Short-chain soluble polyphosphate fertilizers increased soil P availability and mobility by reducing P fixation in two contrasting calcareous soils

Soils sampling and soil description

Experiments were conducted using two calcareous soils (loam and clay soils). According to FAO/WRB soil taxonomy, both were classified as Calcisol Fluvisols. Loam soil was taken from experimental station of Shihezi University (44°18′ N, 86°02′ E) and clay soil was collected from 147 State Farm (44°37′ N, 86°10′ E) in Shihezi region. Soil samples were air dried, stones and small visible plant residues were manually removed. All samples were ground to pass through a 2 mm sieve prior to measurement of soil properties. The selected soil physical and chemical properties are presented in (Table 1).

Different P sources description

Three types of P sources including (i) mono-ammonium phosphate (powder MAP, NH₄H₂PO₄) with purity of 99% and density of 1.803 g cm⁻³ and 26.9% of P (Shengao Chemical Reagent Co., Ltd., Shanghai, China); (ii) phosphoric acid (fluid PA, H₃PO₄) with purity of 85% and density of 1.685 g cm⁻³ and 27.1% of P (Fuyu Special Chemicals Co., Ltd., Dongying, China), and (iii) ammonium polyphosphate (fluid APP, H₆P₄O₁₃) with purity of 85% and density of 2.1 g cm⁻³ and 31.3% of P (Aladdin Industrial Corporation, Mooresville, NC, USA) were employed in this study.

Experiment I soil cylinder experiment

Soil cylinder experiment was carried out in a transparency Plexi glass cylinders (350 mm height, 64 mm internal diameter). Five holes (2 mm diameter) were drilled at bottom of the cylinder to maintain aerobic condition. A 10-mm-thick layer of fine sand was placed at the bottom of the cylinder and placed two pieces of filter paper on the surface of the sand layer. A total of 1,302 g loam soil and 1,331 g clay soil were placed into each cylinder. Another two pieces of filter paper was placed on the top of the cylinder. Soil bulk densities of 1.35 g cm⁻³ and 1.38 g cm⁻³ were designed for loam and clay soil, respectively. Three P types (i.e., powder MAP, fluid PA, and fluid APP) were applied to the each cylinder following the methods of: (i) Single application, the total amount of P fertilizer was uniformly mixed into the upper 20 mm of soil layer and watered four times over the 4-week-long incubation period; (ii) split repeated application, the fertilizer was initially dissolved with 50 mL milli-Q water in a 250 mL beaker individually except MAP which was applied directly and then fertilizer solution was injected into soil at the center site of the cylinder using a syringe, P fertilizer was four times evenly added at equivalent ratio during entire 28-days period. The irrigation water was poured into a Markov bottle, and hanged the bottle 1.5 m above each cylinder (Fig. 1). The flow rate of the emitter was controlled at 6 drops per second by connecting each emitter to individual Markov bottle. The total water column irrigated were 401.6 mL and 515.8 mL cylinder⁻¹ for loam and clay soils, respectively. P application rates were 58.5 mg P cylinder⁻¹ for loam and 59.8 mg P cylinder⁻¹ for clay soils, respectively. The P fertilizer application rate was two to three times higher than in normal farmland (Du et al., 2013; Wang & Chu, 2015). Each treatment replicated three times giving a total of 42 cylinders (containing 6 CK treatments with no fertilizer either in loam or clay soil, 18 cylinders for single application of MAP, PA and APP either in loam or clay soil and another 18 cylinders for split application of MAP, PA and APP either in loam or clay soil). All cylinders were covered with a piece of parafilm, and incubated at room temperature during incubation period.

After 28-day incubation, all cylinders were placed into a −80 °C freezer for 12 h to keep soil column hard enough. The frozen cylinders were cut into the given thick soil slices by placing the cylinder to a high-speed spindle of the lathe with a sharp blade at millimeter-scale (Type: CW6163C, made in Dalian, China) (Supplementary Figures). Each cylinder was vertically cut into 20 slices with a 5-mm thickness for each slice, then another 10 slices with a 20-mm-thickness were cut. Each slice was immediately placed in a plastic bag and transferred to fridge.

All collected soil slices were used to determine soil moisture, soil WE-P (water extracted phosphorus) and Olsen-P. Briefly, to determine WE-P, an aliquot 2 g soil was taken in a 100 mL centrifuge tube with 50 mL triple de-ionized (TDI) water. Shaking the soil sample for 1 h at 25 °C, following centrifugation for 15 min at 900 rpm, decant supernatant was separated and WE-P was measured using malachite green colorimetric method at an absorbance of 610 nm (Masson et al., 2001). Olsen-P was determined in 0.5 M NaHCO₃ extracts (pH 8.5), using the molybdenum blue method (Olsen et al., 1954).

An exponential decay equation model was employed to describe the P downward movement in soil cylinder.

(1) $$\mathrm{Y}=({\mathrm{Y}}_0-\mathrm{P}\mathrm{l}\mathrm{a}\mathrm{t}\mathrm{e}\mathrm{a}\mathrm{u})\times \exp (-\mathrm{K}\times \mathrm{X})+\mathrm{P}\mathrm{l}\mathrm{a}\mathrm{t}\mathrm{e}\mathrm{a}\mathrm{u}$$

where X represents the depth of the soil column expressed as mm. Y indicates the content of phosphorus in soil column in mg/kg. Y₀ is the Y value when X (depth) is zero. It is expressed in the same units as Y; Plateau is the Y value at infinite depths, expressed in the same units as Y; K is the rate constant, expressed in reciprocal of the X axis depth units (mm⁻¹); Half-depth is in the depth units of the X axis. It is computed as ln(2)/K, which means the movement depth when the concentration of P drop down to the half.

Experiment II soil incubation experiment

To explore the influences of different P types on phosphorus transformation, a 560-day microcosm incubation experiment was set up. Herein only loam soil was used as described in the previous section. In brief, a total of 253 g soil was thoroughly mixed with different P fertilizers (MAP, PA and APP) at the addition rate of 40.3 mg P pot⁻¹. The mixed soil was placed in a plastic pot (5.7 cm in height, 6.6 cm in diameter) with soil bulk density of 1.30 g cm⁻³, all pots were incubated under room condition for 560 days. Tap water was added periodically per week to keep water holding capacity of 60% (WHC). Each treatment was replicated four times giving a total of 16 pots (including 4 CK treatments having no fertilizer).

After 65-, 140-, 230-, 320-, 560-day incubation, soil samples were collected from each pot. All soil samples were air dried and grounded for determination of different soil inorganic P fractions according to Guppy et al. (2000) and modified by Aulakh et al. (2003). Briefly, soil P sequential extraction followed: (1) resin-P: an aliquot 0.500 g soil sample was mixed with 30 mL deionized water in a 500 mL centrifuge tube, two Cl-saturated anion exchange resin strips were added to remove resin-P, the sample was shaken for 16 h. After shaking the two Cl-saturated anion exchange resin strips were taken in a 100 mL centrifuge tube and 30 mL 0.7 moL/L NaCl was added subsequently, shaking the sample for 1 h. (2) NaHCO₃-P: the soil residue in the previous step was treated with 30 mL 0.5 moL/L NaHCO₃ to remove NaHCO₃-P, shaking the sample for stay overnight, and centrifugation for 30 min. (3) NaOH-P: the soil residue was treated with 30 mL 0.1 M NaOH and 1 mL of 4 M NaCl to remove NaOH-P, shaking the sample for 16 h, and subsequently centrifugation for 30 min. (4) HCl-P: the soil residue was treated with 30 mL 1moL/L HCl, shaking stay over. (5) Residue-P: aliquot 0.5 g anhydrous MgSO₄ and 5 mL of H₂SO₄: HClO₄ (20:1) acid solution to remove residue-P. Inorganic P concentration in each of the extracts was determined using malachite green colorimetric method at an absorbance of 610 nm (Masson et al., 2001). The total P content of soil samples was determined using molybdenum blue spectrophotometry method after digestion with HClO₄–H₂SO₄ (Masson et al., 2001).

The proportion of applied P fertilizer transformed to different P fractions in soils was figured out referring to Aulakh et al. (2003). Individual parameters were calculated as

(2) $$\begin{array}{c}\text{Amount of fertilizer P recovered as Pi in soil}({\text{kg P ha}}^{-1})\hfill \\ =[\text{amount of Pi in P − fertilized treatment}({\text{kg P ha}}^{-1})]\hfill \\ -[\text{amount of Pi in no − P control}({\text{kg P ha}}^{-1})]\hfill \end{array}$$

(3) $$\begin{array}{c}\text{Total P recovered in soil}({\text{kg P ha}}^{-1})\hfill \\ =[\text{sum of P recovered in different fractions in the soils}({\text{kg P ha}}^{-1})\hfill \\ \times (2.325\times 106{\text{kg ha}}^{-1})\hfill \end{array}$$

where 2.325 × 10⁶ kg P ha⁻¹ is soil mass of 0–150 mm soil layer computed using field bulk density of 1.55 g cm⁻³.

(4) $$\begin{array}{c}\text{Total fertilizer P recovered in soil}({\text{kg P ha}}^{-1})\hfill \\ =[\text{total P recovered in P − fertilized treatment}({\text{kg P ha}}^{-1})]\hfill \\ -[\text{total P recovered in no − P control}({\text{kg P ha}}^{-1})]\hfill \end{array}$$

(5) $$\begin{array}{c}\text{Fertilizer P present in any specific soil fraction}(\mathrm{\%})\hfill \\ =\frac{\left[\left(\mathrm{P}\mathrm{i}\mathrm{i}\mathrm{n}\mathrm{P}-\mathrm{f}\mathrm{e}\mathrm{r}\mathrm{t}\mathrm{i}\mathrm{l}\mathrm{i}\mathrm{z}\mathrm{e}\mathrm{d}\mathrm{t}\mathrm{r}\mathrm{e}\mathrm{a}\mathrm{t}\mathrm{m}\mathrm{e}\mathrm{n}\mathrm{t}\right)-\left(\mathrm{P}\mathrm{i}\mathrm{i}\mathrm{n}\mathrm{n}\mathrm{o}-\mathrm{P}\mathrm{c}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}\right)\right]\left(\mathrm{k}\mathrm{g}\mathrm{P}\mathrm{h}{\mathrm{a}}^{-1}\right)}{\mathrm{T}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l}\mathrm{f}\mathrm{e}\mathrm{r}\mathrm{t}\mathrm{i}\mathrm{l}\mathrm{i}\mathrm{z}\mathrm{e}\mathrm{r}\mathrm{P}\mathrm{r}\mathrm{e}\mathrm{c}\mathrm{o}\mathrm{v}\mathrm{e}\mathrm{r}\mathrm{e}\mathrm{d}\mathrm{i}\mathrm{n}\mathrm{s}\mathrm{o}\mathrm{i}\mathrm{l}\left(\mathrm{k}\mathrm{g}\mathrm{P}\mathrm{h}{\mathrm{a}}^{-1}\right)}\hfill \end{array}$$

Statistical analyses

Data were analyzed using the SPSS 11.5 statistical program (SPSS Inc., Chicago, IL, USA) with two-way ANOVA at a significance level of p < 0.05. A Duncan multiple range test was carried out to test the significant differences between different treatments. Microsoft Excel 2003 and Graphpad Prism 5.0 software (GraphPad Software, Inc., San Diego, CA, USA) were used for data processing and images making. All results in figures and tables were presented as mean of three or four replicates with a standard deviation (SD), a statistical significance level of p < 0.05 was used for all analyses.

Different P application methods on P mobility and availability

Soil Olsen-P decreased with soil depth (0–100 mm) increasing across all P sources (Fig. 2). Compared to the single application treatment, the significant higher values of Olsen-P were always appeared in the repeated PA and APP application at 0–40 mm depth (p < 0.05). The vertical movement of Olsen-P reached the same value with CK (baseline) were 80 mm and 95 mm, respectively, in the PA and APP repeated application treatment, while those were merely 42 mm and 50 mm in the single application treatment in loam soil. Similar tendency was also observed in clay soil, indicating that repeated P application significantly promoted P vertical migration.

An exponential decay model was employed to describe Olsen-P movement in soil cylinder (Table 2). Rate constant (K) and half-depth were key parameters to reflect the effects of different application methods on P movement. The steeper the response curve (higher K value) and lower half-depth, the shorter distance the added-P moved, and vice versa. In clay soil, K value (Rate constant) in the repeated P application was lower than that in the single application treatment across all treatments. The values of half-depth increased by 39.7%, 49.8%, 158.6% in the repeated application treatment, respectively, relative to the single MAP, PA and APP application treatments.

Different P types on P mobility and availability

Compared to the treatments of MAP and PA application, soil P mobility was significantly increased by APP application (Fig. 3). The distance of WE-P downward movement in the single P application treatment in loam soil followed order of APP (80 mm) > MAP (60 mm) > PA (35 mm). Similarly, the distance of WE-P downward movement in clay soil followed the order of APP (83 mm) > MAP (62 mm) > PA (55 mm). Moreover, the exponential decay equation model showed that K value in loam soil followed the order of APP (0.054 mm⁻¹) < MAP (0.065 mm⁻¹) < PA (0.115 mm⁻¹), and the half-depth followed the order of APP (12.70 mm) > MAP (10.66 mm) > PA (6.02 mm) when P fertilizer applied as single application method (Table 3). Likewise, K value in the repeated application treatment (loam soil) followed order of APP (0.039 mm⁻¹) < MAP (0.044 mm⁻¹) < PA (0.062 mm⁻¹), the half-depth followed the sequence of APP (17.92 mm) > MAP (15.88 mm) > PA (11.24 mm). The half-depth in the APP treatment were 4.84% and 31.5%, respectively, greater than the MAP and PA treated clay soils. These results indicate that APP fertilization significantly improved P mobility and availability.

Inorganic P transformation

During whole incubation time, resin-P and NaHCO₃-P (labile P) notably decreased, while NaOH-P (moderately labile P) and HCl-P (recalcitrant soil P) showed an increasing tendency (Fig. 4). Furthermore, soil resin-P, NaHCO₃-P and NaOH-P in the APP added treatment were significantly higher than in the PA and MAP treatments. For instance, at the end of incubation (560 d), resin-P (5.33 mg kg⁻¹) and NaHCO₃-P (73.7 mg kg⁻¹) in the APP addition treatment increased by 51.6% and 191% respectively, relative to the MAP and PA application treatments (on average). In addition, When P fertilizer was added as polyphosphate, NaOH-P was increased by 24.6% and 39.8%, respectively, relative to the MAP and PA treatments. The influence of the different P types on non-labile P distribution proportion followed order of MAP (87.8%) > PA (73.7%) > APP (54.5%) (Table 4), indicating that polyphosphate addition significantly retarded the transformation of the added-P from labile to non-labile P forms, thus reduced P fixation.