Modelling Crop Yield And Nitrogen Balance Based On Crop-Soil Models

Irrational agricultural management practices（cropping systems,fertilization,tillage etc.）not only affect crop production but also disturb nitrogen balance which can increase potential water and atmospheric pollution.In addition,climate change can directly impact crop yield and nitrogen cycling in agroecosystems.Therefore,improving agricultural management practices and exploring adaptation measures under future climate scenarios are crucial to maintain high crop production,decrease environmental risks and then achieve sustainable development of agriculture.In this study,we evaluated the effects of different cropping systems,management practices and climate scenarios on crop yield,agricultural nitrogen cycling as well as environmental risks usingDSSAT（Decision Support System for Agrotechnology Transfer）,DNDC（Denitrification-Decomposition）and a developed nitrogenbalance model based on CANB（Canadian Agricultural Nitrogen Budget）and OECD（Organization for Economic Cooperation and Development）models.The main conclusions as follows:1.Sensitivity analysis of DSSAT modelling.Based on two long-term field experiments（spring wheat at Swift Current and maize at Woodslee）,DSSAT sensitivity analysis was explored to assess the impacts of the input parameters on crop yield,soil water and nitrogen leaching.The input parameters include（1）soil drainage upper limit（DUL）,（2）soil drainage lower limit（LL）,（3）planting date（PD）,（4）plant population（PP）,（5）fertilizer N rate（FNR）and（6）daily precipitation（PREC）.The order of input parameters that influenced wheat grain yields and above-ground biomass at Swift Current was PREC>FNR>DUL>PD>LL>PP,whereas maize yields and above-ground biomass at Woodslee followed the pattern FNR>PREC>DUL>PD>LL>PP.At both locations,soil water storage was found to be sensitive to DUL and PREC,soil N leaching（below 1.2m）was sensitive to DUL,FNR and PREC parameters.At Swift Current,average N leaching losses over 39 years were less than 7 kg N ha^-1 under long term continuous wheat cropping system using the default parameters.However,when FNR was set up to 90 kg N ha^-1,N leaching could reach 20 kg N ha^-1.At Woodslee,average soil N leaching losses under normal years was less than 5 kg N ha^-1 for maize field using the default parameters.However,higher soil N leaching was found in the wet year of 2011(20-85 kg N ha^-1).These results suggested that fertilizer N application rates could be increased to 90-110 kg N ha^-1for spring wheat at Swift Current and to 150-210 kg Nha^-1for maize at Woodslee to obtain maximum yields.2.Modelling the impacts of climate change on spring wheat and maize yields as well as N leaching based on DSSAT model.Climate change impacts on spring wheat and maize yields,soil nitrogen leaching,and exploring potential adaptation measures were determined using the DSSAT model at Swift Current and Woodslee.Climate scenarios（RCP4.5 and RCP8.5）were generated by the Can RCM4 model for baseline（1971-2000）and future periods 2050s（2041-2070）and 2080s（2071-2100）.Simulation results indicated that the spring wheat yields increased by 8,8,11,15%and maize yields decreased by 15,25,22and 41%under RCP4.5 2050s,RCP4.5 2080s,RCP8.5 2050s and RCP8.5 2080s scenarios,respectively.More N leaching occurred during the growing season at Swift Current and during the non-growing season at Woodslee.Annual average N leaching increased significantly by 19,57,73,129%at Swift Current and 84,117,208,317%at Woodslee under RCP4.52050s,RCP4.5 2080s,RCP8.5 2050s and RCP8.5 2080s climate scenarios,respectively.Increasing nitrogen fertilizer from 50 to 100 kg N ha^-1could significantly increase wheat yield but with increased N leaching at Swift Current.The fertilizer N rate of 150 kg N ha^-1is required to obtain high maize yields with low N leaching under future climate conditions at Woodslee.Early planting could contribute to a higher wheat yield under baseline and future RCP8.5 scenarios at Swift Current whereas late planting date could reduce the negative impacts of climate change on maize production at Woodslee.3.Calibration and evaluation of a regionalized DNDC model.The DNDC model was calibrated and evaluated using measured crop yields,soil temperature,moisture and N₂O emissions for a winter wheat-maize-soybean rotation system（1999-2006）under conventional and no tillage practices at Woodslee,Ontario,Canada.For CT treatments,the statistical values of d index,NSE and nRMSE（i.e.,d index ranged 0.92-0.98,nRMSE ranged from 0.69-0.93 and NSE ranged from 3.4-6.4%）indicated“excellent”agreements for all three crop yields.The NT treatments also showed“good”to“excellent”performance.For example,the values of d index,nRMSE and NSE ranged from 0.86-0.94,0.54-0.69and 5.5-11.8%,respectively.In addition,the DNDC model provided a reasonable prediction of soil water contents（0-0.1 m）,but it overestimated water contents during dry conditions.The statistics showed“good”agreement between the simulated and measured N₂O emissions.However,the DNDC model overestimated N₂O emissions especially for NT treatment（by 17.4%）in 2005.4.Estimating climate change impacts on crop yield and N₂O emission using the DNDC model.The impacts of climate change on crop yields and N₂O emissions were evaluated by comparing baseline（1971-2000）and future predications（2071-2100）under CT and NT treatments,based on a winter wheat-maize-soybean rotation at Woodslee,Canada.When CO₂ concentrations were at baseline levels in the scenarios,winter wheat yields significantly decreased by 15.8%for CT and by 15.9%for NT at RCP4.5 scenarios whereas they decreased by 32%for CT and 32.1%for NT with RCP8.5scenarios.When the effect of increased CO₂ was taken into consideration,average soybean yields significantly increased by 17.4%and 33.5%under RCP4.5 and RCP8.5 scenarios,respectively.Using the baseline CO₂ concentrations,the mean annual N₂O emissions for CT treatments significantly increased by 56.1%,for winter wheat,10.7%for maize and 44.6%for soybean for RCP4.5 whereas it increased by 87.6%,9.5%and 31.2%for wheat,maize and soybean,respectively under RCP8.5.The mean annual N₂O emissions under NT only increased for winter wheat（19.2%for RCP4.5 and 31.7%for RCP8.5）whereas no significant differences were found for maize and soybean compared to the baseline scenario.Compared to current cultivars,when the new winter wheat cultivar was evaluated,N₂O emissions were significantly reduced by 32.5%for CT and 25.1%for NT with the RCP4.5 scenario whereas they decreased by 35.9%for CT and 33.9%for NT with the RCP8.5 scenario.There were also trends for decreased N₂O emissions for new maize and soybean cultivars.5.Evaluating nitrogen balance in China’s farmland based on CANB and OECD models.Spatial and temporal variations of soil nitrogen balance were analyzed from 1984 to 2014 in China’s farmland using CANB and OECD methods.The national N input,output and surplus ranged from 26.2 Tg N(197.8kg N ha^-1)to 54.5 Tg N(315.7kg N ha^-1),20.3 Tg N(153.2 kg N ha^-1)to 41.7 Tg N(241.5 kg N ha^-1)and 5.9Tg N(44.6kg N ha^-1)to 12.8 Tg N(74.2 kg N ha^-1)respectively between1984 and 2014.The increased percentage of total N input,output and surplus gradually decreasedbetween 1980s（1984-1989）,1990s（1990-1999）,2000s（2000-2009）and 2010s（2010-2014）.The amount of nitrogen surplus in2010s was slightly lower than that of in 2000s.Compared with 2000s,the nitrogen surplus rate decreased,nitrogen use efficiency and nitrogen uptake ratio increased in2010s.The north central of China had the highest N input and output per unit area which were288.9-363.4 kg N ha^-1and 213.3-290.4 kg N ha^-1respectively whereas the lowest N input(209.2-245.6kg N ha^-1)and output(156.6-178.2kg N ha^-1)per unit area were in the southwest of China from 1990s to2010s.However,thehighest Ninput and output per unit areawere in the southeast and the middle and lower reaches of the Yangtze River,and the lowest Ninput and output per unit area were in the northeast and southwestof China in 1980s.The N loss of N₂,N₂O,NO,NH₃,leaching and runoff increased significantly from 1980s to 1990s,and then the growth rate gradually reduced from 1990s to 2000s.The north central of China had the highest lossper unit areaof N₂,N₂O,NO,leaching and runoff（higher in Beijing and Tianjin）and the highest loss of NH₃ in the middle and lower reaches of the Yangtze River（higher in Jiangsu and Shanghai）.The highest and lowest N surplusper unit area from 1980s to2010s were in the southeast(92.8-125.8 kg N ha^-1)and the northeast of China(13.1-26.8 kg N ha^-1),respectively.However,the highest total N surpluswas in the middle and lower reaches of the Yangtze River.The N surplusper unit area haddecreased trends from 2000s to 2010s exceptionfor the southeast and southwest of China.For example,the N surplusper unit area decreased by39.9%in the northeast of China.Rational reduction of chemical N fertilizer would decrease soil residual N and environmental risk.