Management strategies for maintaining rice yield within rice-based cropping systems

Abstract

A five-year study was undertaken to develop management strategies for maintaining yield of consecutive annual rice (Oryza sativa L.) crops. Stubble management technique, N application rate and N application time were all important determinants of yield; together they accounted for up to 80% of yield variance. Incorporating stubble, rather than burning it, had little initial effect on yield but, from the third crop onwards, stubble-incorporation plots out-yielded stubble-burn plots by an average 0.35 t ha−1 per crop. The higher yield was due to an average 11% increase in panicle number. Yield of unfertilized rice fell from 9 t ha−1 to 3.7 between the first and fifth crops, with much of the fall occurring between the first two crops. The highest-yielding treatment in the fourth crop (which received 105 kg N ha−1 at permanent flood (PF) plus 105 kg N ha−1 at panicle intiation (PI)_did not achieve the yield of the first rice crop receiving 50 kg N ha−1 at PF (8.3 t ha− compared with 9.6). This suggests that N applications in excess of 200 kg N ha−1 were necessary to maintain yield. Fertilizing at sowing had little effect on yield. However, fertilization with a five-year average of 94 kg N ha−1 prior to PF resulted in higher average yield (6.8 t ha−1) than that obtained from plots receiving all their fertilizer at PI (6.1 t ha−1). The difference between PF- and PI-fertilized plots became greater with increasing number of crops and was because of increased floret sterility in crops receiving all their N at PI. It was concluded that reasonable yield of annual rice crops could be maintained with a combination of stubble incorporation during autumn-winter plus relatively heavy N applications just prior to PF.