Previous PageTable Of ContentsNext Page

II. Regional studies

  Review papers

A review of browse and its role in livestock production in southern Africa*

B.H. Walker

Department of Botany, University of the Witwatersrand, Johannesburg, South Africa


1. Introduction

2. Major vegetation types and dominant woody species

2.1. Determinants of vegetation structure and composition

2.2 Description of the major vegetation types

3. Biomass and seasonal production of woody vegetation

3.1 "Valley Bushveld" (Karroid) (Aucamp 1979, Aucamp et al, 1978).

3.2 Colophospermum mopane and arid shrub and tree savanna (Kelly and Walker, 1976)

3.3 Colophospermum mopane (Martin 1974)

3.4 Mixed tree and shrub savanna

3.5 Shrub vegetation of Kalahari sand deposits (Bailiaiae plurijuga vegetation type) (Rushworth 1975).

3.6 Dry miombo woodland (Martin 1974)

3.7 Alluvial vegetation along the Lutope River, Sengwa Research Area, Zimbabwe. Rainfall = c 500 mm (Goodman 1974)

4. Chemical composition and nutrient value of browse

4.1 Karroid vegetation

4.2 Arid shrub and tree savanna, and Colophospermum mopane

4.3 Mixed tree and shrub savanna

4.4 Shrub vegetation in the Baikaiae plurijuga region

4.5 The miombo woodlands

5. Utilization and palatability of woody vegetation

6. The role of browse in livestock production

Conclusion

References


1. Introduction

The aim of this paper is to present a review of the above-ground biomass and nutritive value of browse in southern Africa, and its role in livestock production. The coverage is necessarily limited, firstly because a detailed account of all the different vegetation types is simply too vast and, secondly, because for some regions the data are not available.

The review is therefore a selected coverage, restricted to those types and species deemed to be important either because they are very abundant or because they are palatable and valuable as a food source. The area covered (Figure 1) includes South Africa, Lesotho, Swaziland, South-West Africa (Namibia), Botswana, Zimbabwe and Mozambique. The map of the vegetation types distinguished for this review is derived from several sources, but chiefly from Acocks (1975) and Wild and Grandvaux-Barbosa (1967).

Figure 1. Woody vegetation types of Southern Africa.

S = Salisbury, B = Bulawayo, M = Maputo,

J = Johannesburg, D = Durban, W = Windhoek.

2. Major vegetation types and dominant woody species

Eight major vegetation types have been distinguished for the purposes of this paper. Seven are shown in Figure 1, and the eighth is the rather specialized and restricted alluvial vegetation of all the rivers in the savanna regions. The grasslands, macchia, forests, pans and swamps are excluded, since they are insignificant from the point of view of browsing ungulates.

The categories of vegetation are necessarily very broad and include many types which are in fact quite distinct. For example, Acocks (1975) gives a total of 72 main vegetation types for South Africa, of which 13 are savanna types, important for browse. He then distinguishes 56 sub-divisions of this bush and savanna vegetation on the basis of floristic composition.

2.1. Determinants of vegetation structure and composition

The range of inherently different vegetation types can be largely accounted for by variation along two main gradients-rainfall and soil type. Soil type ranges from deep, sandy infertile soils (such as the Kalahari sand deposits) at one extreme to shallow and heavy-textured soils at the other. There is a minimum development of woody vegetation under dry conditions with shallow, clay soils, and it becomes increasingly important (as indicated in Figure 2) along these two gradients.

Figure 2. Approximate distribution of the major types of woody vegetation according to rainfall and soil texture.

Deep, sandy soils with a rainfall greater than 700 mm give rise to closed woodland. In the drier regions (< 700 mm) an increase in the sand content of the soil leads to a marked increase in the ratio of the woody: grass vegetation.

Human disturbance is a third major factor which cuts across these two gradients and gives rise to quite different vegetation types under otherwise similar conditions. In particular, fire has been a major tool in removing or opening up woodlands, and it is necessary to consider its role.

Fire has been a natural feature of southern African savannas throughout their evolution (Phillips, 1965, West, 1965), and the species have evolved mechanisms to cope with it, and benefit from it. It is, nevertheless, a major determinant of both the amount of woody vegetation and the species composition, in that less fire-tolerant species increase in abundance in the absence of fire.

Natural fires are caused by lighting, and occur during the thunderstorms which mark the beginning of southern Africa's rainy season. They are, in fact, frequently doused by the rainstorms which follow. Man's use of fire has generally resulted in the vegetation being burned early in the dry season. The grass then flushes to produce a little green growth, which is heavily grazed. No root reserve replenishment can take place and the grasses die off again. This practice has greatly weakened the vigour of the grass sward, and resulted in an increase in the density of woody vegetation. Fires at the end of the dry season do minimal damage to grasses (which are still dormant) but can severely damage shrubs and trees, which have already flushed.

Three aspects of fire are important in the context of this paper.

i. Vegetation on sandy soils is much more tolerant to fire than that on heavy-textured, shallow soils. The former has a large, well-developed underground root system (up to 90% of the total biomass), and removal of the top-growth results in the removal of only a small proportion of the total plant nutrient reserves. In addition to the greater loss of reserves, the heavy-textured soils tend to form a surface cap when bare (Walker, 1974) and frequent fires, keeping the soil bare of litter, can cause a reduction in the proportion of rainfall which enters the soil.

ii. In arid and semi-arid areas fire generally has an adverse effect, and lack of fire does not lead to the development of woodland (due to the soil moisture balance).

iii. In moist savannas protection from fire leads to the accumulation of moribund grass and this increased fuel load results in intense fires, which damage all components of the vegetation. Frequent,  light fires keep much of the woody vegetation short, where it is available as browse. Furthermore, regrowth from woody vegetation after a fire (especially on sandy soils) is very vigorous, producing larger, softer and more palatable shoots. For example, Rushworth (1974) showed that the mean numbers of new coppice shoots per shrub, for six species in Kalahari sands, was 14.8 and 1.9 in a burned and unburned area, respectively.

2.2 Description of the major vegetation types

The following brief accounts are based on the more complete descriptions of Rattray (1961), Phillips (1969), Wild and Fernandes (1967), Acocks (1975) and Barnes (1979). In the lists of species under each type, the symbols T, S, TS and ST indicate that the species occurs as a tree, as a shrub, as a tree and to a lesser extent a shrub, and as a shrub and to a lesser extent a tree, respectively.

2.2.1 The Karoo and karroid vegetation

Acocks (1975) recognizes 38 sub-divisions of Karoo and karroid vegetation; a vast region (Figure 1) in which the production of sheep and goats is the main form of land use. In this account the mountain Karoo types are omitted, and the others are considered to comprise two main types. By far the larger is the Karoo proper together with the various other false (induced) Karoo types. The other is the rather specialized valley bushveld).

2.2.1.1 The Karoo

The Karoo is an arid, hot region with a rainfall ranging from 50 mm to 250 mm per annum. Much of the area receives the rain either in late summer or winter, enhancing the aridity of the summers.

The vegetation consists of a low, sparse cover of succulent and non-succulent shrubs, with little grass. The region has deteriorated markedly over the past 100 years, owing to overgrazing by sheep. In some areas the Karoo has advanced up to 250 km eastwards into what was previously grassland. In general, the boundary of the karroid vegetation, as given in Figure 1, is now in the vicinity of longitude 26° E, between latitudes 28° and 33° S. The original boundary, as estimated by Acocks (1975), was in the vicinity, of longitude 24° E, some 200 km to the west. The floristic composition of the whole region is maintained by the grazing pressure, with a generally disproportionately high number of succulents which have increased at the expense of the more palatable non-succulents. It is a very rich flora, and only a few of the more characteristic species are listed below.

Woody (W) and succulent (S): Pentizia spinescens (W), P. incana (W), P. globosa (W), P. lanata (W), Eriocephalus spinescens (W), E. ericoides (W), E. pubescens (W), Pteronia glauca (W), P. erythrochaeta (W), P. glomerata (W), Galenia fruticosa (W), Zygophyllum microphyllum (S), Sphalmanthus blandus (S), S. vigilans (S), Mesembryanthemum karrooënse (S), Psilocaulon utile (S), Ruschia multiflora (S).

Grasses: Aristida spp., Tragus racemosa, Eragrostis lehmanniana, E. obtusa, Stipagrostis ciliata, S. obtusa, Oropetium capense, Sporobolus fimbdatus, and many more.

2.2.1.2 Valley bushveld

This vegetation type is restricted to the southern parts of the Eastern Cape Province of South Africa, and along river valleys up the East Coast where rainfall is in the vicinity of 500 mm per annum.

It is a dense, shrub community with many succulents and a very subordinate grass layer. It has a total crown cover of c. 75% in the shrub layer (Aucamp, 1979). The dominant species are: Portulacaria afra, Capparis sepiaria, Rhigozum obovatum, Scutia myrtina, Brachylaena ilicifolia, Schotia afra, Pappea capensis, Grewia occidentalis, Euclea undulata, Ehretia rigida, Euphorbia bothae, Maytenus capitata.

Owing to the reduced grass layer, it is used mainly as browse, by improved strains of goats.

2.2.2 Arid shrub and tree savanna

A heterogeneous mosaic of many communities, occurring at low altitudes (< 700 m) on a variety of soils though mainly sandy loams. The rainfall is low, from 300–500 mm, and temperatures are high. This type of vegetation, together with the distinctive Colophospermum mopane (with which it often forms a mosaic) constitute those areas of Zimbabwe and South Africa which are known as the "lowveld".

As already mentioned, the soils range from sands to loams. With increasing clay content there is a transition to C. mopane. On the sandiest soils, small Terminalia sericea and Rhigozum sp. are dominant, and there is an increase in species of Acacia, Commiphora, Grewia and Combretum as the soil becomes more loamy. A single list of species is somewhat misleading, since some of the associations are quite distinct. nevertheless, the more abundant and important species are:

Woody: Grewiaflava (S), Adansonia digitata (T), Commiphora pyracanthoides (S), Terminalia serfcea (S), T. prunioides (S), Rhigozum obovatum (S), Sesamnothamnus lugardii (S), Combretum apiculatum (S T), C. imberbe (T S), Boscia foetida (S), Acacia mellifera (S), A. erubescens (S), A. tortilis (S T), A. nigrescens, A luederitzii (S), A. karroo (T S), Dichrostachys cinerea (S).

Grasses: Schmidtia pappophoroides, Eragrostis trichophora, Aristida congesta, Aristida spp., Enneapogon scoparius, Heteropogon contortus, Eragrostis superba, Bothriochloa radicans, Brachiaria nigropedata, Panicum maximum, Digitaria eriantha, Cenchrus ciliaris, Pterocarpus brenanii (S T), Dipkorhynchus condylocarpon (S T), Euclea undulata (S), Strychnos spp. (S T), Sterculia rogersii (T S).

2.2.3 Colphospermum mopane

It occurs in two main forms, namely woodland and tree shrub savanna. Well-developed woodland is restricted to the moister (c. 700 mm) Zambesi valley region and other, small isolated pockets. Of much greater significance for browse are mixed C. mopane tree and scrub savannas which occupy extensive areas in northern and central Botswana, the south and south-east of Zimbabwe, western and central Mozambique and the north-eastern Transvaal in South Africa.

It occurs almost exclusively on heavy textured soils, the few exceptions being sands underlain by clays, or saline sandy soils.

Mopane veld (as it is known in southern Africa) is generally regarded as valuable browse, but (for cattle) generally rather low in production. Details are given in later sections. The leaves of C. mopane itself have a very high tannin content when fresh, and are avoided by most ungulates. The dry leaves, however, are palatable and are picked up off the ground by cattle and wild ungulates. At irregular intervals of a few years outbreaks of the moth Gonimbrasia belina result in extensive areas of C. mopane being completely defoliated by the larvae.

The major species in this vegetation type are as follows:

Woody: C. mopane (T S), Acacia nigrescens (T S), A. tortilis (T S), Combretum apiculatum (T S), Kirkia acuminata (T), Dalbergia melanoxylon (T S), Commiphora africana S (T), Boscia albitrunca (T S), B. foetida (S), Cissus cornifolia (S), Dichrostachys cinerea (S), Ximenia americana S (T), Terminalia prunioides T (S), Grewia bicolor (S), G. flavescens (S), G. monticola (S).

Grasses: Bothriochloa radicans, Schmidtia pappophoroides, Cenchrus ciliaris, Enneapogon scoparius, E. cenchroides, Eragrostis rigidior, E. superba, Aristida spp.

2.2.4 Acacia associations

Acacia species dominate over a wide range of soil and rainfall conditions, but for these purposes four areas are distinguished. The herbaceous species vary from area to area but the dominants are essentially the same. They are therefore given in only the first of the four types.

2.2.4.1 The thornveld of south-west and central Zimbabwe, on clays and clay loams derived from basic rock with a rainfall from 500–600 mm. There are many species of Acacia, but the dominants are given below, followed by the most important associated species.

A. nilotica (T S), A. karroo (T S), A. gerrardii (T S), A. rehmanniana (S T), Combretum apiculatum (T S), Combretum apiculatum (T S), C. hereroense(S), Ormocarpum trichocarpum (S), Sclerocarya caffra (T), Ziziphus mucronata (T), Grewia spp, (S).

Grasses: Brachiaria nigropedata, Themeda triandra, Hyparrhenia spp., Heteropogon contortus, Cymbopogon plurinodis, Bothriochloa insculpta, Digitaria eriantha, Eragrostis spp.

2.2.4.2 The extensive tree/shrub savannas of southern Botswana and the north-eastern Cape, occupying much of the central part of the Kalahari region. The soils are sandier than in the previous type, and the rainfall is generally less than 500 mm, and frequently around 300 mm. The vegetation is short and scrubby, with scattered trees. The dominant woody species are:

Acacia erioloba T (S), A. leuderitzii (T S), A. mellifera S (T), A. hebeclada (S), A. karroo (T S), Boscia foetida (S), Commmiphora pyracanthoides S (T), Tarchonanthus camphoratus var litakunensis (S), Grewia flava (S), G. bicolor (S).

2.2.4.3 The central and northern Transvaall in South Africa occurring on heavy textured soils with a rainfall between 500 and 700 mm. The dominant woody species are:

Acacia tortilis (T S), A. nilotica (T S), A. karroo (S T), A. gerrardii (T S), A. robusta subsp. robusta S (T), Ziziphus mucronata (T), Grewia spp. (S), Ehretia rigida S (T).

2.2.4.4 The Zululand thornveld, east of the Drakensberg mountains along the coastal zone, again mainly on heavy textured soils, with a rainfall between 500 and 700 mm. There are many different Acacia species as well as a rich variety of associated sub-dominants, as follows:

Acacia nilotica (T S), A. karro (T S), A. caffra (T S), A. gerrardii (T S), A. robusta (T S), A. sieberana (T S), Auphorbia ingens (S T), Sclerocarya caffra (T S), Albizzia versicolor (T S), Dichrostachys cinerea (S), Maytenus senegalensis (S), Cussonia spicata (T S), Spirostachys africana (T S), Dombeya rotundifolia (S T).

2.2.5 Mixed tree and shrub savanna on sandy soils

This is the most heterogeneous of the vegetation types recognized for this analysis. It occurs on sandy to sandy-loam soils, at higher altitudes than the arid shrub and tree savanna (> 700 m), with the rainfall varying from around 450 to 750 mm. Acocks (1975) recognizes two distinct sub-divisions and these are perhaps worth noting.

2.2.5.1 Combretum apiculatum savanna

The soils are always very shallow and often stony. Small, dense, uniform trees (almost the junction of shrubs and trees) of C. apiculatum are dominant throughout. It characteristically occupies the ridges of undulating country, particularly in the drier and lower altitude regions. In this sense it fits more closely into the "lowveld" and mopane areas. It occurs over quite extensive areas on shallow, outcropping soils in a mosaic with the other sub-division of this type. The main woody species are:

Combretum apiculatum (S T), Acacia caffra (S T), Dichrostachys cinerea (S), Ximenia caffra (S), Lannea discolor (S T), Kirkia acuminata (T), Sclerocarya caffra (T).

2.2.5.2 Mixed Terminalia, Burkea, Combretum savanna

This is an amalgam of four sub-divisions of Acocks (1975) and various other communities recognized by Wild and Fernandes (1967). It is a very variable mixture and does contain some quite distinct associations. They all occur, however, on sandy to sandy-loam soils. The number of woody species is very large, and there are well over a hundred that rate as important. The most significant and abundant of these, together with the more important grasses, are:

Woody: Terminalia sericea (T S), Burkea africana (T S), Sclerocarya caffra (T), Combretum molle (T), C. apiculatum (T S), C. zeyheri (S T), Ochna pulchra (S), Grewia bicolor (S), G. flava (S), Grewia spp. (S), Mundulea sericea (S), Peltophorum africanum (T), Pterocarpus rotundifolia (S T), P.angolensis (T), Maytenus senegalensis (S), Pseudolachnostylis maprouneifolia (T S), Dombeya rotundiloia (S T), Piliostigma thronningii (T), Afzelia quantensis (T), Diospyros lycioides (S).

Grasses: Aristida graciliflora, Aristida spp. Eragrostis pallens, Eragrostis spp., Stipagrostis uniplumis, Anthephora pubescens, Ditigaria eriantha, Brachiaria nigropedata, Loudetia simplex, Schmidtia pappophoroides.

Of particular importance in this type is the presence in many sandy soil areas of the evergreen geophyte Dichapetalum cymosum. It contains monofluoracetate and is highly poisonous to ungulates. Wild herbivores do not eat it but cattle do, with the result that the areas where it occurs can only be used for cattle during the growing season (January–April), when it is completely over-shadowed by more palatable food.

2.2.6 Baikiaea plurijuga woodlands

B. plurijuga is the dominant tree occurring on the deep, aeolian sand deposits from the Kalahari which covered extensive areas of western Zimbabwe, northern Botswana and southern Zambia during the Pleistocene. Rainfall varies from c 600–1000 mm. The woodlands form a mosaic with patches of scrub, and the two components will be presented separately.

2.2.6.1 Woodland

The deep sands and relatively high rainfall give rise to a closed woodland with a canopy hight of around 10–15 m. The herbaceous and shrub layers are consequently greatly reduced in comparison to all the other types discussed thus far. B. plurijuga itself (known as "Rhodesian Teak") is useful as a timber tree but is unpalatable. These woodlands are therefore of little value as a source of browse, and serve mainly as sheltering areas for herbivores. The dominant woody species are:

Baikiaea plurijuga (T), Guibortia coleosperma (T), Burkea africana (T), Pterocarpus angolensis (T), Lonchocarpus nelsii (T), Dialium engleranum (T), Ricinodendron rautenenii (T), Terminalia sericea (T S).

2.2.6.2 Scrub

Patches of mixed scrub, containing a large variety of woody species, occur in areas where woodland cannot develop (soil depth or frost pockets), and wherever the woodland has been removed, either by felling for timber or by fire. In contrast to the woodlands the scrub areas form a valuable supply of browse, and the value of the woody species is enhanced by the fact that the grasses on these sands are particularly coarse and unpalatable. The most important species are:

Woody: Terminalia sericea (S), Baphia messaiensis (S), Combretum collinum (S), C. zeyheri (S), Bauhinia macrantha (S), Baikiae plurijuga (S), Ochna pulchra (S), Erythrophleum africanum (S), Croton pseudopulchellus (S), Acacia fleckii (S), A. ataxacdntha (S).

Grasses: Aristida graciliflora, A. stipitata, Eragrostis pallens, Digitaria eriantha.

The total number of species is very large. Rushworth (1975) recorded 56 woody and 172 herbaceous species in two sites covering a few ha.

The poisonous Dichapetalum cymosum described in the previous type also occurs throughout the Baikiaea area, and thus limits its usefulness for cattle ranching

2.2.7 The Miombo region

The miombo woodlands of central and southern Africa extend over a large area and include many different associations. Since this region will be dealt with in detail elsewhere in this symposium, it will be covered only briefly here. The characteristic genus of the region is Brachystegia. It does not extend south of around latitude 21 ° S, except for a coastal belt in Mozambique which reaches around latitude 26° S.

Rainfall varies between ± 700 mm and 1500 mm, and soils vary from granite sands to basalt and dolerite derived clays. At the dry end of the spectrum the vegetation is dominated by B. boehmii and a rich variety of other tree and shrub species. It is an open woodland with a well-developed herbaceous layer. With higher rainfall B. spiciformis dominates, together with Julbernardia globiflora and other tall trees. The vegetation is virtually a closed woodland with few shrubs and a sparse herbaceous layer. Where the ground water table is high a more open woodland results, usually dominated by Parinari curatellifolia.

Much of this region has been cleared of woody vegetation for arable crop production, or partially cleared and maintained in an open savanna or park-like form for cattle ranching.

The total number of woody species is very large, and those given below are only a representative sample, including the most abundant, and some of the few palatable species.

In general, the browse produced in the miombo region is less palatable to ungulates, and less valuable as a food source, than that in the drier vegetation types. With the exception of the "Chitemene" agriculture system of Zambia, and use by cattle of the new flush of leaves at the end of the dry season, it is not generally considered to make a significant contribution to livestock production.

Woody: Brachystegia spiciformis (T), B. boehmii (T), B. glaucescens (T), Julbernardia globiflora (T), Parinari curatellifolia (T), Diplorhynchus condylocarpon (T S), Combretum molle (T S), Cussonia spicata (T), Uapaca kirkiana (T), Monotes glaber (T), Faurea saligna (T S), Albizzia antunesiana (T S), Strychnos spinosa (T S), Vangueria infausta (T S).

Grasses*: Hyparrhenia spp., Hyperthelia dissoluta, Setaria sphacelata, Andropogon gayanus, Diheteropogon amplectens, Digitaria spp., Sporobolus pyramidalis, Brachiaria brizantha

*Come in where woodland is cleared.

2.2.8 Alluvial and riverine woodlands

These occur throughout the entire region, and collectively constitute a most important vegetation type. The size and development of alluvia increase with increasing length of the river, and consequently they are most significant in the lower altitude regions which have drained the higher plateaux. The rivers here are larger, and they are mostly in the arid regions. Alluvial vegetation is of particular importance in wildlife regions, where it supplies a vital food source during the dry season. The dominant trees are large and spectacular, and many of them produce edible fruits and pods. Some of the more important species are:

Woody: Cordyla africana (T), Xanthocercis zambesiaca (T), Kigelia africana (T), Acacia albida (T), A. xanthophlosa (T), A. nigrescens (T), Combretum imberbe (T), Trichelia emetica (T), Lonchocarpus capassa (T), Acacia tortilts (T), Combretum mossambicense (S), Schotia brachypetala (T), Figus sycamoris (T), F. capensis (T), Ficus spp. (T), Dispyros mespiliformis (T), Garcinia livingstonii (T), Salvadore persica (S), Canthium sp. (S), Balanites aegyptiaca (S). 

Grasses: Panicum maximum, Urochloa mossambicense, Digitaria spp., Setaria sphacelata.

3. Biomass and seasonal production of woody vegetation

Although there is a considerable amount of information on the biomass and productivity of the grass layer in all vegetation types in southern Africa, there is very little on the woody vegetation. It is unfortunate that in the many clearing and thinning experiments which have been conducted, the amount of woody material removed was not recorded. However, a few studies have been conducted which span the range of vegetation types being considered, and which give a reasonable overall view. Their results are presented below, according to the vegetation type in which they were conducted.

An important point must be made before any of the results from case studies are considered. The initial flush of leaves and twigs, which represents the major seasonal above-ground increase in biomass, is in fact not growth. It is a relocation of material stored in the roots and other storage organs during the previous season. Growth begins after the flush, and the current seasonal production of photosynthate is largely moved away from the growing points to branch, stem and roots. The common reference to early season growth and productivity is therefore misleading, and the results presented below should be interpreted with this in mind.

3.1 "Valley Bushveld" (Karroid) (Aucamp 1979, Aucamp et al, 1978).

The study was conducted in a particular area in which the succulent shrub Portulacaria afra made up 74% of the biomass. Total biomass was 50000 kg ha-1.

Biomass available as browse was 2000 kg ha-1 (4% of the total). This small percentage is due to the extremely dense and spiny nature of the vegetation.

Growth of the vegetation is slow, and P. afra took 275 days to recover to its original biomass after a 50% defoliation.

The rainfall in the Valley Bushveld is higher than in the Karoo proper, and the latter is now also much degraded through overgrazing and soil erosion. Biomass and seasonal production of the Karoo are therefore much less, though no data are available.

3.2 Colophospermum mopane and arid shrub and tree savanna (Kelly and Walker, 1976)

The area concerned fits best into the C. mapane vegetation type, but it grades into the arid savanna type and includes many shrub species from the latter. To some extent it therefore represents both of these types. The study involved a comparison of production on differently managed sites, and the results demonstrate that production varies strongly due both to inherent site differences (mainly soil depth) and to management.

Nine sites were examined under four degrees of utilization, viz. nil (a tsetse fly control corridor), light (game reserve), moderate (commercial cattle ranch) and intense (subsistence agriculture with cattle and goats). Above-ground biomass is given in Table 1.

Table 1. Above-ground standing crop in kg ha-1under different degrees of utilization in an arid shrub and tree savanna in southeastern Zimbabwe rainfall = c. 500 mm per annum.

 

Degree of Use

Nil

Light

Moderate

Intense

Above-ground biomass

Site No

1

2

3

4

5

6

7

8

9

Trees

 

18 101

  6 750

18 173

  9 766

25 623

21 728

28 246

8 967

20 556

Shrubs

 

  3 266

  1 976

  2 040

  1 338

  2 672

  3 214

  2 536

   800

  1 495

Total woody

 

21 367

  8 726

20 213

11104

28 295

24 295

30 782

9 767

22 051

Herbaceous (incl. sub shrubs)

 

  2 235

  1 581

  2 358

  1 843

  1 427

  1 057

  1 792

1 300

  1 201

TOTAL

 

23 602

10 307

22 571

12 947

29 722

25 999

32 574

11 067

23 252

Source: Kelly and Walker (1976)

Seasonal and daily production of leaves and twigs (from the first rain until maximum standing crop), which represents the total amount of browse produced, are presented in Table 2. The two years of data for the herbaceous layer are included in a comparative measure, and to indicate the large interseasonal variation. (It is interesting to note that the combined production of leaves and twigs is approximately equal to the production of herbaceous vegetation which accounted for 52%, on average, of total seasonal production). Despite having a much lower standing crop than the trees, shrubs contributed 61% of the total woody production.

Table 2. Arid shrub and tree savanna. Above-ground seasonal production (s p, kg ha-1 season-1) and daily productivity (d p, g-2 m d-1) of woody species leaves and twigs in 1971-72.

Degree of

Utilization

Site

Herbaceous

Shrubs

Trees

Total Woody

   

1970-71

1971-72

     
   

SP

DP

SP

DP

SP

DP

SP

DP

Nil

1

  650

2 084

781

.41

1 340

.71

2 121

1.12

 

2

  773

1 371

508

.27

   534

.28

1 042

.55

Light

3

1 165

2 347

503

.26

1 339

.70

1 842

.97

 

4

   965

1 821

334

.18

   806

.42

1 141

.60

 

5

1 117

1 422

681

.36

   954

.50

1 635

.86

Moderate

6

1 187

1 011

801

.42

   836

.44

1 637

.86

 

7

–

1 776

787

.41

   920

.48

1 707

.90

Intense

8

  262

1 294

172

.09

   422

.22

   594

.31

 

9

  228

1 181

340

.18

1 498

.79

1 837

.97

Source: Kelly and Walker (1976).

3.3 Colophospermum mopane (Martin 1974)

This area of C. mopane occurs in west-central Zimbabwe with a slightly higher rainfall (400-600 mm) on deeper soils. The vegetation is better developed than in the previous area and the total standing crop of woody material was estimated to be much higher, at 67 783 kg ha-1. Of this; however, only 2528 kg ha-1 was current growth (leaf and twigs), which is just slightly higher than the value obtained for the first site. 95.8% of the total biomass was wood.

The herbaceous production in this site varied from 345 to 1010 kg ha-1 between two successive years. It is at best less than half the leaf and twig biomass.

Shrubs (< 2.5 m) and fallen living trees made up c 3% of the total woody biomass. Much of the remaining 97% would therefore be unavailable to browsing ungulates.

3.4 Mixed tree and shrub savanna

These findings come from the South African Savanna Ecosystem Projects at Nylsvley, which is a large multidisciplinary study in the Northern Transvaal of South Africa. The site is dominated by Burkea africana, with Terminalia sericea and Ochna pulchra being very abundant. Soils are very sandy and rainfall is c. 700 mm per annum. Rutherford (1979a) has described the biomass structure of the woody vegetation in detail, and the following account is extracted from his work.

The overall structure of the biomass is given in Table 3. The percentage contributions of wood, twig and leaf exclude Grewia spp. (which have a very different morphology).

Table 3. Overall above-ground woody biomass of mixed tree and shrub savanna, dominated by Burkea africana, in sandy soil in the northern Transvaal, South Africa. Rainfall = 700 mm per annum.

Tree

Shrub

 

Total

Wood

13 493 (93.7%)

1 444 (77.4%)

14 937

Twigs

    135 ( 0.9%)

   101 ( 5.4%)

    236

Leaf

    773 ( 5.4%)

   327 (17.5%)

1 100

Total

14 402

1 827

16 273

Dead Wood

         ( 7.5%)

        (27.1%)

 

Source: Rutherford (1979a).

The total woody biomass of 16 237 kg ha-1 is somewhat less than the 21367 kg ha-1 described for the first C. mopane site, and much less than occurred in the second one, but Rutherford (op. cit.) describes a different Burkea africana community in South West Africa (Namibia), where total biomass was 22 290 kg ha-1.

A feature of all these savanna areas is the large spatial variation. Leaf area index of woody plants for the Nylsvley site varied from just less than 0.6 to 1.1 between adjacent transects.

It is again interesting to note that, as in the first savanna example, the combined production of leaves (1 100 kg ha-1) and twigs (236 kg ha-1) is approximately equal to the production of the herbaceous layer (c 1400 kg ha-1).

The contribution of shrubs to the total woody biomass was 12% for this Burkea savanna, and 11% for the first Colophospermus savanna.

3.5 Shrub vegetation of Kalahari sand deposits (Bailiaiae plurijuga vegetation type) (Rushworth 1975).

This study was conducted in the Wankie National park, Zimbabwe, in shrub vegetation dominated by Tenninalia sericea, but with many other species. The rainfall is around 600 mm per annum.

The values presented here are the averaged values of Rushworth's (op. cit.) burned and unburned sites, since they were in most cases not significantly different.

The total biomass and above-and below-ground biomass of the vegetation are presented in Table 4. Perhaps the most noteworthy feature of this vegetation is the very high proportion of the woody vegetation which is below ground. The below-ground measurements were furthermore only made to a depth of 0.5 m, so they are actually an underestimate.

Table 4. Biomass structure (in kg ha-1) of shrub vegetation dominated by Terminalia sericea within the Baikiaea plurijuga vegetation type in Wankie National Park, Zimbabwe. Rainfall= c.600 per annum.

 

Above ground

Below ground

Total

 

Woody

Grass

Woody

Grass

Woody

Grass

May

6 651 (14.2%)

1 049

40 184 (85.8%)

1 091

46 835

2 140

August

3 356 (10.0%)

   661

30 200 (90.0%)

   689

33 560

1 350

December

10 696 (17.1%)

   816

51 854 (82.9%)

   849

62 550

1 665

Mean

5 137 (13.8%)

   845 (49%)

32 088 (86.2%)

   880 (51%)

37 225

1 725

Source: Rushworth (1975).

For the above-ground woody vegetation, leaf biomass made up an average of 31.6% of the total, and stems made up 67.4%. This compares to corresponding values of 23% and 77% for the shrub layer in the Burkea savanna and around 20% and 80% for the Colophospermum savanna.

Above-ground woody biomass (mean of 5 225 kg/ha-1) is much less than for the other two savanna vegetation types, owing to the fact that there were no trees. With regard to leaf biomass, however, this vegetation type produced 1 810 kg/ ha-1, compared to c. 1 500 kg/ha-1 in the first Colophospermum community, 2 528 kg/ha-1 in the second, and 1100 kg/ha-1 in the Burkea community. If one considers leaves available to browsing ungulates (i.e. excluding trees) then these three figures drop to c 650, 76 and 327, respectively. Maximum leaf biomass of shrubs amounted to around 2 l/2 times the biomass of the herbaceous layer (c 700 kg/ha-1).

It is difficult to compare the biomass of the savanna types with that of the Valley Bushveld, as the latter is an unusually dense, woody community. The annual production of the Valley Bushveld, though, is low (c.f. the rate of regrowth), and the percentage with second C. mopane area, where most of the biomass was wood, in the form of large trees. In the savannas, unlike the Valley Bushveld, much of the leaf material below c 2.5 m is available.

3.6 Dry miombo woodland (Martin 1974)

This is the same study which produced the C. mopane biomass described in section 3 above. Rainfall is slightly higher (c 600 mm), and soils are sandy. The total biomass was 22 257 kg/ha-1 of which 21 161 kg/ha-1 was woody vegetation. The biomass of current woody growth (leaves + twigs) was 1 150 kg/ ha-1 (half that for the C. mopane), and the herbaceous production was 1096 and 1679 kg/ha-1 in two successive years. Shrubs (< 2.5 m high) and fallen, living trees made up c. 8% of the total woody biomass. Much of the remaining 92% would therefore be unavailable to ungulates.

3.7 Alluvial vegetation along the Lutope River, Sengwa Research Area, Zimbabwe. Rainfall = c 500 mm (Goodman 1974)

This study concerned vegetation production and feeding ecology of wild herbivores, and paid particular attention to the distribution of available plant biomass at different heights above the ground. The averaged results from three sites are presented in Table 5.

Table 5.Vertical distribution of available plant biomass (kg ha-1) in an open alluvial woodland in west-central Zimbabwe.

 

Herbaceous

Woody (ht. above ground)

 

Grass

Forbs

0–1m

1–2.5m

2.5–5m

>5m

Total Woody

May

3 107

730

119

268

163

1 141

1 691

Sept

2 842

513

105

252

345

1 153

1 855

Jan

3 663

450

259

482

389

1 124

2 254

Source: Goodman (1974)

Each plant was assessed for the proportion of its biomass which was available to a browsing ungulate, and the total amount adjusted accordingly. The values in Table 5 are therefore not comparable to those in the preceding studies. A significant finding of the study was that the woody vegetation supplies much less utilizable fodder than does the herbaceous layer (virtually all of which is available to grazers). If one considers that very few animals can use material above 2.5 m, then the woody vegetation supplies only around one-tenth of the amount of food available in the herbaceous layer. This helps to explain the relatively small proportion of the biomass of natural communities of wild ungulates which is comprised of browsers.

3.7.1 Estimation of woody biomass and production

In four of the studies just described regression equations were developed to estimate biomass and production from various measures, such as stem area, plant height, canopy volume, etc. While many of them are fairly similar in form, it is apparent that each species needs to be examined separately. There are too many for presentation in this paper, but most of them are either given in or referenced in the review by Rutherford (1979b).

It is again emphasized that no one has yet accurately measured woody vegetation production. The initial leaf flush does not constitute production and quantitative data for the amount of material translocated to the roots, and lost from there as a result of exudation, leaching and fine root mortality, are lacking for any African savannas. What has been measured is some proportion of this, in the form of above-ground biomass increments (reflecting previous season's effects), together with some rather dubious estimates of root mass (with little regard to species or live: dead ratios).

4. Chemical composition and nutrient value of browse

Information on the chemical composition of browse in southern Africa is highly selective, and scanty. A number of the references which are available are not worth including as they give values based on a single sample—which is completely inadequate and serves only to confuse the picture. In this review I have chosen not to present the data species by species, but rather by vegetation types. This is because, for southern Africa as a whole, the variation within a species (between different areas) can be far greater than between species within one community. A recent species by species review, giving data for 31 species, has been compiled as part of a countrywide assessment project for Botswana (DHV Consulting Engineers 1979). The review includes data from all over southern Africa, and highlights the variability between areas (and presumably analytical laboratories).

4.1 Karroid vegetation

Table 6 is a selection from the fairly comprehensive list by Louw et al (1968), giving only the more common species for which three or more samples were available. These authors present data on a further seven elements, but since the Karoo is only marginally included in this review, and is of least interest in this symposium, they are not presented. In general there are very big differences between species in all chemical components. Crude protein varies from around 5% to 20%, and is generally about 3% higher (in the same species) in summer than it is in winter.

Table 6. Average chemical composition of selected north-western karoo species.

In the second form of Karroid vegetation the valley bushveld, Aucamp (1979) has shown that crude protein ranges from a minimum of 10.5% in September–November to 14.5% in February–May. In another study, Aucamp et al (1978) used oesophageal fistulated Angora and Boer goats to determine the quality of the food they selected. The composition of their diets did not vary much over a year, and the mean values are presented in Table 7. Crude protein varied from 10.1% (October) to 17.1% (March) for Boer goats, and from 9.4% (October) to 17.8% (March) for Angora goats. Metabolizable energy remains high all year, and the selected diet allows for high animal performance all year.

Table 7. Food quality of valley bushveld browse as selected by Boer and Angora goats.

 

CPa

(%)

DCPb

(%)

CEc

(MJ kg-1')

DEd

(%)

MEe

(MJ kg-1)

DDMf

(%)

Boer goat

12.6

67.5

19.6

44.6

5.9

44.0

Angora goat

12.3

67.0

19.3

46.5

6.2

46.1

aCrude protein, bdigestible crude protein, ccrude energy, ddigestible energy, emetabolizable energy, fdigestible dry matter. Source: Aucamp et al (1978).

4.2 Arid shrub and tree savanna, and Colophospermum mopane

Barnes (1979) presents data on the crude protein and digestion coefficients for same common browse species in southeastern Zimbabwe. They are presented here in Table 8.

Table 8. Crude protein of leaves and twigs and digestibility of leaves of common browse species in an arid shrub savanna in southeastern Zimbabwe.

Species

Crude protein (% of dry matter)

Digestion coefficient for leaves (%)

 

Leaves

Twigs

Crude Protein

Dry Matter

Melhania acuminata

15.9

6.9

71.1

48.6

Grewia flavescens

15.1

6.9

70.1

53.5

Grewia sp.

14.4

7.4

69.2

35.3

Combretum apiculatum

12.8

6.5

66.6

47.4

Colophospermum mopane

12.3

5.0

65.6

41.3

Commiphora mollis

10.9

5.1

62.4

46.1

Source: Barnes (1979)

Two points are particularly noteworthy: Firstly, the crude protein of the leaves is generally high, and usually more than twice that of the twigs. Secondly, the digestibility of the leaves as a whole (determined by in vitro digestion) is low. Considering that the digestibility of the crude protein content is quite high, it means that the digestibility of the remaining constituents is even lower than the values given under "dry matter" would suggest. According to Barnes (op. cit.) this indicates that these browse plants are primarily useful as a source of protein, and not as a source of energy.

Further data for this vegetation type are presented in Table 9. Crude protein values for these species are comparable to those in Table 8, and demonstrate the marked seasonal changes. Protein and mineral elements are in highest concentration in the early growing season (November to January), and are lowest at the end of the dry season. The January and May data are included for comparison with the data in subsequent tables.

Table 9. Chemical composition of selected browse species from arid and mixed savanna in the Northern Transvaal, South Africa.

Species

%

Crude protein

%

Ether extract

%

Crude Fibre

%

CA

%

P

Acacia erioloba

12.1

1.63

32.0

0.99

0.26

Boscia albitrunca JAN

17.0

 

32.5

1.61

0.12

                         MAY

13.4

 

31.5

1.10

0.07

                        MIN

12.5

 

23.7

0.84

0.06

                       MAX

18.8

 

37.3

1.61

0.12

Colaphospemum mopane              JAN

13.7

 

28.1

1.51

0.19

                        MAY

11.4

 

25.6

2.28

0.20

                      MAX

16.6

 

28.1

3.23

0.23

                    MIN

8.4

 

21.9

1.15

0.12

Combretum apiculatum       JAN

14.0

 

24.7

1.36

0.24

                  MAY

10.3

 

31.3

1.56

0.12

              MAY

15.2

 

45.1

2.63

0.24

         MIN

5.1

 

34.7

1.08

0.06

Dichrostachys cinerea

10.4

 

26.7

 

0:45

Acacia nilotica

12.9

 

15.2

   

Grewia flava

11.5

 

30.7

1.56

0.14

Grewia occidentalis

13.2

 

31.0

1.96

0.14

Olea africana

10.9

 

22.6

3.39

0.21

Tarchonanthus camphorates

10.6

 

25.5

2.00

0.30

Rhus lancea

12.9

 

22.3

   

Source: Bousma (1976)

4.3 Mixed tree and shrub savanna

There are no comprehensive data for this vegetation type, although many of the species occur in other vegetation types as well and have been analysed there. Bate (1979) has determined the nitrogen content of species in the Burkea savanna of the Northern Transvaal. By multiplying these values by 6.25, the crude protein of the leaves can be obtained. The seasonal pattern for the three dominant woody species is presented in Table 10.

Table 10.Seasonal change in percentage crude protein of the leaves of the three dominant species in the Burkea (mixed tree and shrub) savanna, South Africa.

 

% Crude protein

 

Oct–Dec

Feb–April

July–Aug

Burkea africana

15

12

6

Ochna pulchra

15

9

3

terminalia sericea

9

9

2

Source: adapted from Bate (1979)

4.4 Shrub vegetation in the Baikaiae plurijuga region

For the shrub communities described in Sections 2.6 and 3.4, the overall chemical composition of the leaves + twigs is given in Table 11. The values are comparable with those given earlier, but it is interesting to note the very high levels of crude protein in September. This point will be dealt with in the next section.

Table 11. Overall chemical composition of shoots (leaves + twigs, as % dry wt.) In shrub vegetation on Kalahari sands, Zimbabwe. The range is given in parentheses.

Month

Ash

Crude

Protein

Ether

Extract

Crude

Fibre

N-Free

Extract

Ca

Mg

K

P

February

4.7


(1.6–9.0)

16.8

(9.5–23.9)

2.9

(0.8-4.4)

26.0

(17.0-37.4)

49.7

(35.6–61.2)

0.94

(0.36–1.45)

–

–

–

–

0.19

(0.14–0.30)

May

4.3

(2.2–6.7)

14.7

(10.6–22.4)

2.8

(1.0–7.9)

28.9

(15.8–40.8)

49.0

(29.9–68.9)

0.88

(0.27-1.69)

0.24

(0.13–0.40)

0.93

(0.5–1.7)

0.10


(0.04–0.15)

September

4.3
(2.4–6.1)

22.0
(12.8–39.4)

1.9
(0.8–4.0)

23.0
(15.1–34.6)

48.8
(28.6–67.6)

0.71
(0.17–1.61)

0.22
(0.11–0.3)

1.25
(0.6–2.26)

0.26
(0.17–0.40)

January

4.2

(1.6-8.6)

17.4

(12.6–24.8)

2.7

(1.1–7.31

31.5

(17.3–40.5)

44.3

(29.1–57.5)

–

–

–

–

–

–

–


–

Source: Rushworth (1975)

The composition of shoots (twigs + leaves) of individual species in the dry and wet season is given in Table 12. Marked differences occur between species in both the macro-components and the individual elements. In particular, the percentage of Ca is very variable. The composition of the stems of these same species is given in Table 13. Crude protein values are still fairly high, though much less than for leaves and twigs. Ether extract is lower, as is the N-Free extract, but the percentage crude fibre is much higher. The concentration of Ca in the stems is also higher than it is in the leaves.

Table 12. Chemical composition of shoots (leaves + twigs) of selected species from the shrub vegetation in table 11. D= dry season, w = wet season.

Species

Season

% Ash

% Crude protein

% Ether extract

% Crude fibre

% N-Free extract

% Ca

% Mg

% K

% P

Acacia ataxacantha

D

3.5

14.2

4.4

15.8

62.1

0.93

0.28

0.76

0.10

 

W

4.6

18.7

1.9

17.3

57.5

       

A. fleckii

D

6.7

18.5

3.5

21.1

50.2

1.61

0.36

0.83

0.10

 

W

5.2

23.6

2.5

22.5

46.2

       

Baikiaea plurijuga

D

2.3

20.4

2.9

28.7

45.7

0.28

0.13

0.93

0.15

 

W

2.6

21.0

3.2

40.0

33.2

       

Baphia massaiensis

D

3.1

19.0

1.9

33.0

43.0

0.57

0.16

1.02

0.10

 

W

2.8

17.9

2.0

32.8

44.5

       

Bauhinia macrantha

D

4.9

13.3

1.9

25.1

54.8

1.49

0.31

0.80

0.10

 

W

3.3

17.7

1.6

30.7

46.7

       

Burkea africana

D

2.2

12.5

1.5

27.1

56.7

0.46

0.14

0.78

0.07

 

W

1.6

14.9

1.9

33.9

47.7

       

Combretum collinum

D

6.1

13.2

2.5

20.5

57.3

1.69

0.35

0.88

0.09

 

W

4.3

14.9

3.4

21.6

55.8

       

C. Zeyheri

D

4.9

11.3

2.9

28.8

47.9

0.92

0.23

0.98

0.09

 

W

3.6

16.5

2.2

26.3

51.4

       

Croton pseudopulchellis

D

4.2

11.2

7.9

26.1

50.6

0.7

0.19

1.7

0.11

 

W

6.4

12.6

5.0

25:5

50.5

       

Erythrophleum africanum

D

2.0

14.8

6.5

34.3

42.4

0.27

0.18

0.72

0.09

 

W

2.0

16.3

7.3

37.6

36.8

       

Lonchocarpus nelsii

D

5.7

22.4

1.5

40.5

29.9

0.92

0.20

1.18

0.13

 

W

5.0

24.8

2.3

38.8

29.1

       

Ochna pulchra

D

2.4

10.6

1.0

36.0

50.0

0.75

0.15

0.53

0.09

 

W

1.7

12.9

1.4

40.5

43.5

       

Pterocarpus angolensis

D

4.2

14.8

1.2

31.8

48.0

0.95

0.40

1.09

0.10

 

W

3.9

20.4

2.4

38.4

34.9

       

Terminalia sericea

D

3.4

10.6

1.3

15.8

68.9

0.63

0.22

0.82

0.08

 

W

2.7

17.2

3.4

22.7

54.0

       

Dichapetalum cymosum

D

6.7

15.9

1.9

36.6

38.9

0.83

0.29

1.09

0.08

 

W

8.6

14.3

1.1

34.7

41.3

       

D. rhodesicum

D

6.6

12.6

1.7

40.8

38.3

1.09

0.23

0.88

0.04

 

W

8.4

14.4

1.9

39.9

35.4

       
                     

Table 13. Chemical composition of the stems of the species included in table 12.

Species

% Ash

% Crude protein

% Ether protein

% Crude fibre

% N-Free extract

% Ca

% Mg

% K

% P

Acacia ataxacantha

3.1

15.8

1.8

41.7

37.6

2.62

0.14

0.58

0.10

A. fleckii

4.5

11.0

2.3

53.7

28.5

4.94

0.39

0.34

0.05

Baikiaea plurijuga

2.5

8.3

2.7

49.5

37.0

1.30

0.19

0.66

0.12

Baphia massaiensis

2.6

10.8

2.5

47.7

36.4

1.31

0.08

0.83

0.08

Bauhinia macrantha

3.4

5.7

2.5

50.2

38.2

2.44

0.22

0.48

0.06

Burkea africana

1.9

8.6

2.5

46.1

40.9

0.98

0.15

0.56

0.10

Combretum collinum

4.7

6.9

2.2

47.5

38.7

4.49

0.32

0.53

0.09

Combretum zeyheri

3.0

5.1

2.8

55.1

34.0

2.52

0.19

0.49

0.08

Croton pseudopulchel-lus

2.4

7.6

2.1

65.3

22.6

1.50

0.14

0.56

0.06

Erythrophleu-m africanum

2.5

8.9

1.3

37.8

49.5

0.38

0.14

0.68

0.08

Lonchocarpus- nelsii

6.3

14.4

2.4

50.0

26.9

3.60

0.22

1.15

0.15

Ochna pulchra

3.3

8.3

1.7

41.4

45.3

1.34

0.18

0.60

0.10

Pterocarpus angolensis

3.9

8.5

2.2

38.3

47.1

0.51

0.16

1.39

0.05

Tenninalia sericea

4.1

4.0

0.9

62.4

28.6

2.64

0.17

0.60

0.05

Dichapetalum cymosum 5.7 11.1 1.1 48.9 33.2 0.70 – 1.4 0.10
Dichapetalum rhodesicum 2.7 14.0 1.5 45.9 35.9 0.60 – 1.1 0.20
Overall mean % 3.5 9.3 2.0 48.8 36.3 1.98 0.19 0.75 0.09
Range % 1.9–6.3 4.0–15.8 0.9–2.8 37.8–65.3 22.6–49.5 0.3–4.94 0.08–0.39 0.34–1.40 0.05–0.2

4.5 The miombo woodlands 

Considering the extent of this vegetation type, in which woody plants are so dominant, there are surprisingly few data on the chemical composition of the species concerned. Rees (1974) has presented the most comprehensive set, for the Northern Province of Zambia. Table 14 presents a summary of her findings.

Table 14. Digestible Dry Matter (DDM) and Crude Protein (CP) content of browse selected by cattle in a Miombo region in northern Zambia

Species

Growth

August

October

   

DDM (%)

CP (%)

DDM (%)

CP (%)

Baphia bequarti

new

61.5

28.1

57.1

23.5

Brachystegia glaberrima

old

33.8

9.6

   
 

new

   

34.4

12.0

B. spiciformis

old

59.6

10.6

   
 

new

   

52.1

18.3

B. utilis

old

54.7

10.2

   
 

new

   

46.4

17.8

Diplorhynchus condylocarpon

old

57.7

  7.6

   
 

new

   

62.7

15.2

Julbernardia paniculata

old

54.8

11.1

   
 

new

61.0

12.2

57.4

11.3

Monotes spp.

old

55.7

  7.7

   
 

new

   

41.8

12.3

Ochthocosmus lemaireanus

old

38.2

  9.1

   
 

new

45.9

11.3

66.4

15.3

Parinari curatellifolia

old

42.1

  9.2

   
 

new

49.9

10.2

39.0

11.2

Uapaca nitida

old

51.5

  8.9

   
 

new

54.9

11.9

64.6

18.9

Grass

   

  3.0

 

  3.0

Source: adapted from Rees (1974)

The results are similar to those for the arid savanna and Baikiaea shrub communities. Rees (op. cit.) uses Alderman's (1966) relationship to derive metabolizable energy (ME) from digestible dry matter (DDM). The values of ME range from 1.23 to 2.41 kcal kg-1

In general, protein content of southern African browse is high, but only around half of the dry matter is digestible. Mineral content is high, with K or Ca being highest at around 1 to 1.5%, followed by Mg at around 0.2 to 0.4%, and then P at around 0.1%.

5. Utilization and palatability of woody vegetation

As demonstrated in the previous section, the proportion of woody vegetation made up by leaves and new twigs is very low. In addition, owing to the morphology and size of woody plants, only a relatively small proportion of the leaves and twigs are available to browsing ungulates. The use of total woody biomass is consequently generally less than 1%. In shrub communities, however, the use of current annual growth can be much higher, and there is evidence that high levels of use have a deleterious effect on woody plant growth.

Taylor and Walker (1978) compared the cattle and game sections of a ranch in an arid shrub and tree savanna in southeastern Zimbabwe, and showed that a mixed population of browsing and grazing game animals resulted in a significantly lower density of small woody plants (3.5 m–2) than occurred in the same plant community stocked with cattle and a few game animals (4.2 m–2). In three different plant communities, the mean percentage utilization of all available browse was 7.9% in the cattle section and 17.2% in the game·section. Donaldson (1978), in a C. mopane savanna in Southern Africa, has also shown that goats control the density of woody plants, allowing a greater biomass of grass for cattle production.

There are marked differences in the palatability of woody species, and the mean utilization value of 17.2% just quoted was made up of values ranging between 0 and 64%. It is significant that the less abundant species were utilized most, suggesting that at high stocking rates, browsing animals may alter the woody species composition to something which is less favourable to them. Martin (1974) showed an extreme case in a C. mopane community, where there were 1.6 kg/ha available of the shrub Combretum mossambicense, of which 86% was utilized, and over 2000 kg/ha–1 of C. mopane available, of which about 1% was utilized.

Palatability of the same species varies from one vegetation type to another, and the literature is full of apparent contradictions as to which species are palatable (and therefore valuable) and which are not. Nevertheless, there is a general tendency for the most palatable species to be selected wherever they occur. The total list of woody species which have been recorded as being eaten by large herbivores is by now very long, and it would serve no useful purpose to present it. I have therefore eliminated those species which have a very restricted distribution, or are very rare, or are generally unpalatable. Some species of low palatability are included because they are very abundant, and therefore still contribute significantly to herbivore diets (e.g. Brachystegia spiciformis).

The indicated palatabilities in Table 15 can only be used as a rough guide. Some authors (e.g. Field, 1978) claim quite different results, but I have based them on a general survey of the literature (particularly Taylor, 1973, Martin, 1974, Grunow, 1980, and the authors they quote) and on my own observations. The list undoubtedly needs to be enlarged, and the indicated palatabilities refined. The list is based on the palatability of the leaves and twigs since these provide by far the bulk of the food. Some species, however, are particularly important because of their fruits or seeds. Notable amongst these are Acacia erioloba, A. albida (large pods) and many other Acacia species, such as A. nilotica, which have very palatable, smaller pods. The large fleshy fruits of Sclerocarya caffra, Uapaca kirkiana and Ficus species contribute a substantial amount but for only a very short period. The smaller fruits of species such as Ziziphus mucronata and the Grewia species dry out and remain on the branches, providing a smaller but prolonged food source.

Table 15.The more important Browse species in southern Africa, excluding the Karoo and Succulent

3 = very palatable, 2 = palatable, l = eaten but not selected 0 = unpalatable, v= occurs in the vegetation type but browse value not known. Two symbols separated by a / indicate that both degrees of palatability have been noted.

Vegetation type

Species

Arid savanna

C. Mopane

Acacia savanna

Mixed savanna

Baikiaea shrub

Miombo

Alluvium

Acacia albida

           

2/3

A. ataxacantha

v

 

v

1

1

 

V

A. erioloba

   

2

2

   

1

A. karoo

2

 

2

2

  2  

A. nigrescens

1

1

1

     

V

A. nilotica

   

1

1

     

A. tortilis

2

2

2

2

   

3/2

Albizzia harveyi

1

 

v

1

   

2

Balanites aegyptiaca

2

1

         

B. maughamii

2

2

       

3

Baphia massaiensis

       

1

   

Bauhinia macrantha

     

1

2

1

 

Boscia albitrunca

2

2

v

       

Brachystegia spp.

         

1/0

 

Burkea africana

     

1/0

1

1/2

 

Canthium gilfillani

      3      

Capparis tomentosa

1/3

1

 

v

   

3/1

Colophospermum mopane

1

1/2

1

       

Combretum apiculatum

2

2

1

1/2

     

C. eleagnoides

1

2

 

1

     

C. fragans

2

3

         

C. hereroense

v

   

1

     

C. imberbe

1

 

1

     

v

C. mossambicense

3

         

3/2

C. molle

     

2/3

 

2

 

C. zeyheri

     

2

2

2

 

Commiphora africana

2

2

v

       

C. mollis

1

   

1

     

c. pyracanthoides

1

1

         

Dalbergia melanoxylon

2

2

2

       

Dichrostachys cinerea

2/0

1

1

v

2

   

Diospyros mespiliformis

1

 

v

     

v

Dombeya rotundifolia

     

1

 

v

 

Diplorhynchus condylocarpon

     

3

2

3

 

Euclea divinorum

0/1

0

 

0

     

E. undulata

0

 

v

2

 

v

 

Erythrophleum africanum

     

v

1

v

 

Erythroxylum zambesiacum

 

2

 

v

   

2

Friesodelsia obovata

 

2

       

2

Gardenia spatulifolia

3

 

2

      2

Grewia bicolor

1

2

1

2

     

G. flava

3

 

1

2

     

G.flavescens

2

2

1

1

 

1

 

G. monticola

1

 

v

1

 

v

 

Julbernardia globiflora

         

2/1

 

Kirkia acuminata

0

0

         

Lannea discolor

   

1

v

v

2

 

L. stuhlmannii

2

3/2

v

       

Lonchocarpus capassa

3

3

2

     

1/3

Maerua parvifolia

1

1

1

       

Maytenus heterophylla

     

2

     

M. senegalensis

0/1

           

Monotes glaber

         

1/2

 

Mundulea sericea

     

3

     

Ochna pulchra

     

0

1

v

 

Pappea capensis

2

   

2

     

Piliostigma thonningii

   

2

2

 

2

 

Pseudolachnostylis maprouneifolia

     

1

 

2

 

Pterocarpus brenanii

1

1

         

Schotia brachypetala

2

         

2

Securidaca longipenduculata

     

2

 

2

 

Securinega virosa

     

2

 

2

 

Sclerocarya caffra

2/1

2

1

1

     

Strychnos innocua

2

2

         

S. pungens

     

3

     

S. spinosa

0/1

   

0/1

 

1

 

Terminalia prunioides

v

1

 

v

     

T. sericea

0

   

0/1

0/1

0

 

Vitex mombassae

     

v

 

1

 

Ximenia americanum

2

2/3

2/3

     

2

X. caffra

     

2

   

v

Ziziphus mucronata

2/0

 

2

2

 

v

 

The degree of utilization of woody species (assuming some constant stocking rate) is a function of their relative palatability, their growth form and the structures of the herbivore community. Palatability is a complex characteristic and there is no single factor which can be used to assess it. Characteristics which seem to be positively related to it are crude protein content, a high percentage of minerals (especially Na) and moisture content. Negative characteristics include a high fibre content and the presence of tannis and aromatic substances.

Growth form can influence the degree of browsing by allowing some maximum amount of use beyond which the remaining browse is unobtainable, either through height or a physical barrier. The thorns of Acacias are an obvious example, but species such as Gardenia spatulifolia, Balanites spp. and Carissa bispinosa, all of which have thick woody spines, are much more effective in limiting browsing.

The structure of the herbivore community can be complex in a wildlife area, but is a function of the proportion of cattle and goats in farming areas. Goodman (1975) examined the relationship between the structure of a wild ungulate community and the degree of browse, utilization in five layers, in an alluvial vegetation. The mean overall percentage utilization, over all plant species, decreased from 15.5% between/-lm, to 12.2% between 1–2.5m, to 3.0% between 2.5–5.0m. The proportion of available browse, on the other hand, increased with height above ground, from 0.23 to0.42 and then dropped slightly to 0.35 at the highest level. This relationship agreed with the results from feeding observations weighted by animal biomass, which showed that 85.4% of browse was taken from the 0–lm layer, 10.4 from the 1–2.5m layer and 4.2% from the 2.5–5m layer. This clearly demonstrates that there was an imbalance in the relationship between herbivore and vegetation biomass, and that one or both would have to change over time, assuming that this degree of utilization is high enough to significantly affect the growth of the woody plants. Matching the herbivore community to the vegetation structure is essential if a stable system is to be achieved.

6. The role of browse in livestock production

In the Karoo and Valley Bushveld livestock production is necessarily based almost entirely on browse. Grass does contribute significantly in much of the false Karoo, but soil erosion has been such that it is doubtful if a grassland will ever be restored. In the Karoo proper shrubs are the best adapted plants and will remain the dominant vegetation. Climate and soils in the valley bushveld are conducive to the dense succulent type of vegetation which occurs there, and to attempt to replace it with grass would be prohibitively expensive, or would fail.

It is perhaps not surprising, therefore, that the one area in southern Africa where non-grasses are planted for livestock fodder is in the Karoo. On an area basis, the spineless cactus (Opuntia ficus-indica) is by far the most important. There are some 50 000–60 000 ha of this species, which have been planted mainly as a fodder reserve in times of drought (Dr G. de Kock, pers. comm.). Recently, however, there has been a change towards planting species of saltbush (Atriplex spp) which are considered to be a more palatable and valuable fodder.

In the savannas of southern Africa, however, livestock production is based essentially on grass, with browse playing a supplementary role to varying degrees.

In the wetter savannas the best form of land use is either arable cropping or (for livestock) clearing the woody vegetation and grazing the area with cattle. Barnes (1979) has shown that three ans shrub clearing (particularly shrubs) results in up to 400% increase in grass growth. Where the rainfall is above about 750mm, and where capital and management expertise are available, then it is desirable to remove the woody vegetation and introduce legumes into the grassland, or to replace the natural grassland with improved, fertilized pastures. However, when economics dictate the use of natural vegetation then, even in these wetter savannas, browse can play an important role. The woody vegetation needs to be thinned, and then reduced in height, as described by Rees (1974) for Zambia.

In the semi-arid and arid savannas it is uneconomical to clear the woody vegetation, and without irrigation fertilized pastures are not possible. The carrying capacity of the vegetation is such that the increased returns from the extra grass which results from clearing may not even cover the cost of clearing. A major problem has been the increase in woody scrub vegetation which has resulted from the replacement of a mixed herbivore community, which made considerable use of the woody vegetation and controlled its density (cf. Taylor and Walker 1978), by cattle. In addition, overgrazing has weakened the competitive effect of the grasses, and further enhanced the growth of woody plants. Control of woody plants is therefore a prime objective over vast areas of these rangelands. Complete removal of the shrubs, however, is not only uneconomical, but undesirable, as woody plants have a number of advantages in this region. Four in particular are pertinent here.

  1. The time of the new leaf flush. The woody vegetation breaks dormancy and begins to produce the new season's flush of leaves from August to October (depending on species and region), well before the rains begin. Grass growth only begins with the rains in November. For 2–3 months, therefore, at the time when grass biomass and quality are minimal, the woody vegetation produces a high quality food source. This period coincides with late pregnancy in most ungulates, and the browse is consequently critical. During this period it has its highest food value and is most palatable. 

    It is interesting to note that the development of new leaves is very rapid, and does not constitute growth as such. It is a redistribution of stored carbohydrate and protein from the roots and woody tissue, and is independent of rainfall.
  2. Food quality. The crude protein content of most browse is generally considerably higher than that for the grasses at all times other than the early growing season (cf. Section 4). The grass layer supplies a surfeit of energy but is usually limiting in protein supply.
  3. Vegetation stability. Both woody and grass vegetation are adversely affected by drought, and in bad years some plants will always die. Generally, however, the woody vegetation varies less inter-seasonally. Furthermore, because a) woody litter remains for longer than grass litter (Kelly and Walker, 1976), and b) even if new leaf growth is much reduced the existing structure of the woody biomass remains essentially the same, the woody component provides protection to the soil from wind and splash erosion, which is particularly important during dry periods when the grass cover declines. It must be emphasised, however, that the grass sward is far more efficient at soil protection and enhancing rainfall infiltration. The value of the woody vegetation lies in the bad years when the grass cover fails. Finally, because of their growth form trees, and even shrubs, cannot normally be completely consumed, and they therefore provide a minimum plant reserve for subsequent recovery.
  4. The influence of tree canopies on grass species composition, Barnes (1979) has shown that although removal of woody plants leads to higher grass production, this is often accompanied by a change to less palatable grasses. Grossman and Grunow (1980) divided the grasses of the Burkea (mixed tree) savanna into fodder and non-fodder species, and found that although total grass biomass was much higher in open areas without trees, the biomass of the fodder grasses was about the same. 

    From all the existing evidence it is apparent that there is some optimum ratio of woody: grass vegetation, and that complete clearing is undesirable. The optimum ratio depends on the types of plant species, and also on the herbivores.

Conclusion

In searching for the best of livestock production in the drier savannas of southern Africa, an essential pre-requisite is the inclusion of browsing animals—both to take advantage of and to control the woody vegetation. Elsewhere (Walker, 1979) I have reviewed the evidence for and against game ranching, and have concluded (along with others) that pure game ranching is inferior to cattle ranching. The ideal form of use seems to be a combination of cattle and game. Veterinary restrictions, which up to now have prohibited it, can be overcome, particularly where browsers are to be re-introduced. This development needs both short and long-term objectives. In the short-term there is a need to make the best use of what is available. This will involve increasing the proportion of existing browsing animals by building up population of species such as giraffe (Giraffae camelopardis), kudu (Tragelaphus strepsiceros), eland (Taurotragus oryx) and improved goats. The correct proportions can be determined by the relative degrees of utilization and growth of the woody and grass vegetation, as described by Taylor and Walker (1978), and by the use of a simple linear programming model, an example of which is given in Walker et al (1978).

In the long-term the need is for the development of new domestic species of large browsers. The two major disadvantages of wild game (Walker 1979) are inferior food conversion ratios (compared to cantle) and difficulties associated with management (rotational grazing, treatment for diseases and parasites, dehorning, etc.) The evidence in favour of large browsers, over such a vast region, warrants a large-scale, major programme aimed at new domestic species. The few attempts to date have been completely inadequate. Species which are considered desirable on the basis of food preferences, carcass size, water relationships and disease resistance must be captured in large numbers and included in carefully designed selection and breeding programmes, applying rigorous culling based on food conversion and growth rates. Parallel research programmes into the optimum vegetation structure for maximum food production are required. In this regard we need to know more about the palatability and productivity of the different browse species, which of them are "water wasters" as opposed to "water savers", and which of them compete most closely with (and therefore suppress) the grasses.

References

Acocks, J.P.H. (1975). "Veld types of South Africa". Memoirs of the Botanical Survey of South Africa, No. 40, 2nd edn. Botanical Reserach Institute, Pretoria.

DHV Consulting Engineers (1979). "Country wide animal and range assessment project, Republic of Botswana, Final Report". The Netherlands, Amersfoort.

Aucamp, A.J. (1976). "The role of the browser in the bushveld of the Eastern Cape". Proc. Grassld. Soc. Sth. Afr. 11: 135–138

Aucamp, A.J. (1979). "Die produksiepotensiaal van die Valleibosveld as weiding vir Boer-en Angorabokke". D.Sc. thesis, Univ. of Pretoria, S. Africa

Aucamp, A.J. et al (1978). "The browse value of the Eastern Cape Valley Bushveld". Proc. Grassld. Soc. Sth. Afr. 13: 91–93.

Barnes, D. (1969). "Cattle ranching in the semi-arid savannas of East and Southern Africa". In: Walker, B.H. (ed.) Management of semi-arid ecosystems. Amsterdam, Esevier Scientific Publ. Co..

Bate, G.C. (1979). "Nitrogen in South African savannas". Proc. Symp. dynamic changes in savanna ecosystems. Pretoria, S. Africa. (Unpubl.).

Bonsma, J. (1976). Bosveldbome en Weistreke. Pretoria, S. Africa, J.L. van Schaik Bpk.

Donaldson, C.F. (1979). "Goats and/or cattle on mopanes veld". Proc. Grassld. Soc. Sth. Afr 14: 119–124.

Field, D.I. (1978). A handbook of basic ecology for range management in Botswana (2nd ed.). Gaborone, Botswana. Land Utilization Div., Min. of Agric.

Goodman, P.S. (1975). "The relation between vegetation structure and its use by wild herbivores in a riverine habitat". M.Sc. thesis, University of Rhodesia.

Grossman, D., Grunow, J.O. (1980). "Biomass cycles, accumulation rates and nutritional characteristics of grass plants in canopied and uncanopied subhabitats of Burkea savanna". Proc. Grassld. Soc. Sth. Afr 15: (in press).

Kelly, R.D.; Walker, B.H. (1976). "The effects of different forms of land use on the ecology of a semi-arid region in south-eastern Rhodesia". J. Ecol. 64: 553–576.

Louw, G.N.; Steenkamp, C.W.P.; Steenkamp, E.L. (1968). Chemiese samestelling van die vernaaste plantspecies in die dorre, skyn-dorre, skynsukkulente en sentrale bo Karoo. Tech. Bull. No. 79. Dept. of Agric. Tech. Services, Govt. Printer, Pretoria. S. Africa.

Martin, R. (1974). "Structure, biomass and utilization of vegetation in the Mopane and Miombo woodlands of the Sengwa Wildlife Research Area. Certificate in Field Ecology Report, University of Rhodesia.

Phillips, J. (1965). "Fire — as master and servant: Its influence in the bioclimatic regions of Trans-Saharan Africa". Proc. 4th Ann. Fire Ecology Conf. Tall Timbers, Florida.

Rattray, J.M. (1961). "Vegetation types of Southern Rhodesia". Separate from Kirkia 2: 68–93.

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

Rushworth, J.E. (1975). "The floristic, physiognomic and biomass structure of Kalahari sand shrub vegetation in relation to fire and frost in Wankie National Park, Rhodesia". M.Sc. thesis, University of Rhodesia.

Rutherford, M.C. (1979a). "Above-ground biomass subdivisions in woody species of the savanna ecosystem project study area, Nylsvley". South African National Scientific Programmes Rep. No. 36 Pretoria, S. Africa. C.S.LR.

Rutherford, M.C. (1979b). "Plant-based techniques for determining available browse and browse utilization: review." Bot. Rev. 45: 203–228.

Taylor, R.D. (1974). "A comparative study of land-use on a cattle and game ranch in the Rhodesian lowveld". M.Sc. thesis, University of Rhodesia.

Taylor, R.D., Walker, B.H. (1978). "A comparison of vegetation use and condition in relation to herbivore biomass on a Rhodesian game and cattle ranch". J. Appl. Eco1.15: 565–581.

Walker, B.H. (1974). "Ecological considerations in the management of semiarid ecosystems in south-central Africa". Proc. 1st. Int. Congr. Ecology. The Netherlands, Pudoc.

Walker, B.H. et al (1978). "A procedure for multidisciplinary ecosystem research: with reference to the South African Savanna Ecosystem Projet". J. Appl. Ecol. 15: 481–502.

Walker, B.H. (1979). "Game ranching in Africa". In: Walker, B.H. (ed.) Management of semi-arid ecosystems. Elsevier Sc. Publ. Co., Amsterdam.

West, O. (1965). "Fire in vegetation and its use in pasture management, with special reference to tropical and subtropical Africa". Mimeo. Publ. No. 1/1965. Commonw. Bur. Pastures and Field Crops. Hurley, Berks., England.

Wild, H.; Grandvaux Barbosa, L. (1967). "Vegetation map of the Flora Zambesiaca Area". Salisbury, Rhodesia M.O. Collins (Pvt.) Ltd.

Wild, H.; Fernandes, A. (1967. "Flora Zambesiaca: Supplement to the Vegetation Map of the Flora Zambesiaca Area". Salisbury, Rhodesia Govt. Printer.

Previous PageTop Of PageNext Page