Coupling effects of phosphorus fertilization source and rate on growth and ion accumulation of common bean under salinity stress

Experimental procedures

A field experiment was performed in moderate saline soil at a private farm, El Qanater El Khayreya, Qalyubia Governorate, Egypt (31°09′07″E 30°11′35″N) during the 2019 season. daily temperatures ranged from 16.8 to 36.4 °C with an average of 25.4±4.8 °C, while night-time temperatures 9.2 ± 3.8 °C, with a minimum and a maximum of 2.2 and 18.2 °C. The average daily relative humidity was 68.7 ± 10.3% and ranged between 36.6% and 83.7%.

The common bean seeds (Phaseolus vulgaris L. cv. Nebraska) were selected, uniform and representative concerning size and color. The seeds were washed with distilled water, sterilized with 1.0% sodium hypochlorite solution for 2 min, washed with distilled water, and dried for 1 h at room temperature. Rhizobial inoculants have been used as peat slurry that contains 10⁷ Rhizobium per gram. The recommended seed rate for the common bean of 95 kg ha^–1 was used. Common bean seeds were sown in the field on the 10^th March 2019. Before the first irrigation, plants were thinned and one plant per hill was left. Before the cultivation, ammonium sulfate and potassium sulfate, at rates of 50 kg N ha^–1 and 125 kg K ha^–1 were supplemented according to the recommendations for the species. Before sowing, the physical and chemical soil characteristics were analysed with the methods described by Black et al. (1965) and Jackson (1973), and the results were presented in Table 1. Soil paste extract was used to determine electrical conductivity (EC). EC_e was 5.21 dS m^–1 that defines the soil as moderately saline (Hazelton & Murphy, 2016). Traditionally, saline soils are classified as having an EC_e value greater than 4 dS m^–1. Salinity is an EC_e-based rating system for soil. When the EC_e value varied between 4 and 8 dS m^–1, the soil is classified as moderately saline. As a result, salt has an effect on the yields of many crops (Hazelton & Murphy, 2016). Extractable soil phosphorus (P) content was 7.02 mg kg^–1 soil, which classed the soils as being low in P (Holford & Cullis, 1985).

The experiment was conducted using a split-plot arrangement in a randomized complete block design with three replicates. P source was the first factor, assigned to the main plots, the rate of P was a second factor, allocated to the sub-plots (split-plots). P sources were as follows: (1) single superphosphate (SP), and (2) urea phosphate (UP). P rates were following: 0.0, 17.5, 35.0, and 52.5 kg P ha^–1. Each replicate included eight plots, i.e., two P sources × four P rates, eight treatments in total as it was shown in Tables 2–7. P fertilizer was added to the soil during its preparation for cultivation. Single superphosphate (SP) contained 15.5% phosphorus pentoxide P₂O₅, 19.5% calcium, 11.5% sulphur, and 2.0 pH, while urea phosphate (UP) contained 17.5% total nitrogen, 44.0% P₂O₅, 2.0 pH, and 10% solution. P fertilizer was applied to the soil before sowing. Experimental plots consisted of 5 rows, 4 m long and 0.75 m wide, with the hills which were spaced 4–5 cm apart. Plots were irrigated using the reference crop evapotranspiration (ET0) values. All other agricultural practices were performed according to the recommendations of the Ministry of Agriculture and Land Reclamation, Egypt for common bean and referred literature (Abdelhamid et al., 2013).

Measurements

At 56 day after sowing (DAS), the following characteristics were analyzed on the samples consisted of ten plants per treatment: leaf area per plant, root dry weight per plant, shoot dry weight per plant, and total dry weight (DW) per plant. Total dry weight yield (t ha^–1) was determined by harvesting all plants from each experimental unit.

Chlorophylls and carotenoids were determined in the fresh leaves of 56 day-old plants. Pigments were extracted by homogenizing of 100 mg of fresh leaf tissue in pestle and mortar, in 80% acetone solution (10 mL) followed by centrifugation for 20 min at 3,000 g. The samples were storage at 4 °C for night, then optical density was measured at 480 nm for carotenoids, and 645 and 663 nm for chlorophyll a and b, respectively, following the method described by Arnon (1949).

DNA and RNA were extracted from leaves of common bean using a bench drill according to cetyltrimethylammonium bromide (CTAB) protocol (Doyle & Doyle, 1990) and (Ferreira et al., 2019) for DNA and RNA extractions, respectively. The grinded plant tissues were dissolved inside the plastic bags (20 × 10 × 0.01 cm of virgin, low-density polyethylene) at rotation of 2,500 rpm, until tissue is fully dissolved. The drill bit was cleaned with 70% alcohol for each sample change. The extraction buffer consisted of 2% CTAB, 100 mM, Tris-HCl, pH 8.0, 50 mM EDTA, pH 8.0; 1.4 M NaCl, 2% PVP-40; 1% Na sulfite, all reagents were add before use. Two mL of extraction buffer were added to plastic bags containing 300 mg of leaf tissue and macerated with Biorema drill (Agdia, Inc., Elkhart, IN, USA).

Dried roots and shoots of 56 day-old plants were used for determining the ion accumulation. The samples were dried at 70 °C, in an electric oven, till constant weight was accessed, and finally ground. Nutrients concentration was determined in the shoot and root of common bean plants. Macronutrients namely, nitrogen (N), P, K⁺, calcium (Ca²⁺), and magnesium (Mg²⁺), and micro-nutrients: iron (Fe³⁺), manganese (Mn²⁺), zinc (Zn²⁺), and copper (Cu²⁺), in addition to sodium (Na⁺) were determined. Total N was analyzed with the micro-Kjeldahl method. P was determined colorimetrically, after P extraction by sodium bicarbonate (Olsen, 1954), with stannous chloride-ammonium molybdate (King, 1951). K⁺ was measured with a flame photometer (ELE Flame Photometer, Leighton Buzzard, UK). Ca²⁺, Mg²⁺, Fe³⁺, Mn²⁺, Zn²⁺, Cu²⁺, and Na⁺ concentrations were determined by atomic absorption spectrophotometry (Chapman, 1965). N, P, K⁺, Ca²⁺, Mg²⁺, and Na⁺ were expressed as gram per kilogram (g kg⁻¹), while Fe³⁺, Mn²⁺, Zn²⁺, and Cu²⁺ were expressed as milligram per kilogram (mg kg⁻¹).

Statistical analysis of the data

The analysis of variance for a split-plot design was used to statistical elaboration of the experimental data (Gomez & Gomez, 1984). Before the analysis, the data were tested for the homogeneity of error variances with the Levene test (Levene, 1961), and for normality distribution with Shapiro and Wilk method (Shapiro & Wilk, 1965). Statistically significant differences between means were determined at the 5% probability level (P ≤ 0.05) using Tukey’s HSD (honestly significant difference) test. Correlation coefficients r were calculated to determine the relationship between biomass dry weight yield and each of the mineral concentrations. The GenStat 17^th Edition (VSN International Ltd., Hemel Hempstead, UK) software was used for statistical analysis.

Photosynthetic pigments of common bean

The interactive effect of phosphorus fertilization source and rate on total chlorophylls (TC) and carotenoids (Car) of common bean plants grown under salinity stress were shown in Table 2. Besides, the significance of the F test of analysis of variance was shown in Table 2. Urea Phosphate (UP) significantly increased TC and Car compared to superphosphate (SP) by 17.7% and 30.0%, respectively. The main effects of P rates were highly significant (P ≤ 0.01) and affected TC and Car. Application of P in a dose of 35.0 kg P ha^–1 compared to the control significantly increased TC and Car by 18.5% and 29.2%, respectively. TC increased gradually with increasing P rate and application of 35.0 kg P ha^–1 in SP and UP source caused the highest values by 20.9% and 16.7%, respectively compared to the control. Also, Car increased progressively with increasing P rate, the application of 35.0 kg P ha⁻¹ resulted in the greatest values in SP and UP source by 32.4% and 27.0%, respectively.

DNA and RNA of common bean

Table 3 shows the interactive effect of phosphorus fertilization source and rate on the nucleic acids (DNA and RNA) of common bean plants grown under salinity stress. Moreover, the significance of the F test of analysis of variance was shown in Table 3. Urea phosphate significantly increased RNA and DNA compared to SP by 9.8% and 10.8%, respectively. The main effects of P rates were highly significant (P ≤ 0.01) and affected RNA and DNA. Application of P with 35.0 kg P ha^–1 compared to control significantly increased RNA and DNA by 40.9% and 50.1%, respectively. RNA increased gradually with increasing P rate and application of 35.0 kg P ha^–1 in SP and UP source caused the highest values by 39.8% and 42.0%, respectively compared to the control. Also, DNA increased gradually with increasing P rate and application of 35.0 kg P ha^–1 resulted in greatest values in SP and UP source by 42.3% and 57.2%, respectively.

Ion concentration of common bean

The interactive effect of different sources and rates of phosphorus fertilization on macro-nutrients (N, P, K⁺, Ca²⁺, Mg²⁺), and Na⁺ of common bean plants grown under salt-affected soil (5.21 dS m^–1) were shown in Table 4. The significance of the F test of analysis of variance was shown also in Table 4. The main effects of P source were significant (P ≤ 0.05) in shoot N, shoot Ca²⁺, shoot Mg²⁺ and root N, highly significant (P ≤ 0.01) in shoot N, shoot K⁺, shoot Na⁺, and root P, while was non-significant in root K⁺, root Ca²⁺, root Mg²⁺, and root Na⁺. Application of UP increased all macro-nutrients in both shoot and roots, while reduced Na⁺ in shoot and root as well, compared to SP. The main effects of P rates were highly significant (P ≤ 0.01) and affected in all aforesaid traits as shown in Table 4. However, P reduced Na⁺ in shoots and roots and this reduction was in parallel with increasing P rate to the dose of 52.5 kg P ha^–1. Under the interactive effect of P source and P rate, application of P increased N, P, K⁺, and Mg²⁺ of shoots till 35.0 kg P ha^–1, whereas shoot Ca²⁺ increased gradually to the dose of 52.5 kg P ha^–1. In roots, application of P enhanced N, K⁺, Ca²⁺ to the dose of 35.0 kg P ha^–1, while 52.5 kg P ha^–1 decreased these nutrients compared to 35.0 kg P ha^–1. However, the P application increased P and Mg to 52.5 kg P ha^–1. Application of P boosted P concentration in shoots by 26.3% and 42.1% under SP, and UP, respectively when P in a dose of 35.0 kg P ha^–1 was applied. Moreover, P application resulted in a reduction in shoot Na⁺ by 27.6% and 29.3% in SP, and UP, respectively, while the reduction in roots Na⁺ was by 21.1% and 22.2% in SP, and UP, respectively. The response in Na⁺ concentration as P application reduced Na⁺ gradually with increasing P rate regardless of P source. The highest P rate resulted in the lowest Na⁺ concentration either in shoots or roots. The response due to P application may refer to P rate more than P source.

Table 5 shows the interactive effect of different sources and rates of phosphorus fertilization on micro-nutrients (Fe³⁺, Mn²⁺, Zn²⁺, Cu²⁺), in shoots and roots of common bean plants grown under salt-affected soil (5.21 dS m^–1). The significance of the F test of analysis of variance was shown also in Table 5. The main effects of P source were significant (P ≤ 0.05) in shoot Fe³⁺, shoot Mn²⁺, root Zn²⁺, and highly significant (P ≤ 0.01) in root Cu²⁺. P application in UP source increased Fe³⁺, Mn²⁺, Zn²⁺, and Cu²⁺ in shoot and Zn²⁺ in root compared to SP source (Table 5). The main effects of P rates were highly significant (P ≤ 0.01) and affected in all micro-nutrients in shoots and roots as shown in Table 5. Application of P increased Fe³⁺, Mn²⁺, Zn²⁺, and Cu²⁺ in shoot till 35.0 kg P ha^–1, while 52.5 kg P ha^–1 decreased these nutrients compared to 35.0 kg P ha^–1. However, P reduced Fe³⁺, Mn²⁺, and Cu²⁺ in roots and this reduction was in parallel with increasing P rate till 52.5 kg P ha^–1. Under the interactive effect of P source and P rate, Fe³⁺, Mn²⁺, Zn²⁺, and Cu²⁺, the F test was no significant in both shoots and roots. Fe³⁺ and Mn²⁺ in shoots and its increase was gradient with increasing P rate till 52.5 kg P ha^–1. However, the application of P in either SP or UP increased Zn²⁺ and Cu²⁺ till 35.0 kg P ha^–1, while 52.5 kg P ha^–1 decreased both Zn²⁺ and Cu²⁺ compared to 35.0 kg P ha^–1. Generally, P application increased micro-nutrients in roots of common bean and this reduction was in parallel with increasing P rate.

Interactive effect of different sources and rates of P fertilization on K⁺:Na⁺ and Ca²⁺:Na⁺ ratios in shoots and roots of common bean plants grown under salt-affected soil were shown in Table 6. The significance of the F test of analysis of variance was shown in Table 6. The main effects of P source were significant (P ≤ 0.05) in K⁺:Na⁺ and Ca²⁺:Na⁺ ratios in shoots and UP increased significantly these ratios in both shoots and roots compared to SP. The main effects of P rates were highly significant (P ≤ 0.01) and affected in all aforesaid traits as shown in Table 6. In general, the application of P increased K⁺:Na⁺ and Ca²⁺:Na⁺ ratios in both shoots and roots of common bean, and this increase coincided with increasing P rate to 52.5 kg P ha^–1. However, Ca²⁺:Na⁺ ratios in roots scored the highest value at 35.0 kg P ha^–1. Under the interactive effect of P source and P rate, only K⁺:Na⁺ ratio in roots, the F test was highly significant (P ≤ 0.05). In general, UP was significantly higher or without significant difference with SP in both K⁺:Na⁺ and Ca²⁺:Na⁺ ratios either in shoots or roots.

Growth and biomass yield of common bean

The growth parameters which included plant height, leaf area per plant, shoot dry weight (g), root dry weight (g), total dry weight per plant (g) and total dry weight per ha (t) were shown in Table 7. In addition, the significance of the F test of analysis of variance was shown in Table 7. The main effects of P source and rate were significant (P ≤ 0.05) on all mentioned traits. UP increased all traits compared to SP. The main effects of P rates were highly significant (P ≤ 0.01) and affected all aforesaid traits in Table 7. UP source of P increased total dry weight yield (t ha^–1) by 18.7% compared to SP. Increasing P rates resulted in increasing all mentioned traits gradually to the dose of 35.0 kg P ha^–1. For example, 35.0 kg P ha^–1, increased total dry weight yield (t ha^–1) by 93.8% compared to the control. Also, the interaction effect of P source and P rate was highly significant (P ≤ 0.05) for all mentioned traits except plant height. The interactive effects increased all growth traits, consequently increased the main target of total dry weight (t ha^–1). All traits increased gradually with increasing P rate and application 35.0 kg P ha^–1 either in SP or UP source increased significantly all traits. Application of 35.0 kg P ha^–1 either in SP or UP increased total dry weight (t ha^–1) by 84.9% and 101.9%, respectively compared to control plants. We found a significant association between leaf area per plant and total dry weight (t ha^–1), where the correlation co-efficient was highly significant (r = 0.994).

Correlation matrix

Pearson’s correlation coefficients (above diagonal) between plant dry weight (shoot dry weight and/or total dry weight per plant) and all nutrients measured in shoots of common bean plants grown with different sources and rates of phosphorus under salt-affected soil were shown in Table 8. There was a positive significant association between shoot dry weight and total dry weight with r = 1 (P ≤ 0.01). There was a positive significant correlation between total dry weight and N, P, Ca²⁺, Mg²⁺, Fe³⁺, Mn²⁺, Zn²⁺, and Cu²⁺, which were associated with one another. Besides, there was a negative association between either shoot dry weight or total dry weight and Na⁺ (P ≤ 0.01).

Table 9 showed Pearson’s correlation coefficients (below diagonal) between plant dry weight (root dry weight and total dry weight) and all nutrients measured in roots of common bean plants fertilized with different sources and rates of phosphorus and grown under salt-affected soil. There was a positive significant association between root dry weight and total dry weight with r = 1 (P ≤ 0.01). There was a positive significant correlation between total dry weight and each of P, K⁺, Ca²⁺, and Zn²⁺, which were associated with one another. Also, there was a negative association between total dry weight and each of N, Na⁺, Fe³⁺, Mn²⁺, and Cu²⁺ (P ≤ 0.01).

Response curve of biomass yield to phosphorus fertilizer rate

Linear and quadratic responses of common bean plants total dry weight (t ha^–1) grown under salinity stress to different rates of phosphorus (kg ha^–1) using different forms of phosphorus, i.e., superphosphate (A), urea phosphate (B), and at an average combined of superphosphate and urea phosphate (C) are shown in Fig. 1. Under superphosphate (A), with a P level increased of 1.0 kg ha^–1, the total dry weight was expected to increase by 14 kg ha^–1. The R² value, defined as the regression sum of squares divided by the total sum of squares, increased from 74.8% (linear) to 84.8% (quadratic), so 84.8% of the variation in total dry weight yields was explained by quadratic regression model. In the quadratic curve, total dry weight, equal 1.675 t ha^–1, was the maximum if P was applied at a level of 51.5 kg ha^–1, and the expected TDW should be 1.675 t ha^–1. Under urea phosphate (B), with a P level increase of 1.0 kg ha^–1, the total dry weight was expected to increase by 18 kg ha^–1. The mentioned value increased from 74.6% (linear) to 86.1% (quadratic). Thus, if P was applied at a level of 42.5 kg ha^–1, the maximum expected total dry weight would be 1.875 t ha^–1. Under average combined of superphosphate and urea phosphate (C), with a P level increased of 1.0 kg ha^–1, the total dry weight was expected to increase by 16 kg ha^–1. This has increased from 74.7% (linear) to 85.6% (quadratic). Hence, if P was applied at a level of 45.9 kg ha^–1, the maximum expected total dry weight would be 1.768 t ha^–1.