Optimizing Irrigation and Nitrogen Fertilizer Regimes to Increase the Yield and Nitrogen Utilization of Tibetan Barley in Tibet

Abstract

Nitrogen (N) fertilization plays a pivotal role in the nitrogen transport process and yield formation of field-grown Tibetan barley (Hordeum vulgare L., qingke in Chinese); however, little is known about its interaction with irrigation regimes. Here, we performed a control experiment to investigate the effects of irrigation regimes (primary irrigation and double irrigation, mentioned as W1 and W2) and N levels (0, 90, 120, and 150 kg ha⁻¹, mentioned as N0, N9, N12, and N15) on the nitrogen accumulation, translocation, and utilization of Tibetan barley in the Tibetan Plateau during the spring barley seasons in 2022. The results showed that the highest yield (6242.28 kg ha⁻¹) and aboveground biomass (12,354.13 kg ha⁻¹ for anthesis; 15,827.9 kg ha⁻¹ for maturity) were achieved in W2N15 as compared to other treatments. The maximum grain N accumulation (117.66 kg ha⁻¹), the N translocation (54.16 kg ha⁻¹), and the post-anthesis N accumulation (63.5 kg ha⁻¹) were achieved in the W1N15 treatment. The N utilization efficiency increased with irrigation frequency and decreased with N application; however, the conclusion given by the N agronomic efficiency is contrary to this trend. The grain yield had significant positive correlations with the grain N accumulation (W1: r = 0.98; W2: r = 0.97) and N translocation (W1: r = 0.84; W2: r = 0.94), but significant negative correlations with the N harvest index (W1: r = −0.95; W2: r = −0.95) and N utilization efficiency (W1: r = −0.9; W2: r = −0.85). The path analysis revealed that the factors related to N utilization (β = 0.875) and the factors related to N translocation (β = −1.426) were the significant direct contributors towards grain yield. The influence of N application (total effect = 0.922) on the grain yield was much stronger than that of the irrigation regime (total effect = 0.324). Our findings can guide future efforts in designing sustainable water and N fertilizer management strategies for Tibetan barley in the Tibetan Plateau.

1. Introduction

The Tibetan Plateau within the Tibet Autonomous Region covers an area of approximately 1.22 million ha, and it is an important part of the farming and pastoral zone of China [1]. While water is available in ample amounts in Tibet, the spatiotemporal distribution of water resources is uneven, with about 60% area lying in arid and semi-arid regions, having an annual rainfall ranging from 50 mm in the north-west to 1000 mm in the south-east, and an annual evaporation, from 300 to 3000 mm [2,3,4]. The nitrogen fertilization needed to achieve high yields results in over-fertilization [5]. The amount of fertilizer in Tibet was 60, 300 tons (pure nutrient), and the average amount of N per hectare was 258.45 kg, which makes the single-season nitrogen (N) fertilizer input significantly higher than that in other provinces (considering a lower multiple cropping index) [6]. In terms of the regional distribution of fertilization, internal disparities exist within the counties. The quantity of fertilizer application can reach up to 600 kg ha⁻¹ in Lhasa and its surrounding suburbs, whereas in some remote areas, fertilizer consumption is only 0 kg ha⁻¹ [7]. Seasonal drought [8], regional water scarcity [9], and improper fertilization [10] have intensified the contradiction between resources’ supply and demand, imposing stress on locally adapted crops and agricultural systems. The exploration of effective fertilization schemes and irrigation methods is an important measure for efficient Tibetan barley production and ecological environment security in Tibet.

Though the uptake of water and nutrients in crops are two separate biochemical processes, the soil moisture regime significantly affects the root system architecture and its physiological activities, making interconnections complex in the utilization of water and nutrients [11]. An excessive nitrogen application rate may not have a major impact on crop yield, it can have a considerable impact on the amount of nitrogen residue in the soil after crop harvest and decrease the use efficiency of nitrogen [12,13]. In particular, the rainy season comes soon after barley harvest in Tibet, which may easily exacerbate the residual nitrogen leaching into deeper soil layers. The response to irrigation and nitrogen levels for the production of crops and investigation into the utilization of water and nitrogen under various schedules was studied. For example, Naghdyzadegan Jahromi et al. [14] treated the barley with different nitrogen and water application limits, and they found that the highest yield was achieved at 75% potential evapotranspiration (ET0) combined with an N supply of 210 kg ha⁻¹ in both years. In another study, Ghasemi-Aghbolaghi and Sepaskhah [15] studied the effect of different nitrogen levels (0, 90, and 180 kg ha⁻¹) and partial root-zone drying irrigation on barley, and they found that the maximum water use efficiency was observed in variable alternate furrow irrigation with nitrogen application of 180 kg ha⁻¹. Those findings under controlled conditions serve as a base to elucidate drought tolerance mechanisms activated with N fertilization.

Tibetan barley (Hordeum vulgare L.), also known as “qingke” in Chinese and “nas” in Tibetan, is widely planted on the Qinghai-Tibetan Plateau [16]. It grows well at an elevation of 1400 to 4700 m due to its ability to be cold and drought tolerant and its early maturity [17]. According to the 2022 statistics of the Tibetan region, the total cultivated area of Tibetan barley in the region amounted to 14.74 million hectares, accounting for 77.54% of the total crop production and 53.2% of the total cultivated area of all crops [18]. Under future scenarios of global climate change, barley yield variability in Tibet is not only about agricultural production but also about food security. The optimization of water and N application rates could improve the canopy photosynthetic efficiency [19] and dry matter accumulation [20], which would ultimately contribute to enhancing the crop grain yield [21]. The interaction effects between the N rates and irrigation could be important in different climate conditions. Previous studies have investigated the effects of nitrogen application rates under water-saving irrigation on grain yield in Ethiopia [22], Iran [14,23,24], and the Mediterranean region [25], but limited research on nitrogen fertilizer and irrigation water management is available in the Tibetan Plateau. Hence, we conducted a control experiment to analyze barley yield, nitrogen transport processes, and nitrogen utilization under various irrigation and nitrogen fertilizer regimes. Furthermore, we elucidated the underlying mechanism governing the relationship between irrigation–nitrogen interactions and yield formation. In this study, we hypothesized that increasing the irrigation frequency and N application rate could improve the dry matter production post-anthesis so as to ensure grain yield, promote the absorption and utilization of nitrogen, increase nitrogen accumulation in Tibetan barley, and reduce N agronomic efficiency and N utilization efficiency. To confirm this, we analyzed the effects of reducing nitrogen application rates under two irrigation frequencies on (1) grain yield and yield components of Tibetan barley, (2) nitrogen accumulation indicators, (3) nitrogen translocation indicators, and (4) nitrogen utilization. We expect that our results can provide a theoretical basis and technical reference for enhancing Tibetan barley yield and nitrogen use efficiency in the Tibetan Plateau.

2. Materials and Methods

2.1. Experimental Site

The present study was conducted in 2022 at the Tibet Academy of Agricultural and Animal Husbandry Sciences in Lhasa (29°38′ N, 91°01′ E and 3662 m elevation), Tibet Autonomous Region of China. The experimental field is flat with uniform fertility and the preceding crop is Tibetan barley. The region has a plateau temperate monsoon semi-arid climate with an average of 426.5 mm annual precipitation and an 8.0 °C air temperature for the last three decades. An in-field weather station (Automatic Weather Station, Shende Technology Co., Ltd., Zhengzhou, China) recorded a total rainfall of 187.6 mm and mean daily minimum and maximum air temperatures of 11 °C and 22.7 °C, respectively, during the experimental period (Figure 1). The soils have been classified as meadow soils (umbric gleysols/haplic phaeozem in the World Reference Base for Soil Resources, WRB) [26], and the soil properties of the topsoil layer (0–20 cm) were determined as described by Bao [27] and are presented as follows: bulk density of 1.39 g cm⁻³, field capacity of 0.284 cm⁻³, pH of 7.67, soil organic carbon of 18.61 g kg⁻¹, total nitrogen of 2.23 g kg⁻¹, total phosphorus of 1.4 g kg⁻¹, and total potassium of 5.75 g kg⁻¹.

2.2. Experimental Design and Crop Management

The Tibetan barley cultivar “Zangqing 3000”, a widely planted lodging-resistant variety in Tibet, was sown on 16 April 2022, and harvested on 2 August 2022 with a density of 225 kg ha⁻¹, with a row spacing of 25 cm. The treatments consisted of two levels of irrigation water regimes: primary irrigation (irrigating at the seedling stage, W1) and double irrigation (irrigating at both the seedling and jointing stages, W2), and four nitrogen application rates: 0, 90, 120, and 150 kg N ha⁻¹ (N0, N9, N12, and N15, respectively). The irrigation water source was groundwater extracted from a depth of 8 m, and the water consumption was measured using a water meter. Specifically, 15 mm of irrigation was applied to all plots prior to sowing in order to ensure emergence post-sowing. After that, irrigation treatment was applied solely on 2 May for W1, whereas for W2, irrigation was conducted on both 2 May and 15 June. About 50 mm of irrigation volume was applied in each irrigation event, which is a common volume among local farmers.

Twenty-four plots of 5.6 × 5.5 m were established on 15 April 2022 using a randomized block design with three replicates. Each experimental plot set a 1.0 m buffer zone between plots to minimize the effects of adjacent plots. Fertilizers (65 kg P ha⁻¹ as calcium superphosphate, and 62 kg K ha⁻¹ as potassium sulfate) were broadcast and incorporated into the soil as a base fertilizer before sowing. Nitrogen as urea (46.5% N) was added to the soil surface at the seedling (2 May) and jointing stages (15 June) in two equal doses. Other field management practices (such as weeding and plowing) were similar to the local field operations.

2.3. Sampling and Measurement

The anthesis of Tibetan barley occurred from 28 June to 30 June, lasting 3 days, whereas the maturity phase extended from 17 July to 2 August, amounting to a total of 16 days. During both periods, samples of one-meter rows were selected randomly from each treatment and cut down to ground level. We determined the spike number per sample in each replicate. The plants were divided into leaves, sheaths, stems, and panicles at anthesis and into leaves, sheaths, stems, panicles, and grains after maturity. Then, samples were oven-dried at 65 °C to a constant weight and weighed for determination of the aboveground biomass (AB, kg ha⁻¹) and grain yield (GY, kg ha⁻¹). The 1000-grain weight was determined by manually threshing the panicles and counting the number of grains per sample. Then, samples were pulverized, the powder was digested with concentrated sulfuric acid (H₂SO₄, 98% Sinopharm Chemical Reagent Co., Ltd., Shanghai, China) and hydrogen peroxide (H₂O₂, 30% Sinopharm Chemical Reagent Co., Ltd., Shanghai, China), and total nitrogen analysis was performed using the Kjeldahl method (Hanon K9840 Kjeldahl apparatus, Hanon Instruments Co., Ltd., Jinan, China), as described by Yan et al. [28].

The following various parameters referring to nitrogen accumulation and translocation were calculated according to Zheng et al. [29] and Wang et al. [30] as below:

Nitrogen accumulation (NA, kg ha⁻¹) = N concentration (%) × AB (kg ha⁻¹);

Nitrogen distribution ratio (NDR, %) = [NA of vegetative organ/NA of all organs (kg ha⁻¹)] × 100;

Nitrogen translocation (NT, kg ha⁻¹) = NA of vegetative organs at anthesis (kg ha⁻¹) − NA of vegetative organs at maturity (kg ha⁻¹);

Contribution of N translocation amount from the vegetative organ to the grain (NRC, %) = [NT (kg ha⁻¹)/NA of grain (kg ha⁻¹)] × 100;

Nitrogen translocation ratio (NTR, %) = [NT (kg ha⁻¹)/NA of vegetative organs at anthesis (kg ha⁻¹)] × 100;

Post-anthesis NA (kg ha⁻¹) = NA of a plant at maturity (kg ha⁻¹) − NA of a plant at anthesis (kg ha⁻¹);

Post-anthesis NRC (%) = [Post-anthesis NA of a plant (kg ha⁻¹)/NA of a plant at maturity (kg ha⁻¹)] × 100.

The nitrogen harvest index, nitrogen utilization efficiency, agronomic efficiency of nitrogen, and nitrogen partial factor productivity were calculated in accordance with Papakosta and Gagianas [31]; Azizian and Sepaskhah [32]; and Dai et al. [33].

Nitrogen harvest index (NHI) = NA of grain (kg ha⁻¹)/NA of plant at maturity (kg ha⁻¹);

Nitrogen utilization efficiency (NUtE, kg kg⁻¹) = grain yield (kg ha⁻¹)/NA of a plant at maturity (kg ha⁻¹);

Agronomic efficiency of nitrogen (NUE, kg kg⁻¹) = [grain yield with nitrogen treatment (kg ha⁻¹) − grain yield with no nitrogen (kg ha⁻¹)]/nitrogen fertilizer applied (kg ha⁻¹);

Nitrogen partial factor productivity (NPFP, kg kg⁻¹) = grain yield with nitrogen treatment application (kg ha⁻¹)/nitrogen fertilizer applied (kg ha⁻¹).

2.4. Statistical Analysis

Microsoft Excel (Microsoft Corporation, Redmond, WA, USA, 2016) [34] was used to collate the data; statistical analyses and graphics were conducted using the R software (R Core Team, version 4.2.2, 2022) [35]. Specifically, the normality of the data and the homogeneity of variances were checked by the Shapiro–Wilk and Bartlett tests, respectively. Two-way ANOVA was used to evaluate the main and interaction effects of treatment (irrigation, nitrogen, and their interactions) on the yield parameters, aboveground biomass, NA parameters, and NT parameters of Tibetan barley. To identify significant effects, the differences were tested by Fisher’s least significant difference (LSD) at p < 0.05. We then calculated the Pearson correlation coefficients between GY and grain NA, NT, NTR, NRC, NHI, NUtE, post-anthesis NA, and post-anthesis NRC, respectively (abbreviations are explained in Supplementary Table S1). The analyses above were conducted with the “agricolae” package (R Core Team, version 1.3-7, 2023) [36].

Multiple regression analysis was carried out using both forward and backward selection to test the effect of factors on GY [37,38]. Model comparisons and selection were assessed using AIC (Akaike’s information criterion) while the significance of the measured variables was assessed via ANOVA with F tests, the model with the lowest AIC being the most strongly supported. From the candidate set, the conditional coefficient of determination (R²) for our models was calculated as a metric of explanatory power. This part was carried out with the “MASS” package (R Core Team, version 7.3-60, 2023) [39]. Furthermore, we selected important predictors based on the AIC and applied the partial least squares path model (PLS-PM) to check for the effects of parameters referring to nitrogen on GY [40,41].

The conceptual PLS-PM included the direct effects of these factors on GY, as well as the indirect influences where irrigation and nitrogen application affected yield formation via changing NA-related factors (grain NA, post-anthesis NRC), nitrogen utilization factors (NHI, NUtE), and/or NT-related factors (NT, NTR) under control conditions. The inner model was evaluated by the goodness-of-fit (GOF) statistics. Detailed statistical parameters for outer model are depicted in Supplementary Table S2. This part was performed using the “pls-pm” package (R Core Team, version 0.5.0, 2023) [42].

3. Results

3.1. Aboveground Biomass, Grain Yield, and Yield Components

The aboveground biomass was in the range of 7901.78–15,827.91 kg ha⁻¹ for both stages of Tibetan barley (Figure 2). The results showed that the nitrogen application of 150 kg ha⁻¹ under double irrigation produced the maximum biomass; whereas the minimum biomass was obtained in the W1N0 plots for both stages. Barley biomass was significantly influenced by the N15 treatments, which resulted in a higher biomass (ranging from 3.76% to 44.62% in W1 and from 22.35% to 66.12% in W2) at anthesis, compared to other N treatments (Figure 2a,b). Similar changes in aboveground biomass were reported at maturity, the percent increment in total aboveground biomass due to 150 kg N ha⁻¹ was 49.69 and 32.29% over the control in the W1 and W2 treatments (p < 0.05), respectively.

The ANOVA results indicated that nitrogen had a significant effect on all traits. The irrigation regime significantly impacted the 1000-grain weight (p < 0.05) and GY (p < 0.001), but no significant effects were observed on the spike number. Notably, the interaction between nitrogen (N) and water (W) significantly affected all studied traits (p < 0.05), with the exception of the spike number (

Table 1. Variance analysis for the yield and yield components of Tibetan barley.

Source of Variation	1000-Grain Weight	Spike Number	Grain Yield
W	*	ns	***
N	**	*	***
W × N	*	ns	*

*, **, ***, and ns indicate p < 0.05, p < 0.01, p < 0.001, and no significant difference, respectively.

).

The effects of the N application rates and irrigation regimes on Tibetan barley production are shown in Figure 3. The highest 1000-grain weight of 55.41 g was obtained in W1 from the application of 150 kg N ha⁻¹. Similarly, the highest 1000-grain weight in W2 (59.71 g) was recorded from the N15 treatment. However, the lowest 1000-grain weights (51.29 and 53.77 g) were obtained from 120 kg N ha⁻¹ in W1 and 90 kg N ha⁻¹ in W2, respectively (Figure 3a). The magnitude of the variation in spike number change with N levels was not consistent, varying between 232 and 257 × 10⁴ ha⁻¹ for barley, and no significant difference was observed between the irrigation treatments (Figure 3b). Nitrogen application increased the grain yield in both irrigation strategies. Compared with N0, there was an increase in GY of 17.29–36.8% in W1 and 12.59–30.57% in W2 (p < 0.05), respectively. Under the same nitrogen application rate, W2 significantly increased the yield by 4.45–12%, compared to W1 (Figure 3c).

3.2. Nitrogen Accumulation and Distribution Ratio

The ANOVA results indicated that the nitrogen accumulation was significantly affected by irrigation (W) and nitrogen (N); however, they were not influenced by the interaction effect of W × N (

Table 2. Variance analysis for nitrogen accumulation of Tibetan barley under different treatments.

Source of Variation	Anthesis			Maturity
	Leaf	Leaf Sheath	Stem	Leaf	Leaf Sheath	Stem	Grain
W	*	***	***	**	**	***	**
N	***	***	***	***	***	***	***
W×N	ns	ns	ns	ns	ns	ns	ns

*, **, ***, and ns indicate p < 0.05, p < 0.01, p < 0.001, and no significant difference, respectively.

). The effect of irrigation had significant effects on the nitrogen accumulation in the leaf sheath and stem of Tibetan barley at anthesis (p < 0.001) and in the stem at maturity (p < 0.001). The ANOVA for NA from different plant parts revealed that the effect of nitrogen was significant for all traits at both stages (p < 0.001).

The effects of nitrogen and water regimes on the NA of different plant parts are shown in Figure 4a for anthesis and Figure 4c for maturity. NA in Tibetan barley differed among vegetative organs, as well as growth periods. The NA in the leaves, leaf sheaths, and stems of Tibetan barley increased with the increase in the nitrogen application rate, and the total nitrogen accumulation in the N15 (150 kg ha⁻¹) and N12 (120 kg ha⁻¹) treatments was significantly higher than that in the N0 (no nitrogen application) and N9 (90 kg ha⁻¹) treatments (p < 0.05) (Figure 4a,c). The mean of NA under the N15 treatment reached 45.1, 33.41, and 36.23 kg ha⁻¹ in leaves, leaf sheaths, and stems at anthesis, respectively, which were higher values than those obtained under the N0 and N9 treatments. Regarding irrigation treatments, the NA of vegetative organs under W2 was 0.82–38.98% lower than that under W1 (Figure 4a). In terms of vegetative organs, the maximum NDR observed in leaves was 38.07–48.57% (Figure 4b), which was significantly higher than others (p < 0.05). Consistent with the changes in NA at anthesis, nitrogen treatments at maturity (N9, N12, and N15) significantly increased the NA in vegetative organs (p < 0.05). However, the NA variable among vegetative organs, with the maximum value of NA being 84.74–117.66 kg ha⁻¹ in grain, accounting for 64.48–75.33% of the entire organs (Figure 4c), and the NDR decreased with the increase in nitrogen application (Figure 4d). At the same N application levels, NA showed a decreasing trend from W1 to W2. The highest reduction (24.29%) was observed in N12 on the stem while the least reduction (2.46%) was exhibited in N15 on the grain (Figure 4d). The results imply certain positive correlations between N application and NA; the increase in NA was mainly affected by the fertilizer N application rate.

3.3. Nitrogen Translocation

Nitrogen translocation from the vegetative organs to the grains is shown in

Table 3. Nitrogen translocation of Tibetan barley under different treatments.

Treatments	NT (kg ha−1)	NTR (%)	NRC (%)	Post-Anthesis NA (kg ha−1)	Post-Anthesis NRC (%)
W1N0	39.32 bc	54.18 ab	44.55 ab	48.97 c	55.45 ab
W1N9	50.16 a	51.41 abc	47.5 a	55.44 bc	52.5 b
W1N12	53.26 a	48.19 cd	47.6 a	58.65 ab	52.4 b
W1N15	54.16 a	45.47 d	46.06 ab	63.5 a	53.94 ab
W2N0	33.84 c	54.78 a	39.94 b	50.9 c	60.06 a
W2N9	41.86 b	50.1 abcd	41.2 ab	59.79 ab	58.8 ab
W2N12	49.09 a	49.63 bcd	45.37 ab	59.08 ab	54.63 ab
W2N15	53.8 a	48.66 cd	46.88 a	60.97 ab	53.12 b
ANOVA
W	*	ns	ns	ns	ns
N	***	**	ns	*	ns
W × N	ns	ns	ns	ns	ns

NT: N transportation; NTR: N transportation rate; NRC: Contribution of N remobilized to grain; Post-anthesis NA: N accumulation after anthesis; Post-anthesis NRC: Contribution of N translocation amount from the vegetative organ to the grain after anthesis. W1 and W2 denote primary irrigation and double irrigation, respectively. N0, N9, N12, and N15 represent the N application rates at 0, 90, 120, and 150 kg N ha⁻¹, respectively. *, **, ***, and ns indicate p < 0.05, p < 0.01, p < 0.001, and no significant difference, respectively. Different lowercase letters indicate significant differences between treatments at p < 0.05 with the LSD test.

. The irrigation regime significantly affected the NT (p < 0.05) but had no significant effects on NTR, NRC, post-anthesis NA, or post-anthesis NRC. The effects of N were significant for NT (p < 0.001), NTR (p < 0.01), and post-anthesis NA (p < 0.05), while the effect of the interaction W × N was not significant for translocation processes. Under the same irrigation level, N application decreased NTR but considerably increased NT (Table 3). However, NRC shows a trend of first increasing and then decreasing with the increase in N under W1, and the highest value was recorded in N12 (47.6%). Furthermore, post-anthesis NA ranged from 48.97 to 63.5 kg ha⁻¹, and post-anthesis NRC varied between 52.4 and 60.6%, suggesting that post-anthesis NA was the primary source of grain nitrogen which was assimilated by Tibetan barley at a sufficient nitrogen supply.

3.4. Nitrogen Utilization

The results of the analysis of variance (ANOVA) in

Table 4. Effects of water and nitrogen combination on nitrogen utilization of Tibetan barley.

Treatments	NHI	NUtE (kg kg−1)	NUE (kg kg−1)	NPFP (kg kg−1)
W1N0	0.73 b	35.12 d
W1N9	0.69 c	32.72 e	8.2 c	55.63 b
W1N12	0.66 ef	32.32 e	9.92 ab	45.49 d
W1N15	0.64 f	32.01 e	10.47 a	38.93 f
W2N0	0.75 a	42.51 a
W2N9	0.71 b	37.58 b	6.69 d	59.81 a
W2N12	0.68 cd	36.1 c	7.68 c	47.52 c
W2N15	0.67 de	36.4 c	9.74 b	41.62 e
ANOVA
W	***	***	***	***
N	***	***	***	***
W × N	ns	***	**	*

NHI: Nitrogen harvest index; NUtE: Nitrogen utilization efficiency; NUE: Agronomic efficiency of nitrogen; NPFP: Partial factor productivity from applied nitrogen. W1 and W2 denote primary irrigation and double irrigation, respectively. N0, N9, N12, and N15 represent the N application rates at 0, 90, 120, and 150 kg N ha⁻¹, respectively. *, **, ***, and ns indicate p < 0.05, p < 0.01, p < 0.001, and no significant difference, respectively. Different lowercase letters indicate significant differences between treatments at p < 0.05 with the LSD test.

showed that the values under N and W were significant for NHI, NUtE, NUE, and NPFP (p < 0.001), along with the effect of the N × W interaction which significantly affected all of the studied traits except for NHI (Table 4). Under the same irrigation level, N application considerably decreased NHI and NPFP, and decreased NUtE only in W1. Contrasting with this, however, NUE increases with the increase in N application, with minimum and maximum values of 8.2 and 10.47 kg kg⁻¹, respectively, for W1, and 6.69 and 9.74 kg kg⁻¹, respectively, for W2. Under the same N application conditions, NHI, NUtE, and NPFP in the W1 treatment were significantly lower by 2.72–3.64%, 10.49–17.39%, and 4.26–6.99%, respectively, compared to those in the W2 treatment (p < 0.05).

3.5. Relationship between Yield and N Parameters

Different correlations between GY and grain NA, NT, NTR, NRC, NHI, NUtE, post-anthesis NA, and post-anthesis NRC were observed in Tibetan barley (Figure 5). The correlation analysis showed that the grain yield of barley was negatively associated (p < 0.001) with NHI (W1: r = −0.95; W2: r = −0.95) and NUtE (W1: r = −0.9; W2: r = −0.85) and was positively related (p < 0.001) to grain NA (W1: r = 0.98; W2: r = 0.97) and NT (W1: r = 0.84; W2: r = 0.94). Also, GY was negatively associated with NTR (r = −0.8) and positively correlated with post-anthesis NA (r = 0.8) at the 0.01 level only in W1.

Variables found to correlate with GY in Pearson’s correlation analysis were entered into multiple regression analyses with GY as a dependent variable (

Table 5. Multiple regression between grain yield and N parameters.

Item	Coefficient	p Value
Intercept	6939.42	0.00764 **
Grain NA	94.18	0.000086 ***
NT	−102.6	0.00917 **
NHI	15,009.78	0.06902 ns
Post-anthesis NRC	−22,236.7	0.00528 **
NTR	−20,058	0.00883 **
NUtE	150.52	0.000496 ***

Grain NA: Grain nitrogen accumulation; NT: Nitrogen translocation; NHI: Nitrogen harvest index; Post-anthesis NRC: Contribution of N translocation amount from the vegetative organ to the grain at post anthesis; NTR: Nitrogen translocation ratio; NUtE: Nitrogen utilization efficiency; **, *** indicate significance at the 0.01 and 0.001 probability levels, respectively. ns indicate not significant.

). The stepwise regression procedure iteratively adds and subtracts model predictors in order to identify the subset with the best performance (AIC = 192.58). Among the eight variables, NRC and post-anthesis NA were excluded from the table because it was eliminated in the stepwise regression process (Table 5). Despite not being significant in the regression procedure, the NHI was included in our model given its known importance in the formation of yield (Supplementary Table S3). The multiple R² and adjusted R² for the model were 0.9952 and 0.9934, respectively.

3.6. PLS-PM Analysis

Different pathways were built to explore the relationships between NA-related factors, NT-related factors, nitrogen utilization factors, and yield under different treatments by the PLS-PM analysis (Figure 6a). The goodness of fit was good for the PLS-PM model (GOF = 0.854).

Path analysis findings suggested that fertilizer combined with irrigation explained 87.6%, 97.8%, and 97.3% of the total variance in NAF, NUF, and NTF, respectively. Irrigation had a significant positive direct effect on NAF (β = 0.272), NUF (β = 0.385), and NTF (β = 0.262). Nitrogen significantly affected the NAF (β = −0.896) and NTF (β = −0.951) of Tibetan barley but had little effect on NUF (β = −0.134, p > 0.05). The PLS-PM indicated that fertilizer, irrigation, NAF, NUF, and NTF can explain 98.6% of the variation in GY under different treatments (Figure 6a, Supplementary Table S4), whereas NAF and fertilizer had no remarkable direct influences on GY (p > 0.05). NUF and NTF (β = 0.875 and −1.426, p < 0.001, respectively) positively and negatively affected GY, respectively. In addition, fertilizer and irrigation had indirect effects (0.761 and 0.089, respectively) on GY (Figure 6b). This indirect fertilization effect was mediated by NAF, NUF, and NTF; however, irrigation and N application showed total effects of 0.324 and 0.922, respectively (Figure 6b), indicating that nitrogen fertilizer was the most critical indicator of the regulation of the aforementioned factors. The loadings plot (Supplementary Figure S1) depicts the contribution of N parameters to each group’s classification.

4. Discussion

4.1. Effect of Water and Nitrogen Regimes on Aboveground Biomass and Grain Yield

Nitrogen can improve the quality of plants, which is one of the key nutrient elements that is involved in the carbon assimilation process and carbohydrate metabolism [43], acts as a signal for physiological metabolism [44], and even affects cellular osmoregulation [45]. A changed water input and nitrogen application might be related to variations in the yield and yield components of barley [25,46]. Several studies in barley showed that the 1000-grain weight and spike number revealed a trend of first increasing and then decreasing or gradually increasing with N application [47,48]. In the present study, however, increases in the yield were significant in N treatments, while increases in the 1000-grain weight and spike number were not consistent in all of them (Figure 3). This might be due to the fact that the accumulation of photosynthetic assimilates varies among the genotypes of barley, the cultivars of barley of the previous study differed from the cultivars of our study, and changes in the 1000-grain weight and panicle number exhibit different N sensitivities. In addition, the maximum nitrogen application in this experiment was 150 kg ha⁻¹, which may not reach the highest nitrogen application for increasing the yield. The interaction of the irrigation regimes and nitrogen fertilizer had a significant effect on the yield and yield components. Increasing nitrogen application or the irrigation amount can compensate for a decrease in crop yield caused by either reduction, but excessive nitrogen application or irrigation can reduce the compensatory effect [23,49,50]. In this regard, Albrizio et al. [51] showed in an experiment on barley under different irrigation regimes and nitrogen fertilizer treatment that the interaction of irrigation and nitrogen fertilizer did not have a significant effect on the biomass and GY. In our study, the main effect of nitrogen led to a significantly higher variation, as compared to the variation caused by irrigation (Figure 3c and Figure 6b). We suspect that this response could be attributed to the higher water-table depth on the near surface at later growth stages [52], which replenishes the soil moisture content; there was no discernible impact on yield even in W2 treatment, leading to a relatively weak influence of irrigation.

In the Tibet alpine and cold area, the aboveground biomass (AB) of Tibetan barley increased with increasing N input; but the increasing amplitude gradually narrowed with increasing N input. The AB of Tibetan barley with a high N application (N15) increased by 3.76–49.69% in W1 and 6.51–66.12% in W2, respectively, when compared with other N treatments (Figure 2a,b). Similar results were obtained in the studies of Ghasemi-Aghbolaghi and Sepaskhah [15], Naghdyzadegan Jahromi et al. [24], and Sajad et al. [53]. The optimized regime comprising an N input of 150 kg N ha⁻¹ with double irrigation can be used to increase the grain yield of Tibetan barley in the Tibet valley region.

4.2. Effects of N Application Rates and Irrigation Regimes on N Accumulation and Distribution

The N uptake and N accumulation occur dependently in each organ of barley, and their content varies depending on the plant growth stage. The N allocation proportion of barley varies across different organs, indicating variable N accumulation [54]. It can be seen from our study that the mean of NA and NDR are highest in the leaves, followed by stems, and lowest in the leaf sheaths alone at anthesis (Figure 4a,b). At maturity, NA significantly reduced among vegetative organs, and N accumulated and allocated at its highest concentration in grains (Figure 4c,d), which is consistent with the results of previous studies [55,56]. The NA and N distribution ratio of barley respond differently to fertilizer N input. Some scholars have concluded that increasing the N rate usually results in a higher NT, NA, and NTR in the barley organs [57,58]; however, others have suggested that nitrogen translocation is higher in fertilization treatments and N translocation efficiency is not affected by fertilization treatments [59]. In our studies, NA decreased with decreasing N fertilization with the greatest decline in N0 (mean decrease of 16.39–36.91% compared with N9), while the NDR showed a significant downward trend with increased N application (Figure 4d). This may be attributable to the intense solar radiation and long sunshine hours in the Tibetan Plateau, where Tibetan barley maintains high rates of photosynthesis during grain filling, exceeding the demands required to fill the grains to their maximum capacity. This can affect N translocation. Barati et al. [60] showed that a water deficit increased the N remobilization efficiency and the contributions of N in various vegetative organs to the grain N (except in rainfed conditions), and our findings also support this (Table 3). This present study suggested that NRC were lower in W2 (43.35% for mean among N treatments) than in W1 (46.43% for mean among N treatments) (Table 3), implying that a water deficit would weaken the availability of N fertilizer but enhance the remobilization of pre-stored N to the grain.

4.3. Effects of Irrigation and Nitrogen Fertilizer on Nitrogen Utilization

In the crop production system, NUE is used to define the relationship between the crops produced and the amount of nitrogen fertilizer involved in that production and to objectively reflect actual agricultural production [61]. It has been found that the effect of the water and nitrogen regime on NUE is related to the crop species and the supply levels of water or nitrogen, so NUE values are influenced by multiple factors [62,63]. The results of this study indicate that NUE was significantly decreased by 10.49–17.39% in W1 compared with W2 under the same N application, suggesting that increased irrigation frequency might improve NUE (Table 4). This is because the high soil moisture content can stimulate microbial and enzyme activities, thereby enhancing the nitrification efficiency mediated by nitrifying bacteria through the biochemical decomposition of N fertilizer. This, in turn, promotes the available nitrogen content in the topsoil layer surrounding the roots, ultimately enhancing Tibetan barley yield.

It has been suggested that the grain yield was much higher compared with the increase in total biomass, and therefore the NHI was increased [64,65]. When N application increases, as in this study, there is a decrease in NHI (Table 4). One possible reason is that the sensitivity to nitrogen supply levels varies across the barley organs, which affects the production of photoassimilates and the distribution of assimilates to the reproductive organs. NUtE was described for the crop uptake of N derived from fertilizer (partly from soil) and the efficiency of translating this uptake into yield [66,67]. Our study found that the NUtE of Tibetan barley showed a significant descending trend with an increase in N application (except for W2N15), while the change in NPFP was consistent (Table 4). The decrease in these two parameters as a consequence of N fertilization in our experiment is in agreement with the results reported by other authors [51,68]. Although a high N application amount could enhance the nitrogen accumulation capacity, the nutrients could not be fully absorbed, which inhibited N transport from the vegetative organs to seed to a certain extent, and resulted in smaller NUtE and NPFP values (Table 3 and Table 4). The NUtE and NPFP of treatments with W2 had increased by 13.22% and 5.9% on average when compared with W1, and this is consistent with earlier findings in wheat [69] and barley [70]. The decrease in both traits under water-stressed conditions confirms that high water availability increases the capability of plants to absorb and efficiently use N.

4.4. The Relationship between Grain Yield and Nitrogen-Related Components

Although the nutrient uptake and the water uptake are independent processes, the connections between N and W are adjacent and complex due to soil water availability affecting the migration, transformation, and accumulation processes of crops [71,72]. A higher nitrogen or protein content in the grain is related to higher nitrogen uptake by barley [73]. Multiple stepwise regression analysis further revealed that grain NA, NT, NTR, NHI, NUtE, and post-anthesis NRC are the most significant predictors of GY (Table 5). Using PLS-PM, we further found that NUF had a significant positive effect on GY, while NTF had a significant negative effect (Figure 6a). This was different from the previous study of Przulj et al. [74], who observed a positive correlation of grain yield with N translocation in some years but a negative correlation was found in others. One possible reason is that there are many trade-offs between the N transport processes and the grain yield and many factors (such as genetic and environmental) that can affect them [75,76]; thus, more research is needed to explore the relationships between N accumulation and Tibetan barley yield. In addition, nitrogen utilization and the translocation process also vary according to the amount of water and fertilizer. Our results suggested that the total effect of fertilizer is larger than that of irrigation in barley yield (Figure 6b). This may be due to the higher water table during the Tibetan barley growing season, which, to a certain extent, offsets the other positive effect of irrigation on the yield [77]. The change in the nitrogen translocation and utilization processes caused by nitrogen application was the main reason for the yield variation.

5. Conclusions

Increasing nitrogen application and irrigation frequency significantly or extremely significantly increased the grain yield, aboveground biomass, and N translocation in the grain and vegetative organs of Tibetan barley, respectively. An increase in the N applied possessed the ability to decrease NHI and NPFP together with a significant raising of NA and NUE, while the increase in the irrigation level elevated NHI, NUtE, and NPFP, with NA and NUE being significantly decreased.

Irrigation and NUF pose positive direct effects (β = 0.234 and 0.875, respectively), while NTF has a significant negative direct effect (β = −1.426) on the grain yield. Furthermore, N application (total effect = 0.922) presented a stronger total effect on GY compared with the irrigation treatment (total effect = 0.324). Under our experimental conditions, the study achieved the highest yield (6242.28 kg ha⁻¹) of Tibetan barley in high-N conditions at double irrigation, with consideration of the actual production process. In treatments with an N rate of 150 kg ha⁻¹, the yield of barley was maximum; however, this value may not reach the peak of the quadratic curve, and validation experiments of different nitrogen and water regimes are needed to verify this in the future. Theoretically, the above findings make a useful supplement to the study of irrigation water and nitrogen-use efficient intensive cereal systems, and they have guiding significance to the actual processes of barley production in Tibet. There is an urgent need for further studies on supplemental experiments and a reasonable increase in nitrogen application.