PhoD Harboring Microbial Community and Alkaline Phosphatase as Affected by Long Term Fertilization Regimes on a Calcareous Soil

: Organic phosphorus (Po) may play a vital role in phosphorus availability via its mineralization by alkaline phosphatase (ALP), being encoded by phoD gene, in calcareous soil. Understanding the effects of long-term fertilization on the community of phoD harboring bacteria and the related alteration of the P availability owing to the changes in ALP secretion may offer a chance to elucidate the Po contribution to soil available P. Based on a long-term experiment, we analyzed the phoD gene harboring microbial diversity, abundance and composition, ALP and Po forms, and their relationship. The treatments involved were control without any fertilizers (CK), synthetic nitrogen and potassium (NK), synthetic nitrogen, phosphorus and potassium (NPK), NPK and crop stalk return (SNPK), and NPK plus organic manure (MNPK). Fertilization increased the abundance and diversity of phoD gene harboring microbial over CK. Those receiving NPK and NPK treatments integrated with organic supplements signiﬁcantly improved the relative abundance of Proteobacteria but decreased Gemma-timonadetes at the phylum level, while all fertilized treatments appreciably increased the relative abundance of Lysobacter but decreased that of Gemmatirosa and Aﬁpia , at the genus level. SNPK and MNPK treatments noticeably increased the relative abundance of Methylobacter but reduced Pseudomonas and Streptomyces relative to those receiving synthetic fertilizer treatments. Long-term fertilization markedly raised ALP activity, which was signiﬁcantly and positively correlated with the relative abundance of the phylum Proteobacteria as represented by the genera Methylobacterium and Lysobacter . ALP was closely associated with moderately labile Po, followed by enzyme P, recalcitrant Po, and labile Po. The changes in phoD bacteria and ALP were mainly driven by soil organic carbon, Olsen P and pH. We concluded that the long-term fertilization, especially the addition of organic supplements, profoundly modiﬁed the soil properties and subsequently changed the diversity and relative abundance of phoD gene harboring bacteria, which promoted the activity of ALP, and thus the mineralization of various forms of Po (mainly moderately labile Po) to enhance the P availability.


Introduction
Phosphorus exists in soils in either inorganic or organic forms.The former is the major form in mineral soils with low organic matter content such as calcareous soils, where it can reach 60% to 80% of the total phosphorus [1,2], and even higher than that in some acidic soils [3].Organic phosphorus, on the other hand, has long been believed less important because of its being low in both content and bioavailability [4].Therefore, a great deal of research has focused on the transformation of inorganic phosphorus and its bioavailability [5,6].However, with extensive return of crop residues, amendment of organic manure and synthetic fertilizers, the level of soil organic matter increased dramatically [7][8][9], the content of soil organic phosphorus also increased accordingly [10,11].As a result, much attention has been paid to the availability of soil organic phosphorus, which can be mineralized to bioavailable inorganic forms by the phosphatases secreted by microorganisms [12].In calcareous soils, the major enzyme responsible for the conversion of organic phosphorus is alkaline phosphatase (ALP) [12,13], which is mainly secreted by bacteria in soils [14], accounting for about one-half of the total enzyme activity [15].So far, three homologous genes (phoA, phoX and phoD) encoding ALP have been identified [16,17], of which phoD gene is the most widely distributed [18], and is reported more sensitive to the alternation of fertilization regimes [6,16,18].
Fertilization significantly altered the community composition and diversity of phoD gene harboring bacteria [19][20][21][22][23], this is largely attributed to the significant effects of fertilizers' management on soil properties, which greatly modified the bacterial diversity.On acidic soils, pH is a key factor affecting phoD gene harboring bacterial diversity [18,24], which may increase with increasing soil pH value.For example, on a sandy loam pasture soil (humic gleysol, pH 5.5) in Ireland, Tan et al. (2013) [16] reported that long-term application of calcium superphosphate increased soil pH and the phoD gene harboring bacterial diversity (Chao 1, ACE, and Shannon indices).A similar observation has also been made in acidic Hapli-Udic Luvisol in Shenyang, China, with application of synthetic diammonium phosphate fertilizer (Chao 1 index) [21].However, a decline in soil pH could correspondingly reduce the phoD gene harboring bacterial diversity.For instance, Chen et al. (2017) [20] documented that long-term application of synthetic fertilizers of nitrogen, phosphate and potassium (NPK) in red soil significantly reduced the soil pH value from 5.53 to 4.30, leading to a significant decrease in diversity of the phoD gene harboring bacterial (Chao 1 index).In some cases, the phoD gene harboring bacterial diversity may remain constant where pH stays unaffected by fertilization even in acidic soils.In red soil, Chen et al. (2017) [20] reported the integrated application of organic manure and synthetic fertilizers (MNPK) did not significantly change the soil pH and the diversity of the phoD gene harboring bacteria.On alkaline soils, the phoD gene harboring bacterial diversity may mainly be governed by SOC content [22].In a lime concretion black soil and a Karst (Leptosols) soil with pH greater than 7, the combination of synthetic phosphate fertilizer with either mineral N or organic manure significantly increased the diversity of the phoD gene harboring bacteria expressed in Shannon and Chao1 indices or OTUs richness; the authors ascribed the observation to the increase in SOC content as a result of fertilization [6,22].The phoD gene harboring bacterial diversity was also observed to remain unchanged on an alluvial loam soil where fertilization showed no impact on SOC content [23].However, in a rainfed winter wheat-summer fallow cropping system in a calcareous Anthrosol on Loess Plateau, China, Liu et al. (2020) [25] reported that various application rates of synthetic phosphate fertilizer in combination with nitrogen substantially enhanced the SOC level but showed no such effects on the phoD gene harboring bacterial diversity, expressed as Shannon and Chao 1 indices, Po content and ALP activity, in the rhizosphere soils.On the same soil, Li et al. (1997) [26] found that the increase in SOC level as a result of root exudates of winter wheat in the rhizosphere soil could also lead to an increase in microbial activity and ALP activity, which were significantly higher than that in the corresponding non-rhizosphere (bulk) soil.Significant positive correlations between SOC and Po content and ALP activity in this soil were also documented by many other researchers [27,28].Long-term application of fertilizers, either in synthetic forms or its combination with organic supplements, could markedly enhance SOC and ALP activity [28,29].Obviously, these results are not consistent with that of Liu et al. (2020) [25].In addition, the results of a 12-year-trial in the same soil showed that the content of organic phosphorus in soil treated with synthetic NPK fertilizers was significantly greater than that in control soil receiving no fertilizers, while integrated application of NPK with organic supplements (dairy manure) considerably enhanced either SOC or Po contents, but not in a synchronous way [9].The authors speculated that the accumulated Po with amendment of organic manure might be largely in recalcitrant forms.On the one hand, most phosphorus contained in organic manure applied is in inorganic forms [30] due to the strong mineralization of Po before its incorporation into soils.On the other hand, organic manure amendment significantly raised ALP [26,31,32]; this way, most of the labile Po contained in organic manure was mineralized by ALP into available inorganic forms, the Po forms of those with lower activity or which were recalcitrant were left over and gradually accumulated in the soils receiving it [33].However, the contribution of different forms of organic P to soil available P through mineralization is somewhat inconsistent.For example, Luo et al. (2017) [6] reported that the labile Po was the main form mineralized by ALP in lime concretion black soil, which has been subjected to long term treatments of synthetic fertilizers, organic manure and their combination [6]; Yin et al. (2001) [34] studied the relationship between the changes of soil Po components and the P absorption of ryegrass after applying pig manure and synthetic phosphate fertilizer on a calcareous yellow fluvo-aquic soil; the results showed that the contribution of various forms of Po to the total phosphorus uptake by ryegrass followed an order of: labile Po > moderately labile Po > moderately stable Po.The main forms of Po mineralized by ALP were labile Po, moderately labile Po, and moderately stable Po in either the rhizosphere soil or bulk soil of wheat after applying different amounts of organic manure and synthetic fertilizers on the Anthrosol [26]; Nevertheless, Wang et al. (1997) [27] reported that soil available P significantly and positively related to the contents of labile Po and moderately labile Po, while moderately stable Po and recalcitrant Po did not contribute to available P. These varying observations, some of which even obtained from the same type of soil, suggested that the forms of Po and their contributions remain to be examined further.Therefore, we characterized the composition and structure of phoD bacterial community by MiSeq sequencing method based on a long-term fertilization experiment in calcareous Anthrosol.The content of various forms of organic P and ALP were also determined.The purposes were: (1) to clarify the effects of different long-term fertilization on the community composition and diversity of the phoD gene harboring bacteria and the activity of ALP in calcareous soil; (2) to determine the accumulation of main forms of organic P and its relationship with the community composition and diversity of the phoD gene harboring bacteria.The findings may provide new insights into improving soil P availability by optimizing fertilization management.

Site Description and Sample Collection
The long-term experiment was initiated in October 1990 at the "Chinese National Soil Fertility and Fertilizer Efficiency Monitoring Base of Loessial Soil" (34 • 17 51 N, 108 • 00 48 E, with an altitude of 524.7 m a.s.l.), located in the southern verge of Loess Plateau, Yangling, Shaanxi Province of P. R. China.The experimental site has a mean annual temperature of 13.0 • C and mean annual precipitation of ca.550 mm, which mainly falls from June to September.The soil at the site is a silt clay loam (clay 16%, silt 52% and sand 32%; Anthrosols with a terric horizon derived from manure and loess material, WRB, 2014).At the beginning of the experiment conducted, the chemical properties of the plough layer soil (0-20 cm) were: 7.44 g kg −1 organic C, 0.93 g kg −1 total N, 9.57 mg kg −1 Olsen P, 191 mg kg −1 exchangeable K, 91.66 g kg −1 carbonates and had a pH of 8.62 (1:1 soil/water ratio).A total of 13 treatments were established in the experiment, with each plot measuring 14 × 14 m, each treatment replicated only once for practical reasons; the detailed information has been described by Yang et al. (2012) [35].
Winter wheat (Triticum aestivum L.) and summer maize (Zea mays L.) double crops each year have been practiced, and five out of the thirteen nutrient management treatments were involved in this study: (1) Control, no fertilizer or manure input (hereafter CK); (2) NK, receiving only synthetic nitrogen (N) and potassium (K); (3) NPK, receiving synthetic nitrogen (N) phosphorus (P) and potassium (K); (4) SNPK, receiving synthetic NPK fertilizers and crop stalk return (S); (5) MNPK, incorporation of NPK plus organic manure.The N, P, and K (element) from treatment (2) to (4) were supplied with synthetic fertilizers of urea (46% N), single super phosphate (12-16% P 2 O 5 ), and potassium sulphate (50% K 2 O), respectively, at rates of 165.0, 57.6, 68.5 kg/hm 2 , respectively, for wheat, and 187.5, 24.6, 77.8 kg/hm 2 , respectively, for maize.SNPK treatment received straw return once a year before winter wheat planting, with 4500 kg/hm 2 wheat straw each year from 1990 through 1998.Since 1999, it has received the aboveground parts of maize stalks harvested from the plot in the preceding season, with a mean annual weight of 4392 kg/hm 2 (ranging from 2630 to 5990 kg/hm 2 ).The added straw was chopped into small pieces with lengths of ca. 3 cm and incorporated into the plough layer soil (0-20 cm).While the MNPK treatment received the same amount of synthetic N, P and K as that of NPK treatment in both wheat and maize seasons except that 70% of the N was substituted by organic manure in winter wheat season, that is, the dairy manure with an average dose of 11.8 t dry weight ha −1 year −1 (CV = 61%).The total C and N contents of the applied manure averaged 304.9 g C kg −1 dry weight (CV = 18%) and 15.5 g N kg −1 dry weight (CV = 59%), respectively.The phosphorus and potassium brought in the soil with organic manure amendment were not accounted.All fertilizers were evenly casted on the surface of the corresponding plots and then incorporated into the soils with rotavator to a depth of ca.20 cm just 1 to 3 days before wheat sown and manually incorporated into the soil of ca. 10 cm away from maize stalk, along the row, to a depth of about 10 cm, at approximately 5 weeks after the maize sowing.Winter wheat was normally planted in the middle of October and harvested in the first decade of next June.Summer maize was then sown immediately and harvested three months later, in late September or early October.The plots were irrigated with groundwater 1 or 2 times during the winter wheat growing season and 2 to 4 times of the summer maize growing season with 90 mm of water on each occasion when necessary.Unless otherwise specified, all the crop residues aboveground were removed from the plot.The fields were conventionally tilled using a rototiller.
Soil samples were collected after wheat harvest in June 2019, each treatment plot was divided into three sections of equal size, and eight soil cores (1.5 cm in diameter and 20 cm in height) were collected and bulked from each section to give three replicated samples.The samples were sieved through a 2 mm screen and then stored in three portions; one was stored at 4 • C for biochemical analysis, one was stored at −80 • C for DNA extraction, and the third portion was airdried for chemical analysis.

Soil Chemical and Biochemical Analyses
Soil organic carbon (SOC) was determined by the potassium dichromate (K 2 Cr 2 O 7 ) oxidation at 170-180 • C followed by titration with 0.1 mol L −1 ferrous sulfate [36].Soil TN was assayed by the Kjeldahl method after H 2 SO 4 digestion in the presence of K 2 SO 4 -CuSO 4 -Se catalyst [37].Total P was digested with nitric (HNO 3 ) and hydrofluoric (HF)-perchloric (HClO 3 ) acids [38].Soil available P (Olsen P) was extracted with sodium bicarbonate (0.5 M, pH 8.5) [39], Phosphorus concentrations in the solutions/extracts were determined by the Murphy and Riley procedure (1962) [40].Exchangeable potassium (Exch.K) was extracted with 1 M ammonium acetate and measured by a flame photometer.Soil pH was measured in a 1:1 water/soil ratio (w/w) of carbon dioxide (CO 2 ) free distilled water suspension by glass electrode.Enzyme-P was determined with biologically based phosphor (BBP) method [41] and the soil organic P fractions were determined using Tissen-Moir fractionation protocol [42].

Potential ALP Activity Assay
ALP activity was estimated by measuring the release of p-nitrophenol (PNP) from paranitrophenyl phosphate (PNPP) as described by Tabatabai and Bremner (1969) and Tessmer (1977) [43,44].One gramme of fresh soil (<2 mm) was slightly shaken with 0.2 mL toluene for a few seconds, and then incubated in modified universal buffer (pH 11.0) containing the p-nitrophenol phosphate (Sigma-Aldrich, St. Louis, MO, USA) for 1 h at 37 • C. Reactions were stopped with 0.5 M NaOH, and solutions were filtered with Whatman 42 filter paper.The formation of p-nitrophenol was determined with spectrophotometer at 410 nm [45][46][47].The ALP activity was expressed as microgram per h per g of soil.

DNA Extraction
Total microbial genomic DNA samples were extracted from 0.25 g soil (dry weight equivalent) using the DNeasy Power Soil Kit (QIAGEN, Inc., Hilden, Netherlands), following the manufacturer's instructions, and stored at −80 • C prior to further analysis.The quantity and quality of extracted DNAs were measured using a Nano Drop ND-1000 spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA) and the purity and quality of the genomic DNA were checked on the 0.8% (w/v) agarose gel electrophoresis.

PCR Analysis of phoD Genes
PCR amplification of the phoD gene was performed with the primer ALPS-F733 (5 -TGGGAYGATCAYGARGT-3 ) and ALPS-1083 (5 -CTGSGCSAKSACRTTCCA-3 ) [18,21], and the size of amplicon is 370 bp.There were three replicates of each sample for phoD gene amplification, the sample-specific 7-bp barcode was integrated into the primers for multiplex sequencing.The PCR components included were 5µL of Q5 reaction buffer (5×), 5 µL of Q5 High Fidelity GC buffer (5×), 0.25 µL of Q5 High-Fidelity DNA Polymerase (5 U/µL), 2 µL (2.5 mM) of dNTPs, 1 µL (10µM) of each Forward and Reverse primer, 2 µL of DNA template, and 8.75 µL of ddH 2 O, thermal cycling including initial denaturation at 98 • C for 2 min, followed by 30 cycles (denaturation at 98 • C) for 15 s, then annealing at 55 • C for 30 s, and extension at 72 • C for 30 s, and finally extension at 72 • C for 5 min [48].The cloned plasmid was continuously diluted (every 10 times) to the standard curve, and then the copy number of the phoD gene was automatically obtained according to the standard curve.The R 2 value of the amplification efficiency was 0.9967.PCR amplicons were purified with Agencourt AMPure Beads (Beckman Coulter, Indianapolis, IN, USA) and quantified using Pico Green dsDNA analysis kit (Invitrogen, Carlsbad, CA, USA).After performing a separate quantification step, the amplicons were combined in equal amounts, and paired-end 2 × 300 bp sequencing was performed at Shanghai Personal Biotechnology Co., Ltd.(Shanghai, China) using the Illumina MiSeq platform.

Illumina MiSeq High-Throughput Sequencing and Data Analysis
The compositions and diversities of phoD-harboring bacterial community were assessed using high-throughput sequencing technique.The universal primers ALPS-F733 and ALPS-1083 were selected for PCR amplification of phoD gene.The Quantitative Insights into Microbial Ecology (QIIME, v1.8.0) pipeline was used to process sequencing data.In short, the raw sequencing reads that exactly match the barcode were assigned to each sample and identified as valid sequences.Low-quality sequences were filtered by the following criteria [49,50]: sequences with a length of <150 bp, sequences with an average Phred score of <20, contains ambiguous bases and sequences > 8 bp single nucleotide repeat sequence.The paired end reads were assembled using FLASH [51].After removal of the chimera, UCLUST clustered the remaining high-quality sequences into operational taxa (OTU) with 97% sequence identity [52].The default parameters were used to select a representative sequence from each OTU.OTU taxonomy classification was performed by BLAST searching a representative sequence set [53] against the Greengenes database using the best match [54].An OTU table was further generated to record the abundance of each OTU in each sample and the classification of these OTUs.The samples containing less than 0.001% OTU of the total sequence were discarded.In order to minimize the difference in sequencing depth between samples, by averaging 100 uniformly resampled OTU subsets within 90% of the minimum sequencing depth, an averaged, rounded OTU table was generated for further analysis.The data that support the findings of this study have been deposited in the BioProject with accession number of PRJNA923783.

Statistical Analyses
Statistical analyses were performed with the IMB SPSS statistical software package version 21.Data were analyzed with one-way ANOVA, LSD method was used for multiple comparisons, and the significance level was p < 0.05.Shannon diversity was calculated con-sidering OTU tables, redundancy analysis (RDA) was performed by the Canoco 5 software to explore the relationships between Po fractions and selected soil properties, and microbial variables.Random forest means predictor importance (percent increase in mean square error (MSE) is used to characterize the main predictors.The community composition of phoD-harboring microbes was used as the first axis scores of the redundancy analysis (RDA) of the OTUs profiles of the phoD gene and all soil properties.

Effects of Different Long-Term Fertilization Treatments on Soil Properties
The chemical properties of the tested soil were significantly affected by long-term fertilization (Table 1).The contents of SOC and TN in all fertilization treatments were significantly greater than the control treatment, with the order in significance of MNPK > SNPK > NPK > NK > CK.Addition of phosphorus-containing fertilizers (synthetic phosphorus or organic supplements such as straw/stalk and dairy manure) significantly increased soil total phosphorus, available phosphorus (Olsen P), Enzyme-P and NaOH-Po contents as compared with those receiving no P fertilizer (CK and NK), especially so for treatment receiving organic supplement (MNPK).Phosphorus addition also markedly improved C.HCl-Po content.Potassium fertilizer application also markedly increased the available potassium content of the soil.Application of synthetic fertilizers and organic materials substantially decreased pH value of the cultivated soil, especially for MNPK and SNPK treatments where pH values were markedly lower than the NK and NPK treatments.Only MNPK treatment enhanced soil C/N ratio over other treatments.NK treatment increased C/P ratio; by contrast, treatments in those receiving both N and P fertilizers (socalled balanced fertilization in the soil in question, i.e., NPK, SNPK, MNPK) significantly decreased the C/P ratio.Notes: C/N and C/P represent the ratios of soil organic carbon (SOC) to soil total nitrogen (TN) and soil total phosphorus (TP), respectively; Enzyme-P means organic phosphorus that can be mineralized by enzymes measured with biologically-based phosphorus method (BBP); NaHCO 3 -Po, NaOH-Po, and C.HCl-Po refer to the labile organic P, moderately labile organic P, and recalcitrant organic P; those are extracted with sodium bicarbonate, sodium hydroxide, and concentrated hydrochloric acid, respectively, with Tiessen-Moir fractionation protocol.Data presented are mean± standard error.Different lower-case letters in the same row mean significant differences between treatments at p < 0.05 level.

The Correlation between Soil Po Fractions, Total Po, and Olsen P
Labile Po, moderately labile Po, recalcitrant Po, total Po (the summation of the above Po fractions), and enzyme-P were all significantly and positively correlated with Olsen P with the coefficients of determination R 2 of 0.57, 0.97, 0.61, 0.82, and 0.99, respectively (p < 0.001) (Figure 1).

The Correlation between Soil Po Fractions, Total Po, and Olsen P
Labile Po, moderately labile Po, recalcitrant Po, total Po (the summation of the above Po fractions), and enzyme-P were all significantly and positively correlated with Olsen P with the coefficients of determination R 2 of 0.57, 0.97, 0.61, 0.82, and 0.99, respectively (p < 0.001) (Figure 1).

Effects of Fertilization Regimes on ALP Activity and phoD Gene Harboring Microbial Diversity, Abundance, Composition
Long-term fertilization considerably boosted soil alkaline phosphatase (ALP) activity (Figure 2a), numbers of phoD gene copies (phoD gene abundance, Figure 2b) and Shannon index (phoD gene harboring microbial diversity) (Figure 2c).Treatments for those receiving synthetic fertilizers in combination with organic materials (SNPK and MNPK) significantly increased ALP activity and phoD gene abundance as compared with those receiving no or only synthetic N, P and K fertilizers.MNPK resulted in a dramatic increase in all the three parameters.Treatments for those receiving both N and P fertilizers had significantly higher Shannon index than those receiving no phosphate fertilizer (NK, CK) with the order of MNPK > SNPK > NPK > NK & CK, in significance.Long-term fertilization extensively affected the relative abundance of bacteria encoded by the phoD genes at both the phylum and genus levels.The dominant phoD-harboring bacterial phyla in all treatments were Proteobacteria, Actinobacteria and Gemmatimonadetes, which accounted for more than 90% of the total bacteria at the phylum level (Figure 3a), Proteobacteria was observed to be the dominant phoD-harboring phylum in plough layer soils of all treatments, particularly in the balanced fertilization treatments where it accounted for 82%.The Actinobacteria seemingly showed less response to fertilization treatments, while the Gemmatimonadetes tended to be decreased in relative abundance in soils receiving balanced fertilization such as NPK, SNPK and MNPK (Figure 3a).Ten dominant genera, which account for more than 70% of the total bacteria at the genus level, were illustrated in Figure 3b, including Pseudomonas, Aquabacterium, Methylobacterium, Gemmatirosa, Afipia, Pleomorphomonas, Streptomyces, Lysobacter, Bradyrhizobium, and Rhizobacter.Of which, Pseudomonas was found the dominant phoDharboring genus in plough layer soils of all treatments, particularly in NK soil where it accounted for 22%.Balanced fertilization significantly increased the relative abundance of Aquabacterium, the addition of organic materials significantly increased the relative abundance of Methylobacterium (10-12%) as compared to those receiving no or unbalanced fertilizers; fertilization markedly lessened the relative abundance of Gemmatirosa and Afipia, the latter was further reduced by the addition of organic materials, especially so by organic manure amendment.Synthetic fertilizers enhanced the relative abundance of Streptomyces, and all fertilization treatments considerably improved the relative abundance of Lysobacter.There was no exhibited significant treatment effect on Pleomorphomonas, Bradyhizobium and Rhizobacter (Figure 3b).Furthermore, the analysis results of PERMANOVA also showed that the bacterial community in CK and all the other fertilized treatments can be differentiated (Table 2).

The Correlation between ALP Activity and Diversity of phoD Harboring Bacteria Abundance of the Dominant Phyla, and Genera
Shannon index (diversity of the phoD gene harboring bacteria) and phoD numbers (phoD gene abundance) were positively and linearly correlated wit activity with the determination coefficients (R 2 ) of 0.85 and 0.76, respectively For the dominant phyla, the relative abundance of Proteobacteria was positive arly correlated with ALP activity (R 2 = 0.61, p < 0.001), but that of Gemmatimon  Shannon index (diversity of the phoD gene harboring bacteria) and phoD gene copy numbers (phoD gene abundance) were positively and linearly correlated with soil ALP activity with the determination coefficients (R 2 ) of 0.85 and 0.76, respectively (Figure 4).For the dominant phyla, the relative abundance of Proteobacteria was positively and linearly correlated with ALP activity (R 2 = 0.61, p < 0.001), but that of Gemmatimonadetes (R 2 = 0.38) and Actinobacteria (R 2 = 0.28) were negatively and linearly correlated with ALP activity (p < 0.05) (Figure 5a-c).Of the dominant genera, the values of relative abundance of Methylobacterium (R 2 = 0.52) and Lysobacter (R 2 = 0.52) were positively correlated with ALP activity (p < 0.01).By contrast, Afipi (R 2 = 0.76) and Gemmatirosa (R 2 = 0.34) were negatively correlated with ALP activity (Figure 5d-g).

Reliable Predictors of ALP Activities and Abundance, Diversity, Composition of phoD Gen Harboring Microbes
The importance of each factor in predicting the ALP activity, the abundance, div

Reliable Predictors of ALP Activities and Abundance, Diversity, Composition of phoD Gene Harboring Microbes
The importance of each factor in predicting the ALP activity, the abundance, diversity, and community composition of the phoD gene harboring bacteria was identified by random forest analysis (Figure 7).We used the percentage increases in the mean squared error (%MSE) of these four variables, where higher values denoted the more important of a given factor in predicting the specific variable.For the ALP activity, the most important impact factors were the SOC followed by Olsen P, pH, abundance, diversity, and community composition of phoD gene harboring bacteria (Figure 7a).For the diversity of the phoD gene harboring bacteria, the most crucial factors were Olsen P followed by SOC, TN, pH, MLPo, and Enzyme P (Figure 7b).The most essential factors of phoD gene harboring bacteria abundance were SOC followed by TN, Olsen P, pH, Enzyme P, and MLPo (Figure 7c).pH was the main predictor of the composition of the phoD community based on random forest analysis, followed by SOC, Olsen P, LPo, Enzyme P, and MLPo (Figure 7d).
Agronomy 2023, 13, x FOR PEER REVIEW 14 of 22 a given factor in predicting the specific variable.For the ALP activity, the most important impact factors were the SOC followed by Olsen P, pH, abundance, diversity, and community composition of phoD gene harboring bacteria (Figure 7a).For the diversity of the phoD gene harboring bacteria, the most crucial factors were Olsen P followed by SOC, TN, pH, MLPo, and Enzyme P (Figure 7b).The most essential factors of phoD gene harboring bacteria abundance were SOC followed by TN, Olsen P, pH, Enzyme P, and MLPo (Figure 7c).pH was the main predictor of the composition of the phoD community based on random forest analysis, followed by SOC, Olsen P, LPo, Enzyme P, and MLPo (Figure 7d).

The Relationship between Po Fractions and Selected Soil Properties, and Microbial Variables
The first two axes of redundancy analysis (RDA) explained 87.88% and 3.11% of the total variation of the Po fractions (Figure 8).The labile Po, moderately labile Po, recalcitrant Po, and enzyme P were all positively related to the microbial variables and all the soil properties except pH.Compared with labile Po and recalcitrant Po, moderately labile Po and enzyme P had stronger correlation with ALP and Olsen P, especially so for

The Relationship between Po Fractions and Selected Soil Properties, and Microbial Variables
The first two axes of redundancy analysis (RDA) explained 87.88% and 3.11% of the total variation of the Po fractions (Figure 8).The labile Po, moderately labile Po, recalcitrant Po, and enzyme P were all positively related to the microbial variables and all the soil properties except pH.Compared with labile Po and recalcitrant Po, moderately labile Po and enzyme P had stronger correlation with ALP and Olsen P, especially so for moderately labile Po.Labile Po has the least correlation with ALP and Olsen P. Recalcitrant Po has the strongest correlation with microbial variables among all the Po fractions.

Responses of Soil Chemical Properties to Fertilization Regimes
Our results showed that long-term balanced fertilization, especially with org supplements, significantly increased SOC and TN contents over control, but soil showed the opposite trend.This can be attributed to the greater effects of balanced fe zation on the growth of crops, thereby bringing about a larger amount of organic m from crop residues, i.e., roots, root exudates, stubbles, and other debris into soils; and rect organic carbon input through crop straw/stalk return or manure for treatments organic supplements, such as SNPK and MNPK [55].Reduction in pH can be due to addition of synthetic fertilizers, especially nitrogen fertilizer [56], and the organic generated from mineralization of organic matter investment through the ways st above.Addition of phosphorus-containing fertilizers, either from synthetic phosphate tilizer or organic supplements (straw and dairy manure), significantly increased the tents of soil TP, Olsen P and all the Po fractions, especially so for MNPK treatment.is expected because the annual addition of phosphorus to the soil from only synth phosphate fertilizer (NPK) far exceeds that required for crop growth [57], the crop res

Responses of Soil Chemical Properties to Fertilization Regimes
Our results showed that long-term balanced fertilization, especially with organic supplements, significantly increased SOC and TN contents over control, but soil pH showed the opposite trend.This can be attributed to the greater effects of balanced fertilization on the growth of crops, thereby bringing about a larger amount of organic matter from crop residues, i.e., roots, root exudates, stubbles, and other debris into soils; and direct organic carbon input through crop straw/stalk return or manure for treatments with organic supplements, such as SNPK and MNPK [55].Reduction in pH can be due to the addition of synthetic fertilizers, especially nitrogen fertilizer [56], and the organic acid generated from mineralization of organic matter investment through the ways stated above.Addition of phosphorus-containing fertilizers, either from synthetic phosphate fertilizer or organic supplements (straw and dairy manure), significantly increased the contents of soil TP, Olsen P and all the Po fractions, especially so for MNPK treatment.This is expected because the annual addition of phosphorus to the soil from only synthetic phosphate fertilizer (NPK) far exceeds that required for crop growth [57], the crop residue return (SNPK) and organic manure incorporation further enhanced the surplus of P, and especially manure contains a large amount of P in either organic or inorganic forms.The sequestration of soil Po in any fraction may be the corollary of the increase in soil organic matter/carbon and phosphorus as stated above [25].The significant increase in soil available P (Olsen P) is also due to the improvement of SOC, which is believed to play an important role through activating recalcitrant phosphorus [58] and decreasing the fixation of inorganic P [59].
Our results demonstrated that soil Olsen P was significantly and positively correlated with either the total Po or its fractions with varying correlation coefficient.This may suggest that the transformation of Po to inorganic P under the mineralization of ALP at varying transformation rates/quantities [60].

ALP Activity
Long-term fertilization significantly increased the ALP activity and phoD gene abundance as compared to control; especially, those receiving organic supplements showed substantially greater effects than those receiving only synthetic fertilizers, i.e., NK and NPK (Figure 2).Our observation is same with the findings of Hu et al. (2018) [22].This can be attributed to the significant effects from the application of synthetic fertilizers or their combination with organic supplements on soil properties (Table 1), thereby providing more energy and nutrients for soil microorganisms to secrete more ALP (Figure 7).The more significant effect of treatments receiving organic supplements may be due to these treatments not only providing abundant nutrients, but also a large amount of exogenous organic carbon raising SOC content (Table 1).Saha et al. (2008) [61] reported that long-term farmyard manure input drastically increased ALP activity, some other studies have also documented that increase in soil organic matter could improve ALP activity in soils collected from pig farms [62,63].Mandal et al. (2007) [64] believed that the increase in ALP activity in soils treated with organic supplements might be on account of the increased diversity of P solubilizing bacteria.By sharp contrast, on a lime concretion black soil, Luo et al. (2017) [6] reported the application of either the synthetic fertilizers alone (NPK) or its combination with organic supplements significantly reduced the soil ALP activity.This inconsistency might be due to the decreased soil pH value from its initial of 7.25 (CK) to 7.09 (HMNPK) and 5.86 (NPK), which altered the living conditions of microorganisms in that study, therefore, resulting in a significant decrease in soil ALP activity.

phoD-Harboring Bacteria
Our results showed that the Shannon index of phoD harboring bacteria for treatments of P-containing fertilizer were significantly greater than those without P fertilizer (NK, CK).This may be due to the significantly improved soil Olsen P, SOC, TN and pH, etc. (Figures 7 and 8), which changed the relative abundance of some dominant bacteria, such as Proteobacteria phylum (Figure 3a) and Aquabacterium genera (Figure 3b) in these treatments, thereby leading to the changes in the diversity of the phoD gene harboring bacteria as expressed with the Shannon index.
The dominant phoD-harboring bacteria phyla were Proteobacteria, Actinobacteria, and Gemmatimonadetes in the investigated soils (Figure 3).Our observation was largely in concert with other studies, where the dominant phyla recorded were Proteobacteria and Actinobacteria [6,22].The similarity of these results suggested that these two bacterial phyla may be dominant in a variety of soils and environments where they may play other important ecological functions in addition to the decomposers of organic phosphorus.These key phoD harboring bacteria phyla also reportedly play important roles in soil nutrient cycling, for example, Rhizobiales (Proteobacteria phylum) served as a decomposer in carbon cycling [65], and as a nitrogen fixer in nitrogen cycling [66].Actinobacteria plays an important role in both nitrogen fixation and nitrification in nitrogen cycling [66,67].The major bacteria, Gemmatimonadetes, identified at phylum level in our study is somewhat inconsistent with that of other studies.This may be partly ascribed to the higher soil pH (>8) of our soils.Previous studies have shown that Gemmatimonadetes prefer high pH soils to acidic ones [68][69][70]; it dominates in alkaline and highly saline soils and almost accounted for ca.17% of all the bacteria [71].On the other hand, it may be due to the difference in primers used between our study and studies mentioned above, where they were ALPS-F730/ALPS-1101 [6,18,22] vs. ALPS-F733/ALPS-1183, in our case.Furthermore, the Gemmatimonadetes phylum may play some role in soil P transformation, since it can accumulate a large amount of polyphosphate in its body [72].The observed significant increase of Proteobacteria in relative abundance in soils receiving balanced fertilization over CK and NK (Figure 3a), might be ascribed to the copiotrophic feature of Proteobacteria, which favors the abundant nutrients and organic carbon these treatments provided [67], while the reduction of Gemmatimonadetes in relative abundance in these balanced fertilization soils might be related to its poor competitiveness over the other phyla such as Proteobacteria.
Ten dominant bacterial genera were observed co-existing in soils subjected to different fertilization treatments, in our case, suggesting that these dominant genera of phoD harboring bacteria may be ubiquitous in all environmental conditions [18]; it may also indicate that the key factors controlling Po mineralization in soils were similar [24].However, these bacteria genera responded differently to various fertilization regimes in their relative abundance.Fertilization significantly improved the relative abundance of Lysobacter while decreasing the relative abundance of Gemmatirosa and Afipia (Figure 3b).Our observation on Lysobacter was similar to that of other studies on alkaline saline alluvial soil, acidic black soil, or alkaline loess soil [73][74][75].This may be due to the fact that bacteria of the genus Lysobacter have a good ability to colonize the root periphery of crops [76,77].Application of fertilizers promoted the growth of roots and thus amplified the habitat of this bacteria.The decrease in the relative abundance of Gemmatirosa and Afipia might be associated with their relatively lower growth than the other bacteria in nutrient-rich environment.Compared with the synthetic fertilizers, the incorporation of organic materials significantly decreased the relative abundance of Pseudomonas and Streptomyces but increased that of Methylobacter.Our results were consistent with that reported by Chen et al. (2019) [78], who also documented the decline of relative abundance of Pseudomonas and Streptomyces.The drop in relative abundance of Pseudomonas may be related to soil P level, since the Pseudomonas is the major P-solubilizing bacteria in croplands [79], and dominant in soils with low P level [80], the higher P content as a result of amendment of organic supplement may depress its growth.However, Hu et al. (2018) [22] reported that integrated application of only higher amount of organic manure or crop straw and synthetic fertilizers increased the relative abundance of Pseudomonas against NPK while the incorporation of lower doses of organic materials showed no effect in a calcareous Karst soil (pH 7.28).This discrepancy might be partly ascribed to the similar total P content between fertilized treatments in case of Hu et al. (2018) [22] and partly to the initial high soil organic carbon (24.96 [81] on an acidic red soil (pH 6.0), these authors reported that organic amendments enhanced or had no effects on Streptomyces.These variations are possibly due to soil pH changes as a result of fertilization.Luo et al. (2017) [6] found that the abundance of Streptomyces was only significantly and positively correlated with soil pH.In our case, soil pH decreased with the addition of organic materials, especially so for manure treated soil, while in cases of Luo et al. (2017) [6], Hu et al. (2018) [22] and Liu et al. (2021) [81] the application of organic manure enhanced soil pH relative to that of synthetic fertilizers.Organic materials improved the abundance of Methylobacterium in our study is similar to the results of Zeng et al. (2022) [82] on an Udic Cambosols.Methylobacterium, as a copiotrophic bacteria, is widely distributed in nutrient-rich environment and significantly and positively associated with SOC [82].The incorporation of organic materials greatly raised SOC content (Table 1).

Relationship between the Abundance of phoD-Harboring Bacteria and ALP
The abundance of phoD genes was significantly and positively correlated with ALP activity in present study (Figure 4), indicating that the diversity of phoD gene harboring bacteria plays a key role in the secretion of ALP.We observed that the soil ALP activity was positively correlated with the Proteobacteria phylum but negatively with Gemmatimonadetes and Actinobacteria phyla, suggested Proteobacteria are the major source of ALP secretion, of which the genera of Methylobacterium and Lysobacter are the major contributors.On lime concretion black soil (pH 7.6), Luo et al. (2017) [6] found that the genera Bradyrhizobium, Streptomyces, Modestobacter, Lysobacter, Frankia, and Burkholderia showed significant positive correlations with ALP activity.On the Hapli-Udic Luvisol (pH 6.39), Chen et al. (2019) [21] reported that the genera Streptomyces, Rhizobacter and Varicvorax were the main contributors to ALP.In the results of Li et al. (2021) [83] on brown coniferous forest soils (pH 4.16-5.06),Singulisphaera, Isosphaera, Pirellula, Chrysosporum, Planctomyces, Hapalosiphon, and Anabaena showed positive relationship with ALP.The large variation in ALP secretion bacteria genera may be due to the differences in soil types coupled with various fertilization regimes, which modified soil properties/environments in different ways and to various degrees.For example, Li et al. (2021) [83] claimed that the ALP was primarily regulated by soil TN and the N:P ratio.While Chen et al. (2019) [78] explained that the continuous long-term inputs of mineral N fertilizer led to a 10-fold increase in soil acidity, which may trigger the more active phoD-harboring bacteria to secrete ALP.Luo et al. (2017) [6] stated that higher soil organic carbon favored the genera those efficiently involve in the synthesis of the enzyme.
The bacteria those were significantly and negatively correlated with ALP may be due to the increased phosphorus levels after fertilization inhibiting bacterial activity, thereby reducing ALP secretion [16,46,47].

Relationship between P Forms and ALP
ALP activity significantly and positively impacted the soil Olsen P level (Figure 6f), which is consistent with the results of Hu et al. [22], suggesting the transformation of Po to available P by ALP.However, Fraser et al. [47] and Luo et al. [6] demonstrated that Olsen P was negatively correlated with ALP activities.A possible explanation for these inconsistent results is that the relationship between Olsen P and ALP activities may depend largely upon the P level in the soil.For example, under low-phosphorus conditions, the stress of P induces bacteria to produce more phosphatase [84].While in P-sufficient soils (especially treatments receiving organic supplements), the increasing number of microorganisms could also secrete more phosphatase to mineralize Po [22], thereby increased the Olsen P content in the soils.The tested soil of Fraser et al. [46] was the clay Humic Vertisol, located in Manitoba, Canada with a low available P content; its Olsen P content only reached a maximum value of 12.0 mg/kg even in treatment given an organic supplement.The soil available P in the study of Luo et al. (2017) [6] was also low; it was 20 mg/kg in treatment amended with organic manure.In our cases, the long-term application of NPK fertilizers or their combination with organic supplements (crop straw/stalk or diary manure) significantly increased the soil available P contents (Table 1).The improvement in the number of microorganisms could secrete more phosphatase to mineralize organic P and, thus, increase the Olsen P content.
We observed significant positive correlations between ALP and labile Po, moderately labile Po, recalcitrant Po, total Po and enzyme P, respectively (Figure 6a-e).On the lime concretion black soil studied by Luo et al. (2017) [6], only labile Po showed a significant positive correlation with ALP, and the authors believed that labile Po is the most easily hydrolyzed Po fraction by the enzymes than the other fractions.On red soil (pH 4.73), Cui et al. (2015) [85] found that ALP activity was mainly associated with moderately labile Po (MLPo) and moderately resistant Po (fulvic acid associated Po) contents, rather than labile Po and resistant Po.These discrepancies may be due to differences in soil properties, such as soil pH, organic carbon, etc., which are also key parameters shaping soil microorganisms [18].

Conclusions
Long-term fertilization considerably altered the soil properties and consequently increased the diversity and the abundance of phoD gene harboring bacteria, and the activity of ALP as well.The major phoD gene harboring bacteria, those secreting ALP, belongs to the Proteobacteria phylum, in which the Methylobacteriu and Lysobacter were the major contributors.The enhanced soil ALP promoted the soil Po mineralization, thus improving the soil phosphorus availability.Our results highlighted soil available phosphorus level, organic carbon and pH were the main factors affecting ALP and related bacteria that are responsible for phosphorus transformation.

Figure 1 .
Figure 1.The correlation between soil Olsen P and labile organic P (labile Po, panel (a)), moderately labile Po (b), recalcitrant Po (c), total Po (d) and enzyme P (e).Note: Labile Po, moderately labile Po and recalcitrant Po refer to organic phosphorus, which are extractable with sodium bicarbonate (NaHCO3), sodium hydroxide (NaOH) and concentrated hydrochloric acid (C.HCl) with Tiessen-Moir fractionation protocol; total Po is their summation.Enzyme-P means organic phosphorus that can be mineralized by enzymes measured with biologically-based phosphorus method (BBP).

Figure 1 .
Figure 1.The correlation between soil Olsen P and labile organic P (labile Po, panel (a)), moderately labile Po (b), recalcitrant Po (c), total Po (d) and enzyme P (e).Note: Labile Po, moderately labile Po and recalcitrant Po refer to organic phosphorus, which are extractable with sodium bicarbonate (NaHCO 3 ), sodium hydroxide (NaOH) and concentrated hydrochloric acid (C.HCl) with Tiessen-Moir fractionation protocol; total Po is their summation.Enzyme-P means organic phosphorus that can be mineralized by enzymes measured with biologically-based phosphorus method (BBP).

Agronomy 2023 , 22 Figure 2 .
Figure 2. Alkaline phosphatase (ALP) activity (a), the number of phoD gene copies (b), and the phoD gene harboring microbial diversity as estimated by Shannon index (c) in soils subjected to diverse long-term fertilization treatments.Data presented are treatment means (n = 3), different lower-case letters on top of the bars indicate significant differences between treatments (p < 0.05).

Figure 2 .
Figure 2. Alkaline phosphatase (ALP) activity (a), the number of phoD gene copies (b), and the phoD gene harboring microbial diversity as estimated by Shannon index (c) in soils subjected to diverse long-term fertilization treatments.Data presented are treatment means (n = 3), different lower-case letters on top of the bars indicate significant differences between treatments (p < 0.05).

Figure 3 .
Figure 3. Relative abundances of different phoD-harboring bacterial at phylum (a) an levels expressed as the percentage of the total community in plough layer soils subjected long-term fertilization treatments.Different lower-case letters indicate significant dif tween treatments at p < 0.05 level.

Figure 3 .
Figure 3. Relative abundances of different phoD-harboring bacterial at phylum (a) and genus (b) levels expressed as the percentage of the total community in plough layer soils subjected to different longterm fertilization treatments.Different lower-case letters indicate significant differences between treatments at p < 0.05 level.

Figure 4 .
Figure 4. Soil alkaline phosphatase activity (ALP) as a function of Shannon index (a), phoD gene copy numbers (b).

Figure 4 .
Figure 4. Soil alkaline phosphatase activity (ALP) as a function of Shannon index (a), phoD gene copy numbers (b).

Figure 6 .
Figure 6.The correlation between labile Po (a), moderately labile Po (b), recalcitrant Po (c), total (d), enzyme P (e) and Olsen P (f) and soil alkaline phosphatase activity (ALP).Note: labile Po, m erately labile Po and resistant Po refer to organic phosphorus; these are extractable with sodi bicarbonate (NaHCO3), sodium hydroxide (NaOH) and concentrated hydrochloric acid (C.H with Tiessen-Moir fractionation protocol; total Po is their summation.Enzyme-P means orga phosphorus that can be mineralized by enzymes measured with biologically-based phospho method (BBP).

Figure 6 .
Figure 6.The correlation between labile Po (a), moderately labile Po (b), recalcitrant Po (c), total Po (d), enzyme P (e) and Olsen P (f) and soil alkaline phosphatase activity (ALP).Note: labile Po, moderately labile Po and resistant Po refer to organic phosphorus; these are extractable with sodium bicarbonate (NaHCO 3 ), sodium hydroxide (NaOH) and concentrated hydrochloric acid (C.HCl) with Tiessen-Moir fractionation protocol; total Po is their summation.Enzyme-P means organic phosphorus that can be mineralized by enzymes measured with biologically-based phosphorus method (BBP).

Figure 7 .
Figure 7. Random forest analysis to identify the main predictors of ALP (a), phoD gene harboring microbial diversity (b), phoD gene abundance (c) and phoD community composition (d).% IncMSE means percentage increase in mean square error of a given parameter.LPo, MLPo and RPo refer to labile Po, moderately labile Po, and recalcitrant Po, respectively.SOC, TP and TN stand for the soil organic carbon, total soil nitrogen and phosphorus, respectively.

Figure 7 .
Figure 7. Random forest analysis to identify the main predictors of ALP (a), phoD gene harboring microbial diversity (b), phoD gene abundance (c) and phoD community composition (d).% IncMSE means percentage increase in mean square error of a given parameter.LPo, MLPo and RPo refer to labile Po, moderately labile Po, and recalcitrant Po, respectively.SOC, TP and TN stand for the soil organic carbon, total soil nitrogen and phosphorus, respectively.

Agronomy 2023 ,Figure 8 .
Figure 8. Redundancy analysis (RDA) to explore the relationship between Po fractions and sele soil properties, microbial variables (phoD gene harboring microbial diversity, phoD gene abund and phoD community composition).The direction of the arrows indicates correlations with the two canonical axes, and the length of the arrows and the angles between explanatory and resp variables represent the strength of the correlations.

Figure 8 .
Figure 8. Redundancy analysis (RDA) to explore the relationship between Po fractions and selected soil properties, microbial variables (phoD gene harboring microbial diversity, phoD gene abundance and phoD community composition).The direction of the arrows indicates correlations with the first two canonical axes, and the length of the arrows and the angles between explanatory and response variables represent the strength of the correlations.

Table 2 .
Significance test using the nonparametric multivariate statistical approach (PER MANOVA) to assess the effects of the fertilization on the phoD-harboring bacterial communities.

Table 2 .
Significance test using the nonparametric multivariate statistical approach (PER MANOVA) to assess the effects of the fertilization on the phoD-harboring bacterial communities.