Estimation of Available Phosphorus Under Phosphorus Fertilization in Paddy Fields of a Cold Region Using Several Extraction Methods: A Case Study from Yamagata, Japan

Abstract

Assessing available phosphorus (P) in paddy fields is challenging due to waterlogging-induced reducing conditions. This study tested the applicability of the Truog, Bray 2, and Mehlich 3 extraction methods in both air-dried and incubated soils, as well as the ascorbic-acid-reduced Bray 2 (AR Bray 2), which simulates reducing conditions, for evaluating rice growth under P fertilization. In addition, to investigate the chemical characteristics of the extraction methods, active Al and Fe and P sequential extractions were measured. Soil samples from four representative regions in Yamagata Prefecture were used. Pot cultivation tests using ‘Haenuki’ and ‘Tsuyahime’ cultivars were conducted with varying P fertilizer levels. Variations in P availability across soil types were influenced by levels of active Al and Fe. Sequential extractions identified NaHCO₃-P and NaOH-P fractions as important for P availability. Bray 2 in both soils and AR Bray 2 were the most effective methods, showing a strong saturating exponential correlation with rice growth and P uptake, whereas Mehlich 3 and Truog showed weaker correlations. Bray 2 and AR Bray 2 show potential but require further evaluation for practical application due to the small number of soils. Future efforts should prioritize developing methods that account for P dynamics under reducing conditions, thereby improving P management strategies and supporting sustainable rice production.

1. Introduction

In cold-climate regions, declining temperatures can inhibit rice growth, making early development critical for ensuring stable yields. Phosphorus (P) plays a vital role in early growth, but low temperatures often impair P uptake, resulting in delayed development [1]. The importance of efficient P fertilization in cold regions lies in elevating extractable P to critical thresholds and improving grain yield [2]. Consequently, large quantities of P fertilizer are frequently applied in northern Japan such as the Tohoku region. Rice fields account for approximately 30% of the total P fertilizer consumption in Japan [3,4]; however, excessive P accumulation has become a growing concern. A nationwide survey revealed that 80% of rice fields exceeded the diagnostic threshold for plant-available P using the Truog method in air-dried soils (44 mg P kg⁻¹) [5]. Similarly, in Hokkaido, Bray-2-extractable P in field-moist soils exceeded recommended values in most fields [6]. P accumulation is also suspected in Yamagata Prefecture [7]. To reduce fertilizer usage while maintaining yield, it is essential to estimate the amount of available P that rice can actually utilize.

Several extraction methods have been developed to estimate available P, each targeting different chemical forms. The Truog method, widely used in Japan, extracts P bound to calcium and magnesium under acidic conditions [8]. Bray 2 extracts P associated with calcium, aluminum, and partially iron [9], while Mehlich 3, commonly used in North America and Europe, targets Ca-, Al-, and Fe-bound P using a mixture of acids and chelating agents [10,11,12]. These methods are not directly comparable due to differences in the extracted P fractions. The absence of standardized diagnostic approaches makes it difficult to establish clear fertilization guidelines.

Assessing available P in paddy soils is particularly challenging due to the reducing conditions caused by waterlogging. In flooded soils, P exists primarily as complexes or precipitates with Ca-, Fe-, and Al-bearing minerals. The biogeochemical behavior of P is largely influenced by changes in redox potential and pH [13,14]. Methods simulating reduced conditions, such as prolonged submergence incubation or chemical reduction using ascorbic acid, have been developed to better estimate Fe-P solubility [15,16]. However, the extent to which these methods correlate with rice growth remains poorly understood.

Iron redox cycling plays a key role in P solubilization and uptake under flooded conditions [17]. Sequential extraction methods, such as the Hedley fractionation, are widely used to identify different P pools, including both inorganic and organic forms [18]. Among these, NaHCO₃-extractable P—typically representing around 10% of total P—and NaOH-extractable P, which is associated with Al and Fe oxides, are considered important indicators of available P in submerged rice systems [19,20,21,22]. These fractions are particularly relevant under reduced conditions where Fe-P becomes more soluble.

This study aims to identify a more reliable method for evaluating available P in paddy fields of a cold region. We hypothesize that P solubility and availability are closely related to specific P fractions extracted through sequential fractionation and simulated reduction techniques. Therefore, we assessed the effectiveness of extraction methods including Truog, Bray 2, and Mehlich 3, as well as their relationships with rice growth parameters, particularly in cold-climate soils from Yamagata, northern Japan.

2. Materials and Methods

2.1. Soil Sampling and Analysis

Soil samples were collected from four major rice-producing regions in Yamagata Prefecture, Japan: Shonai (Field Science Center, Yamagata University; 38°42′ N, 139°49′ E), Mogami (Shinjo City; 38°50′ N, 140°19′ E), Murayama (Tendo City; 38°20′ N, 140°21′ E), and Okitama (Yonezawa City; 37°56′ N, 140°06′ E). According to climatic data from the Japan Meteorological Agency, the average annual temperatures at these sites are 12.9 °C, 11.0 °C, 11.4 °C, and 11.4 °C, respectively, with corresponding annual precipitation levels of 2191, 2006, 1088, and 1445 mm [23]. Soil sampling was conducted in tilled paddy fields during March and April 2022. Surface soils (0–15 cm) were collected from five points arranged along diagonal transects in each field, including the center. All fields were under conventional management, following standard fertilization and irrigation practices. Based on the Japanese soil classification system, the sampled soils were classified as Gray Lowland (Shonai), Wet Andosol (Mogami), Gley Lowland (Murayama), and Brown Forest (Okitama) [24]. According to the World Reference Base for Soil Resources, these correspond to Fluvisols for Shonai and Luvisols for the other three regions.

All samples were air-dried and passed through a 2 mm sieve before analysis. Standard physical and chemical properties were determined as follows (

Table 1. Physical and chemical properties of soils collected from four regions in Yamagata Prefecture.

Location	pH	EC	Total		CEC	Exchangeable Cation			Available N	Particle Size Distribution
			C	N		Ca	Mg	K		Clay	Silt	Sand
		(μS cm−1)	(%)		(cmolc kg−1)	(cmolc kg−1)			(mg kg−1)	(%)
Shonai	5.9	126.4	2.0	0.1	12.3	7.4	1.6	0.6	50.3	7.6	18.8	73.6
Mogami	5.8	35.2	7.2	0.5	27.7	5.9	1.2	0.6	79.6	26.0	48.5	25.6
Murayama	5.6	28.0	4.2	0.3	24.5	8.7	2.2	0.9	95.3	26.1	46.1	27.7
Okitama	5.7	25.5	3.0	0.3	15.0	9.1	1.9	0.4	77.5	13.0	29.9	57.1

). Soil pH (in a 1:2.5 soil-to-water suspension) and electrical conductivity (EC; in a 1:5 suspension) were measured using glass electrodes(D-210PC-S, HORIBA, Ltd., Kyoto, Japan). Cation exchange capacity (CEC) was determined using the ammonium acetate method (1 M, pH 7.0), followed by displacement with potassium chloride [25]. Exchangeable Ca, Mg, and K were extracted with 1 M ammonium acetate and quantified using atomic absorption spectrophotometry (AAS; Z-2300, Hitachi, Ltd., Tokyo, Japan). Available nitrogen (N) was estimated by incubating soil samples at 30 °C for four weeks, extracting with 2 M KCl, and analyzing the extract colorimetrically [26].

To investigate the relationship with the phosphorus extraction method, active aluminum (Al) and iron (Fe) were extracted using three methods: acid ammonium oxalate, sodium pyrophosphate, and dithionite–citrate–bicarbonate (DCB) [27,28]. Extracted Al and Fe concentrations were measured by AAS. Soil texture was analyzed using the pipette method [28].

2.2. Soil Analysis for Comparison of Available Phosphorus Analysis Methods

Available P was analyzed using four extraction methods: Truog, Bray 2, Mehlich 3, and ascorbic-acid-reduced Bray 2 (AR Bray 2). Previous studies suggested that AR Bray 2 had the potential to better estimate available P in flooded conditions [15,16]. For the Truog method, 0.001 M H₂SO₄ was added to air-dried soil at a 1:200 soil-to-extractant ratio, allowed to infiltrate for 30 min, and then filtered [8]. For Bray 2, 0.1 M HCl and 0.03 M NH₄F were added at a 1:20 ratio, followed by 1 min of shaking and filtration [25]. Mehlich 3 extraction involved the addition of a mixed solution containing 0.2 M CH₃COOH, 0.25 M NH₄NO₃, 0.015 M NH₄F, 0.013 M HNO₃, and 0.001 M EDTA at a 1:10 ratio, shaken for 5 min, and then filtered [10]. For AR Bray 2, 10 mL of 0.057 M ascorbic acid was added to 1 g of air-dried soil and shaken for 16 h. Subsequently, 10 mL of Bray 2 extractant at double strength was added, shaken for 1 min, and filtered [16]. All filtrates were analyzed using the molybdenum blue method with a spectrophotometer (UV-2550, SHIMADZU CORPORATION, Kyoto, Japan).

Soil incubation to simulate reduced soil conditions was conducted following the vacuum pouch incubation method [29]. Approximately 10 g of air-dried soil was placed in an aluminum-vapor-deposited pouch (LAMIZIP AL-11, 202 mm × 110 mm, SEISANNIPPONSHA Ltd., Tokyo, Japan) with 100 mL of deionized water. The pouch was vacuum-sealed and incubated at 30 °C for four weeks. After incubation, available P was determined using the Truog, Bray 2, and Mehlich 3 methods.

Total P was extracted from 5 g of air-dried, finely ground soil by sequential acid digestion using 1 mL of concentrated H₂SO₄, 5 mL of HNO₃, and 20 mL of HClO₄. After digestion, 30 mL of 1 M HCl and 50 mL of hot water were added, and the mixture was heated to just below boiling. The supernatant was filtered using quantitative filter paper (ADVANTEC No. 6, ADVANTEC TOYO KAISHA, Ltd., Tokyo, Japan), and P concentrations were determined colorimetrically [25].

P fractions were characterized using a modified Hedley sequential extraction method [18]. Soil was sequentially extracted with distilled water (1:30 w/v), 0.5 M NaHCO₃ (pH 8.5), 0.1 M NaOH, and 1 M HCl, with each extraction performed under continuous shaking for 16 h at 25 °C. The supernatant P concentrations were measured colorimetrically. Water-extractable P (H₂O-P) represents the most readily available fraction, while NaHCO₃-extractable P (NaHCO₃-P) is moderately available. NaOH- and HCl-extractable P represent more strongly bound forms with lower solubility [30]. The sum of these four fractions was considered the total inorganic P, and the difference between total and inorganic P was defined as residual P (Resid-P). All analyses were conducted three times for each of the four soils and we thus indicated the standard errors in the figures and tables.

2.3. Pot Experiment to Assess the Effects of Accumulated Phosphorus

A pot experiment was conducted in the experimental field of the Faculty of Agriculture, Yamagata University, using soils collected from four regions (Shonai, Mogami, Murayama, and Okitama). Air-dried soil (2.5 kg) from each site was placed into Wagner pots (Wagner’s pot 1/5000a, FUJIWARA SCIENTIFIC CO., Ltd., Tokyo, Japan). Two rice cultivars were used: ‘Haenuki’, a widely grown variety in Yamagata Prefecture, and ‘Tsuyahime’, a high-quality variety that has expanded in cultivation area in recent years and is subject to reduced- or no-fertilizer cultivation under the Yamagata Prefecture certification system [31]. Fertilization treatments were designed in accordance with regional guidelines [32]. Based on these guidelines, the recommended fertilization rates differ between the two rice cultivars. The recommended rates for Haenuki and Tsuyahime are 60:60:60 and 30:30:30 kg ha⁻¹ for N:P:K, respectively. Three P application rates were tested across all soils: P0 (no P), P1/2 (half of the standard rate), and P1 (the standard rate). The standard rate, P1, for each variety was aligned with the recommended fertilizer rate. For Haenuki, the P0, P1, and P2 treatments were set at 0, 60, and 120 mg P₂O₅ pot⁻¹, respectively, while for Tsuyahime, the rates were set at 0, 30, and 60 mg P₂O₅ pot⁻¹. To better evaluate crop response in low-P conditions, two additional fertilizer rates were applied exclusively to the Shonai soil: P4/3 (160 mg P₂O₅ pot⁻¹) and P2 (240 mg P₂O₅ pot⁻¹) for Haenuki and P4/3 (80 mg P₂O₅ pot⁻¹) and P2 (120 mg P₂O₅ pot⁻¹) for Tsuyahime. Nitrogen and potassium were applied at fixed rates: 120 mg N and 120 mg K₂O pot⁻¹ for Haenuki and 60 mg N and 60 mg K₂O pot⁻¹ for Tsuyahime. Ammonium sulfate, superphosphate, and potassium chloride were used as sources of nitrogen, P, and potassium, respectively. Fertilizers were mixed into the air-dried soil, which was then fully saturated with water and homogenized. Each treatment was replicated three times.

Seedlings were transplanted on 20 May 2022, and the experiment was continued until the heading stage on 10 August 2022. The average temperature during the growing period was 25.1 °C, with a total sunshine duration of 196.4 h. Pots were kept under continuously flooded conditions, and weed and pest control were managed according to standard practices. Because P uptake intensifies after root establishment and peaks during the heading and flowering stages, sampling was conducted at heading to evaluate the effect of P status on total P uptake by the whole plant. At this stage, aboveground biomass was harvested, oven-dried at 75 °C for 48 h, and weighed to determine shoot dry weight (SDW). To assess P uptake, approximately 500 mg of dried, finely ground plant tissue was digested. One milliliter of deionized water and 4 mL of concentrated sulfuric acid were added to a Pyrex test tube containing the sample, which was mixed using a touch mixer (VMX-3000V, AS ONE CORPORATION, Osaka, Japan). The tube was then heated at 250 °C in an aluminum block for 30 min. Hydrogen peroxide was added until the digest became clear, and the final volume was adjusted to 100 mL [33]. The P concentration in the digest was determined using the molybdenum blue method with a spectrophotometer.

2.4. Statistical Analysis

Analysis of variance (ANOVA) was used to evaluate the effects of P fertilization on SDW, P concentration, and P uptake (PU). Before ANOVA, normality and homogeneity of variance were tested using Shapiro–Wilk and Levene tests, respectively. Relationships between available P and SDW or PU were modeled using exponential saturation functions:

$$\mathrm{S}\mathrm{D}\mathrm{W}=\mathrm{a}\times \left[1-\exp \left(\mathrm{b} \times \mathrm{Ext}-\mathrm{P}\right)\right]$$

(1)

$$\mathrm{P}\mathrm{U}=\mathrm{a}\times \left[1-\exp \left(\mathrm{b} \times \mathrm{Ext}-\mathrm{P}\right)\right]$$

(2)

where a represents the maximum value of SDW or P uptake and b determines the rate at which SDW or P uptake responds to increasing Ext-P. In both models, Ext-P represents the available P concentration determined from air-dried soils prior to P fertilization, as shown in Figure 1. Pearson’s correlation coefficients were calculated to assess the relationships among P extraction methods, sequential extraction fractions, and soil Fe and Al. Statistical analyses were performed using R software version 3.3.3 [34].

3. Results

3.1. Comparison of Available Phosphorus Analysis Methods

The results of available P extractions from both air-dried and incubated soils are presented in Figure 1. Across all extraction methods and soil types, incubated soils consistently yielded higher P concentrations than air-dried soils. Truog-extractable P showed particularly large increases in Murayama and Shonai soils under incubation, with 13.2-fold and 5.3-fold rises, respectively. In contrast, increases in Mogami and Okitama soils were more moderate, approximately twofold. For Bray 2, increases were comparatively smaller, with 3.5-fold in Murayama, 1.8-fold in Shonai and Mogami, and 1.2-fold in Okitama. Mehlich 3 showed the most substantial change in Murayama, exhibiting a 20.8-fold increase under reduced conditions. The AR Bray 2 method, which incorporates ascorbic acid as a reducing agent, produced results that closely aligned with those of incubated Bray 2, supporting its validity as an alternative for estimating available P under reduced soil conditions in paddy fields. As shown in Figure 1, the available P concentrations prior to fertilization were used to assess the relationship between initial soil P status and plant response variables such as dry matter production, P concentration, and P uptake.

Figure 2 summarizes the results of the sequential P extraction. Total P concentrations varied considerably among soils, ranging from 677 mg P kg⁻¹ in Shonai, the lowest, to between 1668 and 2359 mg P kg⁻¹ in the other three soils. The proportion of inorganic P (H₂O-P, NaHCO₃-P, NaOH-P, and HCl-P) accounted for 51.2% to 61.5% of total P. H₂O-P was negligible in all samples (<0.2%), while NaHCO₃-P consistently comprised 8.8–11.8% of total P. NaOH-P was most abundant in Murayama (29.2%) and least in Mogami (15.6%). In contrast, HCl-P was lowest in Murayama (13.1%), indicating variation in the distribution of P fractions across soil types.

Active aluminum and iron contents (

Table 2. Contents of active aluminum (Al_o, Al_p, Al_d) and active iron (Fe_o, Fe_p, Fe_d) in soils from four regions in Yamagata Prefecture.

Location	Active Al									Active Fe
	Alo			Alp			Ald			Feo			Fep			Fed
	(g kg−1)									(g kg−1)
Shonai	2.62	±	0.06	0.93	±	0.02	2.56	±	0.02	7.07	±	0.09	2.88	±	0.01	13.38	±	0.07
Mogami	17.99	±	0.19	16.15	±	0.21	14.18	±	0.07	11.04	±	0.03	9.74	±	0.03	19.45	±	0.05
Murayama	3.58	±	0.05	1.81	±	0.26	3.24	±	0.09	11.13	±	0.03	8.04	±	0.17	17.99	±	0.19
Okitama	2.45	±	0.02	1.46	±	0.03	2.68	±	0.04	3.19	±	0.02	2.43	±	0.01	9.35	±	0.18

Al_o: oxalate-extractable Al; Al_p: pyrophosphate-extractable Al; Al_d: dithionite–citrate-extractable Al; Fe_o: oxalate-extractable Fe; Fe_p: pyrophosphate-extractable Fe; Fe_d: dithionite–citrate-extractable Fe. Values are means ± standard error (n = 3).

) showed distinct regional differences. Mogami soil exhibited the highest concentrations of oxalate-extractable, pyrophosphate-extractable, and dithionite–citrate aluminum (Al_o: 18.0 g kg⁻¹, Al_p: 16.1 g kg⁻¹, Al_d: 14.2 g kg⁻¹) and iron (Fe_o: 11.0 g kg⁻¹, Fe_p: 9.7 g kg⁻¹, Fe_d: 19.5 g kg⁻¹). Murayama soil also showed high levels of active Fe (Fe_o: 11.1 g kg⁻¹), whereas Shonai and Okitama had comparatively lower values across all fractions (Fe_o: 3.2–7.1 g kg⁻¹; Al_o: 2.5–2.6 g kg⁻¹).

Correlation analysis (Figure 3) showed strong relationships among extraction methods in air-dried soils. DS Truog and DS Mehlich 3 were nearly perfectly correlated (r = 0.99), and DS Bray 2 also showed a strong correlation with DS Truog (r = 0.94). In contrast, negative correlations were observed between Fe-related parameters (Fe_o, Fe_p, Fe_d) and extractable p values, particularly for DS Truog and DS Mehlich 3, suggesting reduced extractability in Fe-rich soils under oxidized conditions. In incubated soils, correlations among extraction methods remained high but were slightly attenuated, likely due to redox-induced shifts in P availability. IS Truog and IS Mehlich 3 were also strongly correlated (r = 0.99), while IS Bray 2 showed a weaker relationship with IS Truog (r = 0.81). Notably, IS Bray 2 was highly correlated with NaOH-extractable P (r = 0.98), which estimates moderately bound Fe- and Al-associated P.

3.2. Effect of Phosphorus Fertilization on Rice Growth and Development

As summarized in

Table 3. Shoot dry weight (SDW), phosphorus concentration (P concentration), and P uptake (PU) of Haenuki and Tsuyahime at the heading stage.

Variety	Location	Treatment	SDW			P Concentration			PU
			(g pot−1)			(%)			(mg pot−1)
Haenuki	Shonai	P0	40.4	±	2.7	0.20	±	0.004	80.5	±	2.7
		P1/2	43.2	±	1.9	0.21	±	0.010	92.1	±	1.9
		P1	44.1	±	1.4	0.20	±	0.005	88.5	±	1.4
		P4/3	44.1	±	1.1	0.21	±	0.004	91.7	±	1.1
		P2	42.7	±	0.1	0.21	±	0.011	90.2	±	0.1
	Mogami	P0	58.2	±	2.9	0.24	±	0.011	141.3	±	2.9
		P1/2	58.3	±	1.5	0.25	±	0.008	146.7	±	1.5
		P1	58.8	±	2.2	0.22	±	0.007	126.2	±	2.2
	Murayama	P0	62.9	±	0.7	0.22	±	0.004	139.4	±	0.7
		P1/2	64.9	±	1.5	0.23	±	0.007	149.7	±	1.5
		P1	63.9	±	1.2	0.21	±	0.003	137.7	±	1.2
	Okitama	P0	50.6	±	2.5	0.34	±	0.020	175.2	±	2.5
		P1/2	50.7	±	1.2	0.34	±	0.003	174.1	±	1.2
		P1	53.0	±	0.6	0.31	±	0.020	162.3	±	0.6
Tsuyahime	Shonai	P0	38.0	±	1.2	0.26	±	0.011	97.0	±	2.0
		P1/2	39.2	±	0.6	0.25	±	0.012	98.2	±	4.3
		P1	36.7	±	1.5	0.23	±	0.003	85.9	±	4.4
		P4/3	38.5	±	1.0	0.24	±	0.008	93.0	±	5.4
		P2	39.0	±	0.4	0.25	±	0.003	96.0	±	2.1
	Mogami	P0	55.8	±	1.6	0.21	±	0.002	119.5	±	4.5
		P1/2	59.8	±	3.5	0.24	±	0.018	141.0	±	1.9
		P1	59.2	±	1.1	0.23	±	0.003	133.7	±	4.2
	Murayama	P0	58.2	±	1.4	0.24	±	0.003	137.8	±	4.7
		P1/2	59.0	±	2.0	0.25	±	0.010	145.9	±	1.9
		P1	57.3	±	1.8	0.23	±	0.008	134.4	±	7.8
	Okitama	P0	52.2	±	2.0	0.30	±	0.006	155.3	±	3.0
		P1/2	46.7	±	1.6	0.32	±	0.005	149.7	±	5.7
		P1	50.6	±	1.2	0.33	±	0.016	165.6	±	4.7
Variety			***			***			ns
Soil			***			***			***
Treatment			ns			*			ns
Variety × Soil			**			***			**
Variety × Treatment			ns			ns			ns
Soil × Treatment			ns			ns			ns
Variety × Soil × Treatment			ns			ns			ns

Values are presented as means ± standard errors (n = 3). No significant effects of P fertilization were observed on SDW, P concentration, or PU. ns: not significant; * p < 0.05; ** p < 0.01; *** p < 0.001.

, SDW, P concentration, and PU at the heading stage were unaffected by P fertilization in all soils, indicating that available P levels were almost sufficient for rice growth. Among the four soil types, Shonai soil consistently produced the lowest SDW: 42.9 g pot⁻¹ for Haenuki and 38.3 g pot⁻¹ for Tsuyahime, representing reductions of approximately 20–52% compared to the other soils. PU was strongly proportional to SDW, further confirming the lack of a clear P response under sufficient P availability conditions. Furthermore, no statistically significant interactions were detected with P application (Treatment), indicating that neither rice variety nor soil type significantly altered the overall response pattern to P fertilization. Although P concentration was statistically affected by P treatment (p < 0.05), the differences were relatively small and did not translate into notable changes in plant growth or P uptake.

Exponential saturation models were used to assess the relationship between available P and both SDW and PU (Figure 4 and Figure 5), using soil P concentrations measured before P fertilization (as shown in Figure 1). No marked differences were observed between the two rice varieties; however, Tsuyahime appeared to reach a growth plateau at lower P concentrations, suggesting more efficient P utilization.

In DS Truog, IS Truog, DS Mehlich 3, and IS Mehlich 3, model fits were poor, with low R² values making curve estimation infeasible for SDW. In air-dried soils, only DS Bray 2 produced model fits with marginal usability. Similar limitations were observed for PU, particularly in DS Mehlich 3, IS Mehlich 3, and IS Truog. Both IS Bray 2 and AR Bray 2 exhibited high R² values in their relationships with SDW and PU, outperforming the other methods. These results indicate that IS Bray 2 and AR Bray 2 are better suited for evaluating plant-available P under reduced soil conditions.

4. Discussion

4.1. Correlation of Extraction Methods for Available P

All three extraction methods (Truog, Bray 2, and Mehlich 3) showed significantly higher available P concentrations in incubated soils compared to air-dried soils (Figure 1). This increase was particularly pronounced for the Truog and Mehlich 3 methods. The Truog method primarily extracts Ca-bound P, while Mehlich 3 is known to extract both Ca-P and Al-P [11,12,35]. Bray 2 targets Ca-P, Al-P, and partially Fe-P [9]. Under reducing conditions, Fe-P is solubilized, whereas Ca- and Al-bound P tends to remain insoluble [36,37]. Therefore, the marked increase in Truog- and Mehlich-3-extractable P under incubation suggests that these methods also accessed solubilized Fe-P under reducing conditions. In contrast, Bray 2 may have been less affected by redox changes due to its inherent ability to extract some Fe-P even in air-dried soils.

The AR Bray 2 method is a simplified technique to simulate reduced soil conditions using ascorbic acid [36]. AR Bray 2 has demonstrated strong agreement with long-term incubation results [38]. In our study, AR Bray 2 was highly correlated with IS Bray 2 (r = 0.92; Figure 3), confirming its potential for estimating P availability under reduced conditions using air-dried samples. Previous studies also suggest that available P measured in moist soils correlates more strongly with plant uptake than P extracted from air-dried samples [39].

Sequential P fractionation revealed clear differences among soils (Figure 2). H₂O-P accounted for less than 0.2% of total P, consistent with previous findings [12,40,41]. NaHCO₃-P represented approximately 10% of total P across all soils, aligning with earlier reports [19,20,21]. Notably, NaOH-P, associated with Fe oxides, was highest in Murayama soil (30%) and lowest in Mogami soil (16%). The strong correlation between NaOH-P and IS Bray 2 (r = 0.98) indicates that IS Bray 2 effectively captures moderately labile Fe-P fractions under reducing conditions, indicating greater sensitivity to P fractions that are mobilized under such conditions. NaOH-P has previously been identified as a key source of plant-available P in paddy soils [42]. A single extraction with alkaline reagents such as NaHCO₃ or NaOH can thus serve as a useful indicator of P availability in submerged rice fields [43].

The solubility of P in flooded soils is strongly influenced by the presence of active Al and Fe [44,45]. Although Mogami soil contained high total P, available P was low, and the extraction efficiency of Truog and Mehlich 3 was poor (Table 1; Figure 1), suggesting that the presence of active Al may have suppressed P solubility. Under reducing conditions, Al- and Fe-associated P becomes more available, but this availability is also influenced by Fe crystallinity. A moderate positive correlation was also observed between IS Bray 2 and Fe_p (r = 0.48), suggesting possible interactions with organic Fe complexes (Figure 3). The ratio (Al_o + Fe_o)/2 is known to be related to soil P retention capacity, and ammonium oxalate-extractable Al and Fe are considered major determinants of P forms in paddy soils [37,46,47]. Organic matter can form stable complexes with Fe and Al, inhibiting their crystallization and enhancing P sorption due to increased non-crystalline Al and Fe levels [48,49]. The high levels of active Al in Mogami soil may have promoted the formation of such amorphous Al complexes, limiting P desorption. The relative contributions of Al and Fe to P retention vary with soil type—Fe tends to dominate in Gley and Gray Lowland soils, while Al is often more influential in Andosols [36]. These differences in mineralogy significantly affect P extractability and availability [30,47,48]. Therefore, assessing P availability in paddy soils requires consideration of soil-specific properties, particularly active Al and Fe levels, and their interactions with extraction methods. Our findings highlight the utility of NaHCO₃-P and NaOH-P as indicators of available P in Fe- and Al-rich paddy soils under reduced conditions.

4.2. Effects of P Treatment on Rice Growth

P fertilization had no significant effect on SDW, P concentration, or PU in either Haenuki or Tsuyahime (Table 3). P application is typically associated with increased biomass production, particularly under P-deficient conditions. However, when soil P is sufficient, the effects of additional P are negligible. Previous studies have shown that nitrogen has a stronger influence on rice phenology than P or K and that rice generally has a low sensitivity to P once a threshold level is reached [50].

In our study, all four soils appeared to contain sufficient P, and no consistent increase in tissue P concentration was observed with increasing P application. This plateau effect is consistent with a proposed feedback mechanism by which rice maintains stable internal P concentrations once they reach levels sufficient for physiological functioning [51]. When available P exceeds the crop requirement, further P input does not translate into increased growth or uptake [52]. Among the four soils, Shonai exhibited lower SDW compared to the others, although it is unclear whether this was due to P availability or other soil factors. Regression analysis of available P versus SDW and PU showed that Tsuyahime achieved maximum PU at lower P concentrations than Haenuki, suggesting greater P uptake efficiency.

Previous studies have demonstrated that SDW and PU are often related to available P through saturation-type functions [53,54,55,56]. In our analysis, IS Bray 2 and AR Bray 2 showed the strongest correlations with both SDW and PU (Figure 4 and Figure 5), underscoring the importance of evaluating P availability under reduced conditions. Reduced soils provide a more accurate reflection of P availability for rice compared to air-dried soils [15]. In particular, IS Bray 2, which targets moderately labile Fe- and Al-P, showed high predictive power for rice growth, consistent with previous findings [53]. Although incubated soil P analysis is effective, its four-week incubation period limits its practical application. As a simpler alternative, AR Bray 2 has shown strong correlations with both rice growth and with incubated P values in this study. However, further validation across diverse soil types is needed to establish AR Bray 2 as a reliable diagnostic method.

Truog and Mehlich 3 results showed considerable variability between air-dried and incubated soils. Truog and Mehlich 3 methods, although widely used, showed greater variability and lower reliability, especially under oxidized conditions in this study. These findings emphasize the importance of using extraction methods that reflect reduced conditions in paddy soils to better assess bioavailable P and its contribution to rice growth. Interestingly, Truog had higher R² values for PU in air-dried soils than in incubated soils (Figure 4 and Figure 5). Many Japanese studies have explored the relationship between Truog-P and rice performance. For example, Truog-P levels at the maximum tillering stage were correlated with plant P status [55]. Mehlich 3, widely used globally, has been reported as accurate and reproducible when standardized [57]. The distinct characteristics of P extracted by different methods further emphasize the need for method selection based on soil properties.

5. Conclusions

This study highlights the importance of considering reduced conditions and soil-specific factors when evaluating available P in paddy soils. We demonstrated that different extraction methods—including incubation, chemical extraction (Truog, Bray 2, Mehlich 3, AR Bray 2), and sequential fractionation—each provide distinct insights into the forms of available P. Among the P fractions analyzed, NaHCO₃-P and NaOH-P were particularly important for assessing P availability under reduced conditions, where Fe- and Al-bound P are more readily mobilized. The strong variability observed among the four soil types indicates that no single extraction method can universally assess P availability. Our findings underscore the necessity of a multifactorial approach that integrates both chemical extraction data and soil-specific properties such as active Al and Fe content. These insights can support the refinement of soil diagnostic tools and inform site-specific P management strategies in rice production systems. However, further research is needed to validate these findings across a broader range of soil types and environmental conditions. By accounting for the complex interplay between redox conditions, mineralogy, and P dynamics, future studies can contribute to more efficient P use and promote sustainable rice cultivation in diverse agroecosystems.