Grain quality of flooded rice is affected by season, nitrogen rate, and plant type

Abstract

Quality of grain, next to yield, is the most important factor for rice (Oryza sativa L.) production in semi-arid tropical Australia. Studies were undertaken in the Burdekin River Irrigation Area of northern Australia to improve rice grain quality through nitrogen fertilisation. This paper reports the results of 4 experiments comparing the response of 3 rice genotypes differing in maturity and stature to 5 rates of applied nitrogen (0, 70, 140, 210, and 280 kg/ha) over 4 seasons (2 wet and 2 dry seasons). The components of grain quality studied were endosperm chalkiness, whole grain millout, grain size, alkali digestion (gelatinisation temperature), and grain protein. This paper also examines the suitability of the 3 genotypes as parental material in breeding programs aimed at selecting for specific grain quality attributes.

Starbonnet was identified as a potential parent in breeding programs which aim to specifically select for reduced chalkiness and high millout in low N environments. Selection for lower chalkiness, and higher millout and protein concentration, should occur in a wide range of target environments to account for the seasonal variation observed in these parameters. Grain size appeared to be affected more by genetic than agronomic factors, since grain length and breadth were largely unaffected by N rate, yet genotypic differences were found for both parameters in all experiments. Newbonnet grain was long and slender, suggesting this genotype would be a suitable parent in breeding programs aimed at improving grain appearance. The response of alkali digestion to N rate and genotype was small for all seasons.

The importance of developing N fertiliser strategies that optimise both grain yield and quality was highlighted by differences in the responses of grain protein and grain yield to N rate. A number of linkages were examined among various components of grain quality. However, the magnitude of these linkages was small, suggesting that selection for one quality component should not be at the expense of selection for another.