Grain yield and grain moisture associations with leaf,stem and root characteristics in maize

XU Chen-chen,ZHANG Ping,WANG Yuan-yuan,LUO Ning,TIAN Bei-jing,LIU Xi-wei,WANG Pu,HUANG Shou-bing

College of Agronomy and Biotechnology,China Agricultural University,Beijing 100193,P.R.China

Abstract Improving grain yield (GY) and reducing grain moisture (GM) are urgent demands for directly harvesting kernels with combine harvesters in maize production. GY and GM are both related to leaf,stem and root characteristics,but the relationships are not fully understood. To better understand these relationships,we conducted a field trial involving 12 maize hybrids with two sowing dates in 2017 and 10 maize hybrids with one sowing date in 2019. GY ranged from 6.5-14.6 t ha-1 in early-sown varieties and 9.3-12.7 t ha-1 in late-sown varieties in 2017,and 5.9-7.4 t ha-1 in 2019,respectively,with corresponding GM variations of 29.8-34.9%,29.4-34.5% and 31.9-37.1% at harvest. A large maximum leaf area contributed to a high yield,a fast leaf senescence rate accelerated grain dehydration in the late growth period,and a compact root structure resulted in both of high-yield and fast-grain dehydration. A strong stem improved lodging resistance but maintained a high GM at harvest,and it is challenging to combine high GY and low GM in maize. High GY co-existed with low GM in some varieties that should have a rapid grain filling,a relatively long grain-filling duration,and a rapid grain dehydration in the late growth period. A high daily temperature in the late growth period also improved GY and reduced GM by influencing grain filling and dehydration,suggesting that adjusting the sowing date should be an alternative strategy to combine high GY and low GM in kernel harvesting.

Keywords:maize,grain yield,grain moisture,stem strength

1.Introduction

Maize (ZeamaysL.) accounts for a large percentage of human food and animal feed worldwide. Agronomists and crop breeders are constantly attempting to increase maize yield through optimizing growth phenology (Tsimbaet al.2013;Longet al.2017;Luet al.2017),leaf productivity (Eik and Hanway 1966;Winter and Ohlrogge 1973;Zhenget al.2009),stalk lodging resistance (Peifferet al.2013;Zhanget al.2018),root system architecture (Miet al.2016),and grain filling traits (Gasuraet al.2013;Liuet al.2021).Along with the demand for higher maize yield,increasing the mechanization of maize production in order to reduce the labor intensity is urgently needed due to the rapid loss of agricultural labor in China (Zhang Xet al.2017).These needs will become more urgent with the centralized management of smallholder farming fields (Wanget al.2015) as well as the aging of the Chinese population (Peng 2011). Compared to mechanization in tillage and sowing,mechanization in kernel harvesting is more difficult in maize in the double cropping regions of China. Maize kernel harvesting is highly dependent on grain moisture (GM) (Wanget al.2019;Zhanget al.2021;Chuet al.2022).Grain yield (GY) is dependent on leaf productivity (Eik and Hanway 1966;Caoet al.2021) and stalk lodging resistance (Reymondet al.2003). GM and GY are both closely related to root system architecture (Hammeret al.2009). Therefore,optimizing the relationships among leaf,stem dynamics,root,GM,and GY is of the utmost importance,but the knowledge regarding these relationships is currently lacking.

To achieve high GY,modern maize varieties combine the traits of large leaf areas (Ciet al.2012),slow leaf senescence (Chenet al.2013) and compact plant structures (Wanget al.2011). In high-yield maize populations,the optimized canopy structure intercepts more solar radiation in the middle and upper leaf layers,increasing their contribution to GY. The weaker light conditions in the lower canopy layers of dense plant populations are believed to reduce maize stem strength and increase lodging (Xueet al.2016),reflecting the close relationships between leaf and stem. As an important support and transportation organ,the stem can in turn directly affect the aboveground parts of the plant,but its impacts on leaf productivity and GY are not fully understood. Roots are equally crucial and complementary with the above-ground parts for maize yield,though they function differently (Robinson and Peterkin 2019). The ability of the root system to absorb nutrients and water depends on the morphology and spatial arrangement of the root system (Liet al.2019). Several studies reported that modern maize hybrids have wide and compact roots to absorb more water and nutrition for above-ground growth and then give a high yield (Hammeret al.2009;Chenet al.2014). However,the relationships between GM and root structure are not fully understood in maize.

Evidence has clearly indicated that GY is closely associated with leaf area dynamics in maize (Wanget al.2019;Tianet al.2020);a proper large leaf area index and leaf area distribution in space during and after the flowering stage can result in a high GY (Huanget al.2017). The relationships between GM and leaf area dynamics are not fully understood. A rapid leaf senescence is supposed to retard grain filling and accelerate grain dehydration (Kanget al.1986;Heet al.2004;Chenet al.2015). In addition,grain filling and dehydration co-exist during the grain filling period and determine the final grain yield and moisture at harvest (Kanget al.1986;Brooking 1990;Borrás and Westgate 2006). The interaction between grain filling and dehydration is likely related to leaf senescence dynamics,but the relevant knowledge is not clearly known.

Apparently,grain filling and grain moisture dynamics are coupled with leaf senescence,stem degradation and root system architecture,and there seems to be innate relationships among these factors. Such knowledge is important for the selection or breeding of high-yield,lodging-resistant and kernel-harvesting varieties of maize. Therefore,the objectives of this study were to 1) explore the relationship between grain yield and grain moisture and 2) determine the relationships among leaf area,stem strength,root system architecture,GM,and GY.

2.Materials and methods

2.1.Experimental site

All field experiments were carried out at the Wuqiao Experimental Station (37°41´N,116°36´E) of China Agricultural University in a loam soil,Hebei Province,China in 2017 and 2019. Weather data from the maize growing season in 2017 and 2019 are shown in Fig.1.

Fig.1 Monthly average temperature,precipitation and sunshine duration in the maize growth seasons in 2017 (A) and 2019 (B) at the experimental site.

The mean precipitation of 366.4,429.5 and 442.5 mm,mean temperatures of 25.0,23.1 and 22.5°C,and sunshine duration of 7.4,6.2 and 6.7 h were recorded during the two growing seasons (May-September and June-October) in 2017 and one growing season (July-October) in 2019,respectively. Soil of 0-40 cm depth contained 40.1 and 15.1 mg kg−1organic matter,0.8 and 0.9 mg kg−1total N,13.2 and 13.0 mg kg−1available phosphorus (Olsen-P),and 119.0 and 78.8 mg kg−1ammonium acetate extractable potassium (K) in 2017 and 2019,respectively.

2.2.Experimental design and plant management

Experiments were arranged in a randomized complete block design,at a plant density of 60 000 plants ha-1,with three replicates. In 2017,12 varieties,including two widely adaptable varieties in China,five varieties adapted to the North China Plain,two varieties to Northwest China and three varieties to Northeast China,were manually sown at two sowing dates (May 19 and June 21). In 2019,10 varieties,including two adaptable varieties in China,five varieties adapted to the North China Plain and three varieties to Northeast China,were sown on July 1. Detailed information about the maize varieties is presented in Table 1.Grain harvesting was conducted from September 18 to October 30,depending on the physiological maturity of each variety. Each plot was 8.0 m long and 5 m wide,with a row spacing of 0.6 m. Fertilizer and water management were optimized throughout the maize growing season. Weeds,insects and diseases were well controlled. At sowing,78.75 kg N ha−1,90 kg P2O5ha−1and 84.4 kg K2O ha−1were applied. An additional 101.25 kg N ha−1was applied at the silking stage. Irrigations were applied at sowing and drought time. Herbicide was sprayed at the 4-leaf stage,and pymetrozine and imidacloprid wettable powder were sprayed at 5-day intervals from the 5-to 7-leaf stage.

Table 1 Information on the maize hybrids used in 2017 and 20191)

2.3.Sampling and measurements

Grain yield (GY),grain moisture (GM),thousand-kernel weight (TKW),grain-filling rate (GFR),root top angle,and root crown width (RCW) were measured in 2017 and 2019. Leaf area at the silking and maturity stages,rind penetration strength (RPS) and bend strength (BS) at the maturity stage were measured in 2017.

Leaf areaAt the silking and maturity stages,the lengths and widths of individual green leaves of three randomly selected plants in each plot were measured. Leaf area was calculated as follows (Montgomery 1911):

Leaf area=Ʃ(Leaf length×Maximum leaf width)×0.75 for expanded leaves

Leaf senescence=(Leaf area at silking stage-Leaf area at maturity)/Leaf area at silking stage

The leaf area index (LAI) was leaf area at the unit soil area.

Stem rind penetration strength (RPS) and bend strength (BS)After measuring leaf area,leaves and sheaths were separated from the stem. The stem RPS and BS were measured from the 3rd to 7th nodes with a stalk strength tester (YYD-1,Zhejiang Top Instrument Co.,Ltd.,Hangzhou,China) (Liuet al.2015;Shiet al.2016). RPS was equal to the force needed for penetration through the skin of stem and BS was equal to the force needed to snap or break the stem.

GFR,GM,grain moisture reduction rate (GMRR),TKW,and GYThe ears of three plants that were carefully selected at the flowering stage were harvested at 15,30 and 45 days after pollination and maturity in 2017 and at 10,30 and 45 days after pollination and maturity in 2019. The kernels from four kernel rows were carefully separated from the ear cob with tweezers,counted,weighed,and finally ovendried at 80°C to a constant weight. The TKW (g) was calculated over different growth periods based on the kernel number and dry weight. GFR (mg d-1) was calculated as GFR=ΔKW/T,where ΔKW represents the increase in kernel weight during the time period T (d) between two ear harvests. GM was calculated as GM (%)=(Grain fresh weight-Grain dry weight)/Kernel fresh weight×100.GMRR was calculated as GMRR=ΔGM/T,where ΔGM represents the decrease in the GM during the time period T (d) between two ear harvests.

To determine the final GY,all ears in four adjacent rows 5-m long in an undisturbed area of each plot were harvested and counted. All ears in each plot were weighed,and the mean ear weight was calculated in order to select 20 representative ears. The fresh weight of the 20 ears was equal to the mean ear weight multiplied by 20. All kernels from the 20 ears were threshed and oven-dried at 80°C to constant weight. Based on the harvesting area and dry grain weight,the GY per hectare was calculated and adjusted to a moisture content of 14%.

RTA and RCWThe root crown together with attached soil of three carefully selected plants were excavated from every plot at the maturity stage by the shovelomics method. To ensure consistency between the roots of different plants,the entire root crown of all the plants contained a length of 40 cm that was perpendicular to the row,a width of 20 cm that was horizontal to the row,and a depth of 30 cm.After excavation,the root crowns were rinsed carefully with water to remove the soil,followed by 30 min air-drying.Afterwards,the root crowns were imaged with a digital camera from a height of 80 cm. A white plastic disk with a diameter of 2.5 cm was included in every image to show the scale. The root images were analyzed with the Digital Imaging of Root Traits (DIRT) Software Platform (http://www.dirt.biology.gatech.edu) to determine the root angle and root crown width (Buckschet al.2014). RTA is the angle between the root system and the horizontal plane at 10% of the excavation depth. RCW is the maximum root width calculated after scanning.

2.4.Statistical analysis

Analysis of the variance in leaf area,stem RPS and BS,TKW,GM,RTA,RCW,and GY was performed using SPSS 20 (SPSS Inc.,Chicago,IL,USA). Differences were assessed with the Tukey test and were considered statistically significant atP＜0.05 orP＜0.01. Correlations were also analyzed using SPSS 20.

3.Results

3.1.GY

GY varied considerably with significant differences among varieties in 2017 and 2019 (Fig.2). Yields ranged from 6.5 to 14.6 t ha-1in early-sown maize in 2017 (Fig.2-A),9.3 to 12.7 t ha-1in late-sown maize in 2017 (Fig.2-B),and 5.9 to 7.4 t ha-1in 2019 (Fig.2-C),and averaged 12.4,11.3 and 6.6 t ha-1,respectively.

Fig.2 Grain yield of maize in all the experiments as a function of variety. A,sown on May 19,2017. B,sown on June 21,2017. C,sown on July 01,2019. Bars indicate standard error of the mean (n=3). Values with different letters are significantly different at P＜0.05.

3.2.TKW and GM

The mean TKW across varieties increased from 33.0 g at 15 days after silking (DAS) to 320.4 g at maturity in earlysown maize in 2017 (Fig.3-A),from 40.5 to 320.3 g in latesown maize in 2017 (Fig.3-C) and from 5.1 g at 10 DAS to 302.7 g at harvest in 2019 (Fig.3-E). The variance of TKW among varieties increased with time during the grain-filling period (Fig.3-A,C and E). Correspondingly,the mean GM across varieties decreased with time,ranging from 79.4% at 15 DAS to 33.2% at maturity in early-sown maize (Fig.3-B) and 76.8 to 31.5% in late-sown maize (Fig.3-D) in 2017,and 86.4% at 10 DAS to 34.3% at harvest in 2019 (Fig.3-F).The differences in both TKW and GM among varieties were significant during the entire grain-filling period.

Fig.3 Thousand-kernel weight (TKW) and grain moisture (GM) at different time points after silking as a function of variety. A and B,sown on May 19,2017. C and D,sown on June 21,2017. E and F,sown on July 1,2019. ＊ and ＊＊,significant differences at P＜0.05 and 0.01,respectively,among hybrids;ns,not significant.

3.3.Correlation analysis between GY,GM and GFR

Data for the GY and GM of all the varieties in 2017 and 2019 were pooled together.There was no significant negative correlation between GY and GM at maturity (r=-0.31;Fig.4-A),but GY was significantly positively correlated with GFR from silking to maturity (r=0.68＊＊;Fig.4-B).

Fig.4 Correlation analysis between yield and grain moisture at maturity (A) and yield and grain-filling rate from silking to maturity (B). ＊＊,significant at P＜0.01.

3.4.LAI at the silking and maturity stages

The LAI of different varieties ranged from 3.2 to 4.5 at the silking stage and from 1.5 to 3.5 at the maturity stage in the early-sown maize (Fig.5-A and B). The ranges of LAI were 2.6-4.4 at the silking stage and 0.6-2.6 at the maturity stages in late-sown maize (Fig.5-C and D). The difference in LAI among varieties was much larger at the maturity stage than that at the silking stage,corresponding to the different leaf senescence rates for different varieties.

Fig.5 Leaf area index (LAI) at silking (A and C) and maturity (B and D). A and B,sown on May 19,2017. C and D,sown on June 21,2017. Bars indicate standard error of the mean (n=6). Values with different letters are significantly different at P＜0.05.

3.5.Stem RPS and BS

The average values of stem RPS and BS across varieties both decreased from the 3rd to the 7th stem nodes in both early-and late-sown maize in 2017 (Fig.6). The mean RPS decreased from 59.1 N at the 3rd stem internode to 45.3 N at the 7th stem internode in the early-sown maize (Fig.6-A),and from 54.6 to 37.3 N in late-sown maize (Fig.6-C). Similar to RPS,the mean BS decreased from the lower stem node to the upper nodes. Moreover,the RPS and BS values of early-sown maize were higher than those of late-sown maize.

Fig.6 Rind penetration strength (RPS) and bend strength (BS) from the 3rd to 7th stem node of varieties sown on May 19,2017 (A and B) and June 21,2017 (C and D). ＊ and ＊＊,significant differences at P＜0.05 and 0.01,respectively,among hybrids;ns,not significant.

3.6.RTA and RCW

RTA at the maturity stage (Fig.7-A) ranged from 29.0 to 53.3° in 2017 and from 8.3 to 33.8° in 2019. The average values of RCW at the maturity stage were 238.1 and 161.0 mm in 2017 and 2019,respectively (Fig.7-B). There was a large of difference in RTA and RCW among varieties at the maturity stage in both years.

Fig.7 Root top angle (RTA) and root crown width (RCW) at maturity in 2017 and 2019. ＊＊ indicates significant differences at P＜0.05 and 0.01,respectively,among hybrids;ns,not significant.

3.7.Correlation analysis

The GY at harvest was significantly and positively correlated with leaf area at silking and with stem BS and root system size at maturity (i.e.,RTA and max RCW;Fig.8). Also,the correlation between GY and mean temperature during the entire grain-filling period was significantly positive. The correlations between GY,leaf area at maturity,and leaf senescence rate after flowering were nonsignificant. The GM at harvest was significantly and positively correlated with leaf area and stem BS at maturity. The correlations between GM and RTA at maturity,and the leaf senescence rate after flowering were significantly negative.

Fig.8 Correlation analysis between temperature,leaf area and senescence,stem strength,root top angle and root crown width,grain yield and grain moisture.GY,grain yield;GM,grain moisture at harvest. LAa,leaf area at silking;LAb,leaf area at maturity;LS,leaf senescence;BS,bend strength at maturity;RPS,rind penetration strength;RTA,root top angle;RCW,root crown width;T,mean temperature from silking to maturity. ＊＊,P＜0.01;＊,P＜0.05.

4.Discussion

4.1.GY correlations with leaf,stem and root characteristics

The present study confirmed that GY strongly depends on leaf,stem,root characteristics,and their interactions in maize. GY correlated with leaf area more closely at the silking stage than at the maturity stage,revealing the important contribution of the maximum leaf area to GY,consistent with the findings of Eik and Hanway (1966) and Huanget al.(2017). The non-significant correlations between leaf area at the maturity stage,leaf senescence after silking,and GY indicated that delayed leaf senescence or green leaf duration made limited contributions to maize GY increase (Kosgeyet al.2013;Antoniettaet al.2014),inconsistent with the results of Thomas and Howarth (2000) and Gregersenet al.(2013). This inconsistency is likely related to growing environments and soil nutrients (Naruokaet al.2012;Antoniettaet al.2016). A long green leaf duration can improve GY under relatively dry and hot conditions but not under cool and wet conditions (Naruokaet al.2012). In the present study,the wet and cloudy conditions probably slowed assimilate production and partitioning for early-sown maize,and the low temperature reduced yield increase for late-sown maize (Fig.1). The rainfall in 2019 was concentrated in August,corresponding to the maize flowering stage. Excessive rainfall may disturb maize pollination and resulted in a lower kernel number (data now shown). Additionally,genotypic differences of leaf senescence are mainly located at the lower canopy layer,where assimilate production was assumed to be less than assimilate consumption in the late growth period of maize because of the limited photosynthetic capacity (Antoniettaet al.2014). Hence,it is difficult to increase the yield potential of maize varieties with long green leaf durations at the experimental site. Using maize varieties with a long growth period is probably not a suitable strategy for improving crop yield particularly in the double cropping system of winter-wheat and summer-maize in the experimental site. This may result in the lack of a relationship between leaf senescence and GY.

The significantly positive correlation between GY and stem BS revealed that a strong stalk benefited yield increase in maize (Singh 1970). The effects of stalk on GY were associated with lodging resistance and assimilate transport (Zhang Jet al.2017;Zhanget al.2018). In the present study,lodging did not occur in either experimental year,eliminating lodging effects on GY. Previous results showed that strong stalks had more developed vascular bundles,thus increasing assimilate allocation to kernels in maize (Wanget al.2006;Zhang Jet al.2017;Guiet al.2018).Besides,strong stalks normally accompany a large leaf area and a stable maintenance of plant structure,thereby improving leaf productivity,particularly post-silking (Xueet al.2017). However,previous studies also indicated that strong stalks competed with grain-filling for photosynthates in maize,which is not in favor of GY increase (Djordjevic and Ivanovic 1996). Therefore,there seems to be a threshold level of stalk strength that is able to maintain lodging-resistance as well as high-yield in maize.

The present study indicated that a compact root structure with a large root number was better able to increase maize grain yield (Hammeret al.2009). This root architecture can not only reduce inter-plant competition,but also increase soil nutrient acquisition,especially in heterogeneous soil (Trachselet al.2013;Chenet al.2014). Additionally,a large root number can improve root lodging resistance in maize (Miet al.2016).

4.2.GM correlations with leaf,stem and root characteristics

GM at maturity is a result of leaf area dynamics,stem and root traits,and growth period in maize (Cross 1991;Salaet al.2007;Wanget al.2019),consistent with results in the present study. GM was positively correlated with leaf area at maturity and negatively correlated with leaf senescence,indicating that a longer growth period will result in high GM at harvest,consistent with Bekavacet al.(1998). These results imply that maize varieties with short growth periods or fast leaf senescence can reduce GM at harvest in the experimental region.

The present results showed that GM was positively correlated with stem BS,indicating a strong stem can result in a high GM at maturity (Eyherabide and Hallauer 1991). The relationship between GM and stem BS is likely related to stem water content (Bekavacet al.1998,2007). The water content in the stem can improve stem BS,in which water can be transported from stem to kernelsviathe more developed vascular bundles,thus maintaining a high GM at the maturity stage for the strong-stalked varieties (Bekavacet al.2007;Zhang Jet al.2017). Djordjevic and Ivanovic (1996) showed that GM was negatively correlated with stalk rind thickness,indicating that stem anatomical structure not only determines stem strength but also affects GM to some extent.In practice,strong basal stems and low GM are favorable at harvest of maize,but the present study revealed a conflict between strong stem and low GM. This conflict is assumed to be reduced by adjusting stem anatomical structure in the breeding new varieties.

A large and active root system can delay leaf senescence (Naruokaet al.2012) and result in high GM at maturity (Bekavacet al.1998). In the present study,however,there was a negative correlation between root crown width and GM,which indicated that a wider root system was related to lower GM at maturity. The wide root architecture is better able to acquire soil nutrients at the shallow soil layer (Miet al.2016),but this root system cannot transport sufficient water from soil to crop plants because of the dry weather condition at the experimental site (Fig.1).

4.3.Relationships between GY and GM

Kernel weight increases with a decline of kernel moisture (%) during the maize grain-filling period (Kanget al.1986;Hadi 2004;Borrás and Westgate 2006),with negative correlations between GY and GM at harvest (Cross 1991;Filipovićet al.2014). The correlation in the present study was,however,not as close as in the previous studies,probably due to the used varieties and growing environments (Wardet al.2016;Tianet al.2019). Maize hybrids with a rapid grain filling and a long grain-filling duration can lead to a high GY,and a long grain-filling duration generally results in a high GM (Kanget al.1986;Hadi 2004;Zhouet al.2019). However,a general pattern cannot describe all the genotypic variability (Borráset al.2009). For instance,varieties JD66 and DH605 used in the present study had a high GY and a low GM at harvest as a result of a rapid grain filling as well as a rapid dehydration in the late grain-filling period. These results suggest that high GY and low GM can co-exist in one hybrid,meeting the demand of directly harvesting kernels with a combine-harvester.

Besides,growing environment during grain-filling period was another important factor affecting maize GY and GM (Gambínet al.2007;Zhouet al.2017;Gaoet al.2018). The present results showed that daily temperature from flowering to maturity was more closely related to GY and GM than precipitation. A high temperature can improve kernel weight by maintaining a high leaf productivity at the late growth stage (Tianet al.2019) when temperature is relatively low at the experimental site (Fig.1). Meanwhile,high temperature was significantly positively correlated with kernel dehydration rate (r=0.56＊＊,data not shown),reducing GM at harvest. Therefore,there are presumably suitable strategies to improve GY and reduce GM at harvest through matching grain filling and dehydration with proper temperature,such as using varieties with short growth periods in combination with an early sowing date.

5.Conclusion

The present results revealed that a high grain yield can coexist with a low grain moisture at harvest in a given maize variety,in association with the proper leaf,stem and root characteristics. A large maximum leaf area contributed to a high yield,a fast leaf senescence accelerated grain dehydration in the late growth period,and a compact root structure resulted in both high-yield and fast grain dehydration. Lodging resistance is necessary for kernel harvesting,but the strong stem strength associated with high grain moisture increased the difficulty of grain dehydration. Hence,stem traits should be more focused from the perspective of stem composition and anatomical structure. Both grain yield and moisture were influenced by growing environments post-flowering;temperature was an important parameter. Presumably,maize hybrids with high GY and low GM should have (1) high grain-filling rate,(2) large leaf area at silking and fast leaf senescence,(3) large and compact root system,and (4) strong stem.

Acknowledgements

The authors want to thank the staff of Wuqiao Experimental Station of China Agricultural University for their excellent field management. This work was supported by the National Natural Science Foundation of China (31701361) and the National Key Research and Development Program of China (2016YFD300301).

Declaration of competing interest

The authors declare that they have no conflict of interest.