How widespread are yield declines in long-term rice experiments in Asia?
Introduction
Long-term experiments (LTEs) provide one of the means to measure sustainable management systems in agriculture (Rasmussen et al., 1998). They are records of the past and may serve as early warning systems for the future. This paper provides an updated review of yield trends in LTEs with intensive rice systems of tropical and subtropical regions of Asia, cropping systems that emerged in the 1960s and 1970s during the Green Revolution. The focus is on systems with two to three crops grown per year, mainly double- and triple-crop rice monoculture and rice-wheat systems, which account for about 50% of the global rice supply. Older LTEs, usually single rice crop systems, exist in temperate regions such as Japan (Matsuo et al., 1976), but are not included in this study.
Perhaps the first evidence of a yield decline in rice–rice systems of the tropics dates back to 1979, when F.N. Ponnamperuma presented data showing trends of declining maximum yields at a rate of −125 to −450 kg ha−1 per year in cultivar trials conducted at IRRI during 1966–1978 (Ponnamperuma, 1979). Flinn et al. (1982) and Flinn and De Datta (1984) also found evidence of declining yields, especially at IRRI (International Rice Research Institute), but also at some other LTEs elsewhere in the Philippines. Cassman and Pingali (1995) summarized these results and extended the analysis to include more LTEs at these and other sites. Their results showed statistically significant yield declines in six of seven LTE data sets at IRRI. In data sets at Philippine LTEs outside of IRRI, yield declines occurred in three of ten cases.
In 1971–1972, two LTEs with rice–rice cropping were established in the All India Coordinated Research Project (AICRP) on Long-Term Fertilizer Experiments (Nambiar and Ghosh, 1984). Cassman and Pingali (1995) estimated yield trends for the period 1972 to 1981 at these two rice–rice experiments from a published graph, but without access to the raw data. They concluded that there were statistically significant yield declines in both the wet (WS) and dry (DS) seasons at an experiment in Hyderabad, and a statistically significant yield decline in the WS only at Bhubaneswar. However, the experiment at Hyderabad was often affected by water shortages and was eventually terminated because of these problems. A more recent analysis of the experiment at Bhubaneswar with data until 1994 did not find a yield decline in the treatments with the highest fertilizer input (Sahoo et al., 1998). Tan et al. (1995) reported a statistically significant yield decline in the WS, but a statistically significant yield increase in the DS in a rice–rice LTE in the Mekong Delta.
Several authors have analyzed yield trends in long-term rice-wheat experiments. An AICRP rice-wheat-cowpea LTE at Pantnagar, Uttar Pradesh, was established in 1971 on a virgin Mollisol. Yields of rice declined from high initial levels, but there was no substantial decline in wheat yields (Nand Ram, 1998). No statistical analysis was reported. In a rice-wheat LTE at Masoda, Uttar Pradesh, yield declines were reported for wheat (−62 kg ha−1 per year) and rice (−25 kg ha−1 per year), but trend lines were fitted using three-year moving averages of yields and no significance testing was reported (Yadav et al., 1998). Brar et al. (1998), Singh et al. (1988) and Regmi (1994) also reported yield declines in the rice phase of long-term rice-wheat experiments in India and Nepal, mostly caused by depletion of soil nutrients, but with no thorough statistical analysis.
Given the large number of long-term experiments in Asia, the available evidence in the literature regarding trends of grain yield under intensive rice cropping is insufficient and overly focused on experiments at IRRI. It is also unclear whether yield declines are more common in experiments where yields are initially high. If true, it might indicate a lack of sustainability of intensive rice cropping at the higher yield levels that will be needed to meet future rice production needs. In this paper, two questions will be addressed. First, how widespread are statistically significant yield declines in LTEs conducted in tropical and subtropical Asia? Second, is the magnitude of the yield decline inversely correlated with initial yields in the experiment? The extent of yield or productivity declines in farmers’ fields is beyond the scope of this paper.
Section snippets
Long-term experiment data
The data used for this analysis come from 30 LTEs with rice–rice and rice-wheat cropping systems conducted at 24 different sites in China, India, Indonesia, Bangladesh, Vietnam, the Philippines, and Malaysia, and they represent a wide variety of soil types (Table 1, Table 2). These LTEs represent all the experiments for which we were able to obtain data sets of at least nine years. It should be noted that there are other long-term experiments of sufficient duration for which we were unable to
Yield trends in rice–rice experiments
In the DS, only two of the 21 data sets from the 15 sites outside of IRRI showed declining yield trends that were statistically different from zero at a 5% level of significance or less (Table 3, Fig. 1). Eight of these 21 data sets showed positive yield trends, with one being statistically different from zero (Omon-LTFE, +4.63% per year). In contrast, all three DS experiments at IRRI had statistically significant yield declines ranging from −1.45 to −1.61% per year. For example, the total
Conclusions
During 10 to 30 years of intensive rice cropping, yields in the most favorable climatic season have remained stable in many LTEs conducted outside of IRRI, mostly at yield levels of 4 to 7 Mg ha−1 with recommended rates of N, P and K applied. Yield declines in the less favorable climatic season (WS) were more abundant than those in the more favorable season (DS), but only a small number of these declines in either season outside of IRRI were statistically significant. There was an inverse
Acknowledgements
We would like to acknowledge the valuable contributions made by many researchers conducting the long-term experiments, helping us to obtain data, or providing comments on an earlier draft of this paper, including S.K. De Datta, K.G. Cassman, D. Olk, J.P. Descalsota, J. Alcantara, E. Laureles, and J.K. Ladha (IRRI), E.M. Imperial (BIARC), J.B. Bajita (PhilRice), Djafar Baco, R. Le Cerff, and R. Mufran (Maros and Lanrang), Li Jiakang and Liang Guoqing (Soil & Fertilizer Institute, Chinese Academy
References (42)
- et al.
Trends in irrigated-rice yields under intensive cropping at Philippine research stations
Field Crops Res.
(1984) - et al.
An analysis of long-term rice yields in a wetland soil
Field Crops Res.
(1982) - et al.
Effect of 12 years’ rice/wheat cropping sequence and fertilizer use on soil properties and crop yields in a sodic soil
Field Crops Res.
(1989) - et al.
Yield trends of a long-term NPK experiment for intensive rice monoculture in the Mekong river delta of Vietnam
Field Crops Res.
(1995) - Bisht, P.S., Pandey, P.C., Lal, P., 1990. Changes in soil fertility, crop productivity and nutrient status in long-term...
- Brar, B.S., Yadvinder-Singh, Dhillon, N.S., Singh, Bijay, 1998. Long-term effects of inorganic fertilizers, organic...
- BRRI, 1995. Annual report. Bangladesh Rice Research Institute, Gazipur,...
- Cassman, K.G., De Datta, S.K., Olk, D.C., Alcantara, J., Samson, M., Descalsota, J.P., Dizon, M., 1995. Yield decline...
- et al.
Scientific evidence of yield and productivity declines in irrigated rice systems of tropical and subtropical Asia
FAO Int. Rice Comm. Newsl.
(1997) - Cassman, K.G., Pingali, P.L., 1995. Extrapolating trends from long-term experiments to farmers fields: the case of...
Changes in yield response to major nutrients and in soil fertility under intensive rice cropping
Soil Sci.
Fertilizer inputs, nutrient balance and soil nutrient-supplying power in intensive, irrigated rice systems II. Effective soil K-supplying capacity
Nutr. Cycl. Agroecosyst.
Fertilizer inputs, nutrient balance and soil nutrient-supplying power in intensive, irrigated rice systems III. Phosphorus
Nutr. Cycl. Agroecosyst.
Scale dependent correlations among soil properties in two tropical lowland ricefields
Soil Sci. Soc. Am. J.
Fertilizer inputs, nutrient balance, and soil nutrient-supplying power in intensive irrigated rice systems I. Potassium uptake and K balance
Nutr. Cycl. Agroecosyst.
Yield stability and fertilizer efficiency of long-term triple cereal cropping in paddy fields of China
Biol. Fertil. Soils
Cited by (195)
Comparative proteome analysis of phosphorus-responsive genotypes reveals the proteins differentially expressed under phosphorous starvation stress in rice
2023, International Journal of Biological MacromoleculesThe economic sustainability of rice farming and its influence on farmer decision-making in the upper Mekong delta, Vietnam
2023, Agricultural Water ManagementFloods, soil and food – Interactions between water management and rice production within An Giang province, Vietnam
2021, Agriculture, Ecosystems and EnvironmentDistribution and determinants of organic carbon and available nutrients in tropical paddy soils revealed by high–resolution sampling
2021, Agriculture, Ecosystems and Environment