A field experiment was conducted to investigate the impact of replacement of mineral P with manure on soil P availability and microbial phosphorus transformation. Two treatments were implemented as part of the experiment: one (the control) consisting of conventional fertilization with mineral nitrogen (N, 90 kg N·ha⁻¹·year⁻¹), P (90 kg P₂O₅·ha⁻¹·year⁻¹), and potassium (K, 90 kg K₂O·ha⁻¹·year⁻¹) fertilizers and cattle manure (84 t ha⁻¹·year⁻¹), and the other (PoR treatment) consisting of Po fertilization with manure (134.4 t ha⁻¹·year⁻¹) in place of the mineral P fertilizer. The total input of P under the PoR treatment was same as under the control. Implementation of the experiment began in 2017 in a solar greenhouse located in Raoyang County, Hebei Province, China (38°15′N, 115°44′E). The soil was silt loam. Each treatment was replicated across three plots, and the plant system was a tomato and muskmelon rotation.

On 10 June 2020, surface soil (0–20 cm) was collected from each plot in accordance with procedures described in a previous study (Sun et al., 2015). The soil samples were sieved though a 2 mm sifter to mix thoroughly and remove impurities, such as plant roots and stones.

Soil pH, total carbon (TC), total nitrogen (TN), soil organic matter (SOM), ammonia (NH₄ ⁺–N), nitrate (NO₃ ⁻–N), available P (AP), and total P (TP) were measured as described in a previous study (Sun et al., 2022).

Soil total DNA was extracted from 0.5 g fresh soil using a FastDNA Spin Kit for Soil (MP Biomedicals, Santa Ana, CA, United States).

The soil bacterial community was characterized using high-throughput sequencing as described in a previous study (Sun et al., 2018). In brief, primer sets 515F/806 R targeting the V4 region of the bacterial 16S rRNA gene were used for polymerase chain reaction (PCR) analysis. PCRs were performed in a 50-µL mixture containing 25 µL PCR premix (TaKaRa Ex TaqR), 1 µL forward primer (10 µM), 1 µL reverse primer (10 µM), 1 µL DNA template (20 ng), and 22 µL PCR-grade water under the following conditions: initial denaturation at 94°C for 10 min; 30 cycles in a series of denaturation at 94°C for 1 min, annealing at 50°C for 1 min, and extension at 72°C for 1 min; and a final extension at 72°C for 10 min. After quality checks and purification, the PCR products were sequenced using an Illumina HiSeq 2,500 system.

Soil bacterial functional profiles with respect to P cycling were predicted using PICRUSt2 (Phylogenetic Investigation of Communities by Reconstruction of Unobserved States) (Douglas et al., 2020). Genes involved in P transformation were extracted to reveal the potential capacity for microbial P transformation under each of the treatments applied.

Soil alkaline phosphatase activity was measured using a Soil Alkaline Phosphatase (S-AKP/ALP) Activity Assay Kit (Beijing Solarbio Science & Technology Co., Ltd. China), and expressed in the form of nM p-PNP (p-nitrophenyl phosphate) h⁻¹g⁻¹ dry soil. The abundance of the phoD gene, which encodes for alkaline phosphatase, was determined using real-time fluorescent quantitative PCR (qPCR), which was conducted using primer sets phoD-F733/phoD-R1083 (5′-TGGGAYGATCAYGARGT-3′/5′-CTGSGCSAKSACRTTCCA-3′) under the following amplification conditions: 10 min at 95°C, followed by 40 cycles of denaturation at 95°C for 15 s and annealing at 55°C for 1 min. After amplification, melting curve analysis and gel electrophoresis were performed to measure the specificity of the reaction (Chen X. et al., 2019).

The microbial dissolving potential for calcium phosphate was determined for each treatment using an incubation method, in accordance with procedures described in our previous study (Sun et al., 2022). Briefly, 1 g soil was added to 90 ml sterilized water and mixed thoroughly. Next, 1 ml of the mixture was added to 50 ml sterile PVK liquid medium and incubated at 28°C (180 rpm) for 5 days. A blank control with 1 ml sterilized water was included. Dissolved P in the culture was measured at 0,12, 24, 48, 72, 96, and 120 h.

VSEARCH (version 2.21.1) (Rognes et al., 2016) was used for bioinformatic analysis of the high-throughput sequencing data following the protocols described in previous work (Liang et al., 2020; Sun et al., 2022). The paired-end reads were first merged, and low-quality and chimeric reads were then removed. Subsequently, the clean reads were further denoised and zOTUs (zero-radius operational taxonomic units) were generated using the UNOISE algorithm (version 3). The taxonomic details of each zOTU were determined using the RDP Classifier tool based on the SILVA rRNA database (version 138). After removal of zOTUs not annotated as bacteria, the zOTU tables were subsampled to extract 48,000 reads per sample for statistical analysis.

Statistical analysis was conducted and figures generated using R (version 4.0.2) as described in our previous study (Sun et al., 2022). The Kruskal–Wallis rank sum test was used to test for significant differences between treatments for each variable. Principal coordinate analysis (PCoA) was performed based on Bray–Curtis distance using the “vegan” library.

PoR treatment significantly lowed soil pH, but had no significant impact on soil contents in terms of total carbon (TC), total nitrogen (TN), soil organic matter (SOM), ammonia nitrogen (NH₄ ⁺–N), or nitrate nitrogen (NO₃ ⁻–N) (Table 1). Levels of soil total P (TP) were similar under the two treatments (Table 1), while PoR treatment significantly increased the level of soil available P (AP), which was 35.51% higher in samples having undergone PoR treatment compared to the control (Table 1).

Tomato and muskmelon yields did not differ significantly between the control and PoR plots (Table 1), indicating that PoR treatment has a similar productivity to that of conventional fertilization.

The Shannon index was calculated as a measure of variation in bacterial diversity. The results showed that there was no significant difference in bacterial α-diversity between samples having undergone control and PoR treatment (Figure 1A), indicating that PoR treatment had little impact on bacterial α-diversity. However, soil bacterial community composition was impacted by PoR treatment. Proteobacteria was the most dominant phylum in the control soil, but it featured in a significantly lower proportion in the PoR soil. Bacteroidota was also diluted under PoR treatment, while Firmicutes and Gemmatimonadota were enriched (Figure 1B). This variation in bacterial community structure is further illustrated by the corresponding 2D PCoA plot (Figure 1C), which shows the clear separation in bacterial community between control and PoR samples. ANOSIM (Analysis of Similarities) also confirmed that there was a significant difference between control and PoR samples in terms of bacterial community (p < 0.05).

Through functional prediction (using PICRUSt2), 38 genes involved in P transformation were detected (Park et al., 2022). PoR treatment significantly decreased the abundance of the phoU and pst genes, while it significantly increased the abundance of the phoR and phnA genes and genes coding for alkaline phosphatase and C-P lyase (Figure 2A). Considering genes in terms of functional groups, those involved in Pi solubilization and Po mineralization were significantly increased under PoR treatment (Figure 2B). In contrast, the abundance of genes involved in P uptake and transport was significantly decreased under PoR treatment (Figure 2C).

Soil alkaline phosphatase activity was significantly higher under PoR treatment than under control treatment (Figure 3A). A similar pattern was observed in the abundance of the phoD gene, which was significantly increased by PoR treatment (Figure 3B).

Taking calcium phosphate as the sole P resource, the incubation experiment indicated that, over the entire incubation period, the levels of dissolved P in the culture liquid followed an approximate bell curve, peaking at 48 h (Figure 3C). At the 48 h time point, the dissolved P content of the culture with PoR-sourced microbes was significantly lower than that of the culture with control-sourced microbes, showing that the potential P solubilization of the microbial community was significantly lower under PoR treatment than under the control treatment.