Root Distribution, Agronomic Performance, and Phosphorus Utilization in Wheat as Mediated by Phosphorus Placement under Rainfed Coastal Saline Conditions

Abstract

Rainfall variations between seasons could affect phosphorus translocation from rainfed saline soil to wheat plants. Whether deep-banded P application increases wheat yield compared to traditional P placement under rainfed coastal saline conditions remains a question. This study investigated the impact of season, P placement, and genotype on root distribution, agronomic performance, and P utilization in wheat grown under rainfed coastal saline conditions. Four wheat genotypes (two tall genotypes (Alice and Shavano) and two dwarf genotypes (AK58 and LX99)) were grown in a saline field with five P placement treatments (Top-dressed High P input (TopHP), Deep-banded High P input (DeepHP), Top-dressed Reduced P input (TopRP), Deep-banded Reduced P input (DeepRP), and no P supply (No P)) for two consecutive seasons. Root length density (RLD), agronomic traits, nutrient concentrations in grain and straw, and P utilization efficiency were determined. Statistical analysis was employed to compare the P utilization across treatments. TopHP increased RLD at a 0–20 cm depth, while deep-banded P increased RLD at a 20–40 cm depth. The wet season (2021–2022) resulted in higher grain yields, more fertile spikelets, and fewer non-fertile spikelets in all four genotypes than the dry season (2020–2021). The highest 1000-kernel weights occurred in DeepHP or TopHP. Deep-banded P outperformed top-dressed P placement in terms of P utilization efficiency for LX99, Shavano, and AK58 (not Alice) in both seasons. Nutrient concentrations/accumulations showed inconsistent patterns due to significant genotype × P placement interactions. PCA analysis revealed that first two PCs accounted for 56.19% and 60.13% of the variance in the 2020–2021 and 2021–2022 seasons, respectively. The first component (PC1) represented root spatial distribution and straw weight, while the second component (PC2) represented 1000-kernel weight, grain number per head, and grain yield. Altered P utilization efficiency mediated by P placement was associated with changes in wheat root distribution, agronomic traits, and nutrient concentrations in straw and grain. The increased wheat yield in the wet season (2021–2022) was attributed to higher rainfall.

1. Introduction

Exploiting saline land for agriculture purposes has significant potential to feed the growing population. In China, for example, the total area of saline land is about 9.9 × 10⁷ ha, while that of coastal saline land is about 6.9 × 10⁵ ha. Farming on coastal saline land often relies on rainfall due to the scarcity of fresh water. Sustainable crop production in rainfed areas requires proper fertilizer management and suitable crop genotypes to optimize the use of rainfall during growth. Coastal saline fields, characterized by shallow high-salinity groundwater, inherently hinder crop growth due to sodium accumulation in plant tissues, restricted root growth, and limited soil nutrients and water availability [1,2,3]. Calcareous soils usually have low available phosphorus (P) due to the binding of orthophosphate ions with calcium carbonate. Large amounts of P are required in crop production to increase soluble inorganic phosphate (Pi). Proper P placement to match root growth with soil moisture dynamics and nutrient availability could enhance crop yields, given the pivotal role of roots in transporting water and nutrients from soils to aboveground plant parts.

The quantity and placement methods of P inputs can influence root growth, aboveground plant development, and P utilization efficiency [4,5,6]. A recent wheat meta-analysis revealed that banded P increased grain yield and P uptake in 76% and 87.5% of the studies, respectively, relative to broadcast P application [7]. Applying an appropriate amount of P fertilizer in deep bands at the right soil depth, especially in soils with low P availability, could enhance P utilization efficiency. Banded P often stimulates root growth as roots encounter P-rich patches, improving P use compared to conventionally broadcast P [6,8,9]. Crops with deep roots can access deeper soil water during drought conditions, while those with shallow roots can better use limited surface nutrients such as P [10]. Thus, P placement is crucial for deploying effective root ideotypes in complex soil environments [11].

Crop roots bridge the upper (low water potential) and deeper (high water potential) soil layers, foraging for water and nutrients [12]. Roots might exude water into dry soil at night due to water potential differences [13,14], facilitating nutrient acquisition [15,16]. Rainfall affects root growth and nutrient mobilization in the upper soil layer, which can quickly dry out between rain events [4]. The frequency of re-moistening of the soil surface influences P uptake by plants and hence shoot growth [17,18]. Consequently, variations in seasonal rainfall would affect crop access to applied P.

Plants under salt stress often exhibit increased root and shoot Na/K ratios and altered nutrient composition [1,19]. Varying P application amounts and placement depths can also affect plant uptake of essential elements. For instance, increased P nutrition can enhance Mg and Ca uptake [20,21], and applying P fertilizer monocalcium phosphate can promote root growth and total Ca accumulation in plants under salt stress [22].

The effect of deep-banded P on root spatial distribution, agronomic performance, and phosphorus utilization in wheat grown on coastal saline fields has not been fully explored. Thus, we conducted a field trial on a coastal saline field to test the following hypotheses: (1) P placement affects root distribution, agronomic performance, and P utilization efficiency in wheat; and (2) P utilization varies among wheat genotypes and across seasons.

2. Materials and Methods

2.1. Wheat Genotypes

Four wheat genotypes were used for this study: two modern dwarf wheat genotypes (AK58 (Bainong Aikang 58) and LX 99 (Liangxing 99)) and two tall genotypes (Alice and Shavano). These four varieties were selected with the assumption that tall genotypes with large root systems may have better performance under rainfed conditions relative to the dwarf genotypes with small root systems, as Alice and Shavano were found to have a relatively larger root system compared to that of AK58 and LX 99 in a preliminary study.

2.2. Experiment Set Up

Field trials were conducted during the 2020–2021 and 2021–2022 wheat growing seasons in a coastal saline field (37.94° N, 117.83° E) at the west line of Bohai Gulf within Wudi, Shandong Province, China. This experimental site experiences semi-drought conditions with pronounced seasonal fluctuations due to its temperate continental monsoon climate. The average annual rainfall in this area is 500–700 mm, and annual evaporation is 1800–2000 mm.

Table 1. Temperature and precipitation during the 2020–2021 and 2021–2022 wheat growing seasons.

2020–2021	Average High Temperature (°C)	Average Low Temperature (°C)	Precipitation (mm)	2021–2022	Average High Temperature (°C)	Average Low Temperature (°C)	Precipitation (mm)
Sep. 2020	28	17	0	Sep. 2021	27	18	178.3
Oct. 2020	21	8	0	Oct. 2021	19	8	89.4
Nov. 2020	14	2	62.9	Nov. 2021	14	1	78.2
Dec. 2020	4	−6	0	Dec. 2021	7	−4	4.2
Jan. 2021	4	−8	0	Jan. 2022	4	−6	0.8
Feb. 2021	11	−2	1.7	Feb. 2022	7	−5	1.7
Mar. 2021	16	4	0.9	Mar. 2022	16	3	43.6
Apr. 2021	19	8	20.5	Apr. 2022	23	9	1.8
May 2021	26	14	3.3	May 2022	28	14	6.7
Jun. 2021	32	20	28.9	Jun. 2022	32	20	61.9

presents the temperature and precipitation data for the 2020–2021 and 2021–2022 seasons. Initial soil characteristics included pH 7.9, 0.9 dS m⁻¹ EC, 1.0 g kg⁻¹ total N concentration, 6.3 mg kg⁻¹ available Olsen-P, and 110.6 mg kg⁻¹ exchangeable K.

The field trial encompassed the four wheat genotypes and five P placements (Top-Dressed High P (TopHP, 15 g P m⁻² placed on soil surface), Top-Dressed Reduced P (TopRP, 7.5 g P m⁻² placed on soil surface), Deep-Banded High P (DeepHP, 15 g P m⁻² placed at 20 cm soil depth), Deep-Banded Reduced P (DeepRP, 7.5 g P m⁻² placed at 20 cm soil depth), and No P added (–P)), using monocalcium phosphate (Ca(H₂PO₄)₂. Each treatment had three replications, resulting in 60 plots (4 genotypes, 5 P placement methods, and 3 replications). Each plot, measuring 6 m², received a specific P placement and equal amounts of N (20 g m⁻²) and K (12 g m⁻²) fertilizer. Seeds were manually sown in mid-October in both seasons, with 250 seeds m⁻² for all genotypes.

2.3. Agronomic Traits

Thirty plants were randomly selected from each plot at maturity to determine fertile spikelet number per head, non-fertile spikelet number per head, and grain number per head. Grain yield, straw weight, and 1000-kernel weight were determined after oven drying at 70 °C for 48 h.

2.4. Root Spatial Distribution

At maturity, roots were harvested separately from 0–20, 20–40, 40–60, 60–80, and 80–100 cm soil depths using a stainless root auger (15 cm diameter), soaked in deionized water overnight, and washed to remove soil particles adhering to the root surface. Root length was measured using the WinRHIZO Root Analysis System (Regent Instruments, Montreal, Canada). Root length density (RLD; cm cm⁻³) was calculated as root length (cm)/soil core volume (cm³).

2.5. Elemental Measurement

Plant samples were harvested, separated into straw and grain, oven dried, and ground into a fine powder. A 50 mg subsample was digested with 13 mL HNO₃ and 2 mL H₂O₂ to determine elemental concentrations using Agilent 5800 ICP-OES.

2.6. P Utilization Parameters

Three indexes were used to evaluate P efficiency: P utilization efficiency (PUE), P agronomic efficiency (PAE), and P physiology efficiency (PPE), calculated as follows:

PUE = (P uptake in treatment − P uptake in control)/(P input amount) × 100%

PAE (kg kg⁻¹) = (Grain yield in treatment − Grain yield in control)/(P input amount)

PPE (kg kg⁻¹) = (Grain yield in treatment − Grain yield in control)/(P uptake in treatment − P uptake in control)

2.7. Statistical Analysis

A season × P placement × genotype interaction model was determined via three-way ANOVA using the SPSS 16.0 package (IBM, New York, NY, USA). A one-way ANOVA was then conducted for each genotype to examine the effect of P placement within a specific season. Multiple comparisons (Duncan’s) of the measured parameters were performed among different P placements for the same genotype within the same season. A significance level (α) of 0.05 was set. Principal component analysis (PCA) was conducted using Origin 2018 software (Originlab, Northampton, MA, USA) to extract valuable information on the investigated traits, including root distribution parameters, agronomic traits, and nutrient concentrations. Pearson’s correlation analysis was conducted using the SPSS package to examine relationships among variables.

3. Results

3.1. Root Spatial Distribution under Different P Placements

Root length density (RLD) at maturity was used to compare root distributions among the P placement treatments. Season, genotype, and P placement significantly affected RLD at all five soil depths (Supplementary Tables S1–S5). Overall, all four genotypes had larger root systems in the 2021–2022 season than in the 2020–2021 season (Figure 1 and Figure 2). Significant differences in RLD were observed between tall and dwarf genotypes in the 2020–2021 and 2021–2022 seasons (Figure 1 and Figure 2). Tall genotypes (Alice and Shavano) had more significant root distributions than dwarf genotypes (AK58 and LX99); for instance, in the 2020–2021 season, the RLD at 0–20 cm depth for Alice (4.29–5.20 cm cm⁻³) and Shavano (4.32–4.93 cm cm⁻³) exceeded that of AK58 (3.07–3.51 cm cm⁻³) and LX99 (3.01–3.64 cm cm⁻³).

A separate ANOVA for each season was conducted to examine the effects of P placement, genotype, and their interactions. For all four genotypes, TopHP resulted in a higher RLD at a 0–20 cm depth, while deep-banded P resulted in a higher RLD at a 20–40 cm depth than top-dressed P (Figure 1 and Figure 2). P application increased RLD in deeper soil, particularly 60–80 cm and 80–100 cm, relative to No P.

3.2. Agronomic Traits

Season, genotype, and P placement significantly affected straw weight and grain yield (Supplementary Tables S6 and S7). DeepHP produced the highest grain yields for AK58, LX99, and Shavano, while TopHP produced the highest grain yield for Alice in both seasons (

Table 2. Agronomic traits at maturity of four wheat genotypes in the TopHP, TopRP, DeepHP, DeepRP, and No P treatments.

Season	Genotype	Treatment	Fertile Spikelet Number per Head	Non-Fertile Spikelet Number per Head	Grain Number per Head	1000-Kernel Weight (g)	Grain Yield (g m−2)	Straw Weight (g m−2)
2020–2021	AK58	TopHP	15.9 ± 0.4 bc	0.9 ± 0.1 b	42.6 ± 2.3 b	42.4 ± 0.7 a	382.3 ± 17.6 a	361.5 ± 16.7 a
		DeepHP	16.8 ± 0.9 a	0.3 ± 0.1 d	48.1 ± 1.1 a	43.3 ± 0.6 a	391.3 ± 17.0 a	364.3 ± 28.8 a
		TopRP	15.6 ± 0.1 c	0.3± 0.1 c	44.6 ± 1.9 ab	41.8 ± 0.6 a	358.5 ± 17.7 a	358.8 ± 22.4 a
		DeepRP	16.1 ± 0.1 b	1.0 ± 0.2 b	44.3 ± 0.4 b	41.9 ± 0.4 a	379.3 ± 25.8 a	360.7 ± 9.2 a
		No P	15.1 ± 0.1 c	1.4 ± 0.2 a	42.6 ± 1.5 c	41.5 ± 1.7 a	314.6 ± 12.6 b	343.8 ± 13.8 a
	LX99	TopHP	15.7 ± 0.3 a	1.4 ± 0.3 b	38.6 ± 1.2 b	41.8 ± 0.5 ab	438.7 ± 14.3 b	541.0 ± 14.6 a
		DeepHP	16.2 ± 0.2 ab	0.7 ± 0.1 c	41.6 ± 0.5 ab	42.0 ± 0.4 a	460.2 ± 5.5 a	533.8 ± 15.5 b
		TopRP	16.4 ± 0.2 a	1.7 ± 0.5 b	41.2 ± 0.8 ab	40.6 ± 0.7 bc	419.0 ± 3.8 c	525.0 ± 13.0 b
		DeepRP	16.3 ± 1.3 a	1.6 ± 0.2 b	44.1 ± 5.5 a	40.6 ± 0.9 bc	414.3 ± 8.6 c	508.8 ± 33.3 c
		No P	14.1 ± 1.1 b	3.0 ± 0.5 a	30.2 ± 2.3 c	40.1 ± 0.4 c	373.3 ± 9.0 d	470.8 ± 17.1 d
	Alice	TopHP	15.2 ± 0.5 b	1.7 ± 0.4 bc	36.1 ± 1.4 a	39.6 ± 0.6 a	537.2 ± 31.0 a	709.2 ± 33.0 a
		DeepHP	15.4 ± 0.2 a	1.3 ± 0.3 c	37.3 ± 0.9 a	39.2 ± 0.6 a	497.0 ± 15.5 b	669.2 ± 19.3 ab
		TopRP	15.9 ± 0.6 b	2.1 ± 0.3 b	36.7 ± 2.5 a	38.9 ± 0.4 a	463.2 ± 13.9 c	632.7 ± 25.5 b
		DeepRP	16.0 ± 0.6 a	1.6 ± 0.2 bc	38.1 ± 2.3 a	38.9 ± 1.1 a	441.1 ± 2.7 c	576.3 ± 23.1 c
		No P	13.8 ± 0.2 a	3.1 ± 0.4 a	33.4 ± 3.0 b	38.6 ± 0.6 a	402.5 ± 10.6 d	522.2 ± 22.1 d
	Shavano	TopHP	14.6 ± 0.4 a	1.3 ± 0.1 a	34.2 ± 1.2 a	31.0 ± 0.2 a	440.6 ± 29.5 a	593.1 ± 28.6 ab
		DeepHP	14.4 ± 1.3 a	1.5 ± 0.3 a	34.3 ± 0.5 a	31.2 ± 0.8 a	442.6 ± 13.5 a	615.2 ± 21.1 a
		TopRP	15.1 ± 0.7 a	2.8 ± 0.3 ab	34.0 ± 1.8 a	30.3 ± 0.4 ab	396.4 ± 23.7 b	546.2 ± 34.9 b
		DeepRP	14.1 ± 0.6 a	2.2 ± 0.4 b	34.1 ± 0.5 a	30.9 ± 0.2 ab	412.8 ± 11.9 ab	589.2 ± 9.0 ab
		No P	14.2 ± 0.7 a	3.1 ± 0.3 a	31.4 ± 1.0 a	30.1 ± 0.3 b	345.3 ± 17.0 c	537.8 ± 42.8 b
2021–2022	AK58	TopHP	16.2 ± 0.4 b	0.9 ± 0.1 b	43.4 ± 2.4 b	43.5 ± 0.5 ab	483.4 ± 30.2 a	406.7 ± 36.0 a
		DeepHP	17.3 ± 0.9 a	0.5 ± 0.1 c	49.5 ± 1.1 a	43.9 ± 0.3 a	487.2 ± 20.3 a	426.8 ± 7.1 a
		TopRP	15.6 ± 0.1 b	1.3 ± 0.2 c	45.5 ± 1.9 ab	43.1 ± 0.3 b	408.1 ± 19.7 b	406.6 ± 15.8 a
		DeepRP	16.4 ± 0.1 b	1.0 ± 0.2 b	43.2 ± 0.4 b	43.1 ± 0.3 b	428.7 ± 18.8 b	434.0 ± 16.9 a
		No P	15.4 ± 0.3 a	1.4 ± 0.3 a	42.6 ± 2.1 a	42.3 ± 0.6 ab	364.0 ± 15.9 c	387.3 ± 32.3 a
	LX99	TopHP	17.0 ± 0.3 a	0.5 ± 0.1 c	40.3 ± 1.3 a	43.3 ± 0.8 a	488.9 ± 7.8 b	608.9 ± 6.3 a
		DeepHP	16.3 ± 0.2 ab	0.5 ± 0.2 c	43.0 ± 0.5 a	43.5 ± 0.7 a	526.2 ± 23.3 a	603.0 ± 40.8 a
		TopRP	17.2 ± 0.2 a	1.1 ± 0.3 b	42.1 ± 0.9 a	41.7 ± 0.5 b	478.6 ± 12.5 bc	558.9 ± 32.5 ab
		DeepRP	17.2 ± 1.4 a	1.0 ± 0.2 b	44.9 ± 5.7 a	41.9 ± 0.7 b	456.6 ± 20.8 cd	522.8 ± 20.8 b
		No P	14.9 ± 1.2 b	2.3 ± 0.3 a	37.2 ± 2.3 b	41.5 ± 0.7 b	434.5 ± 8.7 d	517.2 ± 14.4 b
	Alice	TopHP	15.1 ± 0.6 b	0.7 ± 0.3 c	35.2 ± 3.0 ab	41.3 ± 0.8 a	630.6 ± 27.4 a	822.8 ± 33.5 a
		DeepHP	16.5 ± 0.2 a	0.6 ± 0.1 c	38.2 ± 1.0 a	41.1 ± 0.6 a	588.2 ± 19.1 b	751.3 ± 38.8 b
		TopRP	14.2 ± 0.6 b	1.2 ± 0.3 a	35.5 ± 2.5 b	39.6 ± 0.6 b	518.2 ± 7.8 c	731.9 ± 28.5 bc
		DeepRP	16.4 ± 0.7 a	1.7 ± 0.2 b	39.0 ± 2.2 a	39.3 ± 1.2 b	514.8 ± 15.4 c	676.5 ± 23.7 c
		No P	14.2 ± 0.5 a	1.4 ± 0.2 b	34.2 ± 3.1 b	39.2 ± 0.7 b	489.8 ± 17.5 c	590.8 ± 26.2 d
	Shavano	TopHP	14.9 ± 0.4 a	1.3 ± 0.1 a	34.9 ± 1.2 a	32.6 ± 0.4 a	533.1 ± 30.3 a	686.4 ± 36.9 b
		DeepHP	14.8 ± 1.4 a	1.2 ± 0.4 a	31.9 ± 6.2 a	32.3 ± 0.5 ab	534.3 ± 34.5 a	776.3 ± 51.7 a
		TopRP	15.4 ± 0.8 a	2.9 ± 0.7 ab	34.7 ± 1.9 a	31.8 ± 0.3 b	461.9 ± 6.5 b	634.9 ± 43.2 b
		DeepRP	14.4 ± 0.7 a	2.2 ± 0.4 b	34.6 ± 0.5 a	32.0 ± 0.3 ab	474.7 ± 25.3 b	716.1 ± 56.1 ab
		No P	14.5 ± 0.6 a	3.1 ± 0.3 a	32.0 ± 1.1 a	30.8 ± 0.4 c	415.6 ± 19.0 c	619.7 ± 9.5 b

Note: Different letters within the same column indicate significant differences at the 0.05 level.

). In the 2020–2021 and 2021–2020 seasons, DeepHP increased grain yield by 24.4% and 33.8% for AK58, 23.3% and 21.1% for LX99, and 28.2% and 28.6% for Shavano, while TopHP increased grain yield by 33.5% and 28.7% for Alice, respectively, compared to No P.

Significant genotypic differences occurred for agronomic traits, including fertile spikelet number per head, non-fertile spikelet number per head, grain number per head, and 1000-kernel weight (Supplementary Tables S8–S11). Moreover, fertile spikelet head per head, non-fertile spikelet number per head, and 1000-kernel weight differed significantly between seasons. Generally, the four genotypes produced more fertile spikelets and fewer non-fertile spikelets in the 2020–2021 season than in the 2021–2022 season (Table 2). The highest 1000-kernel weight for all four genotypes occurred in the DeepHP or TopHP treatment.

3.3. Nutrient Concentrations and Accumulation in Straw and Grain

ANOVA indicated that season, genotype, and P placement significantly affected straw Na, K, Ca, Mg, and P concentrations (Supplementary Tables S12–S16). Due to significant interactions between genotype and P placement, no consistent change pattern in straw nutrient concentrations occurred among genotypes in response to P placement (Supplementary Table S17). Similarly, season, genotype, and P placement significantly affected grain Na, K, Ca, Mg, and P concentrations (Supplementary Tables S18–S22). Grain Na concentrations were significantly lower, while grain P concentrations were significantly higher than their respective straw concentrations (Supplementary Table S23).

No consistent change pattern emerged from single nutrient concentrations in straw or grain in response to P placement (Supplementary Table S24). The genotype P placement interaction significantly affected the total nutrient accumulation in straw; for instance, TopHP produced the highest straw K accumulation for three genotypes (AK58, LX99, and Alice) but not Shavano. TopHP produced the highest total grain P accumulation in AK58, while DeepHP produced the highest in LX99, Alice, and Shavano (Supplementary Table S25). The P-applied treatments had higher total P accumulation in aboveground plant parts than No P (

Table 3. Total elemental accumulation in aboveground plants of four wheat genotypes at maturity in the TopHP, TopRP, DeepHP, DeepRP, and No P treatments.

Season	Genotype	Treatment	K (g)	Na (g)	Ca (g)	Mg (g)	P (g)
2020–2021	AK58	TopHP	16.76 ± 0.67 a	0.67 ± 0.01 b	1.23 ± 0.03 c	1.48 ± 0.03 a	1.71 ± 0.02 a
		DeepHP	15.68 ± 1.02 a	0.63 ± 0.03 c	1.24 ± 0.06 c	1.20 ± 0.01 c	1.28 ± 0.02 c
		TopRP	11.96 ± 0.35 bc	0.40 ± 0.01 e	1.07 ± 0.02 d	1.01 ± 0.02 d	1.23 ± 0.03 c
		DeepRP	11.08 ± 0.42 c	0.52 ± 0.01 d	1.49 ± 0.02 a	1.29 ± 0.04 b	1.46 ± 0.03 b
		No P	12.26 ± 0.24 b	0.76 ± 0.02 a	1.37 ± 0.01 b	1.18 ± 0.06 c	1.21 ± 0.01 c
	LX99	TopHP	18.01 ± 0.93 a	0.55 ± 0.03 a	1.29 ± 0.05 b	1.49 ± 0.10 b	1.65 ± 0.11 b
		DeepHP	12.19 ± 0.90 b	0.52 ± 0.02 a	1.43 ± 0.12 ab	1.68 ± 0.10 ab	1.86 ± 0.09 a
		TopRP	12.17 ± 0.15 b	0.39 ± 0.01 b	1.53 ± 0.07 a	1.73 ± 0.08 a	1.72 ± 0.06 ab
		DeepRP	12.68 ± 1.31 b	0.55 ± 0.06 a	1.50 ± 0.17 ab	1.51 ± 0.14 b	1.45 ± 0.10 c
		No P	12.98 ± 0.90 b	0.39 ± 0.02 b	1.52 ± 0.14 ab	1.47 ± 0.13 b	1.33 ± 0.08 c
	Alice	TopHP	17.70 ± 0.81 a	0.82 ± 0.02 a	2.27 ± 0.07 a	1.95 ± 0.14 a	1.73 ± 0.15 a
		DeepHP	15.46 ± 0.27 b	0.82 ± 0.01 a	1.89 ± 0.04 c	1.79 ± 0.05 b	1.90 ± 0.05 a
		TopRP	13.58 ± 0.15 cd	0.86 ± 0.02 a	2.13 ± 0.04 b	1.71 ± 0.04 bc	1.68 ± 0.09 a
		DeepRP	15.11 ± 1.64 bc	0.72 ± 0.06 b	1.72 ± 0.08 d	1.55 ± 0.11 d	1.70 ± 0.12 a
		No P	12.21 ± 0.34 d	0.57 ± 0.01 c	1.87 ± 0.04 c	1.59 ± 0.04 cd	1.67 ± 0.10 a
	Shavano	TopHP	10.68 ± 0.76 b	0.98 ± 0.10 b	1.43 ± 0.15 bc	1.33 ± 0.11 bc	1.30 ± 0.11 a
		DeepHP	12.60 ± 0.71 a	1.08 ± 0.10 b	2.41 ± 0.20 a	1.89 ± 0.16 a	1.42 ± 0.13 a
		TopRP	12.72 ± 1.09 a	0.95 ± 0.11 b	1.74 ± 0.20 b	1.53 ± 0.19 b	1.39 ± 0.12 a
		DeepRP	12.76 ± 0.46 a	1.34 ± 0.10 a	1.27 ± 0.09 c	1.22 ± 0.08 c	1.21 ± 0.12 ab
		No P	9.00 ± 0.80 c	1.00 ± 0.13 b	1.70 ± 0.21 b	1.22 ± 0.14 c	1.04 ± 0.11 b
2021–2022	AK58	TopHP	20.19 ± 2.40 a	0.76 ± 0.06 a	1.48 ± 0.08 c	1.83 ± 0.07 a	2.09 ± 0.29 a
		DeepHP	18.70 ± 1.12 a	0.77 ± 0.05 a	1.53 ± 0.10 c	1.51 ± 0.05 b	1.66 ± 0.01 b
		TopRP	13.16 ± 0.27 b	0.44 ± 0.04 c	1.29 ± 0.07 d	1.18 ± 0.03 d	1.52 ± 0.05 bc
		DeepRP	13.57 ± 0.51 b	0.62 ± 0.03 b	1.88 ± 0.04 a	1.52 ± 0.09 b	1.74 ± 0.19 b
		No P	13.89 ± 0.92 b	0.80 ± 0.03 a	1.67 ± 0.05 b	1.39 ± 0.03 c	1.35 ± 0.05 c
	LX99	TopHP	19.99 ± 1.56 a	0.60 ± 0.04 a	1.51 ± 0.10 a	1.68 ± 0.06 c	1.97 ± 0.05 b
		DeepHP	13.71 ± 1.47 b	0.56 ± 0.04 a	1.67 ± 0.18 a	1.98 ± 0.17 a	2.20 ± 0.14 a
		TopRP	13.15 ± 0.62 b	0.30 ± 0.03 c	1.74 ± 0.16 a	1.96 ± 0.24 ab	2.08 ± 0.16 ab
		DeepRP	13.39 ± 0.22 b	0.61 ± 0.03 a	1.61 ± 0.03 a	1.63 ± 0.05 c	1.74 ± 0.06 c
		No P	14.90 ± 0.73 b	0.41 ± 0.03 b	1.76 ± 0.19 a	1.70 ± 0.10 bc	1.58 ± 0.06 c
	Alice	TopHP	20.97 ± 1.60 a	0.94 ± 0.07 ab	2.82 ± 0.26 a	2.39 ± 0.19 a	2.26 ± 0.07 a
		DeepHP	17.21 ± 0.42 b	0.94 ± 0.01 ab	2.26 ± 0.08 b	2.15 ± 0.02 ab	2.30 ± 0.08 a
		TopRP	15.57 ± 1.40 bc	1.01 ± 0.06 a	2.58 ± 0.13 a	2.01 ± 0.13 bc	1.84 ± 0.08 bc
		DeepRP	17.74 ± 2.31 b	0.84 ± 0.11 b	2.12 ± 0.12 b	1.87 ± 0.17 c	1.99 ± 0.13 b
		No P	14.17 ± 0.98 c	0.65 ± 0.04 c	2.27 ± 0.05 b	1.91 ± 0.07 bc	1.68 ± 0.11 c
	Shavano	TopHP	12.60 ± 1.08 bc	1.14 ± 0.12 bc	1.76 ± 0.22 b	1.61 ± 0.12 b	1.61 ± 0.13 a
		DeepHP	16.24 ± 1.83 a	1.39 ± 0.19 ab	3.18 ± 0.45 a	2.48 ± 0.27 a	1.78 ± 0.23 a
		TopRP	14.66 ± 1.89 ab	1.07 ± 0.11 c	2.11 ± 0.25 b	1.77 ± 0.20 b	1.65 ± 0.10 a
		DeepRP	15.15 ± 1.33 ab	1.67 ± 0.22 a	1.61 ± 0.15 b	1.47 ± 0.22 b	1.46 ± 0.18 ab
		No P	10.47 ± 0.56 c	1.17 ± 0.10 bc	2.00 ± 0.11 b	1.48 ± 0.13 b	1.25 ± 0.14 b

Note: Different letters within the same column indicate significant differences at the 0.05 level.

). Corresponding with total P accumulation in grain, TopHP produced the highest total P accumulation in aboveground plant parts in AK58, while DeepHP produced the highest in LX99, Alice, and Shavano (Table 3).

3.4. Correlations among Root Distribution, Agronomic Traits, and Nutrient Utilization

The PCA explored associations among root distribution, agronomic traits, and nutrient utilization using 21 measured traits, including root distribution parameters, agronomic traits, and elemental concentrations. The PCA for these 21 traits in the four genotypes explained 56.19% of the variance in the first two components in the 2020–2021 season and 60.13% in the 2021–2022 season (Figure 3, Supplementary Figure S1). The first component (PC1) represented 41.11% of the variability in the 2020–2021 season and 43.16% in the 2021–2022 season, accounting primarily for root spatial distribution and straw weight. The second component (PC2) represented 15.08% of the variability in the 2020–2021 season and 16.97% in the 2021–2022 season, accounting primarily for 1000-kernel weight, grain number per head, and grain yield.

Pearson’s correlation analysis further examined the measured traits used in PCA. Spikelet number per head negatively correlated with non-fertile spikelet number and positively correlated with grain number per head and 1000-kernel weight in both seasons (Supplementary Figures S2 and S3). Root length density at all five depths positively correlated with grain yield and straw weight. Root length density at 0–20 cm negatively correlated with grain number per head and 1000-kernel weight. Straw K concentration positively correlated with grain number per head, 1000-kernel weight, straw Mg concentration, and straw P concentration, and negatively correlated with straw weight and root length density at 0–20, 20–40, 40–60, and 60–80 cm in both seasons. Grain K concentration positively correlated with grain P concentration, straw K concentration, grain number per head, 1000-kernel weight, and fertile spikelet number in both seasons.

3.5. Phosphorus Utilization Efficiency

Genotype, P placement method, and genotype × P placement method significantly affected straw and grain P concentrations. The different P placements resulted in different change patterns for the P utilization efficiency parameters (PUE, PAE, and PPE) in the four wheat genotypes (

Table 4. P utilization parameters of four wheat genotypes grown under TopHP, TopRP, DeepHP, and DeepRP treatments.

Season	Genotype	Treatment	PUE (%)	PAE (kg kg−1)	PPE (kg kg−1)
2020–2021	AK58	TopHP	5.2 ± 0.2 b	4.5 ± 0.4 b	86.5 ± 7.9 d
		DeepHP	3.1 ± 1.1 c	5.1 ± 0.5 b	213.4 ± 5.1 b
		TopRP	2.4 ± 0.5 c	5.9 ± 1.1 b	233.7 ± 14.1 a
		DeepRP	7.2 ± 0.5 a	8.6 ± 1.8 a	119.2 ± 21.1 c
		No P	–	–	–
	LX99	TopHP	3.5 ± 1.5 ab	4.4 ± 0.6 a	208.4 ± 6.0 b
		DeepHP	6.0 ± 2.3 a	5.8 ± 0.4 a	163.6 ± 5.6 c
		TopRP	5.2 ± 0.6 a	6.1 ± 1.6 a	116.3 ± 18.8 d
		DeepRP	1.6 ± 0.4 b	4.5 ± 1.4 a	343.1 ± 37.6 a
		No P	–	–	–
	Alice	TopHP	2.8 ± 0.4 c	9.0 ± 1.7 a	313.9 ± 24.4 a
		DeepHP	4.0 ± 0.4 b	6.3 ± 1.0 ab	158.7 ± 20.2 b
		TopRP	2.3 ± 0.4 c	8.1 ± 2.1 ab	353.0 ± 40.5 a
		DeepRP	5.3 ± 0.9 a	5.1 ± 1.5 b	96.5 ± 12.9 c
		No P	–	–	–
	Shavano	TopHP	1.7 ± 0.5 a	6.3 ± 1.9 a	369.6 ± 57.1 a
		DeepHP	2.3 ± 0.3 a	6.5 ± 0.7 a	250.6 ± 13.3 b
		TopRP	4.1 ± 1.7 a	6.8 ± 2.0 a	142.2 ± 29.9 c
		DeepRP	2.8 ± 2.1 a	9.0 ± 1.1 a	386.7 ± 15.2 a
		No P	–	–	–
2021–2022	AK58	TopHP	4.9 ± 2.1 a	8.0 ± 1.2 a	184.1 ± 80.0 b
		DeepHP	2.8 ± 1.6 a	8.2 ± 0.9 a	400.8 ± 54.1 a
		TopRP	2.3 ± 0.2 a	5.9 ± 0.5 a	261.6 ± 43.2 b
		DeepRP	5.1 ± 2.8 a	8.6 ± 3.6 a	184.3 ± 45.8 b
		No P	–	–	–
	LX99	TopHP	4.2 ± 1.4 ab	3.6 ± 0.3 a	144.2 ± 19.6 a
		DeepHP	6.8 ± 2.5 a	6.1 ± 2.1 a	147.6 ± 35.0 a
		TopRP	6.5 ± 1.2 a	5.8 ± 2.8 a	86.9 ± 25.1 b
		DeepRP	2.0 ± 0.8 b	2.9 ± 1.9 a	139.9 ± 32.2 ab
		No P	–	–	–
	Alice	TopHP	3.9 ± 0.5 a	9.4 ± 3.0 a	240.3 ± 61.6 a
		DeepHP	4.2 ± 0.5 a	6.6 ± 2.4 ab	155.5 ± 50.2 ab
		TopRP	2.2 ± 0.9 b	3.8 ± 2.1 b	167.4 ± 45.0 a
		DeepRP	4.1 ± 0.8 a	3.3 ± 0.3 b	89.8 ± 16.3 b
		No P	–	–	–
	Shavano	TopHP	2.4 ± 0.4 a	7.8 ± 1.2 a	336.9 ± 68.9 a
		DeepHP	2.9 ± 0.3 a	7.9 ± 1.8 a	225.8 ± 14.5 b
		TopRP	5.5 ± 0.3 a	6.2 ± 2.4 a	112.9 ± 26.4 c
		DeepRP	3.0 ± 0.2 a	7.9 ± 2.3 a	278.9 ± 12.3 ab
		No P	–	–	–

Note: Different letters within the same column indicate significant differences at the 0.05 level. Abbreviations: PUE, phosphorus utilization efficiency; PAE, phosphorus agronomic efficiency; PPE, phosphorus physiological efficiency.

). The highest PUE across both seasons occurred for AK58 under TopHP, for LX99 under DeepHP, and for Shavano under TopRP. PUE also varied across seasons, with Alice reaching the highest PUE under DeepRP in the 2020–2021 season and DeepHP in the 2021–2022 season. The highest PAE occurred for AK58 and Shavano under DeepRP in both seasons. PPE varied largely between genotypes, P placements, and seasons.

4. Discussion

4.1. Effect of Season, Genotype, and P Placement on Root Distribution and Agronomic Traits

This study investigated how different P placements influence root distribution, agronomic traits, and nutrient utilization in tall and modern dwarf wheat genotypes. Crop growth in rainfed fields depends mainly on rainfall during the growing season. Frequent rainfall is associated with higher grain yield, whereas lower rainfall often results in reduced yields. All four genotypes exhibited more extensive root systems in the wet season (2021–2022) than in the dry season (2020–2021), suggesting increased rainfall positively impacted growth. Notably, substantial differences in root systematic architecture were observed among genotypes. For instance, the two modern dwarf wheat genotypes (AK58 and LX99) had smaller root systems, characterized by reduced root length density, than the tall wheat genotypes (Alice and Shavano) at all soil depths, consistent with earlier research on root architecture differences between these types of wheat [22,23].

Environmental factors and agronomic practices can influence root growth. Uneven P distribution due to localized P supply can stimulate root development in nutrient-rich zones [24]. For instance, root growth increased near the fertilizer (nitrogen and phosphorus) application sites in the subsoil, with minimal effects in other parts of the profile [25]. Similarly, in the current study, TopHP increased RLD at a 0–20 cm depth, while deep-banded P increased RLD at 20–40 cm depth compared to top-dressed P for all four genotypes, indicating that localized P supply enhances root growth. Different P input amounts and placement depths altered the root response to P in this study, with significant changes in root distribution, particularly at 0–20 and 20–40 cm depths among treatments, suggesting an adaptive root development response to P placement.

Roots forage soils for water and nutrients to support aboveground plant growth, while aboveground plant plants, in turn, allocate some photosynthetic assimilates to support root growth. Roots primarily explore topsoil during early growth, extending downwards as growth progresses. Whether deep-banded P can enhance P uptake and biomass relative to top-dressed P depends on root distribution and P availability in different soil layers. P supply should, in principle, alter root distribution by modifying aboveground agronomic performance. For example, spike number, grain number, and grain weight of two wheat cultivars increased with increasing applied P in the field [26] and in column culture glasshouse conditions [22]. Similarly, in the current study, wheat grown with P supply had fewer non-fertile spikelets and more grain per head than No P. However, the effect of different P placements led to coordinated changes in individual agronomic traits, which varied across seasons, genotypes, and P placements.

The advantage of a particular root ideotype in enhancing grain yield depends on the growing environment. For instance, durum wheat types with deep roots increased yields under low moisture by 37–38% but decreased yield by 20–40% in moisture-rich environments relative to shallow root types [27]. The impact of P supply on grain yield depends largely on soil P availability before sowing. Halvorson et al. (1992) found no significant grain yield differences between broadcast and banded treatments with an initial 10 mg kg⁻¹ NaHCO₃-extractable P in soil (pH 7.8) [28]. The yield response to added P can also vary among soils with different properties [29]. Moreover, weather conditions such as rainfall distribution can lead to different crop reactions in root system development [30]. Our previous data indicated that increased P supply could enhance belowground and aboveground growth in wheat under saline conditions [22]. In the current study, the two tall genotypes (Shavano and Alice) consistently outperformed the two dwarf genotypes (AK58 and LX99) in terms of grain yield in both seasons, suggesting that a more extensive root system provides an advantage in sustaining grain yield under rainfed conditions.

4.2. Nutrient Utilization as Affected by Genotype and P Placement

The superior performance of banded fertilizer compared to broadcast fertilizer for dryland crops may be because broadcast fertilizer becomes less available as surface layer roots become less active and effective in capturing nutrients [31]. Moreover, the interaction between genotype and P placement significantly influenced grain yield. Previous research demonstrated that deep P placement increased grain yield in a drought-tolerant wheat cultivar (CH58) and decreased grain yield in a drought-sensitive cultivar (XY22) [9]. Similarly, DeepHP produced the highest grain yields for AK58, LX99, and Shavano, while TopHP produced the highest grain yield for Alice across both seasons. These results suggest that the beneficial effect of deep-banded P on wheat is genotype dependent.

No consistent change pattern in nutrient concentrations was observed among different genotypes in response to P placement due to significant genotype P placement interactions. Correlations among root distribution, agronomic traits, and plant nutrient concentrations were generally consistent between seasons, indicating a shared mechanism in maintaining the balance between plant growth and nutrient uptake. The interaction between sodium and potassium use efficiency in wheat has been suggested previously [32]. In the current study, straw K concentration positively correlated with grain number per head, 1000-kernel weight, straw Mg concentration, and straw P concentration, while grain K concentration positively correlated with grain P concentration, straw K concentration, grain number per head, 1000-kernel weight, and fertile spikelet number. These findings imply that enhanced K uptake under salt stress could alleviate growth inhibition and yield reduction.

Differential responses in grain and straw P concentrations among genotypes to P placement suggest a genotypic effect on P utilization. While the Deep P treatments exhibited a yield advantage for AK58, LX99, and Shavano relative to the Top P treatments, no such advantage was evident for Alice. Moreover, the effects of P placement on different wheat genotypes led to distinct patterns of P utilization efficiency parameters (PUE, PAE, and PPE). Among genotypes, the highest PUEs were recorded for different P placements: AK58 under TopHP, LX99 and Alice under DeepHP, and Shavano under TopRP. These results highlight the potential benefits of combining appropriate P placement with suitable genotypes to improve PUE.

4.3. P Utilization in Relation to Rainfall under Rainfed Conditions

Nutrient transport in dry soils toward roots can be hindered by ion diffusion, which depends on soil moisture content [33]. Consequently, soil P availability to plants in arid and semi-arid regions is related to soil moisture [34,35,36]. Thus, the effectiveness of deep-placed fertilizers can vary based on seasonal conditions, with greater effectiveness expected for drier years [4]. Nutrient availability and utilization efficiency decrease with lower soil water contents [35]. In dryland settings, P input may result in satisfactory yields in years with timely precipitation, but greater yield losses could occur in drier years. This phenomenon is particularly evident in spring, a critical growth stage for winter wheat in North China (jointing stage), where supplementary irrigation is often used to achieve high yields [37]. Wheat growth in rainfed coastal saline fields can vary between dry and wet seasons. In the current study, higher rainfall occurred just before sowing, during sowing (September and October), and at the jointing stage (March) in the 2021–2022 season relative to the drier 2020–2021 season (Table 1), which positively affected plant growth. Consequently, the overall agronomic traits and PUE for all four genotypes in 2021–2022 (wet season) outperformed those in 2020–2021 (dry season).

Notably, the benefits of deep P placement are related to the soil background conditions, as the critical Olsen-P can vary significantly among different soils [29]. Therefore, determining the optimal P input amount combined with an appropriate placement depth for a specific soil becomes crucial. Furthermore, investigating plant growth responses to different P placements under saline soils with varying salinity levels is essential for enhancing our understanding of reasonable P management strategies for coastal saline fields.

5. Conclusions

This study revealed significant interactive effects between season, genotype, and P placement on root distribution, agronomic traits, and nutrient utilization in wheat under rainfed coastal saline conditions. Root growth and agronomic performance varied between seasons, with growth advantages more pronounced in the wet season (2021–2022) due to the higher soil moisture during wheat growth. Localized P supply had a clear stimulus effect on root growth, with TopHP increasing RLD at a 0–20 cm depth and deep-banded P increasing RLD at a 20–40 cm depth for all four genotypes. Deep-banded P outperformed top-dressed P for three wheat genotypes (excluding Alice) in both seasons. The modified PUE due to P placement was associated with changes in root distribution, agronomic traits, and nutrient concentrations in straw and grain. The positive correlations between straw K concentration and grain number per head, 1000-kernel weight, straw Mg concentration, and straw P concentration, and between grain K concentration and grain P concentration, straw K concentration, grain number per head, 1000-kernel weight, and fertile spikelet number across all treatments highlighted the importance of K uptake in sustaining crop yield under saline conditions. Both P placement and genotype should be considered when aiming to improve PUE. The findings suggest that selecting an appropriate P placement and a well-suited genotype could enhance sustainable wheat production in rainfed coastal saline environments.