The Use of El Nino-Southern Oscillation Information for Crop Management Decisions and Its Economic Implication: Case Study for Rice-Base Farming System of West Java, Indonesia Rizaldi Boer1, and Elza Surmaini2 1 Centre for Climate Risk and Opportunity Management in Southeast Asia and Pacific of Bogor Agricultural University, Bogor-Indonesia E-Mail: rizaldiboer@gmail,com 2 Indonesian Agroclimate and Hydrology Research Institute, Bogor-Indonesia Abstract In The El-Nino-Southern Oscillation (ENSO) ‘warm event’ or El-Niño years, for most cases the dry season starts earlier and the rainfall amount is often substantially reduced in Indonesia’s region. One of indicators used for detecting the occurrence of ENSO events is Southern Oscillation Index (SOI). Based on historical rainfall data in the study areas, i.e. Ciparay and Bojongsoang sub-districts, Bandung District, when SOI phase prior to this season was constantly negative (phase 1) or rapidly falling (phase 3) indicating ElNiño cycle, the chance of having rainfall far below normal in this season increased. Therefore, the use SOI phase information prior to the season would be useful for assisting farmers in making a satisfactory planting decision. The results of this study indicates that farmers who switch their crop to soybean or maize for May planting if April SOI Phase is 1 or 3 will get a higher income compared to farmers that keep planting rice. Cumulatively, over 62 years, the net annual income difference between farmers that use April SOI Phase information and chose soybean for May planting and those that do not is about 1,700 USD at Ciparay and 2,350 USD at Bojongsoang. Key words: SOI phases, dry season, drought, rice based farming system. 1. Introduction Floods and droughts are climate hazards that commonly occur in rice-based farming system in West Java, Indonesia (Boer and Subbiah, 2005; Ditlin 2007). In this farming system, farmers plant rice crop twice a year; the first rice crops are planted in the wet season (between November and February), and the second rice crops are planted in the dry season (between March and June). The first crop is commonly affected by floods and the second by drought. Floods commonly occurred between January and March, while drought typically develops in May. These hazards are frequently associated with ENSO events. In ENSO ‘warm event’ (El-Niño) years, for most cases the dry season starts earlier and the rainfall amount is often substantially reduced (Boer and Subbiah, 2005; Kirono and Partridge, 2002; Moron et al., 2009). Under this condition most of second rice crops will suffer from drought (Alimoeso et al., 2002; Meinke and Boer, 2002). SOI phases have been commonly used to indicate the occurrence of El-Nino or La-Nina. SOI consist of five phases (Stone et al., 1996) namely consistently negative (Phase 1), consistently positive (Phase 2), rapidly falling (Phase 3), rapidly rising (Phase 4) and near zero (Phase 5). Phase 1 and 3 is commonly associated with ENSO ‘warm event’, Phase 2 and 4 with ENSO ‘cold event’, and Phase 5 with normal condition. Many studies have demonstrated that the SOI phases can provide skill in assessing future rainfall probabilities during the subsequent planting season (e.g. Stone and Auliciems, 1992; Stone et al, 1996; Everingham et al. 2009; Fawcett and Stone, 2010). In Australia, the SOI phase information has been used widely in managing climate risk to crop production (Meinke, et al. 1996; Meinke and Stone, 1997; Meinke and Hochman, 2000; Everingham et al., 2003; Wang et al. 2008) and also in some other countries such as India (Selvaraju et al. 2004), Pakistan (Meinke et al. 2001), Argentina (Messina et al. 1999), Zimbabwe (Venus, 2000; Zinyengere, 2010), Canada and USA (Hill et al. 2000). The use SOI phase information prior to the season would be useful for assisting farmers in making a satisfactory planting decision. This study aimed to evaluate the economic benefit of using SOI (Southern Oscillation Index) phase information in March/April to determine a satisfactory planting decision for the second crop (April/May planting). 2. Methods Location selected for this study was Ciparay and Bojongsoang sub-districts of Bandung District, West Java Province, Indonesia. These two sub-districts like the rest of West Java have monsoon rainfall (‘Monsoon’) with a monthly rainfall peak in December. Areas with this type of rainfall are significantly affected by ENSO events (Tjasyono, 1997; Boer and Faqih, 2004; Boer and Subbiah, 2005). Rice based farming systems and climate related hazards at the two sub-districts were evaluated through survey. The likely crops yields across seasons were evaluated using crop simulation models (DSSAT; Hoogenboom, 2010) with long-term, daily rainfall records (1950-2012). As the model requires other climatic data such as solar radiation and temperature, these two data were generated using the observed daily rainfall data (Boer et al., 2004). The DSSAT was validated using experimental data collected from Sukamandi Rice Research Institute and Bogor Research Institute for Food Crops. The experiments were conducted in a number of sites, i.e. Bogor, Geritan, Pati and Binangun, West Java. The differences between observed and simulated yields were assessed using statistical t-test. In general the cropping pattern in rice based farming system in West Java is dominated rice-rice-fallow. The first rice crop is planted in wet season (November- February) with peak planting time in January, and the second crops in dry season (March-June) with peak planting time in May. In El-Nino years, May planting was normally exposed to high drought risk. This study evaluated the variability of crops yields from year to year under three different planting arrangements. The first arrangement, the first crop was planted in 1st of December and the second crop in 1st of April. The second, the first crop was planted in 1st of January and the second crop in 1st of May. The third, the first crop was planted in 15th of January and the second crop in 15th of May. March/April SOI phase information is then used to determine whether the second crop would (April/May planting) be continued with rice or changed with other crops, i.e. maize or soybean or left to be fallowed. With the above arrangement, there were four planting decisions namely: (i) RiceRice system, in which farmers will keep planting rice twice a year irrespective of SOI phase information; (ii) Rice-Maize system, in which farmers will switch their second crop into maize if March/April SOI Phase is 1 or 3, and keep planting rice otherwise; (iii) Rice-Soybean system, in which farmers will switch their second crop into soybean if March/April SOI Phase is 1 or 3, and keep planting rice otherwise; and (iv) Rice-fallow system, in which farmers will not plant a second crop if March/April SOI Phase is 1 or 3, but keep planting rice otherwise. The economic benefit of using SOI phases was assessed by cumulating the difference between gross margins, (i) the Rice-Rice system and the Rice-Maize system, (ii) the Rice-Rice system and the Rice-Soybean system, and (iii) the Rice-Rice system and the Rice-Fallow system. The gross margin is the difference between revenue and cost. 3. Results and Discussion Rice Based Farming System and Climate Related Problems in West Java. Most of agriculture areas in the two sub-districts are irrigated, only a few are rainfed. Farmers normally plant rice twice a year in the irrigated areas and once a year in rainfed areas. Maize is the second main crops after rice. Other crops are cassave, sweet potato, onion, long bean, chilli, cucumber and some other vegetable crops. Soybean is planted only by very few farmers. At present, most farmers do not use their rice field for planting maize or soybean. They normally plant maize in dryland throughout the year. Soybean is also planted throughout the year but there is a big variation of planting area between years. 1. Technical irrigared lands Rice-Rice-Fallow 2. Conventional irrigated lands or rainfed: rice-rice-fallow Rice-Rice-Fallow Rice-Upland-Vegetable/fallow Rice-vegetable-vegetable/Fallow 3. Dry lands agriculture: Maize-Maize-Vegetable Maize-soybean-vegetable Like other rice growing areas of West Java, flood and drought are two common climate hazards in these sites. Based on interview with farmers, flood commonly occur between January and April (flood risk increases when monthly rainfall in these months increase above 300 mm) while drought could start developing in May. Drought will affect second crops when rainy season ends earlier or rainfall in the dry season fall far below normal. Too late planting for the second rice crop due to much delay of the onset of wet season can also cause drought. These conditions commonly occur in El-Nino years. Too late planting for second rice will not cause drought problem when rainfall in the season is much above normal which is commonly occur in La-Nina years. A study at Indramayu district of West Java province show clear connection between ENSO events and rainfall characteristic and its association with drought occurrence (Fig. 1). This result of the study shows that the April SOI phases can provide skill in assessing DS rainfall probability distribution of May-June rainfall anomaly of the study areas (Fig 2). Probability to have above average rainfall (indicating by positive anomaly) in May-June increases when April SOI phase is 2 or 4. Probability to have below average rainfall (indicating by negative anomaly) in May-June increases when April SOI phase is 1 or 3. This suggests that profitability of rice farmers could be increased by using knowledge of atmospheric conditions obtained prior to planting to forward estimate production level and risk. The above results indicate that in rice based cropping system, the timing of the first planting and rainfall condition during the second planting season are the two primary factors that should be considered in making second planting decision. Much delay in first planting due delay in the onset of rainy season will cause much delay in second planting. Under these circumstances, keeping planting rice for the second crop may be exposed to high drought risk. Fig. 3 shows that weak El-Nino that occurred in 2003 caused higher drought impact than that occurred in strong 1997 El-Nino. This was because the second planting for rice delayed between one and two months due to late planting of the fist rice crops caused by late onset of wet season in 2002. In 1997, the El-Nino developed rapidly and very strong compare other El-Nino years. This caused the rainy season ended earlier and rainfall in May and June already disappeared before some farmers start planting their second rice crop. Therefore, total area being affected by drought in the strong 1997 El-Nino was lower than those in other El-Nino years. Strong connection between Sea Surface Temperature (SST) anomaly at Pacific Ocean and rice production in Indonesia has motivated a number of scientists to develop rice production model from the SST anomaly. Falcon et al (2004) found that the rice production anomaly in Indonesia can be well predicted from August sea surface temperature anomaly (AugSSA) over the Pacific Ocean. The equation as the following: Paddy production = 35,311 – 1,318 AugSSTA + 1,675 t – 45.0 t2 (1); Adj R2 = 0.94 Where t is time measured in years, where 1 = 1983/84. Economic Benefits of Using SOI Phase Information. The result of validation of the DSSAT model used to evaluate economic benefits of using SOI phase information in crop management decision was reasonably good (Fig 4). The differences between simulated and observed yields were not significantly different. Further analysis to assess the economic benefits of using SOI phase was then done using the simulated yield. The results of the analysis show that farmers who switch their crop to soybean or maize for May planting if April SOI Phase is 1 or 3 will get a higher income compared to farmers that keep planting rice (Fig. 5). Cumulatively, over 62 years, the net income difference between farmers that use April SOI Phase information in deciding a crop type for May planting and those that do not is about 1,700 USD at Ciparay and 2,350 USD at Bojongsoang if they chose soybean as the second crops, and about 1,524 USD and 1,970 USD respectively if they chose maize. Farmers who choose not to plant a second crop will accrue a lower income than those that plant the crops. Furthermore, farmers who decide not to plant a second crop in April if March SOI Phase is 1 or 3 will get a much lower income compared to farmers who plant the second crops. This outcome suggests that the decision to planting or not to plant a second crop in May is more crucial than in April. Farmers who make what appears to be a proper decision for May planting based on April SOI Phase will, in the long term get consistently a much higher benefit than farmers who keep planting rice irrespective of the April SOI Phase. This approach can be used to assess the economic benefit of changing management options such as introducing new crop sequence (non-rice-rice-non-rice) if length of rainy season shortened or changing second crop with horticulture etc. The knowledge gained in study needs to be communicated to farmers so they can make us of the knowledge to assist them in coping with drought hazards. There are a number of means to communicate technologies and knowledge to farmers and each country will have different experiences in doing this. Barriers and constraints faced by farmers for implementing the options also need to be assessed and removed. 4. Conclusion Yield variability at Ciparay and Bojongsoang is significantly correlated with ENSO (El Nino Southern Oscillation). Consistent use of this information in designing cropping management strategy would give benefit to farmers. Farmers’ who changes the second crop from rice to non-rice crops (soybean and maize) whenever April SOI Phase is rapidly falling or consistent negative will have higher income than farmers’ who keep planting rice. Since many parts of Indonesia are strongly influence by ENSO, the use of ENSO information in designing better cropping strategies in particular season is encouraged. Acknowledgement Our thanks to Holger Meinke for valuable inputs. This work was supported by APN CAPaBLE Project. References Alimoeso S, Boer R, Subroto SWG, Purwani ET, Sugiarto Y, Rahadiyan RMK and Suciantini (2002). Distribution of drought prone rice growing area in Indonesia. Directorate of Plant Protection, Ministry of Agriculture, Jakarta (in Indonesian). Boer R and Faqih M (2004). Global climate forcing factors and rainfall variability in West-Java. Indones J Agric Meteorol 18:1-12 Boer R, Rakiso P , Faqih A (2004). The use of climatic data generator to cope with daily climatic data scarcity in simulation studies. In Proceedings of the 4th International Crop Science Congress Brisbane, Australia, 26 Sep – 1 Oct 2004, http://www.cropscience.org.au/icsc2004/poster/2/6/index.htm#TopOfPage Proceedings of the 4th International Crop Science Congress, 26 Sep – 1 Oct 2004, Brisbane, Australia. Published on CDROM (www.cropscience.org.au) Boer R, Subbiah AR (2005) Agricultural droughts in Indonesia. In Boken VK, Cracknell AP, Heathcote RL. Monitoring and Predicting Agriculture Drought. Oxford University Press, p:330-344. Ditlin. 2007. Pedoman Mitigasi Dampak Fenomena Iklim: Pengelolaan Dampak Kejadian Iklim Ekstrim pada Tanaman Pangan. Direktorat Perlindungan Tanaman, Departemen Pertanian, Jakarta. Everingham YL, Zamykal D, McKinna LI (2009). Rain forecaster—a seasonal climate forecasting tool. Proceeding Australian Society Sugar Cane Technology 31:5064. Everingham YL, Muchow RC, Stone RC, Coomans DH (2003) Using southern oscillation index phases to forecast sugarcane yields: a case study for northeastern Australia. Int. J Climatology 23: 1211–1218 Falcon WP, Naylor RL, Smith WL, Burke MB, McCullough EB (2004). Using Climate Models To Improve Indonesian Food Security, Bull Indones Econ Stud 40: 355– 77 Fawcett RJB, Stone RC (2010). A comparison of two seasonal rainfall forecasting systems for Australia. Australian Meteorological and Oceanographic Journal 60: 15-24 Hill JP, Mjelde JW, Rosenthal W, Love HA, Fuller SW (2000) Comparing the value of Southern Oscillation index-based climate forecast methods for Canadian and US wheat producers. Agric. For. Meteor., 100, 261–272. Hoogenboom G, Jones JW, Wilkens PW, Porter CH, Boote KJ, Hunt LA, Singh U, Lizaso JL, White JW, Uryasev O, Royce FS, Ogoshi R, Gijsman AJ, Tsuji GY, Koo J (2010) Decision Support System for Agrotechnology Transfer (DSSAT) Version 4.5 [CD-ROM]. University of Hawaii, Honolulu, Hawaii. Kirono D, Partridge IJ. 2002. The climate and the SOI. p. 17-24. In Partridge IJ, Ma’shum M(ed) Will It Rain?: The effect of the Southern Oscillation and El Niño in Indonesia. Queensland Government, Department of Primary Industry, Australia. Meinke H, Boer R (2002) Plant growth and the SOI. In Partridge IJ, Ma’shum M (ed) Will It Rain? The effect of the Southern Oscillation and El Niño in Indonesia. Queensland Government, Department of Primary Industry, Australia. p. 25-28. Meinke H, Baethgen WE, Carberry PS, Donatelli M, Hammer GL, Selvaraju R, Stockle CO (200) Increasing profits and reducing risks in crop production using participatory systems simulation approaches. Agric. Syst70: 493–513. Messina C, Hansen JW, Hall AJ (1999) Land allocation conditioned on El Niño– Southern Oscillation phases in the Pampas of Argentina. Agric. Syst., 60, 197– 212. Moron V, Robertson A. Boer R (2009) Spatial Coherence and Seasonal Predictability of Monsoon Onset over Indonesia. J Clim 22: 840–850. Selvaraju R., Meinke H, Hansen J (2004). Approaches allowing smallholder farmers in India to benefit from seasonal climate forecasting. Proceedings of the 4th International Crop Science Congress, 26 Sep – 1 Oct 2004, Brisbane, Australia. Published on CDROM (www.cropscience.org.au) Stone RC, Auliciems A (1992). SOI phase relationship with rainfall in earstern Australia. Int J Climatology, 12:625-636 Stone RC, Hammer GL, Marcussen T (1996) Prediction of global rainfall probabilities using phases of the Southern Osciallation Index. Nature 384:252-255. Tjasyono B (1997) Mekanisme fisis para, selama, dan pasca El Niño. Paper disajikan pada Workshop Kelompok Peneliti Dinamika Atmosfer, 13-14 Maret 1997. Venus V (2000) Towards a framework for maize yield forecasting; a comparison of two promising methods: The Case of Mashonaland West Province, Zimbabwe. Thesis submitted to the International Institute for Aerospace Survey and Earth Sciences in partial fulfillment of the requirements for the degree of Master of Science in Rural Land Ecology Survey for Natural Resource Management. The Netherland Wang E., Xu JH, Smith CJ (2008) Value of historical climate knowledge, SOI based seasonal climate forecasting and stored moisture at sowing in crop nitrogen management in south eastern Australia. Agric For Meteor 148:1743-1753 Zinyengere N (2010) Maize yield prediction using seasonal weather forecasts and a crop growth simulation model. Thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in Agricultural Meteorology, University of Zimbabwe 80000 70000 200 150 100 40000 30000 20000 50 0 10000 0 Okt-90 Nov-90 Dec-90 Jan-91 Feb-91 Mar-91 Apr-91 May-91 Jun-91 Jul-91 Aug-91 Sep-91 60000 50000 DS rainfall far below normal Normal Rainfall Rainfall Planting Area 80000 70000 60000 50000 40000 30000 20000 10000 0 Area (ha) 300 250 400 350 300 250 200 150 100 50 0 Oct-97 Nov-97 Dec-97 Jan-98 Feb-98 Mar-98 Apr-98 May-98 Jun-98 Jul-98 Aug-98 Sep-98 Rainfall Planting Area Drought Area Rainfall (mm) Normal Rainfall Area (ha) Rainfall (mm) 400 350 DS rainfall far above normal Fig 1. Relationship between rainfall characteristics and drought occurren ce in El-Nino (1990/91) and La-Nina (1997/98) years at Indramayu District of West Java Province (Boer, 2002) Probablity of Exceedence 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 -200 Phase 1+3 Phase 2+4 Phase 5 0 200 400 Anomaly May-June Rainfall (mm) Fig 2. Probability of exceedence of May-June rainfall anomaly in the study area by April SOI phases 1+3, 2+4 and 5 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 Anomaly of SST 3.4 (oC) -0.5 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 Cumulatuve Drought Area (000 ha) 3.0 2.5 2.0 1.5 1.0 0.5 0.0 -1.0 1991 1991 1994 1994 1997 1997 2002 2002 2003 25 20 15 10 5 0 2003 Fig 3. Relationship between El-Nino development and cumulative monthly drought rice area in Bandung district. 3 7 6 Simulated Yield (t/ha) r = 0.87 5 4 3 2 1 0 Simulated Yield (t/ha) 4 8 Simulated Yield (t/ha) Soybean Maize Rice r = 0.81 3 2 1 0 0 1 2 3 4 5 6 Observed Yield (t/ha) 7 8 r = 0.90 2 1 0 0 1 2 3 Observed Yield (t/ha) 4 0 1 2 3 Observed Yield (t/ha) Fig 4. Comparison between simulated yields from DSSAT and observed yields for rice, maize and soybean 1st Planting: 1 Dec, 2nd Planting: 1 April Ciparay 0 -1000 -2000 -3000 -4000 Rice-Maize Rice-Soybean Rice-Fallow 2000 1000 0 -1000 -2000 -3000 -4000 Rice-Maize Rice-Soybean Rice-Fallow 1950 1955 1960 1965 1970 1975 1981 1987 1992 1997 2002 2007 2012 1000 Cumulative Income Difference from Rice-Rice System (USD) 2000 Bojongsoang 3000 1950 1955 1960 1965 1970 1975 1981 1987 1992 1997 2002 2007 2012 Cumulative Income Difference from Rice-Rice System (USD) 3000 1st Planting: 1 Jan, 2nd Planting: 1 May 1000 0 -1000 -2000 -3000 -4000 Rice-Maize Rice-Soybean Rice-Fallow Cumulative Income Difference from Rice-Rice System (USD) 2000 3000 Bojongsoang 2000 1000 0 -1000 -2000 -3000 -4000 Rice-Maize Rice-Soybean Rice-Fallow 1950 1955 1960 1965 1970 1975 1981 1987 1992 1997 2002 2007 2012 Ciparay 1950 1955 1960 1965 1970 1975 1981 1987 1992 1997 2002 2007 2012 Cumulative Income Difference from Rice-Rice System (USD) 3000 1st Planting: 15 Jan, 2nd Planting: 15 May 1000 0 -1000 -2000 -3000 -4000 Rice-Maize Rice-Soybean Rice-Fallow Cumulative Income Difference from Rice-Rice System (USD) 2000 3000 Bojongsoang 2000 1000 0 -1000 -2000 -3000 -4000 Rice-Maize Rice-Soybean Rice-Fallow 1950 1955 1960 1965 1970 1975 1981 1987 1992 1997 2002 2007 2012 Ciparay 1950 1955 1960 1965 1970 1975 1981 1987 1992 1997 2002 2007 2012 Cumulative Income Difference from Rice-Rice System (USD) 3000 Fig 5. Income difference between farmers that use and do not use SOI Phase information