The Potential Role of Zinc and Silicon in Improving Grain Yield and Lodging Resistance of Rice (Oryza sativa L.)

Abstract

Understanding the agronomic interventions that ensure higher crop yields and minimize their chances of failure is critical for meeting global nutritional demands. Rice is a staple food crop that is prone to lodging risk, particularly when higher yields are desired. The potential role of a combined application of Zinc (Zn) and Silicon (Si) in determining the grain yield and lodging resistance has been rarely investigated under field conditions. Thus, field trials were carried out to evaluate the grain yield and lodging resistance of rice at two different locations i.e., Qionghai and Wuzhishan, under three levels of Zn (0, 40, and 80 kg ha⁻¹) and Si (0, 120, and 240 kg ha⁻¹). The results showed that Zn application at the rates of 40 and 80 kg ha⁻¹ increased rice yield by 9% and 5% at Qionghai, and by 5% and 6% at Wuzhishan, respectively. The improved grain yield due to Zn application could be attributed to the increased panicles m⁻², splikelets m⁻², and aboveground biomass. Meanwhile, Zn failed to show any remarkable impact on stem and root lodging susceptibility. Conversely, no significant influence of applying Si on grain yield was observed, while its application at the rates of 120 and 240 kg Si ha⁻¹ enhanced the stem and root lodging resistance (denoted by their respective safety factors, for stem (SF_s) and for root (SF_r) by 32% and 22% at Qionghai, and by 11% and 34% at Wuzhishan, respectively, compared to zero Si application. The improved lodging resistance in terms of SF_s and SF_r could be ascribed to the increased stem bending strength and anchorage strength, while self-weight moment of whole plant decreased. In summary, a beneficial role of Si in lodging resistance and Zn in yield enhancement were evidenced in the present study across the two sites. It can be concluded that by combining 40 kg Zn ha⁻¹ with 120 kg Si ha⁻¹, both grain yield and lodging resistance could be simultaneously improved in rice crops.

1. Introduction

Nearly half of the global population relies on rice (Oryza sativa L.) as a staple food [1]. To maintain nutritional security worldwide, it is imperative that rice production keeps pace with the growing population. With the present pace of population increase, the demand for rice is expected to rise by 38% within the next 30 years [1]. Although a precise interpretation of the crop production trends is not easy because of the complex confounding impacts of climate, management practices, genetics, and their interactions, a stagnation or even a declining trend in yields of the three major cereals, i.e., maize, rice, and wheat, has been reported [2,3]. Such declining trends in yield are raising serious concerns over global food security [4]. The pivotal role of plant nutrition-based research in meeting nutritional needs is evident from the fact that more than 60% of the currently cultivated global land is still confronted with disorders (including both deficiencies and toxicities) of various minerals [5]. These issues are particularly aggravated in a country like China that has been under various degrees of food production pressures, as it is not only home to one-fifth of the world’s population but also plays a key role in exporting different food crops to other nations [6]. To cope with the challenge of increased production, Chinese farmers most often non-judiciously apply different inorganic fertilizers, which lead to reduced nutrient use efficiency and environmental pollution [7]. Thus, balanced crop nutrition remains one of the mainstays of agricultural research.

Micronutrient fertilizers have been developed as novel and efficient alternatives to conventional fertilizers [8]. These fertilizers greatly reduce the chances of pollution and the hazards associated with chemical fertilization, while also improving fertilizer application efficiency [9]. Micronutrients may also enhance plant metabolism owing to their distinctive physicochemical features [10]. Moreover, in comparison to the requirements and cost of chemical fertilizers, micro-nutrient fertilizers are more economically affordable and require smaller quantities for crops [9]. Zinc (Zn), being an essential nutrient, is equally crucial for both plants and human beings. In plants, it promotes growth, development, and yield by controlling different cellular, physiological, and metabolic processes, such as production of chlorophyll, and it is a key structural, enzymatic, and regulatory constituent of various proteins and enzymes [11,12]. It is not readily available to the majority of crops, including rice [13]. It also interacts with other nutrients, such as Ferrum (Fe), and thus modulates their uptake and translocation in rice crops [14]. Deficiency of Zn is amongst the most commonly occurring micronutrient adversities, particularly in lowland paddy areas, and results in reduced growth, chlorosis, smaller leaves, and sterility as well as low grain yield and nutritional quality [15,16,17]. Moreover, the ubiquity of Zn-lacking soils worldwide has exposed one-third of the global population Zn deficiency [18], including children less than five years old, who become extremely vulnerable to the risks of various infectious diseases and stunted growth [19]. Meanwhile, rice crops grown in a field with low plant-available Zn will also result in grains having suboptimal Zn concentration for humans [20].

Although the benefits of Silicon (Si) in crop husbandry have been recognized for more than 150 years, it is not categorized as an essential mineral element for a crop, as it can complete its life cycle in its absence [21]. Yet, it is categorized as a beneficial element for many agriculturally prominent crops, as it can enhance their growth, production, and even quality, particularly under stressful conditions, by providing tolerance against various biotic stresses, such as pathogens and insects, and abiotic stresses including drought, salt, and imbalance of nutrients and metals [22]. Rice, being a prominent accumulator of Si and a key scientific model organism, has attracted a plethora of researchers to conduct experiments on its uptake and utilization of Si [23]. Based on its important role in rice husbandry, Si has been given the status of an “agronomically essential element” in Japan and is frequently added to rice fields [24].

Lodging is a key constraint in many agroecosystems as it not only hinders the efficiency of mechanical harvesting but also adversely affects grain yield (up to 80% of high-yielding cropping systems) and the quality of rice [25,26]. It also increases the need for the drying of grains, resulting in an overall increased cost of production [27]. Hence, it is imperative to identify agronomic interventions, including fertilizer management, that might effectively mitigate the risk of rice lodging while concurrently enhancing grain production and thereby boost food security. Unfortunately, crop lodging is most often overlooked in any fertilizer management intervention that aims for higher crop productivity [27,28]. As a result, the need to enhance crop yields has led to the prevalence of crop lodging and the associated risks [29]. Thus, understanding the potential tradeoff between crop lodging and high yield has become one of the most difficult conundrums for researchers. Nonetheless, limited research has been undertaken to address this knowledge gap or elucidate the tradeoff between seed production and lodging hazards. Both stem buckling and root anchorage failure (albeit interconnected) occur differently in response to rising fertilizer application levels [27,28,30]. Meanwhile, it has been well recognized that to reduce the risk of crop lodging more effectively, the causal reasons for the two types of lodging should be distinguished from one another. Some of the key morphological and mechanical features have been already well-explained, along with their precise differences [27].

Until now, most studies have examined how different fertilizer techniques affect stem lodging, while root lodging has been the subject of little research. In addition, only a few researchers have examined lodging of both stem and root in rice plants in response to different fertilizers. Meanwhile, the influence of both Zn and Si on increasing yield and lodging resistance has been studied individually. However, the impact of their co-application on rice yield and lodging has not yet been well documented. Therefore, these trials were carried out with the objectives to (1) determine the influence of co-application of Zn and Si fertilizers on grain yield and its components; (2) study the response of rice lodging and the associated traits to different rates of Zn and Si; (3) unravel whether Zn and Si affect stem and root lodgings in the same manner. This research offers novel insights regarding the fertilizer interventional strategies that aim to benefit from co-application of Zn and Si for rice crops to enhance grain yield and lodging resistance.

2. Materials and Methods

2.1. Experimental Sites, Treatments, and Design

Field trials were conducted at two different sites i.e., Qionghai and Wuzhishan of Hainan province, China, from February to June 2023. Two hybrid rice varieties, i.e., Mianwuyou 128 and Huayou 218, were sown in Qionghai and Wuzhishan, respectively. Various soil parameters were assessed before the start of the trials (Table S1). Two factors, i.e., Zn and Si fertilizers, were evaluated. The first factor comprised three different levels of Zn (0, 40, and 80 kg ha⁻¹), while the second factor included three Si levels (0, 120, and 240 kg ha⁻¹). Each treatment had three replications. Nitrogen (N), Phosphorus (P), and Potassium (K) fertilizers were applied as per local farming practices at a rate of 180 kg N ha⁻¹, 90 kg P₂O₅ ha⁻¹, and 90 kg K₂O ha⁻¹. Urea fertilizer (having 46% N) was added in three splits, i.e., 60% as basal before transplantation of the seedlings, 20% at the active tillering, and the remaining 20% at the start of the panicle formation stage. P and K fertilizers were also added at the time of sowing. The area of each subplot was 40 m² (4 m × 10 m) at both sites, with the same row-to-row and plant-to-plant distances of 20 cm. A uniform plant density of 2.5 × 10⁵ hills ha⁻¹ was maintained in each plot after thinning. The implementation of various crop management practices adhered to the standard recommendations provided by the local agricultural technology extension office.

2.2. Data Collection

2.2.1. Determination of Yield Parameters

For measuring grain yield-related traits, 12 hills were randomly harvested at maturity from a 5 m² area. After counting the number of panicles, the plant samples were separated into straw and panicles. The panicles were hand-threshed, and the filled spikelets were separated from the unfilled spikelets by submerging them in tap water. Three subsamples of 30 g of filled spikelets were used to count their number. Likewise, all the unfilled spikelets were also counted to find their number. The dry weights of straw, rachis, and filled and unfilled spikelets were recorded after oven-drying at 80 °C to a constant weight. Aboveground total biomass was the total dry matter of straw, rachis, and filled and unfilled spikelets. Grain weight, spikelets panicle⁻¹, grain-filling percentage (100 × filled spikelet number/total spikelet number), and harvest index (100 × filled spikelet weight/total aboveground biomass) were calculated. The grain yield was measured from a 5 m² area for each treatment and adjusted to the standard moisture content of 0.14 g H₂O g⁻¹ fresh weight [31].

2.2.2. Determination of Stem and Root Lodging Parameters

In each experiment, eight representative plants from each plot were randomly chosen, labelled, and cut off at the soil surface level approximately 10 days before maturity. The plant samples were then brought to the laboratory for biomechanical analysis after measurement of anchorage strength (S_r). The methods proposed by Wu and Ma et al. were employed to assess various stem and root lodging -related parameters [27,28]. Briefly, the center of gravity of the stem (h_s, cm) as well as its fresh weight (m_s, g) were recorded, excluding the basal stem. Subsequently, the base stem and its corresponding aboveground plant were joined together as a full plant. The center of gravity of the whole plant (h_r, cm), diameter of the stem, mass density, fresh weight (m_r, g), and plant height (cm) of the whole plant were also recorded. A three-point bending test with a device (Multitest 2.5-I equipped with a 100 N load cell, Mecmesin, Slinfold, UK) was then used to evaluate the stem breaking resistance and the highest bending force (F_max). The equation S_s = F_max L/4 (L = 5 cm) was used for determination of the stem’s maximum breaking resistance (S_s, Nm). Basal stem segment mass density (mg cm⁻¹) and average diameter (mm) were also noted.

2.2.3. Determination of the Safety Factor as an Indicator of Lodging Resistance

The susceptibility of the crops towards lodging was assessed by a safety factor (SF) technique that denotes the number of times a support organ like the stem or root can bear the self-weight moment of the organ it supports [32]. The formulae used for the safety factor of the stem (SF_s) and root (SF_r) and other relevant details are described elsewhere [28,32,33]. The usefulness of both SF_s and SF_r in assessing stem and root lodging, respectively, prior to their occurrence under field conditions has been well demonstrated [32]. A higher SF value corresponds to an increased resistance to crop lodging and vice versa.

2.2.4. Data Analysis

Two-way analysis of variance (ANOVA) was employed to evaluate the impact of all levels of Zn and Si on yield-related traits, lodging-associated parameters, and their interactions. Data were analyzed with the help of SAS (Version 9.3, SAS Institute, Cary, NC, USA). Treatment mean comparisons were conducted using the least significant difference (LSD) method with a conservative letter grouping approach at a confidence level of 95%. The R program was used to calculate Pearson’s correlations (version 4.0.0, R Core Team, Vienna, Austria). The statistical analyses were conducted using a significance threshold of 5%. Figures were generated using Sigma Plot (version 12.5; SYSTAT, San Jose, CA, USA) and R software. The observed characteristics were classified into two primary categories: (1) yield related parameters and (2) traits associated with lodging. Principal component analysis (PCA) for each category was performed via R software. Prior to conducting a PCA, the data were verified to ensure normality and homogeneity. The PCA yielded ordination diagrams in biplot format.

3. Results

3.1. Grain Yield and Related Parameters

Our findings revealed that Zn application has a more pronounced effect on grain yield of the tested rice genotype than Si at both the experimental sites (p < 0.05) (

Table 1. Panicles m⁻², spikelets panicle⁻¹, spikelets m⁻² (×10⁴), grain filling (%), grain weight (mg), grain yield (t ha⁻¹), aboveground biomass (t ha⁻¹), harvest index (%) as affected by various Si (kg Si ha⁻¹) and Zn (kg Zn ha⁻¹) application rates at the two experimental sites.

Sites	Zn Rates	Si Rates	Panicles m−2	Spikelets Panicle−1	Spikelets m−2	Grain Filling	Grain Weight	Grain Yield	Aboveground Biomass	Harvest Index
Qionghai
	0	0	344.5 b	125.3 a	4.32 a	74.7 b	26.3 b	10.15 a	17.6 b	48.3 a
		120	363.7 ab	130.7 a	4.67 a	75.12 b	27.40 a	10.52 a	18.68 ab	51.11 a
		240	410.8 a	112.1 a	4.61 a	84.85 a	27.64 a	10.41 a	22.88 a	47.43 a
		Mean	373.0 B	122.7 A	4.54 B	78.23 A	27.12 A	10.36 B	19.72 B	48.96 A
	40	0	419.2 a	125.1 a	5.24 a	74.98 c	27.58 a	11.04 a	23.68 ab	46.05 a
		120	384.0 a	121.0 a	4.64 a	77.08 b	26.72 a	11.72 a	20.04 b	47.91 a
		240	430.0 a	120.0 a	5.16 a	82.20 a	27.43 a	11.13 a	25.25 a	46.05 a
		Mean	411.0 AB	122.0 A	5.01 A	78.09 A	27.25 A	11.30 A	22.99 A	46.67 A
	80	0	417.1 a	128.5 a	5.31 a	73.71 b	27.16 a	10.69 a	23.29 a	45.91 a
		120	464.3 a	110.2 b	5.10 a	86.48 a	27.59 a	10.92 a	24.06 a	50.53 a
		240	408.3 a	135.6 a	5.53 a	77.88 b	27.37 a	11.02 a	23.56 a	50.23 a
		Mean	429.9 A	124.8 A	5.31 A	79.35 A	27.37 A	10.88 AB	23.64 A	48.89 A
Wuzhishan
	0	0	204.9 b	126.6 b	2.59 b	73.43 ab	28.60 a	8.01 b	14.02 a	43.97 b
		120	227.0 a	130.5 b	2.96 a	72.29 b	28.29 a	8.66 a	12.95 a	48.93 a
		240	200.6 b	157.7 a	3.16 a	75.24 a	28.57 a	8.58 a	15.02 a	50.62 a
		Mean	210.8 B	138.3 A	2.91 B	73.65 B	28.48 A	8.41 B	14.00 B	47.84 A
	40	0	233.2 a	146.7 a	3.42 a	70.70 b	28.39 a	8.71 a	15.44 a	47.60 a
		120	224.9 a	151.6 a	3.40 a	75.75 a	28.32 a	8.79 a	15.26 a	46.82 a
		240	216.7 a	153.9 a	3.35 a	76.63 a	28.21 a	9.03 a	15.72 a	47.50 a
		Mean	224.9 AB	150.8 A	3.39 A	74.36 AB	28.31 A	8.85 A	15.47 A	47.31 A
	80	0	209.5 a	137.7 a	2.89 a	72.11 c	28.30 a	8.83 a	13.26 a	49.00 b
		120	237.2 a	140.8 a	3.35 a	76.59 b	28.18 a	9.01 a	15.14 a	45.79 b
		240	237.3 a	150.6 a	3.57 a	86.94 a	27.53 a	8.95 a	15.58 a	52.57 a
		Mean	228.0 A	143.0 A	3.27 AB	78.55 A	28.00 A	8.93 A	14.66 AB	49.12 A

Two hybrid rice varieties, i.e., Mianwuyou 128 and Huayou 218, were sown in Qionghai and Wuzhishan, respectively. Within a column for each Zn treatments, means followed by different small alphabetical letters show the significant differences among the three Si treatments according to the least significant difference test at the 95% level of confidence. Similarly, means followed by different capital alphabetical letters show significant differences among the three Zn treatments.

). Applying 40 kg and 80 kg Zn ha⁻¹ at Qionghai and Wuzhishan produced maximum grain yields of 11.3 t ha⁻¹ and 8.93 t ha⁻¹, respectively, when averaged across all the three rates of Si. In contrast, no application of Zn at both sites resulted in the minimum grain yield when averaged across the three Si rates. In the case of Si, the grain yield under different Si rates was not significantly influenced at various Zn rates, except under 0 kg Zn ha⁻¹ at Wuzhishan (at which zero kg of Si application resulted in significantly lower grain yield than the plots where 40 and 80 kg Si ha⁻¹ was applied). Interaction data revealed that application of 40 kg Zn and 120 kg Si ha⁻¹ at Qionghai resulted in the highest grain yield of 11.72 t ha⁻¹, while no application of Si and Zn at Wuzhishan produced 8.01 t ha⁻¹ of grain yield, which was the lowest value across various treatment combinations of the two experimental sites. Zinc application at the rates of 40 and 80 kg ha⁻¹ enhanced seed production by 9% and 6%, respectively, at Qionghai, while the corresponding increases were 5% and 6% at Wuzhishan in comparison with control plots. Meanwhile, a comparison of the two sites revealed greater grain yield at Qionghai than Wuzhishan, irrespective of different treatment combinations.

Analysis of yield-related attributes revealed subsequent increments of 10% and 15% under 40 and 80 kg Zn ha⁻¹ at Qionghai, and 6% and 8% at Wuzhishan, respectively, when compared with zero Zn application. Likewise, the number of spikelets m⁻² in Qionghai was increased by 10% (40 kg Zn ha⁻¹) and 17% (80 kg Zn ha⁻¹), while at Wuzhishan it was enhanced by 16% (40 kg Zn ha⁻¹) and 12% (80 kg Zn ha⁻¹) when compared with the control plots. Meanwhile, the grain filling when averaged across the three Si rates was not influenced by Zn rates at Qionghai, while at Wuzhinshan, 80 kg Zn ha⁻¹ resulted in significantly higher filling (7%) than the control treatment when averaged across the three Si rates. In the case of spikelets panicle⁻¹, mostly no significant and consistent effect of the different levels of Zn and Si was observed at both sites. Meanwhile, the aboveground biomass in Qionghai was improved by 17% (40 kg Zn ha⁻¹) and 20% (80 kg Zn ha⁻¹), while at Wuzhishan, it was increased by 11% (40 kg Zn ha⁻¹) and 5% (80 kg Zn ha⁻¹), respectively, in comparison with 0 kg Zn ha⁻¹ (Table 1). There was no significant difference in the dry weight of each internode under Zn treatment (Figure 1). Grain weight and harvest index were also influenced by various levels of both Zn and Si in a consistent manner at the two sites, although the differences under various treatment combinations were mostly non-significant (Table 1).

3.2. Lodging-Associated Traits

Some of the lodging-related traits were significantly influenced by various Si application rates at both sites (Figure 2, Figure 3, Figure 4 and Figure 5 and

Table 2. Gravity center height (cm), stem diameter (mm), plant height (cm), fresh weight (g), and mass density (mg cm⁻¹) as affected by the Si (kg Si ha⁻¹) and Zn (kg Zn ha⁻¹) application rates across the two site environments.

Site	Zn Rates	Si Rates	Gravity Center Height (cm)	Stem Diameter (mm)	Plant Height (cm)	Fresh Weight (g)	Mass Density (mg cm−1)
Qionghai
	0	0	48.43 a	4.28 a	91.04 a	62.65 a	4.08 a
		120	48.23 a	4.23 a	94.23 a	64.10 a	4.38 a
		240	47.08 a	4.68 a	94.21 a	75.80 a	4.76 a
		Mean	47.92 B	4.40 A	93.16 B	67.52 B	4.41 A
	40	0	53.92 a	4.09 a	98.01 a	71.39 b	4.19 a
		120	47.40 b	4.55 a	95.21 a	78.11 ab	4.48 a
		240	49.53 b	4.76 a	98.94 a	84.33 a	5.06 a
		Mean	50.28 AB	4.47 A	97.39 A	77.94 A	4.58 A
	80	0	55.52 a	5.19 a	97.53 a	77.00 a	3.92 a
		120	48.83 b	4.60 a	101.43 a	77.05 a	4.61 a
		240	50.78 b	4.58 a	99.07 a	77.06 a	4.60 a
		Mean	51.71 A	4.79 A	99.34 A	77.03 A	4.38 A
Wuzhishan
	0	0	53.73 a	4.30 a	101.32 a	119.74 a	6.10 a
		120	53.53 a	4.10 a	102.75 ab	114.33 a	5.52 ab
		240	53.00 a	4.11 a	98.48 b	107.47 a	5.03 b
		Mean	53.42 A	4.17 A	100.85 A	113.85 A	5.55 A
	40	0	53.93 a	3.99 b	100.41 a	111.96 a	5.26 a
		120	52.53 ab	3.82 b	99.52 a	109.26 a	5.39 a
		240	50.53 b	4.42 a	97.33 a	116.29 a	5.89 a
		Mean	52.33 A	4.08 A	99.09 A	112.51 A	5.52 A
	80	0	54.40 a	4.04 a	101.10 a	115.27 a	5.56 a
		120	52.53 a	4.30 a	98.60 ab	106.89 a	5.21 a
		240	51.53 a	4.51 a	95.23 b	114.98 a	5.67 a
		Mean	52.82 A	4.28 A	98.31 A	112.38 A	5.48 A

Two hybrid rice varieties, i.e., Mianwuyou 128 and Huayou 218, were sown in Qionghai and Wuzhishan, respectively. Within a column for each Zn treatment, means followed by the different small alphabetical letters show significant differences among the three Si treatments according to the least significant difference test at the 95% level of confidence. Similarly, means followed by different capital alphabetical letters show significant difference among the three Zn treatments.

). Compared to the non-treated plot, the lodging resistance of rice was generally improved under applications of 120 kg and 240 kg Si ha⁻¹ at both the experimental sites and all three different levels of Zn fertilizer. For example, under the 80 kg Zn ha⁻¹, SF_s was significantly increased by 52% (120 kg Si ha⁻¹) and 32% (240 kg Si ha⁻¹) and SF_r was significantly increased by 80% (120 kg Si ha⁻¹) and 65% (240 kg Si ha⁻¹) in Qionghai when compared with zero application of Si. Similarly, in Wuzhishan, the increase in SF_s under 80 kg Zn application was 9% (120 kg Si ha⁻¹) and 4% (240 kg Si ha⁻¹), and the increase in SF_r under 80 kg Zn application was 19% (120 kg Si ha⁻¹) and 37% (240 kg Si ha⁻¹) compared with no application of Si (Figure 4 and Figure 5).

In contrast to Si, the three Zn application levels did not exhibit any significant effect on SF (Table S2). The Si application at rates of 120 and 240 kg Si ha⁻¹ increased the stem and root average lodging resistance (represented by SF) by 32% and 22% at Qionghai and by at 11% and 34% at Wuzhishan when averaged across the three Zn application levels, respectively, compared to zero Si application. Si levels showed more resistance to lodging parameters, whereas zinc levels showed no negative or constant effect on lodging parameters. It is evident from this box plot that at both sites, the SF_s values were higher than the SF_r, showing greater risk of root lodging than the stem.

The results further revealed that the gap between SF_s and SF_r was greater at Qionghai, with 222% more resistance of SF_s than SF_r. While at Wuzhishan, the SF_s was higher by only by 22% when compared with the SF_r (Figure 6).

3.3. PERMANOVA, Pearson’s Correlations, and PCA Analyses

Results of the PERMANOVA further confirmed the trends obtained for both Zn and Si (Table S3). A significant impact of Zn on yield-related attributes was found, while Si did not significantly influence these parameters. Conversely, the lodging-related traits were significantly affected by Si rates, while the three Zn rates did not significantly influence these traits. Interestingly, the effect of the two sites on both categories of traits (yield- and lodging-related) was significant. Pearson’s correlations represented in a heatmap showed that grain yield was strongly and positively correlated with aboveground biomass (0.79 ***), spikelets m⁻² (0.81 ***), and panicles m⁻² (0.84 ***) (Figure 7), while PCA analysis showed that grain yield was oriented in the same direction as grain filling, panicles m⁻², and harvest index at Wuzhishan and spikelets m⁻² and aboveground biomass at Qionghai (Figure 8). Similarly, S_s depicted a positive and strong correlation with mass density (0.76 ***), fresh weight (0.87 ***), M_r (0.82 ***), S_r (0.84 ***), SF_r (0.80 **), and M_s (0.79 ***), whereas SF_r exhibited a positive correlation with mass density (0.61 ***), fresh weight (0.79 ***), M_r (0.75 ***), and S_r (0.98 ***), as shown in Figure 7.

Pearson’s correlation further revealed that SF_r was positively correlated with S_r, M_r, stem diameter, gravity center height, fresh weight, mass density, spikelet panicle, grain weight, Ss, and Ms (Figure 7). Conversely SF_r showed a strong negative correlation with grain yield, panicles m⁻², spikelets m⁻², and aboveground biomass. Similarly, SF_s showed a positive correlation with grain filling and panicles m⁻² and a strong negative correlation with gravity center height, M_s, and M_r. Meanwhile, PCA showed that SF_s was positioned in the same direction as that of SF_r, S_s, and S_r at both experimental sites (Figure 8). In terms of lodging-related parameters, S_s was oriented in the parallel direction as the average stem diameter and S_r, suggesting that an increased Si application rate led to enhanced SF_r as well as SF_s across the two site environments (Figure 8).

4. Discussion

4.1. Grain Yield and the Associated Traits under Various Levels of Zn and Si

Our results demonstrated that Zn improved rice grain production significantly compared with no Zn fertilizer in both experimental locations and at the three Si rates. The maximum rice yield was attained while applying 40 kg ha⁻¹ of Zn, and further increment in Zn treatment (i.e., 80 kg ha⁻¹) led to a decline in yield at Qionghai. Such results are in conformity with the available literature, which suggests that within a particular range, a substantial positive association can be found between rice yield and Zn rate; however, excessive Zn applications are detrimental to rice and may not help in increasing yield [34]. Our results also depicted a similar trend, as by applying Zn, not only grain yield but other associated traits such as panicles m⁻², spikelets m⁻², and aboveground biomass were increased when compared with no Zn fertilizer. Previously, Tian et al. [35] stated that an increased seed production of rice is highly dependent on dry matter accumulation and distribution. The yield gain under Zn application can be also explained in light of its key role in regulating various enzymatic activities of plants, such as the chlorophyll formation and the metabolism of carbohydrates. Similarly, Zn is also vital for synthesis of assimilates through photosynthesis, which ultimately facilitates yield formation [36]. In contrast to Zn, the application of Si at various levels did not reveal any significant effect on yield and its related traits. A previous study conducted in Iran revealed a positive response of rice to the application of Si [13]. This necessitates a more detailed and in-depth study focusing on a diverse set of rice genotypes, which can enhance our understanding regarding the impact of Si on rice.

4.2. Crop Lodging as Influenced by the Application of Zn and Si

Our results revealed that application of Si improved rice stem lodging resistance, while application of Zn fertilizer did not exhibit any consistent effect on lodging. In our experiment, compared to 0 kg Si ha⁻¹, 120 kg Si ha⁻¹ treatment significantly increased the SFs (by 22%) and Ss (by 16%) of rice and decreased the gravity center height, plant height, and fresh weight. Meanwhile, correlation analysis and PCA analysis also showed that Ss was significantly and positively correlated with SFs, mass density, and stem diameter of the third internode. Thus, it could be inferred that the increase in stem lodging resistance by Si fertilizer was mainly attributable to the enhancement of mass density, stem diameter, and Ss [37]. In addition, the application of Si can directly increase the contents of silicate and carbohydrates in cells and promote keratinocyte growth and thus can facilitate the syntheses of lignin and cellulose in rice plants [38]. Such increases in lignin and cellulose contents can further enhance the physical strength of the stem and thus increase the stem lodging resistance of rice [39].

The root lodging resistance of rice was also assessed in this study, and we found that application of Si fertilizer increased S_r and SF_r. For example, compared to 0 kg Si ha⁻¹, 120 kg Si ha⁻¹ and 240 kg Si ha⁻¹ increased SF_r by 23% and 32%, and S_r by 18% and 31%, respectively. The Si fertilizer was applied by broadcasting in this study, which might have enhanced the growth of the root and root biomass accumulation to reduce root lodging risk. Several studies have indicated that an increase in root biomass accumulation as well as a larger allocation ratio of dry matter to the roots than the aboveground parts are crucial factors in enhancing resistance to root lodging [28,33,40]. A recent meta-analysis also showed that broadcasting Si fertilizers was more effective than foliar spraying in increasing lodging resistance [30]. In summary, the results of our study suggest that adding 40 kg Zn ha⁻¹ and 120 kg Si ha⁻¹ can be recommended for rice production, as the combination of these two doses can synergistically improve grain yield and crop lodging resistance (both stem and root lodging).

4.3. A Comparison between Stem Lodging and Root Lodging

We also assessed the probability of lodging risk resulting from stem or root lodging in rice by comparing SFs with SFr, based on the technique previously suggested by Wu et al. [32]. Such comparisons help researchers and growers to determine, in a field prone to both stem and root lodgings, which type of lodging occurs most commonly [32]. Such comparisons also provide an opportunity to intervene in a timely fashion by adopting appropriate remedial measures to minimize lodging and the associated threats. Our results showed that the value of SF_r was smaller than the SF_s, as the S_r was smaller than S_s, at both Qionghai and Wuzhishan sites. This implies that rice crops are more susceptible to root lodging rather than stem lodging. The greater resistance of rice crops to stem lodging than root lodging suggests that it is crucial to prioritize the enhancement of S_r as a strategy to enhance root lodging resistance. In this regard, rice genotypes that have a vigorous root system with an enhanced capability to penetrate deep into the soil can resist root lodging more than those with shallow roots [41]. Thus, the breeding and subsequent cultivation of crop genotypes having a robust root system should be exploited in the future [32]. This would become more important in the case of direct-seeded rice, which is usually more prone to lodging due to its shallower roots than transplanted rice, having a well-developed and deeper root system and thus exhibiting better lodging resistance [41]. Meanwhile, partitioning of more assimilates towards the roots by a crop can be a promising strategy that will enable the plant to enhance its tolerance against root lodging [32]. Thus, future research aiming at improving the lodging resistance of rice should be directed towards understanding the root system of rice.

5. Conclusions

Application of Zn improved rice grain yields significantly compared to control plots at both the experimental sites. The higher grain yield of the Zn-applied plots was complemented by the maximum number of panicles m⁻², spikelets m⁻², and shoot biomass. Conversely, Zn application failed to exhibit any substantial influence on both stem and root lodging susceptibilities. On the other hand, Si did not remarkably affect yield, but it increased both stem and root lodging resistance significantly compared to zero Si application at both experimental sites, regardless of Zn application rates. The enhanced lodging resistance under Si application could be attributed to the increased Ss and Sr as well as the decreased Ms and Mr. In summary, a beneficial role of Si in lodging resistance, and that of Zn in yield enhancement, was evidenced at both locations. Among the various treatment combinations, 40 kg Zn ha⁻¹ and Si application at a rate of 120 kg ha⁻¹ could be recommended for enhancing the grain yield and reducing the lodging risk simultaneously in rice crops. Meanwhile, the root lodging susceptibility was greater than the stem lodging susceptibility, which implies that agronomic and breeding strategies aimed at enhancing root anchorage strength can potentially reduce the risk of lodging in rice crops and must be prioritized for future research.