Zambezia (1979), VII (ii).A SCENARIO OF CHANGES IN SUBSISTENCELAND USE AND ITS RELEVANCE TO THETRIBAL AREAS OF ZIMBABWEJ. R. WHITLOWDepartment of Geography, University of RhodesiaPOPULATION PRESSURES, both human and livestock, and widespread landdegradation are two major problems in the tribal areas of Zimbabwe,placing severe constraints on development prospects in many regions. Whilstsubsistence agriculture forms the foundation of the tribal economy, it issupplemented by an inflow of cash and goods largely derived from migrantworkers and in recent years cash cropping and sales of livestock havebecome increasingly important elements of tribal farming. Therefore exter-nal influences have greatly modified the subsistence agriculture in the TribalTrust Lands, but this has been matched equally by internal influences andin particular increasing population pressure.There is a great need to improve standards of living in the TribalTrust Lands but generally attempts to improve agricultural production andconservation practices have experienced varied and sometimes limitedsuccess. Various factors have militated against the adoption of moreefficient and productive farming methods, including the tribal land tenuresystem, the socio-economic circumstances of individual farmers and theinherent conservatism of tribal populations faced with increasing numbersof people and conditions of land hunger. Political factors have also playeda role in this resistance to change; for example the construction of contourbanks, prohibition of cultivation adjacent to streams and compulsory destock-ing although necessary for conservation and agronomic reasons, have been'imposed' on tribal farmers and, especially where these have involved time,effort and loss of income, such practices have been resisted strongly.Land hunger and food production have attracted renewed attentionwith the changing political circumstances in this country. The present dualfarming economy with commercial and subsistence sectors is unlikely tochange substantially over the next few years, but there is already greatpressure to take over apparently 'under-utilized' commercial farmlands forthe resettlement of African families from congested tribal areas. In therecent Five Year Plan for Rural Development (Ministry of Finance, 1978)areas of highest population pressure and lowest per capita incomes in theTribal Trust Lands have been designated as 'Intensive Rural DevelopmentAreas-' (IRDAs). Concerted efforts to develop these IRDAs are likely to171172SUBSISTENCE LAND USE AND ITS RELEVANCEbe hampered by a lack of funds, personnel and the low agricultural poten-tial of the tribal regions (Whitlow, 1980a). However, this does repre-sent a positive move to improve agriculture and general economic conditionswithin the Tribal Trust Lands and if combined with controlled resettlementschemes should achieve a reasonable level of success.Since population - resource relationships and land use are majorissues in many development projects it is perhaps timely to review thepresent status of agriculture within the tribal areas, especially as this hasa bearing both on resettlement schemes and development prospects. Thereis an extensive literature on tribal agriculture, and so, to avoid tediousrepetition, the present paper will attempt to synthesize the various elementsof the current situation by using a scenario of changes in subsistence landuse under population pressure. Before this is presented and its relevanceto the tribal areas is evaluated, it is necessary to describe some of thegeneral characteristics of the Tribal Trust Lands.GENERAL BACKGROUNDCrop and livestock production in this country are greatly influenced bythe amount and variability of rainfall. The Agro-Ecological Survey (Conex,1976), defines five natural regions on the basis of rainfall; conditions becomeincreasingly marginal for agriculture from Region I to Region V as theamount and reliability of rainfall decrease. Commercial dryland cropping isnot considered a feasible proposition in Regions IV and V, these areas beingutilized largely for livestock and game ranching in the commercial sector.However, the tribal farmer is forced by circumstance to attempt to growcrops in these low rainfall regions.Table IDISTRIBUTION OF LAND AND CULTIVATION ACCORDINGTO AGRO-ECOLOGICAL REGIONS (1977)A. PERCENTAGE OF LANDTribal Trust LandsCommercial Farming Areas*ZimbabweB. PERCENTAGE CULTIVATED ANDTribal Trust LandsCommercial Farming AreasZimbabwe/0,52,91,8Agro-Ecological RegionII8,926,815,4FALLOW38,47,012,642,229,332,7///16,422,118,533,110,620,0IV44,424,837,423,14,516,8V29,823,426,9Total100,0100,0100,0Average17,12,811,224,712,216,4Source: Whitlow (1979a)Ł Former Intensive Conservation AreasJ. R. WHITLOW 173The figures in Table I indicate the present situation with regard to thedistribution of land of varying quality and the proportions of cultivatedland. For comparative purposes the data on commercial farming areas andnational averages are presented; a comprehensive discussion of the datais presented elsewhere (Whitlow, 1979a).For present purposes it is necessary to note the following. Firstly,nearly 75 per cent of the tribal areas occur in Natural Regions IV and V,which are marginal for agricultural production; and secondly, there is anextremely high proportion of the land under the plough in the tribal areaseven in regions not suited to cropping. The implications of these two factsin relation to agricultural development are self-evident. Moreover, the highproportion of land under cultivation is symptomatic of land hunger andpopulation pressure. Kay (1975) has described 57 per cent of the tribal areasas being over-populated or grossly over-populated. This assessment wasbased on the 1969 census when the population of the Tribal Trust Landswas estimated to be about three million; the current estimate of tribal popu-lation is approaching four million people, and so conditions have probablydeteriorated considerably since 1969. High population densities can besustained in the tribal areas but the consequences include recurrent foodshortages after poor rainy seasons and extensive land degradation (Kay,1976). Obviously one objective of general development in the tribal areasshould be to relieve pressure on the land as a means of improving standardsof living and also to prevent further environmental deterioration.Against this general background the scenario on subsistence land usewhich is relevant to the situation in the Tribal Trust Lands can be presented.SCENARIO OF SUBSISTENCE LAND USEDescriptive scenarios of the type present in Figure 1, whilst having certainlimitations, have the advantages of:providing a useful framework for discussion;identifying key problem areas within the broader context of a landuse system;incorporating important concepts on population - resource relationships;being applicable to individual resources and at varying scales from aregional to a local level.Whilst 'scenario analysis' can be used to predict possible future changes ina given system (e.g. Meadows, Meadows, Randers and Behrens, 1972), thepresent scenario is restricted more to the four elements mentioned above.It is based largely on experience in Central Africa (Kay, 1979), althoughit is relevant to subsistence land use systems in general.174SUBSISTENCE LAND USE AND ITS RELEVANCEHUNTINGFISHINGGATHERINGGRAZINGWUIN N/IWA44cROPLANpyHuntingFISHINGGATHERINGGRAZINGCRC>PLYield = 4U per HaYlold = 4U per Haitttt innnmStage 1Stage 2GatheringGrazingŠDEOKIPL;kNDSOILTERIORATKDNGatheringGrazingŠPL/>,NDIMPROVEDLAND USEYield = 2U per Ha Yield = 5U per HaIIIII hhhhh2 2 2 2 9 2A2A2A2A2AStage 3AStage 4Figure 1: A SCENARIO OF A SUBSISTENCE LAND USE SYSTEM PLACED UNDERPOPULATION PRESSURE (after G. Kay, 1979, personal communication)An integral part of the scenario is the concept of carrying capacitywhich can be stated as follows:any area of land will support in perpetuity only a limited numberof people [and livestock] ... if this limit is exceeded, without acompensating change in the system of land use, then a cycle ofdegenerative changes is set in motion, which must result indeterioration or destruction of the land and ultimately in hungerand reduction of the population (Allan, 1949, p.l).The concept has been commented on by various authors and differentmethods have been proposed as a means of assessing carrying capacity (e.g.Allan, 1949; Street, 1969; Ricci, 1978). In subsistence cultivation systemsthere is a strong correlation between increasing population pressure andincreasing agricultural intensity (Turner, Hanham and Portararo, 1977);for example, there is a reduction in the crop - fallow ratio as pressureon the land increases. Carrying capacity is generally expressed in terms ofa given number of people or livestock per square kilometre and dependsin part on land quality; thus higher quality land can support greater numbersof people and livestock than poor quality land. Optimum densities or safecarrying capacities can be identified and related to actual densities toJ. R. WHITLOW I7Sdetermine the extent to which carrying capacity is exceeded or not yetreached. If the critical densities are exceeded then the land degradationprocesses mentioned by Allan (1949) are initiated and these can substantiallylower the carrying capacity of the land thereby reducing the populationsthat can be effectively supported in a given area.There are four stages in the subsistence land use scenario proposedby Kay (1979) and shown in Figure 1. Whilst these are presented as discretestages it should be born in mind that in reality they form a continuum.The four stages are as follows:Stage 1: this is characterized by low population densities and no pressureon the land. Areas which are not cropped provide a variety of foodstuffsand materials which contribute significantly to the upkeep of subsistencehouseholds, especially during the 'hungry months' and drought periods(Puzo, 1978). Also there are areas which are not utilized at all since thesmall populations can be supported on a limited area of land.Stage 2: as the human population increases so the area under cultivationis extended to produce more food. Numbers of livestock also increase, butthe extension of cropland has not reduced the area available for grazingand food supplies can still be supplemented by fishing and hunting. However,there is no unutilized land remaining and whilst crop production levelshave been maintained, soil fertility has declined as a result of leaching,soil erosion and crop removals. The critical thresholds in the physicalenvironment have been reached and a cycle of degenerative changes is setin motion. In the early stages this will have very little impact on humanand livestock populations, but changes in natural systems tend to be closelyinter-related and reinforcing (Chorley and Kennedy, 1979); therefore it isnow just a question of time before declining land quality affects thesepopulations.Stage 3: prolonged cultivation has resulted in declining yields, and so tomaintain production levels the area under cropping is extended at the ex-pense of the grazing lands. The areas taken over for cropping at this stageare likely to be characterized by less fertile soils and steeper slopes, thatis marginal lands; therefore yields from these croplands are unlikely to behigh. Although there has been no increase in human populations, the earlysymptoms of pressure are present. These include the following:Fishing and hunting no longer form a significant part of the subsistenceeconomy; thus certain dietary deficiencies may be initiated amongstthe human population.Gathering of foodstuffs and materials such as woodfuel from the in-digenous vegetation is less affected, except in so far as it requires moretime and effort to obtain these commodities than before (Whitlow,1979b).The main impact of exceeding the carrying capacity is manifested176 SUBSISTENCE LAND USE AND ITS RELEVANCEamongst the livestock and the grazing lands; over-grazing and selectiveremoval of palatable species reduces the number of livestock that canbe effectively supported on a more restricted area of grazing lands.This stage can be referred to as 'subsistence carrying capacity'(Dasmann, Milton and Freeman, 1973) and the livestock experiencestarvation, malnutrition and mortality levels, which, especially indrought periods, are extremely high.Stage 4: this final stage has been described by Kay (personal communication,1979) as follows:improvements in the use of cropland to gain increased returns froma given area can support an increase in population, but this adds furtherpressure to the non-cropland leading to a greater scarcity of all re-sources derived directly from the biosphere . . . also donkeys andgoats tend to replace cattle on degenerate scrub grazing which is stillone stage before livestock are pushed out altogether, and man is reducedto a vegetarian being surviving on a narrow range of crops.Elements of this scenario would seem to apply not only to the tribal areasof Zimbabwe, but also to other developing countries in the tropics wherethe F.A.O. (1967, p.53) have reported that 'increases in agricultural pro-duction over the years have been accomplished more by expanding the areascultivated than by raising productivity'. The extension of cultivation hasbeen one of the main factors contributing to widespread global deforestation(Eckholm, 1977) and has been described as 'merely "buying time" againstthe land hunger problem which must be solved rather through agrarianreform and informed land use policy' (F.A.O., 1967, p.54). This statementis particularly relevant to Zimbabwe at the present time since the resettle-ment schemes which may become operational over the next few years canonly be a temporary panacea for the sensitive political issue of land hunger.APPLICATION OF SCENARIO TO TRIBAL TRUST LANDSThe relevance of the subsistence land use scenario to the tribal areas cannow be evaluated. Discussion will be carried out under the following:1. Changes in cultivated areas2. Utilization of woodland resources3. Livestock in the tribal areas.There are other factors which could be examined which provide furtherevidence of population pressure. For example, unbalanced sex ratios amongstthe tribal population could provide an index of the degree of pressure(Hunter, 1966); hence areas experiencing severe pressure would be morelikely to lose active males through rural - urban migration than areas wherethere was little or no pressure on the land. Kay (1972) has described howselective migration has 'leached' the tribal areas of Zimbabwe of their mostJ. R. WHITLOW 177capable workers Š a fact which has important implications for ruraldevelopment programmes. However, to date, the structure of rural popu-lations has not been used as an index of population pressure in individualtribal areas.1. Changes in cultivated area: Nearly 25 per cent of the Tribal Trust Landsare at present under cultivation or fallow and in 17 per cent of areas theproportion of cultivation exceeds half of the available land (Whitlow,1979a). Moreover, there is a strong (and not unexpected) correlation betweenthe areas of greatest population pressure and areas with a high proportionof cropland (Whitlow, 1980b). Increases in area under cultivation withinthe Tribal Trust Lands can be seen as a response to the following:(a) increases in population, from about 500,000 people in 1900 tonearly 4,000,000 in 1977;(b) declining productivity of croplands necessitating extension ofcultivation to maintain and increase production levels.Another factor which has resulted in the extension of cultivation is theintroduction of cash crops such as cotton and groundnuts. The tendencyhas been to continue growing the staple food crops, with additional areasbeing used for cash crops. Moreover, crops such as cotton constitute soilerosion hazards since they provide only limited ground cover during theearly growing season when intensive rains are common (Stocking andElwell, 1976); this would result in a further decline in productivity andexacerbate the pressure on cultivated lands. Estimates of the changes incultivated land from 1900 to 1977 are presented in Table II. Althoughthese figures are undoubtedly subject to errors they provide an indicationof the general magnitudes of changes in area under cropping. The author'sestimates are based on the sampling of over 8,000 aerial photographs forthe period 1972 - 7; a range is given for these estimates corresponding withvalues of 35 per cent and 45 per cent fallow for the higher and lowerestimates of cropland respectively (Whitlow, 1979d).Table IICHANGES IN CULTIVATED LANDSIN THE TRIBAL AREAS, 1900-1977Year19001945196219751972 - 7Cultivated AreaPer Person (ha)0,430,480,520,600,55 - 0,65Total Area UnderCultivation (ha)215 000480 0001 134 6001 800 0002 304 700 -2 584 000SourceKay (1970)Kay (1970)Stubbs (1979)Walker (1975)My Calculation178 SUBSISTENCE LAND USE AND ITS RELEVANCEUsing the estimates presented in Table II it is possible to calculatethe percentage changes in cultivated land from 1900 to 1977 and to relatethese to the population increase over the same period (Table III). The highervalues of my estimates are taken as more realistic for 1977, since Walker(1975) suggests that his estimate of cultivation is a conservative one andalso the aerial photography covering the more densely settled areas of theTribal Trust Lands was taken during 1972-5; one might therefore expectthat there had been a slight increase in area under cultivation by 1977.Table HItPERCENTAGE CHANGES IN CULTIVATED LAND ANDPOPULATION IN THE TRIBAL AREAS, 1900 -1977________ŠŠ_Š_Š________________________Š_____________ i >Year Cultivated Area Total Area Under Tribal IPer Person (ha) Cultivation (ha) Population ,1900 0,43 215 000 500 000* |1977 0,65 2 594 000 4 000 000 I% increase 50 1 100 700Łrelates to total population in the country since no figures relevant to the present ftribal areas were available for 1900.IIEvidently there has been a considerable extension in the area undercultivation since 1900; but, more significantly, this has involved an increasein area of land cropped per person, a feature which is symptomatic of asubsistence cultivation system under pressure and attempting to maintainproduction levels in the face of declining soil fertility. Thus an increase of Ł-700 per cent in population over the survey period is considerably less thanthe 1 100 per cent increase in the area under the plough. This trend conformswith the changes suggested in the scenario on land use in Figure 1. Inaddition, nearly 25 per cent of the tribal areas are characterized by domedinselberg or 'dwala' terrain which greatly limits the potential for furtherextension of cultivated lands since up to 35 per cent of these regions maybe occupied by rock outcrops (Whitlow, 1980b). Such areas, unfortunately,coincide with some of the most densely populated Tribal Trust Lands inthis country. This means that the final stage of Kay's scenario may alreadybe found in certain areas but these may still be lacking the 'improvementsin the use of cropland' necessary to sustain further increases in population.This, combined with the disruption of medical facilities caused by thepresent war, may result in widespread human suffering and possiblymortalities in these regions.2. Utilization of woodland resources: An important element of the subsistenceland use scenario is the gathering of foodstuffs and materials from theJ. R. WHITLOW179n© TOIAi. OF SAMPlL Ai \vm. -i PIn! %i'n-Figure 2: SHORTAGES OF WOOD IN FIVE TRIBAL AREAS IN ZIMBABWEindigenous vegetation. As the population increases and the cultivated areaextends at the expense of woodlands, so increasing pressure is placed onthe diminishing resources of the biosphere. A recent pilot questionnairesurvey on the use of woodland resources in five'tribal areas has revealedsome significant population - resource relationships with respect to theshortages of timber for firewood, building and fencing (Fig. 2). The detailsof the five tribal areas sampled are given in Table IV. Although the totalnumber of households sampled was small (72), there were distinct differencesbetween the five areas which suggested varying degrees of pressure on thewoodlands. Shortages of timber were assessed on a qualitative scale of'minor', 'moderate' and 'severe'.Table IVPOPULATION DENSITY IN TRIBAL TRUST LANDSNumber ofHouseholdsSampled1515151512Source: Whitlow (1979b), p.5.Sample AreaMangwendeChiwunduraNdangaMarandaSemukweDensity of Population/km1(1977Human5059542422estimate)Livestock428976328618O SUBSISTENCE LAND USE AND ITS RELEVANCEThe trends in Figure 2 are a function of not only population pressure,but also the characteristics of the local woodlands and the availability ofpreferred species (Whitlow, 1979b). In general it would seem that there areshortages of all three commodities of firewood, building and fencing timberin the tribal areas (Fig. 2,F). The shortage of building timber is mostacute followed by firewood and then fencing timber. There are howevermarked differences in the degrees of shortages of these commodities betweenthe five sample areas. The most severe shortages occur in Mangwende andChiwundura tribal areas which have extremely high population densities(Table IV). Shortages of timber especially for building are beginning tobe felt in Ndanga which also has a high population density; a more recentsurvey suggests that the shortages in this area are more acute than revealedin the pilot study (Munzwa, 1979). The tribal areas with the low populationdensities, Maranda and Semukwe, seem to experience only limited pressureon their woodland resources, although the early stages of shortages arealready present in Semukwe. In both Chiwundura and Mangwende TribalTrust Lands it was discovered that shortages of timber had resulted in thesupplementing of firewood with maize and paraffin and that about 50per cent of dwellings were made of brick, partly because of the shortagesof more traditional materials. This is a clear symptom of pressure on thethe woodland resources which can be noted in other densely populated tribalareas. Such a situation would occur in Stage 3 of the subsistence land usescenario.Critical areas of woodland shortages have been identified provisionallyby Whitlow (1979c). These areas constitute about 20 per cent of the TribalTrust Lands and correspond mainly with areas described by Kay (1975) asbeing over-populated or grossly over-populated. In many of these areasthere are only remnants of woodlands surviving on the steep and inaccessibleslopes; one suspects that in the absence of steep-sided hills these regionswould be completely stripped of their woody vegetation. However, whilstit is relatively easy to identify areas where extensive destruction of thewoody vegetation has already taken place, it is more difficult to identifysymptoms of shortages in regions which still have considerable areas ofwoodland remaining. Shortages may well occur in these areas especiallysince there are definite preferences for different species related to theirsuitability for different purposes in the rural households (Whitlow, 1979b).However, these shortages are unlikely to be as severe as in the over-populated regions Š perhaps representing a phase between Stages 2 and 3of the land use scenario.3. Livestock in the tribal areas: Livestock, especially cattle, have a multi-purpose role in the subsistence agriculture of the tribal areas. This variedrole includes provision of draught power, maintenance of soil fertility, foodproduction and, finally, a cultural need (Le Roux, Stubbs and Donnelly,J. R. WHITLOW1811978). There are in fact very good arguments for increasing the numbersof livestock in the tribal areas, but given the current situation of 'overstock-ing' this is not a practicable proposition at present. The problems and pros-pects of livestock in the tribal areas have been discussed by numerousauthors (e.g. Prescott, 1961; Hornby, 1968; West, 1968; Barnes, 1978).The present discussion will be confined to the following:(a) consequences of exceeding carrying capacity on range lands;(b) present distribution of livestock and changes in composition;(c) status of grazing in tribal areas.In terms of the subsistence land use scenario (Fig. 1), Kay (1975, p.8) hasstated that 'the extensions of cultivated area bring increasing scarcity inthe grazing lands, and acute effects of over-population and over-stockingare first evident amongst livestock and in the grazing areas'. Since livestockand grazing conditions provide the early symptoms of population pressure,these will be discussed in greater detail.c 100-10couIIw i:ei5 8n, NurrSoil Erilatioand£1I2 0-CarryingCapacity"T* Rangt Condition111! "^111Animal Sp«cws|Š ty1Soil Erosion ! ^/111 1%.\\\\\ /"XŁŁŁ^1Shntwrnsh^ Rills, Oullits ^*Ł and Wind Erosion\v \X \X x\\ \FINAL STAGE-low biomaM-f«w plant andanimal si*'"Łrosionandsurtaci iŽ10"-giwrtly raducadcarrying capacityTime Seal*Stages in Degradation of RangelandsFigure 3: CONSEQUENCES OF EXCEEDING CARRYING CAPACITY ON RANGE-LANDS (after Savory, 1969)(a) Consequences of exceeding carrying capacity on range lands:Some of the ecological aspects of overstocking and degradation on rangelands applicable to the tribal areas are shown in Figure 3. This was adaptedfrom a scheme described by Savory (1969) with reference to game popu-lations and applies particularly to the lower rainfall regions. The six stages182 SUBSISTENCE LAND USE AND JTS RELEVANCEin the sequence of range land deterioration are as follows.Stage 0: carrying capacity has not yet been exceeded and there is 'apparent'stability within the system, with a variety of plant and animal species andstable soil-water conditions. Natural erosion is at a low level.Stage I: critical thresholds in the system are exceeded as the numbers oflivestock increase beyond the carrying capacity. Selective foraging by thecattle results in the palatable perennial grasses giving way to less favouredannual grasses, weeds and shrubs. Decreasing plant cover promotes an in-crease in soil erosion and the general degradation of the habitat reducesthe number of wild animals that can be supported.Stage H: environmental deterioration continues with degradation processesbeing closely inter-related and reinforcing. The cattle population is, however,still increasing but is surviving at a subsistence level (Dasmann, Milton andFreeman, 1973). This places further pressure on the range lands and sopromotes further degradation.Stage III: a critical stage is reached where land degradation is so faradvanced that starvation, malnutrition and mortalities of livestock arecommonplace especially during dry periods. Bush encroachment may becomea major problem in semi-arid areas, greatly reducing the effective carryingcapacity for cattle, but favouring browsing stock.Stage IV: considerable areas of the range lands have now been stripped oftheir protective plant cover and goats replace cattle on degraded pastures.Erosion rates continue at a high level but by this time much of the topsoilwill have been removed. Poor infiltration and high run-off result in increasingaridity and this, combined with soil erosion, makes regeneration of therange lands extremely difficult.Stage V: this represents the final stage in the degradation of the rangelands which in ecological terms may constitute the 'point of no return'.From being a reasonably balanced and productive system, the final stagesof degradation are characterized by: low biomass; few plant and animalspecies; excessive soil erosion and surface run-off; and a greatly reducedcarrying capacity.Whilst the individual details of this scheme may vary from one tribalarea to another the general pattern is the same. For example, with referenceto a tribal area in the south-east lowveld, Kelly and Walker (1976, p.581)state that 'intensive utilization was resulting in a change of dominance fromperennial to annual grasses, with increasing year to year fluctuations inseasonal production, a large proportion of bare soil and increasing run-offof rain water'. Describing general conditions in the tribal areas of NaturalRegion IV and V, a scientific sub-committee (1969, p.12) stated that:while grazing during the dry season is relatively harmless, severeover-grazing during the growing season, coupled with the povertyof the water regime, has led to perennial grasses giving way toannual grasses and weeds. This general reduction in plant coverand the compaction of the exposed soil surface, has enhanced run-off so that the effectiveness of an already erratic rainfall is furtherreduced..J. R. WHITLOW183Rainfall amount and variability have been identified as key 'limitingfactors' influencing the carrying capacity of range land particularly in lowrainfall regions (Condon, 1968). Hence carrying capacities decrease withdecreasing rainfall, and natural pastures are even more susceptible to seriousdeterioration through overstocking. Moreover, the reclamation of degradedgrazing lands in semi-arid and arid areas is likely to be an extremely costlyand slow process requiring a combination of destocking (i.e. removal ofcause of degradation) and changes in pasture management (Barnes, 1978).Since nearly 75 per cent of the Tribal Trust Lands are in low rainfallregions which have inherently low carrying capacities, it is not unexpectedthat many of these areas exhibit elements of Stages III and IV of the schemepresented in Figure 3.Figure 4: DENSITY OF CATTLE, SHEEP AND GOATS IN THE TRIBAL TRUSTLANDS (1977).(b) Present distribution of livestock and changes in composition:The present distribution (1977) of cattle, sheep and goats is shown inFigure 4 and the relevant statistics are summarized in Table V. Althoughareas in the extreme north-west and north of the country are indicated as184SUBSISTENCE LAND USE AND ITS RELEVANCEhaving low densities of livestock (i.e. where data were available), it is un-likely that cattle occur in these regions because of tsetse fly infestation (Fig.4). The highest densities of livestock, over 50 animals per square kilometre,occur in an arc of tribal areas extending from Plumtree through to Belingweand up to Buhera. Other high density areas include the Tribal Trust Landssouth of Gokwe and around Salisbury and Hartley. These constitute 59per cent of the Tribal Trust Lands (Table V). In the more arid regions inthe south of the country, densities of more than 30 animals per square kilo-metre are not uncommon. These densities must, however, be compared withthe assessed carrying capacities of the different natural regions. Figuresbased on Conex estimates and the original agro-ecological survey of Vincentand Thomas (1960) are given in Table VI (Coleman, 1979).Table VDENSITIES OF CATTLE, SHEEP AND GOATS IN THETRIBAL AREAS, 1977Animals/kmzunder 1111-2021-3031-4041-50over 50TOTALPercentageCattle27,318,116,225,18,84,5100,0Sheep88,011,80,2ŠŠŠ100,0of TribalGoats49,026,011,013,30,7100,0AreaTotal*8,021,010,010,012,039,0100,0* cattle, sheep and goats.Source: derived from Ministry of Internal Affairs, Livestock Returns, 1977,Tabk VIESTIMATED CARRYING CAPACITIESOF CATTLE ACCORDING TO NATURAL REGIONSNaural Region(Conex, 1976)IIIIIIIVVCarrying Capacity(L,S.U.*/ha)4 ha4 ha6 ha8 ha12 haCritical Densities(cattle I kni2-)252517128* L.S.U. = Livestock Unit (1 mature beast or 5 small livestock such as sheepand goats).J. R. WHITLOW 185Comparing these assessed carrying capacities with the actual densities inTable V, one can appreciate that with respect to cattle alone, well over 50per cent of the tribal areas are experiencing varying degrees of overstocking.In the more arid regions carrying capacities have been exceeded up to threetimes the safe stocking levels. These figures certainly suggest that the over-stocking rates given by Cross (1977) are under-estimates, especially whenthe extensions of cultivation described earlier are taken into account. Alsoshown in Figure 4 as an inset map is the rate of increase of livestock popu-lations for the period 1969-77. This indicates that the areas of highestdensities generally correspond with those experiencing a greater than 4 percent increase in livestock per annum. It is unlikely that such high ratesof population growth can be sustained in these regions for much longergiven the present state of land degradation and the current breakdown ofdipping services, which has resulted in the outbreaks of various tick-bornediseases (Norval, 1978). However, the high rates of increase in the tribalareas of Victoria province and southern Matabeleland, where carryingcapacities are extremely low, are cause for concern. The scenario on landuse changes (Fig. 1) suggests that an important symptom of pressure on thegrazing lands is a change in the composition of the livestock herd. Thusgoats replace cattle on degraded pastures, especially where bush encroach-ment is taking place, since they can browse on the woody species and aregenerally hardier animals. There has certainly been an overall trend ofmore rapid increases of goats and, to a lesser extent, sheep in the tribalareas (Table VII). Over the period 1964 - 77 there was over a 200 per centincrease in the goat population compared with about 160 per cent for sheepand less than 90 per cent for cattle. Over the same period the human popu-lation in the tribal areas increased by about 50 per cent (Central StatisticalOffice, October 1978, Table 1), which is considerably less than the livestockincreases!Table VIICHANGES IN LIVESTOCKIN THE TRIBAL AREAS, 1964 -19771964 NumberPercentage1977 NumberPercentagePercentage increase1964 to 1977Cattle1916 00071,53 592 00061,287,5Sheep186 0006,9486 0008,3161,3Goats579 00021,61 787 00030,5208,6Total2 681 000100,05 865 000100,0118,8Source: Central Statistical Office, October, 1978, Table 18.186SUBSISTENCE LAND USE AND ITS RELEVANCEŁS 11-20 map*EJ3 21-30 animals pw »quori$0 Ovtr 30 animal* p#r squart kilofnvtrt£3 Ma data avaiiabwf~l Non rr*ol ArtMAyo Ecologicol RtgKjnsE Fteg«n Four Y Ftagnn Fiw© Mom Cilm Ł Town* and VrikignFigure 5: DENSITY OF GOATS IN THE TRIBAL TRUST LANDS (1977).Also the proportion of goats as a percentage of the total livestock (cattle,sheep and goats) has risen from just over 20 per cent in 1964 to about 30per cent by 1977. The present distribution (1977) of goats is shown inFigure 5. From this it can be observed that the higher densities of goats(over 20/km2) occur in three main areas: east of Fort Victoria, south ofBulawayo and in the Gokwe region. "With the exception of Gokwe, theseareas have been described as bare to very overgrazed by Cleghorn (1966).In these regions it is not uncommon for up to 25 per cent of the cattle todie during severe droughts such as occurred in the 1963 - 4 and 1965 - 6seasons. This critical situation is characteristic of a well advanced Stage 3of the subsistence land use scenario.(c) Status of grazing in the tribal areas:A survey of grazing conditions was carried out in the early 1960s and isreported on by Cleghorn (1966). In this survey the veld condition wasdescribed according to four classes as follows:Bare: almost entirely devoid of grass cover for most of the year;Very overgrazed: grass constituents low in the plant succession andaffording a sparse ground cover. Dry season forage is in short supply;J. R. WHITLOW 187Moderate: the plant succession may have been put back but the veldis still in a productive state and is not over grazed to the extent thatforage is short every dry season; andGood: ungrazed or grazed within its carrying capacity so that the veldremains vigorous and productive. Dry season forage is sufficient exceptduring extremely severe droughts.The results of this survey are presented in Figure 6 and the statistical dataare summarized in Table VIII according to the natural regions defined byVincent and Thomas (1960). In examining this data it should be notedthat whilst a given area may be described as predominantly 'bare' or 'over-grazed' there may be localized areas of better grazing. Invariably the patternof grazing within the tribal areas is one of 'pressure points' around waterholes and night kraals with the result that extensive areas of veld may beunder-utilized; more efficient pasture management is certainly one methodof overcoming pressure on the land and increasing livestock production(Barnes and Clatworthy, 1976). Also the status of the grazing lands isdynamic with the result that areas described as 'moderate' in the early1960s may have become overgrazed since then as a result of increases inlivestock population. Similarly a 'bare' region might have experienced asuccession of droughts thus resulting in a reduction of livestock; con-sequently in better rainy seasons the condition of the grazing lands mighthave improved, although processes of regeneration are generally slowerthan rates of degradation.Table VIIICONDITIONS OF GRAZING LANDS IN THE TRIBAL AREASNaturalRegion*IIIIIIIVVAVERAGEBareŠ4,839,113,0Percentage of landVery Overgrazed21,465,568,229,726,336,4Moderate76,718,012,916,69,814,8Good1,916,518,948,924,835,8Total100,0100,0100,0100,0100,0100,0Percentage17145226100,0Source: Cleghorn (1966).* Based on Vincent and Thomas (1960).From Figure 6 and Table VIII it can be appreciated that over 50 percent of the tribal areas are overgrazed or bare and that these degradedpastures occur mainly in the low rainfall zones of the south-west and south-east of the country. A comparison of Figure 6 with previous maps showingthe varying densities of livestock indicates that there is a strong correlation188SUBSISTENCE LAND USE AND ITS RELEVANCEbetween high livestock densities and poor grazing conditions. The onlyareas where grazing lands are still in good condition are those in theextreme north-west and north of the country, where tsetse fly in effect haveprotected the veld through preventing or limiting the raising of livestock.Figure 6: CONDITION OF GRAZING LANDS IN THE TRIBAL AREAS (1964).CONCLUSIONA scenario of a subsistence land use system placed under increasing popu-lation pressure has been described. Elements of the scenario have beenevaluated with respect to the status of cultivated lands, woodland resources,livestock and grazing areas in the Tribal Trust Lands. From the evidencepresented it can be concluded that there is extreme pressure on the landwith many areas falling into Stage 3 of the scenario, especially in the arcof tribal areas extending from Belingwe through to Buhera. Also there issome evidence for elements of Stage 4 of Kay's scenario operating in certaintribal areas, but without the compensating changes in land use necessary tosustain higher population densities (Kay, 1979; Allan, 1949). The prospectsfor development in these regions, and the Tribal Trust Land as a whole,are likely to be limited by both human and livestock population pressuresJ. R. WHITLOW 189which already have resulted in extensive land degradation and seriousdepletion of resources. The subsistence land use scenario presented in thisarticle has the advantage of providing a holistic framework within whichthe present problems of the Tribal Trust Lands can be evaluated and suit-able solutions formulated.AcknowledgementsI would like to thank Messrs R. G. Wheeler, E. Chingwe and S. Gorogodofor the production of the Figures used in the text. The article is based inpart on research carried out on the behalf of the Natural Resources Boardexamining the problems of deforestation in this country. Professor G. Kaykindly provided an up-dated version of the subsistence land use scenariofirst presented by him in 1975. Mr P. Hawkins gave valuable commentson the original draft of this article.ReferencesALLAN, W. 1949 Studies in African Land Usage in Northern Rhodesia (Cape Town,Oxford, Univ. Press for Rhodes-Livingstone Institute, Rhodes-Livingstone PaperNo. 15).BARNES, D. 1978 'Problems and prospects of increased pastoral production in theTribal Trust Lands', Zambezia, VI, 49-59.BARNES, D. and CLATWORTHY, J. N. 1976 'Research in veld and pasture produc-tion in relation to the Tribal Trust Lands', The Rhodesia Science News, X, 271-8.CENTRAL STATISTICAL OFFICE 1978 Supplement to the Monthly Digest ofStatistics, October (Salisbury, Government Printer).CHORLEY, R. J. and KENNEDY, B. A. 1971 Physical Geography: A Systems Approach(London, Prentice-Hall).CLEGHORN, W. B. 1966 'Report on the conditions of grazing in the Tribal TrustLands', Rhodesia Agriculture Journal, LXIII, 57-67.COLEMAN, M. 1979 Personal Communication.CONDON, R. W. 1968 'Estimation of grazing capacity on arid grazing lands', in G.A. Stewart (ed.), Land Evaluation (Melbourne, Macmillan), 112-24.CONEX 1976 Agro-Ecological Map of Rhodesia, 1:1 At (Salisbury, GovernmentPrinter).CROSS, E. G. 1977 'The Tribal Trust Lands in transition : The national implications',The Rhodesia Science News, XI, 185-9.DASMANN, R. F., MILTON, J. P. and FREEMAN, P. H. 1973 Ecological Principlesfor Economic Development (London, Wiley).ECKHOLM, E. P. 1977 'The shrinking forests', Focus, XXVIII, i, 12-16.F.A.O. 1967 Wood: World Trends and Prospects (Rome. F.A.O.).HORNBY, H. E. 1968 'Overstocking: A modern approach to the problem', RhodesiaAgriculture Journal, LXV, 67-74.HUNTER, J. M. 1966 'Ascertaining population carrying capacity under traditionalsystems of agriculture in developing countries', Professional Geographer, XVIII151-4.KAY, G. 1970 Rhodesia: A Human Geography (London, Univ. of London Press).KAY, G. 1972 Distribution and Density of African Population in Rhodesia (Hull,Univ. of Hull, Miscellaneous Series in Geography No. 12).KAY, G. 1975 'Population pressures and development prospects in Rhodesia', TheRhodesia Science News, IX, 7-13.I9O SUBSISTENCE LAND USE AND ITS RELEVANCEKAY, G. 1976 'Population problems and development strategy in Rhodesia', ScottishGeographical Magazine, XCIII, 148-60.KAY, G. 1979 Personal Gommunication.KELLY, R. D. and WALKER, B. H. 1976 'The effects of different forms of landuse'on the ecology of a semi arid region in south eastern Rhodesia', Journal ofEcology, LXIV, 555-82.LE ROUX, P., STUBBS, A. J., and DONNELLY, P. H. 1978 'Problems and prospectsof increasing beef production in the Tribal Trust Lands', Zambezia, VI, 37-48.MEADOWS, D. H., MEADOWS, D. L., RANDER, J., and ANDREWS, W. W. 1972The Limits to Growth (London, Earth Island Press).MINISTRY OF FINANCE 1978 Integrated Plan for Rural Development July 1978[2nd impression, annexure to Proposals for a Five-Year Programme of Developmentin the Public Sector (Salisbury, Government Printer, 1979) ].MUNZWA, K. 1979 'Household Demand for Woodland Resources: A Study of LandUse Pattern and the Problem of Deforestation in Ndanga' (Salisbury, Univ. ofRhodesia, unpubl. B. A. dissertation in Geography).NATURAL RESOURCES BOARD 1969 The Use of Marginal Lands of Rhodesia(Salisbury, Report of Scientific Sub-Committee).NORVAL, R. A. I. 1978 'The effects of partial breakdown of dipping in Africanareas of Rhodesia', Rhodesian Veterinary Journal, IX, 9-16.PRESCOTT J. R. V. 1961 'Overpopulation and overstocking in the native areas ofMatabeleland', Geography Journal, CXXVII, 212-25.PUZO, B. 1978 'Patterns of man land relations', in M. J. A. WERGER (ed.).Biogeography and Ecology in Southern Africa (The Hague, W. Junk), 1049-112.RICGI, P. F. 1978 'Policy analysis through carrying capacity', Journal of EnvironmentalManagement, VI, 85-97.SAVORY, C. R. 1969 'Range deterioration patterns and game utilization to SouthernRhodesia', in Wildlife Conservation Training Course (Salisbury, Univ. Coll. ofRhodesia and Nyasaland), 70-8.STOCKING, M. and ELWELL, H. 1976 'Vegetation and erosion: A review', ScottishGeographical Magazine, XCII, 4-16.STREET, J. M. 1969 'An evaluation of the concept of carrying capacity', ProfessionalGeographer, XXI, 104-7.STUBBS, A. T. 1977 'The Tribal Trust Lands in transition: Land use', The RhodesiaScience News, XI, 181-4.TURNER, B. L., HANHAM, R. Q., and PORTARARO, A. V. 1977 'Populationpressure and agricultural intensity', Annals of the Association of American Geo-graphers, LXVII, 384-96.VINCENT, V. and THOMAS, R. G. 1960 An Agricultural Survey of Southern Rho-desia. Part 1: Agro-Ecological Survey (Salisbury, Federal Government Printer).WALKER, B. H. 1975 'Ecological constraints to growth in Rhodesia', The RhodesiaScience News, IX, 18-20.WEST, O. 1968 'The vegetation of southern Matabeleland: A study of the reasonsfor its present degradation and possible means of rehabilitation', in The Associationof Scientific Societies of Rhodesia, 1st Rhodesian Science Congress (SalisburyM.O. Collins), 85-103.WHITLOW, J. R. 1979a 'An assessment of cultivated lands in Zimbabwe Rhodesia1972 to 1977', The Zimbabwe Rhodesia Science News, XIII, 233-8. 'WHITLOW, J. R. 1979b The Household Use of Woodland Resources in Rural Areas(Salisbury, The Natural Resources Board).WHITLOW, J. R. 1979c 'Deforestation: Some global and national perspectives'Geographical Association of Zimbabwe Rhodesia Proceedings, XII (in press). 'WHITLOW, J. R. 1979d 'An assessment of cultivated lands in Zimbabwe Rhodesia1963-1977/ The Zimbabwe Rhodesia Science News, XIII, 286 - 90 Khodesia>WHITLOW, J. R. 1980a 'Agricultural potential in Zimbabwe Rhodesia- A factorizedsurvey', Zimbabwe Rhodesia Agricultural Journal, LXXVII, (in press).WHITLOW, J. R. 1980b 'Land use, population pressure and rock outcrops in thetribal areas of Zimbabwe Rhodesia', Zimbabwe Rhodesia Agricultural Journal,LXXVII, 3-11.