WPS4288 Policy ReseaRch WoRking PaPeR 4288 Actual Crop Water Use in Project Countries A Synthesis at the Regional Level Robina Wahaj Florent Maraux Giovanni Munoz The World Bank Development Research Group Sustainable Rural and Urban Development Team July 2007 Policy ReseaRch WoRking PaPeR 4288 Abstract This report aims to synthesize the results of a crop water maximum evapotranspiration, especially for the rainfed use study conducted by country teams of the GEF/World crops. This highlights the importance of improved water Bank project, Regional Climate, Water, and Agriculture: management if agriculture is to play an important role as Impacts on and Adaptation of Agro-ecological Systems in a source of food security and better livelihoods. Africa. It also presents the results of the second phase of The report highlights the vulnerability of maize to the study based on climate change scenarios, conducted water stress and the increased risks to the viability of by the South Africa country team. rainfed farming systems based on this crop. The results The actual evapotranspiration of five commonly grown of the second phase of analysis show that a 2°C increase crops--maize, millet, sorghum, groundnuts, and beans-- in the temperature and a doubling of carbon dioxide in two selected districts were analyzed by six country concentration in the atmosphere will shorten the growing teams. In addition, two country teams also analyzed other period of maize, which will result in decreased crop water crops grown in the districts. The regional analysis shows requirement and use. that the actual yield of the different crops--specifically The authors recommend extending this type of analysis of maize and groundnuts--improves with an increase to other crops as well as to other countries to develop a in actual evapotranspiration, although the gap remains clearer picture of the changing pattern in crop water use wide between actual and potential yield and actual and of the major crops grown in the project countries. This paper--a product of the Sustainable Rural and Urban DevelopmentTeam, Development Research Group--is part of a larger effort in the group to mainstream climate change research. Copies of the paper are available free from the World Bank, 1818 H Street NW, Washington, DC 20433. Please contact Pauline Kokila, room MC3-446, telephone 202-473- 3716, fax 202-522-1151, email address pkokila@worldbank.org. Policy Research Working Papers are also posted on the Web at http://econ.worldbank.org. The authors may be contacted at giovanni.munoz@fao.org. July 2007. (56 pages) The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. Produced by the Research Support Team ACTUAL CROP WATER USE IN PROJECT COUNTRIES: A SYNTHESIS AT THE REGIONAL LEVEL1 Robina Wahaj2, Florent Maraux3 and Giovanni Munoz1 1An earlier version of this Working Paper was published as CEEPA Discussion Paper number 38. 2Land and Water Development Division of FAO (Food and Agriculture Organization of the United Nations). 3Visiting Scientist from CIRAD (Centre de coopération internationale en recherche agronomique pour le développement) to FAO. The views expressed are the authors' alone. This paper was funded by the GEF and the World Bank. It is part of a larger study on the effect of climate change on agriculture in Africa, managed by the World Bank and coordinated by the Centre for Environmental Economics and Policy in Africa (CEEPA), University of Pretoria, South Africa. SUMMARY This report aims to synthesize the results of a crop water use study conducted by country teams of the GEF/World Bank project Regional Climate, Water and Agriculture: Impacts on and Adaptation of Agro-ecological Systems in Africa. It also presents the results of the second phase of the study based on climate change scenarios, conducted by the South Africa country team. The actual evapotranspiration of five commonly grown crops ­ maize, millet, sorghum, groundnuts and beans ­ in two selected districts were analyzed by six country teams. In addition, two country teams also analyzed other crops grown in the districts. The regional analysis shows that the actual yield of the different crops ­ specifically of maize and groundnuts ­ improves with an increase in actual evapotranspiration, although the gap remains wide between actual and potential yield and actual and maximum evapotranspiration, especially for the rainfed crops. This highlights the importance of improved water management if agriculture is to play an important role as a source of food security and better livelihoods. In general, the study results give realistic evapotranspiration and actual yield values for maize, sorghum, millet, beans and groundnuts. The average values for crop water productivity for these crops are within the common published ranges, with maize and sorghum being the most water efficient crops in terms of water use. It is important, however, to highlight the vulnerability of maize to water stress and the increased risks to the viability of rainfed farming systems based on this crop. The first phase of the study provided a framework for the analysis of future crop water use as affected by climate change in Africa. The second phase of the analysis, that includes climate change impact on crop water use, was conducted by the South Africa country team. This analysis was performed for maize, using the methodology developed by the FAO (Food and Agriculture Organization) that is used together with CROPWAT to assess future crop water requirement and use. The results of the second phase of analysis show that a 2°C increase in the temperature and a doubling of CO2 concentration in the atmosphere will shorten the growing period of maize, which will result in decreased crop water requirement and use. It is recommended that this analysis is extended to the other crops as well as to other countries to be able to get a clearer picture of the changing pattern in crop water use of the major crops grown in the project countries. 2 TABLE OF CONTENTS Section Page 1 Introduction 4 2 Methodology 5 3 Actual crop water use 7 4 Crop water use in the project countries 9 5 Water efficient crops: Example from Egypt and South Africa 11 6 Conclusion and recommendations 12 References 15 Appendix 1: Using CROPWAT to study crop water requirement as 35 affected by climate change Appendix 2: Application of draft methodology for `using CROPWAT 50 to study crop water requirement as affected by climate change' to maize crop in South Africa 3 1. Introduction The GEF/World Bank project Regional Climate, Water and Agriculture: Impacts on and Adaptation of Agro-ecological Systems in Africa seeks to investigate the effects of climate change on different agro-ecosystems in Africa. This study is one of the first analyses of climate change impacts on agriculture and how it adapts in Africa. Although there have been some studies of climate impacts in the continent, it is still unclear how Africa will be affected by climate change and how its agriculture will adapt. This study aims to provide holistic empirical evidence incorporating three main approaches ­ crop response simulation modeling, hydrological modeling and economic analysis ­ to fully understand the role that climate plays in Africa today and how that might change with global warming. There is enough scientific evidence to show that any significant change in climate on a global scale will affect local agriculture and therefore the world's food supply. In several geographical areas there has been a considerable number of studies on how and up to which level future climate changes will affect agricultural production. These kinds of studies are subject to all the complexities that characterize natural agro-ecosystems, so no single approach is valid in all circumstances. One of the most common approaches in climate change impact studies is to use agro-environmental simulation models. Several of these have been developed to study specific aspects of this impact, and some studies have explored possible future scenarios combining the use of different existing models. For the agricultural sector the complexity comes from the confrontation between normal climate variability and climate changes and the dynamics of farming systems. Worldwide farming systems represent the responses to constraints and opportunities from the surrounding environmental and socio-economic conditions. A change in these constraints and opportunities generates changes in the farming systems. Therefore, in agricultural production, analyzing the impacts of climate change must be done considering possible changes or adaptation to farming systems. The project, in its launch workshop in December 2003, adopted the Ricardian approach4 to assess the economic impact of climate change on African agriculture. To further increase the understanding of this impact it was recommended to initiate parallel analysis in crop simulation and river basin hydrological modeling. Crop modeling offers a powerful tool for studying in a very short period of time the possible consequences of changing conditions. Thus, modeling is commonly used to analyze the possible effects climate change will have on agriculture. These impacts are manifold and the interaction among different parameters is complex to analyze and not always fully understood; modeling may often be considered an oversimplification. However, it is one of the best existing tools as it allows users to test many scenarios in a short period of time and if the limitations of the approach are properly recognized the results may provide adequate insights into the climate change impacts on agriculture. 4 The Ricardian approach uses statistical analysis of data across geographic areas to separate climate from other factors (such as soil quality) that explain production differences across regions, and uses the estimated statistical relationships to assess impacts of climate change. The approach assumes farmers optimize their farming systems. 4 The FAO (Food and Agriculture Organization) was requested to help the national teams develop a unified approach in crop simulation modeling. Provision of training was envisaged as an FAO input. A first training workshop on Crop response simulation and river basin hydrology modeling was held in June 2003 in Ghana. The country teams used the CROPWAT program to assess potential and actual crop water use of the selected crops in the selected districts. The project countries foreseen to participate in the study and in the regional analysis are presented in Table 1. The study was conducted in two phases. In the first phase a minimum of five crops ­ beans, groundnuts, maize, millet and sorghum ­ were selected for two districts in each country for the analysis on present crop water use. In the second phase, the countries would use climate change scenarios to forecast the future water requirement of these crops. For this second phase a particular methodology was developed to take into account changes in both temperature and CO2 concentrations for the calculation of crop water use under the climate change scenarios with the use of CROPWAT. This methodology is presented in Appendix 1. It has already been used by the South Africa country team and the results were presented in the project workshop held in Spain in December 2004. These results are presented in Appendix 2 of this report. This report presents the final results of the regional synthesis of the first phase of the study that analyzes the water use of the selected crops in the project countries, and the results of the second phase of the study based on climate change scenarios conducted by the South Africa country team. 2. Methodology The project countries have diverse farming systems covering a wide range of agro-climatic zones. The national teams selected at least two districts (see Figure 1) each representing two different agro-ecozones of their respective countries with varying climatic conditions, cropping and farming systems. Figure 2 shows the broad farming systems in Africa5 as defined by the FAO and World Bank (2001), and Table 2 summarizes the information about the main features of the selected districts in the project countries. The following are the main characteristics of the farming systems in the project countries: · Agro-pastoral millet/sorghum: Rainfed sorghum and pearl millet are the main sources of food of this farming system and are rarely marketed, whereas sesame and pulses are sometimes sold. Land preparation is by oxen or camel, while hoe cultivation is common along river banks. Livestock are kept for subsistence and transportation. The main factor for vulnerability is drought, leading to crop failure, weak animals and distress sale of assets. 5A farming system is defined as a population of individual farm systems that have broadly similar resource bases, enterprise patterns, household livelihoods and constraints, and for which similar development strategies and interventions would be appropriate (FAO and World Bank, 2001). Although the country teams did not use this classification of farming systems, they used their national classification of both agro-ecozones and farming systems. 5 · Irrigated: This system is characterized by large and small irrigation schemes with high population density and small farm size. Crop failure is generally not a problem but livelihoods are vulnerable to water shortages. · Cereal root cropped mixed: Here cereals such as maize, millet and sorghum are widespread. Intercropping is common. This farming system is found predominantly in the dry sub-humid zone. Livestock is abundant. The main factor for vulnerability is drought. The agriculture growth potential is high. · Forest based: In this system farmers clear new fields from the forest every year, thus practice shifting cultivation. These fields are cultivated for two to five years ­ first cereal or groundnuts and then cassava ­ and then abandoned to bush fallow for seven to 20 years. However, because of increased population density the fallow periods are progressively being reduced. Physical isolation and lack of roads and markets pose serious problems. · Highland perennial: This farming system is based on perennial crops such as bananas and coffee and is complemented by cassava, sweet potatoes, beans and cereals. Land use is intense and holdings are very small: more than 50% of the landholdings are smaller than half a hectare. · Highland temperate mixed: In this system small grains such as wheat and barley are the main staple, complemented by peas, lentils, broad beans and Irish potatoes. Cattle are used for plowing. The major factor for vulnerability is climate: early and late frosts at high altitude cause a decrease in yield and crop failures are not uncommon in cold and wet years. · Large commercial and smallholder: This farming system lies mainly in the semi-arid and dry sub-humid zones of South Africa and Namibia. It comprises two distinctive types of farms: scattered smallholder farms and large commercial farms. Both types are largely mixed cereals-livestock systems. Small farmers often survive by means of off-farm income from employment. Poor soils and drought are the sources of vulnerability. · Maize mixed: This farming system lies mainly in East and southern Africa with altitudes from 800 to 1500 meters. It also contains scattered small-scale irrigation schemes. The climate varies from dry sub-humid to moist sub-humid. The population density is high and farm sizes are small. The main staple is maize and the main sources of cash are food crops such as maize and pulses, tobacco and coffee; livestock such as cattle and small ruminants; and migrant remittances. Socio-economic differentiation is considerable, due mainly to irrigation. · Pastoral: This system is based mainly on sheep, goats and camels and is located largely in the arid and semi-arid zones. These zones are sparsely populated, with more densely populated areas around irrigation settlements. Socio-economic differences are considerable. The main source of vulnerability is great climatic variability and consequently a high incidence of drought. 6 Among the project areas selected to be studied within the countries, Egypt presents the driest environment (with rainfall less than 100mm per year), whereas the wettest locations are in the Bobo Dioulasso district of Burkina Faso (with an average rainfall of about 1000mm per year). In Egypt, unlike the other countries participating in the study, agricultural production is totally dependent on irrigation. There are also large differences in the climatic conditions and agro- ecozones within some countries, such as Ethiopia, Senegal and Zambia. The majority of the project countries have farming systems with subsistence farmers practicing manual farming using family labor for crop production. In Egypt, South Africa, and the Chipata district in Zambia,6 large farmers practice commercial, mechanized and intensive farming. Maize, millet, groundnuts, sorghum and beans are the main crops grown in the project countries, especially where rainfed agriculture is practiced. In Egypt the main crops are cotton, wheat, maize and citrus fruits. Commercial farmers in South Africa grow fruit such as apples and pears. For this study, the country teams chose the following five crops for the analysis: maize, millet, sorghum, groundnuts and beans. The actual water use of these crops was assessed using the FAO methodology outlined in FAO Irrigation and Drainage (I&D) Papers Nos. 33 and 56 (FAO 1979, 1998) and the CROPWAT program. CROPWAT is a decision support system developed by the Land and Water Development Division of FAO. Its main functions are to calculate reference evapotranspiration, crop water requirements and crop irrigation requirements in order to develop irrigation schedules under various management conditions and scheme water supply and to evaluate rainfed production, drought effects and efficiency of irrigation practices.7 CROPWAT can be used in combination with the CLIMWAT database,8 which includes monthly average data from a total of 3262 meteorological stations from 144 countries, including all the project countries of this study. Thirty years' average climate data from 1961­1991 were used for assessing potential crop evapotranspiration by all the project countries except Burkina Faso and Ethiopia. Some country teams collected the climate data directly from their respective meteorological stations, others used the CLIMWAT database.9 The country teams also collected production data for a number of years. Table 3 shows the years and sources of data used for the analysis. 3. Actual crop water use The methodology used in this study to assess actual crop water use is outlined in FAO I&D 56 (1998) which draws heavily on the procedures for predicting yields when all the climate, soil and 6 In other provinces/districts of Zambia, agriculture production is mainly rainfed, thus yield varies with varying rainfall. 7http://www.fao.org/ag/AGL/aglw/cropwat.stm 8 CLIMWAT is published as Irrigation and Drainage paper No. 49 (FAO 1994) and includes a manual which describes the use of the database with CROPWAT. 9During the training workshop in Ghana in 2003, and later in Egypt in 2004, it was decided that 30 years' average climate data would be used for the analysis. It was also decided that this data would be extracted from FAO's CLIMWAT database. 7 crop parameters are known, as described in FAO I&D 33 (1979) Yield response to water. In fact, this approach is the inverse procedure of the more widely known and used one developed in the same publication, which aims to predict crop yield based on the actual crop water use (ETa) and maximum crop water requirement (ETc). This approach proposes to estimate actual evapotranspiration (or actual crop water use), after having estimated the stress factor from the ratio of actual to potential yield. The complete procedure is explained in the following paragraphs. Crop evapotranspiration or crop water use can be assessed by multiplying the reference evapotranspiration (ETo) by the crop coefficient (Kc) (see Equation 1). The reference evapotranspiration is calculated by using climatic data, with the Penman-Monteith equation: ETc = Kc× ETo (1) where ETc = Crop evapotranspiration ETo = Reference evapotranspiration Kc = Crop coefficient. The ETc calculated through Equation (1) is the evapotranspiration from crops grown under optimal management and environmental conditions, with good water availability and no limitations of any other input. The crop evapotranspiration, also know as actual crop water use, in this report is calculated by using a water stress coefficient Ks and/or by adjusting Kc for all kinds of other stresses and environmental constraints. In this report the actual crop water use is calculated by using the following formulae: ETcactual = Ks× ETc (2) Where ETcactual = Actual crop evapotranspiration Ks = Water stress coefficient. 8 FAO Irrigation and Drainage Paper No 33 (FAO 1979) proposes to assess the yield response factor using the following equation: Ya 1-Ym = Ky1- ETa ETm (3) Combining Equations 2 and 3 and solving for the stress factor: Ks = 1- 1 Ya Ky 1 - Ym (4) where Ya = actual yield Ym = maximum/potential yield10 Ky = yield response factor. Ky describes the reduction in relative yield according to the reduction in ETm caused by soil water shortage ETa = ETc actual = actual crop evapotranspiration ETm = maximum/potential evapotranspiration. ETm = ETc Ks = Water stress coefficient. 4. Crop water use in the project countries The FAO methodology was used by the project countries to assess the actual crop water use of the selected crops in the different districts.11 As expected, the results show that the potential crop water requirement (ETm) is highest in arid and semi-arid areas, and among crops maize has, in general, the highest potential crop water requirement. It is important to note here that the crop water use of the irrigated millet and sorghum is higher than the crop water use of rainfed maize, 10Maximum yield of a crop (Ym) as defined in FAO (1979) is `the harvested yield of a high-producing variety, well-adapted to the given growing environment, including the time available to reach maturity, under conditions where water, nutrients and pests and diseases do not limit yield. Information on yields indicates the maximum yield that is obtained under actual farming conditions, with a high level of crop and water management'. 11 In case of Egypt, it was also possible to analyze crop water use of other additional crops because of the information provided in the report. The South Africa team has analyzed other additional crops as well. 9 which is also logical as there is more water available to evapotranspire to the irrigated crops. See Table 4 for results reported by the country teams. However, the crop water use of some of the crops seems unrealistic. For example, actual evapotranspiration of sorghum in the Miesso district in Ethiopia and of all the crops in the Aguie district in Niger seems to be underestimated for their actual yield. These discrepancies probably stem from the fact that the average yield data are not used for the same years as those of climate data. Moreover, water is only one of the several factors affecting actual yield and low figures for yield do not always explain the reasons behind these values. The FAO methodology uses actual and potential yields for assessing crop water use, and any inconsistency in these two yields will result in misleading values of actual crop water use. For consistency, such results were not included in the regional analysis. 4.1 Actual yield versus actual crop evapotranspiration Crop water use, in general, is directly related to the yield of the crop, if all the other factors remain constant. Although the analysis in this report is based on the actual data collected from different sources where many factors affected the crop yield, the linear correlation between the actual water used by maize and actual yield is very strong (see Figure 3).12 The actual yield of maize increases 13 times with a four times increase in the water use. The trend shown in Figure 3a matches a similar trend presented in a recent study by Zwart and Bastiaanssen (2004) in which the published data from ten countries with a big sample size (n = 233) was used for analyses. In the case of rainfed maize, actual yield increases from 0.8 tons/ha when the actual crop water use is 212mm to 2.26 tons/ha when the actual crop water use is 381mm.13 These values are realistic according to the published values for water use efficiency. Among the selected crops, maize is the least drought resistant and needs water particularly at the flowering stage. The figure also shows that there would not be any more maize yield below 150mm of evapotranspiration. As expected, farmers prefer to grow drought resistant crops (sorghum, pearl millet, groundnuts, etc.) in the dry Sahelian areas (see Table 2). Growing maize in wet areas, or with timely irrigation, can therefore improve the harvest tremendously. This trend is also apparent for groundnuts (see Figure 3b), although the number of samples is too small to draw any conclusion. 4.2 Actual evapotranspiration of maize crop Maize is by far the most common crop grown in all the selected districts. It has replaced some other, better adapted, crops such as sorghum and millet, which over time may ensure better food security. It is grown in climates ranging from temperate to tropical and tolerates high temperatures. The ideal temperature for maize ranges from 15°C (frost free) to 45°C. The actual evapotranspiration of maize for selected districts of the project countries is shown in Figure 4. The results from the country teams show that, in general, the actual water use by maize 12This curve should, in fact be, polynomial. However, since neither crop water use, nor actual yield is at its maximum in the selected districts, only the initial slope of the polynomial curve is apparent in the results. 13The actual crop water use was assessed following the procedures described in the methodology section of this report. 10 is rather low, which does not allow it to produce a good yield. Maize (medium maturity) requires between 500 and 800 mm of water (depending on the climate) to give the maximum yield (FAO 1979). In the selected districts, crop water use by maize is highest for irrigated crops (in Egypt) followed by the crops cultivated in the sub-humid agro-climatic zones of Burkina Faso, Ethiopia and Senegal ­ in these countries maize is grown mainly by farmers in the southern parts, which are more humid. Evapotranspiration by rainfed maize is least in semi-arid and dry zones because of erratic rainfall, leading to low yields. The crop water productivity14 of maize is also high compared with other crops grown in the same districts. The average crop water productivity of maize in the districts studied in this project is 0.8 kg/m3, which is in line with the range 0.8 to 1.6 kg/m3 published by the FAO (1979). 4.3 Actual evapotranspiration of other crops Millet, groundnuts, sorghum and beans are the other crops selected by the project country teams for the analysis. Figure 5 shows the actual evapotranspiration of these crops in the selected districts. Millet is the only crop that is not irrigated in any of these districts. It is drought resistant and therefore can survive long dry spells, which might explain the reason for there being no particular pattern of crop water use related to agro-climatic zones. Moreover, a complementary explication is that a farmer, in general, has several soil and crop management options for dealing with dry conditions. He can manage the crop density, or the sowing date, or the fertilization strategy. All these options, except the sowing date, are not taken into account by CROPWAT, and evidently cannot be presented here as explanatory factors. The actual evapotranspiration of sorghum increases from semi-arid to sub-humid agro-climatic zones (see Figure 5), and in districts with no irrigation it also increases with the increase in total yearly rainfall. Sorghum is one of the most drought resistant crops and is extensively grown under rainfed conditions as a food crop and also as fodder. An attempt is made (see Figure 6) to show the relationship between average rainfall over a year and crop water use. Considering the problem of the inconsistency of the real time field data, this relationship seems to be very good and shows the vulnerability of a crop dependent on variable rainfall. Crop water use by sorghum is highest in the Giza district of Egypt as the crop is irrigated there. If we compare the crop water requirement with actual crop water use for rainfed sorghum (see Table 4), it is in line with the rainfall pattern as well. Sorghum and millet are two important food crops after maize in the project countries. Because they are well suited to the dry conditions of semi-arid zones, they are part of farmers' strategy for guaranteeing food production and ensuring that the risk of failure in dry spells will be low. Groundnuts are also frequently grown under rainfed conditions, however, this crop is less drought resistant than sorghum. The results from the selected districts show that the yield in the Giza district (Egypt) is relatively good (3 tons/ha) for the amount of water the crop consumes (455mm) as compared to its water requirements under optimal conditions (720mm) The reason for this is that irrigation makes it possible to provide water to the crop at the critical growth stages. Although the gap between the crop water requirement and the water actually used in the 14Crop water productivity is the amount of water required per unit of yield. This productivity will vary greatly according to the crop species. 11 Diourbel district of Senegal is close to that of the Giza district, the yield is not as good (0.7 tons/ha in Diourbel) because the crop is rainfed and the flexibility to supply water at the critical growth stage does not exist. Beans grow well in areas with medium rainfall, but the crop is not suited to the humid, wet tropics. Moreover, its ideal minimum and maximum temperatures for growth are 10°C and 27°C (FAO 1979). The results for the selected districts, presented in Figure 5 and Table 4, show that the actual crop water use is rather low compared with the crop water requirement calculated with CROPWAT (see Table 4), even when the crop is irrigated. The water requirement for maximum production of a 60 to 120 day crop varies between 300mm and 500mm (FAO 1979). The maximum crop water use in the selected districts is less than 200mm, which is consistent with the fact that short cycle varieties have been used. Average crop water productivity values for sorghum, beans and millet are in line with the ranges published in FAO (1979) and elsewhere. These values are slightly lower for groundnuts (see Table 5). Although the actual reasons for these low figures are not known they are not necessarily due to water use. 5. Water efficient crops: Examples from Egypt and South Africa This section of the report will briefly touch on water use of various crops within the selected districts. Because of the availability of relevant information for the analysis, the districts selected for this section are from Egypt and South Africa. South Africa also reported water withdrawal for different crops, making it easier to conduct this analysis. 5.1 Water use by crops in Egypt The selected districts/governorates in Egypt are mainly irrigated, thus making it possible to grow two crops a year and also perennial crops. Crop diversity is high: the country team reported about 25 crops grown in the two districts, including summer crops, winter crops, industrial crops and food crops. An attempt is made here to show the percentage of area cultivated under different crops and the percentage of water use (see Figure 7). The figure shows that cotton and tomatoes use a relatively high percentage of water compared with the area cultivated. The intensive use of water resources for these crops is justified as they are cash crops and can give a high return. Unfortunately, data on water diversion to different crops was not available; therefore it is not possible to compare water diversion to water use, but as these crops are mainly irrigated it can be assumed that actual water use is almost completely dependent on irrigation. 5.2 Water use by crops in South Africa A wide range of crops ­ both irrigated and rainfed ­ are grown in the two selected districts in South Africa. The high altitude of the Caledon district makes it possible to grow deciduous fruit, such as apples, pears, peaches and wine grapes. In the dryland hilly areas of this district wheat and barley are the main crops grown. Maize is the main crop grown in the Kroonstad district, followed by winter wheat. Figure 8 presents the crop water use of the irrigated crops, water withdrawal for these crops, and their potential and actual yield in the two districts in South 12 Africa. Apples and pumpkins in the Caledon district and sorghum and groundnuts in the Kroonstad district seem to have smallest gap between potential and actual yield, while maize followed by wheat in the Kroonstad district and lucerne followed by dry beans in Caledon have the largest gap between potential and actual yield. These differences between the potential and actual yield may be due to the way the potential yield was assessed, as it is a debatable criterion and depends on the weight given to non-water factors (such as average fertilization, average diseases management, etc.). According to the criteria used by the evaluator, the potential yield vary considerably in a region under study. It is interesting to note that sorghum in Kroonstad and pumpkins and apples in Caledon have only a reduced gap between their actual water use and water requirement (ETa/ETc). However, the gap between the actual water use and the water requirement of maize in Kroonstad is the second smallest after sorghum (0.29 for maize and 0.31 for sorghum); it is not reflected in the ratio of actual to potential yield of maize in this district. Water withdrawal15 as compared to crop water requirement is highest for dry beans and potatoes in the Caledon district and wheat and sunflower in the Kroonstad district. This ratio is lowest for wine grapes in Caledon and groundnuts and maize in Kroonstad. There seems to be a relatively strong relationship between the ratio of water use to crop water requirement and the ratio of actual to potential yield of different crops in the Caledon district (see Figure 9). 6. Conclusion and recommendations This report aims at synthesizing the results of the crop water use study conducted by the country teams of the GEF/World Bank project Regional Climate, Water and Agriculture: Impacts on and Adaptation of Agro-ecological Systems in Africa. Present crop water use of five commonly grown crops, including maize, millet, sorghum, groundnuts and beans in two selected districts, were analyzed by six country teams. In addition, two country teams also analyzed other crops grown in some districts. The analysis shows that the actual yield of the different crops ­ specifically of maize and groundnuts ­ improves with an increase in actual evapotranspiration, although the gap remains wide between the actual and potential yield and actual and maximum evapotranspiration, especially for rainfed crops. In case of irrigated crops, the yields are better even when the crop water use is relatively low as compared to their respective water requirement as a result of flexibility in water supply at the critical growth stages of the crops. Rainfed maize and sorghum seem to be performing better in terms of crop water use in the sub-humid climate as compared to semi-arid Sahelian climatic conditions due to better rainfall. This corroborates the well-known fact that water is among the main limiting factors in several African farming systems and therefore irrigation could play an ancillary role in agricultural development. In general, study results give realistic values for maize, sorghum, millet, beans and groundnuts. evapotranspiration and actual yield. The average values for Crop Water Productivity (CWP) for these crops are within the common published ranges. Maize and sorghum appear to be the most 15Note that no irrigation efficiencies have been accounted for in this analysis. 13 water efficient crops grown in the districts. Maize, however, is the crop that is the most sensitive to water stress among the crops studied and should therefore be grown only where good availability of water can be guaranteed. It should therefore be grown under irrigation or only in rainfed areas where rainfall is reliable and the crop needs can be properly satisfied. This unfortunately is not the case in most of the districts studied. If information about the current low reliability of rainfall patterns is combined with recent studies on possible changes in surface water availability, the interest in increasing the area under irrigation increases. In fact, de Wit and Stankiewicz (2006) predict that by the end of this century 25% of Africa will have reduced surface flows owing to diminishing rainfall. It could be concluded, on the basis of results obtained, that this study provides a realistic and representative sample of African conditions, regarding: · the diversity of agroclimatic and environmental conditions and the dominant cropping systems, and · crops and their present water requirements, actual water use and expected yields. The results of the first phase of study may be used as an initial picture, to which climate change scenarios can be applied. The estimated present crop water use of the main crops in the project countries provides the basis for the framework for simulating climate change scenarios and their impact on crop water requirement. For this purpose, the FAO has developed a draft methodology (see Appendix 1) that would allow CROPWAT to be used to analyze the effect of climate change on crop water requirement. The analysis proposed in the draft methodology follows three steps: 1. Assessing change in the duration of different growth stages as affected by increased or decreased level of temperature and CO2 concentration in the atmosphere. 2. Calculating crop water requirement by using the projected climate data and the new growth stages from step 1 in CROPWAT. 3. Recalculating the actual crop water use as described in the methodology section of this report. This draft methodology has been used by the South African country team to calculate the impact of climate change on the crop water requirement of maize for three districts ­ Lichtenburg, Kroonstad and Middelburg. The results of this analysis are presented in Appendix 2. In South Africa, crop water use by maize as a result of elevated CO2 concentration in the atmosphere and an increase in temperature of 2°C will shorten the total growing period by at least 16 days, and consequently crop water requirement will be decreased by 10%. It is estimated that actual crop water use will be reduced by 4% on average. It is recommended here that other country teams also conduct this analysis with climate change data. Moreover, it will be useful to carry out a similar analysis for other major crops. 14 REFERENCES De Wit M & Stankiewicz J, 2006. Changes in surface water supply across Africa with predicted climate change. Science 31 311(5769): 5769 http://www.sciencemag.org/cgi/content/abstract/311/5769/1917 Diop M, 2006. Simulating impacts of climate variations on crop water use productivity in Senegal. CEEPA Discussion Paper No. 34, Centre for Environmental Economics and Policy in Africa, University of Pretoria. Diouf O, 2000. Réponses agrophysiologiques du mil (Pennisetum Glaucum (L.) R. BR.) à la sécheresse: Influence de la nutrition azotée. Thèse présentèe en vue de l' obtention du grade de docteur en Sciences. (PhD thesis.) Faculté des Sciences. Laboratiore de Physiologie et d'Agrotechnologies Végétales. Université Libre de Bruxelles. Durand W, 2006. Assessing the impact of climate change on crop water use in South Africa using CROPWAT. CEEPA Discussion Paper No. 28, Centre for Environmental Economics and Policy in Africa, University of Pretoria. Eid H, El-Marsafawy S & Ouda S, 2006. Using CROPWAT to analyse climate change impacts on crop water use in Egypt. CEEPA Discussion Paper No. 29, Centre for Environmental Economics and Policy in Africa, University of Pretoria. FAO (Food and Agriculture Organization), 1979. Yield response to water. Authors, Doorenbos J & Kassam AH. Irrigation and Drainage Paper 33. Rome, Italy. FAO (Food and Agriculture Organization), 1994. CLIMWAT for CROPWAT. Author, Smith M. Irrigation and Drainage Paper 49. Rome, Italy. FAO (Food and Agriculture Organization), 1998. Crop evapotranspiration : Guidelines for computing crop water requirements. Authors, Allen RG, Pereira LS, Raes D & Smith M. Irrigation and Drainage Paper 56. Rome, Italy. FAO & World Bank, 2001. Farming Systems and poverty: Improving farmers' livelihoods in a changing world. Authors, Dixon J & Gibbon D. Rome, Italy, and Washington DC. Giogris K & Tarekegen D, 2006. Climate change impacts on crop yield and crop water use in Ethiopia. CEEPA Discussion Paper No. 31, Centre for Environmental Economics and Policy in Africa, University of Pretoria. Moussa K M & Amadou M, 2006. Use of CROPWAT model to predict SMD and crops yield with climate change. CEEPA Discussion Paper No. 32, Centre for Environmental Economics and Policy in Africa, University of Pretoria. Some L, Dembele Y, Ouedraogo M, Some M B, Kambire L F & Sangare S, 2006. Analysis of crop water use and soil water balance in Burkina Faso using CROPWAT. CEEPA 15 Discussion Paper No. 36, Centre for Environmental Economics and Policy in Africa, University of Pretoria. Tamala KT, 2004. CROPWAT exercise report for Zambia, unpublished project report. Zwart SJ & Bastiaanssen WBM, 2004. Review of measured crop water productivity values for irrigated wheat, rice, cotton, and maize. Agricultural Water Management 69: 115­133. 16 Table 1: Project countries and selected districts for CROPWAT study Country Area (km2) No of districts Nos. of districts selected by country teams* Burkina Faso 273,719 301 4 Cameroon 466,307 10 3 Egypt 982,910 27 2 Ethiopia 1,132,328 65 2 Ghana 239,981 110 - Kenya 584,429 48 6 Niger 1,186,021 36 2 Senegal 196,911 320 2 South Africa 1,221,943 372 4 Zambia 754,773 72 2 Zimbabwe 390,804 60 - * Where no number is shown the CROPWAT study has not yet been conducted. 17 s d d d emst rldo 1) alr sy W 200 ppeorc ppeorc eratep ppeorc tem ngi and nk, ootr ootr sed ba ootr pasto, mraF O xedim xedim xedim Ba (FA realeC xedim realeC xedim Forest Highland xedim realeC xedim rigatedIr rigatedIr zeia zeia zeia Pastoral M M M alu alu alu alu alu ev ev alu alu al/u alu opping cr m anm/ anm/ anm/ anm/ anm/ / ers ers ers ers ers ers Intensi/ / ers Intensi/ anm/ anm/ anm/ anm/ syste farmlla evis ers ers ers ers ten farmlla evis ten farmlla evis ten farmlla evis ten farmlla eivs farm farm evis ten lla lla farmlla eivs ten farmlla evis ten farmlla eivs farmlla ten Dominant Sm Ex/ Sm Ex/ Sm Ex/ Sm Ex/ Sm In/ Sm mechanized Sm mechanized Sm In/ Sm Ex/ ten Sm In Sm Ex/ ops cr ,etllim, ont , ,s , cot nutdn an t, not , es, na wheat, rghum ou ba citrus ghum sn sn s,n tato beans, rice, so gr s ,s ea,t,m s, mille cot, e,ziam, wheat, sor bea hsirI, ea,t s, po Major humg pea yam esbla onion, Sor maize, Millet, ze,iam co pall oeat humg cow Co cassava, Oi pot veget Maize, sor wheat,nottoC maize Maize, es,toamto beatoc humg beatoc ze,ia nas oeat toes,a ey Maize, hari Sor hari M bana pot coffee Maize, tom peas, barl n countries fed/ de / Rain Irrigated fedniaR fed fed Rain Rain fedniaR cotto nfed/iaR def rig.Ir gatriIr gatedriIr Rain fedniaR fedniaR fed 18 rojectp Rain rigatedIr ll the infa 00 00 0 0 00 mm ­00 of 10 008 95 60 20 058 50 006 Ra 15 0004 25 11 emts th gni sy of )sy 0 15 021 - - Leng ing grow period (da farm ne -o zo di ) nna di the ict va Agr hum elian di nna -aridi hum on b- clima Su yrD sahg idmuh- b- (Sah Hum Hi Sahel sava Desert Desert Sub Sem Su Semi-arid iont rma ossaluo mar hk hei Gou info Districts/ Province DioboB N'a bam Fad Am ndaemaB uao -Selrf a bu Gar Kha Giza amdA iaip essoi am M Ki Laik Basic 2: y untr ain on rk erom a Table Co Bu Faso Ca Egypt Ethiopia Keny xed,im zeia alro M past alu anm/ ers farmlla evis ten Sm Ex/ s,n woc,s ttonoc, bea pea on coffee Maize, gepi peas, ral ltu fed/niaR 19 rig.Ir rticuoh crops 00 10 d -arii Sem ni kuea M s ,al m et ) ghr ghr ghr ial root der 0012 land - - - Farming syste xan alro alro alro (Di ghiH rennep Maize xedim Maize xedim ro let/so ro let/so ro let/so Ag past mil um Ag past mil um Ag past mil um Cereal cropped xedim mmercial hollal Large co and sm / / / / / / / m ers ers ers ers ers ers ers /s / farm eivs Dominant cropping syste lla /lau ev farm lla /lau ev farm lla /lau ev farm lla /lau ev farm lla /lau ev farmlla /lau ev farmlla /lau ev rmfa ten Sm man Extensi Sm man Extensi Sm man Extensi Sm man Extensi Sm man Extensi Sm man Extensi Sm man Extensi Large mechanized In , , ce,ir, ,m ts ts , ops , nu nu cr bacco,ot , rcane t, icus stu stu apples ga und und sava su, ghum asep ndn ndn cap e rghum gro etllim, rley, Major ffee,oc eziam, cas mille sor gro humg humg coffee Tea, beans, sor Maize, sor tea, Maize, cow,snaeb kal es,toamto so ourg ourg ba ea, ea, ons,lem wp humg wp Millet, co Sor co Millet, Millet, Wheat, pears ral fed/ fed ltu fed Rain Irrigated Rainfed Rain fed/niaR fed rig.Ir rticuoh crops Rain fedniaR fed Rain %)2.6( dfen Rain g.riIr rai & ll infa mm 1800­ 9001 Ra 0041­0 005 0 70 005 059 0 20 70 700 45 fo do th perigni )sy 08 0 (da 11 001 051 Leng grow ict )e ait en id d ) ) ) zo ro-clima dim -arii -aridi elian elian -aridi elian idmuh- nean rraetdie Ag (Thornthw Sub-hum Hu Sem Sem (Sah Dry (Sah Sem (Sah Sub M Districts/ Province agi e uié a elbru n adl Migori Vih Kwal Ag Gay Dio Ko doelaC (continued) 2: y untr a r gal hut Table Co Keny Nige Sene So Africa der mmercial hollal Large co and sm Maize xedim Maize xedim / / / ers / ers ers rmfa eivs ten farmlal / ev rmfa eivs Large mechanized In Sm nualam ten Extensi Large mechanized/ In s s potatoe wero nfl nutdnuo wheat, su gr ze,ia ze,ia Maize, M M ) %96.0( dfen g.riIr rai & fedniaR fedniaR 0 0 0 21 55 70 90 nean di rraetdie hum b- M Dry Su da nst oo Kr wegnohC aatpihC iab Zam Table 3: Years and source of the data used in the analysis Country Data* Source climate data* Climate Production (including rainfall) Burkina Faso 1962-2003 Cameroon Egypt 1961-1990 1990-2000 Weather station Ethiopia 1991-2001 1992-2002 Weather station Kenya Niger 1961-1990 2003 National metrological service Senegal 1961-1990 South Africa 1961-1990 1993 CLIMWAT/SAPWAT Zambia 1961-1990 CLIMWAT * Where no information is given about the year and source, it was not found in the country reports or supplied otherwise. 22 p use 2 2 0 3 6 7 Cro 14 22 11 29 92 17 17 10 23 0 7 4 28 22 39 13 20 (mcm) 170. 10 water 0 3 4 2 0 7 9 3 6 9 4 1 8 5 3 1 7 7 3 7 9 7 8 0 ETa (mm) 23 26 33 21 23 32 15 15 28 54 11 52 19 45 50 38 25 19 91 13 12 10 79 13 44 12 (ha) 03,4 04,4 06,8 05,7 00,0 06,8 23,1 0 608, 36,4 0 6 98 72 002 301 007 007 009 38 00 Cropped area 61 84 32 13 40 52 32 45 591, 21 17 35 16 14 86 Ks 480. 660. 810. 410. 630. 900. 440. 440. 480. 990. 730. 770. 770. 630. 740. 720. 550. 530. 260. 400. 480. 290. 270. 410. 271. 460. )a Ym (t/h 04. 31. 90. 03. 21. 80. 2 2 2 9 4 39. 63. 4 7 53. 54. 53. 54. 09. 52. 09. 52. 09. 04. 52. )a Ya (t/h 41. 90. 70. 80. 80. 70. 60. 50. 70. 9 3 66. 62. 3 55. 32. 72. 41. 51. 32. 01. 01. 40. 42. 84. 90. Ky 251. 90. 21. 251. 90. 21. 251. 251. 251. 251. 151. 251. 151. 70. 90. 251. 90. 251. 90. 251. 151. 251. 151. 251. 70. 151. countries 0 0 0 2 5 5 7 2 6 7 7 6 8 0 6 2 9 5 3 0 4 9 2 2 4 1 23 project ETm (mm) 48 40 41 51 36 36 26 38 59 55 15 67 25 72 67 53 46 37 35 33 26 37 29 33 35 26 the in 6 7 3 3 8 2 3 0 0 8 9 5 0 5 5 3 9 6 3 2 9 5 3 5 1 4 ETo (mm) 59 55 57 63 50 52 41 57 89 67 18 82 31 83 82 56 46 42 35 49 41 56 46 49 49 41 crops lected dry- dry- sut dnn dry- dry- sut dry- se p zeia humg etllim zeia humg etllim zeia zeia zeia zeia ou humg humg humg dnn Cro M Sor Pearl M Sor Pearl M M M M eansB zeia M eansB zeia zeia zeia Gr Sor M Sor M Sor M eansB zeia M eansB zeia ou the M Gr eansB of a hk use lasso urm hei Go -Diou N' a bu ni water bo bam District Bo Fada Am ndaemaB uao -Selrf am kuea gahi Gar Kha Giza Adam Miesso Ki M Vi crop Actual y 4: Fassoain n untr rk Co Bu rooemaC a Egypt Ethiopia Keny Table 1 2 3 4 5 M) water (MC 19 5 39 21 4 27 20.0 15 463 99 023 5 26 583 3 14 952 5 23 98 440. 8 14 3 5 0 32 36 Crop use 4 1 8 9 ETa (mm) 13 94 21 21 98 12 711 7 11 053 0 14 012 7 18 382 1 18 023 3 2 7 0 6 6 28 31 70 16 22 13 27 4 4 7 (ha) 382 84 503 05 35 009 004 86,9 15,1 09,4 9 50,8 58 14 50 18 95 36 02 20 12 Cropped Area 24,401 93,0 5079, 1870, 71 30,251 2141, 62 11 14 88,531 78 ,95708 83 31 88 23 13 Ks 300. 290. 540. 610. 300. 400. 0. 84 470. 0. 37 310. 0. 85 370. 0. 36 450. 0. 37 750. 800. 190. 290. 310. 280. 470. (2006) )a Ym Diop esure (t/h 09. 52. 09. 52. 05. 09. 0.3 52. 9.0 60. 8.0 40. 8.0 21. 2.1 31. 21. 3 7 6 7 7 watp )a cro Ya (t/h 11. 50. 83. 41. 91. 32. 4.1 01. 6.0 10. 4.0 10. 5.0 70. 9.0 11. 90. 20. (2006), 790. 272. 700. 402. al. etd alutca Ky 251. 151. 251. 151. 90. Ei 251. 1 151. 2.1 151. 2.1 151. 1 70. 1 70. 251. 151. 251. 90. 251. 251. ornoi (2006), deliye ratpi 24 nsra sret 4 1 4 7 7 9 0 1 2 3 0 9 1 5 2 4 1 ETm (mm) 44 32 39 35 32 31 2 44 25 4 28 45 3 95 51 4 05 40 4 93 38 38 37 57 71 48 58 vable fficient achi coe mu potaveporc 3 9 8 4 7 6 ETo (mm) 66 50 58 49 52 47 064 7 39 866 1 55 825 9 61 383 4 39 655 9 7 0 9 4 4 0 Tarekegen stress 54 63 51 75 87 62 74 xima alu mecibucnoill and M mi = Water Act = = = dry- dry- dry- dry- dry- sut sut Ym Ks ETa mcm p zeia zeia humg zeia dnn dnn dry- Giogris, humg Cro M eansB M eansB Sor M Millet eansB Millet eansB Millet eansB ou ou zeia Millet Gr Millet Gr M eansB zeia zeia zeia M Sor M M (2006) noi ratpi iaip gorii a uie elbru n da urb ad adou rationip ranstop District Kwale M Laik Gay Ag Dio Kol doelaC nst oo ddeli Kr Lichtenbu rg M g Am rans eva and pot factor eva crop y cari ce mu onsep (continued) Moussa ren res eldiy 4: a r gal untr Afhut xima Refe M Co Keny Nige Sene So = = Yield Actual = = Table 6 7 8 Sources: ETo ETm Ky Ya Table 5: Average Crop Water Productivity (CWP) of the selected crops in the project districts Crop Average CWP (kg/m3) CWP range published in FAO (1979) and elsewhere* (kg/m3) Beans 0.44 0.30 ­ 0.60 Groundnuts 0.50 0.60 ­ 0.80 Millet 0.21 0.16 ­ 0.66 Sorghum 0.70 0.60 ­ 1.00 * The range for millet is taken from a PhD dissertation (Diouf 2000). The values are calculated on the basis of two years' experiments with different treatments, in Senegal. Ranges for rest of the crops are published in FAO (1979). 25 Figure 1: Selected districts for CROPWAT analysis in the project countries 26 Figure 2: Farming systems in Africa (Adapted from FAO and World Bank, 2001) 27 10.00 9.00 8.00 R2 = 0.84 ah/snot 7.00 6.00 ni deliy 5.00 alut 4.00 ac 3.00 2.00 1.00 0.00 0 100 200 300 400 500 600 actual crop water use in mm (a) Maize 3.0 2.5 2.0 /has ton 1.5 in eldyi 1.0 0.5 0.0 0 50 100 150 200 250 300 350 400 450 500 actual water use in mm (b) Groundnuts Figure 3: Actual crop water use versus actual yield of maize (a) and groundnut (b) crops in the selected districts in Africa 28 Figure 4: Actual crop water use by maize in the selected districts in Africa 600 Irrigated Irrigated 500 m m ni esur 400 teaw 300 porcla 200 tuca 100 0 g ur adt a m os g adl a az nbe ns essoi chtLi oorK M uroG'N asl urbel Ko ma Gi ouiD- Ad ddi a M hkehiS-la-r Fad booB afhK 29 Figure 5: Actual crop water use by millet, sorghum, groundnut and beans in the selected districts in Africa 600 500 m 400 m ni e us er 300 wat alut ac200 100 0 Aguie Diourable Kolda Fada N'Gourma Bobo-Dioulasso Gaya millet 600 Irrigated 500 m 400 m ni e us er 300 atw ualt ac 200 100 0 Miesso Kroonstad Fada N'Gourma Adama Bobo-Dioulasso Giza sorghum 30 Figure 5 (Cont'd) 500 Irrigated 450 400 mm 350 ni 300 use aterw 250 oprc 200 actual150 100 50 0 Diourable Kolda Giza groundnuts 250 Irrigated 200 m m ni e 150 usretaw oprcl 100 tua ac 50 0 Caledon Gaya Aguie Giza beans 31 300 250 200 mm in e usre 150 atw oprc100 50 0 0 200 400 600 800 1000 1200 average annual rainfall in mm Figure 6: Rainfall versus actual crop water use by sorghum in the selected districts in Africa 32 40 % area cultivated 35 % water use 30 25 20 15 10 5 0 na ye dr- age ustr n e o tto tes er at pes oes ean nea re anaB Barl ns Ci Co Da Maiz Gra yb beetr low Bea abbC eppP Whe nf Tomat Mang So ugaS garc Su Su (a) Khafr el-Sheikh district 60 50 % area cultivated % water use 40 30 20 10 0 ana rley dr- e ont s s e o n o te per at eat Ba bag surt ngo bea ane Ci Da Maiz rc rewlo Ban ans Be abC Cot aperG Ma Pep Pot oyS Wh ugaS unfS Tomat (b) Giza district 33 Figure 7: Area cultivated under different crops and crop water use in the selected districts in Egypt 34 350 8 Crop water use 300 Water withdrawal 7 Yield Aatual 6 250 re Yield Potential 5 )a metc 200 s/h ubic 4 on(t on 150 dle 3 Yi Milli 100 2 50 1 0 0 Maize Sorghum Wheat Other winter Sunflower Groundnuts Other fodder cereals crops (a) Kroonstad district 90 water use 80 80 Water withdrawal 70 Yield Actual 70 Yield Potential 60 rete 60 m 50 )ah/ 50 bicuc 40 ons(t 40 dl llioni 30 Yie M 30 20 20 10 10 0 0 ans se s s s m rne oe rew sin s rs he pes beyrD ce at flo egulr Lu ppleA ums Pea Pl aperg r tiurfs gra Pot he he auliC Pumpk eacP e bleaT Ot Win duouci Ot de 35 (b) Caledon district Figure 8: Water withdrawal, water use and yield of different crops in the two selected districts in South Africa 36 0.60 R2 = 0.73 0.50 eldyil 0.40 tia en pot/ 0.30 eldiy al tu 0.20 ac 0.10 0.00 0.00 0.10 0.20 0.30 0.40 0.50 0.60 actual water use/crop water requirement Figure 9: Water use versus yield in Caledon district in South Africa 37 APPENDIX 1: Using CROPWAT to study crop water requirement as effected by climate change Background The GEF/World Bank project "Regional climate, water and agriculture: impacts and adaptation of agro-ecological systems in Africa" seeks to investigate the effects of climate change on different agro-eco systems in Africa. The project, in its launching workshop in December 2003, adopted the Ricardian analysis to assess the economic impact of climate change on African agriculture. To further increase the understanding of the impact of climate change in agricultural production it was recommended to initiate parallel analyses in crop simulation and river basin hydrological modelling. FAO was requested to assist the national teams in the development of a unified approach in crop simulation modelling. Initially it was planned to use crop simulation models like CERES, along with CROPWAT, for the crop simulation component of the project to study impact of climate change on crop water requirement and water stress. However the country teams, during the training workshop on "crop response simulation and river basin hydrology modelling" in Ghana, in June 2003, decided to use, solely CROPWAT for assessing future crop water requirement as effected by the climate change. This decision was based on the fact that the crop growth simulation models, although are suitable for such studies, are very data intensive and require longer training as compared to the CROPWAT programme. This issue was discussed again and the decision was reinforced by the country teams during the training workshop on "quality control for country level and regional analysis and reporting" in Egypt in November 2003. Taking into account the concerns of the country teams, AGLW-FAO agreed to look into the possibility of outlining a methodology to use CROPWAT model for assessing the crop water requirement as effected by climate change. Since then AGLW-FAO had tried to come-up with a simple method to study the impact of climate change on crop water requirement. This note outlines the methodology and algorithms that can be used to assess the changes in the length of crop growth stages in response to changes in average temperature and Carbon dioxide concentration in the atmosphere. These "new" values of the length of crop growth stages can then be used with the climate data generated by Global Circulation Models (GCM) to compute "future" crop water requirements. General considerations: What can and can not be studied using CROPWAT Prospective studies on effects of Climate Change (CC) on crops, as generally considered by the International Panel on Climate Change (IPCC), may consider three different scenarios taking into account changes in (i) Climate/weather parameters (ii) climate/weather parameters plus crop physiological responses to CO2 concentration; and (iii) climate/weather parameters plus crop adaptations. 38 Crops are simultaneously affected by many climatic parameters, and as a consequence, they offer an integrated response to climate change. As there are numerous interactions between parameters, and between the effects of these parameters on plants, it may seem unrealistic to identify specific single direct effects. Moreover, it may also be considered to take into account simultaneous effects, or interactions between two climatic parameters which are linked in CC and have contradictory effects on crop growth. One such example is the impact of increased level of CO2 on crop water use: it leads to a faster establishment of LAI, then to a higher radiation interception, then to a higher potential transpiration. In the same time, higher CO2 concentration modifies stomata closure process, which finally leads to a lower transpiration. Besides, the specific effects of these phenomena may be different for C3 and C4 crops. CROPWAT is a generic tool that predicts crop water requirements, and crop water stress, and these will be evidently affected by climate change. It does not take into account dynamics of and interactions between different processes taking place simultaneously. Evidently, maintaining its initial simplified options, CROPWAT will not be able to take into account correctly above mentioned simultaneous and contradictory effects of climate change on crop growth and crop water requirement. This section proposes some indications on what aspects of climate change can reasonably be taken into account using the CROPWAT programme for their potential impact on the crop water requirements. (i) Weather parameters Purely physical phenomena (wind, radiation, rainfall, temperature and relative humidity of the air are integrated through the Penman Monteith equation for ET calculations), can be taken into account directly with CROPWAT, as input parameters. If we consider also the irrigation scheduling option of CROPWAT, rainfall pattern modifications will be taken into account directly. (ii) Weather parameters plus crop physiological responses to CO2 concentration CROPWAT includes a function for estimating water stress effects on yield, through potential yields and Ky16 coefficients (see FAO 33)17. CC may affect these functions on several aspects including the following: · Potential Yield: Potential yield is expected to be modified, as a result of combined effects of photosynthetic efficiency increase (due to CO2 concentration increase), harvest index modifications, and reduction of the length of the crop cycle (due to temperature increase). Generally, as a result of these effects, the potential yield for C3 crops is expected to increase, while for C4 it is expected to decrease. · Water Use Efficiency: Another effect is a modification in water use efficiency (WUE). Increase of CO2 concentration modifies water use efficiency and water exchanges through changes in stomatal resistance. C4 crops are more efficient than C3 in reducing transpiration with CO2 increase. Thus, resulting in changes in Ky coefficients. 16Ky is a factor that describes the reduction of relative yield according to the reduction in crop water requirement (Etc) caused by soil water shortage. Ky values are crop specific and vary over the growing season according to growth stage. 17FAO 1979. Yield Response to Water. Doorenbos, J. and Kassam, A. H. Irrigation and Drainage Paper No 33. 39 At the same time, as CC will take place gradually in the future18, it is acknowledged here that crops, varieties, cropping systems etc. will have changed a great deal. Added to the difficulty for modifying the empirical Ky/Ymax19 approach, we conclude that although it is possible to try and work out some algorithms for taking into account these effects. However, it would be a huge and time consuming task that may require longer time frame than the GEF/CEEPA Project life and is out of the strict scope and framework of CROPWAT. Therefore these algorithms will not be developed. (iii) Weather parameters plus crop adaptations CROPWAT takes into account plant physiology through an internal clock (the unit is in days), which defines the duration of phenological stages, and associates to these stages, crop coefficient values. Changes in two climatic parameters, Temperature and CO2, will modify the growth and development engines of crops through the CROPWAT internal clock in the following manner: a) Temperature: A certain amount of heat is required to move crops to the next development stage and for a given crop variety, this value is constant from year to year. Increasing temperature will affect the duration of the stages as far as less days will be necessary to reach the accumulated degree-day threshold. This phenomenon is not dependent on C3/C4 metabolism of a crop. Whereas temperature is not explicitly taken into account in crop development by CROPWAT, it is possible to take into account its variations as far as base and maximum temperatures (crop specific) are available. This can be taken into account with CROPWAT by assessing the length of new crop growth stages in a separate spreadsheet and then using these "new" values as input in CROPWAT. See suggested tables, algorithms and justifications in annex 1. b) CO2 concentration: Increasing CO2 concentration will increase the efficiency of plant biomass engine, i.e., conversion of radiation to biomass, and thus will induce faster rate of LAI20 and Kc21 before reaching maximum. This phenomenon is C3/C4 dependant and some theoretical development is proposed for taking it into account (see annex II). Varying CO2 concentration also has numerous direct and indirect effects on crop growth. Many of them are linked to temperature modifications, and earlier studies considering these effects on entire plants have some difficulties in separating their combined effects or in other words to "decouple" these effects. Thus, it is suggested here to take into account only the effect on the changes in the length of growth stages during the initial and development phases of a crop. This will be realised through a) starting the development phase earlier, thus resulting in a shorter initial stage; and b) increasing Kc rate according to LAI growth modifications, hence resulting in a steeper slope. These calculations will also be 18Climate change is a dynamic process/phenomenon and is expected to continue in the subsequent years and in future. However Climate change scenarios generally consider the time frame until 2100. 19Ymax is the maximum yield of a crop, which is defined as the harvested yield of a high producing variety, well- adapted to the given growing environment, including the time available to reach maturity, under conditions where water, nutrients and pests and diseases do not limit yield. 20LAI is not a CROPWAT variable, but Kc may be considered as statistically linked to LAI. 21Kc is a crop coefficient, and is a dimensionless number. Kc tells how much water a crop uses in comparison to the reference crop. For example, Kc = 1.1 means that at the particular crop stage under consideration, a crop uses 10% more water as does the reference crop. 40 done in a spreadsheet outside CROPWAT. The values obtained for lengths of initial and development phases will then be used as input in CROPWAT. See suggested tables, algorithms and justification in annex II The combined effect of these two parameters (a & b) is described in Annex III. NOTE: It is worth noting here that the algorithms suggested in the section will be implemented in the spreadsheet outside CROPWAT and the results/values obtained from these algorithms will be used as input in CROPWAT. 41 Annex 1: Impact of increased Temperature on the duration of crop stages The phenomenon. Crops have an internal clock based on which all the phenological stages duration are governed. Growing degree days, or GDDs, are used to estimate the development of plants during the growing season, which is closely related to the daily accumulation of heat. A certain amount of heat is required for a crop to move to the next development stage, and for a given crop variety, this value is constant from year to year. Each crop has a minimum base temperature (Tbase) or threshold below which development does not occur. To calculate GDD, the base temperature is subtracted from the mean temperature for the day to give a daily GDD. DailyGDD= Taverage -Tbase ( ) Where Taverage = (Tmax -Tmin) 2 Each daily GDD is then accumulated over the growing season. Data bases can be found in the literature, which give precise information for GDD thresholds, considering every crop/variety and every development stage. In tropical conditions, the suggested method is not so effective, because very hot temperatures do not provide efficient conditions for crop growth. Moreover, high night temperatures may have a negative effect on crop growth. However, these phenomena are not well described in the literature. Crop growth simulation models generally consider a threshold temperature (Tdmax) for every crop over which additional degrees are not effective anymore: For GDD calculation, we may consider that If Taverage > Tdmax then Taverage = Tdmax Where Tdmax is the upper threshold. The maximum average temperature (Tdmax) for the growth of most crops ranges between 25 and 40 °C. With CC and global warming, it is expected that daily GDD will increase, and thus the thresholds will be reached in less time. So, a reduction is expected in the duration of every crop growth stage. 42 CROPWAT does not calculate GDD, and does not consider inter-annual variability of stages duration. We propose here a simplified method for linking temperature increase with the reduction of stage duration. Step 1: Quantification of the effects of increasing temperature If D0 is the normal present duration of a crop growth stage (as used in CROPWAT), it is (indirectly) based on an accumulated sum of GDD (GDD)0 so that if (GDD)0 is the theoretical average daily GDD during the stage, and D0 is the initial length of the stage (in days), (GDD) = D0 ×(GDD)0 0 If we have an increase of temperature of T due to CC, we will have consequently a decrease of the number of days D0, so that : (GDD) = D0 ×(GDD)0 = (D0 - D0)×[(GDD)0 + T] 0 Where D0 is the decreasing number of days of the stage necessary to reach the total (GDD)0. So D0 ×(GDD)0 = D0 ×T - D0 ×T As D0*T has low value, we can approximate D0 as follows: D0 = D0 ×T (GDD)0 OR D0 = D0T (T ) average-Tbase Taverage during the stage may be obtained from local conditions Tbase is crop specific, and is available in the literature. D0 is given by CC scenario. For example, if we study possible effects of CC on maize cultivated in Paris, during grain filling, we will have following estimation for the grain filling stage: D0 = 80 days Taverage = 22 degrees Tbase = 6 degrees 43 T = 2 degrees D0 = 80×2 =10days 22- 6 Hence, in Crop Data window of CROPWAT, we will reduce the duration of grain filling stage from 80 to 70 days. Step 2: Taking into account high temperature Preliminary remark: Most likely, the temperature will not be highly affected by CC in the tropical regions. However, and with the preoccupation of proposing realistic algorithms, we introduce here a `security algorithm', CHT = Coefficient of high temperature, through a correction factor built as follows: Let us evaluate how far will be average temperature from Tdmax after modification by CC: Td max = T max -(Taverage+T) We consider that the lower the value of Tdmax, the less efficient will the supplemental degrees be. It will only be fully efficient if Tdmax> 5. If not, then we `'punish'' the D value calculated in step 1 if Td max> 5, CHT = 1. if 0 < Td max<5 CHT = Tdmax/5 if Td max <= 0 CHT = 0 (0 CHT 1) Finally D = D0×CHT 44 Example Following table presents some examples based on data from France and Nicaragua we propose here some realistic examples: Maize, France Onion, France Maize, Nicaragua D0 80 60 80 Taverage 22 22 27 Tbase 6 2 6 Tdmax 30 30 30 T 2 2 2 D 10 6 2 Another Example: Graphic presentation of the effect of increase in temperature by 2 °C Crop: Wheat Country: Egypt Crop Temperatures: T base 4.0 °C Tdmax 30.0 °C Initial/present values of crop parameters and temperature are given in the following table Kc values Duration (days) Taverage (°C) Kc ini 0.3 30 19 Kc dev. 65 15 Kc mid 1.15 40 14 Kc end 0.5 30 17 Total 165 With an increase in average temperature by 2 °C, the length of the growth stages will reduce and so will the total growing period as shown in the following table Kc values Duration (days) Taverage (°C) Days Kc ini 0.3 26 21 - 4 Kc dev. 53 17 -12 Kc mid 1.15 32 16 - 8 Kc end 0.5 25 19 - 5 Total 136 -29 45 Following figure presents this change in the format used in CROPWAT: Impact of temperature change on length of plant growth stages and total growing period 1.20 1.00 0.80 ues valcK 0.60 0.40 0.20 0.00 0 50 100 150 200 Days Initial values Modified values 46 Annex II: Impact of increased level CO2 concentration on the duration of initial development and mid season crop growth stages. The phenomenon: Increasing CO2 concentration will increase the efficiency of plant photosynthetic engine for biomass production. It is commonly admitted that with 100 % increase in CO2 concentration (350 ppm to 700 ppm) photosynthesis efficiency in C3 crops will increase by almost 30 %. According to the projections this level of CO2 concentration could be reached by 2100. This increased photosynthesis is expected to lead to faster growth rate which will eventually result in crop achieving maximum LAI in relatively shorter time. Based on the literature review, we assume that for C4 crops, this phenomenon is not significant. For taking into account the increased photosynthesis efficiency in C3 crops with CROPWAT we assume that the crop is grown under optimal conditions with neither water nor nitrogen stress. FAO 5622 proposes the following equation for equivalence between Kc and LAI during the first growing stage of crops: Kc = Kcmin+ (Kcmax- Kcmin)[1- exp(- 0.7× LAI)] (equation 97 in FAO 56) From this equation, we estimate that Kcmax will be reached when LAI = 3 The first stages for LAI as a function of time are described with the following logistic curve, with three parameters: LAI = LAImax× 1 [1+{Exp( ×(Time -T inf l))}] Where LAImax: is the initial maximum leaf area index and is linked with Kc through the empirical formula, derived from equation 97 in FAO 56. It is dimensionless (-) Time: is the running time, in days (d) : is a shape parameter (the default value taken of = -0.1 (d-1), not varying with CO2 increase 22Richard G. A.; Pereira, L.S.; Raes, D.; Smith, M. 1998. Crop evapotranspiration: guidelines for computing crop water requirements. FAO Irrigation and Drainage Paper No 56. 47 Tinf : l is the number of days at which the logistic curve has an inflexion point. In order to maintain coherence in the LAI curve, Tinfl is linked with LAImax and Kcmax and the duration of the growing phase (Dini + Ddev)23 as follows: ( T inf l = (Dini + Ddev)- Log LAImax3 -1 ) Kcmax LAImax0 is not a CROPWAT variable. However, it is linked here with Kcmax through the following empirical relationship: LAImax = (3)(Kc max+0.2) As we know from the literature, higher conversion efficiency will result in faster biomass accumulation which resultantly will have an increase rate of LAI. It is assumed here that the increase in conversion efficiency is linearly related to the biomass accumulation. Since the biomass is described in terms of volume, which is a three dimensional variable, and the LAI is described in terms of area, which is a two dimensional variable, the rate of increase in biomass and LAI can not be linear. To deduce a logical relationship between biomass and LAI we consider the following: if biomass (-) is the increasing efficiency factor due to CO2 increase affecting biomass, then LAI (-) affecting LAI is : LAI = (biomass)23 It is assumed here that CO2 concentration modifies growth rate of LAI by modifying LAImax proportionally to the conversion efficiency factor (LAI). Furthermore, it is also assumed that the total length of the crop cycle will not be affected by increased concentration levels of CO2. According to the modifications of the curve, the value LAI= 3 will be reached sooner: LAImax0 3-1 Days = LogLAImax1 3-1 23D refers to the duration (in days) of a development stage. For example Dini is the number of days a crop needs to complete its initial stage. 48 As a consequence, the initial phase will be shorter and the development phase will begin earlier. This is reached when the modified LAI equals the same value as it had reached in the former situation at the end of the initial phase. This is deduced from the LAI curve, and is calculated as follows: D0 = Dini -T inf l + (1)× LN LAImaxLAImax0 × [1+ EXP{ ×(Dini -T inf l)}-1] The following graph presents the initial LAI curve, modified LAI curve and, as a consequence, the reduction (Delta Days) of the length of the initial and development growth stages. Impact of CO2 on LAI Curve, (less days for reaching LAI = 3). 4.00 3.00 I LA2.00 Delta D0+D1, 5.3 Delta D0, 2.2 1.00 0.00 0 10 20 30 40 50 60 70 Days initial LAI curve modified LAI curve LAI = 3 LAI for D0 Example: For the following values of the parameters of conversion efficiency and wheat in Egypt: Conversion increase efficiency factor D0= 1.3 LAImax0 = 4.4 Kc min = 0.3 D ini = 30 days D dev = 65 days 49 Kc max = 1.15 D mid = 40 days Kc end = 0.50 D end = 30 days Following values for the length of different crop growth stages are calculated: Growth stage Initial/present Modified duration Difference duration (days) (days) ( Days ) Dini 30 27 -3 Ddev 65 61 -4 Dmid 40 47 +7 Dend 30 30 0 Total 165 165 0 The following figure shows this change in the format used in CROPWAT Impact of CO2on reaching maximum Kc values 1.20 1.00 s 0.80 lueav 0.60 cK0.40 0.20 0.00 0 25 50 75 100 125 150 175 Days Initial values Modified values 50 Annex III: Combining the impact of changes in Temperature and CO2 concentration on the duration of different crop growth stages and total crop growth cycle. To combine the effects of both temperature and CO2 changes on the total length of a crop cycle and individual growth stages, it is proposed to simply add the total number of days reduced in each cycle. Example: Taking the examples in Annex I and II, which demonstrate a reduction in the length of growth stages, and thus reduced total length of the crop, of Wheat in Egypt the following table shows the expected total reduction in crop growth stages. Growth Present Difference Difference Total Modified stage duration due to Temp. due to CO2 ( difference ( duration (Days) ( Days) Days) Days) (Days) Dini 30 -4 -3 -7 23 Ddev 65 -12 -4 -16 49 Dmid 40 -8 +7 -1 39 Dend 30 -5 0 -5 25 Total 165 -29 0 -29 136 51 The following figure shows this change in the format used in CROPWAT: Combined Impact of increasd temperature and CO2 concentration level on crop growth stages 1.20 1.00 seulav 0.80 0.60 Kc 0.40 0.20 0.00 0 50 100 150 200 Days Initial values final modified values 52 APPENDIX 2: Application of draft methodology for "using CROPWAT to study crop water requirement as effected by climate change" on Maize crop in South Africa South Africa team used the draft Cropwat-cc methodology to study future crop water requirement and use as affected by climate change. The study was based on climate predictions24 for the year 2050 produced by two internationally renowned institutes, the Hadley Centre (UK) and National Centre for Atmospheric Research (NCAR, USA). Outputs from 3 General Circulation Models (GCM's), namely Genesis - origin presently unclear, CSM - developed by NCAR, USA, and UKMO - developed by the Hadley Center, UK (termed HadCM2, a revised version of the original GCM HadCM) were used in the anaylsis. HADCM2 scenarios were developed using two assumptions regarding sulphate aerosols, one assuming their presence and gradual decline during the simulation period, the other assuming their complete absence during the simulation period. Sulphate aerosols are a product of fossil fuel (especially coal) burning, and ironically act to reflect radiation and therefore counteract warming due to increasing carbon dioxide (CO2). These scenarios are termed HadCM2S (with sulphate effect), and HadCM2N (sulphate effect excluded). Of the four models, Genesis tended to produce the most unique and contradictory results. The models predict changes in temperature and CO2 concentrations. These changes are introduced in the cropwat-cc module (see Figure A-II-1), and changes in the growth stages and total growing period of maize are derived (see table A-II-1 and Figure A-II-1). Interestingly, simulations based on data from all the models are predicting a significant reduction of the initial and development stage of maize as compared to the present situation, with minimum changes in the mid and late stages. A striking observation is that the range of projections based on the data by different models is very narrow for a district ­ 2 days for - Lichtenburg, 3 days for Kroonstad, and 2 days for Middelburg. The greatest decrease in the total growing period of maize (23 days less than present growing period) is projected for Middleburg district when data from CSM and HadCM2S models were used. Based on the climate data produced by GCM and changes in the length of growth stages of maize ­ calculated based on draft methodology proposed by Cropwat-cc, future crop water use was determined using CROPWAT programme. Rest of the other factors, such as area grown under maize in different districts, potential and actual yield and Ky will remain constant. Results vary considerably (see table A-II-2) depending on the climate change scenario. It is striking to see that scenario based on the data from HadCM2N in Lichtenburg district is projecting an increase in crop water requirement and water use of maize. Results based on the data from all other models are projecting a decrease with sharpest (12%) being in Kroonstad district by HadCM2N. 24These predictions assume an increase in atmospheric carbon dioxide (CO2) from about 370 ppm in the year 2000 to 550 ppm in 2050. 53 Impact of CO2 and temperature on crop stage duration 1.4 1.2 esu 1.0 0.8 valcK0.6 0.4 0.2 0.0 0.0 20.0 40.0 60.0 80.0 100.0 120.0 140.0 160.0 Days Initial values final modified values Figure A-II-1: Combined impact of changes in CO2 concentration and temperature on growth stages of maize. This prototype exercise shows what kind of prospective considerations can be done with the suggested methodology. Evidently, the results have to be analysed very carefully. It would be optimum to use them in a training session, in which the effects of different scenarios could be simulated and discussed with the participants. 54 55 S 7 Had CM2 21 22 65 8 11 chtenburg,iL N 9 for Had CM2 22 23 66 9 11 g 8 GEN 22 23 65 9 11 scenarios 7 Middelbur CSM 21 22 65 8 11 change S 2 Had CM2 22 24 67 9 12 climate 0 Had CM2 N 22 23 66 8 12 four foreiz 2 GEN 23 24 67 9 12 tad ma 2 for Kroons CSM 22 24 67 9 12 S stages 3 Had CM2 23 24 67 9 12 2.0 0.2 N 4 Had CM2 23 24 67 9 12 development g th 2 districts GEN 23 24 67 9 12 grow 2 of concentration Lichtenbur CSM 23 24 67 9 12 2 CO ina Length Middelburg in 0 Orig l 30 32 68 10 14 8°C and easer A-II-1: 0.16 1.15d 0.16 34°C ince v d l ini x mi end TatleD ini de mid en ta ma base Table Kroonstad Growth stage D D D D To Kc Kc Kc T Td Relativ CC e/ in increas seu n thdrawiw % decrease cropwater -4 -3 2 -1 -5 -3 -12 -2 -7 -4 -1 -5 teraw seu phases. total ha)/m cropwater (m 45419621 43556092 43974983 46482960 44844988 27676056 26293283 26733377 24425211 26991982 16685678 15596001 16054290 16505484 15836260 and ha) m/ development Actual )m (cubic maize ETcrop (m 189 182 183 194 187 212 201 205 187 206 188 176 181 186 178 lla rainf e adaptede th Area (ha) 239750 239750 239750 239750 239750 130718 130718 130718 130718 130718 88805 88805 88805 88805 88805 ectivfef using Ks 0.45 0.45 0.45 0.45 0.45 0.51 0.51 0.51 0.51 0.51 0.53 0.53 0.53 0.53 0.53 ), Ym (t) 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 m/ha scenarios Ya (t)* 2.17 2.17 2.17 2.17 2.17 2.7 2.7 2.7 2.7 2.7 2.91 2.91 2.91 2.91 2.91 (cubic change Ky 1.25 1.25 1.25 1.25 1.25 1.25 1.25 1.25 1.25 1.25 1.25 1.25 1.25 1.25 1.25 climate )m ur ETcrop (m 423 406 409 433 418 416 396 402 367 406 353 330 339 349 335 fo evapotranspiration nda )m ETo (m 545 499 504 528 511 531 480 491 447 490 462 404 421 431 407 ctsi crop (ha), distree change years5 thr ate for Area the Clim scenario Original CSM GEN HadCM2N HadCM2S Original CSM GEN HadCM2N HadCM2S Original CSM GEN HadCM2N HadCM2S for average A-II-2: km) Table (cubic District Lichtenburg Kroonstad Middelburg Provincial Policy Research Working Paper Series Title Author Date Contact for paper WPS4263HIV/AIDSandSocialCapitalina AntonioC.David June2007 A.David Cross-SectionofCountries 82842 WPS4264FinancingofthePrivateSectorin ConstantinosStephanou June2007 S.Coca Mexico,2000­05:Evolution, EmanuelSalinasMuñoz 37474 Composition,andDeterminants WPS4265TheStructureofImportTariffsinthe OleksandrShepotylo June2007 P.Flewitt RussianFederation:2001­05 32724 WPS4266TheEconomicCommunityofWest SimpliceG.Zouhon-Bi June2007 S.Zouhon-Bi AfricanStates:FiscalRevenue LyngeNielsen 82929 ImplicationsoftheProspective EconomicPartnershipAgreement withtheEuropeanUnion WPS4267FinancialIntermediationinthe HeikoHesse June2007 G.Johnson Pre-ConsolicatedBankingSectorin 34436 Nigeria WPS4268PowertothePeople:Evidencefrom MartinaBjörkman June2007 I.Hafiz aRandomizedFieldExperimentofa JakobSvensson 37851 Community-BasedMonitoringProject inUganda WPS4269ShadowSovereignRatingsfor DilipRatha June2007 N.Aliyeva UnratedDevelopingCountries PrabalDe 80524 SanketMohapatra WPS4270Jump-StartingSelf-Employment? RitaAlmeida June2007 A.Bonfield EvidenceamongWelfareParticipants EmanuelaGalasso 31248 inArgentina WPS4271Construction,Corruption,and CharlesKenny June2007 C.Kenny DevelopingCountries 33540 WPS4272Migration,Remittances,Poverty, DavidMcKenzie July2007 M.Sasin andHumanCapital:Conceptualand MarcinJ.Sasin 36877 EmpiricalChallenges WPS4273RulesofOriginandtheWebofEast MiriamManchin July2007 L.Yeargin AsianFreeTradeAgreements AnnetteO.Pelkmans-Balaoing 81553 WPS4274AreLaborRegulationsDriving MohammadAmin July2007 S.Narsiah ComputerUsageinIndia'sRetail 88768 Stores? WPS4275CanForeignLobbyingEnhance KishoreGawande July2007 V.Cornago Development?TheCaseofTourism WilliamMaloney 84039 intheCaribbean GabrielV.MontesRojas WPS4276HumanCapital,TradeLiberalization, TomKrebs July2007 V.Cornago andIncomeRisk PravinKrishna 84039 WilliamMaloney WPS4277ClimateChangeAdaptationinAfrica: SungnoNiggolSeo July2007 P.Kokila AMicroeconomicAnalysisof RobertMendelsohn 33716 LivestockChoice WPS4278EndogenousIrrigation:TheImpactof PradeepKurukulasuriya July2007 P.Kokila ClimateChangeonFarmersin RobertMendelsohn 33716 Africa WPS4279TheImpactofClimateChangeon SungnoNiggolSeo July2007 P.Kokila LivestockManagementinAfrica: RobertMendelsohn 33716 AStructuralRicardianAnalysis Policy Research Working Paper Series Title Author Date Contact for paper WPS4280 GovernanceMattersVI:Aggregate DanielKaufmann July2007 R.Bonfield andIndividualGovernance: AartKraay 31248 Indicators,1996-2006 MassimoMastruzzi WPS4281 CreditGrowthInEmergingEurope: SophieSirtaine July2007 S.Sirtaine ACauseForStabilityConcerns? IliasSkamnelos 87006 WPS4282 AreCashTransfersMadetoWomen NorbertSchady July2007 I.Hafiz SpentLikeOtherSourcesofIncome JoséRosero 37851 WPS4283 InnovationShortfalls WilliamMaloney July2007 V.Cornago AndrésRodríguez-Clare 84039 WPS4284 CustomerMarketPowerandthe NeeltjeVanHoren July2007 M.Gamboa ProvisionofTradeCredit: 34847 EvidencefromEasternEuropeand CentralAsia WPS4285 PovertyAnalysisUsingAn J.A.L.Cranfield July2007 P.Flewitt InternationalCross-CountryDemand PaulV.Preckel 32724 System ThomasW.Hertel WPS4286 InstitutionalEffectsasDeterminants JesúsÁlvarez July2007 S.Baksh ofLearningOutcomes:Exploring VicenteGarcíaMoreno 31085 StateVariationsinMexico HarryAnthonyPatrinos WPS4287 ACross-CountryAnalysisofPublic MartinMeleckyy July2007 M.Rosenquist DebtManagementStrategies 82602