citation - icrisatejournal.icrisat.org/mpii/v3i1/pdfs/91-2006.pdf · citation: cooper pjm, dimes j,...

Citation: Cooper PJM, Dimes J, Rao KPC, Shapiro B, Shiferaw B and Twomlow S. 2006. Copingbetter with current climatic variability in the rain-fed farming systems of sub-Saharan Africa: Adress rehearsal for adapting to future climate change? Global Theme on Agroecosystems Report no.27. PO Box 29063-00623, Nairobi, Kenya: International Crops Research Institute for the Semi-Arid Tropics. 24 pp.

Summary

Far greater investment in the rain-fed farming systems of sub-Saharan Africa (SSA) is essential ifchronic food shortages and poverty are to be reduced and progress towards the MillenniumDevelopment Goals is to be achieved. However, in such systems, rainfall variability is the fundamentalfactor that defines production uncertainty, and whilst farmers have learned to cope with currentclimatic variability, they and many associated potential investors are ‘risk averse’ and over-estimatethe impact of rainfall variability on crop and livestock production. As a result they are reluctant tomake such investments when the outcomes seem so uncertain from year to year. Climate changewill result in even greater rainfall variability in many parts of SSA and can only exacerbate thissituation. However, unless the livelihoods of resource poor farmers and the natural resource baseupon which they depend are made more resilient though coping better with current climate variability,the challenge of adapting to future climate change will be daunting for most and perhaps impossiblefor many. ICRISAT is helping to bring together a consortium of national, regional and internationalpartners that will bring new and proven climate risk management tools to address the concerns offarmers and stakeholder investors and will help them build strategic and tactical climate riskmanagement approaches into their planning and activities. Indeed, if a better understanding of theconstraints and opportunities of climatically induced risk is not provided to key stakeholders andfarmers alike, investment in the rain-fed agricultural sector in SSA is likely to remain at its currentlow and inadequate level resulting in persistent poverty and vulnerability of rural populations. TheConsortium experts are convinced that the successful application of these approaches will, by 2015,have facilitated and guided agricultural investments that will play a major role in moving towardsthe achievement of the Millennium Development Goals in several countries in sub-Saharan Africa,and in answering Secretary Kofi Annan’s call for a truly African Green Revolution.

Global Theme on AgroecosystemsReport no. 27

Coping better with current climaticvariability in the rain-fed farmingsystems of sub-Saharan Africa:A dress rehearsal for adaptingto future climate change?

PJM Cooper, J Dimes, KPC Rao, B Shapiro, B Shiferaw,S Twomlow

International Crop Research Institute for the Semi-Arid TropicsPO Box 39063-00623, Nairobi, Kenya.

PO Box 776, Bulawayo, Zimbabwe.Patancheru 502-324, India.

2006

ICRISAT

®

An Open Access Journal published by ICRISAT________________________________________________________________________________________________________

SAT eJournal | ejournal.icrisat.org December 2007 | Volume 3 | Issue 1

AcknowedgementsMany of the ideas presented in this paper were developed whilst the principal author, PJM Cooper,was employed by IDRC, Ottawa, Canada between 2000 and 2004. He is grateful for the support andopportunities provided by IDRC over that period. We would also like to acknowledge input andcomment on an earlier draft of this paper provided by ICRISAT colleagues Jane Alumira, SibiryTraore and Mark Winslow; and from Nuhu Hatibu, Coordinator of the ASARECA Soil and WaterManagement Research Network (SWMnet) and James Hansen of the International ResearchInstitute for Climate and Society, Earth Institute, Columbia University.

PJM Cooper, KPC Rao, International Crop Research Institute for the Semi-Arid TropicsB Shiferaw PO Box 39063-00623, Nairobi, Kenya.B Shapiro International Crop Research Institute for the Semi-Arid TropicsPatancheru 502-324, India.J Dimes, S Twomlow International Crop Research Institute for the Semi-Arid TropicsPO Box 776, Bulawayo, [email protected]@[email protected]@[email protected]@cgiar.org

Authors Contacts



ContentsIntroduction ......................................................................................................................... 1Rain-fed agriculture in sub-Saharan Africa will remain vital for food security ...................... 1Why does investment in rain-fed agriculture remain so low? ................................................ 3Rainfall variability, production uncertainty and climate change ............................................. 3Farmers cope with climate variability, but can they adapt to change? ................................... 5New and proven tools can facilitate investment by a range of stakeholders .......................... 9Conclusions and the way forward ....................................................................................... 15References .......................................................................................................................... 16



1

IntroductionIt is now widely accepted that the impacts of global warming and associated climate change areinevitable in many parts of the world and that adaptation to these changes will be essential. Adaptationstrategies to climate change need to consider both adaptation to longer term changes (50 to 100 years)in key parameters such as rainfall and temperatures and their associated impacts on food production,health, biodiversity and migration, and, of greater and more immediate concern, the predicted, and inmany instances already evident impacts of increased frequency of extreme climatic events (shocks)and greater climatic variability. The second of these two aspects provides the focus for this paper.The need for this focus is reinforced by the fact that such changes will impact most severely on thosealready vulnerable to present day climatic hazards and to within and between season climatevariability. Nowhere is this more true than for those communities in sub-Saharan Africa who relylargely or totally on rain-fed agriculture or pastoralism for their livelihoods. Such communities,already struggling to cope effectively with the impacts of current climatic variability, will face adaunting task in adapting to future climate change. However, whilst it is clear that rural communitiesare indeed the primary ‘investors’ and risk-takers in rain-fed production, there is also a wide range ofassociated ‘investors’ upon whose strategies, decisions and operations they often depend. Farmers andagricultural stakeholders will need to adapt their tactical and strategic planning to these evolvingclimate risks, but given the magnitude of the existing poverty and food security challenges that arefaced in sub-Saharan Africa, adaptation to climate change should not and cannot be divorced fromcurrent development priorities.In this paper we suggest that enhancing the ability of such rural communities and associatedstakeholders to cope better with the constraints and opportunities of present day climatic variability,is in fact a necessary ‘dress rehearsal’ for adapting to future climate change.Rain-fed agriculture in sub-Saharan Africa will remain vital for foodsecurityRecent reviews have considered an impending global water crisis in the context of continuedpopulation growth and predicted climate change. They suggest that the projected trends in worldpopulation growth and dynamics will place substantially greater multi-sectoral demands on water,leading to exacerbated competition between sectors for an increasingly limited supply of abstractedwater (Cosgrove and Rijsberman 2000). In Africa specifically, the projected combined impacts ofclimate change and population growth suggest an alarming increase in water scarcity for manycountries in Africa, with 22 of the 28 countries considered likely to face water scarcity or water stressby 2025 (UNECA 1999), (Figure 1). This in turn will curtail the ability of irrigated agriculture torespond to the expanding food requirements of a global population – particularly those in thedeveloping world. In contrast to the achievement of the Millennium Development Goals, this raisesthe specter of a worsening food security crisis (Rosegrant et al. 2002a).To offset and reverse such a scenario, it has been concluded that much greater emphasis will have to begiven to increasing the productivity of global rain-fed agriculture, which currently provides 60% of theworld’s food. This is especially true in sub-Saharan Africa where currently nearly 90% of staple foodproduction comes, and will continue to come, from rain-fed farming systems (Rosegrant et al. 2002b).In such an endeavor, there are special challenges in Africa’s rain-fed farming systems. It is here thatsome of the poorest and most vulnerable communities live. They manage and largely rely upon rain-



2

Figure 1. Per capita freshwater demand in Africa in 1990 and projected for 2025.

fed agriculture and pastoral systems for their livelihoods, and are the custodians of the natural resourcebase upon which such enterprises depend. Added to the constraints imposed by extreme poverty, andoften a degrading resource base, is the inherent variability of rainfall amounts and distribution.Furthermore, in many instances such communities have been devastated by the HIV/AIDS pandemicthat has further exacerbated their vulnerability through loss of productive labour, knowledge, incomeand the rising dependency burden of taking care of orphans (Yamano and Jayne 2004).Recognizing the importance of rain-fed agriculture in SSA for both individual as well as national foodsecurity, agricultural research and development initiatives have, for decades, developed and promotedagricultural and pastoral innovations that aimed to increase the value and productivity of assets athand, be they land, labour or capital. In many instances, such innovations not only target increasedproductivity, but also attempt to mitigate climatically induced uncertainty of production throughspecific soil, crop and rainfall management strategies.



3

Such research has often shown great potential on research stations and in farmers’ fields, with‘achievable’ yields often several hundred percent greater than those obtained by farmers. However ingeneral terms adoption has been low. Whilst ‘islands of success’ continue to provide hope for the future,little scaling up of such successes is reported, and widespread impact is currently not evident. Indeed, inmany situations, production and the natural resource base upon which it depends are declining. As aresult, cereal deficits in SSA, currently standing at around 9 million tons annually, are projected to morethan triple to 35 million tons by the year 2025 leading to SSA being identified as a “food trade hotspot”.It is highly unlikely that SSA will be in a position to finance such a level of food imports. In such ascenario, either international food aid must be increasingly called upon or policies must be put in placeand decisions taken to greatly accelerate the current trends of investment within the agricultural sectorbeyond the ‘business as usual scenario’ upon which such projections are based (Rosegrant et al. 2002c).Why does investment in rain-fed agriculture remain so low?There are many complex and often daunting inter-related issues that contribute to the current lack ofinvestment and the resultant stagnation of rain-fed production in sub-Saharan Africa. The greenrevolution that made dramatic contributions for improving agricultural productivity and reducingpoverty in Asia and Latin America has largely by-passed sub-Saharan Africa. Lack of investment andstagnation of agricultural production reinforce each other – leading to poverty traps and vulnerabilityof livelihoods to climatic and other shocks (Reardon and Vosti 1995; Collier and Gunning 1999). Themarket-led induced innovation model of agricultural transformation (Ruttan and Hayami 1998) didnot materialize in sub-Saharan Africa mainly because of interplay of market and policy failures (WorldBank, 2000). Whilst agricultural investment by smallholder farmers in risk-prone environments hasoccurred to some extent over the last few decades (LSE 2001), for such investment strategies toblossom and produce the needed impact, favorable policies, institutional arrangements and basicdevelopment infrastructure (including irrigation, roads, electricity and ICT) needed for properfunctioning of markets are required. An enabling investment policy environment would include theexistence of proper incentives, market access, information, input supply systems and institutions(Barret et al. 2002). Low per capita incomes, debt servicing and negative balance of payments at thenational level have undermined the ability of governments to invest in basic infrastructure needed formarkets and the private sector to operate efficiently and effectively. These issues all impinge oninvestment decisions taken by a range of stakeholders within the rain-fed agricultural sector.There is, however, one fundamental factor that underpins all else to a greater or lesser extent andwhich cannot be ignored, and that is the rainfall variability both within and between seasons and theunderlying uncertainty that it imposes on production. It is precisely this uncertainty that constrainsbeneficial ‘investment’ decisions required, not only from farmers and farming communities, but alsofrom a wide range of additional agricultural stakeholders. Such stakeholders often over-estimate theimpact of climate-induced risk. As a result, they show understandable reluctance to invest inpotentially more sustainable, productive and economically rewarding practices when the outcomesand returns seem so uncertain from year to year.Rainfall variability, production uncertainty and climate changeIn systems reliant on rainfall as the sole source of moisture for crop or pasture growth, seasonal rainfallvariability is inevitably mirrored in both highly variable production levels as well as in the risk-averselivelihood and coping strategies that have emerged over time amongst rural populations.



4

This is particularly evident in the semi-arid tropics (SAT) of Africa. Home to approximately 80million of the continent’s most impoverished and marginalized communities (Shapiro and Sanders2002), the SAT are also increasingly under pressure from expanding livestock numbers and an in-migration of peoples from more favourable agro-ecosystems where population pressure, reduced farmsize and resource degradation are resulting in agriculture no longer being a viable livelihood option.Added to this, it is in the SAT of Africa that climate variability and climate extremes have their mostprofound impacts on production. Inherent variability in seasonal rainfall totals increasesdisproportionately as one moves from the sub-humid to the semi-arid regions of East and SouthernAfrica (Rao 2004 Pers. Comm.). (Figure 2).

Figure 2. Rainfall means and their coefficient of variation in eastern and southern Africa.Whilst seasonal rainfall totals and their season to season variability are in themselves important, thenature of within season variability can also have a major effect on crop productivity. This can beillustrated by simulating maize yields for Machakos in Kenya using a crop growth simulation modeldriven by historical daily climate data (Rao 2004). As would be expected, there is a general trend ofincreasing yield as seasonal rainfall totals increase from 100 to 500mm, but there is considerable yieldvariation within that relationship resulting from the contrasting patterns of within season rainfalldistribution experienced in any given year. This is particularly evident in drier seasons receiving below200mm. (Figure 3).The dependence of crop yields on variable rainfall, and the increasing production uncertaintyexperienced in progressively drier environments can be further illustrated by comparing the long termvariability in yields of crops generally grown in wetter environments with those of crops morenormally grown in drier environments. For example, an analyses of national average yields for maize(grown in wetter environments) and millet (grown in drier environments) in Kenya for the period1980 to 2002 indicates both the inherent variability of rain-fed cereal production and the lower and



5

more variable yields of millet (range 250-1100kg/ha with CV of 33%) compared with maize (1200-2050kg/ha with CV of 13%). Such data will, in fact, underestimate the yield variability experiencedby individual farmers since (a) national average yields even out the spatial variability of rainfall withinany given season, and (b) FAO statistics are derived from ‘harvested area’ and thus will over-estimateaverage yields in poor years when crop failure, especially in the drier areas, is widespread (Cooper2004).Overlaid on this challenging scenario is the widely accepted prediction that, whatever happens tofuture greenhouse gas emissions, we are now locked into global warming and inevitable changes toclimatic patterns which are highly likely to exacerbate existing rainfall variability in SSA and furtherincrease the frequency of climatic extremes (IPCC 2001). Indeed, evidence for such climate change isalready emerging in eastern and southern Africa (CCI/CLIVAR 2004; Larsson and Ormann 2004).Adaptation to climate change is therefore no longer a secondary and long-term response option only tobe considered as a last resort. It is now prevalent and imperative, and for those communities alreadyvulnerable to the impacts of present day climatic hazards, an urgent imperative (IISD 2003).Farmers cope with climate variability, but can they adapt to change?Coping StrategiesOver generations, and especially in the more arid environments where rainfall variability impacts moststrongly on livelihoods, people have developed coping strategies to buffer against the uncertaintiesinduced by year-to-year variation in water supply coupled with the socio-economic drivers that impact

Figure 3. Seasonal rainfall totals and simulated maize yields (40kg/ha N), Machakos, Kenya.



6

on their lives. Whilst such coping strategies have been of greatest importance and have evolved overmany generations in the drier and more risk prone environments, they have perhaps only recentlygained importance in many of the wetter and more assured environments as a range of factors(population pressure, declining soil fertility, weed invasion, decreasing farm size, disease, lack ofmarkets or access to markets for high value produce, lack of off-farm employment, etc) are resultingin agriculture becoming a less viable foundation for rural livelihoods (Jayne et al. 2003).Depending on subjective assessment of risks and vulnerability, farm households make certainadjustments in their choice of technologies, and production and consumption decisions. Such copingstrategies can be broadly grouped into three categories: (a) ex-ante risk-management options such aschoice of risk-tolerant varieties, investment in water management, and diversification of both farmingand other associated livelihood enterprises prior to the onset of the season, (b) in-season adjustmentof crop and resource management options in response to specific climatic shocks as they evolve, and(c) ex-post risk management options that minimize livelihood impacts of adverse climatic shocks (eg,distress sale of assets, borrowing, cutting expenditures on non-essential items). Matlon andKristjanson (1988) provide an example of such a matrix to describe such coping strategies in the semi-arid tropics of West Africa and also consider the ‘spatial scale’ at which the various strategies operate.(Table 1).

Whilst this matrix provides a useful general regional picture, it is also well understood there will beregion-to-region, village-to-village and household-to-household variation in coping strategies that haveevolved. For example, in a study of over 100 households in Kezi village of Matabeleland, Zimbabwe,Alumira (2002) confirmed the broad range of contrasting diversification strategies employed betweendifferent types of households headed by either females (de jure or de facto) or males, with theownership or lack of ownership of cattle being a key factor that cuts across household types, andwhich provided considerable additional flexibility.

Table 1 Coping strategies used by farmers in semi-arid West Africa.Time Frame

Scale Before the season During the season After the seasonVarietal selection for stresstolerance/resistanceStaggered planting dates.Low density planting.Intercropping. Run-offmanagement. Delayedfertilizer use.Diversified croppingLand typediversificationPlotfragmentationCereal stocksLivestock/assetsSocial and Off-farmemployment networks

Plant

Farm

Plot

Household,village, region

Replanting with earliermaturing varieties.Changing crops whenre-planting.Increasing or decreasing plantdensity at re-planting or bythinning

Matching weeding labour inputsto expectations of the season

Shifting crops between landtypes

Grazing of failed plotsfor animal maintenance.

Late planting for forage

Asset sales for cerealpurchases.FoodtransfersMigrationemployment.



7

In the even drier environments, cropping is largely impossible and certainly highly risky both withregard to production and environmental degradation. Here pastoralism dominates. In suchenvironments coping strategies assume even greater importance, but are perhaps less diversified dueto the more restricted asset base and the more marginalized nature of such communities. McIntire(1991) notes that mixed species herds, widespread and seasonally available pastures, splitting animalsinto discrete herds and mobility in response to seasonal variation in pasture productivity, are keystrategies. Where the opportunities exist, working as wage labourers, trading commodities andgrowing crops are also common. He argues that the risk associated with livestock production in suchdry environments inflates incentives to invest, and since animals are the only asset, herders tend tohold more than the ‘profit maximizing’ number of head in order to insure that they will remain viableafter any given disaster. Such a strategy often leads to overstocking and overgrazing, and can eventuallyprove a serious threat to the resource base (Cooper and Bailey 1991).Adaptive Strategies, Adaptive Capacity and Livelihood AssetsThus far, we have highlighted ‘diversification’ and ‘response’ coping strategies that have evolved todeal with both expected rainfall uncertainty and the evolving within season fluctuations in rainfall. Inmany parts of the world however, longer-term changes have and are impinging on the livelihoods ofrural communities, and thus the nature and relative importance of such strategies cannot remainunchanged. Adaptation to these longer-term changes is required both by farming communities as wellas those stakeholders with whom they interact and on whom they often depend. The extent to whichpeople and institutes are able to successfully respond to a new set of circumstances will depend upontheir Adaptive Capacity.The concept of adaptive capacity is well illustrated by a village level study conducted in the semi-aridtropics of India over a 25-year period in 10 villages (Bantilan and Anupama 2002). Evidence from thevillages of Aurepalle and Dokur reveals the acute effects of persistent drought and increasing waterscarcity on livelihood strategies. Almost all dug wells in both villages have dried up and villageirrigation tanks (previously filled through run-off) have not filled for a decade. Farmers are nowforced to leave much of their land fallow and the % income derived from agriculturally relatedactivities has declined dramatically – from 88 to 47% in Aurepalle and from 94 to 35% in Dokur.However, farm families have successfully adapted and diversified their livelihood strategies thoughincreased off-farm activity, caste occupations and seasonal job migration. Indeed, in real terms, theyhave higher incomes today as a result. In other words, it would appear that the communities in thesetwo villages (under the particular new stresses that they have experienced) have had a high adaptivecapacity (Table 2).Although households have adapted to changes induced by recurrent drought through diversificationinto off-farm activities in these villages, this may not be a feasible alternative to many smallholderfarmers in isolated and less-favored areas of rain-fed systems in Africa. This implies the need todevelop new options and innovations that enhance the resilience of agricultural production and reducevulnerability to such shocks in rural areas. Research investments to enhance tolerance to droughtstress, improving water productivity, and integrated management of land and water resources (eg,watershed management) have the potential to reduce vulnerability to climatic shocks while alsoimproving productivity. This can be illustrated by the effects of integrated watershed management inKothapally village (relatively close to Auropalle and Dokur in semi-arid Andhra Pradesh, India), whichhas contributed to improved resilience of agricultural incomes despite the high incidence of drought



8

(Table 3). While drought-induced shocks reduced the average share of crop income (in totalhousehold income) in the non-project villages from 44 to 12%, this share remained unchanged atabout 36% in the adjoining watershed project village of Kothapally (Shiferaw et al. 2005).

Table 2. Changes in the % contribution of different sources and the levels of household netincome in two villages in the semi-arid tropics of India. Aurepalle Dokur

Sources of Income 1975 - 1978 2001 - 2002 1975 - 1978 2001 - 2002Crops 30 15 46 3Ag. labour 33 23 46 14Livestock 25 9 2 18Off-farm activities 12 13 1 24Caste occupations 0 28 0 6Seasonal migration for work 0 8 0 20Others 0 4 5 15Net Household Income. (rupees) # 15205 31561 19107 36757# Incomes for the period 1975 – 1978 represent equivalent values of base year incomes at current prices.

Table 3. The effect of integrated watershed interventions on alternative sources of householdincome (Rs 1000).CropIncomeYear Village groupa Statistics Livestockincome Off-farmincome Household Income

Non-project Mean incomeShare of totalincome (%)

2001(averageyear)Watershedproject Mean income

Share of totalincome (%)Mean incomeNon-ProjectShare of totalincome (%)

2002(droughtyear)Watershedproject Mean income

Share of totalincome (%)

12.7 1.9 14.3 28.944.0 6.6 49.5 100.015.4 4.4 22.7 42.536.2 10.4 53.4 100.0 2.5 2.7 15.0 20.212.2 13.3 74.5 100.010.1 4.0 13.4 27.636.7 14.6 48.7 100.0

a The sample size (n=60 smallholder farmers) in each group



9

Central to the concept of adaptive capacity is the idea of livelihood assets. These are the means ofproduction available to a given individual, household or group that can be used in their livelihoodactivities. These assets are the basis on which livelihoods are built. Five types of livelihood assets havebeen described, namely (i) Natural capital, (ii) Social-political capital, (iii) Human capital, (iv)Physical capital and (v) Financial capital (DFID, 1998). Taken together, knowledge of these assetshelp us understand how livelihoods work, and in the context of this paper, how people respond toclimatic variability and adapt to change (IISD, 2003). In general, the stronger, more resilient andmore varied the asset base, the greater the people’s adaptive capacity and the level of security andsustainability of their future livelihoods.It is this recognition that provides the underlying premise of this paper, namely that coping better withcurrent climate variability is a prerequisite for adaptation to the predicted greater rainfall variabilityassociated with climate change.New and proven tools can facilitate investment by a range ofstakeholdersIn recent years, a range of innovative climate analytical tools have been developed and proven. Thesetools allow for a far greater understanding of the temporal and spatial agricultural implications of shortand longer-term climatic variability. Will such an understanding facilitate broad-based investment inthe uncertain sector of rain-fed agriculture?Simply put, we believe that the answer is ‘yes’, and from two complementary perspectives, namely (i)through the shorter-term seasonal climatic forecasting and (ii) through the characterization andmapping of the longer-term agricultural implications of climate variability.Seasonal Climate ForecastingRecent advances in understanding and modeling of the oceanic atmosphere system at the global andregional scales are important developments that result in the evolving potential of seasonal weatherforecasting being evaluated and demonstrated in Asia (Balusubramanian et al. 2002; Huda andPackham 2004) as well as in several regions of SSA (IRI 2005). These developments have animportant role to play in assisting farmers and other investors optimize their immediate decisions andtactical planning with regard to the approaching season. Whilst the potential predictability of climateanomalies for parts of Africa is amongst the highest anywhere in the world, currently thepredictability of seasonal rainfall remains variable within Africa (Washington et al. 2004). (Table 4)The high level of predictability of the October, November and December rains in East Africa areillustrated by the results of a study undertaken at Machakos in Kenya. (Figure 4).Table 4. Qualitative assessment of potential predictability of seasonal forecasts in AfricaRegion Rainfall Period Potential PredictabilityWest Africa July to September HighEast Africa October to December HighEast Africa March to May Low to MediumSouthern Africa January to March MediumNorth Africa LowCongo, Mozambique, Angola Unknown



10

Surveys and pilot studies in eastern and southern Africa show that farmers see opportunity to benefitfrom seasonal forecasts (see Table 5). It is interesting, but perhaps not surprising, to note thesimilarity of management responses to seasonal variation in the semi-arid tropics of SSA betweenthose observed 20 years ago in West Africa (Table 1) and those of today by East African farmers.However, such studies have also shown that farmers are often constrained by the timing, scale andformat of available forecasts, lack of trust or comprehension of the forecasts, and need for competentguidance for livelihood responses as requirements for rural communities to use seasonal forecastseffectively (O’Brian et al. 2000; Ngugi 2002; Patt and Gwata 2002; Rao and Okwach 2005).Effective agrometeorology extension can address these challenges and facilitate the effective use offorecast information by supporting dissemination, interpretation, education, technical guidance, andfeedback to forecast providers. This generally requires close collaboration between agricultural andmeteorological institutions.

Figure 4. IRI weather forecasts and observed rainfall for October, November and December at Machakos,Kenya (1981-2003).

Table 5: Some farmer identified management options for below normal and normal to abovenormal seasons at Machakos, Kenya. (Rao and Okwach 2005)Management Decisions

Dry Season Normal to Wet Season1. Use low plant density (2.2 plants/m2)2. Reduce labor and other input use3. Increased use of drought tolerant crops suchsorghum, millet, green grams, and cassava4. Plough and plant early before the start of the rain5. Adopt water conservation measures6. Reduce area under cultivation

1. Use higher plant density (3.5 to 4.5 plants/m2)2. Apply fertilizer3. Plant hybrid maize varieties such as pioneer4. Adopt intercropping5. Strengthen terraces6. Increase area under cultivation



11

In addition, evidence suggests that farmers are most likely to trust and act on information and advicewhen it comes from sources that they already know and trust (Hansen 2002). Thus, depending on thecontext, the impact of agrometeorology extension may be most effective through existing agriculturalextension services, meteorological services, development NGOs, agribusiness, farmer associations orcommunity leaders. Such evidence reinforces the need for a multi-stakeholder approach to enhancingfarmers’ coping strategies to seasonal climatic variability. In addition to the potential benefits that canaccrue to farming communities themselves, seasonal climate forecasts also assist in national and/orregional disaster preparedness. This is more true if the approach used links seasonal forecasts with theuse of crop growth simulation models that provide probabilistic crop yield and production estimateswell in advance of harvest (Hansen and Indeji 2004). Weather data linked to changes in crop yields canalso be instrumental in expanding crop insurance schemes to smallholder farmers in risk-proneenvironments (Skees et al. 1999).Characterizing and mapping the agricultural implications of climatic variabilityThe use of long-term daily climatic data combined with field based research results, spatial weathergenerators, crop growth simulation models, geographic information systems and improved access toand use of climate analysis software allow for the development of robust frameworks which canfacilitate and guide risk assessment and management, longer term strategic planning, and decisionmaking by all ‘investors’ involved in rain-fed farming. Such work can incorporate various degrees ofcomplexity, and is usually based upon the use of long term daily climatic records. Crops principallyrespond to daily or sequences of daily rainfall, and thus daily rainfall becomes the key parameter inrain-fed agriculture. Such records have been collected throughout SSA for decades, and in thiscontext are now proving to be invaluable. The use of such records allows the determination of the“probability” of occurrence of a wide range of parameters of importance to agriculture, and hence therisk associated with them.At one level of analyses, research can focus on the probability of climatic events of known importanceto agriculture such as the start of the growing season, the frequency of dry spells within the season, thefrequency of high intensity erosive rainfall events, the impact of prolonged wet spells on plant diseaseor the length of the growing season itself (Sivakumar 1988; Virmani and Shurpali 1999). Suchanalyses are becoming increasingly easy to undertake as initiatives to provide more user-friendlysoftware, and the training to go with it, take place. Just such an example in SSA is the Statistics inApplied Climatology (SIAC) course. SIAC has been running since 1982, first at Reading University(UK) until 2000, and now continues at the Institute for Meteorological Training and Research (IMTR)in Nairobi and at the AGRHYMET Regional Centre in Niamey for French speaking countries. SIAChas the specific objective of helping national meteorological services to become more involved in theagricultural development processes in their respective countries. Training focuses on (a) the use ofclimate data management tools, especially in the use of Climsoft, a successor to Clicom, which wasdeveloped by staff from Kenya, Zimbabwe and Botswana, and (b) the use of new, improved and userfriendly software for the analyses of climatic data. Instat and Genstat, for example, have had specialfacilities for climatic analyses added. SIAC is currently evaluating e-learning as an approach to scale upthe impact of its training.A further step is to use crop growth simulation models that are driven by daily climatic data to predictthe impact of long term climate variability on the probability of success of a range of crops and soilmanagement strategies. One such model that is becoming increasingly used in SSA is the Agricultural



12

Figure 5. Predicted (APSIM) and observed maize grain yields on two soil types in Kenya.

Figure 6. Predicted (APSIM) performance of short and long season maize cultivars and the impact of nitrogenfertilizer (ammonium nitrate) at Bulawayo, Zimbabwe. 1952-1998.

Productions Systems Simulator (APSIM). APSIM can simulate the growth and yield of a range ofcrops amongst which maize, sorghum, pearl millet, chickpea, soybean, groundnut, sunflower, cotton,weeds and trees are likely to be of most interest in SSA. APSIM has been well calibrated for most ofthese crops (Figure 5) and allows the comparison and assessment of a range of crop and soilmanagement practices across seasons (Figure 6). Such results can also be expressed in financial termsand be usefully presented in a cumulative probability format (Figure 7) (Dimes 2005).



13

One thing is clear. Such simulation modeling can be invaluable in posing a wide range of ‘what if ’questions, which mirror those asked by farmers and can provide valuable insights and answers framedin the context of the long term characteristics of climate variability in any given location. In otherwords, they can contribute directly to enhanced and more resilient coping and adaptive strategies.Indeed, recent activities that link model input and output directly with farmers and participatoryresearch approaches has proved to be powerful (Dimes et al. 2003).The tremendous value of the type of research described above is however constrained to some extentby the fact that it relies upon ‘point source’ climate data collected at specific weather stations, thusmaking interpolation of the outputs between weather stations problematic. This can be overcome bythe use of modern and proven spatial weather generators such as MarkSim. MarkSim is a spatiallyexplicit daily weather generator that was developed at CIAT and was released in 2004. The climatesurfaces that are produced use data from 10,000 stations in Latin America, 7000 from Africa and 4500from Asia. MarkSim relies on climatic data surfaces interpolated from weather stations and generateslong-term weather records on a grid basis of 18x18km. The characteristics of the long-term dailyweather data that is generated by MarkSim compares well with existing long-term climatic records fora number of meteorological stations in the semi-arid tropics of Kenya, as illustrated in Figure 8.(Farrow 2005).The combined use of crop growth simulation models, historic climatic data sets and weather generatorssuch as MarkSim is a powerful combination that allows both the characterization and the subsequentmapping of the agricultural implications of climatic variability (Jones and Thornton 2002). It is alsopossible to integrate different climate change scenarios into MarkSim and, through crop growthsimulation models, assess their impact on agricultural production (Jones and Thornton 2003).

Figure 7. Cumulative probability distribution of predicted (APSIM) financial returns to recommended N-application (60kg/ha) and a lower rate of N-application (17kg/ha, weeded and unweeded) on maize productionat Masvingo, Zimbabwe.



14

Integrating approachesWith the increasing availability, reliability and ease of use of such tools as described above, it nowbecomes possible for decision-makers and investors involved in agriculture to formulate adevelopment agenda that integrates the following three key aspects of climate risk management,namely:1. Decision-support frameworks that provide a longer-term strategic understanding of the temporaland spatial distribution of climatic variability and its impact on the probability of performance andprofitability of existing and innovative agricultural practices.2. Seasonal climate and agricultural forecasting to enable farmers and other stakeholders to ‘fine tune’long-term strategies in the context of the approaching season and thus to plan tactically and farmmore effectively in the context of variable weather.3. Information on the extent to which climate change is impacting, or is likely to impact, on the natureof climate variability and the implications for rain-fed farming systems and their futuredevelopment and productivity.Who are the investors in the agricultural sector who would benefit from integratedclimate risk management strategies?The demand for improved climate risk management strategies is increasingly being voiced by a broadrange of investor stakeholders who are seeking to identify appropriate short and longer-terminvestment strategies, for example:• National and District Policy Makers who are charged with making short and longer-termagricultural investment decisions on the types of development initiatives to promote and support inany given season and area• The Private Sector and Micro-finance Institutions wishing to have a clear picture of season toseason variability in production and its implications for the establishment and sustainability ofviable market enterprises and financing schemes

Figure 8. Comparison of probability of exceeding rainfall totals for the Long Rains Season (Mar-May) ofobserved data 1959 - 2000 and data simulated using MarkSim for Makindu, Kenya.



15

• Extension Services and Development NGO’s working with farmers who would wish to bettertarget and test innovations that have a long-term acceptable level of probability of success, and whowould wish to advise their clients innovations likely to be most appropriate in the coming season• Farmers and Farmer Groups who would wish to have information on the likely performance of aninnovation in good, average or poor years before singly or jointly making short-term or long-terminvestment in such an innovation• Disaster Relief Agencies and National Policy Makers who would wish to have due warning ofimpending food shortages in any given season coupled with a longer-term temporal and spatialperspective on the probability of such shortages and appropriate post-disaster recovery strategies• National and Regional Meteorological Services who are increasingly seeking opportunities to usetheir information and skills in the agricultural development arena.Conclusions and the way forwardFor agricultural communities and agricultural stakeholders in SSA to adjust to climate change and thepredicted future increases in climate variability, their ability to cope better with the constraints andopportunities of current climate variability must first be enhanced. Tools and approaches are nowavailable that allow for a far better understanding, characterization and mapping of the agriculturalimplications of climate variability and the development of climate risk management strategiesspecifically tailored to stakeholders needs.To this end, ICRISAT is helping build a consortium of 15 national, regional and internationalorganizations, all of whom have endorsed this approach as a way forward in SSA. In addition, at theAfrican Union/NEPAD meeting in Accra in May 2005, ASARECA and COMESA presented theconcept note that laid the foundation for this work. It was officially endorsed by the NEPAD-G8Implementation Plan for the Comprehensive Africa Agriculture Development Programme (CAADP).The consortium members are now planning directly with a range of stakeholders in Eastern, Southernand Western Africa. These stakeholders are being selected on the basis that they meet the followingfour criteria.(1) They are directly involved in activities aimed at improving the livelihoods of ruralcommunities who are largely or totally dependant on agriculture(2) They have expressed a clear need for an integrated approach to climate risk management toenhance the effectiveness, efficiency and targeting of their interventions(3) They are keen to undertake pilot projects to evaluate such climate risk managementapproaches(4) Upon the successful outcome of Pilot Projects, they are well placed to scale-up the climaterisk management approaches that they have validated.The aim of this initiative is to develop an initial set of projects that will directly contribute to thedevelopment and implementation of integrated climate risk management strategies that guide andtarget seasonally responsive investment in rain-fed agriculture. When scaled up, such strategies willprovide a more resilient foundation upon which adaptation to climate change can be built.



16

ReferencesAlumira JD. 2002. Rural household investment behaviour and implications for development pathways: a casestudy in the semi-arid tropics of Zimbabwe. Presented at the CGIAR International Social Research Conference,11 – 14 September, 2002, Cali, Colombia.Balusubramanian T, Huda A, Selvaraju R and Geethalakshmi V. (Eds.). 2002. Proceedings of the InternationalConference on Seasonal Climate Forecasts in Agricultural Management. 24-26th September 2002. Tamil NaduUniversity, Coimbatore-641003, India. 147 pp.Bantilan MSC and Anupama KV. 2002. Vulnerability and Adaptation in Dryland Agriculture in India’s SAT:Experiences from ICRISAT’s Village-Level Studies. Working Paper Series no.13. Patancheru 502 324, AndhraPradesh, India: ICRISAT. 20 ppBarrett C, Lynam J, Place F. 2002. Towards improved natural resource management in African agriculture. InNatural Resources Management in African Agriculture: Understanding and Improving Current Practices.(Barrett CB and Place F, eds.). CABI Publishing.CCI/CLIVAR. 2004. Expert Team on Climate Change Detection, Monitoring and Indices. Southern AfricaWorkshop Report ( New M, ed.). (http://www.clivar.org/organization/etccd)Collier P and Gunning J. 1999. Why has Africa grown slowly? Journal of Economic Perspectives 13(3): 3-22.Cooper PJM. 2004. Compiled statistical database. ICRISAT, Nairobi, Kenya. Derived from FAO Statistics.(www.faostat.org).Cooper PJM and Bailey E. 1991. Livestock in Mediterranean Farming Systems: Once a Traditional BufferAgainst Uncertainty – Now a threat to the Agricultural resource base? In ‘Risk in Agriculture.’ Proceedings ofthe Tenth Agriculture Sector Symposium, World Bank, Washington, D.C. Pages 81 – 120.Cosgrove WJ and Rijsberman FR. 2000. World Water Vision: Making Water Everybody’s Business. London:World Water Council and World Water Vision and Earthscan.DFID. 1998. Sustainable Livelihoods. (Carney D, ed.). London: DFID.Dimes JP. 2005. Application of APSIM to evaluate crop improvement technologies for enhanced water useefficiency in Zimbabwe’s SAT. In Management for improved water use efficiency in the dry areas of Africa andWest Asia: Proceedings of a workshop organized by the Optimizing Soils Water Use (OSWU) Consortium.April 2002, Ankara, Turkey. (Pala M, Beukes DJ, Dimes JP and Myers RJK, eds.). Aleppo Syria; ICARDA andPatancheru, India: ICRISAT. 288 pp.Dimes J, Twomlow S and Carberry P. 2003. Application of new tools: exploring the synergies betweensimulation models and participatory research in smallholder farming systems. Global Theme 3: Water, Soil andAgrodiversity Management for Ecosystem Resilience. Report no. 5. PO Box 776, Bulawayo, Zimbabwe:ICRISAT, and PO Box 102, Toowoomba, Queensland 4350, Australia: CSIRO Sustainable Ecosystems/APSRU. 20 pp.Farrow A. 2005. Climatic data and tools for risk mapping as an aid to FAO Farmer Field Schools in the semi-aridareas of Kenya. Working Week Report, 28 November to 2 December, 2005. CIAT, Kawanda, Kampala, Uganda.Hansen JW. 2002. Realizing the potential benefits of climate prediction to agriculture: issues, approaches,challenges. Agric. Syst. 74:309-330.Hansen JW and Indeje M. 2004. Linking dynamic seasonal climate forecast with crop simulation for maizeyield prediction in semi-arid Kenya. Agricultural and Forest Meteorology 125:143-157.Huda AKS and Packham RG. (Eds.). 2004. Using seasonal climate forecasting in agriculture: a participatorydecision making approach. ACAIR Technical Reports No. 59. 52 pp.



17

IISD. 2003. Livelihoods and Climate Change. A Conceptual Framework Paper prepared by the Task Force onClimate Change, Vulnerable Communities and Adaptation. IUCN, SEI, IISD.IPCC. 2001. Intergovernmental Panel on Climate Change. Climate Change 2001:The Scientific Basis.Contribution of the Working Group 1 to the Third Assessment Report of the IPCC. Cambridge, CambridgeUniversity Press. <www.grida.no/climate/ipcc_tar/wg1/index.htm>IRI. 2005. Sustainable Development in Africa: Is the Climate Right?http://iri.columbia.edu/africa/whatisnew/SusDevAfricafinal.pdfJayne TS, Yamano T, Weber M, Tschirley D, Benfica R, Chapato A and Zulu B. 2003. Small Holder Incomeand Land Distribution in Africa: Implications for Poverty Reduction Strategies. Food Policy, 28(3): 253-275.Jones PG and Thornton PK. 2002. Spatial modeling of risk in natural resource management. ConservationEcology 5(2): 27. [online] http://www.consecol.org/vol5/iss2/art27Jones PG and Thornton PK. 2003. The potential impacts of climate change on maize production in Africa andLatin America in 2055. Global Environmental Change 13 (1): 51-59.Matlon P and Kristjanson P. 1988. Farmer’s Strategies to Manage Crop Risk in the West African Semi-aridTropics. In Challenges in Dryland Agriculture: a Global Perspective. Proceedings of the InternationalConference on Dryland Farming, August 15 – 19, 1988, Bushland Texas, USA. Pages 604 – 606.cIntire J. 1991. Managing Risk in African Pastoralism. In Risk in Agriculture. Proceedings of the TenthAgriculture Sector Symposium, The World Bank, Washington, DC. Pages 129 – 142.Ngugi RN. 2002. Climate Forecast Information: The Status, Needs and Expectations among Smallholder Agro-pastoralists in Machakos District, Kenya. IRI Tech. Rep. 02-04. IRI, Palisades, New York.Larsson N and Ormann L. 2004. A study of deforestation’s impact on soil properties, infiltration and erosion,and change in rainfall pattern in Lake Nakuru catchment, Kenya. Swedish University of Agricultural Sciences,Uppsala, August 2004. SLU Minor Field Studies No. 273. ISSN 1402-3237.LSE. 2001. Livelihood transformations in semi-arid Africa 1960-2000. Policy lessons from farmers’ investmentstrategies in Kenya, Senegal, Niger and northern Nigeria. Working Paper 40. A workshop held at the LondonSchool of Economics, 17 January 2001.O’Brien K, Sygna L, Næss LO, Kingamlono R and Hochobeb B. 2000. Is Information Enough? UserResponses to Seasonal Climate Forecasts in Southern Africa. CICERO Rep. 2000:3, Ctr. for Int’l. Climateand Env. Res., Oslo.Patt AG and Gwata C. 2002. Effective seasonal climate forecasts applications: Examining constraints forsubsistence farmers in Zimbabwe. Global Environ. Change - Human and Policy Dimensions 12:185-195.Rao KPC. 2004. ICRISAT, Nairobi, Kenya. Personal Communication.Rao KPC and Okwach GE. 2005. Enhancing Productivity of Water under Variable Climate. Paper prepared forthe Integrated River Basin Management Conference, Morogoro, Tanzania, 7-9 March 2005.Reardon T and Vosti SA. 1995. Links between rural poverty and the environment in developing countries: assetcategories and investment poverty. World Development 23(9):1495-1506.Rosegrant MW, Cai X and Cline SA 2002a. Global Water Outlook to 2025: Averting an Impending Crisis.IFPRI Food Policy Report. Sept. 2002.Rosegrant MW, Cai X and Cline SA. 2002b. World Water and Food to 2025: Dealing with Scarcity.(Washington DC. IFPRI 2002).Rosegrant MW, Cai X and Cline SA. 2002c. The Role of Rain-fed Agriculture in the Future of Global FoodProduction. IFPRI, Environment and Production Technology Division Discussion Paper No. 90. February 2002.



18

Ruttan VW and Hayami Y. 1998. The induced innovation model of agricultural development. In Internationalagricultural development. Third edition. (Eicher C and Staatz J, eds.). The Johns Hopkins University Press.Shapiro BI and Sanders JH. 2002. Natural Resource Technologies for the Semi-arid Regions of Sub-SaharanAfrica. Pages 261-274 in ‘Natural Resources Management in African Agriculture.’ (Barrett CA, Place F and AboudAA, eds.). CABI Publishing, Oxon, UK. In association with ICRAF, Nairobi, Kenya. ISBN 0-85199-584-5Shiferaw B, Wani S and Sreedevi TK. 2005. Collective Action for Integrated Community WatershedManagement in Semi-Arid India: Analysis of Multiple Livelihood Impacts and the Drivers of Change. Paperprepared for the International Association of Agricultural Economists (IAAE), August 2006, Brisbane,Australia.Skees J, Hazell P and Miranda M. 1999. New approaches to crop yield insurance in developing countries.Discussion Paper 55, International Food Policy Research Institute (IFPRI), Washington DC. 29 pp.Sivakumar MVK. 1988. Predicting rainy season potential from the onset of the rains in southern Sahelian andSudanian climatic zones of West Africa. Agricultural and Forest Meteorology 42:295-305UNECA. 1999. Addis Ababa: Global Environment Outlook 2000 (GEO), UNEP, Earthscan, London.Virmani SM and Shurpali NJ. 1999. Crop establishment risks in Groundnut production in the Anantapurregion, India. ICRISAT Research Bulletin. ICRISAT, Patancheru, India.Washington R, Harrison M and Conway D. 2004. African Climate Report. Report commissioned by the UKGovernment to review African climate science, policy and options for action. www.defra.gov.uk/environment/climatechange/ccafrica-study/index.htmWorld Bank. 2000. Can African claim the 21st century? The World Bank, Washington DC.Yamano T and Jayne TS. 2004. Measuring the impacts of working-age adult mortality on small-scale farmhouseholds in Kenya. World Development 32 (1): 91-119.



91–2006

®

About ICRISAT

The International Crops Research Institute for the Semi-Arid Tropics (ICRISAT) is a nonprofit,

non-political organization that does innovative agricultural research and capacity building for

sustainable development with a wide array of partners across the globe. ICRISAT’s mission is to

help empower 600 million poor people to overcome hunger, poverty and a degraded environment in

the dry tropics through better agriculture. ICRISAT belongs to the Alliance of Future Harvest Centers

of the Consultative Group on International Agricultural Research (CGIAR).

Contact Information

ICRISAT-Patancheru

(Headquarters)

Patancheru 502 324

Andhra Pradesh, India

Tel +91 40 30713071

Fax +91 40 30713074

[email protected]

Liaison Office

CG Centers Block

NASC Complex

Dev Prakash Shastri Marg

New Delhi 110 012, India

Tel +91 11 32472306 to 08

Fax +91 11 25841294

ICRISAT-Nairobi

(Regional hub ESA)

PO Box 39063, Nairobi, Kenya

Tel +254 20 7224550

Fax +254 20 7224001

[email protected]

ICRISAT-Niamey

(Regional hub WCA)

BP 12404

Niamey, Niger (Via Paris)

Tel +227 722529, 722725

Fax +227 734329

[email protected]

ICRISAT-Bamako

BP 320

Bamako, Mali

Tel +223 2223375

Fax +223 2228683

[email protected]

ICRISAT-Bulawayo

Matopos Research Station

PO Box 776,

Bulawayo, Zimbabwe

Tel +263 83 8311 to 15

Fax +263 83 8253/8307

[email protected]

ICRISAT-Lilongwe

Chitedze Agricultural Research Station

PO Box 1096

Lilongwe, Malawi

Tel +265 1 707297/071/067/057

Fax +265 1 707298

[email protected]

ICRISAT-Maputo

c/o INIA, Av. das FPLM No 2698

Caixa Postal 1906

Maputo, Mozambique

Tel +258 21 461657

Fax +258 21 461581

[email protected]

Visit us at www.icrisat.org

®



citation - icrisatejournal.icrisat.org/mpii/v3i1/pdfs/91-2006.pdf · citation: cooper pjm, dimes j,...

Documents