HomeInternational Rice Research Notesvol. 28 no. 2 (2003)

Yields at IRRI Research Farm are Still Close to the Climatic Potential Level

M. J. Kropff | K. G. Cassman | H. H. Van Laar

Discipline: Agriculture

 

Abstract:

Highest dry-season yields obtained at the IRRI Research Farm declined from 9-10 t ha–1 in the late 1960s and early 1970s to less than 7 t ha–1 in the late 1980s; the comparable yield decline in the wet season was from about 6 to 4 t ha–1 during the same period (Cassman et al 1995). In 1991, research teams led by Cassman, Kropff, and Peng initiated investigations into the cause of the yield decline at IRRI. In the wet season of 1991, wet-season yields of 6 t ha–1 were achieved with modified N management to improve the congruence between N supply and crop demand (Cassman et al 1994), and these yields were comparable with wet-season yield levels achieved in the 1960s and early 1970s. Based on these results, Kropff et al (1994a) developed, parameterized, and evaluated the ORYZA1 simulation model for yield potential in rice. Based on simulations from this model, Kropff, Cassman, and van Laar (Kropff et al 1994a, b) predicted that, in most years, dry-season yields could be increased substantially with improved N management. Using a historical weather database from the IRRI Climate Unit, the model predicted that the dry-season yield potential would range from 8.5 to 10 t ha–1 in 8 of 10 years with a mean potential yield of about 9.3 t ha–1, when the crop was transplanted in early January, which is the optimal transplanting date to achieve maximum yield at this site (Kropff et al 1993, 1994b). In the dry season of 1992, a yield of 9.5 t ha Highest dry-season yields obtained at the IRRI Research Farm declined from 9-10 t ha–1 in the late 1960s and early 1970s to less than 7 t ha–1 in the late 1980s; the comparable yield decline in the wet season was from about 6 to 4 t ha–1during the same period (Cassman et al 1995). In 1991, research teams led by Cassman, Kropff, and Peng initiated investigations into the cause of the yield decline at IRRI. In the wet season of 1991, wet-season yields of 6 t ha–1 were achieved with modified N management to improve the congruence between N supply and crop demand (Cassman et al 1994), and these yields were comparable with wet-season yield levels achieved in the 1960s and early 1970s. Based on these results, Kropff et al (1994a) developed, parameterized, and evaluated the ORYZA1 simulation model for yield potential in rice. Based on simulations from this model, Kropff, Cassman, and van Laar (Kropff et al 1994a, b) predicted that, in most years, dry-season yields could be increased substantially with improved N management. Using a historical weather database from the IRRI Climate Unit, the model predicted that the dry-season yield potential would range from 8.5 to 10 t ha–1 in 8 of 10 years with a mean potential yield of about 9.3 t ha–1, when the crop was transplanted in early January, which is the optimal transplanting date to achieve maximum yield at this site (Kropff et al 1993, 1994b). In the dry season of 1992, a yield of 9.5 t ha–1 was obtained with IR72, and a hybrid variety yielded 10.7 t ha–1 under improved N management (Kropff et al 1994b). In a subsequent study, Dobermann et al (2000) confirmed the role of improved N management in restoring yields close to yield potential levels at the IRRI Research Farm.