Occurrence and modeling of un-ripened grain under high temperature and shading condition

Abstract

Evaluated temperature caused by global warming would affect grain filling in rice. Especially, High temperature during grain filling period reduces grain filling duration and increases potential grain growth rate. This effect could be complemented by higher assimilate supply. In other words, grain filling is determined by interaction between sink and source strength. For evaluating the climate change impact on rice yield effectively, it is important to develop a model for the estimating the filled grain ratio considering this interaction. The objects of study are to observe effects of high temperature and shading condition during ripening stage on rice yield and yield component and to develop a model of estimation filled grain ratio. The two rice cultivars Odaebyeo (early maturing) and Hwasungbyeo (medium-maturing) were pot(1/5000a) cultured in a green house controlled to ambient temperature (AT) until heading, and pots were transferred to greenhouses controlled to the target temperatures of AT, AT+1.5, AT+3, and AT+5. A half of pots in each house was subjected to shading treatment with shade net of about 50% transmittance. Mean and distribution of panicle heading date were not statistically different among treatments, and also significant difference was not found in panicle number and spikelet number among temperature and shading treatments. 1000-grain weight showed no significant difference among temperature treatments, while it was decreased significantly by shading treatment. Spikelet fertility tended to decrease with temperature elevation and shading treatments. Ratio of filled grain showed significant decrease with temperature elevation and shading treatment. In summary, heading date distribution, panicle number, and spikelet number were not affected by temperature and radiation condition during grain filling stage. 1000-grain weight was affected more substantially by shading condition rather than by high temperature condition. Decrease in 1000-grain weight due to shading treatment may be attributable to the quick exhaustion of pre-heading reserve carbohydrate compared to no-shading treatment. Spikelet sterility was affected by high temperature but the sensitivity was less than the occurrence of un-ripened grain. Shading condition reduced spikelet sterility, too. The decrease would be assumed to be caused by flowering time delay or hormone action. But we need more study about this phenomenon. Un-ripened grain occurrence was affected by temperature and solar radiation. An Un-ripened grain occurred with higher in Odaebyeo than in Hwasungbyeo. The higher occurrence of unfilled grain in Odaebyeo would have been caused as pre-heading reserve carbohydrate was accumulated less due to shorter vegetative period than Hwaseongbyeo. Base on the literature review and this result, we developed a model for predicting un-ripened grain ratio. We assumed that un-ripened grains occur when carbohydrate supply strength (source strength) is not sufficient to compensate the sink activity increase with temperature increase. Sink activity is defined as the potential grain growth rate that is derived from the potential grain growth curve. The following potential grain growth curve was formulated by incorporating the notion, that grain growth varies according to growth temperature during grain filling and grain position on a panicle which determines the priority of flowering, to the existing sigmoid potential grain model.

where a is maximum grain weight. b, c are coefficient that is related to grain filling duration and rate. b1, and c1 are coefficient to distinctly calculate grain weight of superior and inferior spikelet.

Potential growth of grains on a panicle was computed on each day after intial flowering by multiplying the probability density distribution of flowering date on a panicle by potential growth of grains differing in flowering date on a panicle. Sources are supposed to be composed of current photosynthesis after heading and pre-heading carbohydrate reserve. Pre-heading carbohydrate reserve was assumed to act as buffer that compensate the lack of current photosynthate. Photosynthesis was calculated as a function of solar radiation, temperature and leaf senescence after heading. Grain density of rice is mainly determined by grain weight because the potential volume of grain is unchanged due to the pre-determined glume size. Therefore. The grain that has weight lower than a threshold weight is determined as un-ripened grain. This model was calibrated with the temperature elevation and shading experiment data using Simplex method. The calibrated model predicted the treatment means of temperature elevation and shading treatments with acceptable accuracy. However, this model showed some limitations in detailed prediction of unripened grain ratio for each panicle within treatments.