| Rice(Oryza sativa L.)is one of the great significance staple food in China.More than 65%of the population are feed on rice.Therefore,stable growth of rice productivity plays a key role in ensurance of national food security.Sowing date is the most simple and effective cultivation measures.It directly determines the rice growing duration temperature and radiation resources can be better use,is the high yield and high quality basis of rice.Optimum sowing date can effectively use the temperature and radiation resources,bring into full play the yield potential of rice varieties.Therefore,the selection of appropriate planting date will be able to effectively and efficiently use the local temperature and radiation resources,give full play to the potential of yield,and thus achieve the purpose of improving rice production status.In this study,these three rice growth models(RiceGrow,ORYZA2000 and CERES-Rice)used as the research tools to explore the effects of different ecological sites and sowing dates on the growth duration and yield of rice in China.And the trend of the optimum sowing date in future climatic conditions.First of all,this study used the field experiment data from 1981 to 2011 of rice in single-season rice region and double-season rice region of China,which selected 34 single-season rice cultivars and 44 double-season rice cultivars(24 early rice and 20 late rice)with more than 6 times from 2000 to 2011.Rice growth models(RiceGrow,ORYZA2000 and CERES-Rice)were calibrated and validated by the selected cultivars involving different varieties,different years and different sowing date treatments in single-season rice region and double-season rice region.Then with the well evaluated model,RiceGrow,ORYZA2000 and CERES-Rice,this study aimed to investigate the effects of different sowing date on rice growth days and yield through designing different sowing date changes at each studied eco-sites in single-season rice region and double-season rice region.At the same time,the future climate data generated by HadGEM2-ES climate model in CMIP5.The rice growth models were used to simulate the sowing date of single-season rice,double-season early rice and double-season late rice in the 2030s(2020-2040),the 2050s(2041-2000)and the 2070s(2061-2080),and the change of optimum sowing date in future climatic scenario was studied.These three rice growth models(RiceGrow,ORYZA2000 and CERES-Rice)were calibrated and validated by field experiment data of different rice cultivars in single-season rice region and double-season rice region.The results showed that the RMSE and NRMSE were 2.91 d and 1.32%at the flowering stage in single-season rice of the RiceGrow model,and the RMSE and NRMSE at the mature stage were 2.94 d and 1.13%,respectively.The RMSE and NRMSE were 2.14 d,1.24%and 3.25 d,1.24%,respectively.The RMSE and NRMSE of early rice and late rice were 2.80 d,1.40,and 1.80 d,respectively,and 670.60 kg ha-1 and 7.81%.And 2.79d,0.94%,the RMSE of early rice was 522.14 kg ha-1 and NRMSE was 6.75%.The RMSE of late rice yield was 542.814 kg ha-1 and NRMSE was 8.54%.In addition,these three rice growth models(RiceGrow,ORYZA2000 and CERES-Rice)showed a quite good performance both in predicting rice growth days and yield in different agro-ecological conditions and sowing dates in single-season rice region and double-season rice region in China.Based on daily meteorological data of 254 meteorological sites in single-season rice region in China,three rice growth models were used to simulate the effect of sowing date of 316 sites in single-season rice region from 1981 to 2011.Finally,this study determined the optimum sowing date and suitable sowing date of different ecological sites.Furthermore,generated the HadGEM2-ES climate model in CMIP5 the future climate data,in the 2030s(2020-2040),the 2050s(2041-2000)and the 2070s(2061-2080)combined with the rice growth model to predict the optimum sowing time.The results show that the optimum sowing date is gradually delayed from south to north and from west to east due to the difference of temperature and radiation resource.in the southern Yunnan valley basin single season rice sub-region(?2),Yangtze River middle and lower reaches of the plain single,double-season rice area(?1),Sichuan-Shaanxi Basin Single-season Rice(?2),Qiandong Xiangxi Plateau mountain single,double-season rice sub-area(?1),Two-cropped sub-region of single-season Rice in the Ridge and Valley of Dianchuan Plateau(?2),Heji plain river valley special precocity suburb(?1),Liaohe coastal plain precocity suburb(V2)and the optimal simulations of the three models in each sub-region were:March 10th,May 8th,March 21st,April 2nd,March 30th,April 15th and April 10th.The average yield of the corresponding optimum date was:10600.6 kg ha-1,10641.8 kg ha-1,11183.6 kg ha-1,9599.7 kg ha-1,1076.8 kg ha-1,10484.9 kg ha-1 And 109892 kg ha-1 were 30.5%,12.0%,17.5%,19.9%,24.5%,13.7%and 22.5%respectively compared with the actual sowing date.In addition,HadGEM2-ES global climate scenario model 2030s,2050s,2070s in each sub-area sowing date respectively in advance:?2 sub-region 3.5 days,5.8 days and 8.4 days;?1 sub-region 6.7 days,12.5 days and 18.6 days;0.5 days,2.7 days and 4.9 days;?1 sub-region 2.3 days,4.8 days and 7.6 days;?2 sub-region 1.32 days,2.80 days and 5.01 days;?1 sub-region 1.3 days,2.8 days and 5.0 days;?2 sub-region 3.9 days and 6.7 days.The result of simulating the optimum sowing date is compared with the actual sowing date,the optimum sowing date is different degrees of advance,just delaye in a few of areas.Compared simulating yield between the actual sowing date and the optimum sowing date found that by adjusting the sowing date can achieve the purpose of increasing production in different ecological sites and yield potential is different.Three rice growth models(RiceGrow model,ORYZA2000 model,CERES-Rice model)was used to simulate the effect of growth duration and yield of the 183 sites of early rice and late rice from 1981 to 2011,respectively.The optimum sowing date and suitable sowing date of early rice and late rice were determined.Finally,the future climate data generated by the HadGEM2-ES climate scenario in CMIP5 was used to simulate the effect of future climate change on the optimum sowing date of early rice and late rice.The results show that the optimum sowing date of early rice is gradually delayed from south to north due to the difference of temperature and radiation.The simulated mean values of the three models were:March 2nd,March 1st,February 28th and March 21st,March 20th,March 25th,respectively.Late rice,the optimal sowing is the gradual change from south to north in advance.The simulated mean values of the three models were as follows:July 9th,July 8th,July 11th and June 23rd,June 22nd,June 24th,respectively.The average yield of early rice was 9885.0 kg ha-1 and 10024.3 kg ha-1,respectively.Compared with the actual sowing date,the simulated yield increased by 35.4%and 20.1%,respectively.The average yield was 8971.8 kg ha-1 and 9408.0 kg ha-1,respectively.The actual yield was 30.2%and 17.9%respectively.In addition,HadGEM2-ES global climate model in 2030s,2050s,2070s in each sub-region early rice and late rice sowing were earlier:1.3 days,3.5 days,5.6 days and 4.2 days,7.1 days,10.1 days;2.1 days,4.2 days,6.4 days,5.1 days,7.4 days,9.8 days.The results of this study will provide some guidance and reference for the application of the varieties and the selection of suitable sowing date. |
PDF Full Text Request