1. Introduction

Climate change is considered to be one of the mainenvironmental problems of the 21st century(IPCC, 2007). Agricultural production is greatly affected byclimate: changes in greenhouse gas concentration, radiation, and temperature patterns may have large consequences for potential and rainfed yields(Torriani et al., 2007; Supit et al., 2012). They may also generate economic effects on agricultural prices, production, dem and , trade, regional comparative advantage, and producer and consumer welfare(Li et al., 2011). The ultimate goal of agricultural production is to obtain acrop yield, the generation of which is determined byboth climatic conditions and advancements and uptakes of scientific/technological agricultural applications. In fact, spatiotemporal climatic variations often spur advancements in agricultural technology, becoming an integral part of agricultural development(Turner and Brush, 1987).

Maize(Zea mays L.)is a critical crop for sustaining human life in terms of its role as a major grain commodity, feed commodity, and significant bioethanol energy source(Li et al., 2011). A number of studies havedocumented that climate change in China has affectedmaize phenology(Li et al., 2014), yield(Tao et al., 2006; Li et al., 2011), productivity(Zhao et al., 2011;Yuan et al., 2012), and agricultural climatic resourceutilization(Guo et al., 2013). For example, Xiong etal.(2012)reported that warming trends have negatively affected maize production in China. Nevertheless, despite increasing evidence suggesting that ongoing climate change has measurable impacts on cropdevelopment and productivity, the mechanisms, size and extent of these impacts remain inconclusive(Tao et al., 2008, 2012a; Liu et al., 2010; Olesen et al., 2011;Yu et al., 2012).

Science and technology are at the core of agricultural development(Vanloqueren and Baret, 2009).Technological advancements in agriculture include enhanced agricultural research capacity, application ofnew scientific and technological achievements, promotion of agricultural technology systems, changing management methods, and improving the quality of workers(Zhang, 2010). Advances in technology have beenresponsible for significant increases in agricultural production in China(Xiong et al., 2012), raising the possibility that they have offset the negative effects ofclimatic change on yield(Zhang et al., 2005; Liu et al., 2010; Tao et al., 2012b; Wang et al., 2012). Tao and Zhang(2010)reported that maize yield in Chinahas increased in the past several decades under thecombined effects of climate change and technologicaladvancements. Therefore, the complex impacts of climate change on crop growth and yield are further confounded by changes in technological advancement(Liu et al., 2010; Tao et al., 2012a). However, the roleof technology in adaptation to climate change is evenmore crucial in developing countries where food security remains a struggle for a significant portion of thepopulation, and impending climate change is expectedto make the situation even worse(Chhetri et al., 2012).Thus, how to disentangle the relative contributionsof climate change and technological advancement tochanges in crop yield, i.e., to investigate the attribution of crop yield change, is a key scientific issue forthe objective ev aluation of sustainable development ofagriculture under climate change.

In China, since the 1980s, the contribution rateof technological advancement has been explored basedon the Solow residual value method. Zhu and Liu(1997)used this method to create the C-D production function model, and then estimated the contribution rate of technical advancement from the periodof the Chinese government's first five-year plan to theninth five-year plan. Recently, however, there has beenincreasing interest in the use of different methods todistinguish between climate change and technologicaladvancement(Liu et al., 2010; Tao et al., 2012b; Wang et al., 2012; Yu et al., 2012; Zhang et al., 2013; Xiao and Tao, 2014). For example, Xiao and Tao(2014)used detailed data from a long-term field experimentfrom 1980 to 2009 at four stations on the North ChinaPlain(NCP), together with a crop simulation model, to elucidate the relative contributions of cultiv ar renewal, fertilization management, and climate changeto winter wheat yield. They found that during thisperiod, the contribution of cultiv ar renewal to yieldincrease was 12.2%-22.6%, while fertilization management was 2.1%-3.6%, and climate change was from-3.0% to 3.0%.

Previous studies are important for underst and ing various aspects of agricultural production underclimate change, and also highlight the complexity ofcrop-climate-technology relationships in China. However, few studies have thus far explored the attributionof maize yield increase in China to climate change and technological advancement over the past 30 years onthe national scale. Thus, to improve underst and ing ofclimate impacts, among others, there is a need to sep-arate the relative contributions of climate change and technological advancement to changes in maize yields, and investigate the attribution of maize yield increasein China.

The objectives of the present study are to: 1)quantitatively ev aluate the relative impacts of differentclimate variables on climatic potential productivity ofmaize in the past five decades in China; 2)separate therelative contributions of climate change and technological advancement to maize yield over the past threedecades in China; 3)quantify the attribution of maizeyield increase in China to climate change and technological advancement between 1980 and 2010; and 4)provide a scientific basis for sustainable maize production in China to adapt to future climate change.2. Materials and methods2.1 Data

Daily climate variables gathered from 553 meteorological stations in China for the period 1961-2010 are obtained from the China National Meteorological Information Center. The variables are mean, maximum, and minimum air temperature; precipitation; relative humidity; bright sunshine hours at 2-m height; and wind speed at 10-m height. Thesemeasured daily data are converted to st and ard input parameters for the Agro-Ecological Zones(AEZ)model(see the following section). Field observ ationsof maize have been widely recorded in China sincearound 1980. The observ ational data used in thepresent study, from 1981 to 2010, covering phenology, growth, and development characteristics of maize, areobtained from the Statistics Data of Agricultur e for 60Ye ars in China(Ministry of Agriculture of the People'sRepublic of China, 2009) and 653 agricultural meteorological stations across China. Agro-technicians havedocumented the yearly dates of major events, including the sowing, seedling, heading, and maturity stagesfor each maize growth cycle at each station.2.2 Methods2.2.1 Calculation of climatic potential productivity ofmaize

Estimation of the climatic potential productivity of a crop is fundermental for research on comprehensive grain production capacity, which can provide important theoretical guidelines for the distribution of agricultural production, adjustment to agricultural structure, and reasonable use of climate resources(Yuan et al., 2012).

The methodology of AEZ is adopted in this study.The Food and Agriculture Organization of the UnitedNations, in collaboration with the International Institute for Applied Systems Analysis, developed the AEZmodel. This model has been in use since 1978 fordetermining the agricultural production potential and carrying capacity of the world's l and areas. An agroecological zone, as originally defined, is comprised ofall parts of grid cells that have uniform soil and climatecharacteristics on a geo-referenced map. The suitability of each of these zones for rainfed production ofvarious crops under different input and managementscenarios is then ev aluated. The yield potential of thecrops most suited to each zone where rainfed crop production is possible determines the overall agriculturalproduction potential of that zone. Crops ev aluatedoriginally included food, fiber and fodder crops, as wellas pasture grasses.

In this study, based on the AEZ methodology, theclimatic potential productivity of maize is estimatedthrough a three-level step-by-step correction process:photosynthetic potential productivity, photosyntheticthermal potential productivity, and climatic potential productivity. By computing the climatic potentialproductivity and limiting factors at different levels, we are able to analyze the impact of climate changeon local agricultural production, and identify factorsaffecting yield formation. Details of the calculationprocedure can be found in Zhao et al.(2011).2.2.1.1 Photosynthetic potential productivity

The photosynthetic potential productivity refersto the yield potential of a specific crop decided by solarradiation when water, fertilizer, heat, and other factors are optimized. T otal dry matter can be expressedby(Zhao et al., 2011)

where Y o is the daily total dry matter amount(kghm^-2day^-1); yo is the amount of dry matter production on cloudy days(kg hm^-2day^-1); F is the cloudcover rate; yc is the amount of dry matter productionon sunny days(kg hm^-2day^-1); Rse is the maximumeffective shortwave radiation on sunny days(J cm^-2day^-1); and Rg is measured incident shortwave radiation(J cm^-2day^-1).2.2.1.2 Photosynthetic thermal potential productivity

The photosynthetic thermal potential productivity refers to the yield potential of a specific crop decided by light and temperature when water and fertilizer are optimized, reflecting the highest yield in irrigated farml and at the highest input level. The photosynthetic thermal potential productivity can be obtained through the corrections of maximum dry matter productivity, leaf area index, net dry matter, and harvesting dry matter, based on the photosyntheticpotential productivity. The methods for doing so areas follows(Zhao et al., 2011).

a)Correction of maximum dry matter productivityThe dry matter productivity depends on the temperature during production and the crop variety. According to the findings of Zhao et al.(2011), whenthe daytime temperature is 15, 20, 25, and 30‰°C, thedry matter productivity(ym)of maize is 5.25, 47.25, 47.25, and 68.25 kg hm^-2h^-1, respectively.

b)Correction of dry matter through leaf area indexThis is calculated as follows(Zhao et al., 2011):

where L_c is the correction coeffcient of leaf area, and LAI is leaf area index.

c)Correction of net dry matter

The growth process of crops, as with all plants, involves both photosynthesis and respiration. The difference in the rate of photosynthesis and respirationcan be used to determine crop growth and the accumulation of substances(Zhao et al., 2011):

where N_c is the correction value of net dry matter and T is the average temperature during the growth periodof maize.

d)Correction of harvesting dry matter(grain, sugar, oil, etc.)

This is calculated as follows(Zhao et al., 2011):

where Ymp is the photosynthetic thermal potentialproductivity(kg hm^-2); H_c is the correction value ofharvesting dry matter; G is the day number of thecrop growth period(day); ym is the dry matter productivity; F is the cloud cover rate.2.2.1.3 Climatic potential productivity

The climatic potential productivity refers to theyield potential of a specific crop decided by light, temperature and water under limited precipitation. Theequations for its calculation are as follows(Zhao et al., 2011):

where T_m is the maize water requirement(mm), K isthe coeffcient of maize water requirement, and P ispotential ev apotranspiration(mm);where s_a is the available soil moisture before sowing(mm), k is the empirical correction index, i and j arethe days before sowing which are determined according to the actual condition, and Pa is precipitation(mm);where S_a2 is the effective soil moisture that day(mm), S_a1 is the effective soil moisture the day before(mm), and T_a is the actual water consumption of maize(mm);where Ym is the percentage of yield reduction duringdifferent growth stages of maize, and Ky is the coeff-cient of maize yield responsewhere I_n is the maize yield index during the nthgrowth stage, i_n-1 and i_n are the percentages of maizeyield reduction during the(n-1)th and nth growthstages respectively; and where Yp is the climatic potential productivity(kghm^-2).2.2.2 Assessment of the imp acts of different climate variables on climatic potential productivity of maize

As mentioned above, the photosynthetic potential productivity reflects the effects of radiation(sunlight)conditions on the production and yield generation of crops. The photosynthetic thermal potentialproductivity reflects the effects of radiation and thermal conditions on the production and yield generationof crops. The climatic potential productivity reflectsthe comprehensive effects of different climatic factorson the production and yield generation of crops.

Therefore, the effects of radiation change, thermalchange, precipitation change, and climatic resourcechange on the climatic potential productivity are calculated as follows:

where Y_R is the change of climatic potential productivity under changed radiation(kg hm^-2 yr^-1), α_R is theslope of change in photosynthetic potential productivity over time, Y₁ is the average photosynthetic potential productivity during the period 1961-2010(kghm^-2 yr^-1), and Y₃ is the average climatic potentialproductivity during the period 1961-2010(kg hm^-2 yr^-1);where Y_T is the change of climatic potential productivity under changed thermal conditions(kg hm^-2 yr^-1); α_T is the slope of change in photosynthetic thermal potential productivity over time, and Y₂ is the averagephotosynthetic thermal potential productivity duringthe period 1961-2010(kg hm^-2 yr^-1);where Y_P is the change of climatic potential productivity under changed precipitation(kg hm^-2 yr^-1) and α_c is the slope of change in climatic potential productivity over time;where Y_c is the change of climatic potential productivity under changed climatic resources(kg hm^-2 yr^-1).2.2.3 Evaluating the attribution of maize yield change to climate change and te chnolo gic al advancement

Changes in crop yields are decided by actual climatic conditions and technological advancements(soilimprovement, new varieties, adjustments to agricultural practices, etc.). Because soil is relatively stable, if the different agro-technical measures remain unchanged during the process of agricultural production, the only factor leading to changes in crop yield is climatic fluctuation. Generally speaking, the climaticpotential productivity of a crop refers to the highestyield per unit area when the effects of crop variety, soilfertility, farming techniques, and other measures havebeen fully exerted under local light, heat, water, and other natural climatic factors. This shows that thebasic production levels for actual yield and climaticpotential productivity do not change over time. Therefore, the changes in actual yield and climatic potential productivity are relev ant only to climatic factors, and their trends are inevitably approximate linear relationships. That is to say, the change trend of actual yield should theoretically be consistent with that ofclimatic potential productivity if the contribution ofscientific and technological advancement is not considered. Thus, analysis of the differences between climatic potential productivity change and actual maizeyield change can contribute to identifying the contributions of scientific and technological advancement.However, the difference in values between climatic potential productivity change and actual maize yield islarge. Therefore, in order to eliminate this difference, the relative value was used in this study.

In addition, the decadal variability of actualmaize yield is large. The study period from 1981 to1990 is the base period used for the analysis. Howclimate change and technological advancement influenced maize yield during this period is ev aluated byusing the equations

where YQ is the change of maize yield under climatechange(kg hm^-2 yr^-1), B1 is the change in climaticpotential productivity over time(kg hm^-2 yr^-1), B2 isthe average climatic potential productivity during theperiod 1981-1990(kg hm^-2 yr^-1), A1 is the changein actual maize yield over time(kg hm^-2 yr^-1), A2is the actual maize yield during the period 1981-1990(kg hm^-2 yr^-1), and YK is the change of maize yieldunder technological advancement(kg hm^-2 yr^-1).3. Results

According to the existing data on sown area and maize yield from 1981 to 2010 in each provincein China, the 14 provinces(autonomous regions)ofJiangsu, Anhui, Sh and ong, Henan, Hebei, Shanxi, Shaanxi, Hubei, Sichuan, Gansu, Liaoning, Jilin, Heilongjiang, and Inner Mongolia play an important rolein nationwide maize production, accounting for 85%of the total sown area and maize yield in China.Therefore, the maize yield in the above-mentioned 14provinces(autonomous regions)is mainly analyzed inthis study.3.1 Imp acts of different climate variables onclimatic potential productivity of maize

From 1961 to 2010, the climatic potential productivities of maize all decreased as a result of the reduction of radiation in the 14 provinces, with a magnitudeof less than -20 kg hm^-2 yr^-1(Fig. 1). There were significant regional differences in the changes of climaticpotential productivity of maize caused by the declineof radiation resource change among each province from1961 to 2010. The climatic potential productivitiesof maize in Gansu, Heilongjiang, and Inner Mongolia decreased slightly by 1.4, 4.1, and 2.2 kg hm^-2 yr^-1, respectively. In Jiangsu, Anhui, Sh and ong, and Hubei, the climatic potential productivities of maizedecreased significantly by 18.8, 18.6, 17.6, and 18.9kg hm^-2 yr^-2, respectively. This shows that radiation reduction has severe impacts on climatic potential productivities of maize in these regions. However, the largest reduction of climatic potential productivitywas found in Henan Province, with a change of -19.5kg hm^-2 yr^-2.

From 1961 to 2010, the increase of thermal resources had negative effects on climatic potential productivities of maize in the 12 provinces with an average decrease of 12.7 kg hm^-2 yr^-1, except for Shaanxi and Gansu(Fig. 2). In Shanxi, Hubei, Liaoning, and Jilin, the climatic potential productivities of maize decreased significantly by 15.2, 16.3, 17.6, and 16.7 kghm^-2 yr^-1, respectively. However, the biggest reduction of climatic potential productivity was found inAnhui Province, with a change of -18.1 kg hm^-2 yr^-1.This further illustrates that the increased thermal resources have been unfavorable for the growth and yieldgeneration of maize over the past 50 years. The mainreason is that the summer temperature has exceededthe threshold of suitability for current maize varieties, limiting the normal growth and development ofmaize.

However, the changes in precipitation resourceshad slight positive effects on climatic potential productivities of maize in the 10 provinces, with an averageincrease of 5.9 kg hm^-2 yr^-1over the past 50 years, except in Sh and ong, Shanxi, Gansu, and Heilongjiang(Fig. 3). The biggest reduction of climatic potential productivity was found in Shanxi, with a change valueof -15.9 kg hm^-2 yr^-1. In Anhui and Hubei, the climatic potential productivities increased obviously by13.3 and 13.0 kg hm^-2 yr^-1.

From 1961 to 2010, the combined effects of radiation, thermal, and precipitation resources on the climatic potential productivity of maize in every provincewere negative, except for Gansu where there was anincrease of 0.5 kg hm^-2 yr^-1(Fig. 4). The biggestreduction of climatic potential productivity was foundin Shanxi with a change of -46.7 kg hm^-2 yr^-1, followed by Sh and ong with a decrease of 31.3 kg hm^-2 yr^-1. In Jiangsu, Anhui, Henan, Hubei, Liaoning, and Jilin, the climatic potential productivities of maize decreased distinctly by 20.1, 23.4, 20.4, 22.2, 24.1, and 26.6 kg hm^-2 yr^-1, respectively.

It can be seen from the above analysis that theclimatic potential productivity of maize is more sensitive to changes in radiation and thermal resourcescompared to precipitation resources. Decreased radiation and increased temperature were the main factors leading to the decrease of climatic potential productivity in the study period. This is consistent withthe result reported by Shuai et al.(2014), who foundthat decreasing sunshine duration and increasing maximum temperature caused a reduction of 0.16 t ha^-1inrice yields(approximately 1.8%)in Jiangsu Province.Therefore, it is important to strengthen the cultiv ationof maize varieties with high photosynthetic effciency and high temperature resistance in order to adapt tofuture climate change.3.2 Attribution of maize yield change in Chinato climate change and te chnolo gic al advancement

From 1981 to 2010, the actual maize yields in the14 main producing provinces in China increased significantly, which was opposite to the decrease in climatic potential productivity(Fig. 5). The suggestion, therefore, is that climate change has not been conducive tothe growth and yield formation of maize in the past50 years, but technological advancement has played asignificant role in promoting increases in actual yield.Moreover, the contributions of scientific and technological advancement to maize yield were greater thanthose of actual yield increase. Technological adv ancement contributed to increases in maize yield by 99.6%-141.6%, while climate change contributed from -41.4%to 0.4%. In particular, the actual maize yields in Sh and ong, Henan, Jilin, and Inner Mongolia increased by98.4, 90.4, 98.7, and 121.5 kg hm^-2 yr^-1, respectively.Corresponding maize yields affected by scientific and technological advancement increased by 113.7, 97.9, 111.5, and 124.8 kg hm^-2 yr^-1, respectively. On thecontrary, maize yields reduced markedly with an average of -9.0 kg hm^-2 yr^-1under climate change. InSh and ong, Shanxi, and Jilin, maize yields decreaseddistinctly by 15.3, 23.4, and 12.8 kg hm^-2 yr^-1, respectively. In Jiangsu, Anhui, Henan, Hebei, Shaanxi, Hubei, Sichuan, Liaoning, Heilongjiang, and InnerMongolia, maize yields decreased slightly by 9.4, 7.8, 7.5, 6.6, 1.5, 6.7, 6.8, 10.9, 5.5, and 3.3 kg hm^-2 yr^-1, respectively. In other words, the increases in maizeyield in China were all affected by scientific and technological advancement in the past 30 years. T echnological advancement offset the negative effects of climatic change on maize yield. If there had been noadverse impacts of climate change, the yields of maizein China would have been higher than the period from1981 to 2010, and the magnitude of increase greater.Therefore, it is crucial to invest more in technology and constantly adapt to new climatic conditions in orderto protect agricultural sustainable development underclimate change.

4. Discussion4.1 Relative effe cts of different climate variables on the climatic potential productivityof maize

As a part of agricultural climate resources, the climatic potential productivity depends on the amountof light, temperature, water, and the degree of theirmutual coordination(Zhao et al., 2011). Under globalwarming, resources such as light, temperature, and water have undergone significant changes, causing acorresponding change in climatic potential productivity and ultimately posing a threat to national food production security.

By the end of this century, growing season temperatures may exceed the most extreme seasonal temperatures recorded in the past century(IPCC, 2007).Sustained temperature increases over the growing season could change the duration(from sowing to maturity)of the crop(Roberts and Summerfield, 1987), which is an important cause of yield reduction underclimate change(Tao et al., 2010). Short episodes ofhigh temperature at critical stages of crop development can cause sterility and consequently yield reduction, independent of any substantial changes in meantemperature(Wheeler et al., 2000; McKeown et al., 2005). The mechanisms responsible for yield loss include high temperature during the reproductive phase, which is associated with a decrease in yield due to a decrease in the number of grains and kernel weight. Under high temperatures, the number of ovules that arefertilized and develop into grain decreases(Schoper et al., 1987a, b). Elevated temperatures also negativelyaffect the seedling and vegetative stages. In maize, seedling growth is maximized at a soil temperatureof 26‰; above this temperature, root and shoot massboth decline by 10% for each degree of increase until35‰ when growth becomes severely retarded(Walker, 1969). Above 35‰, the maize leaf elongation rate, leafarea, shoot biomass, and photosynthetic CO2 assimilation rate all decrease(Watt, 1972). Elongation ofthe first internode and overall shoot growth of maizehave been suggested as the most sensitive processesof the vegetative stage to high temperatures(Weaich et al., 1996). A recent analysis suggested that an increase in temperature of 2‰ would result in a greaterreduction in maize yields within sub-Saharan Africathan a decrease in precipitation by 20%(Lobell and Burke, 2010). Zhang et al.(2010)reported past warming reduced production of maize in China by approximately 10%. The present study has shown that theincrease in thermal resources had negative effects onclimatic potential productivities of maize during thepast 50 years in the 12 considered provinces(i.e., except Shaanxi and Gansu), with an average decreaseof 12.7 kg hm^-2 yr^-1, which is consistent with theconclusions of Lobell and Burke(2010), Lobell et al.(2011), and Zhang et al.(2010).

Solar radiation has been declining across manyparts of the world over the last 50 years as a consequence of industrialization, which increases atmospheric aerosols|a process known as "global dimming"(Stanhill and Cohen, 2001; Liepert, 2002; Yang et al., 2013). This phenomenon reduces the totalamount of photosynthetically active radiation(P AR;400-700 nm) and therefore in theory reduces crop yieldpotential via decreasing photosynthesis(Chameides et al., 1999). Several studies have indicated that globaldimming would result in crop yields declining(Stanhill and Cohen, 2001; Swain et al., 2007; Yang et al., 2013), especially in China where atmospheric aerosolshave increased due to industrialization(Zhang et al., 2010). Yang et al.(2013)ev aluated the impact of"global dimming" and climate change on wheat yield and water use in China during recent decades usingthe Agricultural Production Systems Simulator, and found that the decline in solar radiation coincided witha warming trend during the same period. The presentstudy has revealed that, from 1961 to 2010, the climatic potential productivities of maize in the mainplanting regions of China all decreased as a result ofthe decline in solar radiation(decline of 20 kg hm^-2 yr^-1or less). This result is supported by the conclusions of Wilson et al.(1995), Xiong et al.(2012), and Yang et al.(2013).

Precipitation is considered an important factor influencing crop yield variability, particularly incountries with monsoon climate(Challinor et al., 2003). Our study, however, showed that precipitation produced less climate-induced potential productivity changes of maize than radiation and thermalresources from 1961 to 2010 in China. This is becauseno pronounced trends in precipitation were found inthe maize growth period during the study period inmost of the 14 maize planting areas, and any precipitation effects in individual regions were often canceledout when averaged over latitude b and s. Additionally, the majority of maize planting areas in Chinareported by Xiong et al.(2012)are irrigated(over70% in 2000), such that climatic potential productivities and yields were relatively insensitive to changes in precipitation. We found that changes in precipitation resources had small effects on climatic potential productivities of maize in the 10 provinces(i.e., except Sh and ong, Shanxi, Gansu, and Heilongjiang)from 1961 to 2010, with a negligible average increaseof 5.9 kg hm^-2 yr^-1.4.2 Attribution of maize yield change to climate change and te chnolo gic al advancement

Worldwide, the climate and its effect on agriculture have continued to stimulate technological innovations that best suit specific climatic conditions.Therefore, it is essential for the development of future management strategies to better underst and hownon-climatic factors and climatic factors contribute tochanges in grain yield/yield components(Xiao and Tao, 2014). In the present study, we found that theincrease in maize yield from 1981 to 2010 in China wasmainly affected by technological advancement. T echnological advancement contributed to the maize yieldincrease by 99.6%-141.6%, whereas climate changecontribution was from -41.4% to 0.4%. In particular, the actual maize yields from 1961 to 2010 in Sh and ong, Henan, Jilin, and Inner Mongolia increased by98.4, 90.4, 98.7, and 121.5 kg hm^-2 yr^-1, respectively.Correspondingly, the maize yields affected by technological advancement increased by 113.7, 97.9, 111.5, and 124.8 kg hm^-2 yr^-1, respectively. However, maizeyields reduced markedly under climate change, withan average reduction of -9.0 kg hm^-2 yr^-1. Clearly, technological advancement offset the negative effectsof climatic change on maize yield in China. Changesin technological advancement played a dominant rolein maize yield increases in China over the past threedecades.

Generally, crop yield formation is influenced bywater and nutrient supply via the soil, by seasonalweather conditions, by pests and diseases, and by numerous agro-management decisions, such as crop cultiv ar choices, soil tillage, sowing or planting, fertilization, irrigation, and plant protection. Changes intechnological advancement played a dominant role inyield increases in China over the past three decades. Inthe future, adaptation strategies and appropriate management practices are needed to minimize the negativeeffects of climate variability and change, including extreme events, on agricultural productivity. Thus, newvarieties with increased resilience to abiotic and biotic stresses will play an important role in autonomousadaptation to climate change(Easterling et al., 2007;Cairns et al., 2012). F avorable agronomic management practices include shifts in the sowing date, improved irrigation and fertilization management, and no-tillage cultiv ation(Jin et al., 2012; Zhang et al., 2013).

The present study's findings are supported bysome previous reports(Evenson and Gollin, 2003; Xiao and Tao, 2014). As long ago as 50 yr or more, scientists have been able to offset yield losses by up to40% through the development of improved germplasm and management options(Evenson and Gollin, 2003).Xiao and Tao(2014)indicated that, during the period1980-2009, cultiv ar renewal contributed to increases inwinter wheat yield in the North China Plain by 12.2%-22.6%, and fertilization management contributed by2.1%-3.6%. However, today, scientists are faced withan even harder challenge|to meet the needs of future generations in the face of both population growth and climate change. While this challenge is immense, advancements in molecular and phenotype tools combined with the vast accumulated knowledge on themechanisms responsible for yield loss will provide asolid foundation to achieve increases in productivitywithin maize systems.4.3 Limitations

The AEZ model was used to estimate the climaticpotential productivity of maize in China in this study.The AEZ method has proved to be a useful conceptin agronomic studies of climate change(Easterling et al., 2007; Tubiello et al., 2007). More rigorous calibration methods have improved the agreement between simulated climatic potential productivity and census/observed yields in this study. However, theAEZ method does lack the necessary information totake into account the diversity of economic decisions(Seo, 2014), geography, climate, and culture betweendifferent regions. Thus, using the AEZ model to estimate the impacts of different climate variables on climatic potential productivity of maize in differentregions in China over time remains highly challenging. Moreover, based on the actual situation of maizeplanting areas, how to scientifically determine the effectiveness of the various levels limiting the production potential also needs further study. In addition, contributions of climate change and technological advancement to maize yield depend to some extent onthe data used, such as the observed yields. However, most of the actual yields are still based on manualobserv ations and single observation measures at current experimental agricultural meteorological stationsin China. These of course contain some observ ationalerrors that were not considered in this study, whichcould outweigh the effects of climate change and technological advancement on maize yield change.

Notwithst and ing these uncertainties, the presentreported results offer a comprehensive estimation ofthe attribution of maize yield increase in China to climate change and technological advancement over thepast three decades on the national scale, as well asthe relative impacts of separate climate variables onclimatic potential productivity of maize over the pastfive decades, with important conclusions that providegreater insight to crop-climate-technology interactions and sustainable maize production in China. A promising adaptation strategy in agricultural technology willresult from the development of new crop varietiespossessing higher photosynthesis use effciency, highthermal requirements, and resistance to biotic/abioticstresses under global warming(Tao and Zhang, 2010).5. Conclusions

In this study, daily climate variables gatheredfrom 553 meteorological stations in China for the period 1961-2010, detailed observ ations of maize from653 agricultural meteorological stations for the period 1981-2010, and results using an AEZ model, were used to explore the attribution of maize yieldincrease in China to climate change and technological advancement. The relative impacts of differentclimate variables on climatic potential productivity ofmaize in the past five decades were also assessed.

The study found that climate change had a significant adverse impact on climatic potential productivity of maize from 1961 to 2010 in China. Decreasedradiation and increased temperature were the mainfactors leading to the decrease in climatic potentialproductivity. However, changes in precipitation had asmall effect. The maize yields in the 14 main planting provinces of China increased markedly over thepast 30 years, which was opposite to the decreasingtrend of climatic potential productivity. This resultsuggests that technological advancement offset thenegative effects of climatic change on maize yieldsin China. Therefore, adaptation strategies such asbreeding cultiv ars that have higher heat requirements, high photosynthetic effciency, and resistance to biotic/abiotic stresses, together with other improvedagronomic technologies, will be effective to offset negative impacts of climate change in the future. Insummary, the present reported results provide useful insights for policymakers attempting to developeffective and sustainable strategies for agricultural development. T o ensure the continuous increase of maizeyields under climate change, further studies and research, as well as effcient environmental policies and actions are required.