Crops use inorganic and labile organic phosphorus from both high‐ and low‐availability layers in no‐till compost‐amended soils

Organic fertilization in no‐till soils increases soil organic matter and nutrient pools primarily in surface soils. However, little is known about how microbial activity affects crop access to phosphorus (P) forms at the surface, where the organic fertilizer is applied, and the subsurface, the main rooting zone. We aimed to study the changes in organic and inorganic P (Po; Pi) forms and compounds in no‐till compost amended surface (0–5 cm) and subsurface (5–15 cm) soils growing a crop rotation for 2 years in pots. Crops were grown in pots with compost amended to the soil surface, while unamended and compost‐amended pots without crops were used as controls. We measured changes in microbial C (carbon), soluble C, total Po and Pi forms, the moderately accessible EDTA‐NaOH‐Pi (‐Po), and labile NaHCO3‐Pi (‐Po). P compounds in the EDTA‐NaOH extract were measured by 31P‐NMR. Compost addition increased the levels of total Pi and although it had no effect on total Po, increases of inositol, other phosphate monoesters and orthophosphate diesters could be observed. After the application of compost, the amount of total organic C, soluble carbon and all P forms increased in surface soil, while in the subsurface soil, there was a reduction in organic C and an increase in soluble C, total Pi, EDTA‐NaOH‐Pi and NaHCO3‐Pi and the EDTA‐NaOH‐Po and labile NaHCO3‐Po. Growing crops reduced all measured Pi forms and labile NaHCO3‐Po, increased EDTA‐NaOH‐Po in surface soils and had no observable impact on total Po in either organic C‐enriched surface or organic C‐reduced subsurface soils. Crops mostly used Pi from the low P availability C‐reduced subsurface layer, where NaHCO3‐Po pools also decreased. Large reductions in NaHCO3‐Po and increased levels of IHP and other‐monoesters in crop‐growing organic C‐enriched surface layers may suggest microbial formation of monoesters Po and crop use of labile Po pools. In summary, Po formation in C‐enriched surface layers and the mobilization of all Pi forms throughout the soil profile are particularly important findings for the understanding of P dynamics in compost‐amended no‐till systems.


| INTRODUCTION
In many modern agricultural management systems, tillage is reduced because of multiple positive effects on soil structure, moisture retention, biological activity, and erosion reduction.However, the absence of tillage leads to a stratification of the soil with increases of soil organic matter at the soil surface (Alvaro-Fuentes et al., 2009), especially when organic amendments are applied.Repeated application of manure to no-till soils has been shown to increase P bioavailability in near-surface soils (Pavinato et al., 2010).This may affect the retention of P forms and the vertical movement of P. Microorganisms can affect plant uptake of P by regulating the transformation between P forms and P pools (Richardson & Simpson, 2011).Understanding such processes is therefore important for optimizing management decisions.
Mineral soils typically contain a significant proportion of organic phosphorus (Po), varying from 20 to 80% of the total phosphorus (Dalal, 1977;Jantamenchai et al., 2022).In some cases, the Po exceeds inorganic P (Pi).Although Po pools are thought to contribute substantially to plant nutrition by mineralization processes (Steffens et al., 2010), Po dynamics in soils are still not well understood (Haygarth et al., 2018).Agricultural management practices can affect P availability and Po pools (Cooper et al., 2018;Faucon et al., 2015;Foereid, 2017).Increases in Po have been found after applying mineral P fertilizers alone or combined with manures (Li, Romanyà, et al., 2022), whereas, in other studies, no significant changes were observed after applying only manures or plant residues (Annaheim et al., 2015;Jantamenchai et al., 2022).
Increasing organic matter in soil favours Pi mobility (Haynes & Mokolobate, 2001), but its effects on Po pools are less clear.In soils with a low content of organic matter, such as those in the early stages of development, carbon (C) demand may drive the mineralization of Po, and under these conditions, Po may be a source of P for plants (Wang et al., 2016).However, at later stages of development, when soils hold large amounts of organic matter, the organic C/Po ratio increases (Crews et al., 1995;Mcgill & Cole, 1981), suggesting that Po is highly mobilized in high-organic matter soils.Po has been found to increase labile P pools in a range of soils with contrasting organic matter content (Jiménez et al., 2019).However, in such soils, the newly applied mineral P (Pi) is reported to be the main source of P, above all in soils rich in organic matter.Indeed, most P in organic amendments is in mineral forms (Koopmans et al., 2007;Sharpley & Moyer, 2000;Takahashi, 2013), and applying organic amendments in croplands increases pools of organic C and labile Pi (Romanyà et al., 2017;Zhang et al., 2020).Moreover, the long-term application of organic fertilizers has been found to reduce the Po/Pi ratio, mainly in soils with large increases in organic C (Li, Wen, et al., 2022).In contrast, other studies reporting moderate increases or reductions in soil organic C after the application of organic fertilizers observed an increase or no change in the Po/Pi ratio (Romanyà & Rovira, 2009).Therefore, it appears that the effect of organic amendments on the transformation of Po into Pi is affected by the availability of soil organic C.
In agricultural soils, a large proportion of C and nutrients is associated with the soil mineral matrix (Cotrufo et al., 2013).Both roots and soil microbiota can contribute to the mobilization of this mineral-associated organic matter, by releasing organic acids and the subsequent mineralization by enzymatic activity (Barrow et al., 2018;Clarholm et al., 2015;Keiluweit et al., 2015).As soil microorganisms are coupled to soil organic matter and the rhizosphere (Huo et al., 2017), microbial activity in nonrhizospheric agricultural soils is expected to be low due to C limitation.However, this C limitation has also been reported in C-rich forest floors (Vance & Chapin, 2001).Under these conditions, easily degradable organic C, such as from root exudates, can strongly stimulate soil organic matter mobilization and its subsequent mineralization by enzymes-in the case of Po, phosphatases-without the physical limitations imposed by mineral surfaces (Jilling et al., 2021).
Compost in no-till organic systems is applied to the soil surface and thus increases soil organic C and nutrient pools and availability in the top layer.In contrast, compost is likely to have less pronounced effects in the subsurface layers where a significant portion of the crops' root systems will grow.Under these conditions, the effects of root exudates may be different in surface, C and nutrient-enriched layers, compared to subsurface, C and nutrients non-enriched layers.Our hypothesis is that the mobilization of retained P and Po mineralization in crop soils would be mainly enhanced in the surface layers due to their greater C richness, microbial abundances and nutrient pools.
In this research, we examined the effects of compost amendments and crop growth, in no-till soils, on Po and Pi pools as well as their composition in soil.The crops were grown for 2 years in pots that received compost on the soil surface.Our aims were to study the mobilization of P and change in P compounds in (i) uncropped no till unamended agricultural soils, in (ii) no till compost amended surface and subsurface soils and (iii) the effects of growing crops in compost-amended soils after a twoyear crop rotation.

| Soil collection and compost
The soil used for the experiments was collected from an organically managed agricultural field in which horticultural crops had been previously grown for a few years using pipe irrigation.The agricultural field is located near the town of Mollet del Vallès, Catalonia (41°33′44.2″N 2°11′43.3″E), and the soil was a Haplic Luvisol (WRB).The first 20 cm of the Ap horizon was used for the experiments.This soil layer was a loam with 0% stoniness, a pH of 8.01, a calcium carbonate equivalent of 4%, 0.87% of organic C and 0.09% of total N.The collected soil was transferred to the experimental greenhouse at the Torribera Campus of the University of Barcelona (41°27′48.4″N, 2°12′53.0″E).

| Experimental design and soil sampling
For the pot experiments, the soil was sieved using a 1 cm mesh, thoroughly homogenized, and mixed with perlite (Premium Gramoflor from Germany, 2-6 mm) to improve aeration and drainage.The soil: perlite ratio was 2:1 (v/v).The mixture was produced using a substrate mixing machine.The soil was then placed into 16 10-L plastic pots.A subsample of the initial soil was air-dried and kept for further analysis.
For the experiment, twelve pots were filled with soil perlite mixture.In eight halves of the twelve used, 1 L of compost was applied on the soil surface while the other 4 pots did not receive compost.This initial dose of compost is in the higher range of comparable experiments (Kelly et al., 2022), but was required to provide available nutrients for a crop rotation over 2 years and to create two contrasting layers.All pots were automatically dripirrigated to near water holding capacity.Compost-free pots and four of the compost-added pots were kept free of plants for 2 years until sampling was carried out.These non-cropped soils virtually did not grow any plants, as germinating seedlings were readily removed shortly after germination.Inputs of plant residues in such soils were thus extremely low and mostly confined to the top layer.The remaining four with applied compost were used for a crop rotation for 2 years, consisting of sugar beet/tomato/tomato/cabbage/fallow/tomato.After harvesting rooting systems of all crops were kept in soils.In the case of sugar beet, only the commercial part of the beetroots was removed.Tomatoes and aboveground parts of all crops were also removed (Table 1).As large amounts of nutrients were applied with the initial addition of compost, no further fertilization was applied during the experimental period.No crops were grown in unfertilized (compost-soils) as we assumed they would not be able to grow crops for 2 years.
Two years after initiating the study, soils from both experiments were sampled using a 1.4 cm diameter auger.Ten samples from a depth of 0-5 cm (surface) and 5-15 cm (subsurface) were taken from each pot and bulked to one sample per pot and layer.In compostamended soils, the 0-5 cm sample contained most of the remaining compost.This differentiation between the two layers was chosen because the compost was only applied to the very surface.
This experimental design was used to make the following of comparisons: (i) the effects of incubating unfertilized soils for 2 years.This was done by comparing the initial soil with incubated surface (0-5 cm) and subsurface (5-15 cm) soils.(ii) The effects of adding compost in cropless soils and (iii) the effect of growing crops in notill compost-amended soils.

| Soil analyses
Total organic C was determined by dichromate oxidation (Moebius, 1960).Soil microbial biomass C was measured by fumigation extraction (Vance et al., 1987).To determine the microbial biomass C, 4 g of moist soil was fumigated with ethanol-free CHCl 3 , which was removed after 24 h.Fumigated and unfumigated samples were extracted with 20 mL of 0.5 M K 2 SO 4 .Soluble organic C in the extracts was measured by K 2 Cr 2 O 7 oxidation.Microbial biomass C was calculated as the difference in soluble C between fumigated and unfumigated 0.5 M K 2 SO 4 extracts.Total Po was analysed by the ignition method (Saunders & Williams, 1955).Briefly, approximately 2 g of ground airdried soil was extracted with 50 mL of H 2 SO 4 0.5 M.This first extract contained the total Pi.Then, approximately 2 g of air-dried soil was calcinated in a muffle furnace at 550°C for 1 h, and extracted with 50 mL of H 2 SO 4 0.5 M. The P in this extract was analysed as above by the ascorbic acid blue method (Murphy & Riley, 1962) using a CECIL CE-7200 spectrophotometer (Cecil Instruments Limited, Cambridge, UK) at 882 nm.Finally, the difference between both extracts gave the retained total Po.Labile NaHCO 3 -Pi was extracted by the Olsen method with 0.5 M NaHCO 3 (pH 8.5, solution ratio 1:20) and determined by the ascorbic acid blue method before and after the digestion of the soil extracts.Labile P extracts were digested with K 2 S 2 O 8 and NaOH in an autoclave at 120°C (Ebina et al., 1983).Labile NaHCO 3 -Po was determined as the difference between total P in the NaHCO 3 extract and soluble Pi.
Less labile EDTA-NaOH-P was extracted by shaking 10 g of soil (0.15 mm) with 200 mL of a solution containing 0.05 M EDTA +0.25 M NaOH for 16 h at 20°C (Cade-Menun & Preston, 1996;Li, Romanyà, et al., 2022).Each extract was centrifuged at 10,000 g for 30 min.The extract was frozen at −80°C and subsequently freeze-dried over 3 days.The EDTA-NaOH total P content was determined by inductively coupled plasma-mass spectrometry (ICP-MS).The EDTA-NaOH inorganic P (EDTA-NaOH-Pi) content was determined by the acid ascorbic blue colorimetric method.
Before obtaining the NMR spectra, the freeze-dried EDTA-NaOH extracts (300 mg) were re-dissolved in 800 mL 1 M NaOH +200 mL D 2 O, reacted for 30 min, oscillated, centrifuged (1500 g, 20 min) and then transferred to 5 mm diameter NMR tubes. 31P-NMR spectra of the solution were obtained using a GE Omega 500 MHz spectrometer, using a pulse of 90° with a 4.5 s relaxation delay time, 0.68 s acquisition time, 161.98 frequency, and 10,000 scans for all samples.The recycle delay for all samples was determined after the preliminary inversion-recovery experiment, the recycle delay of five times (the longest T1 spin-lattice relaxation times) was chosen to ensure full relaxation between scans, and the cycle delay of all samples ranged from 4.5 to 21.0 s.P composition was identified by the chemical shifts as described by Cade-Menun (2005) and Reusser et al. (2020).Percentages of orthophosphate monoesters and diesters were corrected for degradation of diesters to monoesters during NMR analysis (Cade-Menun, 2017;Li et al., 2020).
To avoid the overlap of the NMR peaks of the most abundant organic P pools as in the case of myo-inositol hexakisphosphate, the deconvolution method was applied (Reusser et al., 2020).Spectral deconvolution fitting was carried out on the orthophosphate and orthophosphate monoester region due to overlapping NMR signals, which was carried out using Matlab R 2022b (The MathWorks Inc.; Natick, MA), as described in Reusser et al. (2020), procedures that include a non-linear optimization algorithm for the spectral deconvolution fitting of the NMR spectra.The 31 P-NMR spectra of the total EDTA-NaOH soil extract solutions collected from all soils including the initial soil are shown in Figure 1.

| Statistical analyses
The data correspond to the means ± standard error of four replicates for each treatment.One-way analysis of variance was used to compare initial soil with surface and subsurface incubated soils.Two-way analysis of variance was used to test the effects of the compost and soil layer (Experiment 1), and the effects of crops and soil layer (Experiment 2) on P forms, soil properties and P compounds.Differences between initial soil and incubated soils were tested in each layer by one-way ANOVA.The same test was used to determine the differences between the top and bottom layers in incubated unamended soils.Relationships between soil properties and P forms were examined by calculating the Pearson correlation coefficients.All the statistical analyses were carried out using SPSS 21.0 (SPSS Inc., Chicago, Illinois, USA).Origin 2020 was used to plot each graph.Prior to statistical analyses, 31 P-NMR data were transformed using centred log-ratio transformation (Abdi et al., 2014). 31P-NMR spectral analysis was conducted using 11.0 MestReNova software.Principal components analysis (PCA) was applied to identify the effects of compost, crop and depth on soil properties and P forms using Origin Software 2020.
F I G U R E 1 Solution 31P NMR spectra of EDTA-NaOH of soil extracts from, initial soil and the incubated soils with and without compost addition, and with and without the presence of crops.The orthophosphate monoester and pyrophosphate spectrum section from 5.1 to −5 ppm is magnified to show the differences between top and bottom layers in each case.

| Effects of incubating an arable soil
Incubating arable soil without compost amendment for 2 years reduced the amount of soil organic C by around one-third in both studied layers (Figure 1, initial soil vs Compost-Plant-, blue box).Soluble C and microbial biomass C were reduced by around 56% in the top and 87% in the bottom layer.During the incubation period, total P did not change, while total Pi significantly increased and total Po decreased.Indeed, the total Po/Pi ratio plummeted from 0.73 in the initial soil to 0.54 in the incubated soils.In contrast with this result, the P extracted by EDTA-NaOH increased considerably in the surface layer and decreased in the subsurface layer (Table 2).EDTA-NaOH-Pi showed a similar pattern as the ETDA-NaOH total P, whereas no significant changes in EDTA-NaOH-Po were observed.Similarly, the proportion of Po in the EDTA-NaOH extracts did not change significantly, ranging from 15 to 25% of the total P.The increase of NaHCO 3 -Pi in both layers was lower, while NaHCO 3 -Po decreased only in the surface layer (Figure 2).The NaHPO 3 Po/Pi ratio decreased in both layers from 0.22 to 0.08 and 0.02 in the incubated soils (Table 3).
The 31 P-NMR spectra of the total EDTA-NaOH soil extract solutions collected from all soils including the initial soil are shown in Figure 3. Orthophosphate, by far the most dominating P form in EDTA-NaOH extracts, mirrored the pattern of EDTA-NaOH-extractable Pi, strongly increasing in the top layer and decreasing in the bottom layer.Pyrophosphate was not very abundant and did not show any significant changes due to its high variability (Figure 4).The changes in the organic P compounds in the incubated soils were not significant, except for a decrease in orthophosphate diesters in the subsurface layer.

| Effects of adding compost
Adding compost to the soil surface increased total organic C in the top layer (0-5 cm) and reduced it slightly in the subsurface layer (5-15 cm).In contrast, microbial and soluble C increased in both layers, although to a much greater extent at the surface layer (Figure 2).
Total Pi increased in compost-amended soils, above all in the surface layer, whereas total Po did not change in either layer (Figure 3).Compared to unamended soils, the reduction in the total Po/Pi ratio was greater in the surface (organic C-enriched) than in the subsurface layer (Cdepleted).An increase in EDTA-NaOH-Po was apparent in both layers, whereas EDTA-NaOH-Pi increased far more on the surface than in the subsurface layer (Figure 3).EDTA-NaOH extracted higher amounts of Pi and Po from the organic matter-enriched surface layer.Regarding the more available NaHCO 3 extracts in compost-amended pots, NaHCO 3 -Po increased extremely in the surface layer (Table 3).In contrast, although NaHCO 3 -Pi also increased to a greater extent in the surface layer, a considerable change was observed in both layers.These changes resulted in the Po/Pi ratio in the NaHCO 3 extracts from the surface layer to increase from 0.08 to 0.37, remaining low in the bottom layer (7%; Table 3).Regarding the P compounds, the addition of compost did not increase the amount of pyrophosphate, but a substantial increase in orthophosphate monoesters was observed in both layers, as well as in other monoesters (Figure 4).No effects on IHP were observed and orthophosphate diesters increased only in the top layer.while total IHP and orthophosphate diesters correlated strongly with microbial biomass, soluble C and NaHCO3-Pi and Po, other monoesters showed slightly weaker correlations (Table 2).Microbial biomass and soluble C were strongly and positively correlated with EDTA-NaOH-Pi and -Po and NaHCO 3 -Pi and -Po forms (Table 2).

Microbial
The first axis of the principal component analysis discriminated between the C-enriched layer (compost+ top soil) and the other studied soils (Figure 5).The separation of the C-enriched layer was because of its high levels of organic, microbial and soluble C, high ratios of total C/total Po, NaHCO 3 -Po, NaHCO 3 -Pi, and NaHCO 3 Po/Pi, and a low total Po/Pi ratio.Other monoesters, total IHP and diesters also contributed to separating this layer from the other soils.Unlike diesters, other-monoesters, total IHP, pyrophosphate, and total Po were important variables on the second axis, which did not discriminate between treatments, showing a large variability among the No-Compost No-Crop soils.

| Effects of growing crops
Although 2 years of crop growth did not significantly increase the total organic C or soluble C in the soil, microbial biomass C was higher in both tested layers (Figure 2).Total Pi declined in both layers, while total Po did not change (Figure 3).Crop growth increased the total Po/ Pi ratio from 0.20 to 0.24% on the surface and from 0.32 to 0.48 in the subsurface layer (Table 3).Similarly, in the EDTA-NaOH extracts, Pi levels decreased.In contrast, EDTA-NaOH-Po increased in the surface layer and no changes in the EDTA-NaOH Po/Pi ratio were observed (Table 3).The amount of NaHCO 3 -Pi and NaHCO 3 -Po decreased in both layers, but in the subsurface layer, this effect was far more pronounced for NaHCO 3 -Pi than for NaHCO 3 -Po.Orthophosphate was reduced in both layers, while pyrophosphate did not show any significant changes (Figure 4).Total IHP and other monoesters significantly increased due to growing crops only in surface soils.Increases in orthophosphate-diesters were not significant.Total IHP correlated strongly with microbial C and soluble C and weakly with NaHCO 3 -Pi and Po (Table 4).In contrast, other monoesters and diesters did not correlate with NaHCO 3 -Pi and Po and showed a weak correlation with soluble C. The correlation between other monoesters and microbial C was strong.
In compost-amended soils with and without crops, microbial biomass and soluble C were strongly and positively correlated with EDTA-NaOH-Pi and -Po and NaHCO 3 -Pi and -Po forms (Table 4).The correlations between NaHCO 3 -Pi and Po and microbial biomass were slightly weaker compared to NaOH-EDTA.The first PCA axis, explaining 62% of the variation, discriminated between the surface layers of compost amended treatments and the subsurface samples (Figure 5).Surface soils, with a high C content, had a low total Po/Pi ratio, high total C/ total Po ratio, and high levels of microbial biomass, soluble C, orthophosphate and TOC/total Po ratio.The second axis (explaining 15% of the variation) discriminated between crop-growing and crop-free soils, but only in the surface soils.Surface crop-growing soils had a high TOC/ total Po ratio, and high content of EDTA-NaOH-Po, IHP and other monoesters and a reduced level of NaHCO 3 -Pi and -Po.

| Effects of compost amendment
The addition of compost increased, as expected, the levels of microbial and soluble C and total C in the surface soil (0-5 cm), the surface layer of the compost+ can be therefore perceived as C-enriched soils.In contrast, in the subsurface layer (5-15 cm), total organic C was reduced relative to the initial soil, although soluble and microbial biomass C increased (Liu et al., 2020).Higher levels of soluble C in the subsurface soil were derived from adding compost to the soil surface and may have accelerated the decomposition of organic C via the priming effect.Other authors have found increased organic C decomposition in soils with high soluble C (Liu et al., 2020).Interestingly, primed decomposition of organic C in the organic Creduced subsurface soil coincided with high increases in total Pi, likely due to leaching with irrigation water.Indeed, adding compost strongly contributed to increasing Pi pools of different degrees of availability.Higher levels of NaHCO 3 -Pi in organic C-enriched soils are reported  in the literature (Shen et al., 2014) and have been related to a reduction in P sorption sites in the soil solid phase caused by an increase in soluble C (Guo et al., 2009).Thus, in our experiment, the increase in soluble C in the organic C-reduced layer may have contributed to the increase in NaHCO 3 -Pi.
Another source of P in the bottom layer was the decomposition of P-containing SOM.Although Po shows sometimes a certain mobility (Darch et al., 2014), in our experiment we detected no leaching of organic P, as the Po in the subsurface layers of the compost-amended treatments a net decrease of Po was observed.
In agreement with our results, other studies report that the addition of organic fertilizers frequently does not change total Po (Annaheim et al., 2015;Jantamenchai et al., 2022).As the total Po balance was not affected by   adding compost, either in surface or subsurface soils, we suggest that the increases in EDTA-NaOH-Po and NaHCO 3 -Po in the subsurface layer of the compostamended soils may indicate a mobilization of Po pools.Indeed, subsurface soil was C-reduced and showed twofold increases in the NaHCO 3 Po/Pi ratio after the addition of compost.Larger pools of NaHCO 3 -Po compared to NaHCO 3 -Pi have been described in soils low in organic C with a low or negative P balance (Romanyà & Rovira, 2007, 2009), suggesting that Po mobilization occurs during mineralization of SOC or accompanying turnover in soils low in organic C.However, in our experiment, increased Po mobilization was also observed in organic C-enriched soils with four-fold increases in the NaHCO 3 Po/Pi ratio and very large increases in NaHCO 3 -Po.Strong correlations between microbial (r = .918,p < .001)and NaHCO 3 -Po and between soluble C (r = .905,p < .001)and NaHCO 3 -Po indicates that labile Po mobilization increases with the availability of C-sources and the abundance of soil microbiota.Therefore, it appears that in both C-enriched and C-reduced soils with higher levels of microbial biomass C and soluble C, P pools can be shifted towards more available forms.

Compost-amended soils having or not having grown crops
The pool of Po in EDTA-NaOH extracts was dominated by other monoesters (Figure 6).The soil content of other monoesters, which are expected to be mineralizable (Xin et al., 2019), may depend on the presence of microbiota (Bünemann et al., 2008).In the present study, the other monoesters significantly increased in compost-amended treatments, both in the surface C-enriched top layer, as well as the subsurface C-reduced subsurface layer, and correlated strongly with microbial biomass C and soluble C.However, since other monoesters were the largest organic compounds in our compost they may have come from the compost itself.Other authors have also found large amounts of other monoesters in compost (Hashimoto et al., 2014).On the other hand, the other monoesters did not decrease in the incubated compostfree or crop-free treatments, even though the microbial biomass C was greatly reduced compared with the initial soil.This suggests that other monoesters may be retained in soils as microbial biomass declines.
Although IHP was the second major compound in our compost, adding compost to the soil surface did not increase IHP in any layer.IHP can be produced (Liu et al., 2018) or decomposed by soil microbiota (Doolette et al., 2010).Indeed, the IHP in our soils was strongly correlated with the levels of microbial biomass or soluble C. On the contrary, total IHP was an important variable for discriminating between C-enriched and C-reduced soils in the first PCA axis, probably because of the IHP content in the compost itself.Higher abundances of IHP have been found during the composting process (Hashimoto et al., 2014) and after the addition of manure to soils (Li, Wen, et al., 2022).It is interesting to note the increased levels of IHP in surface compost-soils as compared to subsurface soils.Such increases may be related to microbial activity or a changed extractability of IHP by EDTA-NaOH extraction with SOM turnover.
Orthophosphate diesters are mainly composed of phospholipids and nucleic acids, mostly derived from microorganisms (Vincent et al., 2013).After compost addition, F I G U R E 6 Overview of the changes in P-forms in the two studied depths and characteristic processes.The size of the circles is proportional to the total P in the NaOH-EDTA extracts, the pie charts represent different P forms characterized by 31P-NMR.
we found substantial increases in orthophosphate diesters only in the organic C-enriched surface soils.Increases in orthophosphate diesters have also been observed after the application of crop residues (Wu et al., 2021); this finding was related to higher structural stability in soils after crop residue application and microbial abundance.In our soils, more abundant orthophosphate diesters were strongly correlated with higher levels of microbial or soluble C, which occurred in organic matter-enriched surface soils.

| Effects of growing crops in compost-amended soils
Here, we present some findings on how growing crops for 2 years may affect P pools and compounds and how these changes relate to soil C dynamics.Growing crops for 2 years did not change the total organic C in the soil, but increased the microbial biomass C. Interestingly soluble C only increased in the subsurface C-reduced layer.P extracted by the crops was about 608 mg pot −1 , which is of the same order of magnitude as the reduction in total P measured in soils from pots growing crops, considering a bulk density of 1.5 g cm −3 .Crop growing did not significantly change the total Po.The concentration of total P decreased by about 57 mg P kg −1 in both layers, indicating that P uptake relative to soil mass was similar in both C-enriched top and C-reduced bottom layers.Nevertheless, the crops obtained three-quarters of their P uptake from the bottom organic C-reduced layer, which had a bigger volume, and lower P availability.This suggests that the crops were able to use P from subsurface organic C-reduced soils despite its lower availability, indicating that the availability of P was not the limiting factor for plant P uptake.According to local reference tables, the availability of Pi in our uncropped compost-amended soils was high in both layers (Villar and Villar, 2016).Growing crops for 2 years reduced Pi availability (NaHCO 3 -Pi) by a similar magnitude in both layers, irrespective of the organic C content and the initial Pi availability.While Pi levels in the organic Crich surface layer remained high after cropping, they were reduced to the medium-low range in the bottom layer.As in other studies (Shen et al., 2014;Zhan et al., 2015), we found higher levels of NaHCO 3 -Pi in organic C-rich versus organic C-reduced soils.However, the lower levels of NaHCO 3 -Pi in the final subsurface C-reduced soils did not change the magnitude of the reduction in NaHCO 3 -Pi after growing crops, but it coincided with increases in soluble C and microbial biomass C.
Reductions in NaHCO3-Pi have been reported in other studies after crop growth, but mostly in soils with low available P, with increases found in soils with high P availability (Barrow, 2022;Romanyà et al., 2017) or with a positive P balance (Zhan et al., 2015).In the latter study, the higher P availability and P reserves after adding manures did not change this trend of reducing NaHCO 3 -Pi.Therefore, we believe that the reducing trend observed in our soils may be related to the low NaHCO 3 -Pi levels in our initial soil.Since we did not observe any reductions in the less labile organic P pools (see EDTA-NaOH Po and total Po) we suggest that reductions of P mostly come from the EDTA-NaOH-Pi and total Pi.Such reductions in Pi, as well as the NaHCO 3 -Pi and Po reductions, are of similar magnitude in both studied layers regardless of the large differences in P pools, total C, soluble C and microbial C that occur between layers.Therefore, it seems that rhizosphere processes will contribute equally to P mobilization in layers with contrasting P and C availability.This process may be mostly largely by plant demand.
Crop growing did not induce any changes in total Po pools in any of the studied soils and the total Po/Pi ratio increased, suggesting higher retention of Po in both surface and subsurface.Po forms have been reported as dominant in low-P soils (Recena et al., 2015).The mobilization of Po in crop-growing soils may thus be confined to less retained Po forms, such as those extracted by EDTA-NaOH and NaHCO 3 and may be related to microbial P recycling.Other studies have reported a higher mobilization of Po due to crop-microorganism interactions (Hallama et al., 2021).Po mobilized by soil microbiota of crop-growing soils has been described as the fraction of the EDTA-NaOH-Po that can be mineralized by enzymes (Hallama et al., 2021), which may be increased by cover crops.Since we found significant increases in EDTA-NaOH-Po in cropgrowing soils it appears that this EDTA-NaOH-Po did not contribute to supplying the crop with P.
The large decrease in NaHCO 3 -Po and the NaHCO 3 Po/Pi ratio in organic C-enriched soil suggests that this form of Po can be used by plants.It has already been proposed that NaHCO 3 -Po contributes to plant nutrition in soils with relatively low levels of NaHCO 3 -Pi (Romanyà & Rovira, 2007).While the total Po/Pi ratio increased in the presence of crops but without any changes in total Po, significant reductions in both parameters were observed in unamended non-cropped incubated soils.This suggests that soil microbiota mineralize Po compounds in the absence of plants.However, this Po reduction coincided with a significant reduction in soil microbial biomass C and the total C/total Po ratio, suggesting a C limitation for the microbiota in unamended non-cropped soils.This contrasts with the increase in the total Po/Pi ratio, soluble C, and microbial biomass C in compost-amended crop-growing soils compared to fallow soils.Inputs of available C by growing crops enhanced microbial processes, contributing to building up more labile Po pools both in the top layers with compost amendment as well as the bottom layer with a lower C content.It has been previously proposed that the demand for C can drive microbial mineralization of Po in organic C-reduced soils (Wang et al., 2016).In the present study, while the total C/total Po ratio was not altered in the bottom layer in the pots growing crops, it increased considerably in the organic C-enriched top layer, which had the highest total C/total Po ratio.This could indicate that in organic C-enriched soils with high P availability, microbes would not mineralize organic C to meet P demand.In fact, no increases in soluble C were found in these soils due to the presence of crops.
Increases in other monoester compounds in organic Cenriched cropped soils amended with exogenous organic C have been related to microbial abundance and activity (Li, Wen, et al., 2022).In the present study, levels of other monoesters increased in the organic C-enriched surface layer of soils with and without crops.This coincided with an increased total C/total Po ratio, and microbial and soluble C due to the addition of compost.The presence of crops further increased the C/total Po ratio and microbial biomass.Root debris and root exudates will likely have contributed to that.Root exudates are composed of carbon-rich compounds such as carbohydrates (Knee et al., 2001;Fernández et al., 2017) and have been reported to enhance soil microbiota in the rhizosphere (Keiluweit et al., 2015).Other studies in intensively managed agricultural soils have found higher levels of microbial biomass P after adding glucose, mainly in organic C-enriched soils (Xu et al., 2020).Such increases coincided with decreases in HCO3-P.In our case, we have also found decreases in NaHCO 3 -Pi and Po and increases in microbial C in both surface C-enriched and subsurface C-reduced cropgrowing soils although the increases in soluble C only occurred in subsurface C-reduced soils and did not coincide with increases in other monoester.This may indicate that in low P and C layers soil microbiota may contribute less to the build-up of other monoester.
Our results revealed that growing crops increased IHP in the top layer.As IHP is known to be an important source of P for plants (Steffens et al., 2010), its increase may contribute to plant nutrition, under the condition of an active microbiome with the capacity to produce phytasehydrolyzing enzymes and a sufficient availability (Jarosch et al., 2019;Liu et al., 2022).Despite their relationship with microbial processes (Condron et al., 2005;Turner & Newman, 2005), orthophosphate diesters were not found to be affected by cropping in the present study.

| CONCLUSIONS
Free-living soil microbiota in incubated agricultural soils mineralize TOC, reduce soluble organic C and microbial C, and transform total Po to Pi in both surface and subsurface layers.In such soils, other monoester Po may be retained as microbial biomass declines.
Compost addition induced a strong increase in soil orthophosphate and Pi fractions and, although it did not add significant amounts of Po, it contributed to a shift in Po pools from more recalcitrant forms to EDTA-NaOH and NaHCO 3 -Po fractions in both the C-enriched surface and C-reduced subsurface layers and coincided with increases in soluble C and microbial biomass C. In the C-enriched surface layer, increases in other-monoester and diester Po coincided with increases in TOC/total Po.
Growing crops in soils with compost amendment promoted the increase of TOC relative to Po, in the C-and Penriched surface layer, while the opposite trend occurred in the subsurface layer.Most crop P uptake occurred in the organic C-reduced layer.Indeed, growing crops reduced orthophosphate and Pi fractions in both depths and although total Po levels did not change, NaHCO 3 -Po was reduced in both layers.However, because the reduction in NaHCO 3 -Po was far greater in crop-growing organic C-enriched surface layer, it may have served as a relevant source of available P for plants.Moreover, crop-growing surface soils resulted in the highest levels of IHP and other monoesters, which increased with microbial biomass and soluble C. In C-reduced subsurface layers growing crops, soil microbiota may contribute less to the build-up of IHP and other monoester Po.

F
Total organic C, soluble C, microbial C, TOC/Po ratio in all treatments.Blue boxes show the comparison of initial soil with incubated soils.ANOVA p-values for the significance of the time factor are shown.Red boxes show the comparison between compostamended soils and controls.ANOVA p-values show the significance of the adding compost and layer factors and their interaction.Black boxes show the effects of growing crops.ANOVA p-values show the significance of growing crops and layer factors and their interaction.
Total Pi/Po ratio, EDTA Pi/Po ratio, NaHCO 3 •Po/Pi ratio, of the two layers of all treatments.

F
Total inorganic P, Total organic P, NaHCO 3 -inorganic P, NHCO 3 -organic P, EDTA-inorganic P, and EDTA-organic P in all treatments.ns, no significance.Blue boxes show the comparison of initial soil with incubated soils.ANOVA p-values for the significance of the time factor are shown.Red boxes show the comparison between compost-amended soils and controls.ANOVA p-values show the significance of the adding compost and layer factors and their interaction.Black boxes show the effects of growing crops.ANOVA p-values show the significance of growing crops and layer factors and their interaction.

F
Total EDTA P, orthophosphate, pyrophosphate, Total IHP, Other-monoester, and diesters in all treatments.ns, no significance.Blue boxes show the comparison of initial soil with incubated soils.ANOVA p-values for the significance of the time factor are shown.Red boxes show the comparison between compost-amended soils and controls.ANOVA p-values show the significance of the adding compost and layer factors and their interaction.Black boxes show the effects of growing crops.ANOVA p-values show the significance of growing crops and layer factors and their interaction.

F
Principal component analyses (PCA) of the soil properties (a) and of the P-forms (b) from different treatments.No-Compost Surface: unamended top soil layer; No-Compost Subsurface: unamended bottom soil layer; +Compost Surface: compost-amended top soil layer; +Compost Subsurface: compost-amended bottom soil layer.T A B L E 4 Correlation coefficients among organic P compounds and, soluble C, microbial biomass C and NaHCO 3 -Pi, NaHCO 3 -Po, and EDTA-NaOH inorganic P (Pi), EDTA-NaOH-P (Po), NaHCO 3 -Pi, NaHCO 3 -Po, soluble C, and microbial biomass C in crop and no crop growing soils.(n = 16).

T A B L E 1
Crop P extractions from pots in crop biomass and harvested yield (beet or tomato) over 2 years.

Treatment Layer Total Po/pi ratio NaHCO 3 Po/pi ratio EDTA-NaOH Po/pi ratio
Values shown are means ± standard error, n = 4. ns, no significance.ANOVA significant factors and interactions are indicated.Differences between initial soil and unamended soils are shown by.