J. Meteor. Res.  2017, Vol. 31 Issue (4): 800-808 PDF
http://dx.doi.org/10.1007/s13351-017-6137-z
The Chinese Meteorological Society
0

#### Article Information

Shuai ZHANG, Fulu TAO, Zhao ZHANG . 2017.
Spatial and Temporal Changes in Vapor Pressure Deficit and Their Impacts on Crop Yields in China during 1980–2008. 2017.
J. Meteor. Res., 31(4): 800-808
http://dx.doi.org/10.1007/s13351-017-6137-z

### Article History

in final form November 25, 2016
Spatial and Temporal Changes in Vapor Pressure Deficit and Their Impacts on Crop Yields in China during 1980–2008
Shuai ZHANG1, Fulu TAO1, Zhao ZHANG2
1. Key Laboratory of Land Surface Pattern and Simulation, Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences, Beijing 100101;
2. State Key Laboratory of Earth Surface Processes and Resource Ecology, Beijing Normal University, Beijing 100875
ABSTRACT: Vapor pressure deficit (VPD) is a widely used measure of atmospheric water demand. It is closely related to crop evapotranspiration and consequently has major impacts on crop growth and yields. Most previous studies have focused on the impacts of temperature, precipitation, and solar radiation on crop yields, but the impact of VPD is poorly understood. Here, we investigated the spatial and temporal changes in VPD and their impacts on yields of major crops in China from 1980 to 2008. The results showed that VPD during the growing period of rice, maize, and soybean increased by more than 0.10 kPa (10 yr)–1 in northeastern and southeastern China, although it increased the least during the wheat growing period. Increases in VPD had different impacts on yields for different crops and in different regions. Crop yields generally decreased due to increased VPD, except for wheat in southeastern China. Maize yield was sensitive to VPD in more counties than other crops. Soybean was the most sensitive and rice was the least sensitive to VPD among the major crops. In the past three decades, due to the rising trend in VPD, wheat, maize, and soybean yields declined by more than 10.0% in parts of northeastern China and the North China Plain, while rice yields were little affected. For China as a whole, the trend in VPD during 1980–2008 increased rice yields by 1.32%, but reduced wheat, maize, and soybean yields by 6.02%, 3.19%, and 7.07%, respectively. Maize and soybean in the arid and semi-arid regions in northern China were more sensitive to the increase in VPD. These findings highlight that climate change can affect crop growth and yield through increasing VPD, and water-saving technologies and agronomic management need to be strongly encouraged to adapt to ongoing climate change.
Key words: agriculture     climate change     vapor pressure deficit     evapotranspiration     water stress
1 Introduction

Climate change has had major impacts on crop growth and yields (Peng et al., 2004; Jagadish et al., 2007; Tao et al., 2008; Lobell et al., 2011a). Extensive studies have been conducted to investigate the impacts of climate change on crop yields in the recent years (Nicholls, 1997; Tao et al., 2006, 2008, 2012; Lobell and Field, 2007; Welch et al., 2010; Lobell et al., 2011a, b). Most of these studies have focused on temperature, precipitation, and solar radiation. Water availability is a crucial limiting factor for crop production in a broad range of regions throughout the world (Lobell et al., 2011b), and can lead to large interannual variability in rainfed crop yields in arid and semi-arid regions (e.g., Shuai et al., 2013; Klink et al., 2014). Precipitation is an important climate variable, but insufficient to describe the water condition for crop production because the latter is also affected by evapotranspiration. Water stress can be induced by increased vapor pressure deficit (VPD) and, consequently, increased evapotranspiration. Transpiration in vegetation, which is related to water exchange between the land surface and atmosphere, is determined by both the conductivity of the stomata and the difference in water vapor pressure between the leaf surface and surrounding atmosphere (i.e., VPD; McNaughton and Jarvis, 1991; Kimball et al., 2002). An increase in the VPD between the stomatal cavity and surrounding air will induce a reduction in the rate of transpiration.

VPD is a widely used measure of atmospheric water demand that depends on both air temperature and humidity (Ray et al., 2002; Lobell et al., 2014). It has been documented that the leaf photosynthetic rate and canopy photosynthetic rate decline when atmospheric VPD increases (Quick et al., 1992; Hirasawa and Hsiao, 1999; Fletcher et al., 2007). For crops especially, atmospheric water content is the main factor impacting upon photosynthesis (Huck et al., 1983; Pettigrew et al., 1990), and even crop yield. Due to stomatal closure in crops, a high VPD induced by drought stress may cause a major reduction in crop yield. Tao et al. (2008, 2012) found that an increase in temperature has affected crop yields indirectly in arid and semi-arid regions of China in the past few decades, by increasing the VPD, evapotranspiration, and water stress (i.e., temperature-induced drought). Lobell et al. (2013) found that the predominant effects of extreme heat on maize production in the United States were associated with increased VPD, which contributed to water stress.

In China, many studies have investigated the impact of VPD on crop growth (Wu et al., 2010), leaf stomatal conductance (Zhang et al., 2011), water use efficiency, and so on (Huang et al., 2007; Wang et al., 2008). However, most of these studies were based on field experiment data on the scale of a single station. Due to the limitation of the representativeness of stations, it is difficult to use the results generated by these studies for interpreting the effect of VPD on crops in other regions. Therefore, larger scale studies across China are needed. Although some studies have investigated the temporal and spatial changes of VPD (Liu et al., 2004), few have investigated the impact of VPD on crop yields based on a complete database of crop yields and climate variables (Lobell et al., 2014).

Crucially, the availability of climate and agricultural production data has improved in recent times (Tao et al., 2012). For instance, county-scale statistical data on crop yields and phenology are maintained across China. In this study, we used these improved datasets to investigate the trends in VPD during the crop growing period, the relationships between yields and VPD, and the effects of the trends in VPD on rice, wheat, maize, and soybean yields, at the county level across China in the past three decades. The aims of our study are: 1) to investigate the temporal and spatial change in VPD during the growing period of major crops across China in the past three decades; 2) to estimate the sensitivity of crop yields to VPD changes; and 3) to estimate the impacts of trends in VPD on major crop yields in the past three decades.

2 Materials and methods 2.1 Data

The annual yields of rice, wheat, maize, and soybean were obtained from the Agricultural Yearbook for every county for the period 1980–2008. As in Tao et al. (2012), the time series of yields over mainland China, including 1670, 2021, 2111, and 1943 counties for rice, wheat, maize, and soybean, respectively, were used in this study. Phenological data for rice, wheat, maize, and soybean, recorded at 300 agro-meteorological experimental stations, have been published by the China Meteorological Administration (CMA). The dataset contains major crop phonological events, for example, sowing date, transplantation date, flowering date, maturation date, and so on. These detailed phonological records allow us to study the impact of climate change on the crop growing period on the county scale (Tao et al., 2012).

We qualitatively checked the time series of the phenology data, yield data, and climate data. Only those stations with observations of more than 20 yr were selected for use in this study.

Daily mean temperature and humidity data, observed at 756 national standard stations (NSSs), were obtained from the CMA. Daily VPD data were computed from the mean temperature and mean humidity as follows (Allen et al., 1998):

 ${{e}}_{\rm{o}}= {\rm{0.6108\,\,exp}}\left[\frac{{{\rm{17}}{\rm{.27}} - {{{T}}_{{\rm{mean}}}}}}{{{{{T}}_{{\rm{mean}}}} + {\rm{237}}{\rm{.3}}}}\right],$ (1)
 $\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!{{{e}}_{\rm{a}}} = {{{e}}_{\rm{o}}}\frac{{{\rm{R}}{{\rm{H}}_{{\rm{mean}}}}}}{{{\rm{10}}{{\rm{0}}_{}}}},$ (2)
 $\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!{\rm{VPD}} = {{{e}}_{\rm{o}}} - {{{e}}_{\rm{a}}},$ (3)

where eo is daily mean saturation vapor pressure (kPa), ea is daily mean actual vapor pressure (kPa), RHmean is daily mean relative humidity (%), and Tmean is daily mean temperature (°C).

Daily weather data from the nearest three NSSs were interpolated by using a triangular irregular network to calculate VPD at the county level (Tao et al., 2012). The density of meteorological and agro-meteorological stations is high over the major production regions, meaning the interpolation process did not introduce errors into our analysis. As in Tao et al. (2012), we used agro-meteorological experimental stations to obtain the major phenological dates for the key crops, including 321 stations for rice, 288 stations for wheat, 258 stations for maize, and 80 stations for soybean. We began our analysis by calculating the means of VPD during the crop growing period for each station.

2.2 Method

For each county across its major production regions, we collected information on crop stations, VPD, crop phenology, and yield data for key crops (i.e., rice, wheat, maize, and soybean). Linear regression analysis was applied to extract linear trends from the time series of VPD and climate variables. The partial correlation between crop yields and VPD, temperature and precipitation was separately investigated, to examine the relationships between yields and VPD and further quantify the impact of VPD trends on crop yields during 1980–2008 at the county level.

The complicated influences of long-term variabilities, such as changes in crop management, were avoided by using the first difference for yield (ΔYield) and VPD (ΔVPD) (i.e., year-to-year changes). We initially computed ΔYield and ΔVPD during the crop growing period (from sowing to maturity), as follows:

 ${\rm{\Delta Yiel}}{{\rm{d}}_{{{i, t}}}} = {{{a}}_{{i}}} + {{{b}}_{{i}}} \cdot {\rm{\Delta VP}}{{\rm{D}}_{{{i, t}}}},$ (4)

where i is the county reference number, t is the year, ΔYieldi, t is the first difference of the yield, ΔVPDi, t is the first difference of the VPD, ai is the average year change caused by management in county i, and bi is model coefficient.

A median estimate of model coefficient b estimated by Eq. (4) is the sensitivity of yield change to VPD. The sensitivity of yield change divided by mean observed yields across the county zone during 1980–2008, expressed as a percentage, is as follows:

 $\frac{{{{{b}}_{i}}}}{{{\rm{Yiel}}{{\rm{d}}_{{i}}}}} \times {\rm{100\% }}{\rm{.}}$ (5)

Here, Yieldi is the mean of observed yields across the stations in county i during 1980–2008 (Tao et al., 2012).

To estimate the impacts of trends in VPD on crop yields during 1980–2008, we multiplied the sensitivity of yield change to VPD by the change of VPD.

The estimated influence of VPD trends on crop yields was weight-averaged during the period 1980–2008, based on the planting area of the crop in 2008.

3 Results 3.1 Partial correlation analysis between yield and VPD, temperature and precipitation

Partial correlation measures the degree of association between two random variables by removing the impacts of other variables. We found that crop yields were negatively correlated with VPD in major crop production regions, except southeastern China (Fig. 1). By contrast, crop yields were generally positively correlated with temperature (figure omitted; see supplemental material). The partial correlation between crop yields and precipitation was more complex. It was generally negative for rice, negative in southeastern and northeastern China, but positive in other regions for wheat. There was a patchy pattern for maize and soybean (figure omitted; see supplemental material). Therefore, VPD could have a major impact on crop yields, particularly in the arid and semi-arid regions (Fig. 1), which was different from the impacts of temperature and precipitation.

 Figure 1 Partial correlation coefficient between yield and vapor pressure deficit (VPD) with fixed temperature and precipitation for the growing periods of (a) rice, (b) wheat, (c) maize, and (d) soybean during 1980–2008.
3.2 Trends in VPD during crop growing periods

Mean VPD during crop growing periods of 1980–2008 was much higher in northwestern and northern China than in southeastern China. VPD was low in southeastern China and high in the arid or semi-arid regions of northern China (Fig. 2). During the rice growing period, VPD increased in most regions [Fig. 3; only statistically significant (p < 0.05) results are shown]. In southeastern and northeastern China, VPD increased generally by greater than 0.03 kPa (10 yr)–1, and by greater than 0.10 kPa (10 yr)–1 in some areas (Fig. 3a). During the wheat growing period, VPD increased across China. In particular, VPD increased by greater than 0.06 kPa (10 yr)–1 in northern China and, by contrast, increased generally by less than 0.03 kPa (10 yr)–1 in southern China (Fig. 3b). During the maize growing period, VPD increased by greater than 0.03 kPa (10 yr)–1 in most areas of northeastern China and eastern China and, by contrast, increased by less than 0.03 kPa (10 yr)–1 in southwestern China (Fig. 3c). During the soybean growing period, VPD increased by less than 0.03 kPa (10 yr)–1 in southwestern China and by greater than 0.03 kPa (10 yr)–1 in northeastern, northern, and southeastern China (Fig. 3d). In particular, VPD during the wheat growing period increased less in comparison with other crops. For rice, maize, and soybean, VPD during the crop growing period increased by greater than 0.10 kPa (10 yr)–1 in parts of northeastern and southeastern China.

 Figure 2 Mean vapor pressure deficit (VPD) for the growing periods of (a) rice, (b) wheat, (c) maize, and (d) soybean during 1980–2008.
 Figure 3 Trends in VPD for the growing periods of (a) rice, (b) wheat, (c) maize, and (d) soybean during 1980–2008.
3.3 Sensitivity of crop yields to VPD

The sensitivity of crop yields to VPD change had a spatially explicit pattern (Fig. 4). With VPD during the rice growing period increasing by 10.0%, rice yield decreased generally by around 10.0% in northeastern and southwestern China; by contrast, the yield increased by about 5.0% in southeastern China (Fig. 4a). With VPD during the wheat growing period increasing by 10.0%, wheat yield decreased by more than 10.0% in northeastern and southwestern China; in contrast, yield increased by about 5.0% in southeastern China (Fig. 4b). With VPD during the maize growing period increasing by 10.0%, maize yield decreased by about 10.0% in most of the major maize cultivation areas from northeastern China to southwestern China (Fig. 4c). Yield decreased most in the arid and semi-arid regions of northern China. With VPD during the soybean growing period increasing by 10.0%, soybean yield decreased by more than 10.0% in parts of northeastern and northern China (Fig. 4d). We found that these major crops were stressed by high VPD in the arid and semi-arid regions of northern, northeastern, and northwestern China. By contrast, crop yields increased along with increased VPD in southeastern China. This is because the climatic limiting factors for crop production in southeastern China are low solar radiation and continuous precipitation (high air humidity) (Tao et al., 2012). Among the four major crops, soybean was the most sensitive to VPD, and rice was the least sensitive. For maize, the number of sensitive counties was the most (Table 1).

 Figure 4 Sensitivity of crop yields to vapor pressure deficit (VPD) for the growing periods of (a) rice, (b) wheat, (c) maize, and (d) soybean during 1980–2008.
Table 1 Impact of vapor pressure deficit (VPD) on crop yields
 Number of counties investigated Number of counties (significant) Sensitivity of crop yields to VPD (%) Change in crop yields due to trends in VPD (%) Rice 1737 227 –0.33 3.39 Wheat 2116 445 –5.29 –5.56 Maize 2182 470 –7.35 –10.58 Soybean 2086 340 –11.13 –16.68

We also found that the sensitivity of crop yields to VPD increased in most areas during 1994–2008, in comparison with the period 1980–93, for all the major crops (Fig. 5), although it also decreased in some areas.

 Figure 5 Change in crop yield sensitivity to vapor pressure deficit (VPD) for the growing periods of (a) rice, (b) wheat, (c) maize, and (d) soybean during 1980–93 and 1994–2008.
3.4 Impacts of trends in VPD on crop yields during 1980–2008

The impacts of trends in VPD on crop yields were based on the sensitivity of crop yields to VPD changes and trends in VPD. In the past three decades, rice yield decreased by around 10.0% due to increased VPD in parts of southwestern China (Fig. 6a). Wheat yield decreased by more than 10.0% in most areas of northeastern and southwestern China but, by contrast, increased in southeastern China (Fig. 6b). Maize yield declined along with increased VPD in most counties (Table 1), decreased by more than 10.0% in northern China, and decreased by around 10.0% in southwestern China (Fig. 6c). Soybean yield declined most along with increased VPD, and decreased by more than 10.0% in northeastern and northern China (Fig. 6d). In summary, crop yields generally decreased along with increased VPD, except rice and wheat in southeastern China. Maize and soybean yields showed more changes than rice and wheat, because maize and soybean were less irrigated and cultivated in the areas where VPD was relatively high (Table 1). Rice yield showed the least change among all the major crops, because rice was usually irrigated and cultivated in humid areas, where VPD was relatively low (Table 1).

 Figure 6 Yield changes of (a) rice, (b) wheat, (c) maize, and (d) soybean in relation to trends in VPD during 1980–2008.

Taking the country as a whole, the VPD trend in the past three decades had measureable effects on crop yields. Sowing area-weighted averages indicated that the VPD trend during the growing period increased rice yield by 1.32%, but reduced wheat yield by 6.02%, maize yield by 3.19%, and soybean yield by 7.07% (Fig. 7). The results indicated that, apart from rice, crop yields declined in relation to the impact of the trend in VPD during 1980–2008.

 Figure 7 Estimated net impact of the trend in VPD during 1980–2008 on crop yields for China as a whole. Values are area-weighted planting averages. The extent of the lines illustrates the standard deviation.
4 Discussion

VPD is an integrated variable for atmospheric water demand that depends on both air temperature and humidity. According to the Clausius–Clapeyron equation, VPD will increase with ongoing climate warming, which suggests that atmospheric water demand or drought will also increase under such a scenario. Therefore, it is necessary to investigate the relationship between increased VPD and crop yields on the regional or global scale.

Rice was the least sensitive to VPD among the major crops because it is more sensitive to changes in precipitation and radiation (Tao et al., 2012). For wheat cultivated in northeastern and southwestern China, VPD was negatively related to yield; whereas, for wheat cultivated in southeastern China, VPD was positivity related to yield. Crops that were positivity related to VPD were usually located in areas where the mean VPD during the crop growing period was relatively lower (Figs. 2a, b). Solar radiation is known to be a major factor affecting crop yield under well-irrigated conditions for some regions in China (Tao et al., 2012). For rice and wheat cultivated in southeastern China, crop yields increased along with increased VPD. Generally, however, yields decreased with increased VPD. In southeastern China, VPD was not the key factor impacting upon crop yield, owing to the abundant precipitation in this region, which may cause the low solar radiation. Therefore, low radiation may be the key limitation for crop yields in southeastern China. For maize and soybean cultivated in northeastern and northern China, VPD was negatively related to yield, particularly in arid and semi-arid regions, such as in northern China. The results are supported by previous studies; for example, Tao et al. (2012) showed that maize, wheat, and soybean were especially sensitive to temperature-increase-induced droughts in Northeast, Northwest, and North China, especially in arid and semi-arid areas. Lobell et al. (2013) indicated that increases in VPD contribute to water stress and affect crop growth and yield in two ways. First, the crops increase their demand for soil water to maintain carbon assimilation at a given rate; and second, the crops reduce the supply of soil water through elevated transpiration rates. Therefore, maize yield declined with increased VPD (Lobell et al., 2013). Shuai et al. (2013) found that VPD affected yield variability through its effects on water stress. Crops that were negatively related to VPD were usually located in areas where the mean VPD during the crop growing period was higher. In parts of the maize cultivation area, the mean VPD during the maize growing period exceeded 0.8 kPa (Fig. 2c). Maize and soybean in China were cultivated under stronger drought conditions than for other crops (Figs. 2c, d). We found that maize had the most counties sensitive to VPD change, and soybean was the most sensitive to VPD change, in the past three decades among the four major crops. The results are consistent with several previous studies (Tao et al., 2008; Liu et al., 2013).

We found that the sensitivity of crop yields to VPD during 1994–2008 generally increased, although it also decreased in some areas, in comparison with the period 1980–93. Lobell et al. (2014) showed that VPD was expected to increase in the US Midwest, and the sensitivity of maize yield to drought had increased in the past several decades. The changes in the sensitivity of crop yields to VPD can be attributed to changes in VPD itself, cultivars and agronomic management practices; for example, increasing crop sowing densities could increase crop sensitivity to VPD (Lobell et al., 2014). Adopting drought-tolerant cultivars and technologies, as well as increasing irrigation areas, might decrease the sensitivity of crop yields to VPD.

Since VPD is dependent on both air temperature and humidity, it is reasonable that the sensitivity of crop yield to VPD may be, to some extent, similar to the sensitivity of crop yield to air temperature or precipitation. Nevertheless, there are differences between them. For example, the partial correlation between crop yield and VPD, temperature and precipitation was different. The difference between the sensitivity of crop yield to VPD and to precipitation was about 10.0% (figure omitted; see online supplemental material). The sensitivity of crop yield to temperature could not fully explain the sensitivity of crop yield to VPD (Fig. 3), particularly in regions where humidity matters, such as southeastern China, and the arid and semi-arid regions of northern, northeastern, and northwestern China.

5 Conclusions

Based on climate data and crop yield data, we investigated the trends in VPD during crop growing periods, the sensitivity of crop yields to change in VPD, and the effects of the trends on rice, wheat, maize, and soybean yield, at the county level across China in the past three decades. We found that regional trends of VPD during the period 1980–2008 had impacts on crop yields, and that the impact had a spatial pattern. VPD during the wheat growing period increased the least. Increases in VPD showed different sensitivity to yields for different crops and in different regions. Crop yield generally decreased along with increased VPD, except for wheat in the middle and lower reaches of the Yangtze River. Soybean was the most sensitive and rice was the least sensitive to VPD among the major crops. Maize yield was sensitive to VPD in more counties than other crops. In the past three decades, in relation to the trends in VPD, wheat, maize, and soybean yields declined by more than 10.0% in parts of northeastern China and the North China Plain, while rice yield showed little change. For the country as a whole, the planting area-weighted averages indicated that VPD trends during growing periods increased rice yield by 1.32%, but reduced wheat yield by 6.02%, maize yield by 3.19%, and soybean yield by 7.07%. Maize and soybean in the arid and semi-arid regions of northern China were more sensitive to increased VPD. The results indicate that climate change can affect crop growth and production by increasing VPD. Water-saving technologies and water-saving agriculture should be strongly encouraged in order to adapt to ongoing climate change.

References
 Allen, R. G., L. S. Pereira, D. Raes, et al., 1998: Crop evapotranspiration—Guidelines for computing crop water requirements. FAO Irrigation and drainage paper 56. Food and Agricultural Organization, Rome, 300 pp. Fletcher L., A. R. Sinclair, T. H. Allen Jr., 2007: Transpiration responses to vapor pressure deficit in well watered " slow-wilting” and commercial soybean. Environ. Exp. Bot., 61, 145–151. DOI:10.1016/j.envexpbot.2007.05.004 Hirasawa C. Hsiao, 1999: Some characteristics of reduced leaf photosynthesis at midday in maize growing in the field. Field Crop. Res., 62, 53–62. DOI:10.1016/S0378-4290(99)00005-2 Huang H., Yu G. R., Sun X. M., et al., 2007: The environmental responses and simulation of canopy conductance in a winter wheat field of North China Plain. Acta Ecol. Sinica, 27, 5209–5221. Huck M. G., Ishihara K., Peterson C. M., et al., 1983: Soybean adaptation to water stress at selected stages of growth. Plant Physiol, 73, 422–427. DOI:10.1104/pp.73.2.422 Jagadish Craufurd, S. Wheeler, 2007: High temperature stress and spikelet fertility in rice (Oryza sativa L.). J. Exp. Bot., 58, 1627–1635. DOI:10.1093/jxb/erm003 Kimball A., B. Kobayashi, K. Bindi, 2002: Responses of agricultural crops to free-air CO2 enrichment . Adv. Agron., 77, 293–368. DOI:10.1016/S0065-2113(02)77017-X Klink K., Wiersma J. J., Crawford C. J., et al., 2014: Impacts of temperature and precipitation variability in the Northern Plains of the United States and Canada on the productivity of spring barley and oat. Int. J. Climatol., 34, 2805–2818. DOI:10.1002/joc.2014.34.issue-8 Liu B. H., Xu M., Henderson M., et al., 2004: A spatial analysis of pan evaporation trends in China, 1955–2000. J. Geophys. Res.: Atmos., 109, D15102. DOI:10.1029/2004JD004511 Liu X. J., Zhang J. Q., Ma D. L., et al., 2013: Dynamic risk assessment of drought disaster for maize based on integrating multi-sources data in the region of the northwest of Liaoning Province, China. Nat. Hazards, 65, 1393–1409. DOI:10.1007/s11069-012-0416-x Lobell B., D. B. Field, 2007: Global scale climate–crop yield relationships and the impacts of recent warming. Environ. Res. Lett., 2, 014002. DOI:10.1088/1748-9326/2/1/014002 Lobell D. B., Hammer G. L., McLean G., et al., 2013: The critical role of extreme heat for maize production in the United States. Nat. Climate Change, 3, 497–501. DOI:10.1038/nclimate1832 Lobell B., D. Bänziger, M. Magorokosho, et al., 2011a: Nonlinear heat effects on African maize as evidenced by historical yield trials. Nat. Climate Change, 1, 42–45. DOI:10.1038/nclimate1043 Lobell B., D. Schlenker, W. Costa-Roberts, 2011b: Climate trends and global crop production since 1980. Science, 333, 616–620. DOI:10.1126/science.1204531 Lobell D. B., Roberts M. J., Schlenker W., et al., 2014: Greater sensitivity to drought accompanies maize yield increase in the US Midwest. Science, 344, 516–519. DOI:10.1126/science.1251423 McNaughton G., K. G. Jarvis, 1991: Effects of spatial scale on stomatal control of transpiration. Agric. Forest Meteor., 54, 279–302. DOI:10.1016/0168-1923(91)90010-N Nicholls N., 1997: Increased Australian wheat yield due to recent climate trends. Nature, 387, 484–485. DOI:10.1038/387484a0 Peng S. B., Huang J. L., Sheehy J. E., et al., 2004: Rice yields decline with higher night temperature from global warming. P. Natl. Acad. Sci. U. S. A., 101, 9971–9975. DOI:10.1073/pnas.0403720101 Pettigrew W., Hesketh J., Peters D., et al., 1990: A vapor pressure deficit effect on crop canopy photosynthesis. Photosynth. Res., 24, 27–34. DOI:10.1007/BF00032641 Quick W., Chaves M., Wendler R., et al., 1992: The effect of water stress on photosynthetic carbon metabolism in four species grown under field conditions. Plant Cell Environ., 15, 25–35. DOI:10.1111/pce.1992.15.issue-1 Ray J. D., Gesch R. W., Sinclair T. R., et al., 2002: The effect of vapor pressure deficit on maize transpiration response to a drying soil. Plant Soil, 239, 113–121. DOI:10.1023/A:1014947422468 Shuai J. B., Zhang Z., Liu X. F., et al., 2013: Increasing concentrations of aerosols offset the benefits of climate warming on rice yields during 1981–2008 in Jiangsu Province, China. Reg. Environ. Change, 13, 287–297. DOI:10.1007/s10113-012-0332-3 Tao F. L., Yokozawa M., Xu Y. L., et al., 2006: Climate changes and trends in phenology and yields of field crops in China, 1981–2000. Agric. Forest Meteor., 138, 82–92. DOI:10.1016/j.agrformet.2006.03.014 Tao F. L., Yokozawa M., Liu J. Y., et al., 2008: Climate–crop yield relationships at provincial scales in China and the impacts of recent climate trends. Climate Res., 38, 83–94. DOI:10.3354/cr00771 Tao L., F. Zhang, S. Zhang, 2012: Spatiotemporal changes of wheat phenology in China under the effects of temperature, day length and cultivar thermal characteristics. Eur. J. Agron., 43, 201–212. DOI:10.1016/j.eja.2012.07.005 Wang X. Y., Li L. G., Xie Y. B., et al., 2008: Stomatal conductance and its scaling in vegetation–atmosphere interaction: A review. Chinese J. Ecol., 27, 454–459. Welch J. R., Vincent J. R., Auffhammer M., et al., 2010: Rice yields in tropical/subtropical Asia exhibit large but opposing sensitivities to minimum and maximum temperatures. P. Natl. Acad. Sci. U. S. A., 107, 14562–14567. DOI:10.1073/pnas.1001222107 Wu R. J., Zheng Y. F., Zhao Z., et al., 2010: Assessment of loss of accumulated dry matter in winter wheat based on stomatal conductance and ozone uptake model. Acta Ecol. Sinica, 30, 2799–2808. Zhang B. Z., Liu Y., Xu D., et al., 2011: Estimation of summer corn canopy conductance by scaling up leaf stomatal conductance. Trans. CSAE, 27, 80–86.