Mixing trait-based corn(Zea mays L.)cultivars increases yield through pollination synchronization and increased cross-fertilization

Hongping Li,Kui Liu,Zhibin Li,Moubio Zhng,Yongen Zhng,Shuyn Li,Xiuling Wng,Jinlong Zhou,Yli Zho,Tinxue Liu,Chohi Li,*

a Agronomy College,Henan Agricultural University,Collaborative Innovation Center of Henan Grain Crops/Co-construction State Key Laboratory of Wheat and Maize Crop Science,Zhengzhou 450046,Henan,China

b Swift Current Research and Development Centre,Swift Current,Saskatchewan S9H 3X2,Canada

c Agricultural Information Institute,Chinese Academy of Agricultural Science,Beijing 100081,China

d Henan Institute of Meteorological Sciences/CMA·Henan Key Laboratory of Agrometeorological Support and Applied Technique,Zhengzhou 450046,Henan,China

Keywords:Cultivar heterogeneity Cultivar mixture Flowering trait synchronization Fertilization complementarity Kernel set

ABSTRACT Abiotic stress such as high temperature at flowering is one of many conditions reducing yield of corn(Zea mays L.).Mixing corn cultivars with diverse functional traits increases within-crop diversity and provides a potential means of mitigating yield losses under stress conditions.We conducted a three-year field study to investigate the effects of cultivar mixtures on kernel setting rate,pollen sources,and yield.This study consisted of six treatments,including two high temperature-tolerant(HTT)monocrops of WK702 and DH701,two high temperature-sensitive(HTS)monocrops of DH605 and DH662,and two HTT–HTS mixtures of WK702-DH605 and DH701-DH662.The anthesis–silking interval(ASI)was 0.9–1.6 days shorter in mixtures than in monocrops.Kernel setting rate was increased in mixtures(86.4%–88.7%)compared with those in monocrops(74.7%–84.1%)as a result of synchrony and complementarity of pollination.Grain yields of the HTT–HTS mixtures increased by 13.3%–18.7%,equivalent to 1169 to 1605 kg ha-1,in comparison with HTS corn monocrops.The results of SSR markers showed that crossfertilization percentage in corn cultivar mixtures ranged from 29.3% to 47.8%,partially explaining yield improvement.Land equivalent ratio(LER)was 1.12 for corn mixtures and the partial land equivalent ratio(e.g.,＞0.5)showed the complementary benefits in corn mixtures.The results indicated that mixing corn cultivars with diverse flowering and drought-tolerance traits increased yields via pollination synchrony.

1.Introduction

Agricultural production is facing unprecedented challenges in a changing climate[1,2],which may increase the frequency of extreme weather such as heat and drought[3,4].High temperature-induced heat and drought stress is a major concern for yield stability[5,6],particularly when the stress conditions occur at the crop flowering stage.For corn crops,an estimated 11.7% of yield deficit(collected from the Emergency Events Database,EM-DAT,https://www.emdat.be/database)was attributed largely to the effects of extreme heat[2].

The Huang-Huai-Hai Plain,one of the six main corn-producing regions in China,has an annual corn seeded acreage of about 15 million hectares and accounts for approximately 29.0% of national corn production(National Bureau of Statistics of the People’s Republic of China,2014,https://data.stats.gov.cn/).In the Huang-Huai-Hai Plain region,high temperature often occurs from late July to early August at corn flowering and causes yield losses[7].High temperature stress during the stages of pollen shed,silk receptivity,pollen–stigma interaction,or early kernel formation impaired reproductive development[8].Pollen morphology was severely damaged and pollen grains shriveled and collapsed when exposed to 36/26 °C treatment for 14 days[9].High-temperature stress advanced tasseling and pollen shedding,shortened the duration of pollen shedding,extended the ASI,and reduced tassel branching and floret numbers,but had no noticeable effects on silking[9,10].Shim and Lee[11]reported that temperature elevation reduces differentiated ovule number.The stages most sensitive to heat stress appeared to be during pollen–stigma interaction and early kernel development in corn,even for high temperature-tolerant genotypes[8].

To mitigate possible high-temperature stress,practices such as adjusting planting dates,breeding heat-tolerant cultivars,increasing irrigation,and using agrichemicals have been recommended[12].However,adjusting planting dates is not feasible in dominant wheat–corn double-cropping systems,where summer corn is seeded immediately after winter wheat harvest in early summer.In addition,applying agrichemicals increases input costs and raises concerns of environmental contamination.There is thus an urgent need to develop alternative strategies for mitigating the frequently occurring high-temperature stresses in corn production.

Intercropping or cultivar mixing has been shown to reduce the risk of lowered yields caused by biotic and abiotic stresses[13–15].Intercropping is a cultural practice of growing two or more species or genotypes simultaneously on the same land.A cultivar mixture is one special type of intercropping using the same crop species with differing functional traits.Cultivar mixtures are considered a sustainable crop intensification practice,on the basis of the increased biodiversity[16–18].A large body of literature has shown that cultivar mixtures provide ecological services such as pest suppression and resource use efficiency improvement,reducing agrochemical inputs and environmental impacts[19].Cultivar mixtures act as insurance to stabilize crop performance in a changing climate[13,20].A literature review indicated that the benefits of cultivar mixtures were further strengthened under stressful conditions by buffering against unpredictable stresses[15].The finding that corn cultivar mixtures led to greater yield improvement than a cereal cultivar mixture[15]suggests that cultivation of corn cultivar mixtures is a viable crop management practice for improving food security,especially in stress environments.

Corn yield is determined partially by the number of kernels,which is controlled mainly by the synchronization of anthesis and silking[21,22].Temperature is one of the critical factors governing the timing of anthesis and silking in corn,and developing reproductive organs are susceptible to heat stress[23].Delayed silking and pollination caused by high temperature was one of the major factors causing kernel abortion and lower yield[23,24].Modern corn breeding has developed advanced cultivars with specific genetic traits to tolerate drought and pests.A heattolerant corn cultivar was less likely to be affected by high temperature,whereas heat-sensitive cultivars often showed reduced kernel set owing to a prolonged ASI or the abortion of late-fertilized kernels as a result of drought stress[23,25,26].A shorter ASI is often used as a selection criteria for breeders to improve drought tolerance of corn[27,28].It is expected that an increased kernel setting rate might be achieved by shortening ASI by mixing cultivars with differing phenological and temperature sensitive traits.Constructing a composite population using the genetic heterogeneity of corn cultivars by cultivar mixing may be an effective pathway towards mitigating biotic and abiotic stresses faced by corn producers.

Selecting cultivars with differing functional traits is the key to achieving the benefits of cultivar mixtures and adapting to unfavorable growth environments.For instance,mixing corn cultivars with compatible flowering phenological traits may be effective in reducing ASI by ensuring a close match between anthesis and silking at the population level.Furthermore,mixing corn cultivars with complementary flowering and heat-resistance traits is expected to generate synergistic effects on pollination,kernel set,and yield.We hypothesized that the yield of corn would be increased in cultivar mixtures compared with corn monocrops through close synchrony between anthesis and silking.Our objectives were to 1)investigate the responses of ASI and pollenshedding duration in corn cultivar mixtures compared with corn monocrops,2)quantify cross-fertilization in cultivar mixtures,and 3)assess effects of cultivar mixtures on yield and land equivalent ratio.

2.Materials and methods

2.1.Site description

A three-year(2016–2018)field study was conducted at the experimental station(33°18′N,114°0′E)of Henan Agricultural University,Xiping county,Henan province,China.The study site is located in a transition zone from subtropical to a warm temperate climate,with a long-term(1970–2018)average annual precipitation of 852 mm,a frost-free period of 221 days and an average temperature of 14.8 °C.Air temperature during the 1970–2015 period was obtained from the on-site weather station,while during the experimental period,2016–2018,temperature was recorded in the corn field at a height of 1.3 m above the ground.

Corn was seeded on June 6,June 7,and June 7 in 2016,2017,and 2018,respectively;and harvested on September 30,September 29,and September 30 in 2016,2017,and 2018,respectively.The soil is classified as lime concretion black soil,with 16.2%sand,20.0%silt,and 42.1%clay.The baseline soil fertility in the 0–30 cm soil layer is as follows:pH 5.4(soil:water 1:2.5),organic carbon 11.8 g C kg-1,total nitrogen 1.0 g N kg-1,Olsen phosphorus 20.1 mg P kg-1,and exchangeable potassium 94.7 mg K kg-1.Fertilizers were side dressed to a 10 cm depth at seeding at rates of 180.0 kg N ha-1,40.0 kg P ha-1,and 74.0 kg K ha-1based on soil test recommendations.At V6 of corn,when the sixth leaf collar of 50% or more of the plants in the field is visible,an additional 75 kg N ha-1was side dressed.

2.2.Experimental design

2.2.1.Cultivar selection

Four corn cultivars,WK702,DH605,DH701,and DH662,were selected for this study.Among them,WK702 and DH701 are high temperature-tolerant(HTT)cultivars while DH605 and DH662 are high temperature-sensitive(HTS)cultivars.The selection of two pairs of cultivars with different high temperature-tolerance characteristics was based on the results from our previous threeyear(2013–2015)study,which compared the effects of high temperature during flowering on kernel set and yield.We found that under the influence of high temperature,DH605 and DH662 had lower kernel set and grain yield,whereas WK702 and DH701 were hardly affected.The literature also showed that DH605 and WK702 had different tolerances to high temperature and drought[29–31].The growth duration of all the four corn cultivars ranges from 100 to 104 days,and plant height ranges from 259 to 272 cm.

The passport information of cultivars was as follows:DH662,WK702 and DH605 were certified by the China National Cultivar Approval Committee in the years 2009,2012,and 2010,respectively,and DH701 was released by the Cultivar Certification Committee of Shandong Province in 2009.

2.2.2.Experimental design and field management

The experiment was arranged in a randomized complete block design with four blocks.There were six treatments:four corn monocrops,including WK702,DH605,DH701,and DH662;and two cultivar mixtures.Each mixture consisted of two cultivars,with one component cultivar sensitive to high temperature and the other tolerant to high temperature based on regional evaluation trials.Based on the flowering traits and drought resistance of the four cultivars,two mixtures were formed:1)WK702 and DH605 mixture,and 2)DH701 and DH662 mixture.For each corn mixture,the two corn cultivars were seeded in a mixed row as shown in Fig.1 to ensure cultivar heterogeneity per unit of area.Each plot was 20 m long×22 m wide and seeded in 36 rows.The row spacing was 60 cm.The target plant density was 75,000 plant ha-1for all treatments,following locally recommended practices.To achieve high emergence,all plots were irrigated with about 30 mm water immediately after planting.Local best management practices were followed for pest control.

2.3.Plant sampling and analysis

2.3.1.Flowering phenology

At the V6(when the collar of the sixth leaf is visible)stage,approximately 75 representative corn plants in each plot were marked for plant observation and sample collection.The initiation of anthesis or silking was recorded when at least one extruded anther or silk became visible.The dates of 50%anthesis and silking of corn in each plot were recorded for 20 plants in each monocrop treatment and 40 plants in each mixture treatment.Anthesis was monitored and recorded daily until the last day of pollen shedding for each cultivar in monocrops and mixtures.The ASI for each plot was calculated as the difference in days between 50%anthesis and 50% silking.

2.3.2.Pollen germination measurements

Each year,20 tagged plants from each monocrop plot and 40 from each mixture plot(with 20 for each component cultivar)were selected to determine pollen germination rate.Tassels of tagged plants were gently tapped to collect fresh pollen into a new bag at 10:00 AM each day during the corn flowering period.The collected fresh pollen was poured immediately onto the agar germination media in petri dish[32].The in vitro culture was conducted in a laboratory at ambient room temperature of about 27°C and 60%relative humidity.The agar media in each petri dish was divided into four sections for a separate germination count.After 2 h of culture,germination counting was performed under a microscope,with the fields of view chosen randomly within the individual sections.Germinated pollen grains were defined as those with intact pollen tube at least one grain diameter in length[32].Bursting pollen was defined as those grains showing an irregular mass of cytoplasm and starch grains extruded from the pollen cell wall.The non-germinated pollen showed no activity.Germination percentage was calculated as the following:

2.3.3.Kernel set and LER calculation

To determine the kernel set,we first collected 30 ears for each cultivar from the center rows of each plot two weeks after 50%silking,as our previous study indicated that unfertilized and aborted florets can be clearly identified at this time.Kernel set was then determined by counting the kernels two weeks after 50% silking.We also counted the unfertilized florets to determine the kernel setting rate,which was calculated as the ratio of the number of kernels to the sum of kernels and unfertilized florets two weeks after 50% silking.

At physiological maturity,grain yield was determined by hand harvesting of four non-border rows in each plot.Corn kernels were oven dried and yield was adjusted to a moisture content of 15.5%.

Land equivalent ratio was calculated as the following:

where LERAand LERBare the partial LERs of component cultivars A and B,respectively[33].

2.3.4.SSR marker screening

To identify the pollen source of harvested kernels,we conducted a DNA marker experiment to distinguish cross-fertilized kernels from self-fertilized kernels in the corn mixture.Kernel endosperm tissue was selected for DNA extraction using an improved alkali-boiling method[34].Endosperm of a size visible to the naked eye(about one-tenth the size of a sesame seed)was sampled from the top of each kernel.Each endosperm sample was placed into a well of a 96-well PCR plate.A 100-μL volume of 0.1 mol L-1sodium hydroxide(NaOH)was added to each well.All wells were sealed with a silica gel pad and centrifuged at 99 °C for 12 min.An aliquot of 100 μL 1×TE buffer(pH 2.0)was then added to each well.After 30 min at room temperature,2 μL of the solution in each well was taken as a template for PCR amplification.To screen differential SSR markers between cultivars,a total of 65 pairs of primers were used for PCR amplification,and then the differential primers were subjected to polyacrylamide gel electrophoresis(Table S1).To quantify the cross-pollination rate in the cultivar mixture,approximately 70 ears were selected from each component cultivar in the mixture in each of the three study years.Kernels of the same cultivar collected during the three-year study period were thoroughly mixed into a composite sample,from which approximately 2000 kernels were randomly selected for analysis.To reduce the border effects,all ears were selected from plants in non-border rows in each plot.The crosspollination rate of one component cultivar was calculated as the ratio of kernels carrying a marker allele from the other component cultivar to all tested kernels.

2.4.Statistical analysis

Data collected from the three-year study period were analyzed separately each year using Proc Mixed Model of SAS[35].For grain yields,ASI,and duration of pollen shed,data were analyzed considering treatment(six levels including four monocrops and two mixtures)a fixed factor and block a random factor.Each mixture consisted of two cultivars differing in high-temperature sensitivity and flowering phenology.To identify differences among the four cultivars planted in two cropping systems(monocrop vs.cultivar mixture),kernel setting rate,germination percentage,kernel number per plant,and 1000-kernel weight were fitted with a twofactor factorial design.Cropping system(monocrop vs cultivar mixture)and cultivar(four cultivars of WK702,DH605,DH701,and DH662)were treated as fixed factors,and block as a random factor.For all data analysis,model assumptions(normal distribution and constant variance of error terms)were verified by examining the residuals.When the treatment was significantly different at an alpha level of 0.05,the least squares means were compared and reported.Orthogonal comparisons were constructed to determine whether response variables differed significantly between monocrops and cultivar mixtures across all cultivars,or between monocrop and mixture for a given cultivar.

3.Results

3.1.Weather

Fig.1.Corn planting patterns showing plant arrangements of corn monocrop and mixture.‘‘×”denotes a high temperature-sensitive cultivar,‘‘o”denotes a high temperature-resistant cultivar.

In the study region,high temperature(greater than 35°C)often occurs from July 15 to August 15.During this high-temperature period,corn may be at either late vegetative growth stages(from V10 to VT.V10:the tenth leaf collar being visible;VT:the last branch of the tassel being completely extended)or early reproductive growth stages(R1 and R2:from silks being visible outside the husks to kernels containing～85% moisture with a blister-like appearance).Such high temperatures have become more frequent and intense since 2013(Fig.2).In the four decades from 1970 to 2009,cumulative days with a daily maximum temperature(Max T)greater than 35°C averaged 3.7 days per year,but they averaged 12.5 days per year from 2010 to 2018.During the three-year study period,2016–2018,this number was 20.3 days.Particularly in 2017,there were nine days with Max T above 40.0 °C from the middle of July to the middle of August.The maximum in-plot temperatures recorded at a height of 1.3 m above the ground were 41.1 °C,43.1 °C and 37.8 °C in 2016,2017,and 2018,respectively.The cumulative daily thermal times with a maximum temperature above 35 °C were 32.9,91.5,and 21.1(in °C d)in 2016,2017 and 2018,respectively.

3.2.ASI,duration of pollen shed,and pollen germination

Cultivar mixtures showed significantly shortened ASI compared with the monocrops of the corresponding component cultivars in all three years(Fig.3).In 2016 and 2017,the ASIs of HTT cultivars were significantly shorter than those of HTS,whereas there was no difference in 2018.Across all individual corn monocrops,the mean ASI varied from 1.1 to 1.7 days,while across all cultivar mixtures,the mean ASI was shortened to＜0.5 days.During the three-year study period,the HTS cultivar DH662 generally started pollen shedding two days earlier than silk extrusion.Mixing DH662 with the HTT cultivar DH701 resulted in the largest reduction in ASI at the mixed-population level.An orthogonal contrast indicated that cultivar mixtures showed shortened ASI compared with monocrops(P＜0.05).

Mixing cultivars significantly increased the duration of pollen shedding compared with the HTS component cultivars in all three years;however,mixtures showed increased duration of pollen shedding in two of three years,compared with the HTT component cultivars(Fig.3).Among all four cultivars,the duration of pollen shedding varied from a minimum of four days for the HTS cultivar DH662 to a maximum of eight days for the HTT cultivar DH701.Among all the cultivars,DH662 shed pollen first while WK702 sheds pollen last.Mixing cultivars with different pollen shedding durations lengthened the duration of pollen shedding in mixtures.Averaged across the entire study period,the duration of pollen shedding was two days longer in mixtures than in monocrops with a relatively short pollen shedding duration.

Fig.2.Weather at the experimental site.(A)Daily maximum temperature and number of days with daily maximum temperature greater than 35 and 40°C between July 15 and August 15 from 1970 to 2020 at the experimental site.(B)the maximum and minimum daily temperatures measured at the height of corn ears(1.3 m above the ground)and precipitation during the experimental period of 2016–2018.The period of July 15 to July 25 corresponds to the pre-anthesis period and the period of July 26 to August 2 corresponds to the flowering period.Max T,maximum temperature.Min T,minimum temperature.

Fig.3.Flowering traits including anthesis–silking interval(A,B,C)and duration of pollen shed(D,E,F)for four corn cultivars(WH702,DH605,DH701,and DH662)and two cultivar mixtures(WK702 and DH605,DH701 and DH662),2016–2018.Bars marked with different letters indicate a difference at the 0.05 probability level.

Mixing cultivars did not increase the percentage of pollen germination compared with the HTT component cultivars,but increased(P＜0.05)the percentage of pollen germination compared with the HTS in all three study years(Figs.4,S1).For the HTS cultivar DH605,the pollen germination percentage increased by respectively 11.6%,10.8%,and 7.9% in mixtures compared to their monocrops in 2016,2017,and 2018.For HTS cultivar DH662,the pollen germination percentage increased by respectively 17.3%,14.9%,and 7.4%in mixtures compared to their monocrops in 2016,2017,and 2018.Among all four cultivars,the percent of pollen germination varied from a minimum of 40.7%for the HTS cultivar DH662 in 2018 to a maximum of 70.0% for the HTT cultivar DH701 in 2017.The percentages of pollen germination were respectively 37.0%,38.8%,and 30.2% higher for the HTT cultivars WK702 than for the HTS cultivar DH605 in 2016,2017,and 2018,and was 44.1%,44.3%,and 43.6% for the HTT DH701 than for the HTS DH662 in 2016,2017,and 2018.Mixing cultivars with differing pollen germination percent significantly increased the percent of pollen germination of HTS cultivars in mixture.

3.3.Kernel setting rate,kernel number per plant,and 1000-kernel weight

The kernel setting rate was higher(P＜0.05)in cultivar mixtures than in monocrops in all three years,with a greater increase for HTS cultivars than for HTT cultivars(Fig.5).During the threeyear study period,kernel setting rate was higher in 2017 than in 2016 and 2018.Across all four study cultivars,the kernel setting rate averaged 79.3% with a range from 74.7% to 84.1% in monocrops and averaged 88.0%with a range from 86.4%to 88.7%in mixtures.The largest increase was observed for the HTS cultivar DH662 in 2018,which increased the kernel setting rate by 24.1%in the mixture(81.6%)compared with the monocrop(65.8%).In comparison,the smallest increase in kernel setting rate was observed for the HTT cultivar DH701 in 2017,which increased by only 4.8% in the mixture(to 92.8%)relative to the monocrop DH701(88.6%).

Fig.4.Germination percentage of pollen of each study cultivar planted in monocrop and mixture,2016–2018.Bars marked with different letters indicate a significant difference at the 0.05 probability level.

Fig.5.Kernel setting rate(A,B,C),kernel number per plant(KNP)(D,E,F),and 1000-kernel weight(G,H,I)of four corn cultivars under monocrops and cultivar mixtures,2016–2018.WK702 and DH701 are high temperature-tolerant while DH605 and DH662 are high temperature-sensitive.The two cultivar mixtures refer to 1)WK702 and DH605 and 2)DH701 and DH662.Bars marked with different letters indicate a significant difference at the 0.05 probability level.

Similar to the kernel setting rate,kernel number per plant also increased(P＜0.05)in cultivar mixtures compared with that in monocrops in each of three years(Fig.5).For example,kernel number per plant of HTT cultivar WK702 was 15.0%,10.7%,and 13.0%higher in mixtures than in monocrops in 2016,2017,and 2018,respectively.Kernel number per plant ranged from 436.1 for the WK702 monocrop in 2018 to 514.0 for the DH662 in the mixtures in 2017.For the HTS cultivars,the kernel number per plant was 8.8%–11.2% higher in mixtures than in monocrops.In comparison,for the HTT cultivars,the kernel number per plant was 10.9%–12.8% higher in mixtures than in monocrops.The results of an orthogonal contrast indicated that cultivar mixtures had a larger positive effect on kernel number per plant for the HTS cultivars(12.0%)than for the HTT cultivars(9.9%)(P＜0.05).

Cultivar mixtures showed no effects on 1000-kernel weight in comparison with monocrops in any of the three study years,but cultivars themselves showed a significant difference in 1000-kernel weight(Fig.5).The 1000-kernel weight averaged 342 g across three cultivars of DH605,WK702,and DH701,which was 8.9% higher than that for the cultivar DH662(P＜0.05).

3.4.Grain yield and land equivalent ratio

Cultivar mixtures showed increased grain yield compared with their corresponding monocrops in all three years(P＜0.05),with an exception that there was no yield difference between mixtures and monocrops for the HTT cultivar DH701 in 2018(Fig.6).The grain yield was highest in 2017 and lowest in 2018.For the HTS cultivars,yields in mixture increased by 16.0%and up to 19.5%in 2018 compared to monocrop,while for the HTT cultivars,grain yield in mixture increased by 8.4% and up to 13.5% in 2018 compared to monocrop.Averaged across the three-year study period,the yield of the DH605 and WK702 mixture(9942 kg ha-1)was 13.3%higher than that of the HTS cultivar DH605 monocrop(8773 kg ha-1).Similarly,the yield of the DH701 and DH662 cultivar mixture(10207 kg ha-1)was 18.7% higher than the HTS DH662 monocrop(8602 kg ha-1).Cultivar mixtures also showed increased yield compared with the HTT cultivar monocrops.For instance,the averaged yield of the DH605 and WK702 cultivar mixture increased by 11.1% compared with the WK702 monocrop(8950 kg ha-1),whereas the yield of the DH701 and DH662 cultivar mixture increased slightly(5.6%)relative to the DH701 monocrop(9663 kg ha-1).An orthogonal contrast indicated that cultivar mixtures had a higher yield than monocrops(P＜0.05),with a mean 19.8% increase.

Averaged across the three-year study period,LER was 1.12 for the WK702 and DH605 mixture,with an equal partial LER of 0.56 for both component cultivars.The LER was also 1.12 for the DH701 and DH662 mixture,but with a higher partial LER for the HTS cultivar DH662(0.59),than for the HTT cultivar DH701(0.53).The partial LERs for all study cultivars were greater than 0.50 in any given study year.

3.5.Pollen sources in cultivar mixtures

Fig.6.Grain yield of corn in monocrops and mixtures.WK702 and DH701 are high temperature-tolerant,whereas DH605 and DH662 are high temperature-sensitive,2016–2018.Bars marked with different letters indicate significant difference at the 0.05 probability level.

SSR marker showed that 30.9% of kernels displayed differing SSR markers for the WK702 and DH605 mixture and 41.2% for the DH701 and DH662 mixture(Fig.7).Across all four cultivars,the self-fertilization percentage was 1.5%–133.1% higher than the cross-fertilization percentage,which ranged from 29.3% to 47.8%(Fig.7).For the WK702 and DH605 mixture,the crossfertilization percentage was 29.3% for the HTT cultivar WK702 and 43.6% for the HTS cultivar DH605.For the DH701 and DH662 mixture,the HTT cultivar DH701 had a lower cross-fertilization rate(33.5%)compared with the HTS cultivar DH662(47.8%).In addition,2.4%–4.1% of kernels were fertilized from unknown sources other than the component cultivars.

4.Discussion

4.1.Cultivar mixtures showed increased synchrony of pollen shedding with silking

Flowering is one of the most important phenological indicators of crop adaptation to a changing growth environment,especially in a stressful environment[36].In the study region,high temperature(days with maximum temperature≥35°C)often occurs from late July to early August.The trend of high temperature occurrence indicated that high-temperature stress has become more frequent and intense in recent years.For example,during early and middle July in 2018,the maximum temperature above 35 °C lasted more than 10 days at the experimental site,and a rain storm occurred from July 27 to August 1(Fig.2).The similar ASI of HTT cultivars to HTS cultivars in 2018 may have been due to rain during flowering,possibly leading to a shorter duration of pollen shed and lower pollen germination percentage,contributing to the lower grain yield in 2018 than in 2016 and 2017.Field observations indicated corn tassel malformation in 2016 and 2017 due to the prolonged high temperature-induced heat and water stresses.This event likely impaired pollination and yield for the HTS cultivars DH662 and DH605.Management practices such as mixing high temperature-tolerant and-sensitive cultivars could be developed as an alternative strategy for coping with the high-temperature environment often occurring at corn flowering.

At the crop population scale,mixing corn cultivars with different genetic backgrounds altered the phenology of flowering traits(silking and anthesis)in comparison with a single pure corn cultivar.Studies[37,38]have found that time to silking was more affected by plant biomass growth(ear biomass accumulation)than time to anthesis.Numerous studies[22,26,39,40]have found that stressful environmental conditions reduced plant growth,lessened assimilate supplies,and then delayed silking.In the present study,the HTS cultivars(DH662 and DH662)had delayed silk emergence compared to the HTT cultivars in the stressful,high-temperature years.Delay of silking often leads to reduced ear biomass,and drought-tolerant cultivars normally require less ear biomass accumulation for silk appearance[28,37].We speculate that the yield reduction in the HTS cultivars in the monocrop treatment was due to the uneven silk emergence and reduced assimilate supplies under stressful environments.

Fig.7.Specific SSR markers and their crossing in corn cultivar mixture of(a)WK702 and DH605 and(b)DH662 and DH701,and(c)the percentage of kernels fertilized by self-pollination,by cross-pollination,and from unknown pollen sources in the mixture.C,cross-fertilization;M,marker;bp,base pair;Other,receiving pollen from sources other than the tested cultivars in the mixture.WK702 and DH701 are high temperature-tolerant and DH662 and DH605 are high temperature-sensitive.

Mixing the HTS and HTT cultivars extended the pollen shedding duration at the population level and supplied adequate pollen for late-extruded silks.However,pollen quantity may not be a critical yield-limiting factor under drought stress conditions[41,43].Pollen viability may also not be severely reduced by short-period high temperature stress under field conditions,as a result of buffering capacity from the corn canopy[39,44].We speculate that reduced assimilate supplies result in kernel abortion and reduced kernel number and yield in a stressful environment such as drought.

In addition,a close synchrony between anthesis and silking is critical for maximizing kernel setting rates and yield.For the HTS cultivars,silks in the lower parts of ears usually had delayed extrusion,resulting in a lower fertilization rate.In a high-temperature stressful environment,lengthened ASI was more evident in the HTS cultivars than for the HTT cultivars,causing a lower kernel fertilization rate.A prolonged duration of pollen shedding likely increases the rate of kernel abortion.Kernels fertilized at the end of pollen shedding may be disfavored in competing for nutrient supplies with kernels fertilized at the beginning of pollen shedding,thereby increasing the kernel abortion rate under biotic and abiotic stresses at the later reproductive growth stage[45].Similarly,Shen et al.[24]reported that delayed pollination was one of the factors causing corn kernel abortion.The higher kernel number per plant in mixtures than in monocrops was likely due to uniform pollination.

The difference in silking and anthesis phenology was attributed mainly to diverse genetic background and reflected adaptability to stressful environments such as drought[46]and heat[23].Studies[23,47]showed that the HTT cultivars had the capacity to increase biomass partitioning to the ear,resulting in rapid silk extrusion.In the present study,anthesis occurred earlier than silking in all four cultivars;thus,rapid silking in the HTT cultivars effectively shortened ASI,compared with the HTS cultivars.A shortened ASI suggested a larger amount of pollen shedding at the critical kernel fertilization stage,which might result in a uniform kernel fertilization.

For the HTS cultivars,high temperature often results in asynchrony between pollen shedding and silk extrusion[23].The duration of pollen shedding determined by genotype and environmental conditions is normally short(4–6 days),and synchronization of anthesis with silking is vital to achieve greater kernel setting rate and a lower kernel abortion rate[40,48].Under certain stress gradients,the risk of low kernel setting rate under high temperature could be effectively mitigated by mixing cultivars with differing tolerance to high-temperature stress[13].Mixing corn cultivars with complementary flowering traits could facilitate fertilization processes via synchrony between anthesis and silking,particularly under stressful environments affecting the development and occurrence of both anthesis and silking.At the population level,the synchronization of anthesis and silking was increased in HTS-HTT mixtures in comparison with HTS monocrops.Cárcova et al.[25]reported that synchronous pollination increased kernel number per plant up to 31%.We argue that the synchrony of anthesis with silking in the cultivar mixture is one of the main mechanisms governing kernel setting and increasing grain yield.It is vital that cultivar mixtures have a short ASI through cultivar mixtures for uniform kernel fertilization and greater kernel number per plant.

ASI showed a strong and consistently negative correlation with corn grain yield[38,46].Variation in ASI accounted for 70%of variation in corn grain yield under drought stress[46].In the Huang-Huai-Hai study region in the transition climate zone,where the temperature in corn growing seasons varies widely as a result of being located in the transition climate zone,a short ASI is an ideal trait of considerable importance in stabilizing corn yield.

Mixing corn cultivars with diverse stress tolerance traits may be an effective adaptation strategy for increasing kernel set and coping with climate change,given that mixtures can increase fertilization synchrony within and between ears and mitigate fertility defects of sensitive cultivars.To achieve synchrony between anthesis and silking and further increase cross-pollination and kernel set,we need to consider the dynamics of both tassel and silk development of each cultivar when selecting candidate mixtures based on regional cultivar evaluation trials.This practice will require a better understanding of corn flowering traits under local conditions before cultivar mixtures are assembled.

4.2.Cultivar mixtures showed increased cross-fertilization

In corn cultivar mixtures,up to 47.8% of cross-fertilization was observed.Cross-pollination is affected by wind speed,wind direction,ASI,and distance between pollen receptors and donors[42].The cross-pollination occurs mainly within 10 m and decreases exponentially with increased distance[49].The lowest crosspollination rate in our study was 29.3%,which is higher than those reported by others[42,50].Ma et al.[49]reported that the crossfertilization rate of corn with the first adjacent row averaged 21.3% downwind and 9.2% upwind from the pollen source.In the corn cultivar mixture treatment,one cultivar was planted near another to mimic the mixed-row intercropping pattern.Such a mixed-row planting pattern might explain the higher crossfertilization rate reported in our study.In addition,the close flowering synchrony of the two mixed cultivars could also account for a higher cross-pollination rate and kernel number per plant.

Prolonged ASI resulted in a reduction in the period from extruded silks to pollen,especially for the seen last ear under stressful conditions,which produced a reduced quantity of pollen when receptive silks were presented,leading to asynchronous pollination.A previous study[48]found that the larger the proportion of synchronous fertilization among tip-to-base ovaries along an ear,the greater was the kernel set per plant.Corn kernel number was reported[51]to increase by 15%through synchronous pollination,while a pollination gap of 2 and 4 d reduced kernel number by up to 51%.We found that the percentage of cross-pollination of cultivar mixtures was higher for HTS cultivars than for HTT cultivars.Heat-tolerant cultivars shed a higher percentage of active pollen than temperature-sensitive cultivars,explaining the increased synchronous pollination and kernel set for the HTT cultivars.This finding indicated that the HTS cultivars had a greater chance to receive pollen from the HTT cultivars.Thus,corn inter-cultivar mixtures may help reduce,by cross pollination,the fertility defect when one cultivar cannot achieve self-pollination owing to environmental stress.Cultivar mixtures could also achieve intercultivar complementation by eliminating their respective asynchronous timings of anthesis and silking.

Mixing corn cultivars with differing genetic origins likely increased grain yield through the xenia(cross-pollination)effect[52].Weingartner et al.[53]reported an average yield increase in corn of 2.6% as a result of the xenia effect.In the present study,the average 1000-kernel weight across four study cultivars increased by 1.0% in corn mixtures relative to corn monocrops,probably owing to cross-pollination effects.An increase in kernel weight contributed to yield improvement in the cultivar mixtures.

4.3.Cultivar mixture improved grain yield

Grain yield in corn mixtures was 19.8% higher than in corn monocrops.This yield gain was comparable to a yield advantage of 0.15–0.17 t ha-1in multiple types of corn mixtures reported by Wang and Zhao[54].Yield benefits in cultivar mixtures were also reported in wheat and rice studies,showing a yield increase of 6.0%–45.6% in cultivar mixtures[36,55].The yield increase in cultivar mixtures might also be attributed to enhanced ecological services such as improved resource use efficiency and reduced disease severity.Borg et al.[13]also reported a greater over-yielding from mixing cultivars with contrasting phenological traits.To maximize yield benefit in cultivar mixtures,it is critical to select cultivars that have required functional traits used to mitigate sitespecific issues faced by local producers.

Modern breeding efforts reduce plant-to-plant variation in phenotypic traits such as silk appearance for the same cultivar,but a clear genotypic difference exists among different cultivars[27].The plant-to-plant variation in stress-sensitive genotypes increased under stressful environments,as evidenced by inconsistent silking and elongated silk extrusion period[47].These dysfunctional flowering traits under a stressful environment posed a challenge for sustaining yields in a monocrop setting.It also suggests difficulties under foreseeable climate change,which increases the chances of pest outbreaks and abnormal weather.Mixing cultivars with differing functional traits increased withinspecies biodiversity,which might effectively mitigate the negative effects of biotic and abiotic stresses on kernel fertilization and grain yield.

To maximize the benefits of cultivar mixtures,plant phenological compatibility must be considered when cultivar mixtures are assembled[13].In the present study,cultivars with similar plant height were chosen when cultivar mixtures were assembled to make practical field operations feasible.We recommend also conducting a pre-screening study to select ideal component cultivars with required phenotypic traits.

The LER of 1.12 in the cultivar mixture indicated that 1.12 ha of monocrop land would be needed to obtain the same grain yield produced by 1 ha of land with a cultivar mixture.The greater LER in our study may be due mainly to flowering trait synchrony and complementarity,as indicated by a shorter ASI in mixtures.In intercropping of multiple crop species,complementary and compensation effects have often been observed,owing to mutualistic and competitive relationships[13].In this study,we observed more complementary than compensation effects,given that the partial LER was larger than 0.5 for each component cultivar.The complementary effects could be the result of increased resourceuse efficiency,which might partially compensate for the limited assimilate supply during the grain-filling period under stressful conditions.The complementary effect demonstrated that the two cultivars benefited from each other in mixtures,producing a higher yield per unit of land in cultivar mixtures.

4.4.Implication of cultivar mixtures for agricultural sustainability

Developing adaptive management practices through biodiversity is essential for increasing productivity and stability of cropping systems under foreseeable climate change.Studies[56–58]have shown that diversification via intra-or inter-species intercropping showed great potential to mitigate biotic and abiotic stresses,increasing production,stability,and profitability.However,in the Huang-Huai-Hai study region,a few elite corn cultivars dominate the corn area,substantially eroding corn genetic diversity.The loss of genetic diversity increases the vulnerability of corn production under increasing biotic and abiotic stress conditions.With the availability of cultivars differing in genotype,yield potential,and tolerance to biotic and abiotic stresses,it is possible to diversify current corn cropping systems by cultivar mixing.Cultivar mixtures also provide an opportunity to increase the area of elite cultivars with fewer concerns of losing genetic biodiversity,thus increasing the resilience of the current cropping systems.

The present study showed that intra-species mixtures with differing phenological and high-temperature traits produced higher yield than a single pure cultivar in the entire study period,when high temperatures occurred in two of three years.Cultivar mixtures with differing traits showed increased kernel set and yield as a result of both synchronized and complementary flowering traits,exemplifying an agronomic approach allowing corn producers to cope with a stressful growth environment.The success of using trait-based cultivars to mitigate environmental stresses can also be extended to cope with biotic stress such as pest outbreaks by assembling pest-susceptible and-resistant cultivars.Overall,cultivar mixtures diversify intra-specific biodiversity,increasing the resilience of cropping systems to biotic and abiotic stresses.

5.Conclusions

A cultivar mixture of corn with different traits in flowering and high-temperature tolerance showed differences in pollination,kernel set,and grain yield relative to a corn monocrop.The kernel setting rate in cultivar mixtures was increased via shortening of the anthesis–silking interval compared with those in corn monocrops.Close synchrony of anthesis and silking in cultivar mixtures was critical to facilitate uniform pollination,likely reducing kernel abortion rates.The cross-fertilization rate in corn mixtures was up to 47.8%,contributing to the 13.3%–18.7%yield gain in mixtures compared with monocrops.Selecting cultivars with complementary traits based on cultivar evaluation trials is critical for maximizing the benefits of cultivar mixtures.We conclude that mixing a HTS cultivar with a HTT cultivar proved to be an effective strategy for increasing corn yield in a corn region with periodic high-temperature stresses.

CRediT authorship contribution statement

Hongping Li:Writing-original draft,supervision and reviewing.Kui Liu:Formal analysis and writing-review&editing.Zhibin Li:SSR marker screening,and investigation.Moubiao Zhang:Weather data curation and investigation.Yongen Zhang:Formal analysis.Shuyan Li:Weather data analysis.Xiuling Wang:Formal analyses and methodology.Jinlong Zhou:Software and supervision.Yali Zhao:Field data curation and validation.Tianxue Liu:Methodology,project administration and supervision.Chaohai Li:Conceptualization,funding acquisition,project administration,resources,supervision and reviewing.

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

We thank Prof.Jinhua Tang for providing the core primers and screening for differential SSR markers.We thank Prof.Jiuran Zhao for providing another set of core primers.This work was supported by National Natural Science Foundation of China(31801308),Henan Provincial Higher Education Key Research Project(21A210024),CMA·Henan Key Laboratory of Agrometeorological Support and Applied Technique(AMF202109).

Appendix A.Supplementary data

Supplementary data for this article can be found online at https://doi.org/10.1016/j.cj.2022.05.007.