HOWS_III Chapter 13—6/5/2016—NANDHINI.P—554083—IBTauris

1 2 3 4 5 6 7 13 RainfedAgriculture, andHunger 8 in Tanzania 9 10 11 Terje Oestigaard 12 13 14 INTRODUCTION 15 16 The2011drought in East Africa wasone of theworst in decades,affecting 17 up to 13 million people in Djibouti, Ethiopia,Kenya,Somalia andparts of 18 Tanzania.Manyofthe affected people were pastoralists,but this natural 19 catastrophe also highlighted theroleofthe life-giving rainsfor farmers 20 dependentuponrainfed fortheir subsistence,livelihoodand 21 welfare.Rainfedagriculture is by farthe most dominant ruralactivityand 22 livelihoodinsub-SaharanAfrica.Understandingthe challenges of rainfed 23 agricultureistherefore of utmost importance in acontext of increased 24 climate variability, populationgrowthand higher food insecurities. Thus, 25 this chapterwillfocus on theroleofrainfed agriculture during the 26 droughtinEastAfrica in 2011 in asmall Sukuma community alongthe 27 southern shores of Lake Victoria in Tanzania.Although thedirectphysical 28 andnatural consequenceofdroughtisthe absenceofwater or thelife- 29 givingrains,the humanand social implications arealso consequences of a 30 widerrange of cultural factorswhich either strengthen or reduce 31 theconsequencesoffailingrains. Hence, Iwill focusfirstonthe global 32 andlocal discoursesonwater andfoodproductionasabackground 33 forunderstandingrainfed agricultureand drought;then on the 34 agrarianquestionand overallchallenges of rainfedagriculture in 35 Tanzania,including thecurrent land laws andnationalpolitics of nuclear 36 villages;and finally an ethnographicindepth discussionofthe 2011 37 droughtamong theSukumafarmersinUsagara village from awater 38 systems perspective. 39 40 41 GLOBAL AND LOCAL DISCOURSES ON AND FOOD 42 PRODUCTION 43 44 The Sukuma are the largest ethnic group in Tanzania, and are estimated to 45 number more than 5million people. Among the Sukuma: HOWS_III Chapter 13—6/5/2016—NANDHINI.P—554083—IBTauris

RainfedAgriculture, Droughtand HungerinTanzania333

46 disastrous ,epidemics andepizooticsare notonlyhearsay and 47 phantoms in theminds of thepeoplebut well-known factsoflife. Life is 48 precarious.Threatenedbydestruction throughfamine, sickness anddeath,lifeis 49 always at risk.Tostayalive is an achievement, somethingtoworkfor incessantly 50 throughthe wholearray of technological, organizational andideological means 51 offeredbyculture andsociety:through cultivationand livestock-rearing, through 52 cooperationwithkin andneighboursand throughthe veneration of the 53 ancestors(Brandstro¨ m1990: 168). 54 55 TheSukumaweretraditionally agro-pastoralists andcattleweretheir 56 main possessionand form of storablewealth forprocuring allof 57 life’snecessities. Although cattlestill have importanceinSukuma 58 society andcosmology,the role of farminghas increasedatthe expense of 59 cattle (Brandstro¨m 1985,Wijsen &Tanner2002).InMwanzaregion, 60 smallholderagriculture employsabout 85 percentofthe population. 61 Usagaravillage is locatedabout 25 km southofMwanza, a35minutedrive 62 by public transport (Figures 13.1–13.2).In2010, thepopulationofthe 63 ward was8,839.The number of cows was2,057 andthere were 42 oxen 64 used forploughing. 1 Theannualprecipitation in Mwanzaisabout 1100 65 mm/yr,but as seen fromtwo precipitationdataset,there areattimes large 66 annual variations,and themostimportant factorfor farmersisthe amount 67 of arrivingatwhich time andinwhich rainy season.Traditionally, 68 waterwas regulatedbyintricate waterlawsamong theSukuma(Cory 69 1953,Drangert2003).Intoday’s world, rural farmersdependent upon 70 rainfedagriculture arepart of theglobalworld,which puts emphasis on 71 what to grow:foodcrops or cash crops? 72 Globally, about 12 per cent of the total land area is used for agricultural 73 production and about 42 per cent of the world’s population live on this 74 land (Fraiture &Wichelens 2007). Only 19 per cent of cultivated land on the 75 globe is irrigated, but this land produces 40 per cent of the world’s food 76 (Hanjra &Quereshi 2010: 365). Although agriculture accounts for about 77 70 per cent of the world’s withdrawals of water, in many low-income 78 countries the figure is about 90 per cent (WWAP 2012: 46). An estimated 79 60 per cent of Africa’s population live in rural areas. In sub-Saharan Africa, 80 more than 90 per cent of the farmland is rainfed (WWAP 2012: 177, 334). 81 In sub-Saharan Africa, agriculture accounts for 35 per cent of GDP and 82 employs about 70 per cent of the population.Agriculture is thus key for 83 economic development and reduction, and every 1per cent 84 increase in agricultural production reduces the numbers of the absolute 85 poor by 0.6–1.2 per cent (Wani et al. 2009: 1–2). Due to population growth, 86 however, the amount of cultivated land per person has globally declined 87 from 0.4 ha in 1961 to 0.2 ha in 2005 (WWAP 2012: 46), directly related to 88 hunger and . 89 Waterand wateravailability is intrinsically linked to food security. 90 Food security is defined as thepoint at which ‘all people,atall times, HOWS_III Chapter 13—6/5/2016—NANDHINI.P—554083—IBTauris

334 AHistory of Water

91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 Figure 13.1. Tanzania. Map from the Nile Basin Research Programme, University of 122 Bergen, Norway. 123 124 have physicaland economic access to sufficient,safeand nutritiousfood 125 to meet theirdietary needs andfoodpreferences foranactive and 126 healthylife’(FAO1996),and as ‘a necessary if notsufficient basis 127 forpoverty alleviation’ (Cooketal. 2011:2). Accordingtothe World 128 Food Council: 129 130 food security implies two things. First, ... that food is available, accessible, 131 affordable –when and where needed –insufficient quantity and quality. 132 Second,itimplies an assurance that this state of affairs can reasonably be 133 expected to continue ... thatitcan be sustained. To put it simply, food security 134 exists when adequate food is available to all people on aregular basis (World 135 Food Council 1988: 2). HOWS_III Chapter 13—6/5/2016—NANDHINI.P—554083—IBTauris

RainfedAgriculture, Droughtand HungerinTanzania335

136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 Figure 13.2. Map of Usagara and Mwanza in Tanzania (map modified by Terje Oestigaard 160 from Google Maps). 161 162 163 In today’s world thereisageneralconsensus that even with thecurrent 164 land andwater resourcesthere will be enough food forevenaworld 165 populationof9billion.However,the current number of people living in 166 hunger is increasingand today counts formorethan800 millionpeople. 167 In themorepolicy-oriented debatesitiscurrently emphasized that one 168 needsaholisticand global approach to theemergingwater andfood 169 crises, including thewhole valuechain from thefarmtothe fork. 170 Although theseoverall andglobalapproachesare important, in practice 171 they have little relevancefor rural dwellers dependent upon the 172 unpredictable . Many of theprocessestakingplace duringdroughts 173 andevenmoresoduringextremedroughtsare alreadyareality forrainfed 174 farmersinnormalyears,although to alesserextent. In sub-SaharanAfrica 175 it wasestimated in 2008that48per cent of thepopulationsurvivedonless 176 than US$1.25aday. Andasaparadox, in this region hunger is most 177 widespread in ruralareas amongfarmers andhouseholdsproducing food 178 (H˚a rsmar2010).Insub-Saharan Africa,being asmallholderfarmerrelying 179 on rainfedagriculture meansthatyou live throughvarious degrees of 180 poverty. In yearsofgoodrains,one mayget asmall surplus, butbeing HOWS_III Chapter 13—6/5/2016—NANDHINI.P—554083—IBTauris

336 AHistory of Water

181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 Figure 13.3. Failed harvest in Usagara, early 2011 ( E Terje Oestigaard). 204 205 206 dependentuponrainfed agriculture impliesone thingfor sure:droughts 207 will come oneyearoranother (Figure13.3). 208 Onemay differonoridentify different typesofdroughtsfroman 209 analytical perspective, although allofthemare characterized by 210 insufficientwater or wateratthe wrongtimefor agriculturalpurposes. 211 First, thereisunpredictabledrought, which occurs when total 212 precipitationiscomparabletonormalyears,but theharvest is exposed 213 to growth stress as aresultofunpredictable,erratic andunevenrainfall. 214 Secondly, there is full-season drought, whichoccurswhenoverall 215 precipitationpatternsare much lower than in normal yearsand plants do 216 notreceive enough water. Thirdly, thereisterminaldrought,which 217 occurs when initially thereisenoughwater forcultivation,but laterthe 218 soil is exposed to awater deficit.Fourthly, thereisintermittent drought, 219 whichoccurswhenthere is ashort dryspell duringthe growingseason 220 andthe harvestisexposedtodrought only at onestage duringgrowth 221 (UNECA ACPC 2011:19). 222 Variability in rainfall generates dry spells almost every season and hence 223 shorter periods of water stress during the growing season. Dry spells are 224 manageable and investments in water infrastructure can overcome these 225 fluctuations, which may last from two to four weeks. Meteorological HOWS_III Chapter 13—6/5/2016—NANDHINI.P—554083—IBTauris

RainfedAgriculture, Droughtand HungerinTanzania337

226 droughts, on the other hand, occurring on average once adecade in moist 227 semi-arid regions and up to twice per decade in dry semi-arid regions, result 228 in complete crop failure. When such droughts occur, they cannot be 229 counteracted by agricultural water management, and other social coping 230 strategies are necessary, such as food relief and grain banks (Wani et al. 231 2009: 8), if they exist. When the rain and the harvest fail, farmers have to 232 buy their food for survival on the global market, where prices increase and 233 fluctuate during periods of drought. This directs the question to food crops 234 or cash crops. 235 For individual farmers, whether poor or relatively better off, it makes 236 sense to generate as much income as possible by cultivating crops that can 237 be sold on domestic or international markets and then to buy cereals at the 238 same markets. From apoor farmer’s perspective, however, there are no 239 cheap cereals. In today’s world, everyone needs money and volatile and 240 increasing food prices make the need to generate more income ever more 241 urgent. When the poor are barely able to make ends meet –and in most 242 cases they cannot –there is only amarginal net difference between selling 243 cash crops and buying life’s necessities. This net difference is, however 244 small, vital to farmers in sub-Saharan Africa living on an absolute 245 subsistence minimum. Thus, even in arid regions farmers dependent on 246 fluctuating rainfall patterns may choose to grow water-intensive crops well 247 aware of the fact that if the rains fail, they are left with nothing. If they had 248 chosen crops requiring less water, they could have some harvest even in 249 years of bad rain and consequently some food. Subsistence farming means 250 living on the absolute minimum –inother words, in extreme poverty and 251 often starving. The possible net difference if rains and harvests are 252 successful is acalculated reward and risk that comes at ahigh price. 253 However, very few farmers take the risk of planting only cash crops; food 254 crops are asecurity net (Oestigaard 2014). If what and when to grow is a 255 risk mitigation strategy, where to farm is apolitical question institutiona- 256 lized by law or land tenure regimes, and varying land systems may offer 257 different possibilities as risk reducing or coping mechanisms in times of 258 failing rains. 259 260 261 THE AGRARIAN QUESTION AND OVERALL CHALLENGE OF 262 RAINFED AGRICULTURE 263 264 The classic agrarian question in Africa has been how capitalism can enable 265 agriculturaldevelopment that contributes to industrializationand 266 decreased poverty. Thus, agrarian reforms have been seen from broadly 267 three perspectives –the social, economic and political. The social 268 perspective has awelfare focus and emphasizes possibilities to reallocate 269 land to the poor. The economic perspective stresses the emergence of 270 small commercial farmers who create employment with multiplier effects. HOWS_III Chapter 13—6/5/2016—NANDHINI.P—554083—IBTauris

338 AHistory of Water

271 Finally, the political perspective argues for transforming the whole agrarian 272 structure as adevelopment strategy (Maghimbi et al. 2011: 19). 273 In 1895, the German colonisers declared that all land in Tanzania, 274 occupied or not, would become Crown lands. In 1928, the British governor 275 authorized land grants for periods up to 99 years. After independence, all 276 land continued to belong to the state. The Land and Village Land Acts of 277 1999 state that all land on mainland Tanzania ‘shall continue to be public 278 land and remain vested in the President as trustee for and on behalf of all 279 citizens of Tanzania,’ meaning that the state is the ultimate owner of the 280 land, and makes grants for the occupation and use of it (Cl. 4(1), op. cit. 281 Maghimbi et al. 2011: 28). 282 President Nyerere introduced the concept of Ujamaa in the Arusha 283 Declaration in 1967. This policy was populist and was based on what today 284 is commonly seen as autopian belief that amodern state could emerge 285 from communal villages. The aim of Ujamaa was to increase agricultural 286 productivity and at the same time reduce economic inequality. At the heart 287 of the socialist vision of Ujamaa was aparticular view of the world and its 288 commodities. In this view, there was alimited amount of good and goods in 289 the world, and one person’s gain would be another’s loss. Since Tanzania 290 aimed to be ‘self-reliant’ in order to avoid dependence as anewly 291 independent state, this implied azero-sum national economy. Within 292 Tanzania, wealth accumulation in any form would be at the expense of 293 someone else, who would lose something (Sanders 2008: 115), thus 294 running counter to the socialist vision. 295 Nyerereplacedthe nation at thecentreofthe frameworkfor 296 modernization and as aresult repressed cultural, ethnic, social and 297 religious diversity. By abolishing chiefdoms in 1963, the rural population’s 298 relationship to the state was changed. The chiefs had largely lost popular 299 support in the colonial era, but with this radical break, in combination with 300 villagization, the state-peasant relationship was fundamentally transformed 301 (Havnevik 2010). 302 From 1967 to 1973, the number of people in Tanzania living in Ujamaa 303 villagesincreased from abouthalfamillion to about2million, 304 approximately 15 per cent of the rural population. In 1973 the government 305 abandoned the policy of voluntarism and the president ordered that the 306 whole population should be moved into villages. This took place through 307 organized ‘operations’, basically amilitary metaphor for mobilization. 308 ‘Operation Villagization’ forcibly removed about 7million people into 309 villages, or about 50 per cent of the rural population. Through this process, 310 some 8,230 new villages were created (Mesaki 1993). However, as Bryceson 311 remarks, villagization ‘was not about socialist collectivity as much as 312 nuclearizing scattered household settlements into villages where health 313 dispensaries, schools and agricultural marketing services and productive 314 infrastructure could be more efficiently provided’ (Bryceson 2010: 75). 315 Nevertheless, villagization also led to environmental degradation. This may HOWS_III Chapter 13—6/5/2016—NANDHINI.P—554083—IBTauris

RainfedAgriculture, Droughtand HungerinTanzania339

316 have happened in any case, but the process of villagization accelerated it 317 (Kikula 1997). 318 In practice, there was little link between the Arusha Declaration and the 319 new Ujamaa policy of 1967 and the actual implementation of the policy, 320 apart from the nucleated village unit. The farmers were still dependent 321 upon fluctuating market prices for their crops rather than government 322 stimuli, but more important; they still dependent upon the fluctuating 323 rains. Although farmers were promised numerous benefits, the new policy 324 was viewed with considerable scepticism (Havnevik 1993). As Kjekshus 325 wrote in 1977: 326 327 The implementational features havefundamentally undercut someofthe 328 preconditions of the program and harmed the outcome. Already the socialist 329 features of the villageplan havebeencompromised, and it is doubtful whether 330 the move has been aschool in grass-roots democracy that has strengthened the 331 peasantry’s senseofself-reliance and dignity. More fundamentally, it is doubtful 332 whether the change of settlementpatternisconsistentwith the fundamental 333 requirements for economic development in the Tanzanian ecology (Kjekshus 334 1977: 281). 335 336 TheSukumausedtolive in widely dispersedhouseholdsonthe semi-arid 337 cultivation steppe, andthere were no villages in Usukumauntil theforced 338 villagization of 1974.Thispolitical decision to forcethe Sukuma into 339 villagesand to work on communal farmshad severe repercussions.Asa 340 consequence, lifefor theSukumabecameharderinthe postcolonial 341 eraand theenvironmentand land wasdesiccated (Wijsen &Tanner2000: 342 21–2). 343 Thevillagizationprocess disregardedexistingcustomary rights and 344 land tenure systems,thereby completely underminingthe security of 345 customarylandholders.Italso createdother problems that contributed to 346 agricultural decline.Nuclear settlementsand villages createdanartificial 347 land shortage in areaswhere landwas in abundance. With farmers living 348 together,all land in thevicinity wasfarmed,but beyond acertain pointit 349 became uneconomical forafarmertowalkand establish newfarms.Asa 350 consequence, in many villages farmersfaced land shortageswhile 351 unoccupied farmland wasavailable only some tenmiles away (Maghimbi 352 et al.2011: 27–33). 353 Since 1989–90 people have been allowed to move back to their original 354 lands, but most people have stayed in the villages. Nevertheless, although 355 the intentions behind the vision were good, the programme failed to 356 transform Tanzania into amodern self-sustainingstate (Sanders 2008: 113). 357 In Tanzania today, about 70 per cent of farming land is cultivated by hand 358 hoe, 20 per cent by ox plough and only 10 per cent by tractor. Small-scale 359 farms still predominate and are on average as small as 0.9 ha, with only a 360 few farms of 3haormore (Maghimbi et al. 2011: 43–4). Thus, villages do HOWS_III Chapter 13—6/5/2016—NANDHINI.P—554083—IBTauris

340 AHistory of Water

361 not necessarily solve the problems, in fact they may create more 362 problems when not access to land but water was and still is the most 363 essentialfor rainfed agriculture.And did notfollow the 364 villagization process. 365 Water is an increasingly scarce resource, and not only at household level. 366 Importantly and fundamentally, erratic rainfall patterns represent huge 367 uncertainty and risk for rainfed agriculture. Water is the basis of all 368 agricultural practices. The absence and presence of different types of water 369 sources structure all societies, whether those sources are rain, rivers or 370 lakes or acombination of water bodies at acertain place. Too much water at 371 the wrong time of year, such as unpredictable and devastating floods, is as 372 bad as too little water when it is really needed, the result being drought. 373 Everything depends on the arrival of the right rains in the right amount at 374 the right time (Figure 13.4), and their failure and the ensuing dire 375 consequences underscore the fragility and vulnerability of all aspects of life 376 (Oestigaard 2009). This total dependency on an unpredictable resource 377 represents ahuge uncertainty. Farmers aim to reduce and control this 378 uncertainty in the best possible ways, but their means are in practice 379 limited. Farmers rely on tradition, in this case the culturally accumulated 380 knowledge derived from generations of experience involving the soil, 381 crops and crop performance under varying and erratic rainfall patterns. 382 Now, this centuries old knowledge is adapted to the modern world in 383 which life and well-being largely hinge on growing the most valuable crops 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 Figure 13.4. The arrival of the rains in Usagara ( E Terje Oestigaard). HOWS_III Chapter 13—6/5/2016—NANDHINI.P—554083—IBTauris

RainfedAgriculture, Droughtand HungerinTanzania341

406 that produce an income and secure alivelihood beyond subsistence 407 (Oestigaard 2012). 408 409 410 RAINFED AGRICULTURE AS AWATER SYSTEM 411 412 Understanding the water-world and how different types of water create 413 possibilities and limitations for livelihoods and agriculture is fundamental 414 to understandingcurrent agricultural practice and future development, and 415 can be seen from awater systems perspective (see Tvedt and Oestigaard 416 this volume,and forinstanceTvedt 2010a, 2010b):(1) Thephysical water- 417 scapeatagiventime, includingrainand riversetc., (2)human modifications 418 andadaptations to theactualwaterworlds;and (3)culturalconceptsand 419 ideasofwater andwater systems, includinglaws. 420 With rainfed agriculture as aparticular water system, one may adapt 421 and develop this approach (Figure 13.5). There is, of course, no direct and 422 deterministic relationship between people’s ideas about their world and 423 how they cultivate. For instance one may grow millet using ahoe 424 irrespective of whether one believes in Christianity, Islam or the ancestors. 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 Figure 13.5. Rainfed agriculture in awater-systems perspective ( E Terje Oestigaard). HOWS_III Chapter 13—6/5/2016—NANDHINI.P—554083—IBTauris

342 AHistory of Water

451 Rain (layer 1) is, however, aconstant and limiting factor that structures 452 agricultural practice. Thus, by analytically using agriculture –from Latin 453 agricultu¯ ra,from ager field, land + cultu¯ ra culture –one may distinguish 454 the physical work and actual cultivation as belonging to layer 2and the 455 cultural elements as part of level 3. These levels overlap and are 456 mutually dependent. 457 458 459 ERRATIC RAIN (LEVEL 1) 460 461 In Mwanza, as in many other places in Africa, there are two main rainy 462 seasons: one in February–April and the other in October–November. The 463 first is the ‘long’ rainy season and the latter the ‘short’ season. However, 464 ‘rain is scarce in this semi-arid region. When it does fall, it is often erratic 465 and unevenly distributed. Too much or too little at the wrong time can and 466 often does spell disaster for these agricultural people’ (Sanders 2000: 473). 467 Moreover, even when the rains come, they may fall unevenly: one village 468 may receive sufficient rain whereas the clouds may pass over another 469 village without rain falling. Even within avillage, the actual precipitation 470 may vary depending upon particular clouds and rain at agiven time in the 471 rainy season. 472 Although life-givingrainisofutmostimportance, thetwo rainyseasons 473 arenot of equalweight, anditisduringthe long rainyseason that most of 474 thefoodsurplus is secured. Thelongrainy season is thus fundamentalto 475 food security in agiven year.Whenthe shortrainy season comes, 476 additionalfoodisproduced. Allofthiswillbeeaten by thetimethe 477 farmersexpectthe nextlongrainy season.Ifthisfails,the resultsmay be 478 catastrophiclikein2011 when thedroughtinEastAfrica affected up to 479 13 millionpeople, andalso hitparts of Tanzania,the Mwanza region 480 included. In Tanzania, it wasestimated that 500,000wereaffectedbythis 481 drought(WorldVision2011). 482 In Usagara, the long rainy season from February to April was disastrous. 483 The rain almost completely failed and there were only small intermittent 484 showers. The short rainy season in October–November 2010 had also been 485 bad, but not as bad as the ensuing long season. During the autumn of 2010, 486 there had been enough rain to produce some food and small amounts of 487 cotton. Consequently, there was barely enough food to tide people over to 488 the next long rainy season. In 2011, the maize planted in late January died, 489 and farmers hoped for the rain so they could start cultivating rice in 490 particular. The rain, however, did not come and the fields were left virtually 491 barren. Without rain, there was nothing the farmers could do. As one 492 farmer put it, one cannot move the land and farmers have to live where 493 their land is. 494 Still, the farmers work and have to be prepared if the rain comes. 495 The preferred agrarian cycle is to prepare the land and plant in January and HOWS_III Chapter 13—6/5/2016—NANDHINI.P—554083—IBTauris

RainfedAgriculture, Droughtand HungerinTanzania343

496 wait for the rain to come and harvest in March–April and then plant again 497 and harvest in June. In ayear of good rain, generous rain should fall at least 498 three days aweek from March to April. However, in 2011 the rains failed 499 more or less completely,with only some small showers from mid-March 500 onward. When the rain fails, not only have farmers put in alot of effort in 501 vain, but the outcome is afailed harvest. 502 Dependence upon fluctuatingrains is agamble. Thevulnerabilityof 503 rainfedagriculture wasoncedescribedbyanIndianfinanceministerin 504 hisnationalbudgetasa‘gamble on therains.’ He went on to stressthat 505 ‘variations in rainfall,ordisruptions in watersupply, canmakethe 506 differencebetween adequatenutrition andhunger, health andsickness – 507 andultimately –life anddeath’ (Human DevelopmentReport 2006:174). 508 Moreover,‘rainfedfarmingisnot just risky; it involveslongdelays 509 betweeninvestmentand fluctuatingreturns.Sooner or later, steadily 510 accruing loan interest outstrips aborrower’s fluctuatingcapacity to pay’ 511 (Shipton2010: 226).In2011, thefarmers facedthe ‘hungryseason’ with 512 barrenfields andhardlyany food.Whenthe rainscameinUsagaravillage 513 in late August in the shortrainy season,theyweremet with greatrelief. 514 Everybody hadsurvivedthe difficult ‘hungryseason’, butasone old 515 womanremarked, this droughthad been oneofthe threeworst in her 516 wholelife. Famine is serious food deficit, andinthe globalized world 517 whereone needs money,farmersface adifficult choice:shouldone grow 518 food or cash crops? 519 Coping with the uncertainties of erratic rainfall in atime of climate 520 change, when rainfall patterns fluctuate even more markedly, is an 521 immense challenge. Farmers living on the absolute subsistence minimum 522 are dependent on rain and on making the right decisions. How is it possible 523 to be strategic when everything depends on the uncertain rains? Millet is a 524 low water-intensive crop and in times of hardship may secure livelihoods. 525 Rice and cotton, on the other hand, demand much water for cultivation, 526 but if rain comes in abundance, they are the choices that bring the most 527 money and prosperity. Farmers cannot afford to make wrong decisions, yet 528 they do, because everything depends on the rain, which is impossible 529 to predict. 530 The main solution is diversifying risks. Specialization may increase 531 vulnerability, whereas diversification may reduce uncertainties. Some plots 532 of land are used for maize, cassava, beans and vegetables and others for rice 533 and cotton. Herein, however, there is also aparadox. If all land is used to 534 grow millet, which is less water-intensive but generates little cash, there 535 might be enough food if there was little rain. However, if the rains are good, 536 farmers will lose opportunities because of their wrong choice of crop. Thus, 537 when the rain fails, much arable land will be left uncultivated, because 538 farmers will not want to put prepared rice fields under millet in case the 539 rain does come. Hence, food deficits are also culturally made. Ultimately, 540 this situation highlights the impossible choices farmers have to make HOWS_III Chapter 13—6/5/2016—NANDHINI.P—554083—IBTauris

344 AHistory of Water

541 when their lives and subsistence are totally dependent on unpredictable 542 annual rains. 543 When the rains fail completely, all these strategies are in vain. The result 544 is hunger and famine, which have huge impacts on society. People have to 545 get money to buy food. Today, the global market on which they sell their 546 agricultural products during times of good rain is also where they have to 547 buy their necessities at fluctuating and increasing prices when there is 548 drought. Money matters and changes lives. 549 Afterthe long rainy season failedduringthe spring of 2011 in Usagara, 550 duringthe following summer thefarmersand theirfamiliesexperienced 551 hunger andhad averyhardand strenuoustime. Thosewith theleast 552 food gotsomesuppliesofmaize from thegovernment, butotherwise 553 socialmechanismsinthe villagehelpedincreasefood security. Within 554 families andamong neighbours,peoplewith asmall surplussharedfood 555 with others.Those betteroff forvariousreasons also employed the 556 poorest to do work in returnfor smallwages, whichenabled them to 557 procure food.Moreover, Usagara village enjoyed acomparative 558 advantageinbeing locatedclose to Mwanza andother smallercities. 559 People fromUsagara couldthusmigrate as daylabourers to towns or 560 searchfor agricultural work in otherparts of thecountry.Thus, although 561 thedrought brought hardship andsufferingtothe Sukuma of Usagara,all 562 of them survived. 563 This drought was hard, but the Sukuma have throughout history been 564 used to droughts and have adapted to this way of living. Despite the 565 difficulties, globalization and the market economy have brought some 566 advantages in times of crises. The problem with drought, it was explained 567 by one, was not really that harvests failed. The problem is the lack of money 568 that enables farmers to buy food from elsewhere. By being day labourers, it 569 was possible, even in the most recent crisis, to generate some income 570 and thus buy food and survive. This was quite different from the crisis in the 571 early 1980s when it was more difficult to buy food even if people 572 had money. 573 The absence of rain influences all parts of society and life. During the 574 drought the wells run dry. Villagers had to dig new wells at greater depths 575 and further away. This implied that the women had to walk longer and 576 spend more time collecting and carrying water. Girls are the first to be 577 taken out of schools to carry water when there is adrought and more time 578 is needed to secure water for the household’s daily consumption. There 579 was less drinking water for both humans and animals, and several farmers 580 had to sell their cattle (Figure 13.6). 581 Although theshort rainyseasonduringautumn2011was good,rain 582 canbeadouble-edged sword. If it rainscontinuously andintensively fora 583 week,the cropswill be swampedand die, sincetheyalso need sun. 584 Rice cansurviveintense rainsfor longer periods,but maizeand 585 otherstaples are more sensitive.Goodrains are thosethatideally last HOWS_III Chapter 13—6/5/2016—NANDHINI.P—554083—IBTauris

RainfedAgriculture, Droughtand HungerinTanzania345

586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 Figure 13.6. Cattle as commodity among the Sukuma ( E Terje Oestigaard). 609 610 611 two–threedays, followed by some three–four days of sunshine,and so 612 on.Thus, as with rain so with sun: toomuchortoo little at thewrong 613 time is equally bad. 614 The short rainy season during the autumn of 2011 started early. The first 615 rains came at the end of August and continued. Although the heavy rains 616 necessary for rice cultivation did not come, the light rains provided enough 617 water for maize, millet, finger millet, beans and other crops. Thus, after the 618 catastrophic season in spring 2011 and the poor harvest of autumn 2010, 619 the rains in autumn 2011 provided farmers with food for survival and even a 620 small surplus for sale. 621 When the rain started, the farmers opted for different strategies. With the 622 arrival of the first rains in late August, some farmers suspected they would 623 not last. They therefore postponed maize planting, since they did not want 624 to lose the planted seeds or waste time and money. Waiting to plant is away 625 to minimize risks, and by the end of October there were still farmers who 626 had not started planting (the main period of rain during the short rainy 627 season is October and November). Other farmers, however, calculated 628 differently and took achance in the belief that the early rains of late August 629 would continue. Some started cultivating maize at the beginning of 630 September, and because the rain kept coming they had the chance to have HOWS_III Chapter 13—6/5/2016—NANDHINI.P—554083—IBTauris

346 AHistory of Water

631 two harvests in this rainy season. Thus, given equal dependency on the 632 seasonal rains, there are different ways to reduce risk and manage the 633 uncertainties of rainfed agriculture. But in the end, these risk-managing 634 strategies are agamble when everything depends upon unpredictable and 635 erratic rains. 636 637 638 AGRI-FIELD, LAND (LEVEL 2) 639 640 By theend of March2011, when therainhad notarrived andthe 641 prospects forthe forthcoming season looked miserableand people 642 prepared fordifficulttimes,farmers startedtoplant millet in various 643 fields. Thelatest onecan plantmilletwith ahopeofasuccessfulharvestis 644 in April. However, they didnot cultivate milletinthe fieldstheyhad 645 preparedfor ricecultivation, sincetheybelievedthe rain would 646 eventually come.Still,evenifmilletisdrought resistant, it nevertheless 647 requiresvaryingamounts of rain foramonth. Unfortunately therain 648 failedinApril as well, leaving farmerswithalmostnoharvest or food until 649 theonset of thenextseason’s rains. 650 In the past, combinations of cash and food crops were grown: cotton 651 together with chickpeas, millet, cassava and sweet potatoes, while in more 652 swampy areas, which have now dried up, rice was cultivated. Millet was the 653 dominant staple until the 1980s, but since then people have introduced 654 crops they prefer to eat, and today millet is perceived as disgusting because 655 of its murky brownish colour. From the 1980s and 1990s onwards, rice 656 cultivation became more dominant. 657 Rice is preferred as astaple for several reasons. First, it is the dietary 658 preference today. Secondly, if asurplus is produced or cash is needed, it 659 can generate agood income, with prices per kilo ranging between Tsh. 660 1,600 and 1,800. Thirdly, rice can be stored for years after harvest and thus 661 represents both food security as well as potential cash income. Maize, for 662 instance, can only be stored some five or six months after harvest and must 663 therefore be consumed within arelatively short time. Finally, even when 664 there is insufficient rain for planting, as in spring 2011, the rice seeds may 665 survive up to three years. Thus, the resistance of rice to spoilage is also a 666 security mechanism. 667 Cotton is highly drought resistant, whereas maize and sorghum are more 668 vulnerable to water shortage and drought. As such, cotton is preferable 669 during dry periods. However, cultivation of cotton is water intensive and 670 requires asuccessful rainy season, whereas millet needs little water. Where 671 land is alimiting factor, farmers must calculate the proportion of land to be 672 used for subsistence agriculture in relation to cash-crops. When they do not 673 have enough land for cultivation, food-crops for personal consumption are 674 more often given priority over crops that may generate limited cash in an 675 uncertain market. Among the villagers it was generally agreed that the best HOWS_III Chapter 13—6/5/2016—NANDHINI.P—554083—IBTauris

RainfedAgriculture, Droughtand HungerinTanzania347

676 way to improve agriculture is to introduce small-scale irrigation or rain- 677 harvesting techniques, which would enable more secure food-crops, with 678 some additional cash-crops. 679 Moreover,changingcrops is notunproblematic. When smallholders 680 change cropsfromyeartoyear, the soil maylosesomeofits nutritional 681 qualities. Amajor problem with maizeisthatthe cropisattacked by striga 682 –also knownas‘witches’ weed’–ahighlydestructive parasiticplant. 683 Milletisalso highlyvulnerabletostriga, and thegovernmenthas 684 instructed farmerswho grow milletintheir fieldsone year,togrow 685 cotton or cassavathe nextinorder to eradicate the parasiticplant. 686 However, if thenextseasonisalsodry, farmersfaceanadded problem, 687 since theycannot growmillet again. Strigaparticularly affects maize, 688 sorghumand millet, butnot rice,which is locatedinwetterareas. 689 Nevertheless, insteadofusingpesticides,ifafield is infected by striga, 690 farmersmay cultivate cotton or cassavathe next year,since strigadoes 691 notsurvive alongsidethese crops. In short, crop rotation patternsalso 692 influence thecrops chosen in agiven year. 693 The Sukuma were the ethnic group cultivating most cotton and cash- 694 crops in Tanzania. Mwanza region has been at the heart of cash-crop 695 farming in Tanzania. In Usagara before 2011, the majority of farmers had 696 shifted from cash-crops, and in particular cotton, to food-crops. Half a 697 decade ago, about 50 per cent of the farmers were growing cotton, whereas 698 in 2010 only 13 per cent had cotton as the major crop. The main reason for 699 this decline in Usagara was the vulnerability caused by the free market 700 system. If the cotton price is high, farmers switch from food to cotton 701 production. When the prices of fertilizer and pesticides increase, the 702 returns from selling cotton decrease: the growing costs may be higher than 703 the net returns and farmers would not have enough income to pay for these 704 inputs. The price for cotton was so low that food crops become the new 705 cash crops, in particular rice. Moreover, individual farmers cultivating 706 cotton on small farmlands cannot compete economically with large 707 industrial plantations. Thus, increased expenses and decreased cotton 708 prices caused farmers to return to food crops to survive. 709 However, during the short rainy season in the autumn of 2011, farmers 710 started reverting to cash crops and cotton production, and many more 711 were expected to produce cotton again during the main rainy season of 712 2012. The government offers preferential loans with good conditions to 713 farmers who want to invest in cotton production, and farmers pay back 714 their loans after the harvest and sale of the cotton. Overall, then, there has 715 been achange from subsistence agriculture to cash crops and back again, 716 and then possibly areturn to more cash crops. 717 About75per cent of cotton is exported, mainly to China(approx. 718 50 percent),but also to India, Bangladeshand Taiwan.Pricesfor cotton 719 used to be Tsh. 500–600/kg, butroserapidly to Tsh. 1,200in2011. There 720 were twomainreasons forthisdoublingofprices: the appreciation of the HOWS_III Chapter 13—6/5/2016—NANDHINI.P—554083—IBTauris

348 AHistory of Water

721 US dollarand lowglobalproduction.The farmersinUsagara were well 722 awareofthesechanges,although notnecessarily of thereasons forthem. 723 Some farmerssuggestedthe Tanzanian governmenthad increased prices 724 in ordertohelpthem. Themisunderstanding of whypricesincreasemay 725 lead to wrongchoices aboutwhich cropstocultivate. In Usagara,cotton 726 waslargely abandoned because of thefalling prices. With therapid 727 increaseinthose prices,manyfarmerswererethinkingtheir options and 728 considered growingcottonagain.However,because theprices of cotton 729 aregovernedbythe internationalmarket andnot theTanzanian 730 government,farmers mayagain fall prey to fluctuating pricesifglobal 731 productionincreases. 732 Nevertheless, although farmers were mainly relying on subsistence 733 farming, as much as half the crop might be sold to generate income. Thus, if 734 farmers were able to keep all the food they grew, there would perhaps be 735 sufficient for the whole year. Normally, however, they do not hold back 736 food for the next year, because they have to sell some of it to earn money. 737 This highlights the impossible situation the farmers face: all decisions are a 738 gamble on the rain to get the highest yield possible, failure to achieve which 739 may have devastating consequences. But even with asuccessful harvest, 740 farmers have to sell large parts of their food supply to generate income and 741 cash. It is extremely expensive to be poor. 742 743 744 CULTURE AND RELIGION (LEVEL 3) 745 746 The previous descriptions of agricultural practices and choices are all, of 747 course, within the realm of culture. Cultural choices and decisions, 748 together with religious conceptions and understandings,shape all 749 practices, but as shown, the cultural world is also dependent, restricted 750 and enabled by the external, real world. And for subsistence farmers living 751 off rainfed agriculture, rain is the first and most important parameter – 752 without rain, no food. And as seen, the consequences of droughts are partly 753 man-made given the Ujamaa policies and the villagization processes forcing 754 people to live in villages. This created an artificial land-shortage and 755 increased the pressure on land. 756 Importantly as well, food is not just food. There is hardly any other item 757 endowed with so much cultural and religious significance, including 758 taboos, as food. Suffice it to point out here that the cultural values of food 759 often outweigh other rational considerations with respect to water 760 consumption, agricultural production and caloric output. Thus, one key 761 advantage of millet as afood crop is that it requires less water for cultivation 762 than other crops and as such it is asurer food source during times of failing 763 rain and drought. Generally, although millet has been the traditional staple, 764 people do not like to eat it anymore and will not grow it, even though it 765 requires less water for cultivation. This change in dietary preference has HOWS_III Chapter 13—6/5/2016—NANDHINI.P—554083—IBTauris

RainfedAgriculture, Droughtand HungerinTanzania349

766 direct implications for what is grown. Thus, cultural taste and preferences 767 have changed actual farming, and even during famines farmers rarely grow 768 millet. And this is the case even though farmers were encouraged by the 769 government to grow millet. Rather than planting millet on the fields 770 prepared for rice cultivation, the fields were left uncultivated. Most farmers 771 instead gambled that the rain would come and enable them to grow rice, 772 which is preferred both as acash crop for income and because it is 773 perceived as tastier. 774 Regarding cotton, other social mechanisms are at work. Apart from the 775 low selling price, there are further reasons cotton is aless attractive cash 776 crop than others. In Usagara, cotton is sold to alocal cooperative, but 777 payments are frequently delayed. Often, farmers only get paid the first 778 week and then the cooperative claims it needs to borrow money from 779 banks. Subsequent payments may be delayed up to amonth –atime when 780 most farmers really need the money and cannot afford to wait. Other cash 781 crops, such as tomatoes grown in gardens, are paid for immediately and 782 rice in particular is seen as most lucrative as acash crop. Milk is also sold, 783 and keeping livestock, which are also asource of meat, is an additional 784 source of income. However, even though cattle are avaluable resource 785 generating extra income and providing security, particularly when the 786 harvests fail, grazing land is limited and farmers cannot exploit the potential 787 of livestock as much as they would wish. 788 Duringdroughtsand famines, government supportand foreignaid are 789 part of thestrategy andplanningfor survival. Although theexternal help 790 is limited andsurvivalonitishard,suchhelpisneverthelesspartofthe 791 coping strategyofthe subsistenceeconomy.There is aSukumasaying 792 ‘Wewill pass’, andthe Sukuma have always suffered from severe famines, 793 whichoccur everysix or sevenyears. Duringfamines,eachperson 794 receives onebig tin-canofmaize from thegovernment. This is not 795 enough to survive on,but it givesthe farmerssomereliefastheysearch 796 forother jobpossibilities. 797 When drought and food shortage crises occur, the Sukuma usually 798 turned to the ancestors for help. However, although the ancestors were 799 propitiated in other Sukuma areas, this was not the case in Usagara. After 800 they became Christians, propitiating the ancestors has been seen as akin to 801 using witchcraft, and the church condemns both practices. Rainmaking 802 rituals were also seen as sinful by the church and thus villagers did not 803 conduct them. 804 The direct consequences of failing rain were that people suffered and 805 starved and that poverty increased. This was interpreted in different ways. 806 First, global climate change was seen by many as the cause of the scarcity or 807 absence of rain, in particular among the young and those with amodern 808 education. Indeed, through its access to information from television, radio 809 and newspapers in varying degrees, rural Tanzania is part of the global 810 world and the discourse about climate change is central in these rural areas HOWS_III Chapter 13—6/5/2016—NANDHINI.P—554083—IBTauris

350 AHistory of Water

811 too. Secondly, farmers have long been cutting trees in the forests, and 812 consequently the land has become eroded and lost substantial water- 813 retention properties. Deforestation is thus seen as an important human 814 factor affecting the actual water environment. Thirdly, changes in weather 815 and the absence of rain were also seen as stemming from declining 816 traditions and the broken relationship with ancestors, so that it is no longer 817 possible to communicate with and propitiate them, or make sacrifices 818 (Figure 13.7). 819 With regard to the last theme, neglect of tradition and the non- 820 propitiation of ancestors have given rise to numerous interpretations. 821 Some traditional healers and diviners blame the Christians, in particular the 822 Pentecostals, whom they perceive as evil because they have created more, 823 and new forms of, witchcraft. Others saw the drought as apunishment by 824 God because Sukuma values were no longer followed. In general, there was 825 an element, or at least the perception, of generational conflict, since the 826 youth were not honouring traditional values and respecting elders. 827 Women, too, were criticized for dressing inappropriately and not wearing 828 their traditional dress. The lack of rain in 2011 was thus also understood as 829 acollective punishment by God for the moral decay of and misconduct and 830 sin in society. As one informant explained, when people do not follow 831 Sukuma tradition and values, God penalizes them with the absence of rain. 832 Moreover, there has been an increase in witchcraft, and this is also believed 833 to influence rainfall patterns in various ways. People use witchcraft to 834 become rich, and one informant linked the drought directly to the killing of 835 an albino in 2009. Nobody was arrested for the crime. According to this 836 farmer, this misdeed led God to punish the society collectively by 837 withholding rain. 838 In other villages with many traditional healers, however, rainmaking 839 rituals were conducted, and if and when the rain comes, the rainmakers 840 were acknowledged and greatly respected. When rainmaking rituals 841 succeeded and the rain came, the life-giving were also seen as holy 842 and adivine gift from God to all people. Although the practice of 843 rainmaking had disappeared in Usagara, Christians did pray to God for rain, 844 including the Pentecostals. People had long been praying for rain in the 845 church, but it never came. Among Muslims, too, from areligious 846 perspective rain is seen as agift from Allah, but erratic and fluctuating rain 847 was seen as aconsequence of climate change and deforestation, which 848 reduce the overall amount of water. 849 Although rainmaking as aritualhas disappeared, thewhole 850 rainmaking cosmologyhas not. Therewerestill people in the village 851 believedtohavepower to withhold therain. Thereasontheymight wish 852 to harm other people andsociety wasunclear,but somehowthey wished 853 to be respectedand seen as important. In caseswhere rain is believed to 854 be withheld by anti-rain witches,peoplemay consulttraditional healers 855 whodevisemedicines to be sprayedonthe fieldswhere thefarmers HOWS_III Chapter 13—6/5/2016—NANDHINI.P—554083—IBTauris

856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 Figure 13.7. Rock-art in arock shelter in Bukumbi village (see Figure 13.2) where 896 rainmaking and ancestral rituals took place in the past ( E Terje Oestigaard). 897 898 899 900 HOWS_III Chapter 13—6/5/2016—NANDHINI.P—554083—IBTauris

352 AHistory of Water

901 wouldliketoreceive rain.Thus, theabsence or presence of rain wasseen 902 as an ongoingcontest betweenthose with thepower to make it and 903 thosewho couldrestrainit. If therainmakerswerethe more powerful, 904 they wouldpunish therain-witches,including by sending sandstormsto 905 kill them. In 2010,inthe neighbouring village of Nyahorongo, aman 906 predictedbeforethe shortrainy season that therewould be no rain that 907 year. When therainhad still notarrived in November,the villagers grew 908 very angryand beat theman to death. Withholdingthe rain wasnot seen 909 as relatedtothe worldofancestors as such,but as aresultofwitchcraft 910 andmalignant forces. 911 Thus, actual cultivation is not merely about planting, sowing and 912 harvesting, but represents an intricate mixture of rainfall parameters and 913 natural variables, including soil composition and qualities; customary 914 and technological practices; as well as awide and influential spectrum of 915 cultural, ideological and religious concepts. 916 917 918 CONCLUSION 919 920 For smallholder agriculturalists in sub-Saharan Africa dependent upon the 921 arrivals of the life-giving rains, climate variability will sooner or later result in 922 droughts of different lengths and intensity. The risk-coping strategies will 923 inevitably vary; from where and when to start preparing the fields to 924 calculating the possible incomes of food or cash crops and how to balance 925 the different crops requiring varying amounts of water for asuccessful 926 harvest. But all this is of course agamble since nobody can know when and 927 in which amount the rains come. The right decisions have to be made, 928 simply because the farmers cannot afford to make the wrong choices. Yet, 929 sometimes the wrong decisions are made, and it highlights the impossible 930 situation many farmers face and live by when their wealth and health –or 931 suffering and possibly starvation, and in the worst case, death –isultimately 932 amatter of the arrival of the unpredictable rains. Political and institutional 933 factors may reduce and increase the consequences of the farmers’ choices 934 given the possibilities they have. Diversifying risks is the optimal and the 935 preferred choice, but when the rain fails completely all these strategies are 936 in vain. And this is perhaps where the most depressing part of rural poverty 937 and hunger lies; most of these farmers are strong, healthy and eager to 938 work –and they do work hard. They prepare the fields and hope for a 939 bountiful harvest, and when that is done they can only wait and look at the 940 sky to see whether the clouds will come. During the drought in 2011, it was 941 mainly the sun shining, and if there came some few clouds, most often 942 they drifted away passing the village. When the rains failed that season, they 943 knew very well how little food they had left while facing the extremely 944 tough and long ‘hungry season’, only hoping that the rains the next rainy 945 season would be better. HOWS_III Chapter 13—6/5/2016—NANDHINI.P—554083—IBTauris

RainfedAgriculture, Droughtand HungerinTanzania353

946 NOTES 947 948 1Statisticsfrom the ward office. 949 950 951 REFERENCES 952 953 Brandstro¨m,P.1985. ‘The Agro-pastoral dilemma: Underutilizationorover- 954 exploitation of land among the SukumaofTanzania’, WorkingPaper in 955 African Studies 8 .Uppsala: University of Uppsala. 956 ——— 1990. ‘Seeds and soil: the quest for life and the domestication of fertility in 957 Sukuma-Nyamwezi thought and social reality’. In Jacobson-Widding, A., van Beek, W(eds), The CreativeCommunion: African Folk Models of fertility 958 and the RegenerationofLife.Uppsala: Almquist and Wiksell, pp. 167–86. 959 Bryceson, D.F. 2010. ‘AgrarianFundamentalism or Foresight? Revisiting Nyerere’s 960 Visionfor Rural Tanzania’. In Havnevik, K., Isinika, A.C (eds), Tanzania in 961 Transition: From Nyerere to Mkapa.Dar es Salaam:Mkuki na Nyota, 962 pp. 71–98. 963 Cook et al. 2011. ‘Water, food and poverty: global- and basin-scale analysis’, Water 964 International 26(1): 1–16. 965 Cory, H. 1953. Sukuma Law and Custom.Oxford: OxfordUniversity Press. 966 Drangert, J.O. 1993. WhoCares About Water? Household Water Development in 967 Sukumaland, Tanzania.Linko¨ping: Linko¨ping Studies in Arts and Science85. 968 FAO 1996. Report of the World Food Summit.Rome: FAO. Fraiture, C., Wichelens,D.2007. ‘Looking ahead to 2050: scenarios of alternative 969 investment approaches’. IMolden, D(ed.), Waterfor Food, Waterfor Life: 970 AComprehensive Assessment of Water Management in Agriculture.London: 971 International WaterManagement Institute, Colombo andEarthscan, 972 pp. 91–145. 973 Hanjra, M.A., Quereshi, M.E. 2010. ‘Global water crisis and future food security in 974 an era of climate change’, Food Policy 35: 365–77. 975 Havnevik, K. 1993. Tanzania. The LimitstoDevelopment from Above. Uppsala: 976 The Nordic AfricaInstitute. 977 ——— (2010).‘AHistorical framework for analysingcurrent Tanzaniantransitions: 978 the post-independence model,Nyerere’s ideas and someinterpretations’. 979 In Havnevik, K., Isinika, A.C (eds), Tanzania in Transition: FromNyerere to Mkapa.Dar es Salaam: Mkuki na Nyota,pp. 19–56. 980 HumanDevelopment Report.2006. Beyond scarcity: Power, poverty and the 981 global water crisis.New York: UNDP. 982 Ha˚ rsmar, M. 2010. ‘Whyisagriculture so important to reducing poverty?’ NAI 983 Policy Notes 4(2010).Uppsala: The Nordic Africa Institute. 984 Kikula,I.S. 1997. Policy Implications on Environment. The case of villagization 985 in Tanzania. Uppsala:The Nordic Africa Institute. 986 Kjekshus, H. 1977. ‘The Tanzanian villagization policy: implementational lessons 987 and ecological dimensions’. Canadian Journal of African Studies 11(2): 988 269–82. 989 Maghimbi, S. et al. 2011. ‘The agrarianquestioninTanzania? Astate of the art 990 paper’. Current African Issues 45.Uppsala:The Nordic AfricaInstitute. HOWS_III Chapter 13—6/5/2016—NANDHINI.P—554083—IBTauris

354 AHistory of Water

991 Mesaki, S. 1993. ‘Witchcraft and witch-killings in Tanzania’. PhD Thesis. Minnesota: 992 University of Minnesota. 993 Oestigaard, T. 2009. ‘Water, culture and identity. comparing past and present 994 traditions in the Nile Basin Region’.InOestigaard, T(ed.), Water, Culture and 995 Identity. Comparing Past and Present Traditions in the Nile Basin Region . Bergen: BRIC Press, pp. 11–22. 996 ———2014. ReligionatWorkinGlobalisedTraditions. Rainmamking, 997 witchcraft and Christianity in Tanzania.Newcastle: Cambridge Scholars 998 Press. 999 Sanders, T. 2000. ‘RainsGone bad, women gone mad: rethinking gender rituals of 1000 rebellion and patriarchy’, The Journal of the Royal AnthropologicalInstitute 1001 6(3): 469–86. 1002 ——— 2008. ‘Buses in Bongoland’, AnthropologicalTheory 8(2): 107–32. 1003 Shipton, P. 2010. Creditbetween Cultures. Farmers, Financiers, and 1004 Misunderstanding in Africa .New Haven:Yale University Press. 1005 Tvedt, T. 2010a. ‘Water systems, environmental history and the deconstructionof 1006 nature’. Environment and History 16: 143–66. ——— 2010b. ‘WhyEnglandand notChina and India? Water systems and the 1007 history of the industrial revolution’, JournalofGlobalHistory 5: 29–50. 1008 Wijsen,F., Tanner, R. 2000. Seeking aGood Life. Religion and Society in 1009 Usukuma, Tanzania.Nairobi: Paulines Publications Africa. 1010 ——— (2002). ‘I am just aSukuma’. Globalization and Identity Construction 1011 in Northwest Tanzania.Amsterdam: Rodopi. 1012 UNDP. 2012. Assessing Progress in Africa toward the Millennium Development 1013 Goals. The MDG Report 2012. New York: UNDP. 1014 UNECA ACPC.2011. United Nations Economic Commission for Africa, African 1015 Climate PolicyCenter Agricultural Water Management in the Context of 1016 .Addis Ababa: Working Paper9. 1017 Wani, S.P. et al. 2009. ‘Rainfed Agriculture –Past Trends and Future Prospects’. In Wani, S.P., Rockstro¨m,J., Oweis, T(eds), Rainfed Agriculture: Unlocking 1018 the Potential.Oxfordshire:Cabi, pp. 1–35. 1019 World Food Council 1988. Towards Sustainable Food Security: Critical Issues, 1020 Report by the Secretariat, 14th Ministerial Session, Nicosia, Cyprus,23–26 May. 1021 World Vision2011. . Response to Drought.90-Day Report,http: // 1022 www.worldvision.org/resources.nsf/main/press-haiti/$file/hoa-90-day-report. 1023 pdf (accessed 2July 2014). 1024 WWAP.2012. The United NationsWorld WaterDevelopment Report 4: Managing 1025 Waterunder Uncertainty and Risk.Paris: UNESCO. 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035