Increasing crop yield and water use efficiency (WUE) in dryland farming requires a quantitative understanding of relationships between crop yield and the water balance over many years. Here, we report on a long-term dryland monitoring site at the Loess Plateau, Shanxi, China, where winter wheat was grown for 30 consecutive years and soil water content (0–200 cm) was measured every 10 days. The monitoring data were used to calibrate the AquaCrop model and then to analyse the components of the water balance. There was a strong positive relationship between total available water and mean cereal yield. However, only one-third of the available water was actually used by the winter wheat for crop transpiration. The remaining two-thirds were lost by soil evaporation, of which 40 and 60% was lost during the growing and fallow seasons, respectively. Wheat yields ranged from 0.6 to 3.9 ton/ha and WUE from 0.3 to 0.9 kg/m3. Results of model experiments suggest that minimizing soil evaporation via straw mulch or plastic film covers could potentially double wheat yields and WUE. We conclude that the relatively low wheat yields and low WUE were mainly related to (i) limited rainfall, (ii) low soil water storage during fallow season due to large soil evaporation, and (iii) poor synchronisation of the wheat growing season to the rain season. The model experiments suggest significant potential for increased yields and WUE.
Citation: Qin W, Chi B, Oenema O (2013) Long-Term Monitoring of Rainfed Wheat Yield and Soil Water at the Loess Plateau Reveals Low Water Use Efficiency. PLoS ONE 8(11): e78828. https://doi.org/10.1371/journal.pone.0078828
Editor: Eric Gordon Lamb, University of Saskatchewan, Canada
Received: March 13, 2013; Accepted: September 16, 2013; Published: November 26, 2013
Copyright: © 2013 Qin et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: The authors have no support or funding to report.
Competing interests: The authors have declared that no competing interests exist.
Water scarcity is a growing global concern –. For rainfed agriculture, this pressure may become more severe under climate change due to the expected more erratic rainfall and longer dry spells –. Currently, rainfed agriculture covers 80% of the world’s cultivated land and accounts for 60% of crop production . However, crop yield and water use efficiency (WUE) are often low in rainfed agriculture, especially in arid and semi-arid areas due to, for example, degraded soils, erratic rainfall and poor water management . Many of these areas in Africa and Asia face also rapid population growth. Hence, there is a pressing need to increase crop yields and WUE in rainfed agriculture –.
Water use efficiency is commonly defined as crop yield over evapotranspiration (ET), where ET is the sum of soil evaporation (E) and crop transpiration (T) –. The latter (T) is a direct consequence of crop production, while E is ‘unproductive’ water loss. The central question for rainfed agriculture in arid and semi-arid regions is ‘how to transform unproductive water loss (E) into productive water use (T)’. Unfortunately, the partitioning between E and T is often not well-known, due to difficulties and high cost in distinguishing E and T in the field. As a result, water use of crops is commonly reported as evapotranspiration (ET) , , . This lack of information makes it difficult to assess how much of the evaporative water loss can be used for increasing yields by appropriate measures.
Crop growth simulation models have the potential of providing more comprehensive insights into the functioning of soil-crop systems, and can be helpful to explore options for increasing yield and WUE –. These models though are simplified representations of parts of reality and therefore require testing in the real world. Fortunately, numerous field studies have examined the effect of water availability, with or without irrigation treatments, on crop yield and WUE –. In principle, these results can be used to calibrate and test the crop growth simulation models. However, most of these field studies are short-term (2–5 years) and focus on the growing season only, largely ignoring the water balance during the fallow season. Furthermore, crop yields and WUE show large variations due to differences in soils, climate conditions and crop husbandry practices. For example, in the North China Plain, annual precipitation ranges from 400 to 650 mm and wheat grain yields roughly range from 1 to 3 ton/ha/year under rainfed conditions. With 200 to 300 mm of irrigation, grain yield can be increased by 60 to 100% and WUE can be increased by 20 to 40% . Globally, WUE of wheat shows even larger variation ranging from approximately 0.2 to 1.8 kg/m3 . Evidently, this wide range suggests considerable scope for improvement, but the underlying causes of WUE is not always well-known and irrigation water is not available on most places.
The study reported here has the objectives (i) to calibrate a water-driven crop growth model on the basis of monitoring data from a long-term field experiment with rainfed winter wheat, (ii) to analyse the components of the water balance of this field, and (iii) to explore the potential for increasing crop yield and WUE through model experiments. The field experiment was situated in the dryland of the Loess Plateau, Shanxi, China. Winter wheat was grown for 30 consecutive years and soil water content (0–200 cm) was measured every 10 days. The FAO AquaCrop model was chosen as simulation model because it is a water-driven crop growth model that can separately calculate E and T, and simulates the final crop yield as function of water use , . The AquaCrop model uses canopy ground cover as the basis to calculate T and to separate E and T. Crop yield is then calculated as the product of biomass and harvest index (HI). The principles and modules of the AquaCrop model are well-documented in a series of AquaCrop publications , –. Compared to some other crop growth models, AquaCrop requires relatively few input parameters , . Such a limited number of input parameters facilitates model calibration and utilization for different crops and under different management strategies , , , –.
Materials and Methods
The long-term monitoring site is located in Beizhang, Linyi county in Shanxi province on the Loess Plateau (35° 9′ 3.83″ N, 110° 34′ 25.40″ E, Altitude: 491 m) in China. The authority is Dryland Agriculture Research Centre, Shanxi Academy of Agricultural Sciences. We have the permission to conduct the study on this site. We confirm that the field studies did not involve endangered or protected species.
The site has a semi-arid climate with extensive monsoonal influence, which is dry and cold in winter, rainy and hot in summer. Rainfall in June to September accounts for more than 70% of annual rainfall. Average annual rainfall was 517 mm in the period of 1980–2010 with large annual variations, from a minimum of 331 mm to a maximum of 832 mm. Mean annual sunshine duration is 2270 h, annual average temperature 13.5°C and mean annual potential evaporation is 1340 mm. The soil is a typical Loessial soils (Calcic Luvisols) of the Loess Plateau –. Soil slope is <1% and soil texture is silt loam, with a small proportion of clay. The soil was rather homogeneous in texture and key physical properties (Table 1). Maximum soil water holding capacity was a significant fraction of the total annual rainfall (Table 2).
The top soil (30 cm) contained 5.8 g/kg of organic C and 0.55 g/kg of total N. Mean available N, P and K were 63 mg/kg, 14 mg/kg and 142 mg/kg, respectively, in 2007. Since 1983, mean available P and K have been increased by 7 mg/kg and 10 mg/kg, respectively . Total soil organic C was measured by dry combustion combined with elemental C analysis. Total N was measured by Kjeldahl method. Soil mineral N was extracted with 1 M KCl and analysed by the cadmium reduction method. Available P was extracted with 0.5 M NaHCO3. Available K was extracted with 1.0 M NH4OAc.
Experimental Design and Measurements
The long-term crop yield and soil moisture monitoring experiment started in 1980. There were no experimental treatments. The size of the field is 0.5 ha. Winter wheat was planted each year between 25 September and 5 October, depending on the actual climate and soil water conditions. Sowing rate was 150 kg seed/ha. The growing period of the winter wheat is approximately 245 days (from 1 October to 1 June of the next year). The fallow season is about 120 days (from 2nd June to 30th September) (Figure 1). Local bred cultivars (Jinmai) were used for all years. Information about crop parameters is listed in Table 3. Fertilizers were applied at planting at a rate of 127.5 kg/ha of N as urea and 90 kg/ha of P2O5 as superphosphate, for all years. At harvest, grain yields were measured in 5 random plots (2 m2 per plot) selected from two diagonal lines of the field. Dry matter content of grain and straw was measured at the laboratory after drying at 70°C. Soil water contents were measured in 16 different layers up to 2 m depth every 10 days by using the gravity method. The top 10 cm was sampled in 5 cm intervals. From 0.1 to 1 m, samples were taken at 10 cm intervals, and from 1 to 2 m at 20 cm intervals (Figure 2). Each sample consisted of 5 subsamples, taken randomly on two diagonal lines across the field. Observations during soil sampling over years revealed that spatial variations in soil profile were small.
Black and white boxes show 75 and 25% percentile values. Whiskers show maximum and minimum values (the same applies to other figures). The growing season of winter wheat was 245 days from 1 October to 1 June. of the next year, highlighted in green bar. Fallow season was 120 days from 2nd Jun. to 30th Sep., highlighted in red bar.
The soil water changes for fallow season were calculated as the soil water content of October deducted by that of June. Similarly, the soil water changes for growing season were calculated as the soil water content of June deducted by that of October in the previous year.
Water Balance and WUE Estimations
Mean monthly rainfall data were collected at a near-by meteorological station (Linyi station, 35° 10′ 7.02″ N, 110° 46′ 44.59″ E, Altitude: 441 m), which is 25 km away from the experimental field. The water balance for both fallow and crop growing seasons reads as follows:(1)where R is rainfall, I is irrigation, ΔS is the change of soil water content, E is soil evaporation, T is crop transpiration, Rr is runoff and Pe is percolation. All units are presented in mm or in m3/ha.
Irrigation was not applied in this study. Furthermore, runoff and percolation (leaching) were small and disregarded. Hence, the water balance of the fallow seasons was simplified to:(2)
The water balance of the growing season was simplified to:(3)
We used formula (2) and the recorded rainfall and soil water content data to calculate evaporation during the fallow season and formula (3) to calculate ET during the growing season. The partitioning of E and T was done by the AquaCrop model.
Water use efficiency (WUE, kg/m3) was calculated as:(4)where ET is evapotranspiration (mm), the sum of E and T during the growing season.
Calibration and Validation of the AquaCrop Model
The FAO AquaCrop model is a water-driven crop growth model for the simulation of crop biomass and yield as function of water availability , . AquaCrop requires 4 main sets of input data, i.e. climate data (rainfall, minimum and maximum temperature and reference evapotranspiration (ETo)), crop parameters, soil data and field management data. Climate data were collected from a near-by meteorological station (Linyi station, 35° 10′ 7.02″ N, 110° 46′ 44.59″ E, Altitude: 441 m), which is 25 km away from the experimental field. ETo was calculated by FAO Penman-Monteith equation as described in Allen et al. . Soil data (Tables 1 and 2) and field management data were derived from measurements and from recordings at the monitoring site. Crop data and parameters were derived from measurements, literature , – and by calibration (Table 3).
The objective of calibration and validation was to achieve the best match between simulated outputs and monitoring data (soil moisture and crop yield) for all 30 years, using a common procedure , –. We randomly selected 15 years’ data from the monitoring field for calibration and used the other 15 years’ data for validation of the calibrated model. In the calibration step, we used the year-specific climate, soil and initial soil water data as fixed input data. Then we adjusted some of the crop parameters (see Table 3), based on our understanding of crop growth, development and crop responses to water deficits, until the differences between simulated output and monitoring data were minimal. We repeated this process for all selected 15 years and ultimately obtained a satisfactorily index of agreement (d = 0.92). Then, the calibrated model was validated with the other 15 years’ data, and again obtained an acceptable index of agreement (d = 0.93).
The root mean square error (RMSE)  has been widely used to evaluate the performance of a model , –. However, Willmott and Matsuura  pointed out that, compared to RMSE, the mean absolute error (MAE) is a better indicator and therefore the evaluation of model performance should be based on the MAE. In this study, we used the MAE, the mean bias error (MBE) and the Willmott index of agreement (d) to evaluate the model performance.
The MAE measures the weighted average magnitude of the absolute errors and was calculated as follows:(5)where n is the number of observations, Mi is the modelled yield or soil water content and Oi is the observed yield or soil water content.
The MBE indicates whether the model is under or over predicting the observed values and also indicates the uniformity of error distribution. Positive MBE values indicate over prediction, negative values indicate under prediction and a value of zero indicates equal distribution between negative and positive values. The MBE was calculated as follows:(6)
The Willmott index of agreement (d)  has values ranging from 0 to 1. A value close to 1 suggests a good model performance. The agreement index d was calculated as follows:(7)where is the average value of the observed yield or soil water content.
To investigate the potentials of minimizing soil evaporation so as to increase crop yield and WUE, we set up six model experiments as follows:
Experiment 1 (E1): Reference A; winter wheat was planted at 50% of field capacity (FC) (low soil water content)
Experiment 2 (E2): E1+organic mulch (straw)
Experiment 3 (E3): E1+plastic film cover
Experiment 4 (E4): Reference B; winter wheat was planted at 70% of FC (high soil water content)
Experiment 5 (E5): E4+organic mulch (straw)
Experiment 6 (E6): E4+plastic film cover
In the model experiments, we used E1 and E4 as references to simulate crop growth under relatively low (with 50% FC) and high (with 70% FC) soil water content at winter wheat seeding (Table 2). The range from 50 to 70% FC largely represented the initial soil water content for the planting period. The low value is representative for no water harvesting, and the high values is representative for water harvesting during the fallow period via mulching and/or covers. Model experiments E2 and E5 aimed at testing the effects of organic mulch during the growing season, and experiments E3 and E6 aimed at testing the effects of plastic film during the growing season on crop yield and water balance. The effectiveness of the soil evaporation reduction by organic mulch and plastic film cover during the growing season were estimated at 50 and 90%, respectively, which are default values of the AquaCrop model . However, since plastic film will cover only about 80% of the wheat planted field, the overall soil evaporation reduction by plastic film was set at 72%. In practice, the plastic film covers are used for the growing season only and destroyed or removed after harvesting the crops. We ran these six experiments for the period of 1980–2010.
Patterns of Monthly Rainfall and Soil Water Content
Most of the rainfall occurs in the summer from June to September, accounting for more than 70% of the total annual rainfall while winter wheat was seeded in the end of September or beginning of October, and harvested in the end of May or beginning of June. Hence, most of the rain fell when wheat had matured already, i.e., during the fallow period. Therefore, the growing season of winter wheat was poorly synchronized to the rain season. On the other hand, due to concentrated rainfall in June to September, mean soil water content in the upper two metre was highest in October when winter wheat was seeded. Thereafter, the soil water content gradually decreased, due to soil evaporation and crop transpiration. Mean soil water content was lowest in June, at the beginning of the rain season when the winter wheat was harvested (Figure 1).
The changes of the soil water content (for 16 layers in the upper 2 meter of the soil) during the fallow and growing seasons are shown in Figure 2. The changes in water content in the fallow season mirror the changes during the growing season, i.e. water stored in the soil during the fallow season largely equalled to soil water depletion during the growing season. The soil water changes for fallow season were calculated as the soil water content of October deducted by that of June. Similarly, the soil water changes for growing season were calculated as the soil water content of June deducted by that of October in the previous year. Soil water content mostly changed in the top 1 m, accounting for 70% of total soil water change (±120 mm) (Figure 2). We cannot exclude that some of the seasonal variations are caused by slight spatial variations in soil characteristics.
Model Calibration and Performance
The AquaCrop model was calibrated using the climate data, soil data, field management data and monitoring data. The full set of crop parameters is listed in Table 3. The performance of the model on the simulated crop yield and soil water balance is shown in Figure 3. The relationship between observed and modelled grain yield of the calibrated model for the second set of 15-years data (validation step) was almost as good as for the whole set of data (not shown). For the 30 years’ data, the relationship between observed and modelled grain yield had a correlation coefficient of (R2) of 0.77, and the index of agreement (d) was 0.93 (Figure 3A). Similarly, the relationship between observed and modelled soil water content showed a R2 of 0.78 and an index of agreement (d) of 0.93 (Figure 3B). Mean absolute errors (MAE) were 311 kg/ha and 25 mm, respectively. Mean bias errors (MBE) were −168 kg/ha and −12 mm in simulating yields and soil water balance, respectively, suggesting that the model slightly underestimated grain yields and soil water contents by 8 and 9%, respectively. We conclude that the performance of the AquaCrop model was acceptable for doing further simulations.
Diagonal lines represent 1∶1 lines.
Water Balance and Partitioning of E and T
AquaCrop was used to estimate soil evaporation and crop transpiration, and the water balance for both fallow and growing seasons during the period 1980–2010 (Figure 4). There was only a very marginal change in soil water content when considering the water balance of the total season (fallow+growing season) over the 30 years’ period; total rainfall was nearly equal to the sum of soil evaporation and crop transpiration. Crop transpiration accounted for approximately one-third of total seasonal rainfall. The remaining two-thirds were assumed to be lost by soil evaporation (Figure 4A), but we cannot exclude that a small fraction of this evaporative loss was actually lost by leaching and/or runoff. Mean rainfall in the fallow season was 323 mm, of which 207 mm (64%) was lost by soil evaporation and 116 mm (36%) was stored in the soil (Figure 4B). During the growing season, mean crop transpiration was 185 mm (57%) and soil evaporation was 137 mm (43%), of which rainfall and soil moisture contributed 194 (60% ) and 128 mm (40%), respectively (Figure 4C). Mean rainfall of the growing season was slightly larger than mean crop transpiration.
Total season means the sum of fallow season and growing season.
Wheat Yield and WUE
Due to limited amounts of available water and the irregular rainfall pattern, wheat yield and WUE were rather low, ranging from 0.6 to 3.9 ton/ha/year and from 0.3 to 0.9 kg/m3, respectively. Relationships between yield and annual rainfall (p = 0.028), and between yield and growing season rainfall (p = 0.027) were highly significant (Figures 5). The relationship between yield and rainfall during the fallow season was not significant (p = 0.167). Furthermore, there were significant linear relationships between wheat grain yield and ET, T, and WUE, with coefficients of determination (R2) of 0.61, 0.68 and 0.72, respectively (Figure 5). Given the slope (0.01) between yield and T, an increase of 1 mm available water can produce 10 kg grain per hectare.
The significant level is 0.028 for (A), 0.167 for (B), 0.027 for (C). For D, E and F, the significant level is all smaller than 0.01.
In model experiment E1 (winter wheat planted at 50% of FC), mean wheat yield and WUE were 2 ton/ha and 0.6 kg/m3, respectively (Figure 6). In model experiment E2, with organic mulch during the growing season, mean yield increased to 2.3 ton/ha and WUE increased to 0.8 kg/m3. In model experiment E3, with plastic film during the growing season, mean yield increased further to 2.5 ton/ha and WUE to 1.0 kg/m3.
Experiment E1: Planting at 50% of FC; Experiment E2: E1+organic mulch; Experiment E3: E1+plastic cover; Experiment E4: Planting at 70% of FC; Experiment E5: E4+organic mulch; Experiment E6: E4+plastic cover.
Similarly, in model experiment E4 (winter wheat planted at 70% of FC), mean yield and WUE were 2.9 ton/ha and 0.8 kg/m3, respectively. With organic mulch (E5), mean yield and WUE increased to 3.2 ton/ha and 0.9 kg/m3, and with plastic film cover (E6), mean yield and WUE increased further to 3.5 ton/ha and 1.1 kg/m3. Our model experiments show that both organic mulch and plastic film cover could significantly improve yield and WUE, but the impact of plastic cover was bigger than the organic mulch mainly due to high effectiveness in reducing soil evaporation. Moreover, increasing water storage in the soil during the fallow season, so that available soil water content increases from 50 to 70% of FC in autumn at the time of winter wheat seeding, is at least as effective as a plastic cover during the growing season.
We successfully calibrated and validated the AquaCrop model on the basis of the long-term monitoring data (30 years) of rainfed winter wheat on the Loess Plateau of northern China. The full set of crop parameters (Table 3) may provide also guidance to future studies and further model calibrations and validations. We also quantified four key components of the water balance, i.e., rainfall, changes in soil water, soil evaporation and crop transpiration, for both fallow and growing seasons by combining empirical data from a long-term wheat monitoring site with calculated results using the AquaCrop model. In the end, we also explored the potential for increasing wheat yield and WUE through model experiments.
The relationship between observed and modelled wheat yield had a R2 of 0.77, slope of 0.9 and an index of agreement (d) of 0.93 (Figure 3). Mkhabela and Bullock  reported a R2 of 0.66, slope of 0.96, index of agreement (d) of 0.99 between observed and modelled wheat yields. Araya et al.  reported a R2>0.80 when simulating barley biomass and grain yield. Stricevic et al.  reported a R2>0.84 when simulating yields of maize, sunflower and sugar beet. Similarly for simulating soil water content, we found a R2 of 0.78, slope of 0.9 and an index of agreement (d) of 0.93 (Figure 3). Mkhabela and Bullock  reported a R2 of 0.9 and a slope of 0.9 for simulating soil water content. Hence, the performance of AquaCrop for our dryland wheat field is largely comparable with that of other modelling studies.
Water stress limited the crop yield at this site. According to our model simulations, water stress has led to suboptimal yields in essentially all years, for both low and relatively high wheat yields. For example, a very low grain yield (0.6 ton/ha) was recorded in the year 2000, when water stress for leaf expansion and stomatal closure started already at the 54th day after planting. In contrast, water stress for leaf expansion and stomatal closure started to occur only from day 156 after planting in 2003, when grain yield was 3.8 ton/ha. In both cases, water stress occurred before flowering stage (∼200 days after planting). Water stress leads to low grain yield and low WUE, but depending on the stage and duration of the water stress , .
We found significant linear relationships between wheat yield and ET, and between wheat yield and T (Figure 5), in line with some other studies , , . Mean T/ET ratio in our study was only 58% during the growing season, which was 8–12% lower than that reported by Liu et al.  and Kang et al.  but highly in line with that reported by Wang et al. . The higher T/ET ratio in the studies of Liu et al.  and Kang et al.  was probably due to the irrigation treatments where the crop had more water for transpiration. It is well-known that crop yield and WUE are often lower in rainfed agriculture than in irrigated agriculture , , , but depending also on possible nutrient, weed, and disease stresses and irrigation management.
Advanced technologies, such as precision irrigation, are for a long time available but unfortunately not affordable and applicable to the farmers of the Loess Plateau, mainly because of the high cost relative to the low value of cereals . Therefore, we focused on low-cost options, such as straw mulch and plastic film cover because those are the most accessible and low cost materials for farmers to implement in the field. Minimizing soil evaporation could save water for crop transpiration, and thereby increase wheat yield and WUE. Our model experiments suggest that wheat yield can be improved significantly by minimizing soil evaporation via organic mulch and plastic film cover, especially also during the fallow period. Mulching with crop residues can decrease soil evaporation and increase soil water retention. Plastic film cover can significantly increase crop yield and WUE, and promote crop growth during early growth when temperature is low. Our results show that crop yields can be increased by ∼0.9 ton/ha through increasing soil water storage during the fallow period. Crop yields can be increased further by on average ∼0.3 ton/ha through straw mulch and by ∼0.5 ton/ha on through plastic film covers during the growing season. At the same time, WUE increases on average by 0.2 to 0.6 kg/m3. These results are in line with results reported by Deng et al.  and others –.
Evidently, increased rainwater harvesting during the fallow season is an effective option. Straw mulch significantly reduces the evaporative water losses during the fallow season and is conducive to the infiltration of rain water in the soil. Plastic film covers are less applicable during the fallow season, because they may limit the infiltration of rain water and thereby increase runoff. Reduced tillage can also improve soil water storage. According to a recent study of Hou et al. , rotational tillage (rotation of no-tillage and subsoiling) could significantly increase soil water storage during the summer fallow and wheat growing season compared with conventional tillage. They found that rotational tillage increased wheat yields by 10%, and WUE by 7.5%, respectively.
Low wheat yield at the monitoring site was largely due to (i) limited rainfall, (ii) low soil water storage during fallow season because of high water loss via soil evaporation, and (iii) the poor synchronisation of the wheat growing season to the rainfall distribution season. Although water was limited, on average only one-third of the total available water was actually used by the crop for transpiration. The remaining two-thirds was lost by soil evaporation, 60% during the fallow season and 40% during the growing season. Our model experiments suggest that minimizing soil evaporation via organic mulch or plastic film covers can significantly increase wheat yield and WUE. More importantly, these increases can be realized by the application of relatively low cost measures. Further studies are needed to test the effectiveness of these measures in the field.
We thank Xiaoguang Yang, from College of Resources and Environmental Sciences, China Agricultural University and Junqi Zhu, from Wageningen University and Research Centre, for providing the climate data for this study. We also thank Xinxin Wang, from Wageningen University and Research Centre, for helping us with statistics.
Conceived and designed the experiments: WQ BC OO. Performed the experiments: WQ BC. Analyzed the data: WQ BC. Contributed reagents/materials/analysis tools: WQ BC OO. Wrote the paper: WQ BC OO.
de Wit CT (1958) Transpiration and crop yields. Verslag Landbouwkundig Onderzoek, Wageningen, Netherlands: Wageningen University.
Marschner H (1995) Mineral Nutrition of Higher Plants. Academic Press, San Diego: 889 p.
- 3. Piao SL, Ciais P, Huang Y, Shen ZH, Peng SS, et al. (2010) The impacts of climate change on water resources and agriculture in China. Nature 467: 43–51.
- 4. Hoekstra AY, Mekonnen MM (2012) The water footprint of humanity. Proceedings of the National Academy of Sciences of the United States of America 109: 3232–3237.
IPCC (2007) Climate Change 2007: Impacts, Adaptation and Vulnerability. Contribution of Working Group II to the Fourth Assessment Report of the International Panel of Climate Change. Cambridge Univ Press, UK.
- 6. Marris E (2008) Water: More crop per drop. Nature 452: 273–277.
- 7. Milly PCD, Betancourt J, Falkenmark M, Hirsch RM, Kundzewicz ZW, et al. (2008) Climate change - Stationarity is dead: Whither water management? Science 319: 573–574.
- 8. Hanjra MA, Qureshi ME (2010) Global water crisis and future food security in an era of climate change. Food Policy 35: 365–377.
World Water Assesment Programme (2009) The 3rd United Nations World Water Development Report: Water in a Changing World. Paris: UNESCO, and London: Earthscan.
- 10. Rockstrom J, Karlberg L, Wani SP, Barron J, Hatibu N, et al. (2010) Managing water in rainfed agriculture-The need for a paradigm shift. Agricultural Water Management 97: 543–550.
- 11. Vitousek PM, Mooney HA, Lubchenco J, Melillo JM (1997) Human domination of Earth’s ecosystems. Science 277: 494–499.
Smil V (2000) Feeding the world. A challenge for the twenty-first century. MIT Press, Cambridge, Massachusetts, USA.
- 13. Tilman D, Cassman KG, Matson PA, Naylor R, Polasky S (2002) Agricultural sustainability and intensive production practices. Nature 418: 671–677.
Sachs JD (2008) Common Wealth. Economics for a Crowded Planet. The Penguin Press, New York: 386 p.
- 15. Godfray HC, Beddington JR, Crute IR, Haddad L, Lawrence D, et al. (2010) Food security: the challenge of feeding 9 billion people. Science 327: 812–818.
- 16. Zhang H, Wang X, You M, Liu C (1999) Water-yield relations and water-use efficiency of winter wheat in the North China Plain. Irrigation Science 19: 37–45.
- 17. Kang SZ, Zhang L, Liang YL, Hu XT, Cai HJ, et al. (2002) Effects of limited irrigation on yield and water use efficiency of winter wheat in the Loess Plateau of China. Agricultural Water Management 55: 203–216.
- 18. Zhang XY, Pei D, Chen SY, Sun HY, Yang YH (2006) Performance of double-cropped winter wheat-summer maize under minimum irrigation in the North China Plain. Agronomy Journal 98: 1620–1626.
- 19. Siahpoosh MR, Dehghanian E (2012) Water Use Efficiency, Transpiration Efficiency, and Uptake Efficiency of Wheat during Drought. Agronomy Journal 104: 1238–1243.
- 20. Zhang XY, Chen SY, Sun HY, Wang YM, Shao LW (2010) Water use efficiency and associated traits in winter wheat cultivars in the North China Plain. Agricultural Water Management 97: 1117–1125.
- 21. Whisler FD, Acock B, Baker DN, Fye RE, Hodges HF, et al. (1986) Crop Simulation-Models in Agronomic Systems. Advances in Agronomy 40: 141–208.
- 22. de Wit CT, Vankeulen H (1987) Modeling Production of Field Crops and Its Requirements. Geoderma 40: 253–265.
- 23. Steduto P, Hsiao TC, Raes D, Fereres E (2009) AquaCrop-The FAO Crop Model to Simulate Yield Response to Water: I. Concepts and Underlying Principles. Agronomy Journal 101: 426–437.
- 24. Zwart SJ, Bastiaanssen WGM, de Fraiture C, Molden DJ (2010) A global benchmark map of water productivity for rainfed and irrigated wheat. Agricultural Water Management 97: 1617–1627.
Steduto P, Hsiao TC, Fereres E, Raes D (2012) FAO Irrigation and drainage paper 66, Crop yield response to water, Food and Agriculture Organization of the United Nations.
- 26. Raes D, Steduto P, Hsiao TC, Fereres E (2009) AquaCrop-The FAO Crop Model to Simulate Yield Response to Water: II. Main Algorithms and Software Description. Agronomy Journal 101: 438–447.
- 27. Hsiao TC, Heng L, Steduto P, Rojas-Lara B, Raes D, et al. (2009) AquaCrop-The FAO Crop Model to Simulate Yield Response to Water: III. Parameterization and Testing for Maize. Agronomy Journal 101: 448–459.
- 28. Todorovic M, Albrizio R, Zivotic L, Saab MTA, Stockle C, et al. (2009) Assessment of AquaCrop, CropSyst, and WOFOST Models in the Simulation of Sunflower Growth under Different Water Regimes. Agronomy Journal 101: 509–521.
- 29. Farahani HJ, Izzi G, Oweis TY (2009) Parameterization and Evaluation of the AquaCrop Model for Full and Deficit Irrigated Cotton. Agronomy Journal 101: 469–476.
- 30. Dominguez A, Tarjuelo JM, de Juan IA, Lopez-Mata E, Breidy J, et al. (2011) Deficit irrigation under water stress and salinity conditions: The MOPECO-Salt Model. Agricultural Water Management 98: 1451–1461.
- 31. Geerts S, Raes D, Garcia M, Taboada C, Miranda R, et al. (2009) Modeling the potential for closing quinoa yield gaps under varying water availability in the Bolivian Altiplano. Agricultural Water Management. 96: 1652–1658.
- 32. Geerts S, Raes D, Garcia M, Miranda R, Cusicanqui JA, et al. (2009) Simulating Yield Response of Quinoa to Water Availability with AquaCrop. Agronomy Journal 101: 499–508.
- 33. Andarzian B, Bannayan M, Steduto P, Mazraeh H, Barati ME, et al. (2011) Validation and testing of the AquaCrop model under full and deficit irrigated wheat production in Iran. Agricultural Water Management 100: 1–8.
- 34. Salemi H, Soom MAM, Lee TS, Mousavi SF, Ganji A, et al. (2011) Application of AquaCrop model in deficit irrigation management of Winter wheat in arid region. African Journal of Agricultural Research 6: 2204–2215.
- 35. Araya A, Keesstra SD, Stroosnijder L (2010) Simulating yield response to water of Teff (Eragrostis tef) with FAO’s AquaCrop model. Field Crops Research 116: 196–204.
- 36. Abedinpour M, Sarangi A, Rajput TBS, Singh M, Pathak H, et al. (2012) Performance evaluation of AquaCrop model for maize crop in a semi-arid environment. Agricultural Water Management 110: 55–66.
ISS-CAS (2012) China Soil Scientific Database Website. Institute of Soil Science, Chinese Academy of Science (ISS-CAS), China. Available: http://www.soil.csdb.cn/. Accessed 2012 Oct 1.
IUSS (2006) World reference base for soil resources 2006. A framework for international classification, correlation and communication, 2nd edition. World Soil Resources Reports No 103 FAO, Rome.
- 39. Wang XB, Cai DX, Hoogmoed WB, Oenema O, Perdok UD (2007) Developments in conservation tillage in rainfed regions of North China. Soil & Tillage Research 93: 239–250.
- 40. Wang XB, Wu HJ, Dai K, Zhang DC, Feng ZH, et al. (2012) Tillage and crop residue effects on rainfed wheat and maize production in northern China. Field Crops Research 132: 106–116.
Xie WY, Zhou HP, Guan CL, Zhang JJ, Yan XY, et al.. (2011) Spatial Distribution of Soil Active Nutrients in Farmland in Shanxi Province. Journal of Shanxi Agricultural Sciences.
Allen RG, Pereira LS, Raes D, Smith M (1998) Crop evapotranspiration. Guidelines for computing crop water requirements. FAO irrigation and drainage paper no. 56, 300 p.
- 43. Willmott CJ (1982) Some Comments on the Evaluation of Model Performance. Bulletin of the American Meteorological Society 63: 1309–1313.
- 44. Willmott CJ, Matsuura K (2005) Advantages of the mean absolute error (MAE) over the root mean square error (RMSE) in assessing average model performance. Climate Research 30: 79–82.
- 45. Hu K, Li B, Chen D, Zhang Y, Edis R (2008) Simulation of nitrate leaching under irrigated maize on sandy soil in desert oasis in Inner Mongolia, China. Agricultural Water Management 95: 1180–1188.
- 46. Hu KL, Li Y, Chen WP, Chen DL, Wei YP, et al. (2010) Modeling Nitrate Leaching and Optimizing Water and Nitrogen Management under Irrigated Maize in Desert Oases in Northwestern China. Journal of Environmental Quality 39: 667–677.
- 47. Mkhabela MS, Bullock PR (2012) Performance of the FAO AquaCrop model for wheat grain yield and soil moisture simulation in Western Canada. Agricultural Water Management 110: 16–24.
Araya A, Habtu S, Hadgu KM, Kebede A, Dejene T (2010) Test of AquaCrop model in simulating biomass and yield of water deficient and irrigated barley (Hordeum vulgare). Agricultural Water Management. 1838–1846.
Stricevic R, Cosic M, Djurovic N, Pejic B, Maksimovic L (2011) Assessment of the FAO AquaCrop model in the simulation of rainfed and supplementally irrigated maize, sugar beet and sunflower. Agricultural Water Management. 1615–1621.
- 50. Zhang HP, Oweis T (1999) Water-yield relations and optimal irrigation scheduling of wheat in the Mediterranean region. Agricultural Water Management 38: 195–211.
- 51. Huang MB, Gallichand J, Zhong LP (2004) Water-yield relationships and optimal water management for winter wheat in the Loess Plateau of China. Irrigation Science 23: 47–54.
- 52. Kang SZ, Zhang L, Liang YL, Dawes W (2003) Simulation of winter wheat yield and water use efficiency in the Loess Plateau of China using WAVES. Agricultural Systems 78: 355–367.
- 53. Liu CM, Zhang XY, Zhang YQ (2002) Determination of daily evaporation and evapotranspiration of winter wheat and maize by large-scale weighing lysimeter and micro-lysimeter. Agricultural and Forest Meteorology 111: 109–120.
- 54. Kang SZ, Gu BJ, Du TS, Zhang JH (2003) Crop coefficient and ratio of transpiration to evapotranspiration of winter wheat and maize in a semi-humid region. Agricultural Water Management 59: 239–254.
- 55. Wang P, Song XF, Han DM, Zhang YH, Zhang B (2012) Determination of evaporation, transpiration and deep percolation of summer corn and winter wheat after irrigation. Agricultural Water Management 105: 32–37.
- 56. Robert PC (2002) Precision agriculture: a challenge for crop nutrition management. Plant and Soil 247: 143–149.
- 57. Deng XP, Shan L, Zhang HP, Turner NC (2006) Improving agricultural water use efficiency in and and semiarid areas of China. Agricultural Water Management 80: 23–40.
- 58. Chakraborty D, Garg RN, Tomar RK, Singh R, Sharma SK, et al. (2010) Synthetic and organic mulching and nitrogen effect on winter wheat (Triticum aestivum L.) in a semi-arid environment. Agricultural Water Management 97: 738–748.
- 59. Gao YJ, Li Y, Zhang JC, Liu WG, Dang ZP, et al. (2009) Effects of mulch, N fertilizer, and plant density on wheat yield, wheat nitrogen uptake, and residual soil nitrate in a dryland area of China. Nutrient Cycling in Agroecosystems 85: 109–121.
- 60. Chakraboyty D, Nagarajan S, Aggarwal P, Gupta VK, Tomar RK, et al. (2008) Effect of mulching on soil and plant water status, and the growth and yield of wheat (Triticum aestivum L.) in a semi-arid environment. Agricultural Water Management 95: 1323–1334.
- 61. Hou XQ, Li R, Jia ZK, Han QF, Wang W, et al. (2012) Effects of rotational tillage practices on soil properties, winter wheat yields and water-use efficiency in semi-arid areas of north-west China. Field Crops Research 129: 7–13.