Analysis Of The Production Potential Of Winter Wheat, Maize And Lucerne Based On The APSIM Model

The Loess Plateau is the most important dry farming area in China. The crop yield in this area is both low and unstable. It is now more necessary than ever to evaluate crop yield under global climate change, so as to take effective measures that can decrease yield loss. The soil-crop model APSIM, was originally developed by Australian scientists to simulate the biological processes in agricultural systems. It is also an important physiological model for simulating the variation of each key component of arid farming systems. This model works effectively in analysing the crop production potential and risk under climate change profiles. APSIM has been used to analyse crop production risk in the North China Plain, however similar effort in the Loess Plateau has not been undertaken. In this study, based on the localization of APSIM to the Loess Plateau, simulation of the productivity of winter wheat and lucerne was conducted and, by applying5climate change scenarios, the production potential and risk in the future were analysed to obtain a better understanding of climate change impacts on crop yields in the Loess Plateau.There were obvious correlations between the simulated and measured values of lucerne and winter wheat biomasses, and winter wheat grain yield during2003-2005and2009-2010periods. The correlation coefficients were0.79,0.98and0.98, respectively. This suggests that the localized model can represent the effects of climate factors on crop yields of lucerne and winter wheat in the Loess Plateau, in addition to the production characteristics of both crops. The correlation coefficients were0.72-0.83between the simulated and measured values of soil water, and the simulated value showed the change in soil water content of surface soils and the relative stability of deep soil water. Therefore, the model can also sensitively reflect the effects of rainfall and evaporation on soil water. Additionally, the simulation can also revealed that lucerne can use deep soil water with its strong and deep root system Accordingly, under typical conditions with rainfall close to average annual precipitation, APSIM is an excellent model for simulating the crop yields in the Loess Plateau. Under the changes in temperature and rainfall during the1961-2010period, yield variation of winter wheat, maize and lucerne ranged from-14.4%-9.3%,-21.4%-11%and-23.9%-17%respectively. This suggests that the changes in temperature and rainfall are the most important factors leading to the changes in crop yield in the Loess Plateau. When only rainfall changed, the yields of winter wheat, maize and lucerne respectively varied by-13.9%-9.3%,-17.7%-11%and-16.9%-17%, while the crop yields decreased by1.4%,7.5%and8.3%responding to the rise in the temperature. The productivity of winter wheat was the most stable whereas maize was the least stable among these three crops. The trendline slopes of the yield probability of the3crops appeared in the order below:daily rainfall+10%scenario> daily rainfall+10%and daily temperature+1degree scenario> historical scenario> daily temperature+1degree scenario> daily rainfall-10%scenario> daily temperature+1degree and daily rainfall-10%scenario. The implication is that high rainfall is a key factor to maintain greater and more stable crop yield, whereas temperature increase ahs the opposite effect. Therefore, the trend of crop production should be estimated based on the climate change trends. The correlation coefficients between the rainfall during the growth season and the yields of the3crops were0.59,0.24and0.28for winter wheat, maize and lucerne respectively, and0.30,0.08and0.34between average temperature and yield. It suggested that, to a certain extent, winter wheat yield has a close relationship with rainfall, and average temperature can only describe part of yield change of3crops.Conclusively, in this research, based on the localization of APSIM to the Loess Plateau, the production potential of winter wheat, maize and lucerne was analysed in the Loess Plateau. It was demonstrated that the model can sensitively describe the effects of rainfall and temperature changes on the production of winter wheat, maize and lucerne in this rain-fed agriculture area. Moreover, further verification to APSIM still should be undertaken on the basis of field experimental data, providing more data for wider application of the model and promotion of its applicability in the Loess Plateau.