| Rice is an important food crop in China and has a great significance for the national food security and stability. But currently, rice is facing the influence of different meteorological disasters, especially the high temperature. Suitable sowing date selection can effectively reduce high temperature damage to achieve stable rice yield. Therefore, two temperature index (the highest temperature?35? and the average temperature?30? for 3 days or more) were chosed in the analysis of temporal and spatial distribution of high temperature condition in 7 stations in Jiangsu from 1966 to 2105. Then, based on the experimental data in 2012 and 2013, influence of sowing dates on rice growth and yield were analyzed as theoretical support of suitable sowing date selection. On the basis of the above analysis,photosynthesis?respiration, and distribution coefficient of dry matter allocated to panicle were modified to improve the ORYZA2000 model. Finally, using the improved and validated model to simulate the yields of rice under 4 sowing dates in Nanjing and Wuxiandongshan where high temperature are relatively serious during 1966 to 2015 to evaluate the high temperature situation. Based on that, references can be provided while making decisions in suitable sowing date selection. The main conclusions are as followed:(1) The probability of critical temperature of highest temperature ?35?and average temperature?30?in Jiangsu during 1966-2015 were basically between 125?250d. The probability of highest temperature ?35? was lower than the probability of average temperature?30? between 175?225d,and but for Xuzhou station, as the date get closer to 210d, the difference between the two was getting larger. The days and accumulated temperature of maximum temperature ?35? and average temperature?30?for 3 days and more in rice growth season (4-10 months) over the past 50 years were consistent in different stations. The highest value appeared in July. The overall variation trend of temperature was getting warmer and average temperature index was more obvious. Accumulated temperature centers were in the southwest of Jiangsu, and had the tendency of moving to the southeast with the pass of time. The moving range of the accumulated temperature center of maximum temperature was larger than average temperature.(2) The differences of sowing dates caused the differences of meteorological factors in same growing stages. The high temperature trend of the two index were consistent in different sowing dates. In the high temperature sensitive stage,the 1st sowing date in 2012 suffered a mild heat damage, while the 2nd and 3rd sowing date in 2013 both suffered a severe heat damage and the 4th a mild. In heading stage, only the 2nd and 3rd in 2013 suffered a moderate heat damage for 7 days. Differences of meteorological factors in same growing stages of different sowing dates also caused the the differences in length of growing stages and yields. Correlation analysis showed that the differences of yields between the first two and last three sowing dates was significant(p < 0.05) both in 2012 and 2013.Analyzing the relationship between the length of growing stage, yield and maximum temperature,average temperature,precipitation and radiation of each stage and the results showed that temperature, especially maximum temperature was the most obvious.(3) Modifed the model by adjusting the relationship between maximum photosynthetic rate and temperature, modifying the leaf temperature and panicle temperature in respiration rate and correcting the dry matter partition coefficient of panicle under high temperature. After modifying the model, the simulated results of model before and after modification were compared with the measured. The results showed that the root mean square error(RMSE) and normalized root square error(NRMSE) of panicle biomass, green leaf biomass,dead leaf biomass, LAI and yield were decreased in the modified model,while the decrease of NRMSE of yield and panicle biomass were relatively significant for 2.23% and 1.83% respectively. From the 1:1 diagram of the simulated and measures values, that total aboveground biomass was improved in 2013 while the linear regression coefficient of the modified model was more closer to 1. Then,yields were simulated under actual and perennial weather in Nanjing and Wuxiandongshan of 4 sowing dates(120d, 130d, 140d and 151d) from 1966 to 2015 with the modified model. From the entire 50 years, yields of 120d were affected most and the overall decrease was most pronounced, while the fluctuation of yield change rate and yield decline of 151d was the least obvious. So, sowing date around 151d would be more suitable to achieve stable yields in Nanjing and Wuxiandongshan areas. |
PDF Full Text Request