**1. Introduction**

Rice (*Oryza sativa* L.) is one of the world's major crops and it provides food for over three billion people [1,2]. China is the main country of rice production, with rice planting area and yield in the forefront of the world [2,3]. Therefore, increasing rice production is essential for population growth in China and the world [4,5]. Transplanted rice is the most traditional planting method. Moreover, the traditional fertilization method is manually surface broadcast [6]. This method is not suitable for the stable improvement in Chinese agricultural systems because of some serious problems such

as labor scarcity and low profits [7]. Therefore, improving mechanization is the main way to solve this problem.

The mechanical pot-seedling transplanting (PST) is an innovative technology for transplanting rice seedlings in the paddy field. It has the advantage of precise row and hill spacing without injury to the rice plants [8]. The technique not only reduces planting costs but improves the quality of transplanting rice seedlings [9]. Mechanical pot-seedling transplanting can transplant seedlings without root injury. It expects to reduce the transplanting shock and maintain the root activity, resulting in increased nutrient absorption and consequently a vigorous initial growth [10]. However, the application of nitrogen fertilizer under the flooding condition can cause nitrogen fertilizer to be lost through ammonia volatilization and runoff, reducing the utilization efficiency of nitrogen fertilizer [11,12]. Thus, the seedlings transplanted by mechanical pot-seedling transplanting were unable to access the N resource, restricting their performance irrespective of enhanced root activity [4]. To solve this problem, much application of chemical fertilizer is one of the approaches, but it can lead to problems of lower nitrogen use efficiency (NUE) and environmental pollution [11,13].

Alternative is the deep N fertilization in mechanical pot-seedling transplanting. Deep fertilizer application methods can maintain the nutrient and enhance nutrient use efficiency [14]. The nitrogen fertilization at about 5 cm depth notably improves the total above-ground biomass and grain yield, compared to the manual surface broadcast [15]. Moreover, deep fertilization could reduce the amount of fertilizer applications without reducing yield [16]. Deep fertilization is to bury the fertilizer near the rice root, which is beneficial for fertilizer absorption by rice root [17]. The appropriate fertilization depth could not only promote the growth of rice root but also improve the growth of rice plants in the early stage [18]. Some studies found that deep N fertilization could reduce CH4 emissions by 40% and NO emissions by 54% [19,20]. Pan et al. [21] found that deep fertilization significantly increased peroxidase (POD) and catalase (CAT) in direct-seeded rice. Xu et al. [22] observed that ensuring the nutrient supply at the late stage of rice was conducive to improving antioxidant enzyme activities and photosynthetic performance of rice leaves. Moreover, Shu et al. [23] reported that mechanical deep fertilization could delay rice plan senescence by enhancing antioxidant enzyme activities and reducing the malonic dialdehyde (MDA) content in direct-seeded rice. Therefore, deep nitrogen fertilization is a feasible way to lessen environmental problems because of the excess fertilization in rice production.

Mechanical pot-seedling transplanting (PST) coupled with deep nitrogen fertilization is an emerging transplanting rice technology. However, little information is available about the effects of PST coupled with mechanized deep N fertilization on grain yield, nitrogen use efficiency, and antioxidant enzyme activities in rice. The aim of this study was to assess whether PST coupled with mechanized N deep fertilization could increase grain yield, nitrogen use efficiency, and antioxidant enzyme activities in rice.

#### **2. Materials and Methods**

#### *2.1. Mechanical Pot-Seedling Transplanting Machine*

A mechanical pot-seedling transplanting machine was developed by Changzhou YaMeiKe mechanical Co., Ltd. (Changzhou, China) (Figure 1). This method realized the synchronous operation of deep fertilization and transplanting seedlings and applied fertilizer quantitatively and fixed-point deep into the soil on the seedling side.

**Figure 1.** Pictorial view of mechanized transplanting rice machine coupled with N deep placement at the farm of South China Agricultural University, Guangzhou city, China.

#### *2.2. Experimental Treatments and Design*

Field experiments were conducted in early seasons of 2018 and 2019, respectively, at the Experimental Research Farm, College of Agriculture, South China Agricultural University, Guangzhou City, China (23◦13 N, 113◦81 E, altitude 11 m). The soil in the experimental field was sandy loam with 1010 mg kg−<sup>1</sup> total N, 1080 mg kg−<sup>1</sup> total *p*, 20,230 mg kg−<sup>1</sup> total K, 73 mg kg−<sup>1</sup> available *p*, 104 mg kg−<sup>1</sup> available K, and 21,560 mg kg−<sup>1</sup> organic C.

Two rice cultivars were *Yuxiangyouzhan* (*YXYZ*) and *Wufengyou615* (*WFY615*), which are inbred and hybrid rice, respectively, and widely grown in the local area. Moreover, the two rice cultivars have growth periods of 118 and 115 days for both early seasons, respectively. Both field experiments were used in a completely randomized design with three replicates with a plot area of 132 m2 (8 m × 16.5 m). The YaraMila compound fertilizer (total nitrogen contents TN = 15%, N: P2O5: K2O = 15%:15%:15%) was used in our experiment, which was manufactured by YaraMila Fertilizer Company, China. The application rate was 150 kg N ha−<sup>1</sup> (pure N) for the fertilizer application treatment. All seedlings were transplanted by PST and three treatments were designed as follows. The mechanized deep N fertilization was a basal fertilizer in 10 cm soil depth (MAF) and fertilizers were broadcast manually on the soil surface two days before the transplanting as a basal fertilizer (BF). No fertilizer was applied during entire growth stage (N0). Water management strategies were adopted by local farmer's advice. Some chemical reagents such as herbicide, imidacloprid, tricyclazole, and carbendazim were adopted to prevent and control weeds, insects, and diseases.

#### *2.3. Yield and Its Components*

At maturity, rice grains were recorded from the harvested-area of 6 m2. In total, 20 rice plants were collected randomly for each treatment and the averaged values were calculated for the number of productive panicles per hill. Six hills of rice plants were taken to investigate yield components and the yield components measurement were determined according to Pan et al. [21]. To divide the filled seeds, all spikelets were submerged in tap water, apart from the rachis (by manual threshing). To calculate the total number of spikelets, we counted the spikelets in three representative subsamples

of 30 g. The average weight of half-filled spikelets was determined. Spikelets per panicle, grain-filling percentage, and 1000-grain-weight were also calculated from sampled plants and averaged.

#### *2.4. Total Above-Ground Biomass (TAB) and Leaf Area Index (LAI) at Di*ff*erent Growth Stages*

According to the average number of tillers in the plot, six plants were taken in the plot at all critical growth stages including mid-tillering (MT), panicle initiation (PI), heading stage (HS), and the maturity stage (MS). LAI and TAB were determined according to Pan et al. [21]. The soil on the rice plants were washed thoroughly. Then, leaf sheaths plus stems, leaves, and spikes were separated from the plant after the heading stage. The leaf area for all green leaf blades was measured with the Li-Cor area meter (Li Cor Model 3100, Lincoln, NE, USA) and the leaf area per m2 (leaf area index, LAI) was then calculated. The separated part of the rice plants was oven-dried at 70 °C to constant weight and then the above-ground biomass was calculated.

#### *2.5. Nitrogen Use E*ffi*ciency*

Six plants were collected from each treatment in the physiological maturity stage. They were then divided into leaves, stems with leaf sheath, and grains. They were finally dried at 70 ◦C until constant weight, then stored to analyze the total N contents. Plant samples (0.2 g) were digested using the Kjeldhal method to analyze ammonia concentrations via an Alliance-Futura NP analyzer (Alliance Instruments, France) and then the N content was measured. The nitrogen use efficiency including nitrogen recovery efficiency (NRE), nitrogen agronomic use efficiency (NAE), nitrogen partial factor productivity (NPFP), nitrogen harvest index (NHI), and nitrogen grain production efficiency (NGPE) were evaluated by the formulae below:


where N0 up and GY0 represented the total nitrogen uptake of above-ground plant parts and grain yields in the N0 plot, respectively. Nup and GY are the total nitrogen uptake of above-ground plant parts and grain yields in other N-fertilized plots, respectively. FN is the applied N fertilizer rate; Ng is the total nitrogen uptake in grain.

#### *2.6. Determination of Antioxidant Enzyme Activities*

About 25 leaves from each treatment were collected during the MT, PI, and HS stage. All samples were stored in −80 ◦C for enzyme activity determination i.e., peroxidase (POD), catalase (CAT), and malonic dialdehyde (MDA). POD and MDA were determined by the method established by Pan et al. [24]. Fresh leaf segments (<2 mm, 0.25 g) were homogenized in an ice bath in 5 mL of 50 mm borate buffer (pH 8.7) containing 5.0 mm sodium hydrogen sulfite and 0.1 g polyvinylpyrrolidone (PVP). The homogenate was centrifuged at 9000 ×g for 15 min at 4 ◦C. The supernatant was used as enzyme extract. POD activity was assayed by adding 0.1 mL of the enzyme extract to a substrate mixture containing acetate buffer (0.1 mol L<sup>−</sup>1, pH 5.4), ortho-dianisidine (0.25% in ethyl alcohol) and 0.1 mL 0.8% H2 O2 was added to 0.1 mL of the enzyme extract. Absorbance change of the brown guaiacol at 460 nm was recorded for calculating POD activity. One POD unit of enzyme activity was defined as the absorbance increase because of guaiacol oxidation by 1-unit min−<sup>1</sup> (U g−<sup>1</sup> FW min<sup>−</sup>1). Leaf samples (0.5 g) were homogenized in 5 mL of 5% trichloroacetic acid. The homogenate was centrifuged at 4000 ×g for 10 min at 25 ◦C and 3 mL of 2-thiobarbituric acid in 20% trichloroacetic acid was added to a 2 mL aliquot of the supernatant. The mixture was heated at 98 ◦C for 10 min and cooled rapidly in an ice bath. After centrifugation at 4000 ×g for 10 min, the absorbance was recorded at 532 nm. Measurements were corrected for non-specific turbidity by subtracting the

absorbance at 600 nm. MDA concentration was determined by the extinction coefficient MDA (ε = 155 μm cm<sup>−</sup>1). CAT activities were determined according to Dhindsa et al. [25]. The 3 mL reaction mixture contained 50 mM phosphate buffer, pH 7.0, 15 mM hydrogen peroxide, and 25 μL enzyme extract. The decrease in hydrogen peroxide was followed as a decline in A240 using a Perkin-Elmer double-beam spectrophotometer connected to a recorder. The activity was expressed in units where one unit of catalase converts one μmole of hydrogen peroxide per minute.

#### *2.7. Data Analysis*

The experimental data were analyzed using DPS3.11 (Data Processing System for Practical Statistics, Hangzhou, China). In the ANOVA model, the single effect of treatment, cultivar, year, and the interaction effect were fixed, while the replication effect in year was random. The differences amongst means of the experimental treatments were separated using the least significant difference (LSD) at 0.05 probability level (ANOVA). All figures were drawn with Origin 9.0.

#### **3. Results**
