Effects of a Novel Weeding and Fertilization Scheme on Yield and Quality of Rice

Abstract

This study aimed to assess the feasibility of a novel weeding and fertilization scheme, namely, mechanical weeding plus a one-time deep application of a reduced amount of slow-release fertilizer for rice cultivation. The effects of the weeding and fertilization method on rice yield and quality were investigated using a split plot test as the research method. Two weeding methods, namely, chemical weeding (CW) and mechanical weeding (MW), and four fertilization methods were tested, including the conventional fertilization method (quantitative split broadcast application of fast-release N fertilizer (CK)), the quantitative split broadcast application of 80% fast-release N fertilizer (LCK), the one-time base application of slow-release fertilizer (SR), and the one-time deep application of 80% slow-release fertilizer (LSR). The results showed that the rice yield under MW with LSR treatment can maintain a high level—higher than 9.2 t ha⁻¹ per year. This was attributed to the slow-release fertilizer and deep fertilization, which increased the number of stems and tillers in the pre-fertility and spike rate, respectively, resulting in a high panicle number with a 20% reduction of N fertilizer. Furthermore, mechanical weeding improved the seed-setting rate, resulting in a higher number of grains per panicle, a higher panicle number, and an increased thousand-grain weight, thereby maintaining a high yield. At the same time, the quality of rice under MW with LSR treatment improved, specifically reflected in the significant improvement of the processing and appearance quality of rice, a slight increase in protein content, and a reduction in the amylose content, thereby improving its nutritional quality while maintaining good cooking quality.

1. Introduction

Rice is the primary staple food source for over 50% of the world’s population, and its yield is crucial not only for food security but also for the overall well-being of people. However, the scheme and quality of fertilization and weeding operations have significant impacts on rice yield during cultivation [1,2,3]. Currently, chemical fertilizers are still predominantly applied in broadcast form in China, leading to a low utilization rate of fertilizers [4,5]. Chemical weeding is the primary method, but the long-term use of herbicides has resulted in increased weed resistance and damage to the soil over time [6,7]. Additionally, there are existing issues with herbicide spraying machinery, with “running, emitting, dripping, and leaking”, significantly affecting weeding efficacy [8,9]. As a result, farmers often resort to blindly increasing the amounts of chemical fertilizers and herbicides. This leads to soil siltation and acidification, increased toxicity of herbicide residues, weed resistance/tolerance, environmental pollution, and destruction of the ecological balance [10,11]. These issues seriously threaten the quality and safety of China’s rice and agricultural ecological environment. Therefore, it is necessary to accelerate the transformation of traditional rice cultivation management modes that rely on chemical fertilizers and herbicides, change existing fertilization and weeding methods, and promote the efficient use of chemical fertilizers and traditional chemical control methods via modern green prevention and control, while ensuring high and stable yields.

This urgent need has led to a focus on the deep application of slow-release fertilizer and mechanical weeding. The principle of deep application of slow-release fertilizer is based on the application of fertilizers with slower nutrient release rates and longer nutrient release cycles to the soil at certain depths near the roots of seedlings, using specialized machinery so that the fertilizer can be more easily absorbed by the roots [12]. Previous studies have concluded that the deep placement of slow-release fertilizer can improve nutrient retention time, reduce nutrient runoff losses, and significantly improve fertilizer utilization compared to the broadcast application of fast-release fertilizer [13,14]. Another study showed that the deep application of slow-release N fertilizer improved the yield of rice compared to the broadcast application of fast-release N fertilizer, the deep application of fast-release N fertilizer, and the broadcast application of slow-release N fertilizer, as shown by the effective increase in nitrogen accumulation in stems and leaf sheaths, the apparent transfer, the panicle number, and the above-ground dry-matter weight [15,16,17]. Additionally, it has also been shown that one-time deep placement of slow-release fertilizer can result in stable or increased rice yields, even with a 20% reduction in the amount of fertilizer used. At the same time, it can significantly improve fertilizer utilization. The most effective combination was a one-time application of slow-release fertilizer, a 20% reduction in the amount of fertilizer, and deep placement [18]. Mechanical weeding, as an environmentally friendly weeding method, relies on the interaction of weeding implements with soil and weeds in paddy fields to complete the turning, pulling out, pulling off, burying, etc., of weeds. The purpose of weeding is achieved by destroying the growth environment and competition conditions of weeds [19,20]. Previous studies have shown that mechanical weeding can achieve weeding effects similar to chemical weeding (with a weeding rate higher or equal to 80%), and it can increase rice yield by 2–11%, mainly by increasing tiller number, chlorophyll content, and leaf enzyme activity [21,22]. Other studies have shown that proper and repeated mechanical weeding promotes rice plant height and panicle number because tilling the soil by means of the weeding implement can increase the oxygen content of the soil, promote crop nutrient absorption, and enhance rice nitrogen absorption [23,24]. In addition, mechanical weeding can improve the physical and chemical properties of the soil, specifically manifested by increasing its porosity, reducing its bulk density, increasing its organic matter content, and improving its fertility, thus promoting crop growth and development, as well as improving yield [25,26].

Although both of these techniques have shown promising results in practice, they have only been used individually in rice cultivation. Therefore, further research is needed to investigate the synergistic effects of these two techniques. Based on previous research, this paper conducted split-plot experiments with different fertilization and weeding methods, using conventional methods as a control, to explore the effects of “mechanical weeding combined with slow-release fertilizer reduction and deep application” on the yield and quality of rice, as well as its application effects. The research results can provide valuable data to support the integrated innovation of mechanized, simplified, high-quality, high-yield, and efficient cultivation techniques for direct-seeded rice.

2. Materials and Methods

2.1. Test Site and Materials

The experiment was carried out from 2018 to 2019 at the Agricultural Science Institute of Dazhong Farm, Yancheng City, Jiangsu Province, China (33°08′ N, 120°39′ E). The area receives, on average, 2078 h of sunlight annually, and it has a mean annual temperature of 14.6 °C, a frost-free period of 215 days, and annual precipitation of 1040 mm. The experimental soil was a desalinated loamy soil with 22.3 g kg⁻¹ of organic matter, 1.2 g kg⁻¹ of total nitrogen, 45 mg kg⁻¹ of rapidly available phosphorus, and 195 mg kg⁻¹ of rapidly available potassium in the 0~20 cm tillage layer. The previous crop was wheat, with a yield of 8.42 t ha⁻¹. Wheat straw was smashed during the harvest, and then rototilled and deeply buried before rice sowing.

The experimental rice variety employed was Nanjing 5718 (NJ5718), a medium japonica cultivar sourced from the Institute of Grain Crops at the Jiangsu Academy of Agricultural Sciences, China. This particular variety yields high-quality rice, boasting a protein content of approximately 8.8% and a straight-chain starch content of around 10.0%.

The slow-release fertilizer used in the experiment was a blended fertilizer, N:P₂O₅:K₂O = 30:7:13, with slow-release N ≥ 20%, supplied by Shandong Maoshi Ecological Fertilizer Co., Ltd., Linyi, China. The fast-release N fertilizer was urea, with TN (Total Nitrogen) ≥ 46.5%, supplied by Henan Nongxin Fertilizer Industry Co., Ltd., Xinxiang, China.

2.2. Test Method

The rice within the designated experimental zone was cultivated through the dry direct-seeding technique. Sowing activities were conducted on 11 June 2018, and 8 June 2019, employing a cultivator-planting combine that was independently developed by Yangzhou University (Figure 1). The sowing span encompassed a width of 2.5 m, with a pair of 10 cm wide drainage furrows thoughtfully placed at the center. These furrows were established at a spacing of 1.5 m, while the intervals between individual sowing rows were set at 25 cm. The total count of sowing rows tallied up to 10, with a corresponding sowing strip width of 5 cm. The quantity of seeds applied was 150 kg ha⁻¹, and during the two-leaf and one-heart growth stage, the seedlings were prudently thinned down to achieve a foundational seedling density of 300 × 10⁴ ha⁻¹.

Two weeding treatments were implemented during the experiment: chemical weeding (CW) and mechanical weeding (MW). CW treatment was administered thrice, following the “one seal, two kill and three supplement” methodology prevalent in local high-yield farming, using a self-propelled sprayer (3WPZ-700A, Zhongnong Fengmao Plant Protection Machinery Co., Ltd., Tianjin, China). Specifically, Bensulfuron-methyl·Pretilachlor(Jiangsu Fengshan Group Co., Ltd., Yancheng, China) (1500 mL ha⁻¹) was applied during the early stage for chemical weeding, Cyhalofop-butyl (Taoshi Yinong Agricultural Technology (China) Co., Ltd., Nantong, China) (1800 mL ha⁻¹) and Quinclorac (Qingdao Jiner Agrochemical Research and Development Co., Ltd., Qingdao, China) (675 g ha⁻¹) were applied in the middle stage to prevent and control weeds belonging to the Graminaceae family, and Basagran (BASF Plant Protection (Jiangsu) Co., Ltd., Nantong, China) (2100 mL ha⁻¹) was used in the later stage to prevent and control broad-leaved weeds and sedge. MW treatment was performed at days 15, 25, and 35 after sowing, utilizing a mechanical weeding implement (P005-6AHN, Q-HOE Co., Ltd., Ashoro, Japan) based on the “multi-peak” characteristics of weed emergence in dry direct-seeded rice fields. It is worth noting that MW treatment necessitated an initial chemical weeding intervention involving 30% Bensulfuron-methyl·Pretilachlor (1500 mL ha⁻¹) post-sowing to ensure the unhindered germination of seedlings. This precaution is taken due to the higher germination rate of weeds compared to seedlings.

Under both CW and MW treatments, four fertilization strategies were implemented: (1) conventional fertilization method, i.e., quantitative split broadcast application of fast-release N fertilizer (CK); (2) quantitative split broadcast application of 80% fast-release N fertilizer (LCK); (3) one-time base application of slow-release blended fertilizer (SR); and (4) one-time deep (at a depth near the main plant roots) application of 80% slow-release blended fertilizer (LSR). The amount of fertilizer applied was calculated as pure N, with both CK and SR treatments applied at a rate of 300 kg ha⁻¹ and both LCK and LSR treatments applied at a rate of 240 kg ha⁻¹. For CK and LCK treatments, N fertilizer was applied in stages of 30% basal fertilizer, 30% tillering fertilizer, 20% flower-promoting fertilizer, and 20% flower-preserving fertilizer. The basal fertilizer was applied on the day of sowing using a cultivator-planting combine. The tillering fertilizer was applied at the three-leaf and one-heart stage, while the flower-promoting and flower-preserving fertilizers were applied at the inverted four-leaf and inverted two-leaf stages, respectively, using a centrifugal fertilizer spreader (2F-30, China YTO Group Co., Ltd., Luoyang, China). In SR treatment, the fertilizer was applied on the day of sowing, using a cultivator-planting combine, while in LSR treatment, the fertilizer was applied at a depth near the main plant roots within 1 day after sowing, using a side-depth fertilizer spreader (2FC, Nantong FLW Agricultural Equipment Co., Ltd., Nantong, China). The used amount of P and K fertilizer was uniformly applied in each treatment, with a total applied amount of 130 kg ha⁻¹ of P₂O₅ and 270 kg ha⁻¹ of K₂O. The P and K deficiencies in each treatment were supplemented with calcium superphosphate and potassium chloride, respectively, and both were uniformly applied as a base fertilizer at sowing.

The experiment was conducted in the same field, with the weeding treatment as the main plot and the fertilization treatment as the split-plot, replicated three times per treatment in plots of 50 m². Barriers were installed between the main plots to prevent herbicide drift, while ridges were constructed and mulched to separate fertilizers and ensure that each plot was separately drained and irrigated. Disease and pest control was uniformly carried out according to local large-scale cultivation methods.

2.3. Test Measurement Indicators and Methods

2.3.1. Stem and Tiller Dynamics and Spike Rate

For each plot, a continuous single row measuring 1 m was selected as the observation point, and this arrangement was replicated three times to ensure representation. At these selected points, the number of stems and tillers were counted during rice jointing, heading, and maturity stages, as well as the spike rate was calculated.

2.3.2. Yield and Its Components

Before the harvest, three observation points were selected in each plot to investigate the panicle number, with each point investigating five consecutive rows of 1 m. Based on the calculated average number of panicles, five sample points were randomly selected from each plot (with 20 consecutive plants as one sample point) to investigate the number of grains per panicle and the seed-setting rate. In order to determine the thousand-grain weight, 1000 dry seeds were weighed, and this weighing was replicated three times (with an error ≤ 0.05 g). At the maturity stage, each plot was harvested in an area of 10 m², which was then dried, and the final grain yield was adjusted to 14% moisture content.

2.3.3. Rice Quality

About 500 g of grains harvested from each plot were dried at 35 °C till the moisture of the grain was 14% and then stored at 4 °C for three months. Rice quality analysis was performed according to the High-Quality Paddy Standard and Rice Quality Measurement Standards [27,28]. The brown rice, milled rice, and head-milled rice rates were expressed as percentages of the total grain weight. Processing quality encompassed the determination of brown rice rate (BRR), milled rice rate (MRR), and head-milled rice rate (HMRR). Chalkiness was evaluated based on 100 milled grains per plot. Chalky grain rate (CR) was defined as the percentage ratio between chalky rice (number of grains containing white belly, white heart, or white back) and a total of specimen rice. Chalkiness degree (CD) was defined as the percentage ratio between the chalky area of chalky rice and the total specimen rice area. CR and CD were measured by means of a rice-appearance scanner. Gel consistency (GC) is defined as the flow length of rice glue after cooling after alkali gelatinization of milled rice flour under specified conditions. Amylose content (AC) is defined as the mass fraction of amylose in the test sample. Protein content (PC) was measured by means of a grain analyzer (Infratec TM 1241, Foss, Hilleroed, Denmark).

2.4. Statistical Analysis

Tables were generated using Microsoft Excel 2017 for Windows (Microsoft Corporation, Washington, USA). The data analysis was performed utilizing DPS V7.05 (Hangzhou Ruifeng Information Technology Co., Ltd, Hangzhou, China), with means subjected to examination through the least significant difference test at a significance level of p = 0.05 (LSD0.05).

3. Results

3.1. Yield and Its Components

The effects of different weeding and fertilization treatments on yield and its components of the rice variety NJ5718 were elucidated in

Table 1. Effect of different weeding and fertilization treatments on yield and its components of the rice variety NJ5718.

Year	Weeding Treatment	Fertilization Treatment	No. of Grains Per Panicle	Seed-Setting Rate (%)	Panicle Number (×104 ha−1)	Thousand-Grain Weight (g)	Yield (t ha−1)
2018	CW	CK	132.06 ± 1.35 a	86.61 ± 0.50 cd	319.93 ± 6.97 a	28.33 ± 0.02 a	10.05 ± 0.12 a
		LCK	119.39 ± 2.23 cd	86.12 ± 1.02 d	307.24 ± 7.32 cd	27.76 ± 0.02 bc	8.72 ± 0.15 cd
		SR	130.04 ± 4.35 ab	87.03 ± 0.67 c	318.14 ± 6.36 ab	28.14 ± 0.01 a	9.86 ± 0.21 a
		LSR	127.03 ± 1.07 b	86.56 ± 2.21 cd	314.97 ± 9.03 b	28.21 ± 0.02 a	9.65 ± 0.10 ab
	MW	CK	128.72 ± 1.26 ab	88.94 ± 0.75 ab	310.92 ± 10.25 c	28.06 ± 0.03 ab	9.80 ± 0.28 a
		LCK	116.35 ± 2.29 d	88.28 ± 1.65 b	298.56 ± 4.38 e	27.62 ± 0.02 c	8.40 ± 0.11 d
		SR	127.62 ± 2.11 b	89.11 ± 0.97 a	308.02 ± 11.02 cd	27.94 ± 0.03 abc	9.65 ± 0.17 ab
		LSR	121.84 ± 1.89 c	89.24 ± 0.43 a	305.48 ± 3.56 d	28.05 ± 0.02 ab	9.20 ± 0.24 bc
2019	CW	CK	133.52 ± 3.27 a	86.76 ± 1.03 c	320.55 ± 8.67 a	28.12 ± 0.02 a	10.27 ± 0.31 a
		LCK	120.26 ± 2.46 cd	86.17 ± 0.39 c	307.80 ± 10.96 cd	27.44 ± 0.01 b	8.86 ± 0.12 cd
		SR	131.23 ± 1.18 ab	87.05 ± 0,98 bc	318.90 ± 6.54 a	28.03 ± 0.01 a	10.02 ± 0.18 ab
		LSR	127.89 ± 3.21 b	87.04 ± 1.24 bc	315.65 ± 11.12 ab	28.02 ± 0.02 a	9.74 ± 0.09 ab
	MW	CK	129.46 ± 4.32 b	88.92 ± 1.28 a	311.65 ± 7.17 bc	27.84 ± 0.02 ab	9.93 ± 0.14 ab
		LCK	117.17 ± 3.96 d	88.26 ± 0.69 ab	299.51 ± 9.06 e	27.46 ± 0.01 b	8.46 ± 0.13 d
		SR	128.36 ± 1.37 b	89.03 ± 0.79 a	308.46 ± 4.27 cd	27.93 ± 0.01 a	9.81 ± 0.29 ab
		LSR	122.75 ± 3.35 c	89.11 ± 1.13 a	305.85 ± 5.38 d	27.82 ± 0.03 ab	9.36 ± 0.16 bc

Note: Different lowercase letters in the same column for the same year indicate significant differences at the 5% level.

. The results indicate that the yield exceeded 8.0 t ha⁻¹ in both 2018 and 2019, with the highest yield obtained in 2018 under CW with CK treatment, which reached 10.27 t ha⁻¹. Remarkably, no substantial divergence was observed in yield between the two years within each treatment. Regardless of the weeding treatment, the yields followed the trend CK > SR > LSR > LCK, with no significant difference between CK, SR, and LSR. Compared to CK, LCK, SR, and LSR led to a decrease in final yield of 13.2–14.85%, 1.2–2.4%, and 4.0–6.1%, respectively. Under the same fertilization treatment, there were slight differences in rice yield between the CW and MW treatments. Specifically, the yield under CW was slightly higher than that under MW, with an incremental range of 2.2% to 4.9%, albeit without achieving statistical significance.

Regarding yield components, the trends of the number of grains per panicle and panicle number under both weeding treatments (CW and MW) were shown as follows: CK > SR > LSR > LCK. There was no significant difference between CK and SR, and the disparity between LSR treatment and the preceding two was minor. Under the same weeding treatment, compared to CK, the number of grains per panicle of LCK, SR, and LSR decreased by 9.3–9.9%, 0.8–1.7%, and 3.8–5.3%, respectively, as well as the panicle number decreased by 3.9–4.0%, 0.5–1.0%, and 1.5–1.9%, respectively. Under the same fertilization treatment, the weeding treatment had a relatively small effect on the number of grains per panicle, with CW showing a slightly higher overall effect than MW. However, the weeding treatment had a significant effect on panicle number. Remarkably, under MW treatment, the panicle number demonstrated a lower figure compared to CW, indicating a decrease ranging from 2.7% to 3.4%.

Regarding seed-setting rate and thousand-grain weight, minimal variance was observed in seed-setting rate across the four fertilization treatments within the same weeding treatment, with SR and LSR resulting slightly higher than CK and LCK. Except for the decreased thousand-grain weight in LCK treatment, both SR and LSR displayed virtually similar values to CK treatment, thus exhibiting no noteworthy variance. When focusing on the same fertilization treatment, the different weeding treatments exhibited a significant impact on the seed-setting rate, which was higher in MW rather than in CW, with an increase of 2.3–3.1%. The highest seed-setting rate of 89.24% occurred under the combination of LSR and MW. The effect of different weeding treatments on thousand-grain weight was relatively small, showing that CW was slightly higher than MW. Furthermore, the reduction in thousand-grain weight due to MW in contrast to CW was below 1%.

3.2. Stem and Tiller Dynamics and Spike Rate

Drawing insights from

Table 2. Effect of different weeding and fertilization treatments on stem and tiller dynamics and spike rate of the rice variety NJ5718.

Year	Weeding Method	Fertilization Treatment	No. of Stems and Tillers (×104 ha−1)			Spike Rate (%)
			Jointing	Heading	Maturity
2018	CW	CK	471.45 ± 6.38 c	321.37 ± 3.45 a	319.90 ± 10.33 a	67.86 ± 0.12 c
		LCK	434.64 ± 11.19 e	309.01 ± 5.66 e	307.20 ± 6.75 cd	70.68 ± 0.06 b
		SR	531.89 ± 8.37 a	319.16 ± 4.71 ab	318.10 ± 9.12 ab	59.81 ± 0.37 e
		LSR	500.49 ± 10.21 b	316.39 ± 7.39 bc	314.95 ± 10.62 b	62.93 ± 0.21 d
	MW	CK	443.13 ± 14.36 e	313.36 ± 6.44 cd	310.95 ± 11.15 c	70.18 ± 0.77 b
		LCK	403.17 ± 8.23 f	300.79 ± 4.77 f	298.55 ± 10.28 e	74.07 ± 0.50 a
		SR	489.85 ± 10.78 b	309.79 ± 8.29 de	308.00 ± 4.14 cd	62.88 ± 0.26 d
		LSR	459.54 ± 4.07 d	307.19 ± 10.51 e	305.45 ± 7.16 d	66.48 ± 0.28 c
2019	CW	CK	477.98 ± 5.39 bc	321.89 ± 7.32 a	320.55 ± 2.03 a	67.07 ± 0.35 c
		LCK	435.92 ± 11.65 d	310.48 ± 13.26 c	307.80 ± 5.20 cd	70.61 ± 0.16 b
		SR	517.66 ± 6.44 a	320.49 ± 6.22 ab	318.90 ± 11.09 a	61.61 ± 0.09 e
		LSR	484.41 ± 8.24b	318.59 ± 2.47 ab	315.65 ± 8.18 ab	65.17 ± 0.27 d
	MW	CK	455.35 ± 7.32 cd	313.55 ± 9.94 bc	311.65 ± 6.31 bc	68.44 ± 0.15 c
		LCK	403.64 ± 14.01 e	301.24 ± 13.36 d	299.50 ± 11.26 e	74.21 ± 0.30 a
		SR	484.93 ± 7.21 b	309.71 ± 5.17 c	308.45 ± 7.70 cd	63.61 ± 0.41 d
		LSR	454.05 ± 9.33 d	308.33 ± 6.79 cd	305.85 ± 8.53 d	67.47 ± 0.26 c

Note: Different lowercase letters in the same column for the same year indicate significant differences at the 5% level.

, it can be observed that the changes in stem and tiller dynamics under the same weeding treatment were basically consistent between the two years. During the jointing stage, the trends of stem and tiller numbers under different treatments were as follows: SR > LSR > CK > LCK, except for MW treatment in 2019. The number of stems and tillers in SR and LSR increased by 6.5–12.8% and 1.3–6.2%, respectively, compared to CK, while those under LCK treatment decreased by 7.8–11.4% compared to CK. In the year 2020, noteworthy differences in the number of stems and tillers emerged under MW treatment, primarily between LSR and CK. In this scenario, CK exhibited a superior figure to LSR, although the disparity was not statistically significant.

The number of stems and tillers under the same fertilization treatment displayed significant differences between the two weeding treatments, with CW > MW. Moreover, the number of stems and tillers under MW was consistently 4.7–8.2% lower than that of CW. During the heading and maturity stages, there were no significant differences in the number of stems and tillers under the same weeding and fertilization treatment. Under the same weeding treatment, the difference in the number of stems and tillers under different fertilization treatments was shown as follows: CK > SR > LSR > LCK. The difference between CK and SR was not significant, while the number of stems and tillers of LSR was reduced by less than 2%, compared to CK, and that of LCK was reduced by about 4%. The number of stems and tillers under MW decreased by 2.4–3.4% compared to CW under the same fertilization treatment.

After analyzing the variation in spike rate, it was observed that the trends of the spike rate under different fertilization treatments were as follows: LCK > CK > LSR > SR under the same weeding treatment. Notably, LCK prompted a 4.2% to 8.4% increase in spike rate compared to CK. Conversely, SR and LSR led to a decline in spike rate by 7.1–11.9% and 1.4–7.3%, respectively. In contrast, MW displayed a slight spike rate enhancement ranging from 2.0 to 5.6% when compared with CW under the same fertilization treatment.

3.3. Processing Quality

According to the data presented in

Table 3. Effect of different weeding and fertilization treatments on processing quality and appearance quality of the rice variety NJ5718.

Year	Weeding Method	Fertilization Treatment	BRR (%)	MRR (%)	HMRR (%)	CR (%)	CD (%)
2018	CW	CK	79.15 ± 3.23 bc	70.82 ± 4.38 cd	66.18 ± 8.53 cd	29.08 ± 1.02 a	7.95 ± 0.32 a
		LCK	78.00 ± 6.19 c	69.88 ± 7.79 d	65.03 ± 5.44 d	25.88 ± 2.44 cd	6.80 ± 0.58 b
		SR	83.65 ± 4.01 a	75.61 ± 2.17 ab	70.55 ± 9.76 b	28.06 ± 0.85 ab	7.86 ± 0.76 a
		LSR	83.32 ± 8.33 a	74.95 ± 7.34 b	70.17 ± 3.39 b	25.12 ± 4.33 de	6.59 ± 0.24 b
	MW	CK	80.82 ± 2.89 b	71.85 ± 6.52 c	66.96 ± 2.59 c	27.16 ± 2.45 bc	6.51 ± 0.53 b
		LCK	79.27 ± 7.27 bc	70.69 ± 3.01 cd	66.00 ± 3.40 cd	24.10 ± 2.01 ef	5.37 ± 0.47 c
		SR	84.66 ± 1.06 a	76.62 ± 9.33 a	71.94 ± 6.14 a	26.36 ± 3.36 cd	6.29 ± 0.94 b
		LSR	84.15 ± 3.52 a	75.96 ± 6.16 ab	71.18 ± 3.42 ab	23.17 ± 1.28 f	5.14 ± 0.25 c
2019	CW	CK	79.21 ± 10.84 c	71.07 ± 3.29 b	66.58 ± 5.90 c	28.83 ± 0.74 a	8.06 ± 0.52 a
		LCK	78.24 ± 3.75 c	70.17 ± 5.10 b	65.27 ± 6.44 c	25.89 ± 3.86 d	6.91 ± 0.23 b
		SR	83.87 ± 9.22 a	75.82 ± 4.54 a	70.75 ± 3.71 ab	27.81 ± 4.32 b	7.95 ± 0.41 a
		LSR	83.89 ± 2.04 a	75.81 ± 10.03 a	70.32 ± 7.67 ab	25.61 ± 2.31 e	6.74 ± 0.62b
	MW	CK	80.04 ± 3.95 bc	72.99 ± 8.47 ab	67.57 ± 7.35 bc	26.64 ± 3.01 c	6.41 ± 0.35 c
		LCK	78.94 ± 5.38 c	71.31 ± 5.36 b	66.04 ± 4.21 c	23.80 ± 1.89 f	5.29 ± 0.33 d
		SR	84.76 ± 7.47 a	76.74 ± 1.72 a	72.44 ± 3.39 a	25.55 ± 2.54 de	6.26 ± 0.61 c
		LSR	84.33 ± 2.65 a	76.28 ± 4.61 a	71.43 ± 2.83 a	22.92 ± 1.77 g	5.09 ± 0.42 d

Note: Different lowercase letters in the same column for the same year indicate significant differences at the 5% level.

, the trends in variations of BRR, MRR, and HMRR among different fertilization treatments under the same weeding treatment remained consistent across the two years, with slow-release fertilizer treatments (SR and LSR) > fast-release fertilizer treatments (CK and LCK), while no significant differences were recorded between SR and LSR or between CK and LCK. Compared to CK, BRR, MRR, and HMRR under LSR treatment increased by 4.1–5.9%, 4.5–6.7%, and 5.6–6.3%, respectively. Under the same fertilization treatment, the weeding treatment had no significant effect on BRR, MRR, and HMRR, as well as the milling quality of MW was slightly improved compared to CW.

3.4. Appearance Quality

As shown in Table 3, the trends in CR and CD among different fertilization treatments under the same weeding treatment remained consistent across the two years. The CR was ranked as LSR < LCK < SR < CK, while CD exhibited lower values under N fertilizer reduction applications (LCK and LSR) compared to N fertilizer adequate applications (CK and SR). Notably, the LSR treatment resulted in a reduction of 12.3–14.7% in CR and 16.4–20.0% in CD when compared with CK. Moreover, the weeding approach significantly influenced CR and CD under the same fertilization treatment, with MW < CW. This led to a notable decrease of 6.1–8.7% in CR and 18.1–24.5% in CD in comparison to CK. This decline was particularly pronounced under LCK and LSR treatments.

3.5. Nutritional Quality

As shown in

Table 4. Effect of different weeding and fertilization treatments on the nutritional quality and cooking quality of the rice variety NJ5718.

Year	Weeding Method	Fertilization Treatment	PC (%)	AC (%)	GC (%)
2018	CW	CK	7.80 ± 0.06 d	9.57 ± 0.62 d	88 ± 2 ab
		LCK	7.51 ± 0.45 f	9.73 ± 0.45 c	92 ± 1 a
		SR	8.05 ± 0.26 b	9.07 ± 0.20 h	87 ± 1 ab
		LSR	7.78 ± 0.72 d	9.23 ± 0.51 g	91 ± 3 ab
	MW	CK	7.91 ± 0.36 c	9.77 ± 0.14 b	86 ± 1 b
		LCK	7.61 ± 0.18 e	9.93 ± 0.66 a	91 ± 1 ab
		SR	8.15 ± 0.49 a	9.27 ± 0.32 f	86 ± 2 b
		LSR	7.89 ± 0.23 c	9.43 ± 0.55 e	90 ± 1 ab
2019	CW	CK	7.91 ± 0.16 cd	9.74 ± 0.21 d	89 ± 2 ab
		LCK	7.68 ± 0.55 f	9.91 ± 0.14 c	92 ± 1 a
		SR	8.17 ± 0.34 a	9.26 ± 0.56 h	87 ± 1 b
		LSR	7.86 ± 0.27 de	9.40 ± 0.48 g	91 ± 3 a
	MW	CK	8.01 ± 0.40 b	9.96 ± 0.73 b	87 ± 2 b
		LCK	7.77 ± 0.17 e	10.02 ± 0.29 a	91 ± 2 a
		SR	8.25 ± 0.62 a	9.47 ± 0.40 f	87 ± 1 b
		LSR	8.00 ± 0.14 bc	9.63 ± 0.57 e	90 ± 1 ab

Note: Different lowercase letters in the same column for the same year indicate significant differences at the 5% level.

, the pattern of changes in panicle number per plant was basically consistent between the two years, and the trends were as follows: SR > CK > LSR > LCK under the same weeding treatment. Notably, when compared to the CK treatment, the panicle number per plant under SR treatment increased by 3.1–3.5%, while there was no significant difference between LSR and CK. Under the same fertilization treatment, the weeding treatment had a significant effect on the panicle number per plant. Specifically, the MW treatment led to a higher value compared to CW, resulting in an increase of 1.0% to 1.9%.

3.6. Cooking Quality

As shown in Table 4, the pattern of changes in AC was basically consistent between the two years, showing that N fertilizer reduction application treatments (LCK) > N fertilizer adequate application treatments (CK and SR) > slow-release fertilizer treatment (LSR), with significant differences between different fertilization treatments under the same weeding treatment. Compared to CK, the AC of rice under LSR decreased by 3.3–3.6%. Under the same weeding treatment, the MW treatment increased the AC of rice by 1.1–2.4% compared to CW. The pattern of changes in GC was basically consistent between the two years under the same weeding treatment. The fertilization treatment had a relatively weak effect on GC, and there was no significant difference between different fertilization treatments. The difference in GC between MW and CW treatments was also not significant under the same fertilization treatment. The GC was higher than 80 mm under all treatments during the two years, while the GC under the combined treatment MW and LSR was ≥90 mm.

4. Discussion

4.1. Effect of Mechanical Weeding Combined with Slow-Release Fertilizer One-Time Reduction and Deep Application on Yield of Rice Yield

Regarding the differences in stem and tiller dynamics and spike rate, there were no significant differences in the number of stems and tillers at the heading and maturity stages under N fertilizer adequate application treatments (CK and SR). However, at the jointing stage, SR treatment exhibited a higher count of stems and tillers compared to CK. Consequently, the spike rate was lower under SR treatment in contrast to CK. The differences in the number of stems and tillers under 20% N fertilizer reduction application treatments (LCK and LSR) showed that LSR > LCK during different growth periods. In addition, the number of stems and tillers under LSR treatment was higher than that under CK during the jointing stage, while the number of stems and tillers showed the opposite trend during the heading and maturity stages. Therefore, the spike rate under LSR treatment was lower than CK. The trends of the number of stems and tillers under different weeding treatments showed that CW > MW, while the difference tended to decrease with the fertility process. Therefore, the spike rate under MW treatment was higher than CW.

The aforementioned analysis suggests that N fertilizer reduction application within a certain range can increase the spike rate, while slow-release fertilizer application can significantly increase the number of stems and tillers during the pre-fertility stage. However, this effect weakens fertility, which is related to the nutrient-release characteristics of slow-release fertilizers and fertilizer rates [29]. Nevertheless, deep fertilization treatment can ensure a higher spike rate, even with a 20% reduction in N fertilizer application. Mechanical weeding treatment results in a lower number of stems and tillers, but the rate of reduction decreases with fertility. This occurs mainly because mechanical weeding treatment damages seedlings during the pre-fertility stage, and the remaining weeds increase external competition, resulting in a lower number of stems and tillers. However, as external competition decreases with fertility, the decrease in the number of stems and tillers also decreases, resulting in a higher number of stems and tillers under mechanical weeding treatment.

In terms of differences in yield, it’s noteworthy that the weeding treatment exhibited no statistically significant impact on overall yield. Similarly, under N fertilizer adequate application, there was no significant effect of fertilization treatment (CK and SR) on yield. However, under 20% N fertilizer reduction application (LCK and LSR), conventional fertilization treatment (LCK) resulted in a certain degree of yield reduction. On the contrary, one-time deep application of a reduced amount of slow-release fertilizer treatment (LSR) ensured that the yield was basically the same as in the case of N fertilizer adequate application. These findings are consistent with previous studies [30,31].

In terms of differences in yield components, the fertilization treatment had a slight or non-significant effect on all factors under N fertilizer adequate application (CK and SR). Conversely, the weeding treatment had little effect on the number of grains per panicle and thousand-grain weight, as well as a significant effect on panicle number and seed-setting rate. In this context, MW led to an enhancement in seed-setting rate while concurrently reducing panicle number in comparison to CW. In the case of 20% N fertilizer reduction application, the fertilization treatment had a slight effect on the seed-setting rate but a significant effect on the number of grains per panicle, panicle number, and thousand-grain weight. Slow-release fertilizer one-time deep application could significantly increase the number of grains per panicle, panicle number, and thousand-grain weight compared to the conventional fertilization treatment. However, this approach also led to a slight reduction in these indices under LSR treatment, as compared to CK. Moreover, the observed variations in these indicators between different weeding treatments mirrored the outcomes seen in the context of N fertilizer adequate application.

The above analysis provides an explanation for the differences in yield. Reduction of N fertilizer by 20% under conventional fertilization treatment leads to a significant decline in yield, primarily due to reductions in the number of grains per panicle, seed-setting rate, panicle number, and thousand-grain weight. However, the use of slow-release fertilizer and deep fertilization treatment resulted in enhancements across these indicators. Notably, even with a 20% reduction in N fertilizer, the combined use of these techniques maintains seed-setting rates and thousand-grain weights comparable to conventional treatments, aligning with findings from prior research [32]. Additionally, the decline in the number of grains per panicle remains under 5.3%, while the reduction in panicle number remains below 2%, culminating in a lack of significant impact on overall yield. This outcome can be attributed to heightened fertilizer utilization efficiency and minimized fertilizer losses facilitated by the application of slow-release fertilizer and deep fertilization treatment [33]. Furthermore, mechanical weeding had a minimal effect on grain number per panicle and thousand-grain weight, compared to chemical weeding, and improved seed-setting rate, even if it has caused a slight reduction in panicle number. Therefore, while ensuring a high weeding rate and low seedling injury rate, the effect of mechanical weeding on yield was not significant when compared to chemical weeding.

4.2. Effect of Mechanical Weeding Combined with Slow-Release Fertilizer One-Time Reduction and Deep Application on Rice Quality

Regarding the processing quality and appearance of rice, the application of slow-release fertilizer can significantly improve the processing quality of rice within a certain range, and the deep application of fertilizer can ensure a high processing quality of rice, even in cases of a 20% reduction in N fertilizer usage [34,35]. The reduction in the used amount of N fertilizer can decrease the CR and CD, while slow-release fertilizer application has some effect on CR but not CD [36]. Mechanical weeding also has a slight effect on processing quality and can significantly reduce CR and CD.

Turning attention to the nutritional and cooking quality of rice, the application of slow-release fertilizer can increase the PC of rice, whereas the reduction of the used amount of N fertilizer can decrease PC, even if the deep application of fertilizer can weaken the impact of N fertilizer reduction [37]. Furthermore, mechanical weeding can enhance the PC of rice, which may be due to its tilling effect on soil, leading to increased soil permeability and nutrient uptake [38].

Conversely, the application of slow-release fertilizer can significantly reduce the AC of rice, while the reduction of the used amount of N fertilizer and mechanical weeding treatments can increase the AC. This resulted in a lower AC of rice under MW with LSR treatment, but it also led to higher GC under this treatment, as the lower AC resulted in a higher GC. Nonetheless, on the whole, the impacts of weeding and fertilization treatments on GC remain largely insignificant.

5. Conclusions

As far as the effect on the yield, the MW with LSR treatment showed no significant difference compared to CW with CK treatment: it could maintain a high yield level, higher than 9.2 t ha⁻¹ per year. The application of slow-release fertilizer and deep fertilization ensured a high panicle number by increasing the number of stems and tillers in pre-fertility and spike rate, even with a 20% reduction of N fertilizer. Simultaneously, the mechanical weeding approach amplified the seed-setting rate by bolstering the number of grains per panicle, panicle number, and thousand-grain weight, thus ensuring a high yield. Additionally, the MW with LSR treatment significantly improved the processing quality and appearance quality of rice. Furthermore, it slightly increased the protein content of rice, thereby improving its nutritional quality and reduced the amylose content, thus maintaining a high gel consistency and ensuring good cooking quality.

In light of these findings, this weeding and fertilization scheme, combining mechanical weeding and one-time deep application of a reduced amount of slow-release fertilizer, was considered applicable to rice cultivation. It was significant for promoting the efficient use of chemical fertilizers by transforming the traditional chemical control to a modern green prevention and control and reducing the use of chemical herbicides, thereby implementing an environmentally friendly cultivation compatible with the quality safety of agricultural products, and agro-eco-environmental protection.