Yuwei Chai,Qiang Chai,Rui Li,Yawei Li,Changgang Yang,Hongo Cheng,Lei Chang,Shouxi Chai,*

a State Key Laboratory of Aridland Crop Science,Gansu Agricultural University,Lanzhou 730070,Gansu,China

b College of Agronomy,Gansu Agricultural University,Lanzhou 730070,Gansu,China

c College of Bioscience and Technology,Gansu Agricultural University,Lanzhou 730070,Gansu,China

Keywords:Mulch pattern Yield formation Vegetative growth Water use efficiency Hydrothermal interaction

ABSTRACT Straw strip mulching (SM) is a new mulching technology.From 2012 to 2018,SM’s effects on soil moisture and temperature and production performances were compared with other mulching practices,using three treatments:full-cover plastic mulch (PM),no mulch with wheat sown in rows as the control (CK),and SM with 50% to 59%of the field area mulched.Compared with CK,on average over six growing seasons,SM and PM increased grain yield by 27.0% and 21.7%,straw yield by 21.6%and 22.6%,kernels ha-1 by 26.6%and 19.0%,net income by 29.8%and-25.0%,soil temperature at 5 cm by-1.5°C and 0.2°C from overwintering to maturity,and soil water storage at 0–200 cm by 25 and 22 mm,respectively.The increase in soil moisture in SM and PM was greater in the early period (overwintering to jointing) than in the later period (booting to maturity) and at 0 to 120 cm than at 120–200 cm in the early period.Although the mean evapotranspiration of whole growth period across six seasons was similar among treatments,SM and PM increased water consumption during the key formation period of yield components after overwintering by 16 and 32 mm,respectively,while reducing it before overwintering.Compared with CK,SM and PM had the effects of warming during overwintering and cooling after jointing.By increasing water consumption after overwintering and ratio of transpiration to evapotranspiration and providing favorable soil temperature for multiple growth stages and more sufficient soil moisture,SM and PM promoted vegetative growth and increased kernels ha-1,the main mechanisms by which SM and PM increased grain yield relative to CK.Relative to PM,SM is a more economically beneficial and environment-friendly technology for dryland wheat production.

1.Introduction

Shortage of water is the primary limitation on rainfed agricultural development [1].Rainfed agriculture without irrigation accounts for approximately 55% of the total arable land in China[2] and is practiced mainly in northern China,the primary zone for potential increases in future production.Wheat (Triticum aestivum L.) is the most widely cultivated food crop in northern China,and there are 4.3–4.7 million hectares in rainfed agricultural regions with annual precipitation ranging from 250 to 600 mm[3].In these regions,55%to 60%of the precipitation occurs during June to September,and seasonal drought occurs because precipitation is scarce during the winter-wheat growing season and evaporation of soil water is high in the inland climate [4].Mulch on the soil surface has long been recognized [5,6] as the most effective means of reducing soil water loss by evaporation and providing more water for crop transpiration.Mulching thus ensures stability and continuous improvement in crop yields in the region.Currently,mulching with plastic is the most widely used practice in arid and semiarid regions of China [7,8].China is globally the largest user of plastic film mulch,accounting for 60% of worldwide consumption in agriculture,with an annual growth rate of 5.7% [9].

Although numerous reports demonstrate that plastic mulching increases crop yields in dryland areas,the practice has begun to attract widespread criticism because of the difficulty of removal from the field and recycling the resulting waste,as well as the increasingly severe pollution of soils and the environment by residual plastics and negative effects on crop production [10,11].Plastic film is composed primarily of polyethylene,which is resistant to degradation.When residual plastic film accumulates in soil above a threshold,negative effects include decreases in soil pore connectivity,soil porosity,infiltration,and absorption of water and nutrients [12];decreases in organic matter,total and nitrate nitrogen,and available phosphorus in the soil profile [13];and inhibition of root growth[7].Water use efficiency(WUE)and crop yields also decrease with increases in residual plastic film [9].It is thus desirable to identify an alternative to plastic mulching.

Straw resources are in excess in dryland farming areas of northern China,including corn straw,of which 70% is discarded or burned[14].In addition to improving conservation of soil moisture[3,15–17],straw mulch reduces fertilizer input by releasing nutrients [18],improves soil structure after incorporation [4],and avoids the pollution caused by residual plastic and straw burning.For these reasons,straw mulch can be an environmentally friendly,low-cost alternative to plastic mulch in developing sustainable agriculture.However,straw mulching can substantially reduce soil temperatures [9,19].

Low temperatures are a particular problem in the heat-deficient region of wheat monoculture in northern China,with mean annual temperatures of 6 °C to 8 °C and cumulative temperatures ≥10 °C of 2000–2600 °C.Conventional full-cover mulching with straw may result in soil temperatures that are too low during the early growth period,leading to impaired seedling emergence,inhibited growth and development,and ultimately either no increase or reductions in yield of wheat,maize,and other thermophilic crops[20].Full-cover straw mulching also impedes the sowing of wheat and increases straw consumption for mulching.

To reduce the adverse effects of low soil temperature caused by straw mulching in the early growth period of winter wheat,and to reduce the amount of straw applied as mulch and facilitate mechanized sowing,a new technology with partial-cover mulching(the straw strip mulch system,SM)with whole corn(Zea mays L.)stalks was developed by the authors.SM interlaces planting strips without mulch with mulching strips[21].Applying whole corn stalks as mulching material reduces the energy required to chop stalks.In comparison with shredded straw mulching,whole-cornstalk mulching can reduce evaporation of moisture that diffuses to edges of straw [22],because whole-stalk mulch has fewer gaps than shredded straw mulch.Because the surface of corn straw has a thick wax layer with poor moisture permeability [23,24],corn straw may reduce soil evaporation more than wheat and rice straw.The flocculent pith in corn straw can also strongly absorb and retain water [25],further reducing water loss.In addition,SM is suitable for mechanized wheat production in dryland of northwest China where most farm areas are small,uneven,and irregular in shape.For mechanical operation,SM is simpler and more convenient than PM.In our previous studies[26–29],similar or higher potato yields were obtained with SM than with wholefield plastic mulch in drylands.

To determine whether SM is suitable for winter wheat in the semiarid conditions in northwest China,comparative trials with no mulch and plastic mulch must be conducted under multiple environments across several years and locations.Such an approach is necessary because the practical effects of any mulching practice on soil moisture and temperature,growth,and final economic yield are strongly dependent on crop type,demands on soil water and temperature in different growth periods,local climatic conditions,and fluctuations in soil hydrothermal regimes between years[4,25,15,30].In addition,there are complex interactions among soil moisture,soil temperature,and atmospheric factors[1,31].According to Farzi et al.[32],mulching practices can provide favorable soil hydrothermal environments for multiple crop growth stages.In wheat,compared with no mulching,both plastic mulch and straw mulch increased yields [15,33,34].However,both mulches have also decreased yields [35–37].In some studies [3,38–43],yield with full-cover plastic film mulch was higher than that with partial-cover plastic film mulch.Sun et al.[10] found a yield increase of 77.9% under full-cover plastic mulch compared with 38.9% under partial-cover plastic mulch.In many studies [44–50],crop yields with full-cover plastic mulch are higher than those with full-cover straw mulch.However,many studies have shown that yields with full-cover plastic mulch are lower than those with full-cover straw mulch [51–55] or are similar [3,56,57].

In the present study,field experiments were conducted for six years from 2012 to 2018 in the semiarid rainfed agriculture region of Northwest China.The objectives of this work were the following:(1)to compare effects of straw strip mulch and full-cover plastic mulch on soil moisture,soil temperature,growth parameters,agronomic performance,and net economic return;(2) to identify the reasons for differences in soil moisture,grain yield,and WUE among the different treatments;and (3) to evaluate the feasibility of straw strip mulching in winter wheat in semiarid rainfed agroecosystems.

2.Materials and methods

2.1.Experimental site

From 2012 to 2018,experiments were conducted at two sites(Shanghe and Pinxiang) with differing hydrothermal conditions(Table S1;Fig.S1)and winter wheat cultivars suitable for local conditions.The two sites were in Dingxi city in Gansu province on the Loess Plateau of northwest China.The sites have a typical semiarid inland climate,and the rainfed agricultural region has one harvest per year.The soil type of the two sites is loessal soil,according to the USDA texture classification system [58].

Shanghe is located at 34°99′N,105°07′E with altitude 1590 m,mean annual temperature 8.4 °C,mean annual rainfall 444 mm,and mean annual evaporation 1450 mm.Pinxiang is located at 35°11′N,105°19′E with altitude 1750 m,mean annual temperature 7.2°C,mean annual rainfall 391 mm,and mean annual evaporation 1500 mm.Total precipitation during the growth period(sowing to maturity)accounted for only 43.7%–53.9%of the annual precipitation at Shanghe and 49.9%–69.7% at Pinxiang.November to March of the following year is the season with the lowest precipitation,and the cumulative precipitation during this phase accounted for only 5.2%–16.5%of the total precipitation during the growth period at Shanghe and 12.0%–18.1% at Pinxiang.Drought in winter and spring is the main characteristic of the climate in the semiarid rainfed region of northwest China.The basic soil properties of the 0–20 cm depth measured at the initiation of experiments at the two sites were as follows:field capacity,25.1%–25.5%;wilting point,6.8%–6.9%;bulk density,1.249–1.251 g cm-3;organic matter,8.81–10.94 g kg-1;total nitrogen,0.69–0.81 g kg-1;available phosphorus,6.94–8.72 mg kg-1;available potassium,115.5–168.2 mg kg-1;pH,8.1–8.3 (Table S1).

2.2.Experimental design

Three treatments were applied in all experiments:(1) straw strip mulch (SM),which is a partial-cover mulching using whole cornstalks in straw mulching strips that alternate with planting strips without mulch;(2) whole-field plastic mulch (PM);and (3)no mulch with conventional planting as the control(CK).Diagrams of the treatments are presented in Fig.1.

Fig.1.Schematic diagrams of the field layouts at Shanghe and Pinxiang*.(a)Straw strip mulch:at the Shanghe site in 2012–2015,the mulching strip and planting strip were each 30 cm width with 50%area mulched and row spacing of 13 cm in the planting strip;(b)Straw strip mulch:at the Pinxiang site in 2015–2018,the mulching and planting strips were respectively 50 and 35 cm wide,and with 59% of area mulched and row spacing of 15 cm in the planting strip;(c) Whole-field plastic mulch (PM) with hole sowing at both sites in 2012–2018;(d) No mulch with sowing by drill (CK) at the two sites in 2012–2018,as a control;*,each treatment was repeated three times in a randomized block design.The sowing rate (including the area mulched in SM) and the fertilizer rate were the same in all treatments.

From 2012 to 2018,each treatment was replicated three times in a randomized block design,with each plot area ≥50 m2.All treatments in each experiment employed the same sowing rate(375 × 104plants ha-1),fertilizer application rate,sowing date,and field-management measures.The sowing rate in SM was calculated according to the total area of mulching strips and planting strips.Chemical fertilizers were applied by rotary tillage before sowing.Pesticides were used to control weeds,diseases,and pests during the growth period.In the six growing seasons,weeds,diseases,and insect pests did not significantly affect the growth or yield of winter wheat in any treatment.

Experiments were conducted at Shanghe from 2012 to 2015.The winter wheat cultivar was Lantian 26.The mulching strip and the planting strip in SM were each 30 cm in width,and 50%of the area was mulched.Three rows of wheat were planted in each planting strip by hole sowing(bunch planting),with row spacing of 13 cm.To prevent cornstalks from pressing on wheat seedlings,a 2-cm distance was maintained between the edge of a mulching strip and the edge row of a planting strip.When the main stem of winter wheat reached three leaves before the overwintering stage,whole cornstalks were placed as mulch.The mulching amount of straw was 9000 kg ha-1(approximately equal to cornstalk production ha-1in drylands).In PM plots,plastic film(0.008 to 0.01-mm thickness,120-cm width)was placed 2–3 days before bunch planting.The sowing row spacing in PM was 20 cm.PM and SM were seeded with the same seed number per hole and hole spacing(bunch spacing),with 10 seeds sown per hole (diameter 3–5 cm) and 12 cm between holes.In CK,drill sowing was used with row spacing of 20 cm.Fertilizers were applied in all treatments as 120 kg ha-1pure N and 90 kg ha-1P2O5.

Experiments were conducted at Pinxiang from 2015 to 2018.The winter wheat cultivar was Longzhong 2.For easier straw mulching and conditions more suitable for mechanical operation,the design of SM was adjusted.The mulching strip was 50 cm in width,the planting strip was 35 cm in width,and 59% of the area was mulched.Three rows of wheat were drill-sown in each planting strip,with row spacing of 15 cm.A 2.5-cm distance was maintained between the edge of a mulching strip and the edge row of a planting strip.The mulching time and the planting method in PM and CK were the same as at Shanghe.Fertilizers were applied before sowing in all treatments as 150 kg ha-1pure N and 120 kg ha-1P2O5.

2.3.Measurements and methods

2.3.1.Determination of soil moisture

Soil water content(SWC,%)at 0–200 cm was measured at eight growth stages (sowing,overwintering,regreening,jointing,booting,flowering,milking,and maturity).At each stage,soil water content was measured in eight soil layers (0–20,20–40,40–60,60–90,90–120,120–150,150–180,and 180–200 cm).In each plot,soil samples were collected using a 5-cm diameter steel-core auger from three random sampling points between two planting rows.Soil from each sample was added to an aluminum box,and the aluminum box and wet soil were weighed as W1on an electronic balance.The soil was then dried at 105°C for 48 h to constant weight.The weight of oven-dried soil and the box (W2) and the weight of the box (W0) were used to calculate soil water content as follows[36]:

Soil water content (SWC,%)=soil water weight/dry soil weight × 100%=(W1-W2)/ (W2-W0) × 100%

The mean SWC at 0–200 cm was calculated as the weighted mean of the eight soil layers.

2.3.2.Soil water storage

Soil water storage(SW,mm)at 0–200 cm was calculated as follows [41]:

where hiis the soil depth (mm),ρiis the soil bulk density (g cm-3),ωiis the soil water content (%),n is the number of soil layers,and i indexes the 0–20,20–40,40–60...180–200 cm soil depths.

2.3.3.Soil water consumption

Soil water consumption is also known as evapotranspiration(ET,mm),which was calculated as follows [59]:

ET=Pw+ΔW

where ET is water consumption (mm),Pwis total precipitation(mm) during the growth period of winter wheat,and ΔW (mm)is the difference in soil water storage at 0–200 cm between sowing and mature stage.

The field experiments were conducted under rainfed and nonirrigated conditions.Groundwater upward supplement and rainfall losses from surface runoff and leakage below 200 cm were excluded from the water balance equation[60]for several reasons.First,the groundwater table was more than 100 m below the surface at the two sites,too deep for upward flow to reach the root zone [61].Thus,the groundwater amount entering the root zone was considered to be negligible.Second,the experimental field was located in flat dryland at both sites,and there were few rainstorms or waterlogging events during the winter wheat growing seasons in the six years.In addition,seeding and seepage holes in the plastic film in PM plots allowed precipitation to infiltrate into soil,whereas in SM plots,the straw mulching strips intercepted runoff.For these reasons,surface runoff and leakage to deep soil were considered to be insignificant in the three treatments,with nearly all rainfall retained in the field.

2.3.4.Water use efficiency (WUE)

Two types of WUE(kg ha-1mm-1),WUEg(based on grain yield)and WUEb(based on biomass),were calculated using the following equations:

WUEg=G/ET

WUEb=B/ET

where ET is total evapotranspiration(mm)from sowing to maturation;G is grain yield (kg ha-1);and B is biomass (kg ha-1).

2.3.5.Measurement of soil temperature

Soil temperature at 5 cm was measured at six growth stages(overwintering,regreening,jointing,flowering,milking,and maturity).Mercury-in-glass geothermometers with a bent stem were inserted to 5 cm depth between the planted rows in each plot.Soil temperatures were recorded at 07:00,14:00,and 19:00 hours on typical sunny days and at a fixed position at all growth stages.Daily mean soil temperature was calculated as the mean of the three intraday records.

2.3.6.Measurement of agronomic performance

At maturity,the spike number ha-1was determined by sampling survey in 3 m2per plot.Plants,20 to 30,were randomly sampled from each plot to measure agronomic performance,including kernel number per spike and thousand-kernel weight.All plants in each plot were harvested to determine grain yield ha-1.The water contents of grain and plants in harvested samples were measured by oven drying at 105 °C for 30 min and then 85 °C for 48 h.The grain yield ha-1,biomass ha-1(including stems,leaves,and spikes),and thousand-kernel weight were then adjusted to 13%moisture content.

Harvest index (HI) was calculated as follows:

HI (%)=G/B × 100%

where G (kg ha-1) is grain yield and B (kg ha-1) is biomass.

2.3.7.Economic benefit calculation

Economic benefit analysis was based on surveys of the current prices of winter wheat at the local markets and agricultural material outlets.Net income (CNY Yuan ha-1) was calculated by subtracting the inputs from the outputs.

The output value was calculated using a price of CNY 2.4 Yuan kg-1of winter wheat grain over years.The inputs included seed cost,fertilizer cost,pesticide cost and the total cost of labor,soil tillage and harvest.For PM,the input cost for purchasing plastic film was added.For SM,waste cornstalks were excluded from the input cost.

2.4.Statistical analyses

Statistical analyses were conducted with the SPSS statistical software package (version 20.0,IBM SPSS Inc.,Chicago,IL,USA).In each experiment,means were compared by one-way analysis of variance (ANOVA).Differences among experimental years and treatments and the interaction between year and treatment(T×Y)were tested by two-way ANOVA.The differences were evaluated by least significant difference (LSD).Pearson correlations with six years’ data were calculated with SPSS.Figures were constructed with OriginPro 2016 9.3 (Origin Lab,Northampton,MA,USA).

3.Results

3.1.Agronomic performance

3.1.1.Grain yield

Grain yield ha-1differed among treatments(P ＜0.01)and years(P ＜0.01),and the interaction T×Y was also significant(P ＜0.01).The mean yield in SM was similar to that in PM (P ＞0.05).Compared with CK,the mean yields of SM and PM increased by 27.0%(912 kg ha-1) and 21.7% (732 kg ha-1in 2012–2018 (Table 1).

Year greatly affected yield (P ＜0.01).In 2017–2018,which had relatively high precipitation,there were no significant differences in yield among treatments.In the three seasons of 2012–2013,2015–2016,and 2016–2017,yields were low in all treatments.

From 2012 to 2015,SM accounted for 50% of the area mulched and hole sowing was used,whereas from 2015 to 2018,SM accounted for 59% of the area mulched and drill sowing was used.There was no difference in yield (P ＞0.05) between SM and PM in five of the growing seasons (with 2014–2015 the exception).

3.1.2.Yield components and vegetative growth

Based on the two yield formation pathways of yield components and growths,the main reasons for the yield increase in SM an PM were the enlargement of sink capacity (kernels ha-1) and the enhancement of vegetative growth (straw yield ha-1),while thousand-kernel weight and harvest index were relatively stable among treatments(Table 1).Compared with CK,on average across six growing seasons,SM and PM increased sink capacity by 26.6%and 19.0%,straw yield by 21.6% and 22.6%,respectively.The increases in sink capacity in SM depended primarily on the increase in spikes ha-1(17.1%),followed by the increase of kernels per spike (8.4%),while those in PM depended primarily on the increase of kernels per spike (14.9%).

Table 1 Grain yield and yield components of dryland winter wheat under two mulching treatments in northwest China.

Grain yield ha-1was positively correlated with spike number ha-1(r=0.61,P ＜0.01),kernel number per spike (r=0.61,P ＜0.01),kernel number ha-1(r=0.83,P ＜0.01),and straw yield ha-1(r=0.90,P ＜0.01).However,grain yield ha-1was relatively weakly correlated with thousand-kernel weight (r=0.43,P ＞0.05) and harvest index (r=0.49,P ＜0.05) (Fig.2).

3.2.Soil moisture

3.2.1.Soil water storage at 0–200 cm

From 2012 to 2018,mean soil water storage in SM and PM at 0–200 cm from overwintering to maturity was respectively 25 and 22 mm higher than that in CK (P ＜0.05).Mean soil water storage in SM and PM was similar(P ＞0.05)(Table 2).However,the differences in soil water storage among treatments varied with experimental years.In the four growing seasons from 2012 to 2016,mean soil water storage in SM and PM was respectively 34 mm(range,25–43 mm) and 28 mm (range,15–42 mm) higher than that in CK.In the two growing seasons from 2016 to 2018,no differences were found among treatments(P ＞0.05).Mean soil water storage in SM was higher than that in PM by 18.0 mm in 2012–2013 and by 16 mm in 2013–2014 (P ＜0.05).However,from 2014 to 2018,there were no differences between SM and PM(P ＞0.05).

Increases in soil water storage in SM and PM were greater in the early period (overwintering to jointing) than in the later period(booting to maturity).At overwintering,regreening,jointing,booting,flowering,milking,and maturing stages,mean soil water storage at 0–200 cm soil across years in SM was 37,36,33,19,15,16,and 21 mm higher,respectively,than that in CK,and in PM,it was respectively 39,44,29,17,10,8,and 7 mm higher.From overwintering to regreening,the increases in soil water storage in PM were slightly greater than those in SM,whereas at other stages,the increases in SM were greater than those in PM.From 2012 to 2018,the mean coefficients of variation (CV) of soil water storage across growth stages were respectively 15.3% in SM,17.1% in PM,and 13.5% in CK.

3.2.2.Differences in soil water content among soil layers

Soil water content was generally higher in SM and PM than in CK at all soil depths(soil layers)and growth stages(Fig.3).As with the differences in soil water storage at 0–200 cm among treatments,the increases in soil water content in SM and PM were also higher in the early growth period (overwintering to jointing) than those in the late period(booting to maturity),than in CK.Especially in the early period,increases in SM and PM at 0–120 cm were greater than those at 120–200 cm.In SM and PM,mean soil water content at 0–120 cm increased by 1.61% and 1.85%,respectively,compared with that in CK.However,in the later period,the increases tended to be similar in each soil layer.At 180–200 cm,the soil water content was relatively stable among growth stages in each treatment,but increases in SM and PM,compared with CK,were smaller than those in other soil layers.

Fig.2.Correlations among yield,yield components and soil hydrothermal parameters.GY,grain yield ha-1;SN,spike number ha-1;GN,kernel number per spike;TGW,thousand-kernel weight;TGN,total kernel number ha-1;BY,biomass ha-1;HI,harvest index(%);WUEg,water use efficiency based on grain yield and ET;WUEb,water use efficiency based on biomass and ET;ET,evapotranspiration from sowing to maturity;ΔW,change in soil water storage at 0–200 cm between sowing and maturity;ETom,evapotranspiration from overwintering to maturity;ΔWom,change in soil water storage at 0–200 cm between overwintering and maturity; Pw,total precipitation during whole growth period(sowing to maturity);Pfa,precipitation during February to April;SWA,mean soil water storage at 0–200 cm from overwintering to maturity;SWR0-60,soil water content at 0–60 cm at regreening stage;STa,mean soil temperature(°C)at 5 cm across growth stages(overwintering to maturity);STf,STm and STr represented soil temperature at 5 cm at flowering,milking,and ripening stage,respectively;Asterisks indicate significant correlation (*, P ＜0.05;**, P ＜0.01).

Compared with CK,both SM and PM increased the mean soil water content at 0–200 cm depth across years and growth stages,but the increases at the three soil layers of 0–20,20–40 and 90–120 cm were greater (by 1.05%–1.39% in SM,1.04%–1.56% in PM)than those in other soil layers (by 0.70%–0.99% in SM,0.13%–0.86% in PM).Generally,soil water content in SM and PM was higher than that in CK.However,sometimes the opposite was observed,depending on stage,soil depth,and year.Thus,compared with CK,both SM and PM showed both effects,with soil moisture either increasing or decreasing (Fig.3).

3.2.3.Water consumption and water use efficiency

There was no difference in mean water consumption (ET)during the whole growth period (sowing to maturity) among treatments from 2012 to 2018 (P ＞0.05) (Table 3).However,in 2014–2015,2015–2016,and 2017–1018,the ET in CK was higher than that in PM or SM;whereas in 2012–2013,2013–2014,and 2016–2017,the ET in CK was lower than that in PM or SM(P ＜0.05).The mean WUEgin SM and PM across six years was respectively 24.5% and 18.6% higher than that in CK (P ＜0.05).WUEgwas highly positively correlated with grain yield (r=0.88,P ＜ 0.01) but was not significantly correlated with ET.Thus,significant increases in grain yield were the main reason for increases in WUEgin SM and PM.However,ET was positively correlated with grain yield (r=0.61,P ＜0.01),straw yield (r=0.59,P ＜0.01),and biomass (r=0.62,P ＜0.01) (Fig.2).

Precipitation during the growth period (Pw) (sowing to maturity) accounted for 65%–67% of total water consumption (ET),whereas precipitation from overwintering to maturity (Pom)accounted for 61%–70% of water consumption (ETom) (Table 3).Grain yield was positively correlated with Pw(r=0.52,P ＜0.05).In particular,precipitation from regreening to booting (Pfa,February to April) was highly positively correlated with grain yield(r=0.77,P ＜0.01) and straw yield (r=0.66,P ＜0.01) (Fig.2).

In SM and PM,water consumption after overwintering (ETom)increased,but before overwintering,it decreased.The ETom/ET ratio was 82%in SM,86%in PM,and 78%in CK.From overwintering to maturity,mean water consumption(ETom)in SM and PM was 16 and 32 mm higher,respectively,than that in CK.In SM and PM,there was more consumption to the pre-overwintering soil water storage(ΔWom)at 0–200 cm than that in CK.The ΔWom/ETomratio in SM and PM was respectively 4.8% and 8.6% higher than that in CK (Table 3).

3.3.Soil temperature

After the overwintering stage,with advancing growth stages and increases in air temperature,soil temperatures at 5 cm andthe differences of soil temperature among treatments gradually increased (Fig.4).For mean soil temperatures over six years(Table 2),PM had a slight warming effect of 0.2 °C compared with that in CK(P ＞0.05),whereas SM had a significant cooling effect of-1.5 °C (P ＜0.05).Still,compared with CK,both SM and PM showed effects of warming and cooling,depending on the growth stages(Fig.4).In SM,warming(P ＜0.05),cooling(P ＜0.05),and no difference (P ＞0.05) accounted for 8%,56%,and 36%,respectively,of the 36 total measurements.In PM,warming accounted for 25%,cooling for 19%,and no difference for 56%of the 36 measurements.The warming effect in PM was greater than that in SM,whereas the cooling effect in SM was greater than that in PM.The cooling effect occurred mainly after the jointing stage.At regreening,jointing,flowering,milking,and maturing stages,the mean soil temperatures in SM across six years were respectively 0.90,1.97,2.72,1.92,and 2.34°C lower,than those in CK.However,at the overwintering stage,SM also had a warming effect of 0.26–1.0°C(2012–2015)or no difference compared with CK.In PM,soil temperature increased by respectively 0.64,0.80,and 0.17 °C at overwintering,regreening,and maturing stages,compared with that in CK,but decreased by 1.16,1.24,and 0.49 °C at jointing,flowering,and milking stages.

Table 2 Soil water storage (mm) at 0–200 cm in six stages (overwintering to maturing) of winter wheat in northwest China.

Grain yield was negatively correlated with soil temperature at 5 cm from flowering to maturity (r=-0.50 (P ＜0.05) to -0.78(P ＜0.01)),and thousand-kernel weight was also negatively correlated with the mean temperature across stages(r=-0.48,P ＜0.05)(Fig.2).In SM,the proportion of mulched area and the sowing method had only weak effects on soil temperature.From 2012 to 2015,SM with 50% of the area mulched and hole sowing reduced the soil temperature at 5 cm by 1.68 °C,compared with CK,whereas from 2015 to 2018,SM with 59% of the area mulched and drill sowing reduced it by 1.66 °C.

Fig.3.Soil water content (%) at soil depths and growth stages of winter wheat at Shanghe site (2012–2015) and Pinxiang site (2015–2018).SM,straw strip mulch;PM,whole-field plastic mulch;CK,no mulch with wheat sown in rows;12–13 refers to the growing season of 2012–2013,and so on;bar indicates LSD at P ＜0.05;no bar means no significant difference (P ＞0.05).

3.4.Economic benefit

From 2012 to 2018,the mean net income in SM was CNY 2381 and 1295 Yuan ha-1higher than that in PM and CK,respectively(Table 4).From 2012 to 2014,the net income in PM was higher than that in CK but lower than that in CK in the other four years.Low net income in PM was due primarily to higher input costs.The mean output/input ratio was 1.50 in PM,2.16 in CK,and 2.22 in SM.From 2012 to 2017,SM yielded the highest net income.

Table 3 Evapotranspiration and water use efficiency (kg ha-1 mm-1) under three mulching treatments in northwest China.

Table 4 Comparison of inputs,outputs,and net income (CNY Yuan ha-1) among mulching treatments in dryland winter wheat in northwest China.

4.Discussion

4.1.Effect of soil temperature on yield formation

In SM and PM,the soil hydrothermal environment was suitable for producing high yields of winter wheat.Soils were warmer at overwintering stage and cooler from jointing to maturity in both SM and PM than in CK.Soil temperature affects growth and development of crops [62,63].However,the suitable soil temperature range is highly variable in different growth phases of crops[7,64].At overwintering stage,the warmer soil can reduce overwintering mortality and maintain continuous wintertime root growth in winter wheat[20,65].Other studies[50,54,66]have also shown that,in the early growth season with lower atmospheric temperature or in regions with heat deficiency,the higher soil temperature maintained with mulching can promote vegetative growth of crops.However,in middle and late growth periods with higher atmospheric temperature,lowering the soil temperature likely helped to reduce the intensity of ET,luxury ET,and the damage to wheat caused by high temperature,resulting in higher yields[40,51,67].

Compared with no mulch,transparent plastic mulch generally has a warming effect[53,68],whereas the effect with straw mulching is the opposite [19–20,62].However,depending on growth stages and soil depths,both straw mulching and plastic mulching have effects of warming or cooling,compared with no mulch[39,69–72].Similarly,in this study,depending on growth stages and soil depths,both PM and SM not only caused warming and cooling(Fig.4)but also increased or reduced soil moisture(Fig.3).The influences of mulching materials and methods on soil temperature depend on the absorption of solar radiant heat by soil and on hydrothermal exchanges between soil and atmosphere[73–76],as well as on the degree of shading under larger canopies with mulch[77,78].

Fig.4.Differences in soil temperature at 5 cm among treatments at six growth stages of winter wheat at two sites.SM,straw strip mulch;PM,whole-field plastic mulch;CK,no mulch with wheat sown in rows.Bar indicates LSD at P ＜0.05;no bar means no difference (P ＞0.05).

4.2.Mechanisms of soil moisture difference between treatments and hydrothermal interaction

Soil moisture in drylands is determined by ET and precipitation during the growth period [79].In this study,the mean soil moisture and ET across stages and years were similar between SM and PM.However,in SM,only 50%or 59%of the area was mulched,and there were also gaps among cornstalks in mulching strips.For this reason,evaporation(E)inhibition under SM was certainly less than that under PM.Soil moisture and ET in SM were similar to those in PM,possibly owing to cooler temperatures in SM.The reasons are as follows:the mean soil temperature in SM was lower(1.7 °C) than that in PM,and the cooler condition in SM likely reduced ET intensity[77,80].As a result,luxury or excessive water consumption decreased in SM.In addition,with plant growth,shading by the canopy can reduce E[66,77,78].In the later growth period,transpiration (T) gradually dominated water consumption,and especially after flowering,the T/ET ratio was ＞90%[81,82],and accordingly,the difference between mulching treatments in inhibiting E may be greatly narrowed or weakened in the later growth period [49,83].

Compared with CK,mulching increased soil moisture primarily in the early growth stages,but in middle and late stages,soil moistures at some soil depths were slightly lower in SM and PM than in CK or not different.Other studies found similar results[15,40,43,49,67,84].There are two possible explanations for these results.First,in middle and late stages,T in mulch treatments with a larger canopy might have been higher than that in CK [85–87].Second,because of soil resistance to water conduction,some time is needed for soil water to move and equilibrate between high and low water-potential zones [31,88–91].

4.3.Relationship between yield formation and water consumption

Comparisons of yield components among treatments and correlation analyses indicated that enlarged sink capacity (kernel number ha-1) and promotion of vegetative growth (straw yield ha-1)were the main reasons for grain yield increase in SM and PM,whereas differences in thousand-kernel weight and harvest index were relatively small among treatments.The increase in spikes ha-1was larger in SM,whereas the increase in kernels per spike was larger in PM.Similar studies also demonstrate that increases in grain yield under mulching practices are primarily due to increases in effective spikes per unit area [3,40,92] or kernels per spike [40,93].Stronger vegetative growth under mulch increases the area and net accumulation of photosynthesis and thereby increases the supply of photosynthetic product for forming more spikes and kernels [14,45,67,94].

The yields and the formation of yield components were dependent not only on water consumption(ET),but even more on water consumption structures.ET during the full growth period was positively correlated with grain yield and straw yield.Other studies[27–28,38,40,95] also found that the high yield of dryland crops is based on high ET.However,the effects of mulching on ET depend on crop types,years,and regional environmental differences.Sun et al.[10]found that plastic film mulch did not affect ET in regions with annual rainfall ＜400 mm and led only to higher WUE.Similarly,Fan et al.[39] reported that although soil water content at 0–200 cm during the maize growing season was always higher with full-cover plastic film mulch than with partial-cover mulch(with 50% or 66% of the area mulched),ET was not different.

Although increases in ET lead to increases in yield of dryland wheat,the yield increases in SM and PM were due primarily to decreases in E and increases in T and T/ET ratio in favor of production[5–8].Compared with CK,mean grain yield increased by 27.0%in SM and by 21.7% in PM over the six years,but there was little increase (6–9 mm) in mean ET in either treatment.Even in 2015–2016,when SM increased grain yield by 69.1% compared with that in CK,ET decreased by 26 mm(7.6%).Based on the mean ET 9 mm higher in PM than in CK and mean WUEgin PM,the calculated increase in grain yield in PM should be 108 kg ha-1(3.2%)greater than that in CK,whereas the measured increase was 732 kg ha-1(21.7%).Thus,the increase in ET in PM was not sufficient to produce such a high increase in grain yield.

Because PM is a type of full-cover plastic film mulching,it is an ideal approach for studying T and the transpiration coefficient(TC).Although PM has a few open sowing holes without mulch,they account for only approximately 5% of field area.After overwintering,the sowing holes were almost completely covered by plants,and accordingly,PM could be considered as almost completely closed mulch.Thus,in PM,ET was equal to T,and WUEbwas equal to TC.In our recent study [96] of water consumption characteristics of winter wheat in a dryland environment,the E/ET and T/ET in CK were 34.8% and 65.2%,respectively,whereas those in SM were 25.2% and 74.8%.According to the above E/ET ratio,it can be estimated in this study that E,T,TC and the grain production efficiency by transpiration (TCg) were 115 mm,214 mm,41.2 kg ha-1mm-1and 15.8 kg ha-1mm-1in CK,respectively,whereas in SM,the estimates were 84 mm,251 mm,40.1 kg ha-1mm-1,and 17.1 kg ha-1mm-1,respectively.Thus,PM reduced water loss from E by at least 115 mm and 84 mm compared with those in CK and SM,respectively.Under different environments,the E/ET ratio of dryland wheat without mulching generally ranges from 30%to 40%[22,97–103].Wang et al.[81]used the SIMDualKc model to estimate that winter wheat E/ET ratios range from 23.7%to 29.6% during the full growth period.

In PM,the mean TC and TCgwere 12.1 kg ha-1mm-1(equal to WUEb) and 29.2 kg ha-1mm-1(equal to WUEg) (Table 3,respectively.Relative to PM,TC and TCgwere 41.1% and 30.6% higher in CK,37.3%and 41.3%higher in SM,respectively.These results indicate that PM supported more nonproductive luxury T.Why TC and TCgin CK and SM were higher than those in PM is uncertain,but may be due to differences in stomatal diffusion resistance among treatments [104–105].The reasons are as follows:theoretically,many factors affect T,TC,and TCg.Only approximately 1% of the water absorbed by roots is used in photosynthetic production,whereas approximately 99% is released from plants via T.Water is lost by T in several pathways,including stomatal T,cuticular T,and guttation,among others,of which stomatal T is the most important[22,104].In several studies[106–108],the ratio of stomatal T to nonstomatal T differed between plastic film mulching and no mulching.The TC also varies with cultivation condition or regional environment [109–110].Although TC and TCgin CK and SM were higher than those in PM,WUEgand WUEbin PM were still higher than those in CK by 18.6%and 19.2%,respectively,and similar to those in SM.

In addition to increases in T and T/ET,yield increases and efficient water use in SM and PM were associated with phase differences in water consumption.In SM and PM,water consumption decreased before overwintering but increased after overwintering.Formation of yield components in winter wheat occurred after overwintering,so that the increase in ETomwas beneficial in promoting vegetative growth and improving grain yield.Similarly,Li et al.[96] found that soil water consumption decreased before jointing stage in PM and SM but increased in the middle growth period (jointing to flowering),compared with CK.

Water consumption of dryland wheat in this study depended primarily on precipitation in the growth period.A similar conclusion was reached in a study of dryland crops[60].But the contributions of pre-sowing or pre-overwintering soil water storage at 0–200 cm to wheat production cannot be ignored.In SM and PM,the use of pre-sowing and the pre-overwintering stored soil water increased,and may alleviate seasonal drought caused by insufficient precipitation in the early growing period (November to April).In addition,the mean soil moisture in SM and PM at 0–200 cm from overwintering to maturity was higher than that in CK.Especially in the early growth period (overwintering to jointing),soil moisture at 0–120 cm was higher in SM and PM than in CK,assuring that water was available to promote vegetative growth and form more spikes ha-1.

4.4.Benefits of straw strip mulch system

Economic return and simplicity of operation are the main considerations of farmers when choosing a mulch pattern.Costs of labor and mechanical energy consumption and other inputs(seeds,fertilizers,pesticides) were almost the same between SM and PM.However,in PM,there was the additional input cost of plastic.Consequently,net income was higher in SM than in PM.Zhao et al.[3]also reported that net income was higher with straw mulch than with plastic mulch.

In this study,to ensure that all precipitation passed through the plastic into soil,seepage holes were punched through the plastic where water accumulated after each precipitation event.However,when plastic film mulching is applied over a large area,this practice is difficult to implement because of the high labor cost.Thus,with PM,there is no guarantee that all precipitation will permeate into soil through only a few holes not covered by mulch[111,112].By contrast,with SM,all precipitation permeates easily into soil.

In the dryland farming regions of northwest China,almost all corn is planted using full-cover mulching with polyethylene plastic film[39,41,43].If cornstalks are chopped and returned to the field during harvest,the chopped corn straw will inevitably cover the plastic film and make its removal difficult.For this reason,the ears are picked first in most of the field with plastic film mulching,the cornstalks are harvested and removed from the field,and then the residual film is removed.As a result,cornstalks cannot be returned to the field in dryland regions,and over 70%of them are discarded or burned (with some 20–30% used as forage) [14,113],increasing environmental pollution.SM provides a new way to recycle cornstalks for dryland farming ecosystems.Furthermore,because only approximately 9000 kg ha-1of cornstalks are used in SM,it is possible to expand the area with straw mulching.

Under the same planting density,whether the mulched area was 50% or 59% and whether seeds were hole or drill sown,the grain yields in SM were similar to or higher than those in PM.Thus,the sowing method and the proportion of mulched area in the range of 50%–59% did not affect the increase in grain yield in SM.However,operations are more convenient in SM with drill sowing and mulched area at 59% (mulching width 50 cm).

5.Conclusions

In a semiarid rainfed agroecosystems,compared with unmulched planting (CK) of winter wheat,straw strip mulching (SM)and whole-field plastic mulching (PM) increases grain yield ha-1,sink capacity(kernels ha-1),vegetable growth,and soil water supply capacity.The increases in sink capacity and final grain yield ha-1in SM were determined primarily by the increase in spikes ha-1,while those in PM were determined primarily by the increase of kernels per spike.SM and PM caused warming during overwintering and cooling after jointing.The higher soil temperature of the early growth season under lower atmospheric temperature helped to maintain or promote wheat growth of SM and PM,while the lower soil temperature under the higher atmospheric temperature contributed to reduce the intensity of evapotranspiration (ET) and luxury ET.Averaged over six years,SM and PM showed water consumption (ET) during the whole growth period similar to that of CK.However,SM and PM increased water consumption during the key formation period of yield components after overwintering while reducing it before overwintering,and increased transpiration(T) and T/ET while inhibiting soil evaporation (E).SM and PM promoted vegetative growth and increased sink capacity by providing favorable soil hydrothermal environments for several growth stages and optimizing water consumption,the main mechanisms resulting in grain yield increase in SM and PM relative to that in CK.

SM not only achieved similar or higher soil moisture and grain yield than PM,but also resulted in higher net income than PM or CK.SM has the advantages of less amounts of straw applied for mulch than conventional full-cover straw mulching and convenient operation of machinery.Thus,SM is an environmentfriendly mulch cropping system suitable for winter wheat production in semiarid rainfed agroecosystems as an alternative to PM.

CrediT authorship contribution statement

Yuwei Chai:Writing-original draft,Investigation,Formal analysis.Qiang Chai:Supervision.Rui Li:Data curation of soil moisture.Yawei Li:Data curation of soil temperature.Changgang Yang:Investigation.Hongbo Cheng:Determinations of agronomic performances.Lei Chang:Visualization.Shouxi Chai:Conceptualization,Methodology,Writing -review &editing,Validation.

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 State Key Laboratory of Aridland Crop Science,Gansu Agricultural University (GSCS-2019-Z05);and by China Agricultural Research System of the Ministry of Finance (MOF) and Ministry of Agriculture and Rural Affairs(MARA) (CARS-3-2-47).We gratefully acknowledge the anonymous reviewers and editors for their constructive comments.

Appendix A.Supplementary data

Supplementary data for this article can be found online at https://doi.org/10.1016/j.cj.2021.09.004.