1. INTRODUCTION

This report contains the results of a review of existing literature on rangeland and livelihood dynamics in a part of south western Botswana. The purpose of the literature review is threefold:

The study focuses on the Matsheng area. This area lies in the northern Kgalagadi sub-district (44,000 km²) which constitutes, together with Kgalagadi South, the Kgalagadi District. Tsabong is the district capital. The location and the main features of the study area are shown in Map 1.

1.1 The Kgalagadi District

Archaeological and anthropological research has shown that south western Botswana - often called a desert - has been inhabited by the ‘Bushmen’ at least as far back as 200 AD. However, the increased resource pressure currently on the area has mainly been a consequence of the expansion of the livestock experiment in the sector (Cooke, 1985). Prior to livestock invasion, the occupation rate of the area was low due to the non-availability of water ("the old age protection of the Kalahari"; Cooke, 1985). The Kalahari's economic importance dates back to the nineteenth century, when it started to be exploited on commercial basis for trade in skins, feathers and ivory. The second stage in the invasion of the Kalahari arose from the mounting pressure on rangeland resources in the hard veld of eastern Botswana. As Cooke (1985) puts it "in search for solutions or palliates all eyes turned toward the Kalahari sand veld". This invasion, that occurred mainly after independence, resulted in the establishment of permanent cattle posts, a process that was facilitated by borehole technology.

1.2 Kgalagadi North and the Study Area

The Matsheng villages are located approximately 24°S and 22°E. Land in the area is generally used for grazing, crop production, residential, other non-agricultural activities, and collection of veld products including wood resources. The area is among the driest in the country. While rangelands are widespread in Botswana, the arid nature of the Kgalagadi District has made it more sensitive and vulnerable to increased grazing pressure.

It is a relatively poor area, as are most other remote areas in the country. The latest poverty study (BIDPA, 1997) found that northern Kgalagadi has the highest "destitute" incidence in the country after Ghanzi district: 3.5 times that of the country's average. Living conditions are comparable to Gambia while the capital, Gaborone, enjoys South Korean living standards. Causes of low living standards include: limited arable potential, remoteness and distance to the markets, low population density, poor quality of government services and remote area dwellers (RADs) living in marginal settlements. The Matsheng villages are probably slightly better off than the sub-district average because of the facilities available and government support. The area has recently been linked to the tarred network in southern Africa, which will make it less isolated. A 100 km tarred road connects the villages to the Trans-kalahari highway at Kang.

1.3 Data Sources

There is a dearth of specific, up-to-date data on the economic activities of the Matsheng villages. Much of the following account of these activities is therefore based on dated sources, some of which address the Kalahari as an ecological region or generalise about the Kgalagadi and the western sub-regions. It is expected that, when more data become available on the villages (e.g., from this study), general impressions presented here are likely to be modified. For the more area-specific data, the report relies heavily on a baseline study of the Matsheng villages (van der Maas et al., 1994). In addition, regular statistics such as Population Census and Agricultural Statistics have been used.

Reviewing literature as and when it is published is an integral part of the project, and it is expected that many other studies will contribute to the modification of our preliminary conclusions. For example, data on groundwater quality are expected in the near future. New Agricultural Statistics may throw new insights on the region. A more thorough analysis of Population Census data is planned at a later stage.

1.4 Working Definitions

The following working definitions are being used in the study:

Stocking rate is the number of animals, domesticated and/or wild, per unit of rangeland.

Rangelands are land resources mostly used for extensive forms of land use, primarily livestock production, wildlife utilisation, wood and plants gathering as the areas are too dry or infertile for more intensive forms of land use.

Rangeland degradation is associated with changes in vegetation and soils leading to a decline in the total land productivity. It includes both reversible and irreversible changes. Obviously, the irreversible changes are a greater cause of concern.

Total land productivity is the summation of livestock, wildlife and veld-products production per unit of rangeland.

1.5 Structure of the Report

The report is structured as follows. Each subsequent section discusses a cluster of the mechanistic model developed at the Gaborone workshop (Odada et al., 1996). In the concluding section, this model is revisited (Figure 7, p. 49), and a synthesis of the preliminary conclusions derived from the literature review is given (section 11).

5.1 Physical Resource Availability

No resource inventory exists for the sub-district or the study area. The study area has traditionally been well endowed with natural resources such as veld products and wildlife. However natural resources such as wildlife are reported to be depleting rapidly. For example, between 1986 and 1994 a dramatic decline occurred in the number of gemsbok, hartebeest, eland and wildebeest populations (see section 2.6). Because of its scarcity, water plays a key role in human activities in the district. Surface water is very limited, diminishing, seasonal and only able to sustain limited human activities (Section 2.1). Groundwater has eased the water constraint to some extent, but mostly for domestic use and for rich livestock producers, who can afford boreholes.

5.2 Regulations and Policies

Land tenure and land use policies are very important in determining resource access. Policies are discussed in more detail in section 2.9. The entire study area comprises tribal land. The bulk of the land is communal, earmarked for mixed farming within the 20 km radius of the villages and wildlife management areas (WMAs) beyond this artificial boundary. The Department of Wildlife and National Parks (DWNP) has started to allocate hunting quotas for some WMAs. Communities have the option to sub-let these rights or to exploit them themselves. Some livestock ranches are situated to the east of the area. This leaves very little communal land designated primarily for livestock production. This constraint is made more severe because of:

In Kgalagadi District in particular, most Tribal Grazing Land Policy (TGLP) ranches occupy land where local communities have hunted and harvested veld products (Thusano Lefatsheng, 1994). Thus, private ranches improved resource access for the rich livestock owners but reduced access to land, wildlife and veld products for the majority of the population. Although TGLP is silent about the on-farm hunting and gathering rights of the rest of the community, in practice, people are excluded by the mere fencing of the area. The Agricultural Development Policy (ADP) of 1991 opens opportunities for fencing of more rangeland. Discussions on how other people can retain the traditional rights to collect veld products within the fenced-off areas are on-going. These rights are vaguely defined and it is clear that admission of those who wish to collect veld products is subject to the discretion of the land-holder. This thorny issue may soon emerge in the study area as Makopong in the Kgalagadi South Sub-district has been selected as one of the first areas for the implementation of the new fencing scheme.

Informal rules continue to influence resource access. For example, to reduce conflicts and over-utilisation, each settlement appears to have their "own" hunting and gathering areas. Informal rules still play a role throughout Botswana. However, the extent of their application and the degree of influence is known to differ from region to region. It is therefore important for the study to find out which informal resource access rules are still operational and what their impact is.

Access to land for livestock is not meaningful without having access to water. This implies that remote rangelands are not accessible for livestock unless you can draw from borehole water. The poor majority has no alternative but to keep cattle in the vicinity of settlements, frequently, and illegally, using council water for their animals during the dry season.

As individual arable land is effectively given priority over communal livestock and wildlife areas, and population density is low, access to cropland is relatively good. According to the Sub-Land Board, on average 15 farmers apply for fields each month. Arable encroachment into grazing areas is therefore a minor problem. Crop damage by cattle is likely to be more problematic.

During a National Workshop on Community Based Natural Resources Management Institutional Building held in Kang in November 1994, the Kgalagadi District Report stated that communities in the district are not benefiting anything from the parks which limit their access to land and the natural resources therein (Thusano Lefatsheng, 1994).

5.3 Socio-economic Means

Socio-economic issues are discussed in more detail in sections 2.5 and 2.7. Suffice to say here that resource pressure is highest in the vicinity of the settlements and boreholes. These are the areas to which the poor are confined. By virtue of their poverty, the poor are confined to rangelands, where resources are most rapidly diminishing. As a result of their confinement, they accelerate resource depletion and reduce their subsistence generating capacity. This process occurs throughout Botswana. For example, in the mid Boteti area, the contrast was great between degraded rangelands close to villages and remote rangeland with abundant grazing (Arntzen et al., 1994).

5.4 Conclusion

Physical availability, policies/tenure systems and socio-economic means all contribute towards the accessibility of natural resources.

Physical availability has adversely affected access to wildlife. The opposite has occurred for water, although the additional groundwater mostly benefits domestic users and rich livestock owners. The trend of other resources is not documented.

Policies and tenure systems have probably influenced resource access most. The Matsheng villages are boxed in between leasehold farms and areas reserved for wildlife. Most land is communal, set aside for mixed farming and wildlife. There is little space for communal livestock production. The area faces an enormous challenge in realising the potential of wildlife utilisation for the benefits of the communities. Allocation of hunting quota and wildlife user rights has been a recent, first step in this direction.

Socio-economic means, or the lack of them, have had a large impact on the de-facto access to resources. The poor are confined to the villages whilst the large cattle-owners can keep their cattle in remote areas (subject to groundwater availability). An important issue for this study is whether access to resources is becoming increasingly skewed against poorer people. Government policies favouring agriculture and the dynamics of the livestock sector have caused increasingly skewed distribution of resources elsewhere in the country. This process is enhanced by existing income inequalities allowing the rich to access and control more remote resources. Community-based resource management projects offer an important opportunity in the study area to redress the balance.

5.5 Topics of interest for further research which arise from the literature review

There appears to be hardly any rangeland-based processing industry in the area. Only leather works are found in some villages such as Lehututu (van der Maas et al., 1994).

7.1 Wildlife

The study area is located between Gemsbok National Park and the Central Kalahari Game Reserve. Eland, gemsbok, duiker, kudu, hartebeest, springbok, steenbok, wildebeest, ostrich and warthog (Arntzen and Veenendaal, 1986; Michelsen Institute, 1996) are common wildlife species in the Kalahari environment. The study area is part of the Schwelle, which is of crucial importance for wildlife in southern Botswana because of the many pans which hold water during the rainy season. Hunting and gathering territories are more or less divided among the local settlements (Figure 4). Although wildlife utilisation is strictly controlled by government (see 2.7), the resource has rapidly declined in the last decades both in terms of quantity and species (Table 8). The expansion of the livestock sector, hunting, declining surface water and interference with wildlife migration through fencing are likely causes for this reduction. The resource decline has led to:

The reduction in wildlife benefits and the increase in costs may have led, as elsewhere in the country, to a negative attitude towards resource: wildlife is seen as a nuisance rather than a development resource. Painter (1997) further argues that a reduction in number of hunting households does not necessarily lead to less hunting: 88% of the hunters sell at least part of the game meat. The meat market may in fact increase because of increased incomes generated elsewhere.

The improved road network and implementation of the joint Botswana/South African management plan for Gemsbok National Park and plans to promote community-based, non-consumptive wildlife utilisation should offer new opportunities for wildlife utilisation. The challenges are to increase the development benefits of wildlife and to use a larger portion of the economic rent for rural development.

Wildlife resources may be increased through restocking or a moratorium on hunting (Barnes, 1996). In addition, strict land use planning should prevent agricultural encroachment into key wildlife areas and reduce conflicts between wildlife and agriculture (predation, crop damage, health risks).

Table 8: Wildlife decline in western region (Kgalagadi and Ghanzi Districts)

 

SPECIES	SETSWANA NAME	SCIENTIFIC NAME	% CHANGE BETWEEN 1978/1986
hartebeest	Kgama	Alcelaphus buselaphus	- 76.1
wildebeest	Kgokong	Connochaetes taurinus	- 94.2
springbok	Tshepe	Antidorcas marsupialis	- 7.0
ostrich			- 63.3
gemsbok		Oryx gazella	+ 35.9
eland	Phofu	Taurotragus oryx	- 46.6
kudu		Tragelaphus strepsicerus	- 15.1

Note: this period largely includes the drought period of 1982/87.
Source: Van der Maas et al., 1994.

Table 9: Single game hunting quota in Kgalagadi District (1981 and 1991)

 

SPECIES	1981	1997	% CHANGE
wildebeest	2324	0	--
hartebeest	1663	0	--
steenbok	1586	1340	- 15.5
duiker	1456	820	- 43.7
gemsbok	1193	615	- 48.5
springbok	742	1985	167.5
ostrich	562	0	--
eland	466	0	--
kudu	398	60	- 84.9
zebra	200	0	--
others	128	4	- 96.9
Meat weight	677 tonnes	146 tonnes	- 78.4
Total value	P 6.0 mln.	P 1.1 mln.	- 81.7

Source: calculated from DWNP data.

7.2 Veld Products

Gathering is still very common but it has lost importance with sedentarisation and the increases in livestock production and formal employment. Sedentarisation has also led to gathering concentrating in small areas around the settlement with the risk of over-exploitation.

Veld products are a neglected subject. A grapple propagation project is on-going in the RAD settlement of Ukwi. No systematic survey has been done of veld products derived from rangeland in the area. Some surveys such as Painter (1997) cover veld products but fail to mention species. Table 10 summarises a list of local veld products mentioned in various references.

Assessing the commercial potential of veld products, Taylor and Moss (1982) identified eight tree species which occur in high densities in the study area (field checks at Hukuntsi, Lehututu, Lokgwabe and Tshane; cf. Table 11.) The tree diversity appears very limited and only a few were judged to have a commercial potential at that time (early 1980s).

The Michelsen Institute (1996) considers the cultivation of some veld products such as morama, sengaparile, truffles and shrubs like Grewia spp. to be a less risky form of agriculture, which requires less care once the crops have been established. It also makes economic sense as most fields are already fenced off and protected against livestock.

Table 10: Major veld products in the study area

 

SPECIES	SETSWANA	ENGLISH NAME	PRODUCT / USE	COMMENT
Harpagophyton procumbens	Sengaparile	grapple plant	Tubers used for medicinal purposes	Used as dried tubers or in tablet form; distribution of this species in Matsheng area is unreliable
Grewia flava	Moretlwa	brandy bush, raisin tree	Kgadi - traditional beer	raisins are sold locally in some Matsheng villages (Taylor and Moss, 1982)
Grewia bicolor	Mogwana		Khadi- traditional beer	same as for G. flava
	Dipotswa		Berries
	Boroku
	Mongone
Ziziphus mucronata	Mokgalo	buffalo thorn	berries
Boscia albitrunca	motlopi
	mokwa		root used
Citrullus lanatus	kgengwe or tsama melon	desert melon	source of water
Acanthosicyos naudiniana	mokapana	gemsbok cucumber	source of water
Amaranthus tunbergii	thepe	wild amaranth	vegetable	the leaves are collected and cooked as spinach
Ipomea adenoids	seroa		source of water	eaten raw
	borego		vegetable
Cucumis africanus	magabala	horned cucumber	vegetable	seeds can be eaten raw, roasted or boiled
Terfizia pfeilii	mahupa	Kalahari truffle	vegetable	this underground fungus (mushroom) is regarded a delicacy locally and in Europe (Taylor and Moss,1982)
Tylosema esculentum	morama bean	Gemsbok beans	seeds	the seeds are very high in protein and oil
Eragrostis pallens	Motshikiri	thatching grass	thatching
Stipagrostis uniplumis	Tshikitshane	thatching grass	thatching

Table 11: Trees occurring in high densities in the study area

 

SCIENTIFIC NAME	SETSWANA NAME	PLANT DENSITY	PRODUCTS/USES
Acacia erioloba		+ 20/ha
Acacia hebecleda		some areas: 5-40/ha other areas + 40/ha
Acacia luederitzii		2-17/ha
Cassia italica		Logwabe only: + 75/ha
Gnidi polycephala		visual ass.
Grewia flava	moretlwa	visual ass.
Grewia bicolor	mogwana	visual ass.
Terminalea sericea		+ 50/ha

Source: Taylor and Moss, 1982.

7.3 Livestock and Grazing

During this century, livestock products have substantially increased following the increase in livestock numbers. However, more recently, cattle numbers appear to have stabilised and the number of goats has expanded more rapidly (cf. Table 12; goats are most common in RADS settlements). The long-term increase in livestock numbers has increased the availability of livestock-related products such as milk, meat, draught power, manure, savings. Some of these products are apparently insignificant in the study area (for instance: milk and draught power; van der Maas et al., 1994). The increase in goat numbers have increased domestic slaughtering. In this respect, goats have probably replaced wildlife as a source of meat. Donkeys are important for transport and draught power.

7.4 Conclusion

Rangeland resources and products are changing due to changes in the resources (wildlife decline) and in activity patterns (sedentarisation, livestock and formal employment). Livestock products have increased during the last decades, even though some appear hardly utilised (milk, draught power). Wildlife products have substantially decreased. It is not exactly known what the trend in veld products and arable production has been.

It is most likely that these changes are inter-linked. Given the low population density, resources are not under threat of subsistence use (wildlife is a possible exception) but rather from agricultural encroachment. Kgabung's study on the impacts of livestock expansion on wildlife

Table 12: Trends in livestock numbers and performance (1980-1990)

 

VARIABLE	1980 Kgalagadi North and South	1980 Botswana (traditional)	1990 Kgalagadi North	1990 Kgalagadi South	1990 Botswana traditional
Number of cattle (000)	49	2 455	43.6	49.3	2 2211
of which: cows	19	996.5	20.5	19.2	936.9
calves	9	488.5	10.2	13.1	463.8
heifers	6	333.2	4.7	8.4	334.9
oxen	6.5	308.6	3.4	3.3	186.5
tollies	6.2	284.5	4.0	4.5	255.2
bulls	0.9	43.7	0.8	0.8	33.8
Av. herd size	54.4	42.5	48.4	54.8	39.6
Birth rate	56.4	58.4	51.7	61.6	74.0
Death rate	14.3	13.8	17.9	12.7	17.2
Sale rate	10.8	7.7	6.0	7.2	6.7
Home slaughter rate	0.4	0.7	1.1	1.9	1.4
Purchase rate	0.8	1.7	0.7	1.5	0.8
Number of Goats (000)	30	624.0	47.1	45.6	2 030.2
Average herd size	25	13.6	42.8	50.7	29.5
Male 1 yr +	3.8	76.5	5.6	7.6	307.8
Female 1 yr +	17.1	337.3	27.8	26.8	1132.8
Males under 1 yr	3.6	98.0	6.1	5.0	263.2
Females under 1 yr	5.5	112.2	7.6	6.2	326.4
Birth rate	41.3	45.0	41.0	40.2	42.3
Death rate	35.7	34.7	14.6	19.7	12.7
Sales rate	4.7	3.0	4.6	5.6	4.2
Home slaughter rate	8.3	5.1	8.1	7.5	6.4
Purchase rate	0.7	1.9	0.8	2.0	2.0
Number of sheep (000)	22	134	9.8	16.3	266.5
Average herd size	55	9.8	19.6	23.3	15.0
Number of Donkeys (000)	4.2	128.2	3.0	3.6	154.9

Source: Agricultural Statistics 1980 and 1990.

and veld products in the study area will provide valuable details on this issue. For the area west of Shoshong, Rampete (1996) found that livestock and borehole expansion led to a reduction in wildlife and gathering. However, he did not assess the impacts in detail.

Resource dynamics have had important socio-economic impacts including lower incomes, more efforts to hunt wildlife and to collect veld products near villages, reduced resource development potential (wildlife), and increasingly skewed distribution of rangeland products. The transformation from a hunting/gathering based economy to a livestock-dominated economy has had a profound impact on resource base, in particular wildlife, and on people's living conditions. Those who do not have livestock are marginalised as they have come to depend on crops, declining wildlife resources and government hand-outs. The question arises whether and how the majority of people are able to eke out a sustainable living from an agricultural economy as compared to a more diverse economy based on agriculture, wildlife and veld products. A further question is whether this transformation suits the local resource base, in other words does it make optimal use of the available natural resources. This includes the identification of opportunities for processing of rangeland products. A final question is to find out what is driving these changes (e.g., policies, climate change).

7.5 Topics of interest for further study which arise from the literature review

It is expected that the new Trans-Kalahari highway will reduce constraints such as expensive supplementary feeding and marketing costs (Van der Maas et al., 1994).

8.2 Wildlife Utilisation

Residents of the Matsheng villages mainly use dogs, donkeys and spears for hunting. These are sufficient for hunting gemsbok but horses are needed for hunting kudu and hartebeest, and horses can only be afforded by the richer households (van der Jagt, 1995). On average, hunters kill 5 animals per year and about 20% of them hunt without licenses. The meat is usually distributed within the extended family (van der Jagt, 1995).

Wildlife utilisation constraints include:

Problems with the enforcement of the hunting licenses may contribute to excessive hunting and resource depletion.

8.3 Veld Products

There is a dearth of literature referring to gathering that is specific to the Matsheng. One of the few studies was done on the grapple plant in Tsabong. Sekhwela (1994) reports that hardly any grapple plant could be found within the 10 km radius of the village of Tsabong. According to villagers, grapple was plentiful in the past but has disappeared due to over-exploitation. The disappearance of this product was partly a result of the failure by harvesters to fill up the holes after harvest. This suggests that over-mining together with inappropriate harvesting techniques led to the disappearance of this resource.

Generally, the degree of involvement in gathering practices varies with ethnic background. The Basarwa tend to collect a greater variety of species than the Bangologa and the Balala (Molamu et al., 1995). Given the low population density, subsistence use is not expected to affect veld products regeneration adversely, with the exception of the products growing in the immediate surroundings of the settlements.

Collected wood is mainly used for fuel, construction material and medicinal purposes. Even though 98% of the households in the Matsheng villages depend on firewood for fuel, fuelwood does not appear, from the literature, to be scarce. Possibly, bush encroachment together with in increasing commercial trading of firewood may put a strain on certain species in future. As a product of the Matsheng Land Use Plan, four wood lots were established in 1979. According to the plan, a 10 hectare plot was prepared close to each village with 75% of it being assigned to tree planting and 25 % to tree nurseries. Water shortages have led to the general failure of the project. However the nursery has managed to supply the villages with seedlings for village tree planting. Only one of the wood lots appears to function properly. The nursery has been take over by the Ministry of Agriculture in the hope that it can inject the required expertise into the initiative.

8.4 Arable Agriculture

Because of lack of moisture and the types of soils, the area is not suitable for the cultivation of most conventional crops such as maize, sorghum and millet. The only suitable crop identified was lablab beans and cow peas (Van der Maas et al., 1994).

Nonetheless, a number of households are still involved in arable agriculture, probably because of lack of alternatives and government assistance. Not surprisingly, the actual productivity that has been achieved is by far lower than the yield potential that can be achieved. This is a consequence of poor management, wrong crop choice and inappropriate production techniques. For example, the study indicates that only a few households use manure to top up soil fertility.

Unlike in the hard veld of eastern Botswana, households with no cattle are more active in arable agriculture than those with cattle. For example while 57.6% of those without cattle are engaged in arable agriculture, only about 25% of those with more than 40 cattle grow crops. Obviously, given the low returns of the sector, those with alternatives such as cattle do not engage in cultivation. The data further reveals that female-headed households are more dependent on this activity than other types of households. It is suggested that arable agriculture is a last-resort activity undertaken by the poorest.

Some key indicators for cultivation are summarised in Table 13. The area of arable land has more than tripled in ten years, mostly because of expansion of the arable land holdings to around the country's average. This probably reflects the impacts of agricultural subsidies through especially ARAP. People plough less frequently and smaller areas as compared to the national average, although ploughing activities have increased. Surprisingly, maize is the most widely cultivated crop, even though the reported yields/hectare are lowest. The limited potential of the area is reflected in the very low yields. Beans and pulses seem to do better than sorghum and maize.

Table 13: Key indicators of crop production in the Kgalagadi district (1980 and 1990)

Variable	1980	1980	1990	1990	1990
	Kgalagadi	Botswana (traditional)	Kgalagadi South	Kgalagadi North	Botswana (traditional)
Arable land area (in 000 ha.)	1.3	325.3	2.9	2.1	345
Farms with fields (%)
Average land holding/farm (ha)	1.6	4.0	4.8	4.2	4.5
cultivators as % of farms with fields	57.1	93.1	83.3	80.0	91.1
harv area/farm (ha.)	0.5	3.7	3.3	n.a.	4.3
area planted; sorghum (1000 ha.)	0.1	142.5	0.3	0.3	180.7
same, maize (1000 ha.)	0.3	74.9	0.8	0.4	68.6
same, beans/ pulses (1000 ha)	0.2	23.4	0.3	0.2	18.5
total area planted (1000 ha.)	0.6	270.7	1.5	1.0	283.8
Yield/planted area (kg). sorghum	n.a.		33	33	106
Same, maize	0	167	13	13	62
Same, beans/pulses			17	100	86

Source: Agricultural Statistics.

The main constraints facing arable agriculture are:

Horticulture is very limited and where it exists it is rarely done by farmers but more by institutions such as the district council. Horticulture is severely inhibited by water constraints.

8.5 Conclusion

There is limited literature on resource use practices. In general, livestock practices appear fairly poor because of the high production and marketing costs. Most livestock holders are below the minimum sustainable herd size, and therefore aim to accumulate rather than sell animals. Traditional hunting practices may become less appropriate with the change in species availability and the increased distances to be covered. Information about veld products is limited, but a grapple study in nearby Tsabong suggests that commercialisation and poor harvesting practices may contribute to the disappearance of species close to settlements. In spite of the reported limited potential of conventional crop production, the majority of poor households is still very active in this sector, most likely because of lack of alternatives and government support. The choice of crops does not seem to match the physical conditions.

8.6 Topics of interest for further research which arise from the literature review

Special game licences (SGL) are issued free of charge to Remote Area Dwellers; in 1994 153 SGLs were issued in northern Kgalagadi by the Department of Wildlife and National Parks. In 1995, 1996 and 1997, 112, 147 and 204 SGLs respectively were issued for the entire Kgalagadi District (DDP5; pers.comm. Thompson). An SGL entitles a person to hunt the following: 1 kudu, 2 gemsbok, 4 hartebeest, 4 springbok, 30 steenbok and 30 duiker.

To resolve conflicts between wildlife and human activities, so-called problem animals (e.g., predators, destroyers of crops) may be shot without a licence. For certain species, the number of problem animals killed exceeds the quota, which reflects the sustainable harvesting rate.

Away from the settlement at 20 km distance, most of the study area is designated Wildlife Management Areas (WMAs) set aside for community-based schemes. Unfortunately, these areas are not yet gazetted and therefore do not receive full protection. In WMAs, wildlife utilisation is the primary form of land use: livestock and crop production are subordinate. This implies for example that no boreholes can be opened for livestock. The critical issue is to identify viable wildlife ventures which can generate sufficient income for the local population. The following observations have been made in this respect:

The relationship between the local population and DWNP is not good, and there are numerous complaints about physical abuse and unfairness/arbitrariness with licences (Van der Jagt, 1995). As a result, illegal harvesting is likely to occur and wildlife issues are sensitive. However, examples from northern Botswana and Zimbabwe have shown that community-based wildlife management schemes have the potential to forge genuine cooperation between villages and wildlife officials.

The Matsheng baseline study indicates that there is scope for eco-tourism in the area, although the intrusion of livestock in the WMAs seems to present a disincentive to the practical realisation of this potential (van der Maas et al., 1994). One effective way of conserving wildlife in the area is by ensuring that the benefits of wildlife management accrue directly to the populations close to the WMAs. This calls for community participation in wildlife utilisation and protection projects, akin to the CAMPFIRE and ADMADE approaches in Zimbabwe and Zambia. Indeed, the DWNP and the USAID-supported Natural Resources Management Project are experimenting with this idea in Zutshwa (Painter, 1997). The on-going improvements in road infrastructure and camping-site facilities should lead to a fuller realisation of the tourism potential of the District which KDDP4 reported to be "virtually untapped" by 1990. The reported scarcity of wildlife in the WMAs and the declining self-help spirit noted by KDDP4 is of concern.

As to the National Parks south of the study area, entrance fees differ for citizens (P2/day), residents (P10/day and non-residents (P50/day). No hunting or gathering may take place in Parks; no livestock is allowed inside the Parks. The Gemsbok Park currently attracts few tourists on the Botswana side. The joint management plan with South Africa and the new tarred road are expected to increase tourism in the area. Although the Central Kalahari Game Reserve is further away, plans to boost wildlife-based tourism may also benefit the study area as it would attract more investments and tourists to south-western Botswana. The tourism policy provides for granting tourism concessions, in which private companies or communities may be granted exclusive user rights to wildlife resources. Such concessions have been pioneered in northern Botswana, where they have been successful in areas with abundant game. No tourism concessions have yet been granted in the Kgalagadi. This is, however, likely to happen, probably firstly in existing community-based wildlife management schemes such as Zutswha's.

Veld Products Policies

Veld products are by and large unmanaged and consequently suffer from open access and the risks of over-utilisation. The Agricultural Resources Board or ARB (established by the Natural Resources Conservation Act 1974) with the assistance of District Conservation Committees aims to raise awareness about the importance of veld products and the need for their sustainable management. The government tries to manage the grapple resource through a licence system (for extractors and traders and export) and education about proper harvesting techniques (to safeguard the regeneration of the plants). The ARB issues free licences to harvest sengaparile. It has been difficult to monitor the actual amount harvested. Harvesting of other veld products is unchecked, although usually permits are required for their export (e.g., mophane worms).

The establishment of a veld products policy for the next 5 years has been announced in the Eighth National Development Plan. Current sentiments appear to favour granting of veld product user rights to communities together with wildlife rights. Cultivation of veld products is not yet promoted by the government. NGOs such as Thusano Lefatsheng, FONSAG and Veldproducts Research assist communities with improved harvesting and cultivation of veld products.

Grazing and Livestock Policies

Technology development and government policies are largely responsible for the rapid increase of the livestock sector in the area (both subsistence and commercial).

The Tribal Grazing Land Policy (1975) has been the most influential policy. It aimed to promote greater livestock productivity and commercialisation together with better rangeland management. Large herds would move out of communal areas into newly-designated ranches (previously communal land). Ranches would lead to higher livestock productivity; herd movements would reduce pressure on the remaining communal areas. Ranches are found in particular in western Botswana, including on the eastern side of the study area (Figure 4). There is no evidence that livestock productivity or rangeland conditions have improved. Reasons abound:

The livestock sector has been a main beneficiary of government support through subsidies and tax advantages. Fidzani et al., (1996) mention that over 50% of the livestock producers costs are subsidised by government. Subsidies have covered borehole drilling and maintenance and various inputs (Arntzen and Fidzani, 1997). In addition to cost subsidies, farmers benefit from preferential access to the European market. Abolishment of such preferential access will lead to a drop in the beef price of up to 50% (Fidzani et al., 1996; Metroeconomica, 1997). Northern Kgalagadi has some major disadvantages with respect to livestock production:

In smaller settlement, goats are most common. Smallstock receives considerable less government support than cattle. BMC smallstock prices are low, and few farmers bother to sell their animals at Lobatse.

Crops and Food Policies

Crop production has been promoted through programmes such as ALDEP, ARAP and drought relief support measures. The main instrument has been the provision of targeted subsidies.

The Arable Land Development Programme or ALDEP started in the late 1970s and aims to develop the agricultural sector through the provision of subsidies for fencing (to reduce crop damage), draught-power (donkeys to relieve draught-power constraints), water storage (to allow people to stay at the land. ALDEP subsidies are restricted to medium and poor farmers (with less than 40 cattle).

The Accelerated Rainfed Arable Programme or ARAP started in the mid 1980s and provided subsidies for recurrent activities such as destumping, ploughing, row planting, weeding, etc. It helped farmers during the drought, but had serious disadvantages:

No food programme takes into account the specific conditions of the sandveld. Arable policies have led to expansion of the amount of cleared land, and supplemented people's income. There is no evidence that any programme has led to a structural improvement in people's livelihood (BIDPA, 1997).

Crop prices are generally set at income-parity level plus the transport costs. Crop prices are thus higher in northern Kgalagadi than in eastern Botswana. To promote the cultivation of drought- resistant sorghum, until recently the Botswana Agricultural Marketing Board or BAMB offered relatively high sorghum prices at the national level. This practice has been halted because of losses on stocks and illegal imports of sorghum to benefit from the high prices.

10.2 Development and Livelihood Support Policies

Financial Assistance Policy (FAP)

The FAP aims to promote productive activities and contribute to economic diversification. FAP has provided financial support for over 22 projects in the Matsheng villages. Grants are higher for projects in remote areas (urban areas excluded), run by females and projects which create more employment. The project which received support in the study area are mostly brick-moulding, sewing/knitting, welding and a few agricultural ventures (goats, poultry). Their failure rate has been high. FAP offers extra subsidies for productive activities in remote areas. One constraint is that FAP does not fully bridge the gap between a project idea and project implementation. For example, community projects are known to be in need of marketing and technical studies to identify a viable project. The recent extension of FAP to the tourist sector offers important opportunities for the study area.

Remote Area Dweller Programme (RADP)

The RADP policy aims at improving living conditions of rural people living outside established and recognised villages. In order to provide basic social services (successfully but at high costs), people were encouraged to concentrate in settlement but the programme has failed to offer viable land use alternatives for the settled population (Michelsen Institute, 1996). Available land was limited and poorly protected, natural resources are rapidly depleting in the vicinity of the settlements (wildlife and veld products) and permanent water in the villages attracts animals often from outside. RADS have become dependent to a large extent on famine relief and government subsidies. Ukwhi, Ngwatle, Nxaang, Monong and Maake are all RADS settlements in and around the study area.

Targeted Financial Support Programmes

Government provides the old people (65 plus) with a monthly pension of P100. In addition, people without sufficient sources of livelihood receive food assistance through the so-called "destitute" programme. Destitutes tend to be older or single people, who are not supported by the extended family or by friends. The percentage of destitutes is relatively high in northern Kgalagadi.

10.3 Settlement and Public Services Policies

The government aims to improve access to basic services by its population in order to improve the quality of life and to offer a foundation for productive rural activities. Its settlement policy aims to organise settlements in such a way that services can be provided in a cost-effective manner. According to the National Settlement Strategy, only villages of Hukuntsi, Lehututu, Lokgwabe and Tshane are recognised as villages as they have populations of more than 500 people. Others are, strictly speaking, localities. This has implications for the facilities provided. Generally the greater the population, the higher the rank of the settlement and more facilities of higher standard will be provided:

In 1991, the facilities provided were as follows:

For the Northern Kgalagadi Sub-District, Hukuntsi is classified as a secondary village and the sub-district headquarters, and houses most services for regional use. Facilities in this village include a primary and junior secondary school, a primary hospital and a clinic, a post office, telephone services. Many government departments are represented, for example, the police, national identity (omang) office, meteorology, wildlife, agriculture and veterinary offices, roads department, and education. Hukuntsi also has a several shops, brick-moulders and leather works. The village is also served with standpipes. The government also operates a generator which generates electricity for the government department buildings.

Lehututu is a tertiary village with fewer facilities: standpipes, a primary school, a community junior secondary school, a clinic, post office, police, agricultural extension and veterinary offices. There are a few shops as well as brick-moulders and leather works. Lokgwabe village, whose population is slightly lower than that of Lehututu, also has standpipes, a primary school, a health post, meteorology office, police, a few shops, brick-moulding and leather works. Tshane has even fewer serves and facilities. Like the other villages it is served with standpipes, has a primary school, a postal agency, the police and a few shops. In addition, it has a meteorological office.

The other settlements have even fewer services. Hunhukwa has a water reticulation system, a primary school, a clinic and a cattle crush. Phuduhudu does not have standpipes but only a borehole. The only other facility available is a health post. Ukwi has a primary school that only goes up to standard 6 , a health post, a few shops and a cattle crush. Unlike Phuduhudu and Ukwi, Logwabe has standpipes, but has a primary school that only goes up to standard 5 and a health post. Monong and Caa cattleposts are served by standpipes and a mobile clinic. Standpipes are the only facility found in Ohe. Unfortunately the village has the least population in the area, and this will limit its chances for future service provision. A primary school, which only goes up to standard 3, is the single facility in Zutswa.

10.4 District Plans

District Development Plan

The emphasis of the district plan is on improvement of the infrastructure and public services in order to improve the quality of life and to attract commercial activities. The following income and employment creation activities were prioritised by KDC in their DDP4:

Sadly, a recent review of DDP4 by the KDC in preparation for DDP5 revealed general failure to attain set goals: "It is obvious that DDP4 has failed to create the necessary employment and income sources in the private sector..." (MLGLH, 1995). Regarding the performance of the agriculture and wildlife sectors, the following quotations are illustrative: Agriculture: "Altogether the agriculture sector failed to increase or to diversify the production as well as to contribute to income and employment generation" (p.2). Wildlife and tourism: The "Wildlife and tourism sector has failed totally to create income and employment in the district" (p.3).

The new District Development Plan (DDP5) needs further analysis. However, the following points emerged:

Interestingly the plan mentions that the community trust in Zutshwa produces salt (400 tonnes in 1994), which is sold to Livestock Advisory centres.

District Land Use Plan

Land use has to fit into the Kgalagadi District land use plan (Figure 4) and into the Matsheng villages land use plan (Figure 5 and 6). Surprisingly, neither plan has been formally approved. The study area is unique in western Botswana because it probably has the first land use plan, dating back to 1979. In general terms, land use in the larger study area is primarily wildlife management area. The immediate surroundings of the villages are reserved for mixed farming (crops and livestock). To the east and north of the villages, blocks of livestock ranches occur. Further away to the south and north, there are game reserves and parks.

10.5 Conclusion

It appears that policies have played a major role in the transformation towards an agricultural society. The question arises as to what extent this reflects the local resource availability (comparative advantages and sustainability) and serves the needs of the local population (equity).

Because of the various forms of financial support of government and lost control over their immediate natural resources, the population has become dependent on government, and lost its spirit of self reliance. This is a country-wide problem.

Although the facilities are improving, inadequate facilities and remoteness of the area still make it difficult to attract commercial non-agricultural activities, including tourism ventures, and to diversify the rural economy.

Figure 4: Kgalagadi District Land Use Plan

Figure 5: The Matsheng Proposed Land Use Plan

Source: Van der Maas et al., 1994

Figure 6: Extension of the Communal Areas according to the New Land Use Plan

Source: Derived from Kgalagadi DLUPU (1991) and DWNP map of WMA boundaries

Source: Van der Maas et al., 1994

The study need to assess and analyse these and other changes in more detail.

These changes have influenced the actual products derived from rangelands. For wildlife, the overall benefits (meat, karosses, trophies, cash from sales) have substantially decreased whilst the costs have increased (more efforts to hunt, wildlife/livestock conflicts, etc.). Livestock products have become more important. Because of the domination of small herds, farmers aim to accumulate rather than sell or use cattle for milking, draught-power, etc. In other words, the investment/saving function seems most important, especially in the absence of banking facilities. Inadequate knowledge exists about the most important veld products derived from the rangelands. Some such as the grapple plant have a commercial potential.

11.2 Causes of Change

The changes may be related to natural causes (e.g., drying up of surface water?) but are mostly related to socio-economic factors and processes. The primary "drivers' of change are:

Households with access to cattle and/or formal employment (75% of households have at least one absentee member) reduce their hunting and gathering activities. However, they constitute a market for game and veld products, hence livestock and employment do not necessarily reduce the pressure on such resources. The extent to which subsistence and/or commercial use contribute to each form of resource depletion needs to be researched.

Secondary causes of rangeland changes are:

11.3 Livelihood Sources and the Impacts of Rangeland Changes

As a result of the changes, the sources of livelihood have changed. For the study area as a whole, livestock has become the most important source of livelihood. However, it should be immediately emphasised that livestock distribution and hence the benefit distribution is not equitable. Half of the population does not benefit; most livestock owners have too few animals for a sustainable source of livelihood. We tentatively conclude that the livestock sector is of great importance for rural livelihoods but that the sector is unable to ensure decent living standards for the majority of the rural population. The experience in other parts of the country is that small herds are likely to be squeezed out, and few large herds will survive and control most rangelands. Therefore, the livestock sector is unlikely to be the engine for rural livelihoods and development.

The government has directly or indirectly become another major source of livelihood. Government employment is an important source of income for most households; moreover, government subsidies, support to the poorest and pensions have become important to such an extent that people have become dependent on government.

Wildlife utilisation remains important but is less so because of hunting bans and the decline in wildlife numbers. This has affected in particular the poorest and the Basarwa. Very little is known about the importance of veld products for livelihoods. The Basarwa are more frequently involved and by inference more dependent on it. Increasing hopes are pinned on non-consumptive wildlife utilisation to boost rural development.

Crop production is practised but almost as a last resort if no alternative sources of livelihood exist. It does not have the capacity to compensate for wildlife losses.

The importance of livelihood sources, and therefore the impacts of changes therein because of rangeland changes depend on the following socio-economic variable: income, gender, education and ethnicity. These variables need to be incorporated in surveys and rapid assessment methods.

With a few exceptions (salt production and leather works), there are few natural resource processing industries in the study area. Resource-processing opportunities require much more attention in future.

11.4 Livelihood Dynamics

There is evidence that households quickly adapt their sources of livelihood in response to constraints and new opportunities. New opportunities have been offered by livestock (cattle and especially small stock), formal employment and government assistance. Arable production is probably also a relatively new activity, but it relatively unattractive. Those engaged in livestock and/or with access to formal employment can afford to withdraw from the more arduous activities such as hunting and gathering. Such switches in livelihood strategies do not necessarily reduce hunting and gathering as a commercial market for game meat and veld products may emerge to serve the needs of livestock owners and people in formal employment. The result is rural specialisation rather than a significant change in the pressure on the resources.

11.5 Compatibilities and Incompatibilities between Rangeland Uses

Traditionally, rangelands are used for multiple purposes. Increasingly, rangelands are earmarked for one major form of use, either livestock or wildlife. In the study area, most rangelands have been earmarked for wildlife utilisation but management plans do not yet exist; the district land use plan does not yet have legal status. Therefore, livestock encroachment is still taking place from borehole areas to the east and from the settlements outwards. There seems to a direct trade-off between livestock and wildlife: wildlife disappears when cattle move in. The situation is less clear with respect to veld products, and requires further investigation. The study should identify the most important trade-offs and explore opportunities for multiple use, leading to higher and more resilient production. For example, veld products could be cultivated together with crops.

11.6 Signs and Impacts of Global Change

Physical data do not yet prove that temperatures have risen in the last decades. There are, however, phenomena which may be linked to global change such as:

The study will further explore global warming through the analysis of physical data (average and variability), archival work, and interviews with village elders. Important questions to be answered include the link between rainfall and primary and secondary productivity. This includes the issue of foraging (browse/grazing) and watering habits.

The report shows clearly that there have been major changes in land use and state of the vegetation.

11.7 Role of External Factors

External factors alleviate developmental and environmental problems but they may also aggravate them. Income generation elsewhere which supports local livelihoods, is a positive impact. In the process, direct resource pressure is being reduced. Negative impacts include the immigration of livestock owners into the area (to what extent this is happening needs to be examined) and absentee livestock ownership impeding proper management. It is important that the study explores these aspects and considers how the local population can increase its benefits from rangelands because of new opportunities offered by improved roads, accessibility to external markets and upgrading of the region's infrastructure and parks.

ANNEX 1

List of Abbreviations

ADP Agricultural Development Policy

ALDEP Arable Land Development Programme

APRU Animal Production Research Unit

ARAP Accelerated Rainfed Arable Programme

ARB Agricultural Resources Board

BAMB Botswana Agricultural Marketing Board

BIDPA Botswana Institute for Development Policy Analysis

BMC Botswana Meat Company

CSO Civil Society Organisation

DWNP Department of Wildlife and National Parks

FAP Financial Assistance Policy

GoB Government of Botswana

KDC Kgalagadi District Council

KDDP Kgalagadi District Development Plan

LSU Livestock Units

MLGLH Ministry of Local Government & Housing

NORAD Norwegian Agency for Development Cooperation

RAD Remote Area Dwellers

RADP Remote Area Development Programme in Botswana

SGL Special Game Licence

TGLP Tribal Grazing Land Policy

TOS Total Dissolved Solids

WMA Wildlife Management Area

* For these references, please contact:

Dr J W Arntzen
Department of Environmental Science
Private Bag 0022
Gaborone, Botswana
Tel: +267 355 2524
Fax: +267 356 591
e-mail: arntzenj@noka.ub.bw