- 76 - LIVESTOCK PRODUCTION AND FOOD SUPPLY TRENDS IN BOTSWANA JMP ARNrlEN Introduction '!his paper describes trends in li vestock developnent and food supply in Botswana. It deals firstly with a number of topics relevant to livestock production. SUbsequently, links between li vestock proouction and food supply are examined. '!hereafter, constraints to livestock production are reviewed before a brief look at the future is given. Numberof Livestock Table 1 sha.rs m.nnbers of cattle, goats, sheep and chicken in the past twenty years (1966-85). In general, there was a decline in the number of cattle rot a rapid increase in the numberof goats during the four years 1982-85 towards the end of the drooght. Overall, livestock numbers have increased SUbstantially since independence. 'Ibis is especially true for cattle. 'lhe average growth rate in the period 1971-1981 has been 3.6 percent which is CCIIparable to the human population gra.tth rate (estimated to be between 3.4 and 3.7 percent). '!he drOJght has lXlt a (teuporary?) halt to the cattle p:>pUl.ation gra.tth rate. It has caUSed17 percent decrease over the period 1982-1985. Overall, LivestOCk developnent has been heavily influenced by . government intervention thrcugh vetennary caDpaigns, breeding programmes, the Tribal Grazing Land Policy, etc. - 77 - Table 1: Deve10pnent of Numbers of Cattie, Goats, Sheep alXi Chicken (In 'lbaJsards) Cattle Goats Sheep Chicken 1966 1,237 308 151 n.a 1967 1,492 647 212 n.a 1968 1,688 703 231 n.a 1969 1,945 847 279 344* 1970 2,017 875 350 390* 1971 2,092 1,015 376 234* 1972 2,177 765 380 n.a 1973 2,138 I,m 400 295* 1974 2,249 1,350 420 n.a 1975 2,384 n.a n.a n.a 1976 2,512 1,400 420 n.a 1977 2,622 n.a n.a n.a 1978 2,880 616 108 n.a 1979 2,840 557 lOB 740 1980 2,911 638 149 833 1981 2,967 621 121 1,046 1982 636 140 1,146 2,979 1983 783 165 970 2,818 1984 8B9 167 709 2,685 1985 1,138 200 1,020 2,459 *'Ibeseare likely to be underestimates. (n.a. not available) Pigs have never exceeded 10,000 and have therefore been left alt. 1977-1985: 69-73 (Chicken only) SQrrces: Agricultural Statistics M::Ibnald, 1980. - 78 Smallstock, particularly goats, are highly vulnerable to diseases nvst prevalent during wet periods, but also very drought resistant (as browsers). The latter is clearly illustrated by the 83 Percent increase in the national herd of goats since 1981. Goats are suitable elements in a diversification strategy for livestock holders during drought. Emphasis by government and the Botswana Meat Camnision (ItC) on cattle (veterinary services and prices) has tended to result in ccmparative neglect of other livestock. Q1ly recently(l984), 1M: increased prices of goats substantially, and farmers responded quickly by selling 1,600 goats to Bt£ canpared with only 1,000 in 1980( 1M: Annual Report). Management Pbrms Livestock production takes place uooer tenurial arrangements: freehold, leasehold and cannunal tenure. Unfortunately, relevant Agricultural Statistics do not distinguish between leasehold and CCIIIIIIUnal tenure. Statistics for these categories are usually jointly grouped under traditional management. Freehold tenure is labelled 'camnercial'. Although it has been clearly established that large herds under tradi tiona! management reach similar productivity levels as freehold ranches (see Hul:bard, 1982: Carl Bro Int, 1982), a distinction between the t1«l can be used to illustrate differences in Iivestock holding strategies and production mix (see Tables 2). Cattle are the most camoonly held animals, ranking first and secorrl in freehold and 1easeho1d/CCIIIIIl1llalholdings, respectively. Chicken are nost camron in camm.ma1 areas. Sane specialized, large freehold farms, however, hold 21 percent of the total m.miber of chicken. Altha1gh goats rank third according to frequency of animals held under CCIIIIIl1llal forms of tenure, their numbers in freehold ranches are small. Although sheep are similarly not canmon, freehold ranches CMn 13.6 percent of their total number. Freehold farms are far nore - 79 - important than their small number (0.5 percent of total agricultural holdings) suggests. Table 2: Livestock UOOerFreehold am Ccmmmal/LeaseholdTenure (Ranked According to Frequency of Animals Held) Canrrunal/ Frequency of %of Freehold Frequency %of Leasehold animals kept national Farms or animals national (99.5% of Herd/ flock (5%of kept herd/flock total farms) Farms l. Olicken 75 78.8 1.Cattle 87 14.2 2. Cattle 71 85.8 2.lbrsF-s 51 22.5 3. Goats 62 97.3 3.Goats 51 2.7 4. D::>nkeys 99.0 4.Sheep 49 13.6 33 5. Sheep 86.4 5.Olicken 36 21.2 15 6.Lbnkeys 33 1.0 6. Horses 7 77.5 7. Pigs 7.pigs 8 9.7 6 90.3 Source: Agricultural Statistics (1984). Table 3 shaNs a strol'¥Jer market orientation ~ freero1d farmers ccmpared to the average "traditional" £armer, especially with respect to cattle. Sales are considerably higher before am duril'¥J the drought (1981 an:! 1985, respectively). Freehold farmers play an iJJllOrtant role in cattle trading as evident fran the relatively high purchaSe rate. '!he different sets of farmers resporded differently to the recent drought. Freehold farmers reduced cattle number by increasil'¥J net offtake by nine percent ocmpared to only 1.5 - 80 - Table 3: Scme ManagementIndicators of Livestock Carmuna1/1easeho1d Freehold 1981 1985 1981 1985 Cattle Sales rate 7.8 9.6 23.7 30.9 Halle slaughter rate 0.8 2.3 0.8 2.4 Purchase rate 1.3 1.6 12.7 10.9 ~ Sales rate 20.0 3.1 1.3 6.8 Hane slaughter rate 30.7 5.2 1.8 8.6 Purchase rate 8.5 2.5 0.7 3.4 Sheep Sales rate 5.7 3.7 Hane slaughter rate 4.8 15.5 Purchase rate 1.7 7.5 ~te: ~ data available for chicJc:en. - Salrce: Agricultural Statistics, 1985. (See :Uso M:::D.:>nald, 1980) percent for the average CCJnmunal-based traditional farmer. Salewhat surprisingly, in view of the extra cash needs, the average traditional farmers seems to have tAIilt up sizeable goat herds. It will be interesting to establish after the present drought whether this is a fonn of drooght adaptation or a structural diversification of the livestock sector. Either leY, it is important to direct governmental assistance to the actual needs of livestock farmers. Spatial Distribution of Liveatock Althoogh rost of the 1ivestock are kept in the eastern part of &:>tswana, IlDre than half of the cattle are kept in Central and - 81 - tbrth East district (see Table 4). A small prqx>rtion of the cattle are kept in the western part of the ca.mtry (Ghanzi and Kgalagadi). Relatively uore goats and particularly sheep are kept in Kgalagadi District. Livestock holding in small districts such as Kgatleng and Sooth East is Hmi ted due to grazing land shortage. 'Ibis has tended to lead to smaller herds (Gulbrandse, 1984: Arntzen, 1985). '!be spatial distribution of ron-freehold livestock over districts has been stable since 1978 (Arntzen and VeeneOOaa1,1986). Table 4: Spatial Distribution of Livestock (1981) % of Cattle % of Goats % of Sheep % of Farm fbldings Soothern District 13.3 17.6 25.1 17.2 South East 0.9 2.0 1.4 3.1 Kweneng 10.1 11.3 10.7 12.7 Kgatleng 4.4 4.8 2.9 6.1 Centraljtbrth East 53.7 37.8 25.8 44.9 Ngamiland 14.2 18.6 5.0 12.4 Chobe 0.2 0.2 0 0.8 Ghanzi 1.7 2.3 1.4 1.1 Kgalagadi 2.4 5.5 14.3 1.8 Salrce: Agricultural Statistics, 1981 - 82 Cooke (l985) argues that there has been a penetration of cattle into the western parts of the CXJUntry. --' lIIl s penetration has been facilitated by the drilling of deep boreholes am lmpr . ovoo veterinary services in Kgalagadi. While borehole drilling, has just managedto keep up with extra livestock numbers, it has not helped to allevlate. eXlstlng " ove rgrazing condi tions (SaMford, 1978). In stead, overgrazing has spread Wl 'th the increasing spread of oorehole drilling. Participation in Livestock ft:>lding In this section we examine the frequency arx3 extent of , lnvo 1Vettent of . reglonal pcp.llahon , groups . ln livestock holding. Table 5 sh::Msparticipation of agricultural holdings in livestock prcxiuction for different regions of the ca.mtry. Many households effectively do not seem to have agricultural holdings. Of the 135,634 rural households counted during the 1981 pop.liation census, only 84,200 (see Agricultural Statistics) bad agricultural holdings. In other \r, 1981). - 83 - 'Dlble 5: Participation of J\gricultural Holdings In Livestock Production (1984)* % with % with % with % with chicken cattle goats sheep Southern 75.5 78.5 62.3 23.8 South East 87.0 43.5 47.8 13.0 Kweneng 75.5 74.5 72.4 15.3 Kgatleng 86.3 64.7 43.1 7.8 Central/North East 75.5 72.3 63.2 14.1 N:Jamiland 61.5 68.1 56.0 8.8 Olobe 88.9 44.4 11.1 Ghanzi 66.7 66.7 17.8 22.2 Kgalagadi 61.1 66.7 83.3 27.8 Average 75 71 62 15 * A household with a set miniJl\Ulll agricultural involvement Source: Agricultural Statistics, 1984. Distriootion of Livestock AiIInJ Holders Livestock distriootion is generally skewed (nsuIIptionOf Animal-Based Food ItE!lllS (1984) a) In volume (metric tonnes) (1) (2) (3) (4)* Locally Imports Exports Cbnsumption Produced Meat beef 58,000 41 29,019 29,022 goats 1,900 0 7 1,893 sheep 494 13 16 491 chicken 209 11 2 218 others n.a 177 n.a n.a Total 60,603 242 29,044 31,801 or 83g1/p/d Milk +27,800 7,249 0+35,000 or 0.091t/p/d DJgs 2,147 113 0 2,260 b) In value (P.OOO**) Meat beef 120,073 96 60,076 60,093 goats 257 0 1 256 sheep 1,235 48 40 1,243 chicken 439 32 5 466 others n.a 518 2,004 n.a Total 122,004 694 62,126 60,572 Milk +18,785 5,251 12 +24,024 DJgs 3,420 180 0 3,600 * 4 = 1+ 2 - 3. ** IDeal production valued at export prices Sources: Calculations based on J\gricultural Statistics arrl External Trade Statistics. - 86 - Table 6: Productiat, Trade am. O:lnsuDptiat Of Animal-Based Food Items (1984) a) In volume (metric tonnes) (1) (2) (3) (4)* Locally Imports Exports Cbnsumption Produced Meat beef 58,000 41 29,019 29,022 goats 1,900 0 7 1,893 sheep 494 13 16 491 chicken 209 11 2 218 others n.a 177 n.a n.a Total 60,603 242 29,044 31,801 or 83g1/p/d Milk +27,800 7,249 0+35,000 or 0.091t/p/d Eggs 2,147 113 0 2,260 b) In value (P.OOO**) Meat beef 120,073 96 60,076 60,093 goats 257 0 1 256 sheep 1,235 48 40 1,243 chicken 439 32 5 466 others n.a 51B 2,004 n.a Total 122,004 694 62,126 60,572 Milk +lB,7B5 5,251 12 +24,024 Eggs 3,420 18:> 0 3,600 * 4 = 1+ 2 - 3. ** Uxal production valued at export prices Sources: Calculations based on Agricultural Statistics am. External Trade Statistics. - 87 - Table 7: Numbers of Slaughtered Cattle in the Period 1966-1984 Others Total 1984 239,000 51,600 290,600 1983 234,000 46,450 280,450 1982 237,000 34,600 271,600 1981 202,000 35,800 237,800 1980 226,000 86,000 312,783 1979 229,000 58,000 287,000 1978 149,346 70,000 219,346 1977 196,850 50,000 246,850 1976 211,987 50,000 261,987 1975 188,440 42,000 230,000 1974 186,041 38,000 224,041 1973 209,443 34,000 243,443 1972 156,510 40,000 196,510 1971 167,430 31,000 198,430 1970 128,199 26,000 154,199 1969 93,074 23,000 116,074 1968 103,776 20,000 123,776 1967 95,902 18,000 113,902 1966 148,654 17,000 165.654 Sources: ~ Annual Reports: Agricultural Statistics and Hubbard, 1983. - 88 - Table 8: Numbers of slaughtered Goats in the Period 1966-1984 others Total 1984 8,216 55,200 63,416 1983 4,127 43,400 47,527 1982 384 41,000 41,384 1981 296 51,400 51,696 1980 309 33,200 33,509 1979 416 35,300 35,716 1978 323 n.a n.a. 1977 3,533 II II 1976 5,137 II 1975 16,010 II II 1974 42,756 II 1973 4,476 II II 1972 14,838 II II 1971 25,244 " 1970 26,359 II " 1969 17,573 II " 1968 19,853 II " 1967 3,323 " II 1966 538 II " n.a.= Not Available Salrces: Be Annual Reports: J\gricultural Statistics and Hulilard, 1983. - 89 - Table 9: Numbers of Slaughtered Sheep in the Period 1966-1984 Others Total 1984 7,965 8,500 16,465 1983 3,382 13,000 16,382 1982 264 9,300 9,564 1981 178 14,590 14,678 1980 198 33,200 33,398 1979 331 6,100 6,431 1978 440 n.a n.a 1977 1,910 1976 2,990 II II 1975 5,364 II 1974 14,989 II 1973 4,240 1972 8,096 " " 1971 16,622 " 1970 7,427 II II 1969 17,387 1968 12,543 II " II 1967 3,903 " 1966 2,812 " " n.a.= Not Available Sources: ~ Annual Report, l\gricultural Statistics and HuJ:j)ard, 1983. - 90 - rairy Products Rural milk production is IlOStly non-commercial. It is used as the primary diet for young children (Carl Bro Int, 1982). '!he Integrated Farming Pilot Project fourrl that in Mathethe, Southern District, only 20 percent of the cc::MSwere milked. Milk production fluctuates considerably from season to season (0.5-1.6 liters per day). The Agricultural Technological Improvement Project found goats to be an important source of milk in the Tutlnlleregion providing an average 1.5 Liters per farmer per day with a meandaily production per goat of 284ml in 1986. '!he number of milking farmers fluctuated I1Dnthly between 35 and 85 percent (Gray, 1987). Milk production is affected by drought. Drought conditions lead to lstentirely imported. Distribution Of Animal Based Food Incane and cattle CMtlership are closely linked. Similarly, small stock and cattle CMtlership appear positively related (Litschauer and Kelly, 1981). '!able 10 slUlUllarizes availability of meat and milk availability as well as incane fran holdings in different herdsize categories. Benefits in terms of milk and food availability are positively correlated to herdsize. M;)reover, meat fran hane slaughter increases during a period of drought as the COlparison between 1981 and 1984 illustrates. Milk availability increases aIOC>ng small herds, however, further studies are required before final conclusions can be made. - 91 - Table 10: Meat am. Milk Availability per herdsize by holding (1981-1984) a) Cattle Herdsize Number of Meat fran Milk Availability Net Incane Agriculture bane slaugh (mlt/pers/day Iblding* (p) (XOOO) (GriPer/day) 1981 1984 1981 1984 1981 1984 1984 0 26.5 23 0 0 0 0 0 1 - 10 13.4 17 7 25 89 244 82 11-20 14.0 12 16 56 188 298 161 41 - 100 10.3 10 44 48 433 672 611 100+ 5.3 4.7 86 198 306 148 2,093 Total with Cattle 31 62 200 203 424 57.7 57.7 1366 2149 1695 1265 55,835 Canmercial .3 .3 !?L Goats 0 0 0 0 0 34.6 30.8 0 8 22 22 P-2 1 - 10 28.2 21.1 5 52 50 P-1 11- 20 12.3 13.9 11 14 19 30 P13 21-40 5.6 10.6 21 21 10 p29 40+ 4.0 31 30 26 1.B Total wi 33 P4 Goats 46.9 49.6 9 14 19 41 P320 Camnercia1 .2 .2 132 205 *Sales Minus Purchase Sources: Agricultural Statistics 1981, 1984. - 92 Incane is measured simply as sales minus purchases (without assessl1l8nt of expenditures) and is generally positively correlated to herdsize (also found by the Livestock Hanagement SUrvey). Incane fran livestock can be used to suwlement incane losses incurred fran elsewhere due to, for example, drought (Kgathi and cpschoor, 1981). 'Ibis fact probably explains the relatively high incane fran cattle in small cattle herds. In contrast, investments seem to be channelled to the building up of larger herds of goats. Cbnstraints on Livestock Developnent 'Ibe country and its people face a m.nnber of socio-economic and environmental constraints in developing livestock related resources. Low and erratic rainfall in canbination with soil characteristics make large parts of the country to be unsuited for extensive forms of grazing. Carrying capacity estimates range fran 12 16ha/Lsu in the eastern hardveld to 16 - 20 ha/Lsu in the western sandveld. Flexibility and IIDbility are important adaptation mechanisms which have, horever, lost relevance due to increased land pressure. Despite the large size of the country, the prevailing environmental oonditions and other land use activities set long term limits. Land has becane a scarce resource, particularly in small districts. Expansion of livestock can no longer take place unless at the expense of other activities such as hunting and gathering. OVer utilization of grazing throughout the country, reduces herd performance (see LivestOCk Management SUrvey) and endangers long term perspecti ves of livestock. Under such circumstances, grazing has becane the IIDst serious constraint which is difficult to ease. Droughts reduce carrying capacity and IIDst seriously affect herds of less than 20 head (Carl Bra Int, 1982). As a result, the distribution of cattle becanes IIDre skewed as small herd owners lose or sell their cattle with very little chances of being able to ever rebuilt their herds. In addition, land pressure has increased the tendency taerds privatization of groundwater sources and surroonding grazing to - 93 - the detriment of the small holders woo have to keep herds in more congested mixed farming areas (Peters, 1983~ Arntzen, 1985). Expansion into the less congested western parts of Botswana is only affordable for large cattle omers. Future IDplications It is obviously not possible to indicate where the livestock sector will be in another 20 years. '!here are too many uncertainties and factors which are (partly) beyond goverrunent control (e.g. access to export facilities and the regional political situation). In addition, the direction of government intervention cannot easily be predicted in detail. Hcwever, it afPeCl.rs possible to predict what may ha~n if no drastic changes occur. '!he anticipated contribution of rapid population growth will in principle increase livestock m.unbers (as there are few alternatives). Cbnsequently, envirorunental and socio-econanic constraints will be IIDre felt. land may becane more degraded, more econanic inputs will be needed for livestock production (e.g. fodjer, lalxlur) aM access to production assets such as waterpoints will be vital. As these requirements cannot be met by small livestock holders, the result will be an increased drop out rate of small livestock holders and an increasingly skewed distribution of cattle. Goats could play the role of "fX'Orman s I cattle" IIDre than at present, but they are not able to replace cattle as draught power. In general, rural incane distribution is likely to becane more skewed unless the I drop-outs I fran livestock production will find alternative incane sources. Artnzen, J.W. 1985 "Agricultural Developnent and ram use..in ~tern Carmllnal Botswana: 'lhe case of Kgatleng. Natl0r:w- Institute of Research and Documentation. Worklng paper No. SO, University of Botswana. - 94 - Arntzen J.W. and E.M. Veenerrlaal, 1986 "A Profile of Environment and Development in Botswana". Institute for Environmental Studies, The Free Universi ty, Amsterdam. Department of Agricultural Research 1986 "ATIPProgress Report 1985-1986". Government Printer. Gray, 1987 "Goat Production in the future region. Agricultural Technology Improvement Project" , Ministry of Agriculture. Gulbrandsen Q, 1980 Agro-pastoral Production and Canrnuna1 Land Use: A Socio-Econanic Study of the I3ang1.aketse. University of Bergen/Ministry of Agriculture, Botswana. Gulbrandsen, Q. 1985 "Access to Agricultural Land and Canrnunal Land Management in Eastern Botswana". Applied Research Unit, Ministry of Local Government and Lands. Hubbard, M. 1982 "'lhe 1979 and 1980 Agricultural Statistics". In: Hitchcock, R.K. (ed.) "Botswana First Livestock Development Project and its Future Implications" , National Institute of Research and Ih::umentation. Hubbard, M. 1983 "Botswana and the International Beef Trade 1900-1981" Ph.D. 'n1esis, University of Sussex. Litschauer, J.G. arrl W.F. Kelly, 1981 "'!he Structure of Tradi tiona1 Agriculture in Botswana. Division of Planning and Statistics, Ministry of Agriculture. McD:>nald,I. 1980 "A Handbook of Livestock Statistics", Ministry of AgriCUlture. Ministry of Agriculture 1965 Agricultural Statistics. 1985 - 95 Murray, M. 1978 "wildlife Utilization Investigation and Planning in Western Botswana"• Ministry of Commerce and Industry. Claturen, M.A. 1983 "GrOW"thwith Equity: Some explanatory !iyp:)thesis based on the Structure and pattern of Distribution in Botswana". In M.A Cbmmen, et. al Botswana's Economy Since Independence. Tata Ma.Graw-Hall. Opschoor, J.D. 1981 "Environmental Resources Utilization in Communal Eastern Botswana. National Institute of Research and TbclllTlentation. I'larking paper No. 38. University of Botswana. Sandford S, 1978 "D:aling with Drought aJ1d Livestock in Botswana: A" Overseas Development Institute, U.K.