Water use efficiency and economic feasibility of growing rice and wheat with sprinkler irrigation in the Indus Basin of Pakistan

Transcription

1 agricultural water management 87 (2007) available at journal homepage: Water use efficiency and economic feasibility of growing rice and wheat with sprinkler irrigation in the Indus Basin of Pakistan Muhammad Akram Kahlown a, *, Abdur Raoof 1,b, Muhammad Zubair 1,b, W. Doral Kemper a a Pakistan Council of Research in Water Resources (PCRWR), Khyaban-e-Johar, H-8/1 Islamabad, Pakistan b Regional Office, PCRWR 6 km, Thokar Niaz Baig, Main Raiwind Road, Lahore, Pakistan article info Article history: Accepted 27 July 2006 Published on line 26 September 2006 Keywords: Rice Wheat Sprinkler irrigation Indus Basin Pakistan Water saving Evapotranspiration Water productivity Economic feasibility abstract With a population of more than 150 million, Pakistan cannot meet its need for food, if adequate water is not available for crop production. Per capita water availability has decreased from 5600 m 3 in 1947 to 1000 m 3 in Water table has gone down by more than 7 m in most parts of the country. Present need is to identify and adopt measures, that will reduce water use and increase crop production. This study was conducted in farmers fields during to evaluate the water use efficiency and economic viability of sprinkler irrigation system for growing rice and wheat crops. Yields and water use were also measured on adjacent fields irrigated by basin flooding, which were planted with the same crop varieties. Sprinkler irrigation of rice produced 18% more yield, while reducing consumption of water to 35% of that used in the traditional irrigation system. Sprinkler irrigation of wheat resulted in a water use efficiency of 5.21 kg of grain per cubic meter of water used compared to 1.38 kg/m 3 in the adjacent flooded basins. Benefit cost analysis showed that adoption of rain-gun sprinkler irrigation for rice and wheat is a financially viable option for farmers. While these findings show large potentials for improving water use efficiency in crop production they also indicate that a large portion of the water applied in traditional flooded basin irrigation is going to groundwater recharge, which has high value near large cities which draw their water from the aquifer. # 2006 Elsevier B.V. All rights reserved. 1. Introduction Irrigated agriculture produces about 40% of all food, and consumes 69% of all freshwater resources (FAO, 2000). Global population growth is expected to increase the demand for cereals including rice and wheat by 1.27% annually between 2000 and 2025 (Rosegrant and Cai, 2000). To meet the projected demand for food, irrigated agriculture will require an increase of 17% in freshwater resources (Seregeldin, 1999). In many arid and semi-arid countries where population growth is high, and freshwater is in short supply, there is pressure on the agricultural sector to reduce its water consumption and make it available for the urban and industrial sectors. This drives the demand to produce cereals, especially rice and wheat, using * Corresponding author. Tel.: ; fax: addresses: pcrwr@isb.comsats.net.pk, kahlown@hotmail.com (M.A. Kahlown), ropcrwr@lhr.comsats.net.pk (A. Raoof), ropcrwr@lhr.comsats.net.pk (M. Zubair), pcrwr@isb.comsats.net.pk, kahlown@hotmail.com (W.D. Kemper). 1 Tel.: ; fax: /$ see front matter # 2006 Elsevier B.V. All rights reserved. doi: /j.agwat

2 agricultural water management 87 (2007) lower amount of irrigation water. Pakistan is no exception to this challenge. In Pakistan, farmers generally apply water to unleveled bunded units, resulting in long irrigation events, poor water uniformity and over-irrigation (Kahlown and Kemper, 2004). Rice growers in particular tend to believe that it requires standing water during the growing season to maximize yields. These practices result in very low irrigation efficiencies. Studies within Pakistan indicate that cm water is applied per irrigation event, which is considerably higher than the consumptive use between two irrigation events, i.e. approximately 8 cm (Kahlown et al., 2001). On-farm irrigation efficiencies range between 23 and 70% (Clyma and Ashraf, 1975; Kalwij, 1997; Kijne and Kuper, 1995; Kahlown et al., 1998). Recently, there have been attempts to adopt pressurized irrigation methods to grow rice and wheat in various countries (Spanu et al., 1996). Sprinkler systems such as portable rainguns can be used to apply a desired depth of water during presowing and subsequent irrigations. The application of irrigation water with sprinklers has improved on-farm irrigation efficiencies up to 80% under the prevailing climatic conditions in the Indian sub-continent (Sharma, 1984). The potential for the adoption of sprinklers to irrigate rice and wheat has not been evaluated in the Indus Basin of Pakistan, where water is delivered to farms by watercourses based on warabandi, a rotational method for equitable allocation of available water, by turn, with a fixed day, time and duration of supply to each irrigator. Warabandi provides a continuous rotation of water in which one complete cycle of rotation generally lasts 7 days. The duration of supply to each farmer is proportional to the size of the farmers land holding (Bandaragoda, 1998). This practice poses an additional constraint on the adoption of pressurized irrigation methods, which require water to be available for irrigation when soil moisture is in deficit. Therefore, there is a need for on-farm storage of water if sprinkler irrigation is to be adopted in this area. Consequently, farmers would incur additional expenses to pay for sprinkler irrigation system and on-farm storage. In this study, the potential for improving water productivity of rice and wheat by adopting rain-gun sprinklers for irrigation in the Indus Basin of Pakistan was evaluated. In addition to sprinkler irrigated fields, basin irrigated fields were also monitored for comparison. The specific objectives of the study were to apply irrigation water with a portable rain-gun above and below the evapotranspiration (ET c ) requirements for rice and wheat and to compare crop water productivity of rice and wheat with crops grown under basin flooding; and to estimate whether the water savings resulting from sprinkler irrigation can pay for the additional costs incurred. 2. Materials and methods 2.1. Layout and treatments for rice (2002) and wheat ( ) Trials in ( ) were carried out on Monoo Farm, 6 km from the Regional Office of the Pakistan Council of Research in Water Resources (PCRWR) at Lahore. Clay loam soil is dominant in this area. Three different irrigation treatments were evaluated each with three replications and having size of plots, each 36.6 m 36.6 m. Circular areas within these plots covered by rain-gun water were taken as the actual experimental areas while the remaining area was used as a buffer area (Fig. 1). A reservoir (2 m 2 m 2 m) was constructed near the trial sites to store water. A 16 HP diesel operated pump was installed to operate the rain-gun system. Pan evaporation and rainfall data were recorded, and daily crop evapotranspiration requirement (ET c ) estimated (Eq. (1)) using daily pan evaporation data (E Pan ), a pan coefficient (K Pan )of0.7(khan, 2001) andacropcoefficient (K c )(Kaleemullah et al., 2001). ET c ¼ E Pan K Pan K c (1) Water flow meters were installed on the supply lines of each rain-gun to measure the cumulative volume discharged. Irrigation was applied twice a week for sprinkler irrigated plots, as a percentage of crop evapotranspiration demand. Adopted irrigation treatments are summarized in Table 1. Water used in the basin irrigation plot was measured with a cut-throat flume. For the rice (2002) crop, the seedbed was prepared by two disc ploughings and four simple ploughings followed by two plankings and puddling (process of breaking down soil aggregates into uniform mud, accomplished by applying mechanical force to the soil at high water content). The rice variety Super Basmati was transplanted in the first week of June and harvested in last week of October. Fertilizers applied included 155 kg/ha of diamonium phosphate at the time of transplanting and after 25 days urea and zinc sulphate at the rate of 124 and 12.5 kg/ha, respectively. Wheat was sown in second week of November 2002 and harvested in second week of April Seed-bed preparation included one disc ploughing followed by two cultivations and two plankings. A seed rate of kg/ha of Inqlab-91 of wheat variety was used. Fertilizers added were 75 kg/ha nitrogen and 70 kg/ha phosphorus at the planting time Layout and treatments for rice (2003) and wheat ( ) Trials on rice and wheat were again conducted ( ) at the Inam Ilahi Farm, at Mouza Sheikh Da Kot, about 13 km from the Regional Office of the PCRWR at Lahore. Clay loam soil is dominant in the area. Three different irrigation treatments were evaluated each with three replications. Since variations among yields of ( ) treatments were small, it was decided to minimize the variation in water applications among sprinkler irrigated treatments ( ). In addition to this, moisture stress condition was also considered for rice trials. Nine plots, each covering one of the half circle areas shown in Fig. 2, each measuring 561 m 2 (18.9 m radius) were prepared for the trials. The irrigation supply was same as in the previous year. Adopted irrigation treatments are summarized in Table 1. Crop water requirements were usually met using canal water with private tubewell water used when canal water was not available. Irrigation volumes were measured with flow meters for sprinkler irrigated plots, and with a cut-throat flume for surface irrigated plots. Water savings were calculated by

3 294 agricultural water management 87 (2007) Fig. 1 Layout of study for rice and wheat subtracting the volume of water applied for the first three treatments from the amount of water applied for basin irrigated plots. For rice (2003), the seedbed preparation included five ploughings with four plankings. Super Basmati was transplanted at the end of May and harvested in the end of October. The crop was fertilized with 155 kg/ha of DAP at transplanting and 124 kg/ha of urea 25 days later. Wheat was grown on the same plots during second week of November 2003 and harvested in third week of April Tillage included three ploughings including one disc ploughing followed by two plankings. Fertilizer addition and crop variety used were the Table 1 Summary of irrigation application treatments for rice and wheat trials Treatment Rice Wheat T 1 100% of ET c 91% of ET c 75% of ET c 85% of ET c T 2 125% of ET c 100% of ET c 100% of ET c 100% of ET c T 3 150% of ET c 109% of ET c 125% of ET c 115% of ET c T 4 (basin) 192% of ET c 261% of ET c 225% of ET c 228% of ET c

4 agricultural water management 87 (2007) Fig. 2 Layout of study for rice and wheat same as in the previous year. Yield of both the crops were measured for the entire treatment area of each replication. Analysis of variance (ANOVA) was carried out for crop yield and water productivity of rice and wheat. The multiple comparisons of mean differences were also done by post hoc test for yield and water productivity of the crops (Ott, 1988). 3. Benefit cost analyses The benefit cost analysis of this study involves the assessment of the benefits and costs of sprinkler and basin flooding irrigation for rice and wheat. The financial feasibility of adopting rain-gun sprinkler irrigation to grow rice and wheat was determined by benefit cost analyses, based on results obtained from rice trials (2003) and wheat trials ( ), which involved the following (McConnell and Brue, 2005): Gross income ðgiþ ¼average price of crop ðapcþ crop water productivity ðcwpþ (2) Net income per cubic meter ðnicmþ ¼ GI cost of cubic meter of water ðccmwþ (3) Net income per water saved ðniwsþ ¼ NICM net income of basin irrigation ðnictþ (4) Benefit cost ratio ¼ NIWS CCMW 4. Results and discussions 4.1. Rice water productivity of (2002) and (2003) trials Results from rice (2002) trials are summarized in Table 2. The number of irrigations applied to the basin irrigated plots were one-half as compared to the sprinkler irrigated plots, reflecting water availability constraint with warabandi system, although much water was supplied than to the sprinkler irrigated plots. (5) Table 2 Water applied (m 3 /ha) and water savings for rice trials Treatment Number of irrigations Yield (kg/ha) Crop water productivity (kg/m 3 ) Volume of water applied (m 3 /ha) Water saving (%) R 1 R 2 R 3 Average R 1 R 2 R 3 Average 2002 T 1 (100% of ET c ) T 2 (125% of ET c ) T 3 (150% of ET c ) T 4 (192% of ET c ) T 1 (91% of ET c ) T 2 (100% of ET c ) T 3 (109% of ET c ) T 4 (261% of ET c )

5 296 agricultural water management 87 (2007) Table 3 Analysis of variance for rice and wheat crops Crop F ratio (rice) F ratio (wheat) Yield * 848 * Crop water productivity * * Yield 2822 * 4791 * Crop water productivity * * * Significant at 0.01 level. The water savings when compared with basin irrigation for T 1 (100% of ET c ), T 2 (125% of ET c ) and T 3 (150% of ET c ) were 48, 35 and 22%, respectively, higher than basin irrigation. Rain-gun treatment with 100% of ET c produced 3105 kg/ha of rice yield with highest crop water productivity of 0.55 kg/m 3 while conventional basin flooding resulted in the lowest productivity of 0.29 kg/m 3. Treatments with 125 and 150% of ET c increased the yield by only 2.1 and 2.9%, respectively, compared with treatment 100% of ET c but the crop water productivity became much lower as compared to T 1. Similar to rice crop (2002), the number of irrigations in basin irrigated plots (2003) were around one-half of those with sprinkler irrigated plots due to rigidity of warabandi system. However, water applied was again higher than the sprinkler irrigated plots. The water savings compared with basin irrigation for T 1 (91% of ET c ), T 2 (100% of ET c )andt 3 (109% of ET c ) were 65, 62 and 58%, respectively, higher than basin irrigation (Table 2). Rain-gun treatment T 3 produced maximum yield (3359 kg/ha) but the crop water productivity was lower than other sprinkler treatments of T 1 and T 2 by 10 and 3%, respectively. The basin irrigated field produced up to 31% less rice than the rain-gun irrigated fields, possibly due to excessive leaching of nutrients. Crop water productivity of sprinkler irrigated treatments was higher than that of basin irrigation by 225%. Treatment T 1 resulted in highest crop water productivity, but did not differ much from crop water productivity of other two sprinkler irrigated treatments. Despite minimizing the variation in water application from 100 to 91%, no significant impact was observed on crop water productivity. Lysimetric study conducted in lower Indus Basin indicated that yield and crop water productivity of rice maximized with water use of 1530 mm (Kahlown et al., 2006). Although the yield increased by increase in water use, crop water productivity decreased. Similar relationship was found by Bouman and Tuong (2001) reporting that water productivity under continuous flooded rice was g grain per kg of water in India with water input of mm whereas it was g grain per kg of water in Philippines with water input of mm. The ANOVA statistical analysis (Table 3) showed that crop yield and water productivity for both the years were found significant at 0.01 level. While the mean differences among treatments T 1,T 2 and T 3 for yield and crop water productivity of rice were significant at 0.01 level during both seasons (Table 4) Wheat water productivity of ( ) and ( ) trials Results of wheat ( ) trials are summarized in Table 5. Up to 67% more water was applied to the basin irrigated plots Table 4 Statistical comparisons among treatments of rice and wheat Treatment Mean differences (2002) Mean differences ( ) Rice yield Rice water productivity Wheat yield Wheat water productivity T 1 T 2 65 * * 34 * * T 3 90 * * 111 * * T 4 18 * * 298 * * T 2 T 3 25 * * 77 * * T 4 47 * * 264 * * T 3 T 4 72 * * 187 * * Mean differences (2003) Mean differences ( ) Rice yield Rice water productivity Wheat yield Wheat water productivity T 1 T * * 750 * * T * * 799 * * T * * 8 * * T 2 T * * 49 * * T * * 758 * * T 3 T * * 807 * * * Significant at 0.01 level.

6 agricultural water management 87 (2007) Table 5 Water applied (m 3 /ha) and water savings for wheat trials Treatment Volume of water applied (m 3 /ha) Water saving (%) Yield (kg/ha) Crop water productivity (kg/m 3 ) R 1 R 2 R 3 Average R 1 R 2 R 3 Average T 1 (75% of ET c ) T 2 (100% of ET c ) T 3 (125% of ET c ) T 4 (225% of ET c ) T 1 (85% of ETc) T 2 (100% of ETc) T 3 (115% of ETc) T 4 (228% of ETc) than sprinkler irrigated plots. The water savings when compared with basin irrigation for T 1 (75% of ET c ), T 2 (100% of ET c ) and T 3 (125% of ET c ) were 67, 56 and 45% higher, respectively, than the basin irrigation. All rain-gun treatments produced more yield than basin irrigation treatment. Crop water productivity of sprinkler irrigated treatments was also higher than that of basin irrigation. The treatment T 1 resulted in highest crop water productivity of 4.29 kg/m 3 of water and treatment of basin irrigation produced the lowest crop water productivity of 1.3 kg/m 3. Trials on wheat water productivity ( ) generally showed that more water was applied to the basin irrigated plots than the sprinkler irrigated plots (Table 5). The water savings when compared with basin irrigation for T 1 (85% of ET c ), T 2 (100% of ET c ) and T 3 (115% of ET c ) were 74, 65 and 58%, respectively, higher than basin irrigation (Table 5). Rain-gun treatment T 3 (115% of ET c ) produced maximum yield of 5036 kg/ha. Crop water productivity of sprinkler irrigated treatments was 70% higher than that of basin irrigation (Table 5). The T 1 treatment resulted in highest crop water productivity of 5.21 kg/m 3 of water while the basin irrigation treatment resulted in the lowest crop water productivity of 1.38 kg/m 3. The statistical analysis (ANOVA) showed that crop yield and water productivity for both the years were found significant at 0.01 level (Table 3). The mean differences analysis among treatments for crop water productivity as well as yield for the crop grown during both seasons ( and ) ranged from to 3.800, which were significant at 0.01 level (Table 4). The experimental results showed that the rain-gun sprinkler irrigation gave better yield and much higher crop water productivity than the traditional irrigation irrespective of application of irrigation water above or below actual crop water requirement, i.e. 100% of ET c. However, the sprinkler irrigation system is not adopted by the farmers in the irrigated areas of Pakistan as value of water saved is less than the cost of system. Whereas, the rain-gun sprinkler irrigation system is becoming popular in dryland farming areas especially in Potohar plateau because of the higher value of water than the cost of system (Yasin et al., 2003). In all cases, farmers used more water than required to meet ET c, which implies excess water is recharging the aquifer. Depending on the circumstances, the recharge may have a positive or a negative impact on the environment and the socio-economic. In saline groundwater areas, recharge lead to water table rise and salinisation. This has a negative impact on the environment and the socio-economic of farming communities. On the other hand, in fresh groundwater areas, recharge to groundwater is often reused for irrigation when canal water is not available. Furthermore, if recharge occurs in fresh groundwater areas near cities, which rely on groundwater for urban and industrial use, the socio-economic use will be very high. Although such trade-offs are recognized but they were not evaluated, as they were considered beyond the scope of this study. Table 6 Benefit cost analysis of growing rice and wheat based on water saved Treatment Crop water productivity (kg/m 3 ) Gross income (US$/m 3 ) Cost of water (US$/m 3 ) Net income (US$/m 3 ) Net income based on water saved (US$/m 3 ) Benefit cost ratio Rice (2003) T 1 (91% of ET c ) T 2 (100% of ET c ) T 3 (109% of ET c ) T 4 (261% of ET c ) Wheat ( ) T 1 (85% of ET c ) T 2 (100% of Et c ) T 3 (115% of ETc) T 4 (228% of ET c )

7 298 agricultural water management 87 (2007) Benefit cost analysis rice (2003) and wheat ( ) Benefit cost analysis of rice (2003) showed benefits occurred in the form of net product value resulting from the use of raingun method, which included irrigation cost (US$ per cubic meter of water), capital and maintenance cost of the sprinklers, pump and water holding reservoirs. Average price of rice used was US$ 0.24 per kg. Eqs. (2) (5) were used to estimate benefit cost ratios of rice 2003 (Table 6). Since the ratio is greater than 1 for T 1 for all three treatments, therefore, investing in a rain-gun sprinkler unit and irrigating between 91 and 109% ET c is a financially viable option at current costs and production prices. The benefit cost analysis of wheat ( ) showed that the benefits occurred from net product value resulting from the use of rain-gun method (method cost includes irrigation cost, US$ per cubic meter). Average price of wheat was considered as US$ 0.17 per kg. Eqs. (2) (5) were used to estimate benefit cost ratios of wheat ( ) (Table 6). Benefit cost ratios of all two sprinkler irrigated treatments were higher than 1. Therefore, it appears that investing in a rain-gun sprinkler unit and irrigating wheat between 85 and 115% ET c is a financially viableproposition atcurrent market costs and prices. 5. Conclusions Irrigation interval, amount and its uniform distribution greatly affected the water use efficiency and yield of rice and wheat crops. Results showed that, less irrigation water was needed when both rice and wheat were irrigated with rain-gun sprinklers and water use efficiency was much higher than for basin irrigation. Depending on the crop and the seasonal conditions, irrigation requirement can be as low as 26% of the water used in basin irrigation. Average crop water productivity of rice and wheat were 0.55 and 3.95 kg/m 3 with sprinkler irrigation and 0.25 and 1.34 kg/m 3 with basin irrigation. Benefit cost analyses based on water saved indicated that investing in rain-gun system to irrigate rice and wheat is a financially viable option for farmers. Therefore, adaptation of rain-gun sprinkler has begun in Potohar plateau of Pakistan to provide supplemental irrigation to dryland farming. However, wide scale adoption of sprinkler irrigation system in canal irrigated areas of the Indus Basin is not expected in near future due to present level of value of water. references Bandaragoda, D.J., Design and Practice of Water Allocation Rules: Lessons from Warabandi in Pakistan s Punjab. Research Report 17. International Irrigation Management Institute, Colombo, Sri Lanka. Bouman, B.A.M., Tuong, T.P., Field water management to save water and increase its productivity in irrigated lowland rice. J. Agric. Water Manage. 49, Clyma, W.A., Ashraf, M.M., Irrigation practices and application efficiencies in Pakistan. In: Water Management Research Project. Water Management Technical Report No. 39, Colorado State University, Fort Collins, Colorado, USA. FAO, Crops and Drops: Making the Best Use of Water for Agriculture. FAO Advance Edition, Rome. Kahlown, M.A., Raoof, A., Hanif, M., Plant population effect on paddy yield. J. Drain. Water Manage. 5 (1), 1 5. Kahlown, M.A., Shafique, M.S., Iqbal, M., Improved irrigation methods for efficient use of irrigation water under different water-table depths. Mona Reclamation Experimental Project, WAPDA, Bhalwal (Pub. no. 231). Kahlown, M.A., Ashraf, M., Chang, M.H., Water requirement of rice in the lower Indus Basin of Pakistan. PCRWR (Research Report 2005). Kahlown, M.A., Kemper, W.D., Seepage losses as affected by condition and composition of channel banks. J. Agric. Water Manage. 65 (2), Kaleemullah, M., Habib, Z., Muhammad, S., Spatial Distribution of References and Potential Evapotranspiration. Working Paper 24. International Water Management Institute, Pakistan. Kalwij, I.M., Assessing the Field Irrigation Performance and Alternative Management Options for Basin Surface Irrigation Systems Through Hydrodynamic Modeling. Report No. 35. International Water Management Institute, Pakistan. Khan, S.R.A., Crop Management in Pakistan with Focus on Soil and Water. Government of Punjab, Agriculture Department, Lahore, Pakistan. Kijne, J.W., Kuper, M., Salinity and sodicity in Pakistan s Punjab: a threat to sustainability of irrigated agriculture? J. Water Resour. Dev. 11 (1). McConnell, C.R., Brue, S.L., Economics: Principles, Problems and Policies, 16th ed. McGraw-Hill Companies, USA. Ott, L., An Introduction to Statistical Methods and Data Analysis. PWS-KENT Publishing Company, Boston, USA. Rosegrant, M.W., Cai, X., Modeling Water Availability and Food Security A Global Perspective: The IMPACT-Water Model. Working Draft. International Food Policy Research Institute, Washington, DC. Seregeldin, I., Looking ahead: water, life and the environment in the 21st century. J. Int. Water Resour. Dev. 15 (1 and 2), Sharma, S.K., Principles and Practices of Irrigation Engineering. Oxford and IBH Publication Co., New York. Spanu, A.G., Andria, P.R., Lavini, A., Chiranda, F.Q., Yield response of rice to increasing sprinkler irrigation. J. Int. Comm. Irri. Drain. 45, Yasin, M.A., Ahmad, S., Hussain, Z., Moshabbir, P.M., Ahmad, M.M., Adaptation of pressurized irrigation systems. S&T Policy Scientometrics J. 2 (2),