Prediction Of Carbon Sequestration Potential Of Chinese Arable Land Under Long-term Fertilizations

Soil organic carbon （SOC） is the important index as the foundation of crop productivityfor soil fertility. The SOC content is only1/3¹/2of the similar soil type of the world, with largeimprovement potential. SOC is affected by many factors, and fertilization affect the SOC changedirectly. China is a large country with a diverse range of climatic factors and managementpractices. The SOC change and its sequestration potential under different fertilizations in thediversity agro-ecosystem environment in China is still unclear. SOC change is a long, slow andcomplex process, therefore the long-term experiment combine with SOC process model has theadvantage. Forthermore, most of previous studies on SOC potential focus on single site by thestatistics analysis, and lack of the regional study. Therefore we chose14long-term fertilizationexperimental sites combine with the Rothamsted Carbon Model （RothC） to find the SOCsequestration potential under different fertilization management. The objectives were （1） evaluateand select the suitable RothC26.3model for Chinese arable land;（2）explore the SOCsequestration potential regional characteristics in Chinese arable land;（3） set manure carbon inputscenarios, to disinter the SOC evolution and the main effect factors under different fertilization inChinese arable land.The14long-term fertilization experimental sites represent three typical agro-ecosystem,upland, paddy-upland rotation, and double-rice. Seven kinds of fertilization treatments waschosen:（1） no fertilizer （Control）,（2）nitrogen （N） only,（3） N and phosphorus （NP）,（4）N,P andpotassium （NPK）,（5） manure only （M）,（6） inorganic fertilizer+manure （N+M/NP+M/NPK+M/hNPK+M）,（7） inorganic fertilizer+straw （NP+S/NPK+S）. The main results and conclusions arefollowed.1. Suitable RothC model evaluation and selectionOriginal RothC26.3model can be used in upland soil in northern China. The root meansquare error （RMSE） ranged5.29¹4.57%（<15%） under in all the upland sites. The R2varied0.71⁰.93（except0.64in Pingliang and0.54in Changping）.The modified RothC26.3model for paddy soil （RothC26.3_p） was suitable in paddy-uplandrotation sites and the plot with lower than2.6t C ha^-1yr^-1carbon input in double-rice site but thehigher carbon input plots. Although the R2value was not high （0.62⁰.77）, the RMSE ranged3.62～15.00%in paddy-upland rotation site. The RMSE value was relative lower （4.92～8.94%）inplots with lower carbon input but higher （14.15⁵4.73%） in plots with higher carbon input indouble-rice site. The model need further modification in double-rice site.2.charateristics of regional SOC sequestration potential in Chinese arable landUnder the current fertilization management, the SOC sequestration potential was-4.9～57.0tC ha^-1in northern China. Under Control and N plots, the potential ranged-7.5～^-14.1t C ha^-1in all sites except Pingliang and Changping with the total average of-4.9t C ha^-1. Under NP plot,there was no SOC potential in Urumqi, Zhangye and Zhengzhou (-2.1～-5.8t C ha^-1), and0.4～32.7t C ha^-1in other five sites. The SOC potential in NP plot in northeast and northwest China（9.9～32.7） was higher than that in north China(0.4～7.0t C ha^-1), and the average one was8.1tC ha^-1. Under NPK plot, the SOC potential was12.4t C ha^-1, showed SOC maintained the initiallevel in all sites except Urumqi and Zhangye. There was higher SOC potential ranged30.8～57.0t C ha^-1under plots with organic materials application.In paddy-upland rotation system, there was no potential under Control plots (-7.6～^-16.1t Cha^-1). Under NPK plot, the SOC keep the initial level in Suining and Wuchang. The SOC potentialunder plots with organic materials application ranged3.6～35.6t C ha^-1. In double-rice soil, evenunder Control plot, the SOC maintain the initial level in Nanchang and Jinxian but Wangcheng.Under NPK plot, the SOC potential ranged10.0～15.9t C ha^-1.3. the SOC change under different manure carbon input scenariosThe manure carbon scenario prediction （2011²100） results showed, when the additional halfamount of manure carbon input of relevant NPK+M plot, the SOC potential will be turned fromnegetive into positive under Control plot in all sites except Zhangye and Zhengzhou, ranged2.8～6.8t Cha^-1. Under NPK plot, the additional half amount manure carbon input made Urumqi andZhangye site from negative SOC potential into positive. Under NPK+M plot,50%-reducedmanure carbon input, the SOC level can still keep the initial SOC level in all the site but Zhangye.All in all, to maintain the initial SOC level, additional half amount of manure carbon input of therelevant NPK+M should be needed under Control, NPK, and NPK+M. In paddy-upland rotationand double-rice soil system, additional half amount of manure carbon input of the relevantNPK+M will turn soil from negetive into positive in all sites except Chongqing. Under NPK+Mplot, even stop the manure carbon input, it still can keep the initial SOC level.Under NPK plot,1t C ha^-1yr^-1manure carbon input will increase43.9%（31.0%～56.7%）SOC sequestration potential in single-crop region and29.8%（28.0%～33.3%）in double-cropregion in northern China,31.3%（29.1%～33.1%）in paddy-upland rotation, and28.5%（26.9%～30.3%）in double-rice soil.In conclusion, we quantified the regional discrepancy of SOC sequestration potential inupland, paddy-upland rotation, and double-rice systems in China based on database in long-termfertilization sites and RothC model. These results provided the technology support to maintain andimprove soil fertility in different region. Forthurmore, we predicted the SOC change underdifferent fertilization management in typical region in China in future, and proposed thefertilization practices to maintain and elevate SOC.