lawton, R.M. (in prep.). 'Woodland ecology reconnaissance survey supplement'. Taora Rural Development programme. Tolworth UK. Land Resources Development Centre.

Lawton, R.M., Gough, M. (1970).'Elephants or fire — which to blame?' Oryx 10 (4): 244–248.

Lightfoot, C.J., Posselt, J. (1977). 'Eland (Taurotragus oryxy) as a ranching animal complementary to cattle in Rhodesia.' Rhodesia Ag. J. 74 (3): 53–61.

Malaisse, F. (1978). 'The Miombo ecosystem'. In: Tropical Forest Ecosystems. State-of-knowledge report prepared by UNESCO/UNEP/FAO/UNESCO Paris. pp. 589–606.

Morgan, W.T.W. (Ed.) (1972). East Africa, its peoples and resources. OUP. Second edition. 312 p.

Pratt. K.J., Gwynne, M.D. (1977). Rangeland Management and Ecology in East Africa. London, Hodder and Stoughton.

Rees, W.A. (1974). 'Preliminary studies into bush utilization by cattle in Zambia'. J. Appl. Ecol. 11 (1): 207–214.

Rees, W.A. (1978). 'The ecology of the Kafue Lechwe: the food supply.' J. Appl. Ecol. 15: pp. 171–191

Rodgers, W.A. (1976). 'Notes and Records. Seasonal diet preferences of impala from south-east Tanzania'. E. Aft Wildl, J. 14: 331–333.

Skerman, P.J. (1977. Tropical forage legumes, FAO plant production and protection series No. 2. Rome, FAO.

Taylor, R.D., Walker, B.H. (1978). 'Comparisons of vegetation use and herbivore biomass on a Rhodesian game and cattle ranch'. J. Appl. Ecol. 15 (2): 565–579.

Tapnell, C.G. (1953). The soils, vegetation and agriculture of north-eastern Rhodesia (second impression). Lusaka, Govt printer.

Trapnell, C.G., Clothier, J.N. (1957). The soils, vegetation and agricultural systems of north western Rhodesia (second edition). Lusaka, Govt printer.

Rensburg, H.J. Van (1968). Palatability of plants in Zambia. Ministry of Agriculture, Zambia. Lusaka, Govt printer.

Vesey-Fitzgerald, D. (1973). 'Browse production and utilization in Tarangire National Park'. E. Aft Wildl. J. 11: 291–305.

Werger, M.G.A., Coetzeen, B.J. (1978). 'The Sudano-Zambezian Region'. In Biogeography and Ecology of Southern Africa. Netherlands. Dr. W. Junk, the Hague.

White, F. (in press). The UNESCO/AETFAT Vegetation map of Africa.