Evaluation of the Impacts of Regional Climate Factors and Crop Management on Corn Yields in Different Climate Regimes of China Using the DayCent Model

Corn is one of most important agricultural products in China. Understanding impacts of regional climate change, as well as agricultural management practices, on corn yields is critical for maintaining stable corn production. Using the DayCent model and observed climatic data in Sichuan province (a humid and hot environment) and Hebei province (a cold and dry environment) in China, corn yields in 1948-2010 were simulated. The spatial variations of simulated corn yields and the relationship between regional climate variability and warming with corn yields in these two environments were analyzed. The results demonstrated that: (1) corn yields in Zhangjiakou of Hebei and most regions of Sichuan decreased significantly after 2000 compared to other regions; (2) relative humidity and precipitation exhibit a significant negative correlation with observed crop yields in the growing season in Hebei province; (3) air temperature from 23.33 °C to 29 °C constitutes the ideal range influencing the increase of corn yields in Sichuan; (4) the planting of the large amount of silage maize in Sichuan compensated the negative impact of the rising air temperature on corn yields; (5) sensitivity tests for different fertilization levels and OMAD suggest that an increasing fertilization level significantly affects corn yields in Hebei province, a cold and dry environment, while a decreasing fertilization level has a significant negative effect in Sichuan province, a hot and humid environment. The overarching goal of these analyses is to provide the theoretical basic for maintaining stable corn production under regional climate warming and different agricultural management practices.


Introduction
Corn is one of the most important crops globally, and its production has increased continuously from 2008 to 2015 in China, reaching 22 trillion kg in 2016. This increase of corn production has been attributed to enlargement of cultivated areas and advancements in production technology (B. Chen & G. Chen, 2007;Peng, Tang, & Zou, 2009;Yu, Huang, & Zhang, 2012;Bryan, King, & Zhao, 2014;Nendel, Kersebaum, Mirschel, & Wenkel, 2014). Effects of climate factors on corn yields have also been widely studied. Some extant literature has investigated the effect of air temperature (Wolfram & Michael, 2009;Yin et al., 2016;Lobell & Field, 2007;Basche et al., 2016;S. Chen, X. G. Chen, & J. T. Xu, 2016;Lee & Durmaz, 2016;Meng, Carew, Florkowski, & Klepacka, 2016), while others have focused on the effect of other climatic factors, such as precipitation and wind (Yin et al., 2016;Lee & Durmaz, 2016;Wang, Bocoling, & Cherkauer, 2016). A number of studies found that rising air temperature in the growing season, especially in July or in the seeding and maturity phases, increased corn yields (Wolfram & Michael, 2009;Yin et al., 2016). However, scholars have also reported that climate change has resulted in the reduction of corn yield over several countries in the past few decades (Wolfram & Michael, 2009;Lobell & Field, 2007;Basche et al., 2016;S. Chen, X. G. Chen, & J. T. Xu, 2016;Lee & Durmaz, 2016;Almarza, Mabood, Zhou, Gregorich, & Smith, 2008). In fact, Chen et al. (2016) reported that corn yield is projected to decline by 3-12% by 2100 in China due to rising air temperature. In the U.S., it was found that after the average growing-season air temperature exceeded 29 °C, corn yield would decrease sharply (Wolfram & Michael, 2009). Warming since 1981 has resulted in annual combined losses of crops representing roughly 40 Mt or USD $5 billion per year, as of 2002 (Lobell & Field, 2007). A 1% increase in the growing season air temperature reduces corn yield per acre by 9% (Lee & Durmaz, 2016). Increases in minimum and maximum air temperatures were attributed to reduced yields of 1.6-2.7% by decade (Basche et al., 2016). Although different climate characteristics among the study areas explain some temperature responses in corn yields, there is a general lack of comprehensive research on the effect of air temperature on crop yield. In this study, we thoroughly investigate this issue by studying the relationship between air temperature and corn growth in two different regions in China.
DayCent is a process-based biogeochemical model, and is a useful tool to predict yields as it integrates crop growth, carbon and nutrient dynamics, hydrology, management, and climate. Many studies have utilized DayCent to simulate changes in soil C, soil N, and greenhouse gas emissions (Wieder, Bonan, & Allison, 2014;Sheehan et al., 2013;Frey, Lee, Melillo, & Six, 2015;Cheng, Ogle, Parton, & Pan, 2014;Robertson, Grace, Izaurralde, Parton, & Zhang, 2014;Rafique, Fieneu, Parkin, & Anex, 2013;Reay et al., 2012;Mangalassery et al., 2014). Only a few previous studies have examined grain yield change under different climatic conditions and agricultural management practices using this model in China (Campbell et al., 2014;Lee, De, & Six, 2011). Cheng et al. (2014) used the DayCent model to successfully simulate crop yields in China, but focused on analyzing greenhouse gas mitigation potential and not the response of crop yields to future climatic change. In this study, variability in corn yields within each study region and the response to future climatic change were analyzed.
We selected Hebei province (a dry and cold environment) and Sichuan province (a warm and humid environment) in China as study areas, which possess quite different climate backgrounds (Figure 1). The spatial variation of simulated corn yields and the relationship between regional climate change, regional warming, and corn yield are reported in this study. The sensitivity of agricultural management practices in simulating corn yields by DayCent is also discussed. jas.ccsenet.  Note. *. Correlation is significant at the 0.05 level (2-tailed); **. Correlation is significant at the 0.01 level (2-tailed).

Spatial Variations of Corn Yield Simulation
The spatial distribution of simulated corn yields in 1948-2010 in Hebei revealed that corn yields in central and north Hebei were more than yields in the south (Figure 7a). Corn yields in Chengde, Baoding and Shijiazhuang were the greatest in the whole province, while corn yields in Tangshan, Hengshui, Langfang, and Xintai were the least. Corn yields in the entire province increased from 1948 to 2010. North of Zhangjiakou, yields increased from 1948 to 1999, but decreased substantially after 2000 compared to other regions. Additionally, corn yields in the north of Chengde decreased substantially compared to other regions from 1948 to 2010. Reduction of corn production that occurred after 2000 was observed in Zhangjiakou and Chengde, which could be due to continuous rainfall in the pollination period and drought in the grain formation-mature period (Liu, 2008;Zhang, 2014;Liu, 2013).
In Sichuan province (Figure 7b), high-yield corn is mainly located east of Sichuan, especially in the northeastern and southeastern regions of the province, which is associated with topographic features and agriculture distribution. Corn yields in Guangyuan, Bazhong, Deyang, Ziyang and Daxian were the greatest in the whole province, while those in Xiaojin, Yaan, and Mianning distributed in the central of Sichuan were the least. After 2000, corn yields in Bazhong, Daxian, Ziyang, Nanchong, Guangan, Yibin, Luzhou, Neijiang, and Zigong decreased substantially compared to other regions (Figure 8). Zhu and Yang (2007) found that corn yields have continued to decline since 2000 due to large-scale adjustment of planting areas of maize under the policy of returning farmland to forest in Sichuan. This simulated yield reflected the declination of corn yields, in agreement with Zhu and Yang (Zhu & Yang, 2007   hat the sensitiv s than in cold of air tempera growth and dev s reduced with u, 2008). The i rcation corn pl n between T m n corn yields, w ted corn yield 0 in Sichuan w able 2). The an re taken as con rs were sorted  Vol. 11, No. 15; sorted with the control factors, the control factors in the different provinces were divided into three levels from small to large, respectively (Table 2). Then, the significance of the difference between the dependent variables among these three levels of the control factors was analyzed using the LSD method. The results showed that, in Hebei, when the mean growing season air temperature increased from 18.69 °C to 19.41 °C, and then to 20.21 °C, there was no significant difference between simulated corn yields, which indicates that there was not a critical threshold value of air temperature influencing corn yields when the growing season air temperature varied in the range of 18.69 °C-20.21 °C. In Sichuan, when the air temperature rose from 22.29 °C to 22.68 °C, simulated corn yields varied non-significantly, but at 23.33 °C, simulated corn yields changed significantly, which shows that 23.33 °C may constitute a critical threshold value influencing simulated corn yields. The change in the growing season air temperature was positively associated with simulated corn yields with the correlation coefficient of 0.456 (sig. = 0.009 < 0.05) (Table 3). Therefore, when the mean growing season air temperature was over 23.33 °C, simulated corn yields increased with the rising of air temperature in Sichuan province. Wolfram and Michael (2009) also found that annual mean air temperature affects corn yields. They indicated that crop yield would increase with elevated air temperature up to 29 °C. Moreover, above this threshold value of air temperature, further warming would have a detrimental effect on yield. The results of this study are not contradictory to those of Wolfram and Michael (2009), and provide a minimum value of air temperature influencing and increasing corn yields. The annual mean air temperature in the growing season of corn (April to September) in Sichuan and Hebei are 22.83 °C and 19.34 °C, respectively. Therefore, when the mean growing season air temperature reaches 23.33 °C-29 °C in Sichuan, corn yields increase with the rising air temperature. It has been found that the 1990s are the warmest 10 years in 1948-2010 because the number of days over 29 °C of actual air temperature in the growing season was more than 190 days in each year of seven years of 1990s. This long period and continuous high temperature are also probably the factors leading to the reduction of corn yields in the 1990s. Note. *. LSD is significant at the 0.05 level (2-tailed); **. LSD is significant at the 0.01 level (2-tailed).

Sensitivity Analysis
Anandhi (2016) also analyzed the impact of management decisions, such as plant water use, fertilizer application and hybrids on corn yields, and found inadequate N fertility results in lower yields, whereas over-fertilization results in higher emissions of nitrous oxide. Here, we examine the changes in yields that result from increasing or decreasing inorganic and organic N applications.   (Figure 10b). than in a hum on levels were ensitivity of th obably associa fertile than th f the higher N reased slightly to 75%, whic ates that crop y 0% produces di

Sensitivity to OMAD
After corn is harvested each year, soil fertility needs to be replenished to benefit corn growth in the following year. The amount of organic N that was applied in the control runs was 20 kg/ha/yr. The C:N ratio of the organic matter applied was 15. We conducted sensitivity studies by adding or decreasing 50% and 75% of the original OMAD level. There was no difference in the mean control and experimental simulations yields in 1979-2010 in Sichuan ( Figure 11b); whereas, there was a slight increase in 1978-2010 mean yields (59.05 kg/ha/yr) in Hebei (Figure 11a). This is also probably related to the original soil fertilization condition. The higher inorganic N soil fertilization amounts in Sichuan meet the needs of corn growth. There was no need for additional organic N fertilizer in the field. In addition, the reason that OMAD increases are not important to either province is probably different. In Hebei, when corn is harvested in autumn, OMAD cannot efficiently improve soil fertility for drier soil; in Sichuan, OMAD can improve soil fertility for humid soil due to ample precipitation and irrigation.
Generally, fertilization level and OMAD have a larger impact in Hebei (cold and dry environment), where there is less active biophysical activity due to lower soil moisture and warmth. Increased soil fertilization can improve soil biophysical activities and meet the needs of corn growth (Tan et al., 2002).

Conclusions
The DayCent model is used to simulate corn yields under different environmental conditions in China. The model reproduced the inter-annual variation of crop yield with a R 2 of 0.71 and 0.85 in Hebei and Sichuan, China, respectively, demonstrating that the DayCent model is capable of accurately simulating corn yield in these two environments.
The annual mean T max and T min increased at a rate of 0.22 °C/10 yr and 0.23 °C/10 yr in Sichuan, respectively, and 0.34 °C/10 yr and 0.55 °C/10 yr in Hebei, respectively. The annual mean downward shortwave radiation (Dswrf) increased at a rate of 1.92 (langleys/day)/10 yr in Sichuan, and decreased at a rate of -0.82 (langleys/day)/10 yr in Hebei. The annual mean RH decreased at a rate of -0.06 %/10 yr in Sichuan, and increased at a rate of -0.45 %/10 yr in Hebei. The annual mean wind decreased at a rate of -0.11 (miles/hour)/10 yr in Sichuan, and increased at a rate of 0.08 (miles/hour)/10 yr in Hebei. The annual mean precipitation decreased at a rate of -0.01 cm/10 yr in both Sichuan and Hebei.
In this paper, the impacts of climatic factors and management amendment levels on corn yields were explored. The results showed that RH and Prcp negatively affected corn yields in Hebei (a cold and dry environment). That impacts of the rising of T max and T min on corn yields were not significant is because corn growth in Hebei mainly depends on irrigation, and is not constrained by the drought resulting from the rising air temperature. Only T min negatively affected corn yields in Sichuan.
Simulated corn yields decreased significantly after de-trending the warm air temperature in Sichuan. There was no critical threshold value of air temperature influencing corn yield simulation in Hebei; whereas, 23.33 °C was the critical threshold value influencing simulated corn yields in Sichuan.
The sensitivity study suggests that crop yield is more sensitive to fertilization level in a cold and dry environment as compared to that in a humid and hot environment. Fertilizer application can remedy the shortage of soil fertility, and accelerate crop growth in a cold and dry environment.
In Sichuan province (a humid and hot environment), current air temperature is limiting to corn production. 23.33 °C-9 °C constitutes the ideal scale of air temperature influencing increase of corn yields.