Differences in rice yield and biomass accumulation dynamics for different direct seeding methods after wheat straw return

Abstract

Field experiments were performed in 2019 and 2020 to explore yield performance and biomass accumulation dynamics for different direct‐seeded rice methods after wheat straw return. The experimental treatments included two direct‐seeded rice methods (dry direct‐seeded rice and wet direct‐seeded rice) and two wheat straw return treatments (with and without wheat straw return). Wheat straw return had a negative impact on the yield and biomass of direct‐seeded rice. Compared with wet direct‐seeded rice, dry direct‐seeded rice was more negatively affected, with yield and biomass reduced by 6.7%–7.6% and 5.4%–6.5%, respectively. The yield loss was mainly associated with insufficient total spikelet numbers caused by reduced panicle numbers. A lower maximum tiller number caused by the restricted seedling tillering capacity was partly responsible for the inadequate panicle numbers in direct‐seeded rice after wheat straw return. Wheat straw return reduced the maximum and average rate of biomass accumulation, shortened the duration of effective biomass accumulation, and delayed the days to achieve the maximum rate of biomass accumulation for direct‐seeded rice. The negative effect of wheat straw return on direct‐seeded rice was concentrated mainly in the early stage; reduction in the average biomass accumulation rate in the early stage accounted for 61.1%–75.3% of the reduction in total average rate. Wet direct‐seeded rice had a higher average rate and amount of biomass accumulation at the early stage, a greater maximum tiller number, as well as a stronger photosynthetic ability and grain filling ability after heading, which contributed to improving panicle number and total spikelet number, ultimately increasing yield and biomass compared with dry direct‐seeded rice. These results demonstrate that wheat straw return reduced yield and biomass, mainly by inhibiting growth in the early stage of direct‐seeded rice. However, the wet direct‐seeding method decreased the losses in yield and biomass after wheat straw return.

INTRODUCTION

Rice is a staple food and nutritional source for about half of the world's population (Chauhan, 2012). Transplanted rice is the conventional method for rice establishment in Asia (Liu et al., 2015). However, this method requires high labor inputs for raising, pulling, and transplanting of seedlings. Recently, challenges such as escalating labor costs (Shi et al., 2020) have driven a change in rice establishment method from transplanted rice to direct-seeded rice (Farooq et al., 2011; Ohno et al., 2018; Wang et al., 2017). The development of multifunctional seeders (Qin et al., 2017; Sidhu et al., 2015) and improvements in weed management technology (Chauhan & Opeña, 2013; Matloob et al., 2015) have also provided opportunities for expanding the cultivation area of direct-seeded rice.

Rice–wheat rotation is an important agricultural production system in Asian countries such as China and India (Ladha et al., 2003; Prasad, 2005). In this system, wheat straw management is a challenging problem owing to the short turnaround time after combining the harvesting of wheat (Bijay et al., 2008). The recycling and open-burning of wheat straw are customary practices (Li et al., 2018), but they cause a series of problems such as rising labor costs and environmental pollution. Therefore, wheat straw return is an important agricultural measure that is widely practiced in the rice–wheat rotation system (Huang et al., 2013; Liu et al., 2014). Direct-seeded rice coupled with wheat straw return is the simplest and most convenient method for rice planting in the rice-wheat rotation system. However, wheat straw return can cause negative effects, such as an accumulation of microbial allelochemicals and a reduction in soil available nitrogen during wheat straw decomposition, which restrict the growth of rice (Gao et al., 2004; Yu et al., 2013). On the other hand, wheat straw return can improve soil structure and increase soil organic matter content, which promote the growth of rice (Jiang et al., 2012; Zhang et al., 2021). Most studies have also shown that wheat straw return can increase the yield of transplanted rice (Huang et al., 2013; Sharma et al., 2021; Yang, Luo, et al., 2020; Zhang, Hang, et al., 2017). However, unlike transplanted rice seedlings that already possess some level of stress resistance at transplant from the seedling cultivation area to the main field, direct-seeded rice seedlings are always under stress from seed germination to seedling growth, leading to poor seedling establishment. The differences between direct-seeded rice and transplanted rice mean that direct-seeded rice may not show increased yields after wheat straw return. The yield performance of direct-seeded rice after wheat straw return in the rice–wheat rotation system therefore requires further evaluation, confirmation, and data support.

Planting patterns and tillage methods are important factors that affect rice growth and yield formation (Xu et al., 2010; Yuan et al., 2013). Dry direct seeding and wet direct seeding are the two main planting methods for direct-seeded rice in the rice–wheat rotation system in south Asia (Bhatt & Kukal, 2015; Kumar & Ladha, 2011). In dry direct seeding, rice seeds are sown into soil that has been dry ploughed and harrowed without puddling. In wet direct seeding, pre-germinated seeds are sown into puddled soil. The selection of an appropriate direct seeding method is important for increasing the yield of direct-seeded rice (Bazaya et al., 2009; Wang et al., 2021).

Biomass is the final product of crop photosynthesis, and biomass accumulation directly determines the level of crop yield (Kumar et al., 2006). Quantitative analysis of biomass accumulation dynamics using model equations is an important method for investigating crop yield formation. Related studies have been reported in crops such as rice (Ji et al., 2012; Li et al., 2010), wheat (Villegas et al., 2001), cotton (Overman & Scholtz, 2013), maize (Meade et al., 2013), and soybean (Nakano et al., 2021; Van Roekel & Purcell, 2014). For example, Wei (Wei et al., 2021) quantitatively analyzed the biomass accumulation dynamics of different rice variety types after transplant based on the Gompertz equation. They concluded that japonica/indica hybrids had superior biomass accumulation compared with conventional japonica and indica hybrids in the early, middle, and late stages. Ma et al. (2021), taking the effective accumulated temperature as the time scale, simulated biomass accumulation of maize using the Richards equation and quantitatively analyzed dynamic changes in biomass accumulation to provide guidance for achieving high maize yields. However, no studies to date have used model equations to quantitatively analyze differences in biomass accumulation characteristics for different direct-seeded rice methods after wheat straw return.

Here, the biomass accumulation characteristics of different direct-seeded rice methods after wheat straw return were compared, and differences between dry direct-seeded rice and wet direct-seeded rice were quantitatively analyzed. The objectives of this study were (1) to evaluate the yield performance and biomass accumulation characteristics of rice for different direct seeding methods after wheat straw return, (2) to clarify the mechanisms that underlie differences in yield formation and biomass accumulation, and (3) to determine the most suitable direct seeding method for rice after wheat straw return. The findings of this study provide theoretical and practical guidance for high-yield cultivation and regulatory approaches for direct-seeded rice after wheat straw return.

MATERIALS AND METHODS Experimental site and weather conditions

The field experiment was performed in 2019 and 2020 in a typical rice–wheat rotation system at the research farm of Yangzhou University, Jiangsu, China (32°61′N, 120°12′E). The field soil was a sandy loam with a viscous texture and 30.4 g kg⁻¹ organic matter, 1.91 g kg⁻¹ total N, 31.6 mg kg⁻¹ available P, and 154 mg kg⁻¹ available K. Meteorological data for daily mean temperature, sunshine hours, and precipitation during the 2019 and 2020 rice growing seasons were collected at a weather station near the experimental site (Figure 1).

FIGURE 1. Daily mean temperature, sunshine hours, and precipitation during the rice growth season in 2019 and 2020. The 0 days after sowing is June 11 for both years.

Experimental design and treatments

The experiment was performed using a split-plot design with four replications. Main plots were the wheat straw return treatment methods: control without wheat straw return (S0) and wheat straw return (S1). Split plots were the two direct-seeded rice methods: dry direct-seeded rice (M1) and wet direct-seeded rice (M2). The high-quality japonica rice variety Nanjing-9108 was used in this experiment.

In both years, wheat was harvested with a Kubota combine [4LBZ-145G (PRO588I-G)], and wheat straw was returned to the field (7.7 t biomass ha⁻¹ and 8.3 t biomass ha⁻¹ in 2019 and 2020, respectively). The date for wheat straw return in both 2019 and 2020 was 9 June. The wheat straw was removed manually from the main plots that did not receive returned wheat straw. Soil in the dry direct-seeded rice plots was dry without water puddling, and ungerminated dry rice seeds were seeded with a multifunctional seeder that performs synchronous rotary tillage and sowing (Yangzhou University). The row spacing of drill sowing, depth of rotary tillage, and seeding rate were 25 cm, 10–15 cm, and 70 kg ha⁻¹, respectively. Unlike dry direct seeding, wet direct seeding involved first soaking the seeds in water for 20–24 h and then incubating them for 8–12 h. The pre-germinated seeds were then seeded on the surface of drained and puddled soil with a rice hill-drop sowing machine (South China Agricultural University). The depth of rotary tillage was 20–25 cm for this paddy soil. The hill seeding density and seeding rate were 25 × 11 cm and 70 kg ha⁻¹, respectively. At the three-leaf stage, four representative plot areas of 36 m² were selected, and the seedlings in each plot were thinned to 150 m⁻². In both years, the sowing date for dry direct-seeded rice was June 11, and the harvest date was October 20. The sowing dates for wet direct-seeded rice were June 13 in 2019 and June 11 in 2020, and the harvest dates were October 22 in 2019 and October 21 in 2020.

Crop cultivation

After sowing, wet irrigation management was used during the seedling period to ensure the growth and development of the rice seedlings. Thereafter, moist conditions were maintained until the five-leaf stage. The field was flooded after the five-leaf stage, and the water level was maintained at 2–3 cm until the middle tillering stage. Water was then drained for 7–10 days to control unproductive tillers. After the stem elongation stage, an alternate wetting and drying irrigation regime was used until 1 week before the final harvest.

In all plots with N application treatment, urea N fertilizer (45.6% N, 270 kg N ha⁻¹) was applied in three splits in a ratio of 3.5:3.5:3 at pre-sowing, the four-leaf stage, and the panicle initiation stage. In all plots, the P fertilizer (P₂O₅; 135 kg P ha⁻¹) was applied at pre-sowing, and the K fertilizer (K₂O; 135 kg K ha⁻¹) was applied at pre-sowing and again at the panicle initiation stage (50 DAS).

Weed, insect, and disease control followed local recommendations throughout the growing season to minimize yield loss over the 2 years.

Sampling and measurements Tillering dynamics

To accurately sample and understand the growth process of rice, we recorded the dates of sowing, the stem elongation stage, the heading stage, and the maturity stage. In 2018, three adjacent and consecutive rows (1 m length) in each plot were measured at 5-day intervals from the three-leaf stage until the tiller number diminished.

Tiller number ($Y$) was fitted with a logistic equation:[Image Omitted. See PDF]where $Y_m$ is the maximum tiller number, x is days after sowing, and A and B are rate-controlling parameters (Caton et al., 2003). The fitting was performed using Origin 9 (Origin Lab Corporation).

Biomass accumulation

Biomass was determined every 10 days after sowing until maturity. Samples were collected from three adjacent rows (50 cm length) in each plot, oven-dried at 105°C for 30 min to deactivate enzymes, and dried at 80°C in bags to a constant weight for dry weight measurement. At the heading and maturity stages, the samples were separated into leaves, stem-sheaths, and panicles. Each component of the rice plants was oven-dried separately to determine the weight.

Translocation quantity, translocation efficiency, and contribution rate of stem-sheath and leaf for direct-seeded rice after heading stage can be calculated as proposed by Papakosta and Gagianas (Papakosta & Gagianas, 1991).[Image Omitted. See PDF][Image Omitted. See PDF][Image Omitted. See PDF]

Yield and yield components

At the maturity stage, three adjacent rows (1 m length) were sampled randomly from each plot for measurement of panicle traits and yield components, including spikelet number per panicle, filled-grain percentage, and 1000-grain weight. Three adjacent rows (5 m × 75 cm) were sampled randomly from each plot to determine the panicle number per m². Total spikelet number per m² was calculated as panicle number per m² multiplied by spikelet number per panicle. Yield was determined from a harvest area of 8 m² in each plot and adjusted to the standard moisture content of 0.14 g H₂O g⁻¹.

Model analysis and formula calculation

In this study, Origin 9 software was used to fit the biomass accumulation after sowing for each treatment. The Richards equation had a better fit coefficient and was therefore selected for the simulation of biomass accumulation dynamics.

The Richards equation was expressed as:[Image Omitted. See PDF]where $W$ represents the biomass (dependent variable), $t$ represents the days after sowing (independent variable), $A$ represents the final biomass, $B$ represents the initial parameter, $K$ represents the biomass accumulation rate parameter, and $N$ represents the shape parameter. The four parameters ($A$, $B$, $K$, and $N$) of the Richards curve under each treatment were estimated by substituting the biomass and days after sowing into the Richards equation.

After parameters of the Richards equation were estimated, the maximum rate of biomass accumulation (MRBA), average rate of biomass accumulation (ARBA), days to achieve the maximum rate of biomass accumulation (DMRBA), and effective biomass accumulation duration (EBAD) were calculated as follows:[Image Omitted. See PDF][Image Omitted. See PDF][Image Omitted. See PDF][Image Omitted. See PDF]The rice biomass accumulation process could be divided into the early stage (0-t₁), middle stage (t₁-t₂), and late stage (t₂-t₃), and t₁, t₂, and t₃ were calculated as follows:[Image Omitted. See PDF][Image Omitted. See PDF][Image Omitted. See PDF]The average rate of biomass accumulation (ARBA) at each stage was calculated as follows:[Image Omitted. See PDF][Image Omitted. See PDF][Image Omitted. See PDF]

Statistical analysis

Data were analyzed using IBM SPSS Statistics 22 (SPSS), and treatment means were compared by the least significant difference test. The fitted Richards equation was derived using Origin 9. Graphical representations of the data were produced using Origin 9, Microsoft Excel 2019 (Microsoft Corporation), SigmaPlot 12.0 (Systat Software, Inc.), and the R statistical programming language.

RESULTS Yield and biomass

Wheat straw return and sowing method had significant effects on yield, yield components, and biomass at maturity for direct-seeded rice (Table 1). Wheat straw return reduced yield by 5.1%–6.2% for direct-seeded rice. Compared with wet direct-seeded rice, dry direct-seeded rice was more negatively affected, with yield and biomass reduced by 6.7%–7.6% and 5.4%–6.5%, respectively. In terms of yield composition, dry direct seeding after wheat straw return reduced panicle number by 6.0%–8.5% and total spikelet number by 4.2%–5.9%. There were differences in the effects of direct seeding methods on yield and biomass after wheat straw return. Wet direct-seeded rice showed 4.0%–5.7% higher yield and 3.9%–5.6% higher biomass than dry direct-seeded rice after wheat straw return. In terms of yield composition, wet direct-seeded rice had 4.7%–6.5% higher panicle numbers and 6.7%–7.7% higher total spikelet numbers. No significant two- or three-way interactions for yield, yield components, and biomass were found.

TABLE 1 Yield, yield components, and biomass at maturity for each treatment in 2019 and 2020

Treatment	Yield (t ha−1)	Panicles (m−2)	Spikelets per panicle	Total spikelets (m−2)	Filled-grain (%)	Grain weight (mg)	Biomass (t ha−2)
2019
S0M1a	10.1ab	364ab	112b	40,550b	94.7a	26.2a	17.0a
S0M2	10.2a	366a	114ab	41,495a	95.0a	26.2a	17.1a
S1M1	9.33b	335c	114ab	38,054c	95.6a	26.3a	15.9b
S1M2	9.70ab	350b	116a	40,604ab	95.6a	26.4a	16.5ab
2020
S0M1	9.46ab	348ab	113c	39,398a	94.9a	26.1a	15.7a
S0M2	9.67a	355a	114bc	40,571a	95.0a	26.1a	16.0a
S1M1	8.83b	312c	116ab	36,250b	96.1a	26.2a	14.9b
S1M2	9.33a	332bc	117a	39,031a	96.4a	26.3a	15.7a
F-valuesc
Y	30.7**	21.3**	NS	10.0**	NS	NS	41.0**
S	11.8**	53.3**	NS	23.3**	5.80*	NS	18.5**
M	10.1**	12.9**	NS	25.4**	NS	NS	7.81*
Y*S	NS	NS	NS	NS	NS	NS	NS
Y*M	NS	NS	NS	NS	NS	NS	NS
S*M	NS	NS	NS	NS	NS	NS	NS
Y*S*M	NS	NS	NS	NS	NS	NS	NS

Note: M, direct seeding method; S, wheat straw return treatment; Y, year.

Significant differences between averages and significance of F-values are indicated by *(0.01 < p ≤ 0.05) and **(0.001 < p ≤ 0.01).

^aS0M1 and S0M2 represent dry direct-seeded rice and wet direct-seeded rice without wheat straw return, respectively. S1M1 and S1M2 represent dry direct-seeded rice and wet direct-seeded rice with wheat straw return, respectively.

^bDifferent lowercase letters among treatments indicate significant differences at the 0.05 probability level.

^cF-values are provided for interactions.

Tillering dynamics

A dynamic tillering model across the 2 years (Figure 2) showed that the estimated maximum tiller number of direct-seeded rice with wheat straw return was 7.2%–10.3% lower than that without wheat straw return. The estimated maximum tiller number decreased by 8.1%–11.8% in dry direct-seeded rice and 6.4%–8.9% in wet direct-seeded rice. The estimated maximum tiller number of wet direct-seeded rice increased by 3.0%–5.2% compared with that of dry direct-seeded rice after wheat straw return.

FIGURE 2. Tillering dynamics for each treatment in 2019 and 2020. S0M1 and S0M2 represent dry direct-seeded rice and wet direct-seeded rice without wheat straw return, respectively. S1M1 and S1M2 represent dry direct-seeded rice and wet direct-seeded rice with wheat straw return, respectively.

Biomass translocation

Wheat straw return and sowing method had significant effects on translocation quantity, translocation efficiency, and contribution rate of stem-sheath and leaf for direct-seeded rice after heading stage (Figure 3). Wheat straw return increased translocation efficiency by 8.8%–9.6% and contribution rate by 6.7%–9.3% for stem-sheath of direct-seeded rice. The translocation quantity of stem-sheath of wet direct-seeded rice increased by 4.8%–8.2% compared with that of dry direct-seeded rice after wheat straw return. Contrary to the translocation of stem-sheath, wheat straw return reduced translocation quantity by 8.4%–12.0%, translocation efficiency by 3.1%–6.2%, and contribution rate by 5.6%–8.3% for leaves of direct-seeded rice. The translocation quantity of leaves of wet direct-seeded rice increased by 4.6%–9.2% compared with that of dry direct-seeded rice after wheat straw return. From the perspective of the translocation of the stem-sheath plus leaf, wheat straw return reduced the translocation quantity by 6.3%–7.5%, and there was a tendency to decrease the translocation efficiency and contribution rate in dry direct-seeded rice. However, the use of wet direct seeding method could increase its translocation quantity by 8.1%–11.4% and translocation efficiency by 4.9%–6.4% compared to dry direct-seeded rice, and the contribution rate also showed a tendency to increase.

FIGURE 3. The translocation quantity, translocation efficiency, and contribution rate of stem-sheath, leaf, and stem-sheath plus leaf for direct-seeded rice after heading stage in 2019 and 2020. S0M1 and S0M2 represent dry direct-seeded rice and wet direct-seeded rice without wheat straw return, respectively. S1M1 and S1M2 represent dry direct-seeded rice and wet direct-seeded rice with wheat straw return, respectively. Error bars show standard error of replicates (n = 3). Different lowercase letters among treatments indicate significant differences at the 0.05 probability level.

Selection, evaluation, and application of simulation equation for biomass accumulation Selection and evaluation of simulation equations

The relationship between biomass and days after sowing for each treatment was fitted by Origin 9 software, taking days after sowing of each treatment as the independent variable and biomass accumulation after sowing as the dependent variable. To select simulation equations that were biologically meaningful and accurately reflected the biomass dynamics after sowing, 15 non-linear growth models (including Richards, Morgan-Mercer-Flodin [MMF], Gompertz, Mischerlich, Weibull, and Logistic) were fitted using treatment S1M1 as an example. Eight equations with good fitting effects (fit coefficient >0.99) are listed in Table 2. The Richards equation had a higher fit coefficient than the other equations. Therefore, this equation was selected for the dynamic simulation of biomass after sowing.

TABLE 2 Simulation equation of biomass accumulation for dry direct-seeded rice after wheat straw return (S1M1) in 2019

[Table omitted. See PDF]

Fitted equations for biomass accumulation after sowing were established for each treatment based on the Richards equation (Table 3), and their fit coefficients were all greater than 0.999. The Richards equation fitted in 2019 was verified using the data from 2020. Likewise, the Richards equation fitted in 2020 was verified using the data from 2019. A linear analysis of Y = X between the simulated and measured biomass values is shown in Figure 4. The results indicated that the simulation equation had high accuracy (R² > 0.999). Therefore, the biomass dynamics for each treatment could be reflected by the fitted Richards equation over the 2 years.

TABLE 3 Fitted equations of biomass accumulation for each treatment in 2019 and 2020

[Table omitted. See PDF]

Note: S0M1 and S0M2 represent dry direct-seeded rice and wet direct-seeded rice without wheat straw return, respectively. S1M1 and S1M2 represent dry direct-seeded rice and wet direct-seeded rice with wheat straw return, respectively.

FIGURE 4. Relationship between simulated values and measured values for biomass of each treatment in 2019 and 2020. S0M1 and S0M2 represent dry direct-seeded rice and wet direct-seeded rice without wheat straw return, respectively. S1M1 and S1M2 represent dry direct-seeded rice and wet direct-seeded rice with wheat straw return, respectively.

Application of simulation equation

The dynamic rate of biomass accumulation was derived by taking the first derivative of the fitted Richards equation for each treatment (Figure 5). The biomass accumulation rate curve for each treatment showed a single peak, first increasing and then decreasing after sowing. Wheat straw return reduced the biomass accumulation rate of direct-seeded rice throughout the whole growth period. However, this negative effect was not significant for w et direct-seeded rice.

FIGURE 5. Dynamics rate in biomass accumulation for each treatment in 2019. S0M1 and S0M2 represent dry direct-seeded rice and wet direct-seeded rice without wheat straw return, respectively. S1M1 and S1M2 represent dry direct-seeded rice and wet direct-seeded rice with wheat straw return, respectively.

The maximum rate of biomass accumulation and the average rate of biomass accumulation were reduced for dry direct-seeded rice after wheat straw return. Wheat straw return also delayed the days to achieve the maximum rate of biomass accumulation and shortened the effective biomass accumulation duration for direct-seeded rice. Wet direct seeding increased the maximum rate and average rate of biomass accumulation compared with dry direct-seeded rice after wheat straw return (Table 4).

TABLE 4 Parameters of biomass accumulation after sowing for each treatment in 2019 and 2020

Year	Treatment	MR (kg ha−1 day−1)	AR (kg ha−1 day−1)	DMR (day)	ED (day)
2019	S0M1a	305.69	207.17	67.46	154.65
	S0M2	307.45	208.39	67.39	154.74
	S1M1	293.66	199.12	67.85	153.30
	S1M2	303.27	205.57	68.19	153.33
2020	S0M1	292.28	198.14	67.84	152.02
	S0M2	293.56	199.05	67.76	153.85
	S1M1	288.80	195.75	68.28	148.21
	S1M2	293.37	198.94	68.04	152.58

Abbreviations: AR, average rate of biomass accumulation; DMR, days to achieve the maximum rate of biomass accumulation; ED, effective duration of biomass accumulation; MR, maximum rate of biomass accumulation.

^aS0M1 and S0M2 represent dry direct-seeded rice and wet direct-seeded rice without wheat straw return, respectively. S1M1 and S1M2 represent dry direct-seeded rice and wet direct-seeded rice with wheat straw return, respectively.

The characteristics of biomass accumulation at each growth stage are shown in Table 5. The average rates of biomass accumulation at the early, middle, and late stage were lower for direct-seeded rice after wheat straw return. The negative impact of wheat straw return on direct-seeded rice was mainly concentrated in the early stage. The average rate of biomass accumulation in the early stage was reduced by 5.5%–10.3%, accounting for 61.1%–75.3% of the reduction in total average rate. As a result, the amount of biomass accumulation in the early stage was reduced by 3.5%–8.0% for direct-seeded rice after wheat straw return. The amount of biomass accumulation for dry direct-seeded rice was reduced by 4.3%–10.9%, 5.8%–7.0%, and 5.7%–6.2% in the early, middle, and late stages, respectively. However, wet direct seeding increased the average rate of biomass accumulation compared with dry direct seeding after wheat straw return. As a result, the amount of biomass accumulation increased by 2.0%–6.4%, 3.7%–6.2%, and 2.8%–7.6% in the early, middle, and late stages, respectively.

TABLE 5 Characteristics of biomass accumulation in the early, middle, and late stages in 2019 and 2020

Treatment	Early stage		Middle stage		Late stage
	AR (kg ha−1 day−1)	Biomass (kg ha−1)	AR (kg ha−1 day−1)	Biomass (kg ha−1)	AR (kg ha−1 day−1)	Biomass (kg ha−1)
2019
S0M1a	45.63	2141	265.35	10,898	71.84	4789
S0M2	45.59	2137	266.85	10,956	72.22	4826
S1M1	39.78	1908	254.73	10,134	68.85	4513
S1M2	42.07	2029	263.18	10,506	71.20	4641
2020
S0M1	39.92	1921	253.62	9997	68.60	4423
S0M2	40.44	1929	254.66	10,216	68.84	4545
S1M1	37.16	1839	250.67	9421	67.85	4148
S1M2	38.76	1875	254.46	10,005	68.77	4462

Abbreviation: AR, average rate of biomass accumulation.

^aS0M1 and S0M2 represent dry direct-seeded rice and wet direct-seeded rice without wheat straw return, respectively. S1M1 and S1M2 represent dry direct-seeded rice and wet direct-seeded rice with wheat straw return, respectively.

Correlation analysis

Correlation analysis of the characteristic values of biomass accumulation at different stages with biomass at maturity (Figure 6) showed that biomass accumulation rate at the early stage had the strongest correlation with biomass accumulation at maturity. The rate and amount of biomass accumulation in the middle and late stages were both correlated with the rate and amount of biomass accumulation in the early stage for direct-seeded rice. Biomass accumulation at maturity was positively correlated with the maximum rate of biomass accumulation, the average rate of biomass accumulation, and the effective biomass accumulation duration, and it was negatively correlated with days to achieve the maximum rate of biomass accumulation.

FIGURE 6. Correlation analysis among biomass at maturity and the characteristic values of biomass accumulation at different stages (P0.05 = 0.707 and P0.01 = 0.834). BM: Biomass at maturity; E-AR, M-AR, and L-AR represent average rate of biomass accumulation during the early, middle, and late stage, respectively; E-BM, M-BM, and L-BM represent biomass during the early, middle, and late stage, respectively; MR and AR represent maximum rate and average rate of biomass accumulation, respectively; DMR and ED represent days to achieve the maximum rate of biomass accumulation and effective duration of biomass accumulation, respectively.

DISCUSSION Yield performance and biomass accumulation characteristics for direct-seeded rice after wheat straw return

Direct seeding rice coupled with wheat straw return is a convenient and environmentally friendly rice planting method in a rice-wheat rotation system. The results of this study confirmed that yield was reduced for direct-seeded rice after wheat straw return. In terms of yield components, spikelet number per panicle and filled grain percentage of direct-seeded rice after wheat straw return showed an increasing trend, but panicle number showed the opposite trend. These results were consistent with previous studies (Huang et al., 2013; Xu et al., 2009). The increase may be related to the positive effects contributed about by returned wheat straw, because the additional organic matter and nutrients brought by the returned wheat straw were beneficial for rice grain filling (Wang et al., 2015; Xu et al., 2010; Zhang et al., 2016). This study also showed that wheat straw return increased stem-sheath translocation and stem-sheath plus leaf translocation and decreased leaf translocation (Figure 3), which facilitated grain filling and photosynthetic performance after heading stage for direct-seeded rice. The possible reason for the decrease in panicle number was the restricted seedling tillering capacity for direct-seeded rice after wheat straw return. In this study, the estimated maximum tiller number based on the dynamic tillering model was significantly lower for direct-seeded rice after wheat straw return (Figure 2). Previous studies have also reported that rice tiller number responded negatively to straw return in the early growth stage (about 6 weeks after sowing; Gao et al., 2004), and wheat straw return reduces the incidence of low tiller position or early emerging tillers (Xiong et al., 2015). The restricted seedling tillering capacity in direct-seeded rice may be related to the negative effects brought about by returned wheat straw. First, the formation of strongly reducing conditions caused by the reduced soil redox potential after wheat straw return (Tian et al., 2022; Yang et al., 2020) and the accumulation of microbial allelochemicals in soil during decomposition of the returned wheat straw (Fageria et al., 2008; Wang et al., 2015; Yu et al., 2013) limited the growth of rice roots and nutrient uptake in rice plants. Second, A decrease in soil available N content may also restrict tillering capacity. Wheat straw with a higher C:N ratio can cause higher microbial consumption of soil available N during decomposition of returned wheat straw (Li & Zhong, 2021; Zhao et al., 2017), leading to poor seedling establishment for direct-seeded rice after wheat straw return. In addition, the study by Tian et al. (2022) concluded that wheat straw return reduced the N uptake of direct-seeded rice seedling.

The yield performance of direct-seeded rice after wheat straw return has always been controversial in production. Like the results of this study, some studies have also reported that the negative effects brought about by the returned wheat straw led to a reduction in rice yields (Du et al., 2016; Gao et al., 2004). Conversely, some studies have suggested that the positive effects contributed by the returned straw may be able to compensate for the negative effects caused by the returned straw and thus increase yield (Sharma et al., 2021; Wang et al., 2015; Xu et al., 2010; Zhang et al., 2016). Previous studies have concluded that returned wheat straw can promote rice yield, but most of these studies were performed with transplanted rice (Yang, Feng, et al., 2020). Differences in seedling establishment methods between direct-seeded rice and transplanted rice are a possible reason for the differences between our results and those of previous research on transplanted rice. When seedlings are transplanted from their initial production area to the main field with returned wheat straw, transplanted seedlings have at least four leaves. At this point, they have reached the stress resistance stage, developed some stress resistance, and are able to adapt to the negative effects of the returned wheat straw, thereby suffering less damage. However, direct-seeded rice seedlings constantly experience stress from germination and throughout their growth, leading to poor seedling establishment and lower yields for direct-seeded rice after wheat straw return.

The use of growth models has become an important research approach for the analysis of crop biomass accumulation (He et al., 2019; Ji et al., 2012). As far as rice is concerned, equation models such as the Gompertz, Richards, and logistic equations have been fitted to analyze the dynamics and characteristics of biomass accumulation based on differences in rice variety types and ecological systems of rice planting areas (Li et al., 2010; Liu et al., 2019). For example, Wei and Ji found that the Gompertz equation could quantitatively analyze the characteristics of biomass accumulation for transplanted rice after transplanting (Ji et al., 2012; Wei et al., 2021). In this study, the fitting coefficient and the fitting effect of the Richards equation were higher than those of other growth equations such as the Gompertz and logistic equations. Reciprocal testing of the equations using 2 years of data indicated that the Richards equation produced a better fit for biomass accumulation dynamics of direct-seeded rice after sowing. Previous studies have shown that different climatic conditions, variety types, sowing dates, irrigation measures, and fertilization practices have significant effects on the characteristics of biomass accumulation in rice (Bai et al., 2016; He et al., 2010; Su et al., 2020). In this study, the dynamics of biomass accumulation of direct-seeded rice were fitted and analyzed by the Richards equation, and the characteristics and differences in biomass accumulation between direct seeding methods were quantitatively analyzed at the same experimental site with the same cultivation measures. Because of the limited data, the reliability and applicability of the Richards equation analysis of biomass accumulation should be further verified under different cultivation regimes in other ecological regions.

Biomass accumulation rate is an important index for describing the population growth of crops such as rice. Some studies have reported that the biomass accumulation rate curve of rice shows a single peak that increases and then decreases with days after planting (Ji et al., 2012; Su et al., 2020). However, some studies have found three peaks of biomass accumulation rate in rice after planting: the late tillering to panicle differentiation stage, the panicle differentiation to flowering stage, and the late flowering to physiological maturity stage (He et al., 2010). In this study, the biomass accumulation rate of direct-seeded rice showed a unimodal curve in the days after sowing (Figure 5). Wheat straw return delayed the days to achieve the maximum rate of biomass accumulation, shortened the duration of effective biomass accumulation, and decreased the maximum rate and average rate of biomass accumulation for direct-seeded rice. Therefore, reduced biomass accumulation of direct-seeded rice after wheat straw return was due mainly to insufficient growth time and a lower maximum rate and average rate of biomass accumulation.

The biomass accumulation process of rice can be divided into three stages: gradual increase, rapid increase, and slow increase (Ji et al., 2012), which can also be called the early, middle, and later stages (Li et al., 2010). Previous studies suggested that returned wheat straw would inhibit rice growth in the early stage, but the nutrients released during straw decomposition would promote biomass accumulation in the later stage (Sharma et al., 2021; Yang, Luo, et al., 2020). However, this study showed that the biomass accumulation rate of direct-seeded rice after wheat straw return decreased to different degrees in the early, middle, and later stages. Although the biomass accumulation rate increased in the middle and later stages, it was still lower than that observed in the absence of wheat straw. Therefore, the additional organic matter and nutrients contributed by the returned wheat straw did not alleviate stress (poor seedling establishment) on direct-seeded rice after wheat straw return. In other words, the positive effects of wheat straw return did not compensate for the negative effects. A possible reason is that the insufficient tiller numbers caused by constrained tiller growth in the early stage reduced growth potential in the middle and later stages. In conclusion, the negative impact of returned wheat straw on direct-seeded rice at the early stage was an important reason for the reduced growth of direct-seeded rice after wheat straw return.

Differences in yield formation and biomass accumulation for different direct seeding methods after wheat straw return

There have been numerous studies on the effects of different direct seeding methods on rice yield, but their results are varied. Several reports concluded that yield was higher for wet direct-seeded rice than for dry direct-seeded rice (Cheng et al., 2020; Zhang, Yu, et al., 2017). However, for early indica rice in south China, dry direct seeding appeared to promote greater yield improvement than wet direct seeding (Wang et al., 2020, 2021). These differences may reflect differences in cultivation measures such as seedling density, variety, sowing date, and irrigation conditions (Peng et al., 1995). In this study, wet direct-seeded rice had a higher potential for yield improvement than dry direct-seeded rice after wheat straw return. The main reason was that wet direct-seeded rice had a significant advantage in terms of panicle number and maximum tiller number. Therefore, higher tiller capacity was the main reason for the higher yield potential of wet direct-seeded rice. Differences in soil available N content may be one of the reasons for the difference in tillering capacity between wet direct-seeded rice and dry direct-seeded rice: dry direct-seeding increases ammonia volatilization (Mkhabela et al., 2008; Rochette et al., 2009; Xu et al., 2013) and N runoff losses (Zhang et al., 2018) in the early growth stage. In addition, increasing the frequency of rotary tillage may also increase the content of ammonium nitrogen and nitrate nitrogen in soil to a certain extent (Kong et al., 2021). Thus, the soil inorganic nitrogen content of puddled soil after multiple rotary tillage passes in wet direct-seeded rice may be higher than that of soil after one-time rotary tillage in dry direct-seeded rice under these experimental conditions. In addition, wet direct-seeded rice also had advantages in terms of spikelet number per panicle, filled grain percentage, and 1000-grain weight. These may be because wet direct-seeded rice reduced leaf translocation, which facilitated photosynthetic capacity after heading stage; meanwhile, wet direct-seeded rice increased translocation of stem-sheath and stem-sheath plus leaf, which facilitated grain filling after heading stage compared with dry direct-seeding rice.

The degree of response to wheat straw return varied between wet direct-seeded rice and dry direct-seeded rice. In this study, there was a trend toward reduced biomass, panicle number, and maximum tiller number for both wet and dry direct-seeded rice in the absence of wheat straw return, but it did not reach the level of significance. However, the biomass, panicle number, and maximum tiller number were significantly lower in dry than in wet direct-seeded rice after wheat straw return. This result demonstrated that wet direct-seeded rice was more resistant and better able to adapt to the negative effects of wheat straw return. Why did wet direct-seeded rice show less response to the negative effects of wheat straw return? One possible reason was a difference in the extent of these negative effects due to different tillage depths in the two treatments. For the wet direct-seeded rice, wheat straw was fully mixed in the 0–25 cm soil layer after wet tillage, potentially dispersing and reducing the negative effects caused by the returned straw. However, the rotary tillage depth of dry direct-seeded rice was only 0–15 cm, which led to the retention of wheat straw on the soil surface and may have increased its toxic effects. Here, the difference in the depth of the rotary tillage was mainly caused by the different methods of rotary tillage. The difference in the depth of the rotary tillage was mainly caused by the different methods of rotary tillage. For the dry direct-seeded rice system under this experiment, the soil condition for rotary tillage was relatively dry and hard because the rotary tillage was performed directly after the wheat harvest. If the depth of rotary tillage is to be increased, a rotary tiller with a larger size and higher rotary power would have to be employed. However, the soil in the wet direct seeding system was soaked with water and then started to rotary tillage; thus, the soil was softer, and the depth of the rotary tiller can be increased more easily with the same specifications of the rotary tiller. In addition, a previous study suggested that tillage depths greater than 20 cm not only improve the physical properties of the tillage layer but also facilitate the accumulation of organic matter and available nutrients (Qi et al., 2021).

According to the results of this experiment, wet direct seeding was a better direct-seeding method after wheat straw return because wet direct-seeded rice had a higher yield potential than dry direct-seeded rice. However, dry direct seeding has been recommended in rice production practice because it is more efficient and saves water compared to wet direct seeding. Although dry direct-seeded rice has a lower yield, its yield can be improved to some extent by adopting cultivation measures such as increasing the seedling density to compensate for restricted and insufficient tillering capacity, increasing nitrogen fertilization to improve stress resistance of seedlings in the early stage, and using large rotary tillage machinery to disperse the returned wheat straw at least 20 cm into the soil.

Yield performance was better in 2019 than in 2020, and this difference could be explained by weather differences between the 2 years (Figure 1). The duration of sunshine hours was 45.6% higher in 2019 than in 2020 during the rice growing season. In addition, the precipitation was 183.5% higher in 2020 than in 2019 during the rice growing season, especially during the early growth stage, which resulted in N loss.

CONCLUSIONS

Wheat straw return, an environmentally friendly measure for improving agricultural efficiency and soil fertility, has been widely practiced in the rice–wheat rotation system. This study confirmed that yield and biomass were reduced for direct-seeded rice after wheat straw return. Insufficient total spikelet numbers and panicle numbers caused by lower maximum tiller number were the main reasons for the reduced yield. In addition, we found that the biomass accumulation of direct-seeded rice after sowing was well fitted by the Richards equation. And the model was used to quantitatively analyze differences in the yields and characteristics of biomass accumulation and translocation associated with different direct-seeded rice methods after wheat straw return. Analysis of biomass accumulation characteristics derived from the Richards equation showed that the negative impact of returned straw at the early rice stage was an important reason for the reduced growth of direct-seeded rice after wheat straw return. Adoption of the wet direct-seeding method could enhance the biomass accumulation rate, biomass accumulation amount, and maximum tiller number for direct-seeded rice at the early stage, thus improving panicle number, total spikelet number, and yield. However, dry direct seeding, from the perspective of greater efficiency and water savings, is recommended for rice production in practice. The improvement of additional cultivation measures will be the focus of future research. The findings of our study provide theoretical and practical guidance for developing high-yield cultivation and regulatory approaches for direct-seeded rice after wheat straw return.

AUTHOR CONTRIBUTIONS

Jinyu Tian: Conceptualization, Investigation, Formal analysis, Resources, and Writing – original draft. Shaoping Li: Investigation and Formal analysis. Zhipeng Xing: Conceptualization, Writing – review & editing, and Supervision. Shuang Cheng: Investigation. Baowei Guo: Conceptualization and Writing – review & editing. Yajie Hu: Conceptualization and Writing – review & editing. Haiyan Wei: Conceptualization and Supervision. Hui Gao: Conceptualization and Supervision. Ping Liao: Conceptualization. Huanhe Wei: Conceptualization. Hongcheng Zhang: Conceptualization, Writing – review & editing, and Project administration. All authors have read and agreed to the published version of the manuscript.

ACKNOWLEDGMENTS

We are grateful for grants from the Jiangsu Agriculture Science and Technology Innovation Fund (CX[20]1012), the Jiangsu Demonstration Project of Modern Agricultural Machinery Equipment and Technology (NJ2020−58, NJ2021−63), the Jiangsu Technical System of Rice Industry (JATS[2021]485), the Yangzhou University Scientific Research and Innovation Program (XKYCX20_022), and the Project Funded by the Priority Academic Program Development of Jiangsu Higher Education Institutions (PAPD).

CONFLICT OF 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.

DATA AVAILABILITY STATEMENT

The data is available on https://doi.org/10.1002/fes3.425.