Phosphate Fertilizer Effects on Microbial Resource Limitations in Wheat Cropland: Evidence from Ecoenzymatic Stoichiometry

Abstract

The application of phosphate fertilizers significantly influences soil microbial communities and nutrient cycling. Soil enzymes, which are sensitive to nutrient levels, play a critical role in microbial metabolism. However, the impact of phosphate fertilizers on nutrient limitations within the microbial metabolism of agricultural soils remains poorly understood. In this study, soil samples were collected from a depth of 0–20 cm from a wheat crop subjected to a three-year field experiment with six different phosphorus (P) application rates. Soil β-glucosidase (BG) and leucine aminopeptidase (LAP) activities were highest under the P3 (60 kg P₂O₅ ha⁻¹) treatment over the three-year study period. The responses of soil N-acetyl-β-glucosaminidase (NAG) and alkaline phosphatase (AKP) to increasing P additions varied across different years. The EES C:N, C:P, and vector length were significantly greater than 1. Soil nutrient characteristics accounted for 70.71% of the variation in soil enzyme stoichiometry. The vector length and angle of soil enzymes explained by soil nutrient characteristics were 0.65 and 0.73, respectively. Among these factors, ROC exhibited the largest direct and total effect on the soil enzyme vector length and angle. These research findings offer valuable insights for the management of agricultural fertilizers. Consequently, it is recommended to enhance soil carbon levels to alleviate carbon limitations and improve P utilization efficiency.

1. Introduction

Phosphorus (P) is one of the most commonly limiting nutrients for plant growth and productivity in terrestrial ecosystems [1,2,3]. P plays a crucial role in plant biochemical processes, including carbohydrate and protein metabolism as well as photosynthesis [4]. This limitation typically constrains 67% of the primary agronomic productivity of croplands [5], particularly impacting crops such as maize and wheat [6]. In China, P limitation affects annual wheat production by 60–80% [6]. Unlike natural ecosystems, P fertilizers are easily fixed by soil minerals, such as iron and aluminum, rendering them unavailable to crops [7,8]. Consequently, fertilization is widely considered an essential way to supplement available nutrients in the soil [6,9]. The overapplication of P fertilizers is common in agricultural systems worldwide to ensure crop yields. However, excessive use of P fertilizers not only leads to low P use efficiency but also results in P loss into the environment [9]. A meta-analysis indicates that P addition disrupts the balance of carbon:nitrogen:phosphorus (C:N:P) stoichiometry in terrestrial ecosystem [10,11], subsequently affecting microbial growth and community diversity, as well as the biogeochemical cycling of P in agricultural ecosystems [11]. For example, P addition directly influences soil microbes by increasing the soil available phosphorus, but the direction and magnitude of soil microbial responses may vary depending on whether the ecosystem is P-limited [12,13]. Furthermore, enhanced availability of phosphorus stimulated plant growth, which in turn shifted both aboveground and belowground C allocation and litter inputs to the soil [3,14], thereby indirectly affecting microbial communities [15].

Soil microorganisms are an essential component of the soil ecosystem, primarily driving nutrient cycling by secreting extracellular enzymes through interactions within the microbial community and specific species [16,17]. These microorganisms release extracellular enzymes to obtain energy and nutrients [18]. Since these enzymes are produced by cellular metabolism and in response to nutrient availability in the environment, extracellular enzyme activity (EEA) serves as a major link between ecological metabolic theory and ecological stoichiometry theory [19,20]. Changes in the relative abundance of soil extracellular enzyme activity result from soil microbial metabolic activity, which often depends on the biomass and activity of microorganisms [17,21]. Hence, comprehending the mechanisms by which soil microbes impede EEA production is beneficial in reflecting the nutrient-limitation status to constrain their nutritional needs in agricultural ecosystems [22]. To fulfill their nutrient requirements, soil microorganisms produce specific enzymes for C, N, P acquisition [23], namely C-acquiring enzyme (i.e., β-1,4-glucosidase (BG)); N-acquiring enzyme (i.e., β-1,4-N-acetylglucosaminidase (NAG) and leucine aminopeptidase (LAP)); and P-acquiring enzyme (i.e., acid or alkaline phosphatase (AKP)). This enzymatic activity comprises extracellular enzymatic stoichiometry (EES), which is utilized to assess the microbial resource limitations by analyzing patterns of enzyme activities in soil ecosystems [23]. Consequently, the ratio of EES based on natural-log transformed values of EEA is strongly constrained, with a mean stabilized ratio of 1:1:1 at the global scale in all habitats [18]. This ratio is regarded as the balance point of soil ecoenzymatic activities [24]. The deviation of this ratio from 1:1:1 implied that microbial metabolism may be limited by C and/or other nutrients [25]. For instance, soil microbes in agroecosystems subjected to long-term applications of combined organic and inorganic fertilizers experienced the N limitation, as evidenced by an EES ratio of 1.04:1.11:1.00 [23]. Therefore, to determine microbial resource nutrient limitations, EES is regarded as a more practical and conceptual approach for explaining microbial competition for nutrients and predicting nutrient availability across various ecosystems [16,26]. This framework provides a comprehensive characterization of microbial metabolic limitations related to C, N, or P [27,28]. Therefore, the EES approach has been proposed as an effective method for inferring the nutritional status or nutrient limitations of microbes [29,30].

Based on EES approaches, the analysis of the vector length and angle quantifies the relative investments in C versus nutrient acquisition (vector length) or P versus N acquisition (vector angle) by plotting C:P acquisition versus the enzymatic C:N ratio. This method estimates the strategies of resource allocation and provides a clear basis for identifying C-, N-, P-limitation [19,31]. It incorporates the vector length and angle to assess microbial nutrient limitations and defines a vector angle of 45 degrees as the threshold of ecoenzymatic activities that distinguishes microbial phosphorus (P) limitation (>45 degrees) and nitrogen (N) limitation (<45 degrees) [32]. In agricultural ecosystems, the addition of C, N, and P complicates the growth restrictions of microorganisms [31]. Any disturbance in the soil nutrient balance may alter the diversity of the soil microbial community and metabolic function [11,33]. Cui et al. [34] observed that P and N limitations exacerbate C limitation, which may, in turn, promote the decomposition of soil organic matter by microbes [34]. Additionally, Wang et al. [35] reported that microbial C limitation could enhance soil organic matter degradation via microbial activity, thereby providing more C sources to fulfill their nutrient requirements [35]. When microorganisms are P-limited, phosphatase activity is increased (the EEA C:P ratio is decreased), leading to increased decomposition of organic phosphorus compounds [36]. Moreover, soil microorganisms are easily limited by P, which prompts alterations in their nutrient acquisition strategies [13,37]. The microbial demand for C may primarily drive the increase in phosphatase activities, with the soil P mineralization serving as an auxiliary function of C acquisition for soil microbiome [17,34]. It is well established that microorganisms commonly possess a competitive advantage over plants in acquiring soil nutrients [3,38]. Consequently, when microorganisms face nutrient limitations, plants may also encounter similar nutrient restrictions. Therefore, it is important to investigate resource limitations in soil microorganisms to assess the potential benefits of P fertilizers in enhancing soil microbial activity and crop productivity. However, most previous studies have predominantly focused on soil nutrients, soil enzyme activities, or soil microbial composition, with little information on microbial metabolic limitations.

To gain a deeper understanding of the relationship between P fertilizer application and the limitation of soil microbial metabolic resources in wheat fields, this study focused on a wheat field in Yanjin County, Xinxiang City, Henan Province. Over a period of three years, different levels of P fertilizer were applied to the wheat field, and the activities of soil extracellular enzyme related to C, N, and P acquisition by soil microorganisms was measured. Previous studies indicate that the metabolism of both plants and soil microorganisms in wheat fields is susceptible to P limitation, which is associated with the mineralization of soil organic carbon. Therefore, to investigate the trends in resource balance following P fertilization, this study addressed two main concerns through a three-year fertilization trial: (1) the effect of P fertilizer addition on soil enzyme activities and stoichiometry; and (2) the changes in soil microbial resource limitation and the relationship between soil microbial C and P resources in the wheat field.

2. Materials and Methods

2.1. Experimental Site

The field experiment was located in Daliushu Village, Yanjin County, Xinxiang City, Henan Province (35.12° N, 114.7° E) and began in 2021 with a maize and wheat crop rotation system. The study area has a continental monsoon climate, characterized by a mean annual temperature of 13.7 °C and a mean annual precipitation of 606.1 mm. The frost-free period lasts approximately 216 d, and the region receives about 2400 h of annual sunshine. The soil at this site is classified as calcareous fluvo-aquic soil (WRB), with the following physical and chemical properties for the surface layer (0–20 cm): clay content 29%; bulk density 1.33 g cm⁻³; CaCO₃ content 11.2%; pH 8.44; organic matter 9.43 g kg⁻¹; total nitrogen 0.88 g kg⁻¹; available phosphorus 29.63 mg kg⁻¹; available potassium 106.33 mg kg⁻¹.

2.2. Experimental Design

A field experiment was conducted using a wheat–maize rotation. This study focused on winter wheat from 2021 to 2023. Each treatment plot was relatively large (216 m²), reflecting field availability and typical farmer practices, with no replicates included. Six P fertilizer treatments were implemented in this study: P0 (0 kg P₂O₅ ha⁻¹), P1 (15 kg P₂O₅ ha⁻¹), P2 (30 kg P₂O₅ ha⁻¹), P3 (60 kg P₂O₅ ha⁻¹), P4 (90 kg P₂O₅ ha⁻¹), and P5 (120 kg P₂O₅ ha⁻¹). Nitrogen fertilizer (180 kg N ha⁻¹, as urea), phosphorus fertilizer (P₂O₅, as calcium superphosphate), and potassium fertilizer (60 kg K₂O ha⁻¹, as potassium sulfate) were applied as basal fertilizers [6,39]. These inorganic fertilizers were evenly broadcast onto the soil surface and immediately incorporated into the plowed soil (0–20 cm depth) through tillage prior to sowing. Irrigation was conducted according to soil moisture content during the growing period to meet the crop’s needs. The soil moisture content was maintained at approximately 60–70% of the field capacity. Soil moisture content was measured using the EM50 Soil Monitoring System (America). Straw was carefully removed following each harvesting.

2.3. Soil Sample Collection and Analysis

Rhizosphere soil samples were collected from each treatment at the maturity stage of winter wheat during the years 2021, 2022, and 2023. Three sampling points (1 m × 0.5 m per point) were randomly selected from each treatment plot following a random S-shaped route. All plants at each sampling point were gently shaken to remove loosely adhered soil, and the more tightly bound rhizosphere soil was then collected from the roots using a brush. The rhizosphere soils from each sampling point were combined into a single composite, with biological triplicates for each treatment plot. Soil samples were homogenized and sieved (2 mm), and portions were stored at 4 °C for subsequent analysis of extracellular enzyme activities, and at room temperature for the analysis of soil properties.

Soil pH was measured using a glass electrode at a 1:2.5 soil-to-water ratio. Calcium carbonate (CaCO₃) content was determined through high-temperature calcination (1000 °C for 2 h). Under high-temperature conditions, CaCO₃ decomposes into calcium oxide (CaO) and carbon dioxide (CO₂). The content of CaCO₃ was calculated based on the weight loss during calcination. The determination of soil organic carbon (SOC) contents involved the gradual addition of a 1 mol L⁻¹ hydrochloric acid solution to soil samples until no air bubbles were observed, followed by measurement using a C/N analyzer (Multi C/N3100, analytic jena, Jena, Germany). The dissolved organic carbon (DOC) and organic nitrogen (DON) content was measured using a C/N analyzer (Multi C/N3100) through extraction with 0.5 mol L⁻¹ K₂SO₄. The soil readily oxidizable organic carbon (ROC) was determined according to the method outlined by Zhang et al. [40]. Finely ground air-dried soil samples were oxidized with 25 mL of 333 mmol L⁻¹ KMnO₄. The suspensions were horizontally shaken at 60 rpm for 1 h and then centrifuged at 2000 rpm for 5 min. The supernatants were diluted and measured at 565 nm using a spectrophotometer. Soil total nitrogen (TN) was determined by adopting the methodology of Kjeldahl digestion [41]. Soil ammonium (NH₄⁺-N) and nitrate (NO₃⁻-N) contents were evaluated using a flow injection auto-analyzer by extracting with 2 mol L⁻¹ KCl [42]. Soil available nitrogen (AN) content was calculated as the sum of NH₄⁺-N and NO₃⁻-N. Soil available P (AP) concentration was extracted with a 0.5 mol L⁻¹ NaHCO₃ solution, pH 8.5 (soil:extractant solution 1:20, w/v) at 25 °C for 30 min using a reciprocal shaker (180 rpm) and was determined using the molybdenum blue colorimetric method [43]. Additionally, we also calculated the stoichiometric ratios of soil available nutrients DOC to AN (DOC:AN), DOC to AP (DOC:AP), and AN to AP (AN:AP).

2.4. Analysis of Soil EEAs and EES

The ecological enzyme stoichiometry theory assesses microbial nutritional status by analyzing the ratio of the activity of C-related enzymes to that of N- or P-related enzymes. In this evaluation, the activities of enzymes associated with C acquisition (β-1,4-glucosidase, BG), N acquisition (β-1,4-N-acetyl-glucosaminidase, NAG, and leucine aminopeptidase, LAP), and P acquisition (alkaline phosphatase, AKP) were measured. The respective soil enzyme activities were quantified using colorimetric methods with appropriate kits (Suzhou Grace Biotechnology Co. Ltd., Suzhou, China). Detailed information and functions of each soil enzyme as well as specifics of the testing kits were given in Table S1 [38]. Enzyme activity was defined as the nanomoles of substrate released per hour per gram of dry soil (nmol g⁻¹ h⁻¹).

To express the EES, the ratios of C-, N-, and P-acquiring enzymes were calculated [44] using the following formulas:

$$\mathrm{EES}\mathrm{C}:\mathrm{N}=\ln \left(\mathrm{BG}\right)/\ln \left(\mathrm{NAG}+\mathrm{LAP}\right)$$

(1)

$$\mathrm{EES}\mathrm{C}:\mathrm{P}=\ln \left(\mathrm{BG}\right)/\ln \left(\mathrm{AKP}\right)$$

(2)

$$\mathrm{EES}\mathrm{N}:\mathrm{P}=\ln \left(\mathrm{LAP}+\mathrm{NAG}\right)/\ln \left(\mathrm{AKP}\right)$$

(3)

The microbial nutrient limitation index is a metric based on EES ratios that evaluates the metabolic constraints of soil microbes. This index is measured by converting EES ratios into vectors and subsequently calculating both their length and direction. For instance, if the vector points toward phosphorus, it indicates a limitation of P in microbial metabolism. The microbial nutrient limitation index serves as a tool to identify the primary limiting elements for microbial metabolism within the soil and assesses the relative significance of these limiting factors. The formula for evaluating specific vector lengths and angles is as follows:

$$\mathrm{Vector}\mathrm{length}=\sqrt{{\left[\ln \left(\mathrm{BG}\right)/\ln \left(\mathrm{LAP}+\mathrm{NAG}\right)\right]}^2+{\left[\ln \left(\mathrm{BG}\right)/\ln \left(\mathrm{AKP}\right)\right]}^2}$$

(4)

$$\mathrm{Vector}\mathrm{angle}=\mathrm{acrtan}2\left\{\left[\ln \left(\mathrm{BG}\right)/\ln \left(\mathrm{AKP}\right)\right],\left[\ln \left(\mathrm{BG}\right)/\ln \left(\mathrm{LAP}+\mathrm{NAG}\right)\right]\right\}$$

(5)

The scatter of enzyme C:N:P stoichiometry, with horizontal and vertical coordinates set at 1 as baselines [45] and vector analysis [31], was utilized to evaluate soil microbial resource limitation. A longer vector length indicates a greater microbial C limitation, while a vector angle of less than 45 degrees indicates microbial N limitation, conversely, an angle greater than or equal to 45 degrees suggests microbial P limitation.

2.5. Data Analysis

Data processing, statistical tests, and data presentation were conducted using R (version 4.1.2). Essential data were subjected to log transformation or z-normalization prior to statistical analysis. A two-way analysis of variance (ANOVA) was performed to evaluate the effects of P addition treatments, year, and their interactions on soil EEA, EES, vector length, and vector angle. A one-way ANOVA and least significant difference (LSD) were employed to determine the differences in measurable variables between the P addition treatments within each year at a significance level of p < 0.05. Additionally, we determined the differences in measurable variables across the three years within each P addition treatment using a one-way ANOVA and LSD at p < 0.05. Redundancy analysis (RDA) was used to assess the explanatory power of soil nutrient factors on the variance in soil enzyme stoichiometry. Piecewise structural equation modeling (Piecewise SEM) was applied to evaluate the direct and indirect effects of soil nutrient factors on soil microbial resource limitations. The RDA and SEM statistical analysis were performed using R software (version 4.1.2) with the “vegan”, “rdacca. hp” [46], and “piecewiseSEM” [47] packages.

3. Results

3.1. Soil Extracellular Enzyme Activities (EEA) and Stoichiometries (EES)

The two-way ANOVA showed that the soil enzyme activity, enzyme stoichiometric characteristics, and the vector length and angle of soil enzyme in the wheat field were significantly different in different years and P addition treatments (

Table 1. Results of two-way ANOVA showing differences in soil enzyme activity and enzyme stoichiometry between different P addition treatments and different years during the wheat harvest season over three consecutive years. The table presents the F and p values obtained from the two-way ANOVA.

	Year		P Addition		Year × P Addition
	F	p	F	p	F	p
BG	393.149	<0.001	45.405	<0.001	40.857	<0.001
NAG	261.681	<0.001	154.309	<0.001	90.466	<0.001
LAP	2428.113	<0.001	435.115	<0.001	152.35	<0.001
AKP	281.441	<0.001	12.795	<0.001	36.115	<0.001
EEA C:N	103.565	<0.001	76.405	<0.001	45.428	<0.001
EEA C:P	636.292	<0.001	12.364	<0.001	35.017	<0.001
EEA N:P	660.401	<0.001	58.581	<0.001	41.482	<0.001
Vector length	222.292	<0.001	29.275	<0.001	39.138	<0.001
Vector angle	670.689	<0.001	58.82	<0.001	41.858	<0.001

). In 2021 (the first year of P addition treatment), the activities of soil BG and LAP enzyme were significantly higher than those in 2022 and 2023 (the second and third years), respectively (Figure 1). The soil BG enzyme activity exhibited a decreasing trend over the years and decreased with increasing P addition treatment (Figure 1). The enzyme activities of soil BG and LAP were the highest under the P3 treatment across all three years, with their activities initially increasing and then decreasing as P addition increased (Figure 1). In contrast to BG and LAP, the changes in soil NAG enzyme activity varied. The soil NAG enzyme activity decreased over the years under the P0 treatment, while it exhibited a yearly increase under P addition treatments (Figure 1). The response of soil NAG to increasing P addition differed across the years, particularly in 2022, when soil NAG enzyme activity increased with P addition (Figure 1). Additionally, soil AKP activity increased over the years, demonstrating a significant increasing trend from P1 to P4 treatments. After three consecutive years of P addition, it was observed that AKP activity reached a high level, although the AKP activity under the P4 treatment was significantly lower than that of P0, P1, and P5 (Figure 1).

The EES C:N ratio increased with the addition of P in 2021 and 2022 (Figure 2). In contrast, the EES C:N ratio initially decreased and then increased with the P addition in 2023 (Figure 2). The EES C:P ratio was the highest under the P2 treatment in 2021, P1 treatment in 2022, P4 treatment in 2023, respectively (Figure 2). The soil EES C:N ratio exhibited an initial increase followed by a decrease with the increasing of P addition across different years (Figure 2). In both 2021 and 2023, the soil EES N:P ratio also increased initially and then decreased with the addition of P treatments (Figure 2). The EES N:P ratio was the highest in P3 treatment, which was significantly higher than that of other P addition treatments (Figure 2).

3.2. Indicators of Microbial Resource Limitation

The vector length of soil enzymes was significantly greater than 1 (Figure 3). Significant differences were observed across different years under the same P addition treatment. In 2023, the vector length of soil enzymes in wheat without P addition was significantly higher than that of P addition treatment (Figure 3). The vector angle exhibited a similar trend to that of the vector length. Specifically, the vector angle of soil enzyme vector in 2023 was significantly greater than that in 2021 and 2022. In 2021, the vector angle was less than 45 degrees under the P0, P2, and P3 treatments, while it exceeded 45 degrees in other treatments (Figure 3). The vector angle initially decreased and then increased with the increasing of P addition in both 2021 and 2023, whereas it consistently increased with the increasing of P addition in 2022 (Figure 3).

In 2021, the soil vector angle in the wheat field initially decreased and then increased with the increasing length of the vector (Figure 4). In 2022, a significant decreasing trend was observed with the increase in the vector length (Figure 4). In contrast to the trend in 2022, a significant increasing trend with was recorded in 2023 as the vector length increased (Figure 4). The extension of the vector length exacerbates the common limitations of C and P for soil microorganisms.

Collectively, the analysis of soil EEA and the scatter plot both indicated microbial limitations of C, P, and N along the P addition gradient over the three years (Figure 5). Soil microorganisms in the wheat field were mainly limited by C and P. Additionally, soil microorganisms in the wheat field were limited by C and N under P0, P1, P2, and P3 treatments in 2021 (Figure 5).

3.3. Linkages Between Enzyme Stoichiometry and Soil Properties

RDA showed that the interpretation rate of soil nutrient characteristics on soil enzyme stoichiometry was 70.71% (Figure 6A). The SOC, ROC, DOC, NH₄⁺-N, NO₃⁻-N, and AP had significant effects on soil enzyme stoichiometry (Figure 6A). Notably, the relative interpretation rates of soil ROC, NH₄⁺-N, and TN with respect to soil enzyme stoichiometry exceeded 10% (Figure 6B).

SEM analysis revealed that the vector length and angle of soil enzyme explained by soil nutrient characteristics were 0.65 and 0.73, respectively (Figure 7A,B). The effect path of P addition treatment on soil pH was 0.24 (p > 0.05) (Figure 7A,B). P treatment did not significantly impact on SOC, DOC, and ROC but did influence soil NO₃⁻-N. Soil pH significantly affected NH₄⁺-N (0.27, p < 0.05), SOC, and ROC (0.51, p < 0.001, 0.55, p < 0.001) (Figure 7A,B). Additionally, ROC significantly affected soil DON (−0.24, p < 0.001), NO₃⁻-N (0.66, p < 0.001), NH₄⁺-N (0.71, p < 0.001), and soil AP (−0.52, p < 0.001). The path coefficient between soil DOC and DON was 0.79 (p < 0.001) (Figure 7A,B). The path coefficients of soil ROC and NH4-N with respect to the soil enzyme vector length were 1.11 (p < 0.001) and −0.49 (p < 0.001), respectively (Figure 7A). P addition had no significant effect on the soil enzyme vector length, with a path coefficient of 0.12 (p < 0.001) (Figure 7A). The path coefficients of soil ROC, NH₄⁺-N, NO₃⁻-N, AP and P addition treatment on the soil enzyme vector angle were −0.75 (p < 0.001), −0.24 (p > 0.05), 0.28 (p < 0.05), −0.18 (p > 0.05), 0.25 (p > 0.05), respectively (Figure 7B). The direct and total effects of soil ROC on the soil enzyme vector length were greater than those of other nutrient factors (Figure 7C). Furthermore, the direct effect of ROC on soil enzyme vector angle represented the largest negative effect, while the indirect effect of SOC on the soil enzyme vector angle was the largest and positive effect (Figure 7D).

4. Discussion

4.1. Effects of P Addition on Soil Enzyme Activities and Stoichiometry

Crop harvests severely deplete soil nutrients; therefore, P addition may have a more pronounced impact on soil nutrients in croplands [48,49]. The effect of different harvest seasons on soil nutrients fluctuate across years. The research findings indicated that the effect of P addition on soil enzyme activity in wheat fields varied from year to year (Figure 1 and Figure S1). Notably in 2021, the activities of soil BG, NAG, LAP, and AKP enzymes were significantly different from those in 2022 and 2023. In 2021, the soil enzyme C:N and C:P ratios showed a significant increase trend with P addition, accompanied by notable increase in the vector length and angle (Figure 1, Figure 2 and Figure S2). In an agricultural ecosystem, the application of inorganic fertilizers reduces the soil C:N (P) ratio below the threshold necessary for microbial metabolism, leading to C limitation for microorganisms [23]. The P addition increases soil carbon limitation (vector length > 1) during wheat harvest (Figure 3), while microbial metabolism in the soil shifts from nitrogen limitation (vector angle < 45) to phosphorus limitation (vector angle > 45) [28,50]. Nutrient limitation may lead to changes in microbial community structure and the secretion of extracellular enzymes [37,51,52]. In this study, 180 kg ha⁻¹ of nitrogen was added concurrently with phosphorus to meet the nitrogen requirements of both wheat and microbial growth. P addition significantly enhances the absorption of soil nitrogen by plants, thereby promoting the growth of both plants and microorganisms [36,53]. Therefore, in the first year of P addition treatment, soil microorganisms are initially limited by nitrogen (vector angle < 45) due to the increase in P addition, but as a substantial amount of P added, microbial metabolism becomes limited by P (vector angle > 45). This may be attributed to the early P addition, which promotes the absorption and utilization of nitrogen by plants and microorganisms, ultimately leading to nitrogen limitation in the soil [10,48]. A large amount of P addition significantly enhances plant growth and stimulates soil microorganisms, resulting in a significant loss of C and P in the soil [13,50]. This increase in P availability promotes the proliferation of soil microorganisms and alters microbial communities, further intensifying the limitations of soil C and P (Figure 3 and Figure 4; vector angle > 45; vector length > 1) [50]. When soil enzyme activity is constrained by C or P, the C (e.g., straw) input can enhance the respiratory intensity of microorganisms, increase the secretion of ecoenzyme (e.g., AKP) by bacteria, and ultimately accelerate the transformation of organic phosphorus to available P [54]. Therefore, P addition revealed a consistent limitation in the C and P metabolism of soil microorganisms during both the second and third years. Research indicates that the high N addition can enhance P absorption, as N inputs have been shown to improve soil AKP activity [55]. The mineralization of organic phosphorus may be driven by the microbial demand for C [56,57]. Some studies found that phosphatases have lower sensitivity to soil available P, while microorganisms have a greater demand for C in C-limited soils [50]. Microorganisms may preferentially derive C from the process of converting organic nitrogen and phosphorus into inorganic forms [6,10,56]. In addition, N-induced increase in plant phosphorus demand exacerbates P limitation, especially at N levels exceeding those typically found in the environment [58]. Therefore, in C-limited soils, P addition increases the co-limitation of C and P in microbial metabolism.

The research results showed that the soil enzymes activity metrics, including EES C:N, C:P, and the vector length, were significantly higher in P addition (>15 kg P₂O₅ ha⁻¹) than those without P addition (Figure 1, Figure 2 and Figure 3 ). The values for soil EES C:N, C:P, and the vector length were all greater than 1, suggesting that microbial metabolism in the soil is limited by C. As P addition increased, soil enzyme activities C:N and C:P significantly increased, implying that increased P addition contributes to the turnover and metabolic processes of soil microorganisms, resulting in higher C demand and more pronounced C deficiency in soil performance [57]. The vector length exhibits similar trends. Notably, the soil EES C:N, C:P and the vector length significantly decreased with continuous P addition (Figure 2, Figure 3 and Figure S3), indicating that long-term P addition can gradually reduce the C demand intensity of soil microbial metabolism by changing the soil microbial community structure and the composition of soil P, while enhancing soil N and P cycling in the soil [13,59]. Furthermore, the soil EES N:P was significantly lower with P addition compared to without P addition (Figure 3). The soil EES N:P value of less than 1 indicates that soil microorganisms are susceptible to P limitation, and P addition increases the EES N:P ratio [32,34]. Thus, P addition can alleviate the intensity of microbial P limitation in the soil. In a three-year P addition experiment, the changes in soil enzyme activity and vector angle of soil enzyme varied with the increasing P addition, indicating that the intensity of N- and P-acquiring enzymes in the soil is highly affected by the N and P content present in that soil particular year (Figure S4), with a notable sensitivity to N [55]. These findings are consistent with the analysis of C, and P limitations in soil microbial metabolism conducted in this study.

The impact of three-year P addition experiments on soil enzyme activity varies, resulting in distinct trends in the vector’s length and angle throughout the three-year P addition experiment. In the first year of P addition (2021), when the soil enzyme vector length is less than 1.48 (Figure 3), the vector angle decreases as the vector length increases, indicating that soil microbial metabolism is increasingly limited by C and N (Figure 5). As the vector length continues to increase, C limitation intensifies, and soil microorganisms shift from N limitation to P limitation (Figure 4). The initial addition of a small amount of C to urea can promote the activity of N-acquiring enzymes, thereby alleviating soil C limitation. As carbon limitation escalates, AKP activity is increased to decompose organic P into inorganic P, obtaining more C from this process [56,57]. Our study found that the SOC pool was low. The changes observed in 2021 and 2022 differ significantly. When the vector length exceeds 1.48 (Figure 5), the vector angle shows a significant decreasing trend with the increasing of the vector length (Figure 5). In 2022, the vector angle decreases (Figure 3 and Figure 5), and the degree of phosphorus limitation decreases. Microbial P limitation (vector angle > 45 degrees) gradually alleviates with increased P addition, which may lead to reduced secretion of AKP and improved microbial C metabolism [1,60]. The increase in C limitation may reduce the growth and turnover of soil microorganisms, further reducing the limitation of soil P. A notable difference between 2022 and 2023 is that the vector’s length was less than 1.5 (Figure 5). When the lengths of soil enzyme vectors were below 1.5, the vector’s angle increased with the increase in the vector length, with all the vector angles exceeding 45 degrees. The limitation of soil microbial P increased as soil C limitation. The addition of P led to an increase in microbial biomass and enzyme secretion. ALP is more stable and, once released by cells, it is protected and captured by aggregates [61]. Therefore, its activity remains unaffected by microbial synthesis, rendering it insensitive to changes in soil P supply [52]. Our research findings suggest that in C limited wheat field soils, the addition of P makes the soil more susceptible to both C and P limitations. As P addition increases, the soil’s demand for C increases, and continuous P addition alters the stoichiometric relationships between soil enzymes, leading to a more balanced production of C- and P-acquiring enzymes. This can reduce microbial growth and turnover, further influencing nutrient cycling. In carbon-limited soils, P addition is detrimental to microbial growth and turnover under high nitrogen conditions. Therefore, excessive P addition may have negative effects on soil health.

4.2. Factors Determining Microbial Metabolic Limitations in the Wheat Field

The characteristics of soil nutrients can determine soil enzyme activity, thereby determining the nutrient limitation relationship of soil microorganisms [9]. The results of this study found that soil ROC, TN, NH₄⁺-N, AP, SOC, and DOC significantly affect soil EES, with ROC exerting the greatest impact. Correlation analysis further reveals that ROC, TN, and NH₄⁺-N can significantly enhance soil EES C:P and N:P ratios (Figure S4). Additionally, ROC, TN, and NH₄⁺-N show significant positive effects on the soil enzyme vector length and significant negative correlations with the vector angle (Figure S4). Soil microorganisms and plants require a large amount of C, and soil BG enzyme activity increases ROC content. TN and NH₄⁺-N are significantly positively correlated with soil BG and LAP (Figure S4), but they show significant and negative correlations with soil AKP (Figure S4). An increase in AP can enhance AKP enzyme activity. Therefore, this study suggests that soil microbial metabolism has been subjected to long-term carbon limitation, resulting in a high demand for C. The increase in AKP activity serves to mineralize organic P into inorganic P, thereby addressing some of the microbial carbon metabolism needs [49,56,57]. Furthermore, DON is significantly negatively correlated with BG, while AN is significantly positively correlated with BG (Figure S4). In this study, soil microorganisms had a high C demand and could acquire C through the mineralization of organic N to inorganic N, thereby alleviating C deficiency in soil metabolic processes [7,48]. Consequently, we found that the vector’s length of soil enzyme is mainly affected by soil TN, ROC, NH₄⁺-N, DOC:AN, AN:AP (Figure S5 and Figure 6). Additionally, the vector angle of soil enzyme is mainly affected by SOC, TN, ROC, NH₄⁺-N, DON, DOC:AN, DOC:DON, and AN:AP (Figure 6). The imbalance in elemental stoichiometry in our study area may be a significant factor contributing to microbial metabolic limitations, due to the homeostasis of microbial biomass [9,23,52]. Moreover, the ratios of DOC:AN, DOC:DON and AN:AP were significantly correlated with microbial C and P limitations (Figure S5), indicating a strong influence of nutrient stoichiometry on the microbial nutrient acquisition. Notably, microbial C limitation was negatively correlated with DOC:AN but positively correlated with DOC:DON, suggesting that soil nutrient stoichiometry exerts different effects on microbial nutrient limitations (Figure S4).

Soil nutrient availability also regulates microbial C and P limitations in the agricultural ecosystems. The microbial demand for nutrients is determined by the elemental stoichiometry of microbial biomass in relation to the availability of nutrients in the environment [62]. The contents of ROC and NH₄⁺-N were found to be positively and negatively correlated with microbial C and P limitations, suggesting the important roles that soil-available nutrients play in microbial nutrient acquisition [63]. In the agricultural ecosystems, the leaching of available P, which is often immobilized by iron and aluminum, is commonly observed in acidic soils under conditions of high precipitation and humidity [8,64,65]. Soils with high calcium concentrations may occlude P in insoluble forms, while acidic sites may occlude P with iron or aluminum [66]. This phenomenon can lead to the microbial metabolism being limited by P.

The results of SEM analysis found that P addition reduced soil pH during the wheat harvest season, while increasing SOC, DOC, and ROC (Figure 7). It significantly increased the availability of N and P and positively influenced the soil enzyme vector degree. Specifically, P addition had a significant positive effect on soil ROC, DOC, and AP content, as well as on the soil enzyme vector length. Among these factors, ROC has the greatest effect on the soil vector length (Figure 7), with variations in ROC directly affecting both the vector length and the C limitation intensity of soil microorganisms. Additionally, SOC, P addition, DON, NO₃⁻-N, and soil pH all have significant positive effects on soil enzyme vector angle, while ROC demonstrated a significant and negative correlation with the soil enzyme vector angle. Therefore, the findings of this study suggest that P addition in C limited soils affect ROC, AP, and AN, making agricultural soils more susceptible to co-limitation by C and P. Our results advocate for the incorporation of an appropriate amount of biochar into farmland soils, as well as the use of organic fertilizers, straw returning, and other methods to enhance soil C content. Concurrently, the excessive application of P fertilizers should be reduced to solve the problem of soil carbon deficiency. This approach is beneficial for alleviating soil C limitation, accelerating soil C cycling, promoting the growth and community development of soil microorganisms, and ultimately improving soil phosphorus utilization efficiency and wheat yield.

5. Conclusions

This study found that P addition significantly altered soil EEA and soil EES in wheat fields, thereby affecting the balance of energy and nutrient requirements of soil microorganisms. A comparison between P addition and no P addition reveals that P addition increases soil C and P limitations. Although P addition initially limited the soil by C and N, with increasing P addition and long-term use, microbial metabolism in the soil becomes jointly restricted by C and P. There is a significant correlation between carbon and nutrient limitation in soil, and the limitations of soil N or P varies with the increase in soil C limitation in different years. P addition in C-limited soil affects the soil ROC, AP, and AN, making farmland soil more susceptible to the C and P co-limitation. This change may lead to greater limitations for soil microbes in acquiring carbon and phosphorus, as microbial metabolism requires a balanced ratio of carbon, nitrogen, and phosphorus. Moreover, long-term P addition may negatively affect microbial growth and turnover, which in turn can impact plant growth and the stability of the soil ecosystem. Therefore, in soils with obvious C limitations, while increasing the use of nutrient additives, attention should be paid to the addition of C to achieve a balance of C, N, and P in the soil. This will be beneficial for the stability of the microbial community structure, reduce competition for plant nutrients, promote plant growth, and increase crop yields.