Effects of different mechanical direct seeding methods on grain yield and lodging resistance of early indica rice in South China

Key Laboratory of Crop Physiology,Ecology and Genetic Breeding,Ministry of Education/Collaborative Innovation Center for the Modernization Production of Double Cropping Rice,College of Agronomy,Jiangxi Agricultural University,Nanchang 330045,P.R.China

Abstract Direct seeding of rice has become a main planting method due to the low labor input and high economic benefit in South China.Dry direct seeding (DDS) has been widely used for single-season rice planting establishment.However,few studies have examined the performance of early-season indica rice under mechanical dry direct seeding.A two-year field experiment was conducted with two indica rice cultivars (i.e.,Zhongjiazao 17 and Zhuliangyou 819) to study lodging characteristics and grain yield formation under DDS,flooded direct seeding (FDS) and wet direct seeding (WDS) patterns.The results showed that the annual grain yield in DDS was higher by 14.42–26.34% for cultivar ZLY819 and 6.64–24.58% for cultivar ZJZ17 than in WDS and FDS,respectively,and these increases were mainly attributed to the improvement of the panicles.The DDS pattern significantly increased the seedling emergence rate of early indica rice cultivars,and increased total dry weight and crop growth rate.Meanwhile,shorter basal internodes,better stem diameter and stem wall thickness and lower lodging index were found in DDS in contrast to FDS and WDS.In particular,DDS improved the stem lodging resistance.Our results suggested that the appropriate direct seeding method was beneficial for improving the grain yield and lodging resistance of early indica rice.

Keywords:mechanical dry direct seeding,early indica rice,grain yield,lodging resistance,South China

1.Introduction

Rice (Oryza sativaL.) is the staple food for about half of the world’s population (Khush and Gupta 2013),and China is the largest rice producing country in the world.Thus,Chinese rice planting methods play a crucial role in strengthening global food security (Liuet al.2010;Xuet al.2016).High labor input has reduced farmers’ enthusiasm for growing transplanted rice,and high labor input in the traditional system is primarily due to seedling raising and transplanting (Peng 2014;Sunet al.2015).Nowadays,with the shortage of labor and the change of production pattern,direct seeded rice has become a main planting method replacing transplanted rice.Previous studies showed that direct seeding could save about 25% on labor inputs compared with the transplanting method (Bhushanet al.2007;Sanusanet al.2010;Liuet al.2015).Recently,the planting area of direct seeded rice has increased rapidly.By the end of 2015,the planting area of direct seeded rice in Jiangxi Province,China had reached at 420 000 hectares,and the proportion of direct seeded area of earlyindicarice was as high as 60.8%.Earlyindicarice is one of the most important food crops for China,especially in South China,where it has been widely cultivated from March to July.From 2005 to 2014,the direct seeded planting area and grain yield of earlyindicarice accounted for about 19.1–20.9%and 16.3–17.6% in China’s rice production,respectively(Tanet al.2018).However,whether direct seeding can obtain the high and stable yield for earlyindicarice has been controversial in production.

In direct seeding,seeding method is an important factor for growth and development,and yield formation of direct seeded earlyindicarice.The grain yield of direct seeding method was generally affected by tillage system,weed infestation,mechanical condition,and water management factors (Andoet al.2000;Ojha and Talukdar 2000;Furuhataet al.2010).A suitable direct seeding method was of great significance to improve the emergence and grain yield of direct seeded rice (Bazayaet al.2009;Xuet al.2014b;Wanget al.2020).It was reported that the grain yield of direct seeded rice was comparable to that of transplanted rice,and the crop productivity of direct seeded rice of double-season cropping would not be reduced under optimal crop management practices (Catonet al.2002;Polthaneeet al.2008;Wang Z Met al.2016).The grain yield of hill direct seeded rice was significantly higher than that of drill direct seeded,which might be related to a higher productive tiller percentage (Xuet al.2014a).The occurrence of weeds is an important factor affecting the yield of direct seeded rice.Previous studies proposed that early moderately dry conditions after direct seeding could reduce the weed occurrence in paddy fields;while flooded direct seeding (FDS),due to a longterm water barrier,was more likely to breed water borne weeds,resulting in reduced grain yield of direct seeded rice (Wanget al.2018;Xuet al.2018).Meanwhile,mechanical hill-direct-seeding technology,invented by the South China Agricultural University,is an efficient direct seeding method that could not only ensure the orderly and even growth of seedlings,but also increase the seedling emergence rate and obtain high and stable yield (Shenet al.2013;Baimbaet al.2014).Recently,a mechanical hill dry direct seeding method has also developed rapidly for rice plant production in South China,especially inindicacultivar planting systems.

Lodging is one of the most important factors for grain yield potential of direct seeded rice,since it can result in yield loss,worsening grain quality and reduced efficiency of mechanical harvesting as well (Kashiwagi and Ishimaru 2004;Yadavet al.2017).The lodging of direct seeded rice was influenced by breed heredity,natural environment,planting method,and water management (Yoshinagaet al.2003;Takayukiet al.2010).The differences in lodging resistance were closely related to the characteristics of panicle,and physical and mechanical characteristics of rice plant stem (Liuet al.2018).It was generally believed that increasing the plant height,decreasing the lodging resistance ability of rice and replenishing stem wall thickness and internodes could be negatively correlated with lodging index.However,some studies have shown that plant height is not the determining factor of lodging,as increasing the plant height does not always reduce the lodging resistance of stems (Berry and Berry 2015).The differences in these results might be related to the cultivar characteristics (Duanet al.2004).

At present,many scholars have studied the cultivation techniques of direct seeding technology forjaponicarice or single season rice (Joshiet al.2013;Wang Cet al.2016;Panet al.2017).However,there are few studies on the effects of mechanical dry direct seeding planting on lodging characteristics and grain yield of earlyindicarice in South China,which is only widely planted in the oneseason rice production in the Yangtze River Basin (Yanet al.2010).Additionally,water management plays an important role in rice production.Previous studies showed that dry–wet alternating irrigation increased yield by 3.48%compared with flood irrigation,which was beneficial to the improvement of lodging resistance of plants (Wuet al.2019).However,the effects of different direct seeding methods with regard to soil water management measures on rice yield and lodging characteristics are rarely reported.Hence,we hypothesized that a mechanical dry direct seeding method could increase grain yield and enhance lodging resistance.To test our hypothesis,a two-year field experiment was conducted to investigate the effects of mechanical dry direct seeding methods with different soil water management practices on grain yield and its components,emergence rate,total dry weight,harvest index,stem morphology,and mechanical properties of earlyindicarice.Our major objectives of this study were to determine the appropriate direct seeding method and to elucidate the effects of direct seeding method on grain yield and lodging resistance of earlyindicarice.

2.Materials and methods

2.1.Rice materials

In this study,two earlyindicarice cultivars Zhongjiazao 17(ZJZ17) and Zhuliangyou 819 (ZLY819),which are mainly generalized for rice production in Jiangxi Province,were selected for the field experiment.ZJZ17 was originally bred by the China National Rice Research Institute (CNRRI),and is an inbredindicarice with an average entire growth period of 114.3 days.ZLY819 was bred by the Hunan Yahua Institute of Seed Science,and is a hybridindicarice with an average entire growth period of 105.4 days.

2.2.Experimental treatments and cultivation methods

This study was conducted at the Agriculture Science and Technology Station (ASTS) of Jiangxi Agricultural University during 2016 and 2017 in Nanchang City,Jiangxi Province,China (28°46´N,115°50´E,48.8 m a.s.l.,annual average temperature:17.5°C,average annual sunshine:1 720.8 h,annual average evaporation:1 139 mm,and average annual rainfall:1 747 mm).The field experiments were conducted with a split plot design,with the planting establishment of experimental treatments used as the main plot,and the cultivars as subplots (length and width:5 m×4 m).Three main-plot seeding methods were divided into dry direct seeding (DDS),FDS and wet direct seeding (WDS).The experimental treatments had four replications.The DDS plots were treated with dry rotor tilling after harvesting of the previous crop and no excess water was maintained in the paddy field until sowing of the later early-season rice seeds.After sowing,the soil moisture content was monitored by a moisture tester (OK-TSS1) and maintained at 40–50%;the plough layer soil of paddy field was covered with fine soil grains,and the paddy surface was not irrigated for the dryland layer until the third-leaf emergence of the rice seedlings.Then the paddy field irrigation was conducted with the alternate irrigation of shallow dry and wet during rice growth.FDS and WDS plots were both subjected to water rotor tilling.For FDS plots,the depth of soil water layer was kept at about 20–30 mm from sowing to emergence,and the paddy field was treated with the flooded irrigation during rice growth.For WDS plots,the plots were ploughed into a soft and smooth mud seedling bed,and the soil moisture content of paddy was kept saturated but no water layer was accumulated.

The experimental field was a winter fallow cropland.The soil fertility properties across the years were as follows:total nitrogen 2.03 g kg–1,soil organic matter 27.35 g kg–1,available nitrogen 94.52 mg kg–1,available phosphorus(P) 24.78 mg kg–1,available potassium (K) 99.42 mg kg–1,and pH 5.52.The rice seeds of the two cultivars were both sown on April 5 in the two years using mechanical hill direct seeding according to the sowing method of rice precision hole direct seeding machine (2BX-10,invented by South China Agricultural University).The germinated seeds were selected for seeding,and approximately six to eight seeds were seeded per hill.The row spacing of seeding was 16.5 cm×16.5 cm.The plot size was 20 m2for both years.The fertilizer management was the same in both years.Nitrogen (N),phosphate (P2O5) and potassium(K2O) fertilizers were applied at 165,90 and 165 kg ha–1,respectively,in the form of urea,calcium magnesium phosphate and potassium chloride.Fertilizer-N was applied over the whole plant growth period at three times (82.5 kg N ha–1at basal application,33.0 kg N ha–1at mid-tillering stage and 49.5 kg N ha–1at panicle initiation).Fertilizer-K2O was applied at the mid-tillering stage (115.5 kg K ha–1)and panicle initiation (49.5 kg K ha–1).Fertilizer-P2O5was all applied as a basal fertilizer.In all plots,pre-emergence herbicides and hand weeding were used to control weeds.Pesticides and fungicides were sprayed to control pests and diseases if needed.

2.3.Yield and yield components

At maturity,rice grain collected in a sampling area of 5 m2in each sub plot was used to evaluate grain yield,and adjusted to 13.5% safe moisture content by weight.The specific steps of measuring grain yield components were as follows:the 1 000-grain weight was calculated from 1 000-grain weight with three replications;the panicles,grain filling and panicles(containing the number of full grains and the number of empty grains) were measured following the methods of Xuet al.(2010).

2.4.Seedling emergence rate

The seedling emergence rate was determined after 15 days of direct seeding and calculated by the eq.:Seedling emergence rate=Number of seedlings emerged/Number of seeds sown

2.5.Total dry weight (TDW)

In both years,five hills of plants were sampled from each subplot at the tillering stage,panicle initiation,heading stage,and maturity stage to investigate the dry weight,and then all plants were divided into leaf,stem-sheath and panicle.After the drying process at 105°C for 15 min,the dry weight of each part was determined after oven-drying at 80°C for 72 h to achieve a constant weight.Plant dry weight of all five hills parts was calculated according to the biomass of each growing stage.The crop growth rate (CGR) of each treatment was calculated by the eq.:CGR=(TDW2–TDW1)/(T2–T1),where TDW1 and TDW2 represent the aboveground total dry weights at the times T1 and T2,respectively.

2.6.Harvest index (HI)

The HI was calculated by the eq.:HI (%)=Grain weight/Total dry weight×100

2.7.Lodging determination

Lodging-related traits were determined at 20 days after heading stage in 2016 and 2017.Ten representative main stems for the three replications were sampled to measure the related traits in each sub-subplot.The main stem morphological parameters included the plant height,panicle length and each internode (I1,I2 and I3,for the first,second and third internodes above the root,respectively).The morphological characteristics of the basal internode,including diameter of culm,stem wall thickness and culm type index,were measured by Vernier caliper (Mitutoyo-530-108,Shenzhen Xinlangpu Electronic Technology Co.,Ltd.,China),and they were measured at the second internode.The breaking stress of the penultimate internode stem was measured by a stem strength testing machine YYD-1 (HM-ZG02,Shandong Hengmei Electronic Technology Co.,Ltd.,China).The other related parameters were calculated according to previously published methods(Islamet al.2007;Ookawaet al.1992;Zhanget al.2014)as follows:

Plant height=Length from the soil base to the top of the main spike

Stem diameter=(Long axis diameter+Short axis diameter)/2

Stem wall thickness=Axis outer diameter–Axis inner diameter

Culm phenotype index=Culm base diameter/Culm length

Bending weight (g cm)=Fresh plant weight of the plant part above the broken point×Length from broken point to panicle tip

Lodging index (LI,%)=Bending weight/Break force×100

2.8.Statistical analysis

The data were analyzed by analysis of variance (SAS Institute,Inc.,Cary,NC,USA),and the means of the direct seeding methods were examined by Tukey’s honest significant difference (HSD) test to compare the differences at the probability level of 0.05.

3.Results

3.1.Yield and yield components

Analysis of variance results for yield and yield components are presented in Tables 1 and 2.For grain yield,only the interactive effect among year (Y),cultivar (C) and seeding method (SM) was not significant.Panicles were significantly affected by Y,C,SM,Y×SM,and Y×C×SM,while seed setting rate was significantly affected by Y,C,SM,and Y×SM.All the factors had significant effects on spikelets per panicle except for Y×C.Y,C,SM,and the interactive effect among C×SM had a significant effect on 1 000-grain weight.

The seeding method demonstrated significant differences in grain yield among the years and cultivars (Table 1).Overall,the grain yield of DDS was significantly higher than those of WDS and FDS.FDS had the lowest grain yield among the seeding methods.The grain yield in DDS was higher by 14.42–26.34% for cultivar ZLY819 and 6.64–24.58% for cultivar ZJZ17 than those in WDS and FDS,respectively.Significant differences in the TDW and HI were also observed across seeding methods in all years and cultivars,except for ZLY819 in 2016.DDS produced a higher TDW than those of the FDS and WDS.Compared with FDS,DDS had average increases in the TDW by 13.94% for ZLY819 and 14.51% for ZJZ17,respectively,and the difference reached a significant level.There was a significant relationship between grain yield and TDW across the earlyindicarice in 2016 and 2017 (Fig.1-A).DDS had higher HI in the two years and cultivars with a significant difference compared with FDS,except for ZLY819 in 2016.

The different direct seeding methods demonstrated significant differences in yield components for all cultivars and years (Table 2).The panicles in DDS were the highest among the three seeding methods.However,the spikelets per panicle,grain filling and 1 000-grain weight of DDS were lower than those of WDS and FDS in the two years and cultivars with the significant differences compared with FDS.For the earlyindicarice,the higher yield in DDS than in WDS and FDS was mainly attributed to the increasing panicles.

3.2.Seedling emergence rate

For the cultivars and years,the seedling emergence rate was higher in DDS than in WDS and FDS (Fig.2).Compared with FDS,DDS and WDS significantly increased seedling emergence rates regardless of the cultivars or years.However,the seedling emergence differed between WDS and DDS.

3.3.Crop growth rate (CGR)

The CGR from tillering to maturity stage was responsiblefor the high TDW and grain yield of DDS (Table 3).The CGR from tillering to panicle initiation did not show a consistent difference between the two cultivars.The DDS had a higher yield than those of the FDS and WDS,which was attributed to the CGR from heading to maturity stages in the earlyindicarice (Fig.1-B).For example,CGR in DDS from heading to maturity stages were higher by 7.52–11.68% for ZLY819 and 9.00–25.62% for ZJZ17 than those in WDS and FDS,respectively.

Table 1 Effect of different direct seedling methods on grain yield,total dry weight (TDW) and harvest index (HI) of early indica rice

Table 2 Effect of different direct seeding methods on yield components of early indica rice

Fig.1 The relationships of yield with total dry weight (TDW) (A) and crop growth rate (CGR) from heading to maturity stage (B).The data were combined from the experiments of 2016 and 2017 in the early indica rice cultivars.

Fig.2 Effects of different direct seeding methods on seedling emergence rate of early indica rice.ZLY819,Zhuliangyou 819;ZJZ17,Zhongjiazao 17.FDS,flooded direct seeding;WDS,wet direct seeding;DDS,dry direct seeding.Bars indicate SE(n=3).Different lowercase letters indicate significant differences at the level of 0.05.

3.4.Main stem morphology

Year (Y),cultivar (C),seeding method (SM),and most their interactions had significant effects on internode length(Table 4).Plant height was significantly affected by C,SM and C×SM,while panicle length was significantly affected by Y,C and SM.In the two years,plant height was the lowest in DDS among the seeding methods,but no difference was found between FDS and WDS.The panicle length had the same trends.The three internode lengths for the stem of DDS were lower than those of WDS and FDS with significant differences in contrast to FDS.The relatively higher amplitude of I3 was decreased for I1 and I2 in DDS,especially for ZJZ17.

Stem wall thickness and culm phenotype index were significantly affected by C and SM,while stem diameter was significantly affected by Y,C and SM (Table 5).Among the years and cultivars,DDS had the highest stem diameter,stem wall thickness and culm phenotype index,but their values for FDS were the lowest among the seeding methods.The significant differences for all indexes,as affected by the seeding methods,were seen in the ZLY819 for the two years.

3.5.Mechanical properties of the stem

Y,C,SM,and their interactions each had a significant effect on bending weight (Table 6).Break force was significantly affected by C and SM,while lodging index was significantly affected by Y,C,SM,C×SM,and Y×C×SM.The bending weight and lodging index of DDS were lower than those of FDS and WDS in the two years and cultivars,with significant differences compared with FDS.However,the trends in break force among the seeding methods were opposite in contrast with the bending weight and lodging index.

4.Discussion

4.1.Effects of different mechanical direct seeding methods on the growth and grain yield of early indica rice

Finding the suitable soil water management measure for direct seeding method could improve the grain yield of riceand coordinate its yield components (Chenget al.2010;Yanet al.2010;Sunet al.2014).Dry direct seeding with low soil moisture content in paddy fields had a higher benefit such as cost ratio than other direct seeding methods (Aliet al.2006).In this study,DDS could significantly increase the grain yield between the two years and cultivars,which was primarilydue to the panicles.This was consistent with the results of Sunet al.(2014) that showed mechanical dry direct seeding modes improved growth,development and yield of rice.In general,the panicle type of direct seeded rice is relatively small as the spikelets per panicle are few (Shenet al.2014).In this study,DDS had the most panicles among the seeding methods,though the spikelets per panicle,grain filling and the 1 000-grain weight were relatively low.This might be due to better seedling emergence rate and productive tiller percentage.In addition,the early moderate dry before third-leaf stage condition could effectively increase the root elongation to the deeper soil,and improve the aboveground

emergent seedling growth that was conducive to improving seedling to early tiller stage (Yin and Xie 2010;Singhet al.2013).This direct seeding method suggested the speedy growth advantages for the short growth duration in earlyindicarice.Our study further demonstrated that maintaining a high CGR and TDW was crucial for increasing the biomass production and grain yield in direct seeded rice.

Table 3 Crop growth rate (g m−2 d–1) of different direct seedling methods during each growth stage in the early indica rice of 2016 and 20171)

Table 4 Effects of different direct seedling methods on main stem morphology

Table 5 Effects of different direct seeding methods on stem characteristics

Table 6 Effect of different direct seeding methods on mechanical properties of stem

Seedling emergence has always been the primary factor for restricting the high and stable yield of direct seeded rice,which directly affects the population growth of rice plants (Chung 2010).In the present study,we found that DDS could increase the seedling emergence in favor of more uniform and vigorous seedling growth.This might be due to the fact that the seeds were sown in the loose and permeable soil layer environment of DDS with relatively low redox potential and sufficient oxygen in the soil (Farooqet al.2011),and this could promote seed germination and rapid growth of seedlings.Inversely,when seeds were sown under flooded layer or saturated water soil environments,they would undergo hypoxic stress,and the long-term water barrier would easily lead to rotting seeds and seedlings,affecting seedling emergence (Hiromiet al.2014).

4.2.Effects of different mechanical direct seeding methods on lodging resistance of early indica rice

Lodging has been observed more frequently in direct seeded rice than traditional manual transplanted rice in recent years(Berryet al.2004;Sinniahet al.2012).Lodging results in yield reductions due to a decrease in photosynthesis by selfshading,and then reduced rice quality (Setteret al.1997;Wonet al.1999).In addition,mechanical harvesting is also extremely cumbersome.Previous studies had proposed that other than genetic differences among cultivars the lodging occurrence was closely related to heavy rainfall,strong wind and plant type (Ishimaruet al.2008;Corbinet al.2016).Our results offer the scope to improve the lodging resistance of earlyindicarice through dry direct seeding methods.Stem lodging risk increase might be associated with plant height;when the plant height and basal internode were shorter,its lodging resistance was stronger(Kimet al.1993;Quanget al.2004;Farooqet al.2011).Our study has also confirmed these results,suggested that DDS improved stem lodging,but the three internode lengths of the cultivars decreased differentially by DDS.Among lodging related traits,break force and lodging index of lower base internodes had the closest relationships with lodging resistance at the paddy field condition (Islamet al.2007).In this study,the break force was enhanced and the lodging index was reduced under DDS,which improved the stem lodging resistance.Additionally,lodging related traits were only measured at the second internode for this study,because stem lodging usually occurred at lower base internodes (Islamet al.2007).Meanwhile,the mechanical dry direct seeding increased the stem diameter,stem wall thickness and culm phenotype index at the base of the plant,which implied an improvement in the lodging resistance and contributed to high grain yield.To summarize this study,although there were significant differences in plant lodging resistance among the seeding methods,no apparent lodging phenomenon was found in the paddy field,which might be related with the fine weather conditions after heading.

In this study,the significant effects of year and cultivar,and the interactive effects between seeding method×year,seeding method×cultivar,and year×cultivar on the yield and lodging related parameters of direct seeded earlyindicarice were also observed,implying that annual ambient environmental factors and genetic characteristics of cultivars could also contribute to high and stable grain yield under direct seeded conditions in earlyindicarice,in addition to the direct seeding methods.Therefore,we should pay more attention to the consideration of the ambient environment such as sowing date and preferentially select the lodging resistant cultivars based on the locally appropriate direct seeding methods when employing direct seeded rice production in the future.

4.3.High yield cultivation methods of direct seeded early indica rice

Direct seeded rice was influenced by mechanical equipment,water management and the ecological environment (Andoet al.2000).Our results showed that the high TDW and CGR were critical for the yield formation of direct seeded rice.Furthermore,the suitable direct seeding method selected for water management measures with appropriate soil moisture content was beneficial for improving lodging resistance,especially for precision rice hill-drop drilling machine seeding.Lodging conditions were different among the cultivars,ecological environment due to the genetic factors of cultivars and management measures.Breeding new rice cultivars for direct seeded,earlyindicarice in South China should aim for high TDW and CGR with desirable plant traits of short internode lengths and strong lodging tolerance.Moreover,in rice production,reasonable chemical control measures that applied growth regulators,such as the paclobutrazol or Mengfadou (i.e.,a strong stem medicament),should be taken to reduce the stem internode lengths,to enhance lodging resistance and increase grain yield of direct seeded rice.

5.Conclusion

The grain yields of earlyindicarice cultivars in South China were significantly increased by DDS compared with FDS or WDS,mainly due to the panicles.The DDS method significantly increased the seedling emergence rate of earlyindicarice,and improved TDW and CGR.In addition,compared with FDS and WDS,DDS had shorter base internodes,better stem thickness and stem wall thickness,and lower lodging index.In particular,DDS could increase the stem lodging resistance.

Acknowledgements

The authors gratefully acknowledge the financial support provided by the National Natural Science Foundation of China (31760366),the National Key Research and Development Program of China (2016YFD0300501 and 2017YFND0301605) and the Natural Science Foundation of Jiangxi,China (20161BAB214171).

Declaration of competing interest

The authors declare that they have no conflict of interest.