The priority of management factors for reducing the yield gap of summer maize in the north of Huang-Huai-Hai region,China

Beijing Key Laboratory of Maize DNA Fingerprinting and Molecular Breeding,Maize Research Center,Beijing Academy of Agriculture & Forestry Sciences,Beijing 100097,P.R.China

Abstract Understanding yield potential,yield gap and the priority of management factors for reducing the yield gap in current intensive maize production is essential for meeting future food demand with the limited resources. In this study,we conducted field experiments using different planting modes,which were basic productivity (CK),farmer practice (FP),high yield and high efficiency (HH),and super high yield (SH),to estimate the yield gap. Different factorial experiments (fertilizer,planting density,hybrids,and irrigation) were also conducted to evaluate the priority of individual management factors for reducing the yield gap between the different planting modes. We found significant differences between the maize yields of different planting modes. The treatments of CK,FP,HH,and SH achieved 54.26,58.76,65.77,and 71.99% of the yield potential,respectively. The yield gaps between three pairs:CK and FP,FP and HH,and HH and SH,were 0.76,1.23 and 0.85 t ha–1,respectively. By further analyzing the priority of management factors for reducing the yield gap between FP and HH,as well as HH and SH,we found that the priorities of the management factors (contribution rates) were plant density (13.29%)>fertilizer(11.95%)>hybrids (8.19%)>irrigation (4%) for FP to HH,and hybrids (8.94%)>plant density (4.84%)>fertilizer (1.91%) for HH to SH. Therefore,increasing the planting density of FP was the key factor for decreasing the yield gap between FP and HH,while choosing hybrids with density and lodging tolerance was the key factor for decreasing the yield gap between HH and SH.

Keywords:maize,yield gap,management factors priority,hybrid,plant density,fertilizer,irrigation

1.Introduction

Yield potential (YP) is the crop yield of certain cultivars without water and nutrient limitations and with diseases and pests effectively controlled (Evans 1993). In other words,YP is the yield obtained when crop growth is only limited by temperature,solar radiation and the atmospheric carbon dioxide concentration. The difference between yield potential and actual farmers’ yield represents the yield gap,and a significant gap often exists between them(van Ittersumet al.2013; Liuet al.2016; Caoet al.2019).Previous analyses of yield gap have been founded on either investigations (Abdulaiet al.2013; Addai and Owusu 2014),crop modelling (Muelleret al.2012),field trials (Sileshiet al.2010) or synthetic assessments employing multiple methods at different levels (e.g.,household survey and complexity experienced by farmers) (van Loonet al.2019).

Many studies have reported that maize,wheat and rice yields only achieved 53,60 and 62% of the yield potentials,respectively,in China,which are significantly lower than the corresponding values of 89,84 and 85% for developed countries (Cassmanet al.1999; Liuet al.2012; Muelleret al.2012; van Ittersumet al.2013; Chenet al.2014; Liuet al.2015). These results indicate a large yield gap in the main grain crops in China,so there is still great potential to further improve grain yields,especially for maize. A large yield gap is mainly due to low actual yield which is restricted by different factors,including climatic factors,soil characteristics,crop management practices,and other social economic factors (Lobellet al.2009; Liuet al.2016a).For example,one study indicated that the 18% yield gap of maize between experimental and average farmers’ yields in China is mainly due to poor field management practices,which were more important than soil texture and cultivar selection (Zhanget al.2011; Liuet al.2016a,b).

It is difficult to increase yield potential over a short time through genetic improvement (Tollenaar and Lee 2002).Therefore,closing the yield gap between maize yield potential and farmers’ yield is of great importance for increasing grain yield and ensuring food security (Menget al.2013). Figuring out how to close yield gaps has been the focus of many studies,which have mostly concentrated on the optimization of crop management and soil improvement to close the yield gaps (Lobellet al.2009; Liuet al.2016; Wanget al.2017;Zhanget al.2019). However,yield gaps have mainly been analyzed through crop models and field surveys in different countries (Liuet al.2012; Liuet al.2016a,b; Silvaet al.2016,2018; Schilset al.2018),and thus far few approaches have been combined with field experiments. Combining the methods of different levels of field experimental analyses(including basic productivity,farmer practices,high grain yield,high efficiency level,and super high yield level) is expected to lead to relevant insights into the impacts of factors which explain yield gaps,and ways to close yield gaps through different farm management adjustments.

The Huang-Huai-Hai summer maize (HM) region is one of the largest maize-production regions in China. The region,which extends from 32 to 42°N,accounts for 34.7% of the cultivated maize area and 36.8% of the total productivity in China (Li and Wang 2010). The north region of HM contains Hebei,Tianjin and Beijing and occupies approximately 1/3 of the planting area and total production of the HM. Therefore,this area plays a significant role in ensuring food security.The objectives of this research were to (i) demonstrate and quantify the maize yield gap in the north region of HM and(ii) quantify the impact of different management technologies on the reduction of the yield gap.

2.Materials and methods

2.1.Experimental site

A field experiment was conducted in Tongzhou,Beijing(39°56´N,116°41´E),which is within the main agroecological area of the north region of HM,during the 2017–2019 growing seasons. The region is characterized by brown sandy loam soils and a temperate continental monsoon climate. Organic matter,total N,Olsen-P,and available potassium (K) in the upper 30 cm of soil were 17.03 g kg−1,1.08 g kg−1,69.67 mg kg−1,and 241.3 mg kg−1,respectively. The air temperature and rainfall during the study period growing seasons are listed in Table 1.

2.2.Experimental design

Field experiments of different planting modesThe following four management treatments were conducted in a randomized block design with three replications:(1) Basic procutivity (CK),maize yield without fertilization;(2) farmer practice (FP),applying farmers’ practices in the region but conducted in experimental plots; (3) high yield and high efficiency (HH),which was based on improving farmers’practices followed by implementing new key technologiesto improve grain yield and nutrient use efficiency; (4) super high-yield level (SH),designed to test the upper limit of yield potential,where crop yields were maximized through inputs(without considering the costs of the various inputs) so that the crops could make full use of solar radiation. Each plot was 6 m×30 m,consisting of 10 rows spaced at 0.6 m for each row. Cropping system and fertilizer application details are shown in Tables 2 and 3,respectively.

Table 1 Monthly maximum temperature (Tmax),minimum temperature (Tmin) and rainfall during the growing season (May–October)from 2017 to 2019

Factorial experimentsFour fertilizer levels (CK,FP,HH,and SH),two water methods (rainfed and irrigated),three planting densities (60 000,75 000 and 90 000 plants ha–1),and four cultivars (Zhengdan 958,Xianyu 335,Jingnongke 728,and SK567) were involved in this study in 2018 and 2019.The experiment was laid out in a split-plot design with water treatment as the main plots,and fertilizer level as subplots followed by plant density and cultivar,resulting in 96 treatments. Four management factors (fertilizer,water,plant density,and cultivars) were assessed for their individual and cumulative contributions to reducing the yield gap. There was no irrigation during the whole growing season for the rainfed treatments,except at seeding. The disease,pest and weed controls in each treatment were well-controlled.

Supplemented vs.withheld treatment structure(1)Supplementedvs.withheld treatment structure between FP and HH (Ruffoet al.2015). The supplemented and withheld experimental treatments used in this study assessed the individual and combined effects of the four field management factors (fertilizer,hybrids,plant density,and irrigation)between HH and SH.Four supplemented treatments(+fertilizer,hybrids,+plant density,and +irrigation) were conducted by individually substituting the FP level of each management factor for the corresponding HH level,while all other factors were maintained at the FP levels (Table 4).For example,the +plant density treatment was created by substituting the FP level (60 000 plants ha–1) for higher plant density (75 000 plants ha–1),while all other management factors were maintained at lower FP levels. On the other hand,the four withheld treatments (–fertilizer,hybrids,–plant density,and–irrigation) were conducted by individually substituting the HH level of each management factor for the corresponding FP level,while maintaining all other factors at the HH levels. For example,the–plant density treatment was created by substituting the HH level of higher plant population (75 000 plants ha–1) for lower plant population(60 000 plants ha–1),while all other management factors were maintained at their higher HH levels. In this way,the contribution rate of each management factor toward yield gap reduction was tested.

(2) Supplementedvs.withheld treatment structure between HH and SH. Similarly,the supplemented and withheld experimental treatments used in this study assessed the individual and combined effects of the three field management factors (fertilizer,hybrids and plant density) between HH and SH. Three supplemented treatments (+fertilizer,+hybrids and +plant density) were conducted by individually substituting the HH level of each management factor for the corresponding SH level,while all other factors were maintained at the HH levels (Table 5).On the other hand,three withheld treatments (–fertilizer,hybrids and–plant density) were conducted by individually substituting the SH level of each management factor for the corresponding HH level,while maintaining all other factors at the SH levels.

2.3.Dataset

Sample collectionThree plant samples were obtained from the center of each plot at the silking (R1) and maturity stages(R6). Samples were dried at 85°C in an oven to measure the aboveground dry matter weight. The dates of sowing,emergence,silking,maturity,and harvest were recorded.The emergence date was recorded when 60% of the plants emerged. The silking date was recorded when 60% of the ears showed silk emergence,and the physiological maturity date was recorded when a black layer appeared. At R6,all plants in the central four rows of each plot,representing an area of 12 m2,were harvested by hand to measure yield.By using 20 selected ears,we counted the number of kernel rows per ear and the number of kernels per row to calculate kernel number per ear. Grain moisture contents and weights were recorded,and the grain yields and 1 000-kernel weights were calculated at 14.0% moisture,which was measured with a portable moisture meter (PM8188,Kett,Japan). All the above traits were measured according to the standards described by Shiet al.(2006).

Model development and preliminary validationTheHybrid Maize Model was used to simulate maize growth using daily time step weather data,and also to estimate yield potential (YP) under defined growth conditions without biotic or abiotic stresses (Yanget al.2006; Liu B Het al.2017).This model has been used widely in the USA (VanWartet al.2013) and China (Menget al.2013; Houet al.2014). The maize YP in 2017,2018 and 2019 were estimated basedon meteorological data (e.g.,solar radiation,precipitation,and the maximum and minimum temperatures),sowing date,planting density,hybrid maturity,bulk density,and other factors (Liu B Het al.2017).

Table 2 Planting methods,planting density and growing period recorded in different treatments from 2017 to 20191)

Table 3 Fertilizer application rates (kg ha−1) and dates for the different treatments1)

2.4.Statistical analysis

Data analysis and the production of tables and figures were performed using Microsoft Excel 2017. Regressions between the ear number ha–1,kernel number per ear,1 000-kernel weight,and maize yield were plotted. The maize yield was the independent variable and the ear number ha–1,kernel number per ear and 1 000-kernel weight were the dependent variables. The differences in maize yield between different treatments were tested by using one-way analysis of variance(ANOVA) with the LSD test at a 5% significance level. All data analyses were conducted by using SAS 9.1.

3.Results

3.1.Yield gaps between different management treatments

Significant yield differences were found between the different management treatments (Fig.1). In 2017,the maize yieldof SH was significantly higher than those of CK and FP,but no significant difference was found between the maize yield of HH and SH. The maize yield of HH was higher than those of CK and FP,but no significant difference was found between them. The yield gaps (yield increase percentages)between HH and SH,and between FP and HH,were 0.56 t ha–1(5.10%) and 0.43 t ha–1(4.10%),respectively. In 2018,significant differences were found between all the management treatments.The yield gaps (yield increase percentages) between pairs CK and FP,FP and HH,HH and SH were 1.72 t ha–1(19.43%),0.97 t ha–1(9.16%) and 1.74 t ha–1(15.08%),respectively. In 2019,maize yields of CK,FP,HH and SH were significantly different from each other,and the maize yield of HH was the highest. The yield gaps(yield increase percentages) between pairs CK and FP,FP and HH,and HH and SH were 0.77 t ha–1(8.53%),2.31 t ha–1(23.62%) and–1.25 t ha–1(–10.36%),respectively. The average yield gaps (yield increase percentages) for the 3 yr(2017–2019) between pairs CK and FP,FP and HH,and HH and SH were 0.76 t ha–1(7.97%),1.23 t ha–1(12.03%)and 0.85 t ha–1(7.42%),respectively. Similiar trends among the maize yields of management treatments in 2017 and 2018,and significant differences were found between the different planting years.

Table 4 Supplemented and withheld treatment structures

Table 5 Supplemented and withheld treatment structures

In this study,we used the Hybrid Maize Model to estimate YP under defined growth conditions without biotic or abiotic stresses,and the yield potentials were 17.98,16.50 and 18.05 t ha–1in 2017,2018 and 2019,respectively. In 2017,the management treatments of CK,FP,HH,and SH achieved 59.34,58.16,60.54,and 63.63% of that year’s yield potential,respectively. The percentages of yield potentials for CK,FP,HH,and SH in each year were 53.56,63.97,69.83,and 80.36% in 2018,49.89,54.15,66.94,and 60.01% in 2019. respectively. CK,FP,HH,and SH averaged over 2017–2019 achieved 54.26,58.76,65.77,and 71.99% of the yield potential,respectively.

3.2.Yield component changes of summer maize in the north region of HM

By analyzing the data of the factorial experiments,we found that maize yield was significantly correlated with the ear number ha–1,kernel number per ear and 1 000-kernel weight(Fig.2). As shown in Fig.2-A,a non-linear relationship existed between the ear number ha–1(Y) and maize yield (X) (Y=–0.0632X2+1.3814X–0.0944). As the maize yield increased,ear number ha–1first increased and then decreased. The greatest ear number ha–1was 74 541 ears ha–1,while the yield in that case was not the highest,which indicated that the highest density does not result in the highest yield. With regard to the kernel number per ear and 1 000-kernel weight,linear relationships existed between each of them and the maize yield (Fig.2-C and D). According to the linear regressions between the kernel number per ear,1 000-kernel weight and maize yield,the kernel number per ear and 1 000-kernel weight increased by 20.93 kernels and 10.58 g,respectively,with each 1 t ha–1increase in the maize yield.

3.3.Effects of management factors on reducing the yield gap from FP to HH

Under the planting mode of FP,increasing the fertilizer input,plant density and irrigation separately increased maize yield significantly (Table 6). After increasing the fertilizer input,plant density and irrigation separately,the yied gaps were 1 108.64 kg ha–1(11.06%),1 257.04 kg ha–1(13.24%)and 1 177.64 kg ha–1(11.59%),respectively,in 2018; and 2 101.05 kg ha–1(24.07%),2 032.06 kg ha–1(20.88%) and 0 kg ha–1(0%),respectively,in 2019. After changing the hybrids,maize yield increased by 838.56 kg ha–1(8.65%)in 2018 and 1 206.04 kg ha–1(12.39%) in 2019,which were both significantly higher than the corresponding maize yields of the standard technology of FP.

Fig.1 Changes of the yield gap between different management treatments. CK,basic productivity; FP,farmer practice; HH,high yield and high efficiency level; SH,super high yield level. Bars are SD (n=3). Different letters represent significant differences at the 5% level.

Under the planting mode of HH,decreasing the fertilizer input,plant density and irrigation separately decreased the maize yield (Table 6). After decreasing the fertilizer input,plant density and irrigation separately,the yield gaps were 436.56 kg ha–1(3.95%),656.28 kg ha–1(6.26%)and 460.96 kg ha–1(4.42%),respectively,in 2018,but no significant differences were found. As for 2019,the yield gaps were 911.96 kg ha–1(8.71%),1 492.95 kg ha–1(12.79%)and 0 kg ha–1(0%),respectively. Compared to HH,maize yields decreased by 342.46 kg ha–1(3.17%) in 2018 and 957.36 kg ha–1(8.55%) in 2019 after changing the hybrids.

Analyzing the contribution rates of management factors(fertilizer,plant density,hybrids,and irrigation) to the maize yield increase between FP and HH,we found that plant density contributed the most,followed by irrigation,and then fertilizer,while the hybrids contributed the least,with contribution rates of 9.75,8.01,7.51,and 5.91%,respectively,in 2018. As for 2019,the plant density contributed the most with a contribution rate of 16.84%,the second was fertilizer with a contribution rate of 16.39%,the hybrids was the third contributing factor with a contribution rate of 10.47%,and the irrigation contributed the least with a contribution rate of 0. Averaging the contribution rates of 2018 and 2019,the management factors were arranged as follows:plant density (13.29%)>fertilizer (11.95%)>hybrids(8.19%)>irrigation (4%) (Fig.3).

3.4.Effects of management factors on reducing the yield gap from HH to SH

Under the planting mode of HH,with an increase in the fertilizer input,plant density increased maize yield separately,but no significant difference was found between them (Table 7). The yield gaps between HH and SH were 308.83 kg ha–1(2.85%) and 559.96 kg ha–1(5.22%) for increasing the fertilizer input and plant density,respectively,in 2018. After changing the hybrids,maize yield increased by 549.15 kg ha–1(5.18%) compared with HH in 2018,while no significant difference was found between them.

Under the planting mode of SH,after decreasing the fertilizer input and plant density separately,maize yields decreased by 77.16 kg ha–1(0.97%) and 598.79 kg ha–1(4.46%),respectively,but no significant difference was found between them. Maize yield decreased by 1 306.77 kg ha–1(12.70%) compared with the standard techonology of SH after changing the hybrid,and a significant difference was found between them in 2018 (Table 7).

Table 6 Differences in yields in absolute and percentage terms for supplemented or withheld treatments relative to the standard technology (FP (farmer practice) and HH (high yield and high efficiency level)) controls in 2018 and 2019

Table 7 Differences in yields in absolute and percentage terms for supplemented or withheld treatments relative to the standard technology (farmer practice (FP),high yield and high efficiency level (HH)) controls in 2018 and 2019

Analyzing the contribution rates of management factors (fertilizer,plant density and hybrids) to the maize yield increases from HH to SH,we found that the hybrids contributed the most,with a contribution rate of 8.94%,followed by plant density with a contribution rate of 4.84%,and the fertilizer contributed the least with a contribution rate of 1.91% (Fig.4).

4.Discussion

4.1.Yield gaps among different treatments and the impact factors of field managements

Significant gaps exist between potential yields and farmers’actual yields (i.e.,yield gaps) (Liuet al.2016; Caoet al.2019). Many studies have reported that the maize yield only achieves 53% of the yield potential in China,which is significantly lower than the 89% of developed countries(Cassmanet al.1999; Liuet al.2012; Muelleret al.2012;van Ittersumet al.2013; Chenet al.2014; Liuet al.2015).In this study,the management treatments of CK,FP,HH,and SH achieved 54.26,58.76,65.77,and 71.99% of the yield potential,respectively. Farmer practices in this study gave higher values than the average value of China,but lower than the value in Northeast China which were 60–65% (Liuet al.2012,2015; Menget al.2013; Liu B Het al.2017).This was probably because of the differences in the regions and crop models used for simulating yield potential. In this study,there were different yield gaps between the pairs CK and FP,FP and HH,HH and SH. This was mainly due to different field managements which resulted in changes in the different yield components.

Fig.4 The contribution rate of management factors (fertilizer,plant density and hybrids) to the increase of maize yield (from the yield level of high yield and high efficiency level (HH) to the yield level of super high yield level (SH)) in 2018.

Genotype is of great importance for achieving high yield of maize and reducing yield gap. Different hybrids achieve 43.6 to 98.7% of the yield potential under the same planting conditions,which indicates the importance of genotype(Liu Get al.2017). In this study,changing the hybrids significantly reduced yield gap between FP and HH,and the contribution rates of hybrids were 5.91 and 10.47% in 2018 and 2019,respectively. The contribution rate of hybrids was 8.94% for reducing the yield gap between HH and SH in 2018. Meanwhile,in 2019,lodging resulted in lower yield of SH compared with HH. This also indicated the importance of hybrids,as some hybrids were not density-tolerant. As planting density increases the light intensity within the canopy increases,especially for some less compact hybrids,and thus results in lodging (Xueet al.2016,2017a). This factor indicates that the selection and breeding of density and lodging tolerant hybrids will be very important for maize yield breakage and yield gap reduction between actual and potential yields in north Huang-Huai-Hai region.

Plant population directly limits the crop yield potential in a given environment,as the high leaf area index at high density can maximize the interception of solar radiation as early as possible in the growing season (Lobellet al.2009;Liu Get al.2017). Many studies have demonstrated that modern maize hybrids have greater yield potential due to greater tolerance to the stresses associated with a higher plant population (Duvick 1997; Hammeret al.2009; Liu Get al.2017). Increasing plant density appropriately can increase maize grain yield and thus decrease the yield gap(Liu Get al.2017). In this study,under the plant mode of FP,the plant density increased from 60 000 to 75 000 plants ha–1. The plant density contribution rate averaged 13.29%for decreasing the yield gaps in 2018 and 2019. Under the plant mode of HH,the plant density increased from 75 000 to 90 000 plants ha–1. The plant density contribution rate averaged 4.84% for decreasing the yield gap in 2018.However,there was no plant density contribution in 2019 as lodging occurred in the SH treatment. This meant that 90 000 plants ha–1was too high of a density in north Huang-Huai-Hai region which resulted in lodging risk (Xueet al.2017b). The yield gaps between FP and HH,and between HH and SH indicated that the environment tested in this study was capable of supporting plant populations of around 75 000 plants ha–1.

Nutrient (N,P and K) deficiencies are the most common and manageable abiotic stress and limiting factor of maize yield globally (Muelleret al.2012). Proper crop fertilizer management is therefore needed to reduce the yield gap(Ruffoet al.2015). In this study,a reasonable increase of fertilizer from FP to HH increased maize grain yield and reduced the yield gap.The contribution rates of fertilizer were 7.51 and 16.39% from FP to HH in 2018 and 2019,respectively. However,the increase of fertilizer from HH to SH only increased maize grain yield by 192.99 kg ha–1and the contribution rate was only 1.91%. This indicated that the fertilizer use of HH was enough for the north Huang-Huai-Hai region,which was in accordance with previous studies (Chenet al.2011,2014). Overapplication of fertilizer could not increase maize yield and reduce the yield gap further,and might result in negative environmental impacts (Dinneset al.2002; Chenet al.2014; Cuiet al.2018). Besides fertilizer,the application of irrigation can reduce the maize yield gap while the function of irrigation is affected by precipitation(Liu B Het al.2017). In this study,the contribution rate of irrigation was 8.01% for reducing the yield gap between FP and HH in 2018,while there was no contribution in 2019.This difference was mainly due to the fact that precipitation was sufficient and its distribution was uniform in 2019.

4.2.Comprehensive analysis of the contribution rates of different field managements

Field management practices have different effects on reducing the yield gap (Ruffoet al.2015). A comprehensive analysis indicated that the contribution rates of different field management variables showed a decreasing trend of plant density (13.29%)>fertilizer (11.95%)>hybrids(8.19%)>irrigation (4%) from FP to HH (Fig.3). This indicates that the traditional plant density adopted by farmers in China is low,which is in accordance with previous studies(Menget al.2013; Xuet al.2017). Increasing planting density is the most effective cultivation management step for increasing the maize yield of farmers and reducing the yield gap. From HH to SH,the contribution rates of different field management practices showed a decreasing trend of hybrids (8.94%)>plant density (4.84%)>fertilizer (1.91%).This trend indicated that adopting hybrids with good density and lodging tolerance was the most effective approach for increasing yield from HH to SH. In the north Huang-Huai-Hai region,breeding hybrids with good density and lodging tolerance is a limiting factor for high yield breaking(Liu Get al.2017). The contribution rates of different field management practices for HH to SH were less than those for FP to HH,which indicated that reducing the yield gap from HH to SH is more difficult. In addition to these traditional field management variables,other factors such as growth regulators should also be considered in future studies.

5.Conclusion

Understanding yield potential,yield gap and the top priority management factors for reducing the yield gap in current intensive maize production is essential for meeting future food demands with limited resources. Significant differences were found between the maize yields of the different planting modes of CK,FP,HH,and SH. A management factor priority analysis indicated the contribution rate of planting density was the largest for reducing the yield gap from FP to HH,which meant that the low plant density of FP was the key limiting factor for increasing yield. However,the contribution rate of hybrids was the largest for reducing the yield gap between HH and SH,which meant that hybrids with good density and lodging tolerance were the limiting factor for achieving higher yield. Analysis of management factor priorities will help farmers in this region adopt suitable field management practices for increasing maize yield and reducing the yield gap.

Acknowledgements

We thank the National Key Research and Development Program of China (2016YFD0300106) and the National Natural Science Foundation of China (31601247) for their financial support.