Cho Zhng,Jiying Kong,Min Tng,Wen Lin,Dinyun Ding,*,Ho Feng

a College of Hydraulic Science and Engineering,Yangzhou University,Yangzhou 225009,Jiangsu,China

b Institute of Water Saving Agriculture in Arid Areas of China,Northwest A&F University,Yangling 712100,Shaanxi,China

c Collage of Agriculture,Shanxi Agricultural University,Jinzhong 30801,Shanxi,China

Keywords: Plastic film mulching Soil temperature Compensatory effect AquaCrop model Maize growth

ABSTRACT Temperature compensatory effect,which quantifies the increase in cumulative air temperature from soil temperature increase caused by mulching,provides an effective method for incorporating soil temperature into crop models.In this study,compensated temperature was integrated into the AquaCrop model to investigate the capability of the compensatory effect to improve assessment of the promotion of maize growth and development by plastic film mulching (PM).A three-year experiment was conducted from 2014 to 2016 with two maize varieties (spring and summer) and two mulching conditions (PM and non-mulching (NM)),and the AquaCrop model was employed to reproduce crop growth and yield responses to changes in NM,PM,and compensated PM.A marked difference in soil temperature between NM and PM was observed before 50 days after sowing(DAS)during three growing seasons.During sowing-emergence and emergence-tasseling,the increase in air temperature was proportional to the compensatory coefficient,with spring maize showing a higher compensatory temperature than summer maize.Simulation results for canopy cover (CC) were generally in good agreement with the measurements,whereas predictions of aboveground biomass and grain yield under PM indicated large underestimates from 60 DAS to the end of maturity.Simulations of spring maize biomass and yield showed general increase based on temperature compensation,accompanied by improvement in modeling accuracy,with RMSEs decreasing from 2.5 to 1.6 t ha-1 and from 4.1 t to 3.4 t ha-1.Improvement in biomass and yield simulation was less pronounced for summer than for spring maize,implying that crops grown during low-temperature periods would benefit more from the compensatory effect.This study demonstrated the effectiveness of the temperature compensatory effect to improve the performance of the AquaCrop model in simulating maize growth under PM practices.

1.Introduction

Maize (Zea maysL.) accounted for 39% of cereal production in 2020 and total maize yield has increased by 36%over the past decade (FAOSTAT: https://www.fao.org/faostat/en/).China is the second-largest maize producer,with 22% of world production.In 2021,260 × 106t of maize grain was produced in China,accounting for 42% of cereal production [1].The Loess Plateau,a semiarid rainfed agricultural region spanning 64 × 106ha in Northwest China,is a dryland maize belt subject to terrain and climatic conditions [2,3].

Plastic film mulching (PM) technology has been widely used in arid and semiarid areas and assists in regulating soil temperature,conserving soil water storage,and boosting crop yield [4].With increasing agricultural modernization,PM practices have been rapidly developed in China.Over the last two decades,China’s film consumption has grown from 0.72×106to 1.38×106t and its film coverage area from 10.62 × 106to 17.63 × 106ha [5],both statistics leading the world [6].

PM has developed into a necessity for high agricultural production in arid and semiarid areas.PM is consistently reported to improve crop biomass [7,8],yield [9,10],evapotranspiration[11,12],water use efficiency[2,13],soil moisture and temperature[14],and greenhouse gas emissions[15].However,these effects in field experiments varied with climate,soil,crops,and other factors.Tiquia et al.[16]proposed that PM could increase soil CO2content,exacerbating inadequate soil aeration and crop growth.Wang et al.[11]suggested that potato growth was inhibited by increased temperature stress under PM conditions.Chen and Dong [17] and Qi et al.[18]found that cotton grown in PM conditions was acculturated,but failed to increase yield owing to rapid leaf senescence in the late season.With the exception of direct factors like soil evaporation and temperature,other characteristics including crop evapotranspiration,water use efficiency,and photosynthesis were indirect factors under PM and were susceptible to change.Coupling these factors with a mechanistic model is needed to advance the PM theory of high-efficiency production.

Crop models have progressed beyond field experimental studies from simulating a single physiological process to simulating the entire growth process incorporating the factors of atmosphere,crop,and soil.Several crop models have been developed to account for the effect of PM on crop growth.Han et al.[19] added a film mulching module to the DNDC (DeNitrification-DeComposition)model in order to modulate soil temperature,wind speed,humidity,and evaporation rate by adjusting film coverage proportion and time period.Liang et al.[20] developed a new model of WHCNS(Water Heat Carbon Nitrogen Simulator) to investigate the effects of mulching on water consumption,N fate,organic matter turnover,and crop growth in the farmland ecosystem.Li et al.[21]refined the CSM-CROPGRO-Cotton model to improve evapotranspiration estimation under PM.PM practice in the RZ-SHAW model(a hybrid version of the Root Zone Water Quality Model and the Simultaneous Heat and Water Model) [22] could alter the shortwave and longwave radiation transfer,turbulent heat and vapor transfer from the surface,and energy and water balances in the system.

The AquaCrop model [23],a water-driven model published by the Food and Agriculture Organization of the United Nations(FAO),accounts for the effects of diverse agricultural management practices(irrigation mode and schedule,mulching cover,field surface practices,and weed infestation) on water and crop growth dynamics.Since the release of the model in 2009,dozens of studies have been conducted to evaluate the robustness of the model under mulching conditions for simulating canopy development,yield,and evapotranspiration of crops including wheat [24,25],maize[26,27],cotton[28],barley[29],sweet pepper[7],soybeans[30],and potato[31].Owing to the limitations of the model mechanism,the AquaCrop model focuses on the influence of PM on soil evaporation,but does not address the implications of soil temperature change.Incorporating soil temperature into the model would refine the mulching module.

Because most crop models have been encapsulated,modifying the model structure in the source code is impossible.Quantifying the compensatory effect of PM is of practical value for predicting the crop growth and development process under PM conditions and improving the model of mulched crops.Zhang et al.[32] proposed the temperature compensatory coefficient to quantify the compensatory effect of soil temperature increase under PM on growing degree days (GDD).Following that innovation,several studies have been performed to reveal the dynamic of the compensatory coefficient for diverse growth stages.Zhang et al.[33]developed the SUCROS-Cotton model to refine crop growth simulation under PM conditions based on the compensatory effect.Ma et al.[34] enhanced the CERES-Rice model by incorporating the compensatory effect to better reproduce rice growth under PM.Shen et al.[35] embedded the compensatory effect into the DSSATCERES model to improve the simulation accuracy of maize growth.Zou et al.[36] quantified the compensatory effect of PM on the plant height of wheat and maize using a logistic model.

Previous studies have focused primarily on model evaluation in simulating crop growth and soil properties under PM conditions,and several studies of compensatory effect concentrated on changes in simulation accuracy.Lacking has been a systematic analysis of the contribution of a temperature compensatory increment in the crop model under varying growth conditions.In this study,a three-year field experiment was performed with spring and summer maize in two climatic regions under PM and nonmulching(NM)conditions and the AquaCrop model was employed to simulate crop growth and yield production under both original and compensated GDD condition.The objectives were to:(1)characterize the dynamics of soil temperature under PM and NM practices during spring maize and summer maize growing seasons;(2)investigate the variation trends and contributions of the compensation coefficients on GDD and the reference evapotranspiration(ET0) in various growing stages;and (3) evaluate the capability of the compensatory effect for interpreting crop development and production in the AquaCrop model.

2.Material and methods

2.1.Study sites

Two sites in the southern Loess Plateau in the northwest of China were chosen(Fig.S1):For spring maize,the field experiment was performed at the Changwu Agro-Ecological Experiment Station,Chinese Academy of Sciences (35°20′N,107°40′E,and 1200 m above sea level).The mean annual air temperature,precipitation,and frost-free period were 9.1 °C,584 mm,and 171 days,respectively.The distribution of precipitation was extremely uneven,with the mean precipitation of 460 mm throughout the spring maize growing season,accounting for 80% of the annual total.Soil texture was classified as silt loam,with 21%clay,43%silt,and 35% sand according to the Harmonized World Soil Database(HWSD).The mean field capacity,wilting point,bulk density,and pH were 34% cm3cm-3,16% cm3cm-3,1.34 g cm-3,and 8.0,respectively.For summer maize,the field experiment was conducted at Institute of Water Saving Agriculture in Arid Areas of China,Northwest A&F University (34°20′N,108°24′E,and 506 m above sea level).In this region,the mean annual air temperature,precipitation,and frost-free days were 12.9 °C,635 mm,and 213 days,respectively.Similarly,the total precipitation received during the summer maize growing season made up half of the annual total (330 mm),basically meeting the crop demands.The soil of the experimental field was loamy type,with a field capacity of 0.28 cm3cm-3,permanent wilting point of 0.11 cm3cm-3,and dry bulk density of 1.28 g cm-3.The groundwater table in both fields was below 40 m depth,inaccessible to the soil tillage layer.

2.2.Experimental design and data collection

Two field experiments with PM and NM practices were performed during three consecutive growing seasons,2014,2015,and 2016.The NM treatment was a bare and unmulched soil surface,whereas in the PM,the soil surface was nearly completely covered(except for holes allowing seed emergence)with transparent plastic film (0.01 mm thickness).Row and plant spacing were assigned as 60 and 40 cm following local practice,with a seeding density of 52,000 plants ha-1(Fig.S2).Spring maize is usually sown in April and harvested in September and summer maize is typically sown in June and harvested in October.The phenological development of maize sowing and harvesting,and key growth stages for each season were recorded (Table S1).Additional crop management practices were also followed to control weeds,pests,and diseases.

In each plot,three representative maize plants were marked and sampled at each growth stage.The leaf area was measured with a measuring tape to estimate leaf area index(LAI).The canopy cover(CC)was calculated from the LAI using an empirical equation provided by Hsiao et al.[37].The aboveground biomass was weighed after drying in an oven at 75 °C for 72 h.At the maturity stage,ten maize plants were randomly harvested from each plot,and the grain yield was determined by threshing and air-drying to constant weight.Volumetric soil water content was measured about once per week using a Trime-TDR instrument (TRIME-TDRPICO-IPH-T3,Imko,Ettlingen,Germany) at intervals of 10 cm in the 0-120 cm soil profile.The temperature sensor was installed in the middle of each plot and positioned between two rows of maize.Soil temperature was taken at a depth of 15 cm for each plot using an automatic temperature monitoring sensor (5TE) and was recorded hourly by a data logger(EM50,Decagon,WA,USA).Meteorological data was collected by a Dynamet weather station(Dynamax Inc.,Houston,TX,USA),located 100 m from the experimental field.Five climatic variables:air temperature,wind speed,relative air humidity,net solar radiation,and precipitation,were logged in a data logger (CR1000,Campbell Scientific Inc.,Logan,UT,USA).ET0was calculated by the FAO Penman-Monteith method from air temperature,wind speed,relative air humidity,and net solar radiation data[38].The dynamics of air temperature,ET0,and precipitation during the three growing seasons are depicted in Fig.S3.

2.3.Crop model description,calibration and validation

The AquaCrop model,an integrated model connecting soil,crop,atmosphere components,was used in this study.AquaCrop is a water-driven model that describes the effects of water interaction across a range of atmosphere (rainfall,temperature,and evapotranspiration),crop(crop cover,root growth,and biomass production),and field management (irrigation and agronomic practices)conditions.The main processes involved in crop development and production were as follows.

(1) Canopy cover(CC).The crop canopy cover is the vital component in AquaCrop simulation as it determines the amount of water transpired.The development and senescence of the green canopy under unstressed conditions is quantified by three exponential equations based on four parameters,including initial canopy cover (CC0),canopy growth coefficient (CGC),maximum canopy cover (CCx),and canopy decline coefficient (CDC).

(2) Crop transpiration (Tr).Tr is calculated by multiplying ET0by the crop transpiration coefficient (KcTr) and by incorporating the effect of water stress (Ks) and cold stress (KsTr)(Eq.(1).

where Ks represents the soil water stress coefficient,describing the effects of waterlogging and stomatal closure,and KsTris the cold stress coefficient of a growing degree.In the absence of water stress,Ks and KsTrare equal to 1.KcTris proportional to CC* (adjusted canopy cover) and the coefficient of maximum crop transpiration.

(3) Aboveground biomass(AGB).Daily AGB was accumulated as a function of the ratio of crop transpiration (Tr) to the ET0and water productivity (WP) (Eq.(2).

where WP*is the normalized water productivity(g m-2),which can be taken as a constant for a given crop and can be adjusted by atmospheric CO2and soil fertility.

(4) Grain yield formation.The grain yield is accumulated by multiplying AGB and harvest index (HI).HI is simulated as a linear increase over days from flowering to physiological maturity.HI is modulated by parameters including phenology timing,water stress,and pollination failure.Further elements of water,heat,salinity stress,rooting expansion,and soil water budget are also incorporated into the AquaCrop model with the exception of the concepts and principles mentioned above.Further algorithmic details of the Aqua-Crop model are described in Steduto et al.[23],Raes et al.[39],and Hsiao et al.[37].

The six data input files for the AquaCrop model fall into the following four categories: (1) Climate.Three kinds of weather data:air temperature (maximum and minimum),rainfall,and ET0,are minimum inputs for the model.(2) Crop.Crop parameters contain both conservative and non-conservative (user-specific) parameters.The conservative parameters,including threshold air temperatures,maximum crop coefficient,and WP*,were held constant.Other non-conservative and cultivar parameters (e.g.,days of emergence,senescence,and maturity) need to be calibrated depending on cultivars,environmental conditions,and planting methods.(3) Management.The management practices in Aqua-Crop include irrigation and field data.Both the spring and summer maize were grown under rainfed conditions without irrigation.For the field files,the soil fertility was set to non-limiting and the weed control was set at‘‘very good level”(＜5%of the CC).Plastic mulch with 95% soil coverage and no mulch were employed to simulate PM and NM scenarios,respectively.(4) Soil.In the soil profile,the soil hydraulic properties(SAT,FC,and PWP)of each layer were filled in using thein situmeasured value(Table S2).The groundwater was absent owing to the 40-m deep groundwater table.

The simulation mode of crop canopy development based on GDD was used in this study,as the thermal time duration of the crop development stages could be adjusted to the temperature regimes of the corresponding years.The model parameters of spring and summer maize were calibrated using observed data from the 2014 growing season.The model robustness was evaluated using the independent data sets during the 2015 and 2016 growing seasons.The final parameters of spring and summer maize determined in AquaCrop are presented in Table 1.

Table 1Main conservative and non-conservative parameters for maize in the AquaCrop model.

2.4.Temperature compensatory effect of film mulching

PM can alter the soil temperature and change the GDD during crop growing seasons.In the AquaCrop model,only the effect of mulches on soil evaporative loss is considered,and soil temperature changes affecting the crop development simulation are not.Thus,to quantify the compensatory effect,a temperature compensatory coefficient was developed by [32],calculated as follows:

wheremis the compensatory coefficient for theithphenological stage.The GDD is growing-degree days (°C d).GDDa（NM）and GDDa（PM）are the thermal time nodes at theithphenological stage under non-mulching and plastic mulching,respectively.The GDDs（NM）and GDDs（PM）represent respectively cumulative soil temperature without and with plastic mulch at the thermal time node of GDDa（PM）.Wheni=1,both ofandare equal to 0 °C d.

A well-developed canopy can block direct solar radiation from heating the soil,offsetting the effect of PM on soil temperatureincrease[40].Ding et al.[41] investigated the variation ofCmover summer maize seasons and concluded that changes inCmcan be seen prior to the maize filling stage as the solar radiation would be intercepted by the full canopy when the crop is entering the tasseling stage.The focus of this study was,accordingly,Cmfrom the sowing to the tasseling stage.After determination ofCm,the maximum and minimum temperature data entered into the AquaCrop to simulate the PM scenario were updated following Eqs.(4)-(6).The construction process of soil temperature effect in the Aqua-Crop model is presented in Fig.1.

Fig.1.Comparison between flowcharts that include and exclude soil temperature compensatory effect in the crop model.

where ΔTis the adjusted temperature for each 1°C increase in soil temperature under PM(°C),Tbaseis the base temperature of 8°C,Ts(PM)andTs(NM)are the respective soil temperatures (°C) of nonmulching and plastic mulching treatments,Ta(NM)andTa(PM)are the respective air temperatures (°C) for non-mulching and plastic mulching treatments,andTmaxandTminare the calculated maximum and minimum temperatures (°C) of plastic mulching for running the model.

3.Results

3.1.Comparison of soil surface temperature between NM and PM

Differing variational tendencies were observed in spring and summer maize (Fig.2).For the spring maize,the soil temperature increased as the growth period progressed and peaked at about 25°C at the tasseling stage,90-110 DAS.The soil temperature then decreased continuously until the plant reached maturity.In contrast,summer maize showed a higher initial soil temperature with a shorter heating-up phase.After reaching its peak around 30 DAS,the soil temperature presented a declining trend.

Fig.2.Growing season soil temperatures for spring (A) and summer (B) maize from 2014 to 2016.Solid and dashed lines represent soil temperatures under non-mulching(NM) and plastic film mulching (PM) conditions,respectively.

For the spring maize,the mean soil temperature under NM treatment was approximately 20 °C across the growing seasons.But soil temperatures were higher under PM than under NM for both spring and summer maize,especially in the early stages.During the 2014 growing seasons,the difference in soil cumulative temperature between NM and PM was only 56°C,obviously lower than in the 2015 (205 °C) and 2016 (355 °C) growing seasons(Fig.2A).Summer maize showed a higher mean soil temperature of 26.4 °C in the three seasons than spring maize.The total soil cumulative temperature difference between NM and PM in the 2014 maize growing cycle (269 °C) was triple that of the 2015(85 °C) and 2016 (86 °C) growing seasons (Fig.2B).As in spring maize,greater soil temperatures in PM were observed only during the early stages (before 50 DAS),and thereafter PM and NM both maintained stable soil temperatures until harvest.Thus,PM effectively raised the soil temperature and was higher prior to jointing stage.

3.2.Characteristics of compensatory coefficient for air temperature and ET0 change

All of the compensatory coefficients were positive,indicating that cumulative air temperature under PM would be increased by the rise of soil temperature(Table 2).For spring maize,the distribution of compensatory coefficients throughout the sowingemergence and emergence-tasseling periods ranged from 0.54 to 1.63 and 1.19 to 1.52,respectively.Summer maize showed lower compensatory coefficients,ranging from 0.71 to 0.86 during the sowing-to-emergence stage and 0.54 to 0.69 during the emergence-to-tasseling stage.The mean compensated air temperature for spring maize across the three seasons was 218 °C,whereas only one third of the increase in temperature (about 68 °C) was presented for summer maize,in which the compensated air temperature in 2014 was nearly twice as high as in 2015 and 2016.At each growing stage,the increase in air temperature accumulation was proportional to the compensatory coefficient.Compensated air temperatures during the emergence-totasseling stage were higher than those of the seeding stages for both spring and summer maize.The change of potential evapotranspiration induced by the increase of air temperature was also analyzed (Table 2).The mean increase in ET0for spring maize was 21.4 mm,accounting for 6.5% before the tasseling stage,whereas the increase in ET0for summer maize was 6.3 mm,accounting for only 2.4% before the tasseling stage.

Table 2Effect of compensatory coefficient for air temperature and ET0 changes under PM for both spring and summer maize in three growing seasons.

3.3.Evaluation of the original AquaCrop model simulation for NM and PM

Fig.3 depicts the performance of the AquaCrop model for simulating summer maize development and production.The simulated length of total growth period was in close agreement with the three-year measurements.For the maize canopy cover,simulated dynamics of time node,such as maximum CC,start of senescence,and maturity,generally matched the measurements.The PM treatment presented a simulation superior to that of the NM treatment in three seasons,with root mean square errors(RMSE)ranging from 2.7%to 6.3%vs.4.4%to 12.9%.Nonetheless,the simulation of AGB displayed large discrepancies for both NM and PM,except in year 2014(the calibration data).Large underestimates by simulation of measured AGB were observed from 60 DAS onward and lasted until the end of the stage.The accuracies of the simulated AGB for PM were lower than those of NM,with the maximum RMSE being 3.7 t ha-1in 2016.As with the AGB simulation results,simulation also underestimated grain yield.The mean relative errors of NM treatments (-9%) were smaller than those of PM treatments (-18%).

Fig.3.Measured and simulated canopy cover(CC,%),aboveground biomass(AGB,t ha-1),and grain yield(t ha-1)of summer maize without(NM,A,B,and C)and with plastic film mulch (PM,D,E,and F) during three growing seasons from 2014 to 2016.Error bars represent he standard errors of the mean.RMSE are shown for CC (%) and AGB (t ha-1).

Simulations of spring maize showed similar behavior to those of summer maize,indicating that underestimation persisted for both NM and PM throughout the three growing seasons (Fig.4).Closer agreement between simulated and measured AGB was achieved for NM treatment,with a minimum RMSE of 1.0 t ha-1.Under the same GDD parameter setting,the crop development periods were shortened in 2015 and 2016,but these were inconsistent within situmeasurements.In general,crop development of CC,AGB,and yield under the PM condition was vigorous,with greateraccumulation than those of NM.AquaCrop showed a dependable capability for revealing the dynamicity of maize development in response to varied tillage practice,even though its simulation accuracy was unsatisfactory.

Fig.4.Measured and simulated aboveground biomass (AGB,t ha-1) and grain yields (t ha-1) of spring maize without (NM) and with plastic film mulch (PM) during three growing seasons in years 2014 (A),2015 (B),and 2016 (C).Error bars represent standard errors of the mean.RMSE are shown for AGB (t ha-1).

Simulations of soil water storage in the wetting layer(0-40 cm)and maximum root depth (0-120 cm) of summer maize for both NM and PM scenarios were quantitatively evaluated against the measurements(Fig.S4).In general,the variation of total water content in 0-40 cm and 0-120 cm soil layers ranged from respectively 70 to 130 mm and 220 to 340 mm.Simulated dynamics of soil water content corresponded to the variability of the measurements in most seasons.Comparable simulation accuracy was observed in both NM and PM treatments during three growing seasons.A large discrepancy between simulated and measured soil water content in the top 40 cm of the soil layer was observed when the crop entered the tasseling stage(65 DAS)in the 2015 season.Heavy rain(Fig.S3)led to an increase in soil moisture that remained until the end of maturity.

For the spring maize,the soil water content was only measured in 2014 and 2015.In general,less precipitation during the growth period resulted in less soil water storage in both 0-40 cm and 0-120 cm soil layers,with ranges of 40-120 mm and 150-300 mm,respectively (Fig.S5).The simulated total soil water content for both NM and PM treatments followed the trend of the actual measured values.Closer agreement between simulated and measured soil water content was achieved for the NM scenario in two consecutive seasons,with the RMSE ranging from 7.7 to 33.6 mm of the 0-120-cm soil layer and 6.6 to 13.2 mm in the 0-40-cm soil layer.For the PM treatment,the main errors occurred in the 0-120 cm soil depth layer during the late stage,when heavy precipitation caused a rapid increase in soil water content.

3.4.Performance of temperature compensatory effect in simulating maize biomass and yield

A general contribution of compensatory effect to variation in maize AGB and yield simulation was observed during three growing seasons(Fig.5).For the spring maize,the simulated AGB showed a sizable increase based on temperature compensation.Simulation accuracies of AGB were also improved,with the RMSE dropping from 2.5 to 1.6 t ha-1in 2014 and from 4.1 to 3.4 t ha-1in the 2015 growing seasons relative to the simulation without temperature compensation.Compared to the difference in AGB between compensation and no compensation,the increase in simulated yield was not apparent,with the RMSE slightly increased.In the 2015 growing season,the temperature-compensated simulated AGB dynamics were greater in the early stages but accumulated more slowly in the latter stages,approaching the uncompensated condition.The compensated simulated yield was even lower than the uncompensated at maturity.As the compensatory effect increased crop growth,the increased canopy cover intensified crop transpiration and accelerated soil water depletion.Based on the underestimation of the uncompensated conditions (Fig.S5d),the potential water stress was raised,further impairing the formation of AGB and grain yield.Thus,the temperature compensation of plastic mulching would also have a negative effect on crop growth and production under the conditions of low soil water content.

Fig.5.Aboveground biomass (AGB,t ha-1)and grain yields (t ha-1)under measured and simulated conditions with (PM-c)and without(PM-o)implementing temperature compensated effects during growing seasons of year 2014 (left column),2015 (middle column),and 2016 (right column) for both spring maize (upper panel) and summer maize (lower panel).

The simulated increases in AGB and yield for summer maize over three growing cycles were less prominent than those of spring maize under temperature compensation(Fig.5).The compensated simulation accuracy of AGB and yield were only slightly improved than the uncompensated,and the significant difference between simulated andin situmeasured AGB remained in the latter stages of 2015 and 2016.The main reason for the results was attributable to a lower compensatory increment in air temperature (around 68 °C) during summer maize growing seasons (Table 2).Although a limited increase in temperature was insufficient to induce appreciable changes in simulating crop AGB and yield,the temperature compensatory effect allowed better estimation of the promotion by plastic mulching of maize growth and development comparing with uncompensated conditions.

4.Discussion

4.1.Effect of film mulching for soil temperature variation

In general,plastic film mulching increases soil temperature by two means [42,43].The first is the greenhouse effect caused by PM,which absorbs solar radiation upon the mulching film and prevents heat exchange inside and outside the mulching film[44,45].The second is that PM application reduces soil evaporation and maintains soil moisture in the topsoil (＜20 cm),resulting in the reduction of latent heat loss and an increase of soil specific heat[46,47].In this study,the effect of film mulching on temperature varied with environmental factors and crop development period.

As shown in Figs.2 and S3,soil temperature fluctuated with air temperature,but the differences between NM and PM were adjusted by precipitation.Greater temperature differences occurred during periods with less precipitation: 0-58 DAS for spring maize in 2015 and 0-50 DAS for summer maize in 2014.The soil surface temperature under PM also increased during the early to middle period(Fig.2).Likewise,Zhou et al.[48]found that plastic mulched areas received more solar energy during the early stages of maize with an inadequate canopy,resulting in warmer topsoil.Wang et al.[44],investigating the effects of PM on soil temperature,indicated that the mean soil temperature of the mulching treatment rose by 2.3 °C before July and by nearly 1.2 °C after July during the maize growing season.Gu et al.[49]compared the effect of biodegradable and conventional polyethylene film,and suggested that they generally increased soil temperature at early growth stages.

4.2.Variability of the compensatory coefficients

Generally,the GDD response to air temperature for a given crop to reach a growing stage is relatively stable in a given climatic region.Crop development could be accelerated under PM conditions by a rise in soil temperature [34,50].However,the increase in soil temperature is not constant,being influenced by multiple factors including crop variety,weather conditions,sowing date,and mulching methods.The increase in cumulative air temperature from soil temperature increase was determined by a compensatory coefficient,in which both air and soil temperatures of mulching and non-mulching for a given growing stage was covered.

According to Eq.(3),the variation of compensatory coefficients is correlated with air and soil temperature,as well as crop development.Theoretically,a larger soil temperature difference between PM and NM would lead to a greater denominator,yielding a smallerCmvalue.Otherwise,soil temperature difference would be diminished with canopy shading,in which case theCmvalue would increase with the decreasing denominator [36].TheCmvalues of spring maize in this study followed this principle,but the summer maize presented the opposite behavior.The difference could be ascribed to the trends of temperature change.For the spring maize,the temperature differences between PM and NM before seedling emergence (0-15 DAS) were generally higher than those in the tasseling stage (16-95 DAS) (Fig.2A).In contrast,the temperature difference between PM and NM for summer maize was consistent for both the seedling stage (0-12 DAS)and tasseling stage (13-55 DAS),resulting in negligible change inCmbetween the two stages.Comparable studies have been conducted to investigate the variation ofCmamong other crops.Zhang et al.[33] reported that theCmvalues of cotton were 0.51 during the seedling to squaring stage and 0.22 before the anthesis stage.Yang et al.[51] reported that theCmof summer maize was 1.36 from the sowing to seedling stages and 0.64 from the seedling to tasseling stages.Fu et al.[52] reported theCmof summer maize as 0.45 during the seedling stage and 0.20 during the tasseling stage.Zou et al.[36]suggested that theCmshould remain constant for the winter wheat growing season and increase as maize develops.

4.3.Performance of compensatory effect for interpreting crop development and production

Previous studies [13,53] have demonstrated that crop emergence,growth,and development would be improved under PM conditions.The primary benefits could be linked to the retention of soil water and temperature [54].Zhou et al.[48] evaluated the effect of several tillage practices on maize development and found that higher soil temperatures under PM could boost grain quality and yield.He et al.[55] found that wheat canopy ET and spike number were increased under PM practice owing to a rise in soil moisture and temperature.Ding et al.[56] evaluated the performance of PM for soil water consumption in a wheat-maize rotation and found that higher soil temperature led crops to transpire more efficiently,resulting in a greater AGB accumulation rate.Our findings were in agreement with these: crop canopy cover,dry matter,and grain yield were all improved under mulching relative to non-mulching treatment (Figs.3,4).

Owing to the limitations of the model mechanism,it is a challenge to descript the effects of all potential factors on crop growth and development conferred by mulching.The function of mulching in AquaCrop model was focused on reducing soil evaporation losses,indirectly increasing soil water storage.In AquaCrop,air temperature is the key variable,as it determines the degree of temperature(both cold and heat)stress to crop transpiration as well as affecting pollination and harvest index [39].Consequently,the major contribution in the compensatory effect was an increase in air temperature over soil temperature under PM conditions.The crop ET0also increased in proportion to the air temperature.Similarly,Wang et al.[57] suggested that compensatory air temperature accounted for 4.9% of the total temperature for the whole crop growth stage.Yang et al.[51] investigated the compensatory effect of air temperature and found that the mean increased by respectively 1.4 °C and 0.9 °C at the maize seedling and tasseling stages.Fu et al.[52] evaluated the compensatory effect of plastic mulching for spring maize in a cold environment and proposed a mean compensatory effect of 1.4 °C before the jointing stage.

The influence of compensatory effect displayed in Fig.5,showing that the simulated crop development and production of spring and summer maize markedly improved,despite their inconsistency with the measured values in some seasons.The increases of simulation by compensatory effect were compatible with the increase in air temperature,which presented a shape contrast between spring and summer maize (Fig.5;Table 2).For the simulated crop AGB,the higher temperature contributed mainly to the ratio of transpiration(Tr)and ET0(Eq.(1)and Eq.(2).Based on this principle of AquaCrop,the increased temperature alleviated the stress degrees in crop transpiration.First,the cold stress (KsTr) for transpiration would be lessened with the increase of temperature.In AquaCrop,the upper threshold of the growing degree was a conservative parameter and defaulted to 12 °C day-1,indicating that crop transpiration is limited when growing degrees(mean temperature-base temperature)fell below 12°C.Second,KcTris proportional to both the CC and coefficient of maximum crop transpiration.The compensatory temperature accelerated crop development and increased canopy coverage on the same DAS,triggering an increase in KcTr.The potential negative effect occurring in Ks is that the water deficit would be strengthened as increased transpiration depleted soil moisture from the root layer(Fig.5).This effect was also reported in Li et al.[58] and Zhang et al.[59].

4.4.Contributions and limitations

In this study,temperature compensation derived from the compensatory effect was incorporated into the AquaCrop model to improve the simulation of maize growth and production under PM conditions.The main contributions are as follows: (1) the difference in soil temperature between NM and PM occurred before tasseling stage,meaning that the compensatory effect of PM operated mainly before tasseling stage;(2) the variation in the compensatory coefficient for spring and summer maize was opposite at the maize seedling and tasseling stages;(3)compensatory temperature improved biomass and yield simulation in spring maize,but this improvement was less pronounced in summer maize.Further limitations in the model await attention: (1) The effects of film cover rate and method on rainfall infiltration are not incorporated in the model.Some studies[60,61]have shown that PM can inhibit precipitation from infiltrating the soil,lowering soil moisture and reducing crop yield.(2)Although higher temperature in GDD could promote crop development,increased air and soil temperature may be accompanied by heat stress,impairing canopy and biomass growth.Lu et al.[62] reported that high temperature led to mortality in maize roots,resulting in a decline in yield.The results of this study await validation in warmer or drier regions and a corresponding theory should be developed.

5.Conclusions

This study revealed the dynamics of soil temperature under plastic film mulching and non-mulching,investigated the effects of the compensation coefficients on GDD by increasing soil temperature,and evaluated the use of the compensatory effect for predicting crop development and production.

A marked difference in soil temperature between NM and PM was observed in the early stages (before 50 DAS) during three maize growing seasons.The distributions of compensatory coefficients at various growth stages were opposite between summer and spring maize,with an upward trend in spring maize and a downward trend in summer maize.The increase in air temperature accumulation was proportional to the compensatory coefficient in each growing stage,with spring maize exhibiting a higher compensatory temperature than that of summer maize owing to its longer growth period.

Beyond the previous studies that proposed the model evaluation of compensatory effects on simulation accuracy,this work elucidated the contribution of temperature increase to the AquaCrop model and highlighted its potential for evaluating the positive effect of plastic mulching on crop growth and development.The air and soil temperatures under mulching and non-mulching,calculated according to the compensatory coefficient and temperature,could be efficiently monitored and retrieved from remote sensing,allowing an opportunity for mapping spatial and temporal variation of the compensatory effect for mulched agricultural production.

CRediT authorship contribution statement

Chao Zhang:Investigation,Writing -original draft.Jiying Kong:Conceptualization,Methodology.Min Tang:Software,Writing -review &editing.Wen Lin:Data curation,Resources.Dianyuan Ding:Supervision,Funding acquisition.Hao Feng:Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

This work was supported by the National Natural Science Foundation of China(51909228 and 52209071),the‘‘High-level Talents Support Program”of Yangzhou University,‘‘Chunhui Plan”Cooperative Scientific Research Project of Ministry of Education of China(HZKY20220115),the China Postdoctoral Science Foundation(2020M671623),and the ‘‘Blue Project” of Yangzhou University.The authors would like to acknowledge the Institute of Watersaving Agriculture in Arid Areas of China,Northwest A&F University and Changwu Loess Plateau Ecological Agriculture Experimental Station,Chinese Academy of Sciences for providing the experiment facilities.

Appendix A.Supplementary data

Supplementary data for this article can be found online at https://doi.org/10.1016/j.cj.2023.05.008.