Low temperature, particularly during the reproductive stage of the development of rice, limits productivity in the Riverina region of New South Wales (NSW). This study primarily examined genotypic differences in cold damage that are associated with low temperature during reproductive development. The objectives were to: (1) investigate the effects of low temperature on physio-morphological traits of rice plants, with particular emphasis on reproductive traits; (2) examine the consistency of expression of cold tolerance in different screening environments; and (3) quantify the effects of temperature and daylength on the phenological development among
Results from experiments in temperature-controlled rooms and the cold water facility were combined with those from four years of field experiments, which used natural exposure to low temperature to examine the response of over 50 cultivars from diverse origins, including cold tolerant cultivars from Eastern Europe, Japan and California. Plants were exposed to day/night air temperatures of 27/13̊C in temperature-controlled rooms and also exposed to a constant temperature of 19̊C in the cold water facility from panicle initiation (PI) to 50% heading. In field experiments several techniques were used to increase the likelihood of inducing cold damage
such as multiple sowing dates (five to eight sowing dates each year), shallow water depths (5cm) and high nitrogen rates (e.g. 300kgN ha-1).
There was significant genotypic variation in several traits such as anther length, anther area and number of engorged pollen grains per anther when measured at flowering. These flowering traits were negatively related to spikelet sterility at maturity. Cultivars originating from Australia and California were inefficient at producing filled grains when low temperature coincided with reproductive development, despite cultivars having a similar number of engorged pollen grains and similar sized
anthers to those from other regions. It is suggested that this inefficiency may partly be related to a small stigma area.
The different screening methods used in the temperature-controlled rooms, cold water facility and field induced sufficient levels of spikelet sterility to identify genotypic differences and demonstrate consistency in cold tolerance. There was a highly significant relationship for spikelet sterility among the common cultivars in temperature-controlled room versus cold water facility (r2=0.63, p<0.01, n=21), temperature-controlled room versus field (r2=0.52, p<0.01, n=31)
and cold water facility versus field (r 2=0.53, p<0.01, n=21). Screening for cold tolerance in a temperature-controlled room or the cold water facility was preferred to field screening because of the improved reliability of exposure to low temperature in both environments. However, while there was a consistent genotypic response to low temperature in these screens, some cultivars varied in their response under the different screens suggesting that it is important to combine a controlled environment screen with field observations.
In each screening environment, several cold-tolerant cultivars (e.g.
M103, Plovdiv 22 and HSC55) and cold-susceptible cultivars (e.g. Sasanishiki, Doongara and Nipponbare) were identified. Many of the cold-tolerant cultivars identified in this study had shorter growth duration and hence a lower yield potential compared to commercial cultivars. Therefore, to determine whether cold tolerance could be improved while increasing growth duration for the development of adapted cultivars for the NSW rice-growing environment, the cultivars HSC55 and Plovdiv 22 were each hybridised with two NSW commercial cultivars, Illabong and Millin. The progeny (206 F5:7lines) were then evaluated for cold tolerance in temperature-controlled rooms. There was no relationship between growth duration and spikelet sterility. It was concluded that cold tolerance could be improved while maintaining or slightly reducing growth duration of existing cultivars.
Therefore, it should be possible to produce cold-tolerant cultivars with appropriate growth duration for Australian conditions.
The sequential sowing dates in the field that were used to expose plants to low temperature under natural conditions, also provided 24 different conditions of temperature and daylength in which phenological development could be examined. Cultivars such as Amaroo and Millin were identified as mildly photoperiod-sensitive, while M103 and HSC55 were identified as photoperiod-insensitive. Crop phenology models to describe development period from sowing to flowering and from PI to flowering were
developed for seven cultivars using field data. The model for Amaroo was also used to predict an optimum sowing date based on historical weather data from 1955 to 2002 that minimises the possibility of exposure to low temperatures during the young microspore and flowering stages. The analysis suggested that 15 October was the optimal sowing date for Amaroo, and sowing up to November I would also have a low risk of encountering low temperatures when seasonal temperatures were average. Increasing the photoperiod-sensitivity of cultivars, above that of Amaroo, may further reduce the risk of encountering low temperatures while at the same time increase sowing flexibility.
Findings presented in this study have improved our understanding of mechanisms of genotypic response to low temperature, and provided ways to develop cold-tolerant cultivars.