Enhancing initial growth of transplanted rice under phosphorus deficiency: combined effects of tillering quantitative trait locus and P-dipping technique

ABSTRACT

Phosphorus (P) deficiency severely constrains rice production in sub-Saharan Africa. Previous studies showed that P-dipping, involving localized application of P near the root zone, and MP3, a natural allele of OsTB1/FC1 that enhances rice tillering, effectively improve rice growth and productivity in P-deficient soils in Madagascar. In the present study, we investigated the combined impact of these two technologies on the initial growth of transplanted rice in P-deficient soils using potted plants. Our experiments revealed that near-isogenic lines for MP3 and fc1 (the loss-of-function allele of OsTB1/FC1) promoted tillering and increased initial shoot biomass compared with the parental cultivar ‘Takanari’ when combined with P-dipping. The rise in shoot biomass would be attributed to increased P-uptake in the shoots, which were brought about by a continuous P-supply from the root zone due to P-dipping. Consequently, these combinations have the potential to enhance rice productivity in P-deficient paddy fields in sub-Saharan Africa.

GRAPHICAL ABSTRACT

KEYWORDS:

• P-dipping
• P-deficiency
• quantitative trait locus
• rice
• tillering

Introduction

Phosphorus (P) deficiency is a significant obstacle to rice production in sub-Saharan Africa due to low soil P-content and the high P-fixing capacity of active Al- and Fe-oxides (Bekunda et al., 2010; Nishigaki et al., 2019; Saito et al., 2019). P-deficiency restricts rice tillering and stunts plant growth and development, which leads to low grain yield (Andrianary et al., 2021; Dobermann & Fairhurst, 2000). Although the application of fertilizers can enhance rice production, local farmers often lack the financial means to purchase sufficient amounts (Vanlauwe et al., 2014), resulting in inadequate fertilization practices (Tsujimoto et al., 2019). Moreover, P is a finite and nonrenewable resource and may be depleted within this century (Cordell et al., 2009). Therefore, it is crucial to develop effective P-fertilizer application practices (Rakotoson et al., 2022) and cultivate rice varieties adapted to P-deficient environments (Ismail et al., 2007).

Our recent studies in the central highlands of Madagascar have demonstrated that dipping rice seedlings in P-enriched slurry during transplanting promotes the initial growth of rice plants, resulting in an increase in grain yield in P-deficient lowlands (Rakotoarisoa et al., 2020). The P-dipping method creates a soluble P-hotspot around the transplanted roots, enabling efficient P-uptake even in high P-fixing soils and improving P-use efficiency (Oo, Tsujimoto, Rakotoarisoa, Kawamura, et al., 2020). Currently, efforts are under way to disseminate optimized P-dipping practices to local farmers in Madagascar (Oo & Tsujimoto, 2023).

Our previous studies have identified MP3 as a quantitative trait locus (QTL) and a natural allele of OsTB1/FC1 that moderately promotes rice tillering and increases panicle number (Takai et al., 2014, 2023). Given the limited tillering and panicle formation under P-deficiency, this QTL’s function may mitigate yield losses. On-farm trials conducted in the P-deficient lowlands of Madagascar confirmed that the near-isogenic line with MP3 (NIL-MP3) exhibited a 1.2-fold higher tiller/panicle number compared with its parent cultivar ‘Takanari’ (Takai et al., 2021). Furthermore, a NIL carrying fc1, the loss-of-function allele of OsTB1/FC1 (NIL-fc1), markedly enhanced tillering by 2.4-fold compared with ‘Takanari’ in nutrient-rich lowlands, despite lodging during grain filling (Takai et al., 2023). Considering the severe limitation of tillering in P-deficient environments, fc1 may be a more favorable choice to achieve higher tiller/panicle numbers.

This study aims to investigate the combined effect of P-dipping and tillering QTL on the initial growth of potted rice plants in P-deficient soils. By integrating these agronomic and breeding approaches, a more efficient rice cultivation system for P-deficient lowlands would be established.

Materials and methods

Plant materials

Two-year greenhouse pot experiments were conducted at the Japan International Research Center for Agricultural Science (JIRCAS) in Tsukuba, Japan (36°3′N, 140°5′E) from May to July in 2021 and 2022. The average daily temperature in the greenhouse was 25.4°C–33.2°C throughout the experimental periods. The study included three rice genotypes:”Takanari,” NIL-MP3, and NIL-fc1. ‘Takanari’ is a high-yielding indica cultivar in Japan (Imbe et al., 2004). NIL-MP3 (Takai et al., 2021) and NIL-fc1 (Takai et al., 2023) are near-isogenic lines with homozygosity for the Koshihikari allele and the loss-of-function allele of OsTB1/FC1, respectively, in the genetic background of ‘Takanari.’ OsTB1/FC1 is a gene that encodes a TCP family transcription factor that functions as a negative regulator of tiller bud outgrowth (Minakuchi et al., 2010; Takeda et al., 2003). After soaking seeds of the three rice genotypes in water at 30°C for two days, the germinated seeds were sown with one seed per cell in seedling cell trays filled with granular soils. Then, 15-day-old seedlings were transplanted with one seedling per pot on 10 May 2021, and 9 June 2022. Rice plants were cultivated under flooded conditions.

Soil conditions and P-treatments

Volcanic soil was collected from the forest subsoil (20–40 cm layer) in the experimental farm of JIRCAS to ensure a low availability of P. The soil was air-dried and sieved through an 8.0 mm mesh. It exhibited sandy loam characteristics with a pH of 5.7 and a high P-retention capacity of 99%, as reported by Oo, Tsujimoto, Rakotoarisoa, Kawamura, et al. (2020) along with other soil properties. In the 2021 and 2022 experiments, 1/10000a and 1/5000a plastic pots were filled with 1 kg and 3 kg of prepared soil, respectively. Prior to transplanting, ammonium nitrate (NH₄NO₃) and potassium sulfate (K₂SO₄) were applied at rates of 100 mg nitrogen (N) kg⁻¹ soil and 100 mg potassium (K) kg⁻¹ soil, respectively, to ensure an adequate supply of N and K.

A randomized complete block design with five replications was used, combining three P-application methods, i.e. no P-application (control), P-incorporation, and P-dipping, with three rice genotypes. For the P-incorporation treatment, triple super phosphate [TSP; 42% phosphorus pentoxide (P₂O₅)] was added to the soil at a rate of 250 mg P₂O₅ kg⁻¹ soil during puddling. For the P-dipping treatment, a soil slurry was prepared by mixing 10 g of TSP, 100 g of experimental soil, and 50–70 ml of distilled water following the method described by Oo, Tsujimoto, and Rakotoarisoa (2020). Rice seedlings were then dipped in the slurry for a few minutes, and seedlings with slurry attached to their roots were transplanted into the pots. The estimated amount of P transferred through P-dipping at transplanting (P-enriched slurry attached to seedling roots) was approximately 90 mg P₂O₅ pot⁻¹.

Measurements

The number of tillers was recorded, and plant samples were collected at 25 days (early tillering stage) and 39 days (late tillering stage) after transplanting in 2021 and 2022, respectively. To determine shoot biomass, the aboveground parts were cut at ground level and oven-dried at 70°C for two days. The shoot P-concentration was measured using the molybdate blue method (Murphy & Riley, 1962) after dry ashing at 550°C for 2 h and digesting with 0.5 M HCl. The shoot P-content (mg pot⁻¹) was calculated by multiplying the P-concentration by the shoot biomass. After removing the shoots, the roots were thoroughly washed with tap water and preserved in 50% ethanol for subsequent measurements. The roots were scanned using an Epson Pro-selection X980 scanner, and the images were analyzed using WinRhizo Pro (Regent Instrument, Quebec, Canada) to estimate the total root length and root surface area. After root analysis, the root samples were oven-dried at 70°C for two days, and their biomass was determined.

Statistics

Statistical analyses were performed using a general linear model through SPSS 23.0 software (IBM, Chicago, USA). Two-way ANOVA was used to assess the effect of P-application, genotype, and their interaction on the measured variables for each year. P-application, genotype, and their interaction were considered fixed effects, whereas replication was considered a random effect. Significance of fixed effects was assessed using Tukey’s honestly significant difference test at a 5% level.

Results

Tiller development

ANOVA revealed significant and consistent effects of P-application, genotype, and their interaction on tiller number during the sampling periods in both 2021 and 2022 (Figure 1a,b). P-dipping resulted in the highest number of tillers, followed by P-incorporation, whereas the control had the lowest count. Among the genotypes, NIL-fc1 consistently exhibited the highest number of tillers across all P-application treatments. Notably, in 2022, NIL-fc1 displayed a remarkable 6.2-fold increase in tiller number compared with that of ‘Takanari’ under the P-dipping treatment. Conversely, NIL-MP3 showed a significantly higher tiller count than”Takanari” only when combined with P-dipping.

Figure 1. Effects of three different P-application treatments on the number of tillers (a, b), shoot biomass (c, d), and P-content in shoots (e, f) in 2021 and 2022 among”Takanari,” NIL-MP3, and NIL-fc1. *, **, and *** indicate significant effects and interactions between P-application method (P) and genotype (G) at the 5%, 1%, and 0.1% levels, respectively. Different letters represent significant differences among genotypes across P-application treatments at the 5% level based on Tukey’s honestly significant difference test. Error bars represent standard deviation.

Shoot biomass, shoot P-content, and root development

ANOVA revealed significant effects of P-application on shoot biomass in both 2021 and 2022 (Figure 1c,d). P-dipping resulted in the highest shoot biomass, followed by P-incorporation, whereas the control had the lowest shoot biomass. In 2021 and 2022, P-dipping led to 2.4 and 2.3 times greater shoot biomass, respectively, compared with P-incorporation, despite the smaller amounts of applied P (90 mg P₂O₅ vs. 250 mg P₂O₅ pot⁻¹ and 90 mg P₂O₅ vs. 750 mg P₂O₅ pot⁻¹, respectively). In 2022, a significant interaction between P-treatment and genotype was observed for shoot biomass. Under the P-dipping treatment, NIL-MP3 exhibited a 12% increase and NIL-fc1 showed a significant 24% increase in shoot biomass compared with ‘Takanari.’ However, no significant genotype differences were found in shoot biomass under the control and P-incorporation treatments. Corresponding with our findings on shoot biomass, shoot P-content tended to increase by 22% in NIL-MP3 and significantly increased by 37% in NIL-fc1 compared with ‘Takanari’ combined with the P-dipping treatment, whereas no genotype differences were observed in the other P-treatments (Figure 1e,f). P-application also significantly influenced root-related traits, with P-dipping resulting in the highest root biomass, root length, and root surface area, followed by P-incorporation, whereas the control had the lowest values (Table 1). However, no genotypic differences were observed for these traits.

Table 1. Two-way ANOVA for root-related traits at 25 and 39 days after transplantation in 2021 and 2022, respectively, in”Takanari,” NIL-MP3, and NIL-fc1 grown in the pots with the three P-application treatments.

Year	P-application (P)	Genotype (G)	Root biomass (g pot^−1)		Root length (cm pot^−1)		Root surface area (cm^2 pot^−1)
2021	Control		0.11c		2942c		134c
	P-incorporation		0.29b		7545b		382b
	P-dipping		0.63a		8900a		612a
		Takanari	0.36a		6440a		379a
		NIL-e	0.35a		6254a		378a
		NIL-fc1	0.31a		6692a		371a
ANOVA
P			***		***		***
G			n.s.		n.s.		n.s.
P × G			n.s.		n.s.		n.s.
2022	Control		0.05c		1696c		71c
	P-incorporation		1.05b		21268b		1302b
	P-dipping		2.26a		30259a		2110a
		Takanari	1.11a		16882a		1138a
		NIL-MP3	1.32a		19706a		1298a
		NIL-fc1	1.02a		17836a		1129a
ANOVA
P			***		***		***
G			n.s.		n.s.		n.s.
P × G			n.s.		n.s.		n.s.

***indicates significance at the 0.1% level; n.s. indicates no significance.

Different letters within each P-application treatment indicate significant differences among the genotypes at the 5% level based on Tukey’s honestly significant difference test.

Shoot biomass and P-content per tiller

Considering the genotypic differences observed in shoot biomass and shoot P-content under the P-dipping treatment in 2022, we further examined these traits per individual tiller. ANOVA revealed significant effects of P-application, genotype, and their interaction on shoot biomass per tiller and P-content per tiller (Figure 2a,b). P-dipping resulted in the highest biomass and P-content per tiller, followed by P-incorporation, whereas the control showed the lowest values. However, in the case of P-dipping, NIL-MP3 showed decreases of 32% and 26% in biomass and P-content per tiller, respectively, compared with ‘Takanari.’ Similarly, NIL-fc1 showed a substantial decrease of 80% and 78% in biomass and P-content per tiller, respectively, compared with ‘Takanari.’

Figure 2. Effects of three different P-application treatments on shoot biomass per tiller (a) and P-content per tiller (b) in 2022 among”Takanari,” NIL-MP3, and NIL-fc1. *** indicates significant effects and interactions between P-application method (P) and genotype (G) at the 0.1% level. Different letters represent significant differences among genotypes across P-application treatments at the 5% level based on Tukey’s honestly significant difference test. Error bars represent standard deviation.

Discussion

In this study, we demonstrated the efficacy of P-dipping as a superior P-treatment for promoting tillering with significantly lower amounts of applied P compared with conventional P-incorporation. This result aligns with previous studies that attribute the positive effect of P-dipping on initial growth to the abundant P-supply from soluble P-hotspots near the root system (Oo, Tsujimoto, Rakotoarisoa, Kawamura, et al., 2020 and Oo, Tsujimoto, Mukai, Nishigaki, et al., 2021). Furthermore, our findings highlight that the fc1 allele is most effective in promoting the highest number of tillers, and when combined with P-dipping, it synergistically enhances shoot biomass and P-uptake, with the following order observed in 2022: ‘Takanari’ < NIL-MP3 < NIL-fc1. However, it is worth noting that shoot biomass did not increase in the P-dipping treatment in 2021, despite the increase in tiller number. This discrepancy might be attributed to the sampling timing, with 25 days after transplanting in 2021 likely representing the stage of tillering promotion before substantial tiller growth. As plants continued to grow and tillers enlarged with adequate nutrient availability, the continuous P-supply in the P-dipping treatment supported vigorous tillering and subsequent tiller growth in NIL-fc1 and NIL-MP3, leading to the genotypic difference observed in shoot biomass and P-uptake in 2022.

Although NIL-fc1 exhibited the highest number of tillers, the tillers were the smallest and contained the lowest amount of P. Takai et al. (2023) observed the presence of thin tillers and lodging in NIL-fc1 during grain filling under nutrient-rich conditions. Considering our focus on nutrient-poor environments, such as P-deficient soils in Madagascar, where the average rice yield is 2.7 t ha⁻¹ (FAOSTAT, 2021), it is important to consider the potential lodging issues of NIL-fc1 even under P-deficient conditions. In such scenarios, the combination of NIL-MP3 with P-dipping may be a more favorable approach for increasing yields under P-deficiency, as NIL-MP3 maintained approximately 70% of the biomass per tiller in ‘Takanari.’ Further field-based studies are required to investigate this matter and evaluate the combination of genotypes and P-dipping until maturity.

In conclusion, the combination of P-dipping and fc1 or MP3 alleles effectively promoted tillering and initial plant growth, likely through enhanced P-uptake. These findings have the potential to contribute to an increase in rice productivity in P-deficient paddy fields in sub-Saharan Africa.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the Japan International Research Center for Agricultural Sciences.