| Appropriate sowing date can ensure optimal thermal and radiation resources for crop development and growth,and achieve suitable growth standard for crop canopy at critical growth stages.Therefore,adjusting sowing date has become one of most important agronomic practices for high and stable crop productivity.Due to climate variation and the changing types of rice variety,large area and ratio of wheat were sown later than the optimal sowing date in Jiangsu Province,and this have huge impact on regional wheat production.Therefore,it’s urgent to quantify the impacts of late sowing at wheat productivity under different climate and cultivation patterns.Due to the advantages on mechanism,process,and comprehensiveness,crop growth models are widely used in the prediction of regional crop productivity.In this study,based on observations from different sowing date field experiments at four typical stations across the main wheat producing region of Jiangsu Province,the responses of four crop growth models(WheatGrow,CERES-Wheat,Nwheat and APSIM-Wheat)to different sowing date in simulating wheat phenology,yield,grain protein content at maturity(GPC)and partial factor productivity of nitrogen fertilizer(PFP)were evaluated.Then,using scenario simulation approaches and different climate scenarios(baseline period,1.5~℃ HAPPI and 2.0~℃ HAPPI),the spatial variation for the effects of late sowing on yield,grain protein content at maturity and partial factor productivity of nitrogen fertilizer under potential yield and designed target yield condition were quantitatively analyzed in the main wheat producing region of Jiangsu Province.The results of this study can be used to optimize the management practice of wheat following rice,and provide theoretical support for designing adaptive strategies under climate change conditions,and have significant implications for ensuring the high and stable production of winter wheat in Jiangsu Province under future climatic conditions.The results of this study are as follows:(1)When postponing sowing date for different wheat varieties at different ecological stations,wheat growth period were shortened,and grain yield and partial factor productivity of nitrogen showed decreasing trends.The protein content of wheat grains showed significant variations with the sowing date,but didn’t have uniform patterns among different treatments.Multi-model ensemble could simulate the response of flowering date,maturity date,yield and partial factor productivity of nitrogen to different sowing dates reasonably.The standardized root mean square error(NRMSE)for flowering date,maturity date,yield and partial factor productivity of nitrogen predicted by different models were 1.42%-5.42%,1.35%-3.06%,12.87%-19.53%,and 9.31%-17.41%,respectively.Multi-model ensemble significantly improves the simulation accuracy of wheat phenology,yield and partial factor productivity of nitrogen under different sowing dates,this indicated the multi-model ensemble method shown better predictability and applicability across different ecological stations and different varieties.However,the simulations of wheat grain protein content at maturity were poor,regardless of the individual model(NRMSE for WheatGrow=20.16%,CERES-Wheat=22.69%,Nwheat=22.55%and APSIM-Wheat=15.99%)or multi-model average(NRMSE=14.93%),and this implied that the algorithms for predicting nitrogen dynamic and formation of grain protein from all four models would need to be improved significantly.(2)Under potential and designed target conditions,the absolute and relative negative effects of winter wheat yield and partial factor productivity of nitrogen fertilizer decreased gradually with to delay of sowing day in the entire planting area of Jiangsu Province,except for individual Wheat Regions and stations,the grain protein content at maturity also had the same change rule.In different climate scenarios,the relative effects of yield under all late sowing scenarios was higher than the potential condition,but the absolute effects had no such rule;However,the absolute and relative effects of light temperature potential on grain protein at maturity were greater than that of designed target condition.The spatial distribution of the absolute and relative effects of late sowing on wheat yield and partial factor productivity under the baseline period and future warming scenarios were relatively consistent,showing a slight"high in the South and low in the north",while the spatial distribution of the effect of grain protein content at maturity showed a significant"low in the South and high in the north".Under the designed target condition,the relative effect of late sowing of 10,20,30,40,and50 days in the entire producing region for the baseline period were 6.8%,13.3%,18.5%,22.2%and 25.0%for yield,were 3.7%,6.4%,8.3%,9.4%and 9.9%for grain protein content at maturity,and were 6.8%,13.3%,18.5%,22.2%and 25.0%for partial factor productivity of nitrogen.Compared with baseline period,when the number of days for late sowing were small,future climate warming was projected to slightly reduce the loss of yield and partial factor productivity of nitrogen caused by late sowing,but when the number of days for late sowing exceeded a certain number of days future climate warming was projected to slightly increase the negative effects of late sowing,and the loss of grain protein content at maturity will be reduced in any late sowing scenario.The differences in the effects of late sowing on winter wheat productivity were not significant between the 1.5~℃ HAPPI and 2.0~℃ HAPPI scenario.Under same sowing date,the differences in the effects of late sowing between the two warming scenarios were generally less than 1.0%.Under the same climate scenario,the effects of the same late sowing day were significantly different in different subregions.The effect of the same late sowing on grain yield and partial factor productivity of nitrogen were the largest in Taihu wheat producing subregion,and the smallest in Huaibei wheat producing subregion.For grain protein content at maturity,the pattern is reversed. |
PDF Full Text Request