Abstract

Impacts of climate change on agriculture are a major concern worldwide, but uncertainties of climate models and emission scenarios may hamper efforts to adapt to climate change. In this paper, a probabilistic approach is used to estimate the uncertainties and simulate impacts of global warming on wheat production and water use in the main wheat cultivation regions of China, with a global mean temperature (GMT) increase scale relative to 1961–90 values. From output of 20 climate scenarios of the Intergovernmental Panel on Climate Change Data Distribution Centre, median values of projected changes in monthly mean climate variables for representative stations are adapted. These are used to drive the Crop Environment Resource Synthesis (CERES)-Wheat model to simulate wheat production and water use under baseline and global warming scenarios, with and without consideration of carbon dioxide (CO2) fertilization effects. Results show that, because of temperature increase, projected wheat-growing periods for GMT changes of 1°, 2°, and 3°C would shorten, with averaged median values of 3.94%, 6.90%, and 9.67%, respectively. There is a high probability of decreasing (increasing) changes in yield and water-use efficiency under higher temperature scenarios without (with) consideration of CO2 fertilization effects. Elevated CO2 concentration generally compensates for the negative effects of warming temperatures on production. Moreover, positive effects of elevated CO2 concentration on grain yield increase with warming temperatures. The findings could be critical for climate-change-driven agricultural production that ensures global food security.

1. Introduction

Average global temperatures have increased in recent decades and have been predicted to continue rising in the foreseeable future (Solomon et al. 2007). With compelling scientific evidence of a warming planet [because of a human-induced carbon dioxide (CO2) increase], it is ever more important to understand impacts of global climate change on agriculture. Climate provides the background for agricultural production and is critical to its stability. Climate change may have an impact on the agricultural production environment, structure, and productivity. There are two primary effects of climate change on agricultural production—the first is from elevated CO2 concentration, and the second is from the impact caused by climate change (mainly global warming; Solomon et al. 2007). Therefore, quantitative study of the effects of elevated CO2 concentration and increased temperature on crop physiological processes and yield has become key to addressing climate change. Several studies have related changes in agricultural phenomena such as crop yield, growth period, and water use to temperature increases (Gbetibouo and Hassan 2005; Aggarwal et al. 2006a,b; Dhungana et al. 2006; Tao et al. 2003, 2008b; Challinor and Wheeler 2008; Mo et al. 2009; Challinor et al. 2010; Laux et al. 2010; Liu and Yuan 2010; Rotter et al. 2011; Tao and Zhang 2011).

Because of the complex relationship between agriculture and climate change, however, there are great discrepancies and uncertainties in estimated responses of agricultural production to climate change. Prediction uncertainties have frequently been associated with general circulation model (GCM) output, emission scenarios, scale transformations, and crop-model-simulated outputs, which extend to simulated effects of climate change on agriculture production, adaptation, and evaluation mechanisms (Challinor et al. 2007; Lobell and Burke 2008; Tao et al. 2008a; Ruiz-Ramos and Mínguez 2010; Roudier et al. 2011). These uncertainties make it extremely difficult for planners and decision makers to develop or adopt agricultural production measures that are adaptable to climate change (Lobell et al. 2011; Tao and Zhang 2011).

Identifying which crops or regions are most affected by which factors can assist ongoing efforts to adapt agricultural production to climate change (Lobell et al. 2011). Tao et al. (2008a) and Tao and Zhang (2011) estimated uncertainties and simulated the impact of global warming on maize/rice production and water use in China against global mean temperature (GMT) rise from 1961 to 1990. The combined use of representative climate-change scenarios and process-based crop models can reduce uncertainties caused by emission scenarios and model simulation and can lead to better predictions (Lobell and Burke 2008; Tao et al. 2008b, 2009).

Wheat (one of China’s staple foods) accounts for 21.69% of the national food supply. Winter and spring wheat are the dominant cultivated crops in northern and eastern regions, accounting for over 22.28% of total land area in the country. Total national wheat production has consistently increased, from 5384 × 104 t in 1978 to 11 511.5 × 104 t in 2009 (National Bureau of Statistics of China 2010). Given its long growth period (220–250 days) and high irrigation demand (200–300 mm), however, winter wheat production is sensitive to climate change and related factors, such as water shortages and increasing temperature (Fang et al. 2010; Gong et al. 2011). Quantitative analysis of wheat production and water use as affected by global warming and CO2 concentration is therefore critical for food security.

Therefore, this study projects quantitative changes in wheat production and water-use efficiency (WUE) in China under 1°, 2°, and 3°C changes in GMT and 20 climate scenarios, with and without consideration of the CO2 fertilization effect. On the basis of simulations with the Crop Environment Resource Synthesis (CERES)-Wheat crop model under different GCM emission scenarios, we address climate factors with the strongest influences on wheat production.

2. Methods and data

a. Study stations

To support adequate model validation and evaluation, we only selected stations that 1) were located in primary wheat-production regions, 2) represented typical wheat-cropping systems in China, 3) were geographically and climatologically different, and 4) had long-term crop (e.g., management practices, phenology, and crop yield) and weather-observation records. These criteria resulted in four agricultural meteorological stations: Jinan (34°43′N, 113°39′E), Zhengzhou (36°41′N, 116°59′E), Chengdu (30°40′N, 104°01′E), and Harbin (45°45′N, 126°46′E). Jinan and Zhengzhou are on the North China Plain. It has a warm, temperate semihumid monsoon climate, and rainfall from June to September accounts for about 80% of the annual total. According to the local experimental records, winter wheat is sown in the end of October and is harvested in early June, and it needs irrigation 3–5 times (200–400 mm) to ensure its growth. The Chengdu site in the southwestern region is in a subtropical monsoon climate zone with temperate climate, clear seasons, and a long-term, frost-free period. It has sufficient rainfall (~862 mm) but less sunlight. It has more-extreme weather changes, cold in winter and hot in summer. The planting system is winter wheat–summer maize multiple cropping, with two crop yields per year as at Jinan and Zhengzhou. The Harbin site is in the north of northeastern China and the south of Heilongjiang Province. It has a temperate continental monsoon climate, with long winter, short summer, and an average annual precipitation of 555 mm. Rainfall is mainly from June to September, representing 60% of the annual amount. Principal crops include spring wheat and rice. Spring wheat is generally sowed from the end of April to the beginning of May and is harvested during the middle 10 days of August. Irrigation about 2–4 times (150–300 mm) is necessary to ensure growth. The fertilization frequency is generally twice for winter wheat, with 80–140 kg N ha−1 each time. The first application occurs before planting, and the other is applied at the jointing or booting stage (early April). Spring wheat uses the one-off autumn fertilization method (after autumn harvest at the end of August or early September) to resolve the difficulties of the spring-freeze fertilization, at 100–200 kg N ha−1. The selected datasets of 2004–08, 1998–2000, 1992–94, and 1996–99 at Jinan, Zhengzhou, Chengdu, and Harbin, respectively, are used for model checking. The varieties that correspond to these periods are Jimai22, Zhengzhou761, Tianmai, and Hamai, respectively. The variety was a representative cultivar for the broad planting in each area at that time. Other basic cultivation details at the selected stations, such as climate, cropping system, and soil type, are given in Fig. 1 and Table 1.

b. Calibration and validation of CERES-Wheat model

1) CERES-Wheat model

The latest process-based CERES-Wheat crop model of Decision Support System for Agrotechnology Transfer (DSSAT-V4.0.2; Jones et al. 2003) was adapted to simulate wheat productivity in China. The model, which simulates crop yield, focused on three important areas—growth duration, growth rate, and the extent to which “stresses” influence these two processes. Stress may be caused by soil water and nutrients or by extremes in temperature. Phasic development was calculated separately through accumulated temperature and day length, and mass growth was calculated primarily through radiation interception by plant leaves. The soil water balance routine of CERES-Wheat includes soil water quantity from inputs of precipitation and irrigation. Evaporation from plants and soil, runoff, and drainage are outputs.

The CERES model has a proven ability to simulate crop phenology, leaf area index, yield, accumulated dry matter, nitrogen, and soil moisture balance within various cultivation, management, and environmental conditions (Bannayan et al. 2003; Panda et al. 2004; Rinaldi 2004; Saseendran et al. 2004; Popova and Kercheva 2005; Yang et al. 2006; Persson et al. 2010). The model also accounts for effects of CO2 concentration on photosynthetic and evapotranspiration (ET) rates, and is therefore applicable in analyzing impacts of climate change on crop productivity (Smith 1996; Eitzinger et al. 2003; Luo et al. 2003; Anwar et al. 2007; Tao et al. 2006; Tao and Zhang 2011).

For “C3” crops, the effect of CO2 on net assimilation was simulated by multiplying the net rate (Peart et al. 1988). The multiplier values changed linearly from 1.0 at 330 ppm CO2 to 1.25 at 660 ppm CO2 and then to 1.43 at 990 ppm CO2. The stresses caused by soil water, nutrients, or temperature extremes on net photosynthesis in increased-CO2 environments are mediated through their effects on leaf area growth and hence on radiation absorption.

Soil water balance, ET, and photosynthesis were calculated by the Ritchie water balance equation, Priestley–Taylor method [modified by Ritchie (1985)], and canopy curve, respectively. Inputs (daily or seasonal) included data on weather, soil properties, and crop genetic parameters and management practices (e.g., depth, density, and geographic location of planting; dates of sowing, fertilization, and irrigation; and amounts and modes of fertilization and irrigation). Daily weather data such as maximum temperature Tmax, minimum temperature Tmin, solar radiation Srd, and precipitation Pre were chosen as weather inputs. Soil inputs mainly included type, textures, moisture, and nutrients. Field management practices mainly consisted of cultivars, planting (planting density and planting date), fertilizing (fertilization time and fertilizer), irrigation (irrigation time, amount of water, operation, and irrigation efficiency), organic amendments, tillage, harvest, and chemical application. Seven genetic parameters in CERES-Wheat require parameterization: vernalization sensitivity coefficient P1V (percent per day of unfulfilled vernalization), photoperiod sensitivity coefficient P1D (percent reduction per hour near threshold), thermal time from onset of linear fill to maturity P5 (°C day−1), kernel number per unit stem and spike weight at anthesis G1 (plant per gram), potential kernel growth rate G2 (milligrams per kernel-day), tiller death coefficient G3 (standard stem and spike weight when elongation ceases; g), and thermal time between successive leaf tip appearances PHINT (°C day−1).

Model parameterization was the adjustment of parameters so that simulated values compared well to measured ones. Before parameter calibration, we initially needed to do a sensitivity analysis of genetic coefficients to identify the range of parameter changes and to select the most sensitive parameter to calibrate. The cultivation habits and previous references for model application were also helpful in determining the parameters. After repeated testing, we obtained the cultivars for simulation and selected values of the seven genetic parameters (Table 2). The selected wheat cultivar was consistent with local experimental records.

2) Simulation scenario establishment and model checking

To evaluate wheat productivity and water use under the baseline and future climate scenarios, a set of experimental scenarios [one accounting for irrigation (I-treat) and the other accounting for rain-fed (R-treat) conditions] was constructed. The scenarios for calibration and validation were the same as for the experimental-station records. Both I-treat and R-treat had the same basic fertilizer treatments, planting dates, and management practices. Automatic irrigation applications were used once soil moisture conditions were satisfied and the stress factor had reached the 60% threshold (at 40-cm depth). For the “I-treatment,” automatic irrigation was applied as refilling of the soil water profile whenever soil water content fell below 60% of capacity at 40-cm depth. Soil profile data at each station were extracted from the Soil Species of China, available from the National Soil Survey Office. Other management measures such as planting date and fertilization were consistent with local experimental records as introduced in section 2a.

The calibration and validation data mainly included phenology and yield or yield components. Calibrated and validated model performance was measured by RE (relative error) and root-mean-square error (RMSE) analysis as

 
formula
 
formula

where was the model-simulated value, i was the ith sample, was the observed value, and n was total sample amount. Wheat WUE was evaluated for both I-treat and R-treat conditions as

 
formula

c. Projecting wheat-productivity and water-use change

The general framework of the study is presented in Fig. 2. For each station, we first used historical daily weather data from 1961 to 1990 to parameterize the “Long Ashton Research Station weather generator” (LARS-WG; Semenov et al. 1998; Semenov 2007) and to generate 100 yr of daily weather data on the same climate variables, which represented baseline climate conditions. At the same time, projected climate changes and atmospheric CO2 concentrations at the station for GMT changes of 1°, 2°, and 3°C were identified on the basis of GCMs and Special Reports on Emissions Scenarios (SRESs) and their projected time series of annual GMT change (Houghton et al. 2001).

We then found median values of the projected changes in monthly mean climate variables (including Tmax, Tmin, Srd, and Pre). We used them to alter the parameters in the weather generator and to generate 100 yr (2001–2100) of daily weather data for the three GMT changes. On the basis of time series of global warming projected by the 20 climate scenarios and projected CO2-concentration trajectories across scenarios (Houghton et al. 2001), CO2 concentration ranges were 396–490, 473–635, and 552–856 ppm for 1°, 2,° and 3°C, respectively. Monthly fields of mean temperature Tmean, changes of diurnal temperature range DTR, and precipitation on a 0.5° grid from 2001 to 2100 in the 20 scenarios were from the Climatic Research Unit, University of East Anglia (Mitchell et al. 2004). The scenarios comprise all 20 combinations of four SRESs (“A1FI,” “A2,” “B1,” “B2”) and five GCMs [the third climate configuration of the Met Office Unified Model (HadCM3); Parallel Climate Model (PCM); Canadian Centre for Climate Modelling and Analysis (CCCma) Coupled General Circulation Model, version 2 (CGCM2); Commonwealth Scientific and Industrial Research Organisation, version 2 (CSIRO2); and the German “ECHAM4”], using GCM outputs from the Intergovernmental Panel on Climate Change Data Distribution Centre. These scenarios cover 93% of the uncertainty range of global warming in the twenty-first century from Houghton et al. (2001).

Twenty different futures were therefore used to represent uncertainty in climate impacts arising from two distinct uncertainty sources—uncertainty in future emissions of greenhouse gases and uncertainty in climate modeling. These different scenarios thus permit users to assess implications for climate impacts of some major sources of uncertainty in future climate. Each of the 20 permutations should be treated as equally likely (Mitchell et al. 2004).

The new set of LARS-WG parameters, which was specific to the station and scenario for the three GMT changes, was calculated by following the method of Semenov (2007). To be specific, in the LARS-WG scenario file we calculated and renewed absolute changes in daily Tmax and Tmin and relative changes in monthly mean precipitation and Srd between the climate-change scenario and baseline climate.

Last, we ran the validated CERES-Wheat model using both baseline climate conditions and climate-change scenarios for the GMT changes, with and without CO2 fertilization effects, for 100 yr. Changes of growing period, wheat yield, ET, and WUE between the future climate-change scenarios and baseline climate conditions were addressed. We further investigated probability distributions of resultant changes for the three GMT changes. Box plots and cumulative distribution functions were used to describe their statistics and probability distributions. A detailed description of the GCMs and the comprehensive dataset construction can be found online (http://www.ipcc-data.org/) and in Mitchell et al. (2004), respectively.

Because the CERES-Wheat model requires input data of Tmax, Tmin, Pre, and Srd, the climate variables Tmax and Tmin were therefore derived through changes in Tmean and DTR by using the following equations (Tao et al. 2008a):

 
formula
 
formula

Solar radiation was derived by a self-calibrating method through monthly average daily Tmax and Tmin (Hargreaves et al. 1985; Allen 1997):

 
formula

where H is monthly average daily solar radiation, Ho is monthly average daily extraterrestrial radiation, and a and b are empirical parameters.

Prior to applying equation Eq. (6), simulated effects of Eqs. (4) and (5) were tested using observed H, Tmax, and Tmin for 1961–93 from the China Meteorological Administration. The 1961–80 dataset was used for parameterization, and the 1981–93 dataset was used for validation of the equation. The simulation of H in Eq. (6) was generally good, with RMSE across the stations between 1.18 and 2.14 MJ m−2 day−1 (Tao and Zhang 2011).

3. Results

a. CERES-Wheat model uncertainty

One-year experimental data were used to parameterize and calibrate the model for each station. The station-by-station calibrated model was then validated by field data from other years (Table 3). There was good agreement between observed and simulated dates of anthesis and maturity for both winter wheat (Jinan, Zhengzhou, and Chengdu) and spring wheat (Harbin) (Fig. 3). RMSEs of simulated anthesis and maturity dates were respectively 7.59 and 8.65 days (winter wheat) and respectively 3.37 and 4.24 days (spring wheat). Average relative error for anthesis date was less than 4.5%, and that for maturity date was less than 4.0%. There was also good agreement between simulated and observed yields, with RMSE of 328.55 kg ha−1 for winter wheat and 175.26 kg ha−1 for spring wheat. Average corresponding error was less than 10.0%. This favorable agreement is a strong indicator of reliability for the CERES-Wheat model. The model results were then used to assess potential impacts of climate change on wheat production.

b. Climate-change uncertainty

Figure 4 depicts median values of projected absolute changes in monthly mean Tmax, Tmin, Srd, and Pre at all stations during wheat growth for 1°C (GMT+1D), 2°C (GMT+2D), and 3°C (GMT+3D) GMT changes under the 20 climate-change scenarios relative to 1961–90 baseline conditions. For all stations, average monthly Tmax and Tmin under the three GMT changes were higher than those under baseline conditions. Further, mean change of annual mean temperature was greater than the corresponding GMT change. Temperature increase at low-latitude stations such as Chengdu was less than at high-latitude stations such as Harbin. This implies greater climate warming at high-latitude regions than at low-latitude ones. Temperature increase during November–January and February–April was greater than in May–June periods. Also, Tmin increase was greater than that of Tmax. These trends are generally consistent with the observed and projected conditions of climate change (Houghton et al. 2001).

Besides regional and seasonal variations, there were considerable uncertainties in temperature for projected regional climate change, as based on the 20 climate scenarios. For instance, with GMT increase of 2°C at Chengdu station, projected absolute changes relative to 1961–90 of monthly mean Tmax ranged from 0.01° to 1.34°C for January [day after planting (DAP) 61–90] and from 2.91° to 5.51°C for April (DAP 152–181), over the 20 climate scenarios. Changes in monthly mean Tmin were from 0.78° to 2.34°C for January and from 2.41° to 4.17°C for April. Changes in monthly mean precipitation were from −0.9% to 0.55% for January and −0.17% to 1.34% for April. Changes in monthly mean Srd were from −0.7% to 0.46 for January and −0.15 to 0.005 for April. For DAP 45–105, the above four climate-change factors for GMT+2D were significantly lower than for GMT+1D and GMT+3D.

c. Changes in wheat yield, growing period, ET, and WUE with GMT increase

1) Without CO2 fertilization effects

Wheat growth and productivity were simulated under irrigated and rain-fed conditions for 100 yr, in future scenarios and baseline conditions. Without consideration of CO2 fertilization effects, wheat growing period, yield, ET, and WUE generally decreased under irrigated conditions. Under irrigated conditions at the stations, median projected declines in growth period from baseline to the 1°, 2°, and 3°C GMT changes were from 3.62% (Jinan) to 4.37% (Chengdu), from 6.49% (Harbin) to 7.36% (Chengdu), and from 8.28% (Harbin) to 10.66% (Zhengzhou), respectively (Fig. 5). All subsequent ranges cited refer respectively to the 1°, 2°, and 3°C GMT changes. Wheat yield also declined under irrigated conditions, and yield loss increased with GMT temperature change (except for Chengdu, where yield was lowest for the GMT+2D condition because of low Srd). Median-value ranges of projected yield loss were from 1.15% (Zhengzhou) to 9.80% (Chengdu), from 3.51% (Zhengzhou) to 51.93% (Chengdu), and from 7.28% (Zhengzhou) to 36.90% (Chengdu). Changes of yield under rain-fed conditions followed those under irrigated conditions at Chengdu, with projected mean losses of 9.80%, 51.93%, and 36.90% for the GMT changes, respectively. Projected median-value ranges of yield for the other three stations under rain-fed conditions were from −16.48% (Zhengzhou) to −0.37% (Jinan), from −15.39% (Harbin) to 7.20% (Jinan), and from −14.93% (Harbin) to −1.23% (Jinan). Under irrigated conditions, median values of projected ET increased for Harbin (2.06%, 5.40%, and 9.44%) and Jinan (0.62%, 3.04%, and 3.32%) and decreased for Chengdu (0.17%, −9.06%, and −3.64%) and Zhengzhou (−2.05%, −2.09%, and −2.04%).

Corresponding declines of wheat growth period under rain-fed conditions were similar to those for irrigated conditions (Fig. 5). Changes of yield under rain-fed conditions followed those under irrigated conditions for Chengdu, with projected mean losses of 9.80%, 51.93%, and 36.90% for the GMT increases. Projected median yields for the other three stations under rain-fed conditions were from −16.48% (Zhengzhou) to −0.37% (Jinan), from −15.39% (Harbin) to 7.20% (Jinan), and from −14.93% (Harbin) to −1.23% (Jinan). Under rain-fed conditions, there was a general increase in the median projected ET, except for Chengdu. Median projected ETs were from −6.42% (Zhengzhou) to 1.64% (Harbin), from −9.00% (Chengdu) to 10.94% (Jinan), and from −3.65 (Chengdu) to 9.85% (Harbin).

The median projected changes of WUE generally decreased, under both irrigated and rain-fed conditions. This may be attributed mainly to rising ET and yield loss. For instance, median values of projected WUE under irrigated conditions were from −10.11% (Chengdu) to 1.11% (Zhengzhou), from −47.23% (Chengdu) to −1.21% (Zhengzhou), and from −34.53% (Chengdu) to −5.14% (Zhengzhou). Changes of WUE under rain-fed conditions were similar to those for irrigated conditions. Corresponding median ranges were from −13.25% (Zhengzhou) to 2.59% (Jinan), from −47.18% (Chengdu) to 1.39% (Zhengzhou), and from −34.52 (Chengdu) to −6.45% (Zhengzhou).

2) With CO2 fertilization effects

The simulated growth period for CO2 fertilization effects was similar to that of non-CO2 fertilization effects. There was also no noticeable difference between irrigated and rain-fed wheat conditions (Fig. 6). This is sufficient evidence that temperature was the major determinant of wheat growth period under the climate-change scenarios. In contrast to non-CO2 fertilization effects, wheat yields, ET, and WUE under CO2 fertilization effects generally increased with both irrigated and rain-fed conditions, except under rain-fed conditions for GMT+2D at Chengdu and GMT+1D at Zhengzhou.

Median-value ranges of wheat yield under irrigated conditions were from 6.00% (Chengdu) to 9.36% (Zhengzhou), from −24.15% (Chengdu) to 18.08% (Zhengzhou), and from 9.06% (Chengdu) to 26.38% (Zhengzhou). Mean projected changes of wheat yield under rain-fed conditions at Chengdu were 6.00%, −24.15%, and 9.06% for the GMT changes. There was a similar trend for non-CO2 fertilization effects, suggesting that CO2 was not a major factor for yield change. For the three other stations under rain-fed conditions, median yields were from −2.41% (Zhengzhou) to 15.05% (Jinan), from 10.21% (Harbin) to 39.42% (Jinan), and from 24.43% (Harbin) to 48.71% (Jinan).

For irrigated wheat, median projected ETs generally decreased (except for Harbin, with median values of 0.96%, 3.29%, and 6.17%), and corresponding ranges for the other stations were from −4.11% (Zhengzhou) to −0.84% (Jinan), from −8.83% (Chengdu) to 0.20% (Jinan), and from −6.64% (Zhengzhou) to −0.67%. Under rain-fed conditions, median projected ETs increased at Harbin and Jinan and decreased at Chengdu and Zhengzhou. Projected median-value increases were respectively 1.16%, 2.84%, and 8.23% at Harbin and 1.32%, 10.09%, and 8.13% at Jinan. Projected median-value decreases were −0.81%, −8.56%, and −4.78% at Chengdu and −6.71%, 1.12%, and −0.87% at Zhengzhou.

With CO2 fertilization effects, median projected changes of WUE generally increased under both irrigated and rain-fed conditions. WUE median-value ranges under irrigated conditions were from 6.23% (Harbin) to 14.25% (Zhengzhou), from −16.85% (Chengdu) to 25.10% (Zhengzhou), and from 14.03 (Harbin) to 35.74% (Zhengzhou). For rain-fed conditions, projected WUE median values were from 6.68% (Chengdu) to 15.47% (Jinan), from −17.04% (Chengdu) to 42.44% (Zhengzhou), and from 14.77% (Chengdu) to 53.83% (Zhengzhou).

d. Probabilistic changes in wheat productivity and water use

Cumulative probability distributions were used to explain probabilistic changes in wheat productivity and water use. For example, probabilities of shorter wheat growth periods under the 1°, 2°, and 3°C GMT changes, with or without irrigation and CO2 fertilization effects, were 98%, 99%, and 99%, respectively, at Jinan (Fig. 7). The wheat yield under irrigated conditions without (with) consideration of CO2 fertilization effects would decrease (increase) with a probability of 55% (88%), 68% (76%), and 74% (97%) for the 1°, 2°, and 3°C GMT changes, respectively. Also, the ET without (with) consideration of CO2 fertilization effects would increase with a probability of 53% (48%), 61% (55%), and 63% (54%) for 1°, 2°, and 3°C GMT changes, respectively. Those WUE without (with) consideration of CO2 fertilization effects would decrease (increase) with a probability of 58% (78%), 80% (86%), and 85% (98%) for 1°, 2°, and 3°C GMT changes, respectively.

For rain-fed wheat, the yield without (with) consideration of CO2 fertilization effects would decrease (increase) with a probability of 53% (57%), 47% (84%), and 56% (92%) for 1°, 2°, and 3°C GMT changes, respectively (Fig. 8). Those of rise without (with) consideration of CO2 effects are 45% (45%), 70% (70%), and 65% (65%) for 1°, 2°, and 3°C GMT changes, respectively. Also WUE without (with) consideration of CO2 fertilization effects would decrease (increase) with a probability of 55% (69%), 58% (87%) and 71% (93%) for 1°, 2°, and 3°C GMT changes, respectively. Hence with the exception of ET, there is generally a high probability of declining wheat growth period, yield, or WUE under irrigated (rain fed) conditions without CO2 fertilization effects.

4. Discussion

a. Changes of yield and water use with GMT increase

The simulation suggests significant interactions between wheat yield and different climate-change factors (e.g., elevated CO2 concentration, warm temperature, low Srd, and others). There have been several studies on the effects of warming temperature and elevated CO2 concentration on cropping systems (Mearns et al. 1997; Reyenga et al. 1999; Van Ittersum et al. 2003; Trnka et al. 2004; Xiong et al. 2005; Liu and Lin 2008; Gobin 2010). However, there have been few investigations accounting for full interactions between warming temperature, elevated CO2 concentration, and rainfall (Fulco and Asseng 2006).

In this study, changes of wheat yield at Chengdu station under rain-fed conditions both with and without CO2 fertilization effects were consistent with those under irrigated conditions. Figure 4 implies that Srd at Chengdu was generally lower than at other sites under the same climate scenarios. This is consistent with its local climate characteristics, as depicted in section 2a (sufficient rainfall but less sunlight). Yield change was insensitive to variations of temperature and CO2 concentration under low-Srd conditions. Mean projected changes of wheat yield for the 1°, 2°, and 3°C GMT changes were −9.80% (6.00%), −51.93% (−24.15%), and −36.90% (9.06%), respectively (parentheses indicate with-fertilization conditions). The highest yield loss with each increase in GMT temperature was at Chengdu under GMT+2D conditions. The most probable reason for this loss could be that the drivers of climate change (especially Srd) under GMT+2D conditions were weaker than in the other conditions (GMT+1D and GMT+3D) for the period 31–105 DAP (Fig. 4). In CERES-Wheat, mass growth was calculated primarily from plant-leaf intercepted radiation. Hence, a declining Srd implies less radiation interception for dry-matter accumulation and yield production. This result suggests that in projecting the impact of climate change on cropping systems it is important to consider not only elevated CO2 concentration and temperature but also complex interactions among agricultural systems and other factors such as Srd, precipitation, and model-driven processes (Lobell et al. 2011; Asseng et al. 2011).

We showed that CO2 fertilization has a positive effect on wheat yield and WUE. For the latter two variables, there were considerable differences between CO2 fertilization and non-CO2-fertilization treatments. As a C3 crop, CO2 fertilization effects under irrigated (rain fed) conditions offset yield loss by 10.51%–15.80% (13.22%–15.80%), 20.84%–7.78% (25.60%–32.22%), and 33.04%–45.96% (39.36%–9.94%) for the 1°, 2°, and 3°C GMT changes, respectively. The effects of CO2 fertilization on grain yield increased with temperature. This confirmed that elevated CO2 concentration compensates for the negative effects of warming temperature on agricultural production.

Because of different planting systems and trends in climate change, there were large differences in ET under irrigated and rain-fed conditions. This implies that the impact of climate change on agricultural systems varies by region. Because water resources were generally not evenly distributed in space and time, ET deficit was a significant stress during wheat growth (especially at the milking-to-grain-filling stage). Since the 1960s, rainfall has declined by 30 mm per year in China. In fact, future climate-change predictions suggest further drastic reductions of rainfall, especially in northern China (Liu et al. 2005). High WUE is therefore critical for offsetting rising water demand and for sustainable agricultural production and food security.

b. Comparisons with previous studies

Several comparable studies used crop models to simulate the impact of warming temperatures on wheat production. Xiong et al. (2005) simulated climate-change impacts on winter wheat yield in north China, using A2 and B2 greenhouse gas emission scenarios along with the Regional Climate Model–Providing Regional Climates for Impacts Studies (RCM-PRECIS) and CERES-Wheat models. They noted increases in maximum yield, average yield, and yield variation under both A2 and B2 climate-change scenarios for the 2020s, 2050s, and 2080s. On average, grain yield increased by ≈10% as a result of CO2 fertilization effects (Xiong et al. 2005).

Using the same models and climate scenarios, Lin et al. (2005) noted 11.0%–13.3% (from −5.6% to −0.5%), 14.2%–25.1% (from −6.7% to −2.2%), and 25.5%–40.3% (from −8.9% to −8.4%) changes of average wheat yield in the 2020s, 2050s, and 2080s under irrigated conditions with (without) CO2 fertilization effects, respectively. Corresponding changes of average yield under rain-fed conditions with (without) CO2 fertilization effects were 4.5%–15.4% (from −18.5% to −10.2%), 6.6%–20.0% (from −20.4% to −11.4%), and 12.7%–23.6% (from −21.7% to −12.9%) in the 2020s, 2050s, and 2080s, respectively.

Ju et al. (2005) used CERES-Wheat to simulate climate-change impacts on winter wheat yield variability in China, reporting an average yield loss of 20% in 2070. They also noted a slight drop in rain-fed wheat production in favor of irrigated wheat production. Yield loss under rain-fed conditions was also higher that that under irrigated conditions. With changing climatic conditions in the 2070s, the decline of spring or spring–winter hybrid wheat production was greater than winter or winter-hybrid wheat production (Ju et al. 2005). The regions of remarkable yield loss were areas of northeast China with spring wheat and southwest China with winter wheat.

Thomson et al. (2006) used the Environmental Policy Integrated (EPIC) model in the Huang-Hai Plain of China under A2 and B2 conditions. They showed average winter wheat yield increases of 0.2 and 0.8 Mg ha−1, owing to higher nighttime temperatures and precipitation, respectively. Amthor (2001) showed that doubling CO2 concentration from 350 to 700 ppm increased grain yield by 31%. Fulco and Asseng (2006) reported that higher CO2 concentration increased plant productivity and grain yield, especially in water-limited environments with high soil nutrient availability. Doubling atmospheric CO2 concentration also increased yields in acid sandy loams at low and high nitrogen fertilization by 38% and 48%, respectively (Fulco and Asseng 2006).

In our study, median values of projected wheat yield under irrigated (rain fed) conditions with CO2 fertilization effects increased by 6.00%–9.36% (6.00%–15.05%), 12.98%–18.08% (10.21%–39.42%), and 9.06%–26.38% (9.06%–48.71%) for GMT+1D, GMT+2D, and GMT+3D treatments, respectively. These results are consistent with those (10%–25%) reported for C3 crops by Solomon et al. (2007) at 550-ppm atmospheric CO2 concentration. The result for CO2 fertilization effects under free-air CO2-enrichment (FACE) conditions on crop productivity was less than expected, however (Leakey et al. 2009). Although several studies reported that the increase of C3 photosynthesis under FACE conditions was greater than that of biomass or yield (Nowak et al. 2004; Ainsworth and Long 2005), corresponding simulated values were still lower than theoretical and observation values (Long et al. 2004, 2006). According to Long et al. (2006), average increase of light-saturated photosynthesis in 45 cultivars in 11 different FACE conditions was 13%; that of aboveground biomass was 17% and that of yield was 16%, with elevated CO2 concentration. Probable reasons for the projected smaller losses or gains in winter yield relative to previous studies were the warming temperatures, rising CO2 concentrations, declining Srd, and shortening growth periods with increasing GMT. Higher temperatures tend to limit grain yield by reducing growth-period lengths, which in turn limits radiation interception and biomass accumulation (Mitchell et al. 1993; Mearns et al. 1997; Van Oijen and Ewert 1999; Lawlor and Mitchell 2000).

c. Uncertainties

We projected wheat yield and WUE along with their uncertainties, using 20 climate scenarios, a process-based crop model, and probabilistic analysis of median Tmax, Tmin, Srd, and Pre. This approach was helpful to reduce uncertainties associated with single emission scenarios and has been widely accepted in recent literature (Challinor et al. 2005; Marletto et al. 2005; Lobell et al. 2008; Tao et al. 2009; Tao and Zhang 2011). One of the major limitations of the study was ignoring fertilizer and cultivar management measures. Studies have indicated that most seasonal variations in crop response to elevated CO2 were directly or indirectly associated with nitrogen uptake (Leakey et al. 2009; Asseng et al. 2011). Since potential yield varies with changes in climatic conditions, a critical management decision therefore has to do with crop adaptation measures. Hence, future work should focus on assessing the effects of nutrient and cultivar management measures. On the other hand, various genetic coefficient values in simulations can affect projection results. As with local climate conditions and cultivation habits, wheat type and cultivar can vary temporally and spatially. Measured experimental data for each study station were used herein to calibrate and validate the genetic coefficient, which may be helpful in understanding physical and biological bases for the yield changes and reduce uncertainties caused by model simulation. In addition, most crop-growth models are currently based on a single plot or field. Scale transformation for model application to different regions can improve further as an effective simulation tool for studying impacts of climatic change on agriculture production.

5. Conclusions

A probabilistic-assessment, process-based crop model was adapted for projecting impacts of climate change on wheat productivity in China, under baseline and 1°, 2°, and 3°C GMT increases, with and without consideration for CO2 fertilization effects. We also analyzed multiple climate/emission scenarios and related uncertainties. Results indicate a shortening wheat growth period with the GMT changes, under irrigated and rain-fed conditions. Temperature was the main factor driving wheat growth period under the simulated climate-change scenarios. There was high probability of decline in yield and WUE under higher temperature scenarios, without CO2 fertilization effects. In contrast, there was high probability of increasing wheat yield and WUE under higher temperature scenarios, with CO2 fertilization effects. This implies that elevated CO2 concentration generally compensates for the negative effects of warming temperatures on agricultural systems. Further, the positive effects of elevated CO2 concentration on grain yield increased with warming temperatures.

In conclusion, we successfully applied this probabilistic assessment approach, driven by a process-based crop model, to projecting crop productivity and water use in China under various GMT and climate conditions. This approach should be tested using nutrient and cultivar management measures.

Acknowledgments

This study was supported by the National Science Foundation of China (41071030) and was partly supported by the National Key Technology R&D Program (Project 2012BAC19B01) and the National Key Program for Developing Basic Science (Project 2010CB950902) of China. The support of the “Hundred Talents” Program of the Chinese Academy of Sciences is also duly acknowledged. We are grateful for the data support provided by the China Ecosystem Research Network. We also thank Dr. J. P. Moiwo and other anonymous reviewers and editors for raising insightful points/comments on the manuscript.

REFERENCES

REFERENCES
Aggarwal
,
P. K.
,
N.
Kalra
,
S.
Chander
, and
H.
Pathak
,
2006a
:
InfoCrop: A dynamic simulation model for the assessment of crop yields, losses due to pests, and environmental impact of agro-ecosystems in tropical environments. I. Model description
.
Agric. Syst.
,
89
,
1
25
.
Aggarwal
,
P. K.
, and
Coauthors
,
2006b
:
InfoCrop: A dynamic simulation model for the assessment of crop yields, losses due to pests, and environmental impact of agro-ecosystems in tropical environments II. Performance of the model
.
Agric. Syst.
,
89
,
47
67
.
Ainsworth
,
E. A.
, and
S. P.
Long
,
2005
:
What have we learned from 15 years of free-air CO2 enrichment (FACE)? A meta-analytic review of the responses of photosynthesis, canopy
.
New Phytol.
,
165
,
351
371
.
Allen
,
R.
,
1997
:
Self-calibrating method for estimating solar radiation from air temperature
.
J. Hydrol. Eng.
,
2
,
56
67
.
Amthor
,
J. S.
,
2001
:
Effects of atmospheric CO2 concentration on wheat yield: Review of results from experiments using various approaches to control CO2 concentration
.
Field Crops Res.
,
73
,
1
34
.
Anwar
,
M. R.
,
G.
O’Leary
,
D.
McNeil
,
H.
Hossain
, and
R.
Nelson
,
2007
:
Climate change impact on rainfed wheat in south-eastern Australia
.
Field Crops Res.
,
104
,
139
147
.
Asseng
,
S.
,
I.
Foster
, and
N. C.
Turner
,
2011
:
The impact of temperature variability on wheat yields
.
Global Change Biol.
,
17
,
997
1012
.
Bannayan
,
M.
,
N. M. J.
Crout
, and
G.
Hoogenboom
,
2003
:
Application of the CERES-Wheat model for within-season prediction of winter wheat yield in the United Kingdom
.
Agron. J.
,
95
,
114
125
.
Challinor
,
A. J.
, and
T. R.
Wheeler
,
2008
:
Crop yield reduction in the tropics under climate change: Processes and uncertainties
.
Agric. For. Meteor.
,
148
,
343
356
.
Challinor
,
A. J.
,
J. M.
Slingo
,
T. R.
Wheeler
, and
F. J.
Doblas-Reyes
,
2005
:
Probabilistic simulations of crop yield over western India using the DEMETER seasonal hindcast ensembles
.
Tellus
,
57A
,
498
512
.
Challinor
,
A. J.
,
T.
Wheeler
,
C.
Garforth
,
P.
Craufurd
, and
A.
Kassam
,
2007
:
Assessing the vulnerability of food crop systems in Africa to climate change
.
Climatic Change
,
83
,
381
399
.
Challinor
,
A. J.
,
E. S.
Simelton
,
E. D. G.
Fraser
,
D.
Hemming
, and
M.
Collins
,
2010
:
Increased crop failure due to climate change: Assessing adaptation options using models and socio-economic data for wheat in China
.
Environ. Res. Lett.
,
5
,
034012
,
doi:10.1088/1748-9326/5/3/034012
.
Dhungana
,
P.
,
K. M.
Eskridge
,
A.
Weiss
, and
P. S.
Baenziger
,
2006
:
Designing crop technology for a future climate: An example using response surface methodology and the CERES-Wheat model
.
Agric. Syst.
,
87
,
63
79
.
Eitzinger
,
J.
,
M.
Stastna
,
Z.
Zalud
, and
M.
Dubrovský
,
2003
:
A simulation study of the effect of soil water balance and water stress on winter wheat production under different climate change scenarios
.
Agric. Water Manage.
,
61
,
195
217
.
Fang
,
Q. X.
,
L.
Ma
,
T. R.
Green
,
Q.
Yu
,
T. D.
Wang
, and
L. R.
Ahuja
,
2010
:
Water resources and water use efficiency in the North China Plain: Current status and agronomic management options
.
Agric. Water Manage.
,
97
,
1102
1116
.
Fulco
,
L.
, and
S.
Asseng
,
2006
:
Climate change impacts on wheat production in a Mediterranean environment in Western Australia
.
Agric. Syst.
,
90
,
159
179
.
Gbetibouo
,
G.
, and
R.
Hassan
,
2005
:
Economic impact of climate change on major South African field crops: A Ricardian approach
.
Global Planet. Change
,
47
,
143
152
.
Gobin
,
A.
,
2010
:
Modelling climate impacts on crop yields in Belgium
.
Climate Res.
,
44
,
55
68
.
Gong
,
P.
,
Y. Y.
Yin
, and
C. Q.
Yu
,
2011
:
China: Invest wisely in sustainable water use
.
Science
,
331
,
1264
1265
.
Hargreaves
,
G. L.
,
G. H.
Hargreaves
, and
P.
Riley
,
1985
:
Irrigation water requirement for the Senegal River Basin
.
J. Irrig. Drain. Eng.
,
111
,
265
275
.
Houghton
,
J. T.
,
Y.
Ding
,
D. J.
Griggs
,
M.
Noguer
,
P. J.
van der Linden
,
X.
Dai
,
K.
Maskell
, and
C. A.
Johnson
, Eds.,
2001
: Climate Change 2001: The Scientific Basis. Cambridge University Press, 881 pp.
Jones
,
J. W.
, and
Coauthors
,
2003
:
The DSSAT cropping system model
.
Eur. J. Agron.
,
18
,
235
265
.
Ju
,
H.
,
W.
Xiong
,
Y. L.
Xu
, and
D.
Lin
,
2005
:
Impacts of climate change on wheat in China
.
Acta Agron. Sin.
,
31
,
1340
1343
.
Laux
,
P.
,
G.
Jackel
,
R. M.
Tingem
, and
H.
Kunstmann
,
2010
:
Impact of climate change on agricultural productivity under rainfed conditions in Cameroon—A method to improve attainable crop yields by planting date adaptations
.
Agric. For. Meteor.
,
150
,
1258
1271
.
Lawlor
,
D. W.
, and
R. A. C.
Mitchell
,
2000
: Crop ecosystems responses to climatic change: Wheat. Climate Change and Global Crop Productivity, K. R. Reddy and H. F. Hodges, Eds., CAB International, 57–80.
Leakey
,
A. D. B.
,
E. A.
Ainsworth
,
C. J.
Bernacchi
,
A.
Rogers
,
S. P.
Long
, and
D. R.
Ort
,
2009
:
Elevated CO2 effects on plant carbon, nitrogen, and water relations: Six important lessons from FACE
.
J. Exp. Bot.
,
60
,
2859
2876
.
Lin
,
E.
,
W.
Xiong
,
H.
Ju
,
Y.
Xu
,
Y.
Li
,
L.
Bai
, and
L.
Xie
,
2005
:
Climate change impacts on crop yield and quality with CO2 fertilization in China
.
Philos. Trans. Roy. Soc. London
,
B360
,
2149
2154
.
Liu
,
C. M.
,
C. Q.
Zhou
,
S. F.
Zhang
, and
X. L.
Wang
,
2005
:
Study on water production function and efficiency of wheat
.
Geogr. Res.
,
24
,
1
10
.
Liu
,
Y. J.
, and
E. D.
Lin
,
2008
:
Impact of climate warming in the past 20 years on agriculture in different regions of China
.
Adv. Climate Change Res.
,
4
,
51
55
.
Liu
,
Y. J.
, and
G. F.
Yuan
,
2010
:
Impacts of climate change on winter wheat growing process and production under B2 climate scenario in Panzhuang irrigation district
.
J. Geogr. Sci.
,
6
,
861
875
.
Lobell
,
D. B.
, and
M. B.
Burke
,
2008
:
Why are agricultural impacts of climate change so uncertain? The importance of temperature relative to precipitation
.
Environ. Res. Lett.
,
3
,
034007
,
doi:10.1088/1748-9326/3/3/034007
.
Lobell
,
D. B.
,
M. B.
Burke
,
C.
Tebaldi
,
M. D.
Mastrandrea
,
W. P.
Falcon
, and
R. L.
Naylor
,
2008
:
Prioritizing climate change adaptation needs for food security in 2030
.
Science
,
19
,
607
610
.
Lobell
,
D. B.
,
W.
Schlenker
, and
J. C.
Roberts
,
2011
:
Climate trends and global crop production since 1980
.
Science
,
333
,
616
620
.
Long
,
S. P.
,
E. A.
Ainsworth
,
A.
Rogers
, and
D. R.
Ort
,
2004
:
Rising atmospheric carbon dioxide: Plants FACE the future
.
Annu. Rev. Plant Biol.
,
55
,
591
628
.
Long
,
S. P.
,
E. A.
Ainsworth
,
A. D. B.
Leakey
, and
D. R.
Ort
,
2006
:
Food for thought: Lower-than-expected crop yield stimulation with rising CO2 conditions
.
Science
,
312
,
1918
1921
.
Luo
,
Q.
,
M.
Williams
,
W.
Bellotti
, and
B.
Bryan
,
2003
:
Quantitative and visual assessments of climate change impacts on South Australian wheat production
.
Agric. Syst.
,
77
,
173
186
.
Marletto
,
V.
, and
Coauthors
,
2005
:
Evaluation of downscaled DEMETER multi-model ensemble seasonal hindcasts in a northern Italy location by means of a model of wheat growth and soil water balance
.
Tellus
,
57A
,
488
497
.
Mearns
,
L. O.
,
C.
Rosenzweig
, and
R.
Goldberg
,
1997
:
Mean and variance change in climate scenarios: Methods, agricultural applications, and measures of uncertainty
.
Climatic Change
,
35
,
386
396
.
Mitchell
,
R. A. C.
,
V. J.
Mitchell
,
S. P.
Driscoll
,
J.
Franklin
, and
D. W.
Lawlor
,
1993
:
Effects of increased CO2 concentration and temperature on growth and yield of winter-wheat at 2 levels of nitrogen application
.
Plant Cell Environ.
,
16
,
521
529
.
Mitchell
,
T. D.
,
T. R.
Carter
,
P. D.
Jones
,
M.
Hulme
, and
M.
New
,
2004
: A comprehensive set of high-resolution grids of monthly climate for Europe and the globe: The observed record (1901–2000) and 16 scenarios (2001–2100). University of East Anglia Tyndall Centre for Climate Change Research Working Paper 55, 30 pp. [Available online at http://www.tyndall.ac.uk/sites/default/files/wp55.pdf.]
Mo
,
X. G.
,
S.
Liu
,
D.
Chen
, and
Z.
Lin
,
2009
:
Scale effects on actual evapotranspiration and gross primary production over a large basin
.
Hydrol. Sci. J.
,
54
,
160
173
.
National Bureau of Statistics of China
,
2010
: China Agriculture Yearbook. China Agricultural Press, 258 pp.
Nowak
,
R. S.
,
E. S.
Ellsworth
, and
S. D.
Smith
,
2004
:
Functional responses of plants to elevated atmospheric CO2: Do photosynthetic and productivity data from FACE experiments support early predictions?
New Phytol.
,
162
,
253
280
.
Panda
,
R. K.
,
S. K.
Behera
, and
P. S.
Kashyap
,
2004
:
Effective management of irrigation water for maize under stressed conditions
.
Agric. Water Manage.
,
66
,
181
203
.
Peart
,
R. M.
,
J. W.
Jones
,
R. B.
Curry
,
K. J.
Boote
, and
L. H.
Allen
,
1988
: Final report: Impact of the climate change on crop yield in the southern U.S.A.: A simulation study. Institute of Food and Agricultural Sciences, University of Florida.
Persson
,
T.
,
A.
Garcia y Garcia
,
J. O.
Paz
,
B. V.
Ortiz
, and
G.
Hoogenboom
,
2010
:
Simulating the production potential and net energy yield of maize-ethanol in the southeastern USA
.
Eur. J. Agron.
,
32
,
272
279
.
Popova
,
Z.
, and
M.
Kercheva
,
2005
:
CERES model application for increasing preparedness to climate variability in agricultural planning—Risk analyses
.
Phys. Chem. Earth
,
30
,
117
124
.
Reyenga
,
P. J.
,
S. M.
Howden
,
H.
Meinke
, and
G. M.
McKeon
,
1999
:
Modelling global change impacts on wheat cropping in South-East Queensland, Australia
.
Environ. Model. Software
,
14
,
297
306
.
Rinaldi
,
M.
,
2004
:
Water availability at sowing and nitrogen management of durum wheat: A seasonal analysis with CERES-Wheat model
.
Field Crops Res.
,
89
,
27
37
.
Ritchie
,
J. T.
,
1985
: A user-oriented model of the soil water balance in wheat. Wheat Growth and Modeling, W. Day and R. K. Atkin, Eds., NATO-ASI Series A, Vol. 86, Plenum, 293–305.
Rotter
,
R. P.
, and
Coauthors
,
2011
:
What would happen to barley production in Finland if global warming exceeded 4 degrees C? A model-based assessment
.
Eur. J. Agron.
,
4
,
205
214
.
Roudier
,
P.
,
B.
Sultan
,
P.
Quirion
, and
A.
Berg
,
2011
:
The impact of future climate change on West African crop yields: What does the recent literature say?
Global Environ. Change
,
21
,
1073
1083
.
Ruiz-Ramos
,
M.
, and
M. I.
Mínguez
,
2010
:
Evaluating uncertainty in climate change impacts on crop productivity in the Iberian Peninsula
.
Climate Res.
,
44
,
69
82
.
Saseendran
,
S. A.
,
D. C.
Nielsen
,
L.
Ma
,
L. R.
Ahuja
, and
A. D.
Halvorson
,
2004
:
Modelling nitrogen management effects on winter wheat production using RZWQM and CERES-Wheat
.
Agron. J.
,
96
,
615
630
.
Semenov
,
M. A.
,
2007
:
Development of high-resolution UKCIP02-based climate change scenarios in the UK
.
Agric. For. Meteor.
,
144
,
127
138
.
Semenov
,
M. A.
,
R. J.
Brooks
,
E. M.
Barrow
, and
C. W.
Richardson
,
1998
:
Comparison of the WGEN and LARS-WG stochastic weather generators for diverse climates
.
Climate Res.
,
10
,
95
107
.
Smith
,
S. R.
,
1996
: Agricultural Recycling of Sewage Sludge and the Environment. Oxford University Press, 400 pp.
Solomon
,
S.
,
D.
Qin
,
M.
Manning
,
M.
Marquis
,
K.
Averyt
,
M. M. B.
Tignor
,
H. L.
Miller
Jr.
, and
Z.
Chen
, Eds.,
2007
: Climate Change 2007: The Physical Science Basis. Cambridge University Press, 996 pp.
Tao
,
F. L.
, and
Z.
Zhang
,
2011
:
Impacts of climate change as a function of global mean temperature: Maize productivity and water use in China
.
Climatic Change
,
105
,
409
432
.
Tao
,
F. L.
,
M.
Yokozawa
,
Y.
Hayashi
, and
E.
Lin
,
2003
:
Future climate change, the agricultural water cycle, and agricultural production in China
.
Agric. Ecosyst. Environ.
,
95
,
203
215
.
Tao
,
F. L.
,
M.
Yokozawa
,
Y. L.
Xu
,
Y.
Hayashi
, and
Z.
Zhang
,
2006
:
Climate change and trends in phenology and yields of field crops in China, 1981–2000
.
Agric. For. Meteor.
,
138
,
82
92
.
Tao
,
F. L.
,
Y.
Hayashi
,
Z.
Zhang
,
T.
Sakamoto
, and
M.
Yokozawa
,
2008a
:
Global warming, rice production and water use in China: Developing a probabilistic assessment
.
Agric. For. Meteor.
,
148
,
94
110
.
Tao
,
F. L.
,
M.
Yokozawa
,
J. Y.
Liu
, and
Z.
Zhang
,
2008b
:
Climate–crop yield relationships at provincial scales in China and the impacts of recent climate trends
.
Climate Res.
,
38
,
83
94
.
Tao
,
F. L.
,
Z.
Zhang
,
J. Y.
Liu
, and
M.
Yokozawa
,
2009
:
Modelling the impacts of weather and climate variability on crop productivity over a large area: A new super-ensemble-based probabilistic projection
.
Agric. For. Meteor.
,
149
,
1266
1278
.
Thomson
,
A. M.
,
R. C.
Izaurralde
,
N. J.
Rosenberg
, and
X. X.
He
,
2006
:
Climate change impacts on agriculture and soil carbon sequestration potential in the Huang-Hai Plain of China
.
Agric. Ecosyst. Environ.
,
114
,
195
209
.
Trnka
,
M.
,
M.
Dubrovsky
,
D.
Semeradova
, and
Z.
Zalud
,
2004
:
Projections of uncertainties in climate change scenarios into expected winter wheat yields
.
Theor. Appl. Climatol.
,
77
,
229
249
.
Van Ittersum
,
M. K.
,
S. M.
Howden
, and
S.
Asseng
,
2003
:
Sensitivity of productivity and deep drainage of wheat cropping systems in a Mediterranean environment to changes in CO2, temperature and precipitation
.
Agric. Ecosyst. Environ.
,
97
,
255
273
.
Van Oijen
,
M.
, and
F.
Ewert
,
1999
:
The effects of climatic variation in Europe on the yield response of spring wheat cv. minaret to elevated CO2 and O3: An analysis of open-top chamber experiments by means of two crop growth simulation models
.
Eur. J. Agron.
,
10
,
249
264
.
Xiong
,
W.
,
Y. L.
Xu
, and
E. D.
Lin
,
2005
:
The simulation of yield variability of winter wheat and its corresponding adaptation options under climate change
.
Chin. Agric. Sci. Bull.
,
21
,
380
380
.
Yang
,
L. X.
,
J. Y.
Huang
,
H. J.
Yang
,
G.
Dong
,
G.
Liu
,
J.
Zhu
, and
Y.
Wang
,
2006
:
Seasonal changes in the effects of free-air CO2 enrichment (FACE) on dry matter production and distribution of rice (Oryza sativa L.)
.
Field Crops Res.
,
98
,
12
19
.