Coupled Effects of Reduced Chemical Fertilization and Biochar Supplementation on Availability and Transformations of Nitrogen and Phosphorus in Vegetable Farmland Soil: An In Situ Study in Southern China

: Reduced fertilization technology is an eco-friendly strategy to minimize nitrogen (N) and phosphorus (P) surpluses and losses in vegetable production. However, little is known about the performance of chemical fertilizer reduction when supplemented with palm silk biochar (PSB) in subtropical acid soils. A short-term (60 d) ﬁeld investigation under conditions of in situ incubation was conducted in vegetable farmland in southern China. The treatments included no fertilization (Control), 100% conventional fertilization (CF 100 ), 90% conventional fertilization plus 10% PSB-based fertilization the sustainable availability of N and P in vegetable farmland soil via a variety of transformation processes, such as mineralization, immobilization, and loss. biochar,


Introduction
As the mainstay industry of agriculture in China, vegetable farming has become one major source of household income for many farmers [1]. Fertilization plays a decisive role in enhancing vegetable yields and profitability in southern China [2,3]. Based on the data of 2018 Guangdong Statistical Yearbook on Agriculture [4], crop farming in Guangdong Province in 2017 consumed 9.10 × 10 5 t of nitrogen (N) fertilizer and 1.25 × 10 5 t of phosphorus (P) fertilizer (on N and P basis, respectively), over 29% of which were used to improve vegetable production. By comparing the annual change in chemical fertilizer consumption in Guangdong Province, the rate of N fertilizer applied in 2017 was only 4% lower than that in 2007, whereas the rate of P fertilizer applied between 2007 and 2017 increased dramatically by 36% [4,5]. Furthermore, with increasing amounts of chemical fertilizer input, the average annual input intensity of chemical fertilizer (chemical fertilizer use per unit area of planted land) has shown an increase of 12% during the past 10 years [4,5]. However, it has far exceeded the maximum safe ceiling of 225 kg hm −2 , which was determined by developed countries to prevent chemical fertilizer from polluting water bodies.
Long-term excessive fertilization has resulted in increased fertilizer nutrient surpluses in Guangdong Province [6]. In vegetable farmlands with different planting patterns, the overuse of fertilizers for vegetable production may not only decrease nutrient use efficiency, but also cause the risk of N and P losses via runoff or leaching from soils to adjacent waters [2,3]. According to the First National Pollution Source Census Bulletin in Guangdong Province, the total N and P losses from the planting industry in 2007 were 6.64 × 10 4 and 7.40 × 10 3 t, respectively [7], accounting for 3% and 0.3% of chemical fertilizer input during the same period, respectively. Therefore, soil N and P losses caused by long-term excessive fertilization cannot be ignored. Nowadays, the resulting agricultural non-point source pollution has become one important factor affecting ecological safety and human health. How to improve N and P availability in soils with minimal adverse environmental impacts has long been a hot topic for the research of nutrient management in agroecosystems.
Biochar is a solid product of biomass (e.g., crop residues, forestry wastes, animal feces), formed by pyrolysis at a relatively low temperature (<700 • C) within an anoxic environment [8]. Generally, it has a high porosity, large specific surface area, more surface charge, and strong stability [8]. In recent years, biochar has received more and more attention globally as a promising soil amendment for its role in emission mitigation and carbon (C) sequestration in agroecosystems. Biochar amendment can alter soil nutrient availability and its C:N:P stoichiometric ratios, which are critical in the regulation of nutrient transformation in the soil ecosystem [9,10]. Mineral N (i.e., NH 4 + -N and NO 3 − -N) and available P (e.g., Olsen-P) are the two important indexes of soil nutrient availability, and their concentrations are greatly affected by biochar amendment [11,12]. Transformations (e.g., mineralization, immobilization, and loss) of N and P are the two most important aspects of nutrient cycling in biochar-amended soil [11]. The interaction of biochar with soil can directly or indirectly affect the mineralization, immobilization, and loss processes of N and P by altering the biotic (e.g., microbial and enzymatic) and abiotic (e.g., physical and chemical) drivers of N and P cycling [13,14]. Thus far, many scholars and researchers from domestic areas and abroad have studied the biochar effects on the mineralization, immobilization, and losses that influence the concentrations of NH 4 + -N, NO 3 − -N, and Olsen-P in soil. Aiming at the elucidation of the availability and transformations of N and P in biochar-amended soil, several kinds of determination methods can be used and have been summarized, including (1) a soil incubation method to determine N and P mineralization [15,16], (2) an isotope tracer method to determine N immobilization [17,18], (3) an adsorption-desorption method to determine P immobilization [14,19], (4) a leachate collection method to determine N and P leaching [20,21], (5) an airflow enclosure method to determine NH 3 volatilization [22,23], and (6) a static chamber method to determine N 2 O emission [24,25]. However, most of the studies on the availability and transformations of N and P in biochar-amended soil involve controlled experiments in a laboratory and are usually concentrated in a single process [13,14]. In addition, given the fact that a high degree of spatial variability exists under actual field conditions, the results of field pilot studies regarding the biochar effects on the availability and transformation processes of N and P in soil are confronted with uncertainties at different locations [14,24]. Presently, there is a need for in situ measurements of soil N and P dynamics under various field conditions, especially when considering the impact of the combined application of chemical fertilizer and biochar.
There have been many studies that have focused on the combined effects of chemical fertilizer and biochar on regulating the availability and losses of soil N and P under the influences of different soil properties (e.g., soil texture and pH), fertilizer types (e.g., urea and NH 4 NO 3 ), biochar characteristics (e.g., biomass source and preparation condition), and experimental conditions (e.g., biochar application rate and residence time of biochar in soil). For example, applying both urea and biochar decreased NH 4 + -N but increased NO 3 − -N, whereas the co-application of NH 4 NO 3 and biochar reduced both forms of mineral N in soil [12]. In the presence of urea and monoammonium phosphate, biochar increased the concentrations of Olsen-P and labile P in two alkaline chernozemic soils [11]. Hangs et al. [26] observed that the combination of urea with added biochar mitigated the increased N 2 O emission from two neutral agricultural soils; nevertheless, the accelerated urea hydrolysis in the presence of biochar may increase NH 3 volatilization associated with urea fertilization. Although the partial substitution of diammonium phosphate with biochar mitigated the fluxes of N and P outputs from paddy field water [27], there are still few reports on the in situ study of reduced chemical fertilization coupled with biochar supplementation to regulate the availability and transformation patterns of N and P in vegetable farmland soil. In particular, the mineralization, immobilization, and losses of N and P in biochar-amended soil remain unclear.
The combination of reduced chemical fertilization and biochar supplementation has scarcely been tested so far, although the positive impact of their interaction might be desirable. The large-scale and massive use of biochar in agricultural production, however, has been constrained by its high cost [28,29]. Thus, we hypothesized that the partial substitution of chemical fertilizer with biochar at a relatively low application rate may maintain the availability of N and P in vegetable farmland soil via a variety of transformation processes, such as mineralization, immobilization, and loss. A short-term field experiment was carried out to test these hypotheses. Therefore, the aims of the present study were (1) to investigate the coupled effects of a reduced chemical fertilization and biochar supplementation on the availability and transformations of soil N and P using an in situ sequential soil coring method described by Raison et al. [30] and (2) to evaluate the potential of different application rates of biochar to partially substitute for chemical fertilizer when growing vegetables under field conditions. This study provides a scientific basis for achieving a reasonable input of chemical fertilizer for vegetable production and reducing the environmental risk of N and P losses from farmland soil after applying biochar.

Field Experimental Site and Materials
The experimental site (23 •  The soil type is waterloggogenic paddy soil (classified as Fe-accumuli-Stagnic Anthrosols by The Chinese Soil Taxonomy and Hydragric Anthrosols by The World Reference Base for Soil Resources) derived from river alluvium. The soil texture of the plow layer (0-25 cm) was clay loam, consisting of 23% sand, 48% silt, and 29% clay. The soil bulk density in the 0-40 cm layer was 1.28 g cm −3 . Palm silk biochar (PSB; pyrolyzed at 550 • C in an oxygen-limited environment) used in this study was purchased from Guangdong Bailv Jiahe Environmental Technology Co., Ltd. (Guangzhou, China). The basic physicochemical properties of the vegetable farmland soil (0-10 cm) and PSB (analyzed using the methods detailed in Lu [31], Table S1) are listed in Table 1.
Seeds of purple-red eggplant (Solanum melongena L.) were obtained from Huizhou Four Seasons Green Agricultural Products Co., Ltd. (Huizhou, China). Abbreviations: SSA, specific surface area; CEC, cation exchange capacity; TOC, total organic carbon; TN, total nitrogen; TP, total phosphorus; DOC, dissolved organic carbon; Olsen-P, Olsen-extractable phosphorus; WSCa, water-soluble calcium; C:N ratio (or C:P ratio), the molar ratio of total organic carbon content to total nitrogen (or total phosphorus) of a substrate (i.e., soil and PSB) measured in mass. Data are expressed as means ± SD (n = 3).

Field Experimental Set-Up
The experimental area was divided into 15 plots, arranged in a completely random design. Each plot (1.3 m in width × 30 m in length) was surrounded by drainage ditches (0.1 m in width × 0.2 m in depth) for draining excessive water from the field. There were five treatments ( To obtain undisturbed soil core samples from in situ incubation (see details in Section 2.3), three types of stainless-steel sampling tubes (4.5 cm in inner diameter × 45 cm in length), including starting tube (S-tube), top-covered tube (C-tube), and top-open tube (O-tube), were driven into the soil as one set of locations (three sets of locations per plot). The lower end of the tube had a sharp edge for the ease of driving the tube into the soil; the upper end had two opposite small holes on the tube wall, which can be used in combination with a stainless-steel handle to pull the tube out of the soil [32]. The tubes were driven into the soil at a depth of 40 cm and positioned evenly along the fertilizing ditch A, planting ditch, and fertilizing ditch B before applying basal chemical fertilizer and PSB. The three types of tubes were located on a straight line (C···S···O), perpendicular to the planting ditch, and distributed in the middle of two adjacent eggplant seedlings ( Figure 1). The eggplant seedlings (48 plants plot −1 ) were transplanted in the planting ditches at a spacing of 1.4 m between rows and 0.6 m between plants in a row. Table 3 presents the rates of chemical fertilizer and PSB applied in different treatments. To compare the potential of partial substitution of chemical fertilizer by PSB, the CF 90 B 10 , CF 85 B 15 , and CF 80 B 20 (hereinafter collectively referred to as PSB-based fertilization) treatments had the same inputs of total N (0.451 g tube −1 or 154 kg ha −1 ; Table 3) and P (0.164 g tube −1 or 56.0 kg ha −1 ; Table 3) as the CF 100 treatment. Basal chemical fertilizer and PSB, applied one day before seedling transplanting, were well incorporated into the buried tubes (S-tube, C-tube, and O-tube) and also into the fertilizing ditches (A and B) at a soil depth of 5-10 cm; then, all plots were watered appropriately to distribute the fertilizer evenly throughout the plow layer. For topdressing, compound fertilizer was only incorporated into the fertilizing ditches, but not into the buried tubes, and all fertilization plots received the same form and rate of compound fertilizer twice, of which, 43% was applied at 10 days after transplanting (N: 46.4 kg ha −1 ; P: 5.16 kg ha −1 ) and 57% at bud emergence (N: 61.6 kg ha −1 ; P: 6.84 kg ha −1 ). Care was taken to prevent topdressing chemical fertilizer from entering the tubes with irrigation water. The irrigation management of each plot was the same throughout the experiment.

In Situ Incubation and Analytical Methods
To study the mineralization (e.g., ammonification, nitrification, and P mineralization), immobilization (e.g., N and P physicochemical adsorption, N and P microbial immobilization, P precipitation, and P occlusion), and losses (e.g., N and P leaching, NH 3 volatilization, N 2 O emission, and denitrification loss) of soil N and P, the fixed-site monitoring research was conducted according to the in situ sequential soil coring method [30,33].
Briefly, the S-tube filled with intact soil core samples was removed from each set of locations after 1 day of in situ incubation, before transplanting eggplant seedlings at the trefoil stage in mid-May. The C-tube and O-tube were kept in situ for subsequent incubations under two different conditions: the C-tube was top-covered by the plastic membrane to prevent N and P leaching caused by rainfall and/or irrigation, whereas the O-tube was top-open to simulate natural conditions. The C-tube and O-tube filled with intact soil core samples were removed from each set of locations after 60 days of in situ incubation, and then the eggplant fruits were harvested from early August to mid-September. Unfortunately, because of the eggplant lodging caused by typhoon Nida (2016), the yield of each plot at each harvest was not able to be recorded.
For all treatments, the soil samples (0-40 cm) in each type of tubes taken from three locations in each plot chronologically were mixed thoroughly to produce a representative homogenous sample corresponding to each type of tube. The soil sample was equally divided into two parts after the removal of all visible stone fragments and plant roots. One subsample was air-dried, ground, and passed through a 2-mm sieve before the use for pH measurement. The other fresh subsample was passed through a 2-mm sieve and stored under field moisture at 4 • C for less than two weeks until analyzed for moisture content, NH 4 + -N, NO 3 − -N, and Olsen-P. The analytical methods for soil pH, moisture content, NH 4 + -N, NO 3 − -N, and Olsen-P were described by Lu [31]. The pH values of the air-dried soil samples were measured by potentiometry using a Sartorius PB-10 pH meter equipped with a Sartorius pH/ATC electrode (Göttingen, Germany) after stirring at a 1:2.5 soil-to-water ratio (w/v). The moisture content of the fresh soil samples was measured by gravimetry after oven drying at 105 • C for 6-8 h to achieve constant dry weight (view results in Table S3). Before the extraction of NH 4 + -N, NO 3 − -N, and Olsen-P from soil, all soil samples were adjusted to the same moisture content for facilitating the subsequent extraction process. The NH 4 + -N and NO 3 − -N concentrations in fresh soil samples were determined by a SKALAR San++ continuous flow analyzer (Breda, Netherlands) after extraction with 1 mol L −1 KCl (final concentration). The Olsen-P concentration in fresh soil samples was determined by Mo-Sb colorimetry using a MAPADA V-1200 UV-Vis spectrophotometer (Shanghai, China) after extraction with 0.05 mol L −1 NaHCO 3 (final concentration, pH 8.5).
The equations used for the calculation of mineralization and losses of N [Equations (1)-(8)] and P [Equations (9) and (10)] are as follows [33,34]: Relative NO 3 − -N loss index (RNLI) = (NO 3 − -N loss /N loss ) × 100 Net P mineralization (P min ) = Olsen-P C − Olsen-P S Olsen-P loss (P loss ) = Olsen-P C − Olsen-P O (10) where NH 4 + -N C , NO 3 − -N C , and Olsen-P C (mg kg −1 ) are the concentrations of NH 4 + -N, NO 3 − -N, and Olsen-P in the C-tube, respectively; NH 4 + -N S , NO 3 − -N S , and Olsen-P S (mg kg −1 ) are those in the S-tube, respectively; and NH 4 + -N O , NO 3 − -N O , and Olsen-P O (mg kg −1 ) are those in the O-tube, respectively. If net N or P mineralization is positive, soil organic matter, chemical fertilizer (e.g., urea), and PSB undergo mineralization and release mineral N or available P; if net N or P mineralization is negative, mineral N or available P are immobilized in soil.

Batch Extraction Experiments and Analytical Methods
To examine the effects of PSB amendment on the release of water-soluble NH 4 + -N, NO 3 − -N, and PO 4 3− -P in soil with or without chemical fertilizer addition, the batch extraction experiments were carried out using the following four fertilization patterns  Table 1) weighing 6 g each were thoroughly mixed with the same particle size of compound fertilizer, calcium superphosphate, and/or PSB in 100-mL centrifuge tubes, with the addition rates of all three exogenous materials accounting for 0.7% (w/w) of their respective field dosages (g tube −1 ; Table 3). Subsequently, 60 mL of Milli-Q water and two drops of chloroform were added to each tube. The tubes were continuously shaken at 200 r min −1 for 24 h and then centrifuged at 4000 r min −1 for 10 min at 25 • C, after which, each supernatant was immediately filtered through a 0.45 µm pore-size 25 mm diameter Jingteng PES syringe filter (Tianjing, China). The filtrates were stored in sealed vials and kept in a refrigerator at -4 • C until analyzed. In addition, blank experiments without adding soil were made using the same procedure as above to determine the concentration of NH 4 + -N, NO 3 − -N, or PO 4 3− -P released separately and together from compound fertilizer, calcium superphosphate, and/or PSB. For statistical analysis, all extraction experiments were performed in triplicate.
The concentrations of NH 4 + -N (or NO 3 − -N) in the filtrates taken from the corresponding vials were determined by the SKALAR San++ continuous flow analyzer, whereas those of PO 4 3− -P were determined by Mo-Sb colorimetry using the MAPADA V-1200 UV-Vis spectrophotometer [31]. In the aforementioned 11 treatments, the following indexes were calculated using Equations (11) and (12): where Q mix (mg kg −1 ) is the NH 4 + -N, NO 3 − -N, or PO 4 3− -P concentration in soil (including soil mixtures with compound fertilizer, calcium superphosphate, and/or PSB); Q sum (mg kg −1 ) is the summation of NH 4 + -N, NO 3 − -N, or PO 4 3− -P concentration in soil and other individual materials (including compound fertilizer, calcium superphosphate, and/or PSB); C mix (mg L −1 ) is the NH 4 + -N, NO 3 − -N, or PO 4 3− -P concentration at equilibrium in solution released from soil (including soil mixtures with compound fertilizer, calcium superphosphate, and/or PSB); C sum is the summation of NH 4 + -N, NO 3 − -N, or PO 4 3− -P concentration at equilibrium in solution released separately from soil and other individual materials (including compound fertilizer, calcium superphosphate, and/or PSB); V tot (mL) is the total volume of Milli-Q water and chloroform; M tot is the total mass of soil, compound fertilizer, calcium superphosphate, and/or PSB.
The pH, total organic C, total N, and total P of soil (including soil mixtures with compound fertilizer, calcium superphosphate, and/or PSB) were analyzed [31]. The pH was determined by potentiometry, as described above. The total organic C was determined by volumetry using K 2 Cr 2 O 7 -H 2 SO 4 oxidation followed by FeSO 4 titration. The total N was determined by the Kjeldahl method using digestion-distillation followed by H 2 SO 4 titration. The total P was determined by Mo-Sb colorimetry using the MAPADA V-1200 UV-Vis spectrophotometer after digestion with H 2 SO 4 -HClO 4 .

Data Processing and Statistical Analysis
WPS Office 2021 (Beijing Kingsoft Office Software Inc., Beijing, China) was used for data processing and PASW Statistics 18 (SPSS Inc., Chicago, IL, USA) was used for statistical analysis. To investigate the influence of different treatments on C:N:P stoichiometry in soil, the soil total organic C, total N, total P, NH 4 + -N, NO 3 − -N, mineral N, and Olsen-P concentrations (mg kg −1 ) were transformed to a unit of mmol kg −1 , and soil C:N, C:P, N:P, NH 4 + -N:Olsen-P, NO 3 − -N:Olsen-P, and mineral N:Olsen-P ratios for each treatment were calculated as the molar ratios in this study. Data normality was examined using the Shapiro-Wilk test, which showed that most dependent variables were normally distributed. Homogeneity of variances was checked using Levene's test. If the assumption of homogeneity of variances was met, multiple comparisons of means were carried out using one-way analysis of variance (ANOVA) followed by the Tukey test; if the homogeneity of variance assumption was violated, means across multiple groups were compared using Welch's ANOVA followed by the Games-Howell test. Pairwise comparisons of means were performed using the Student's t-test (normality was assumed) or Mann-Whitney U-test (normality was not assumed) for independent samples. Calculation of Pearson's correlation coefficient (r value) was performed using a bivariate correlation analysis followed by the two-tailed t-test of significance.

Soil pH
Compared with the non-fertilization treatment (Control), the soil pH decreased significantly (p < 0.05) in almost all fertilization treatments (except CF 90 B 10 for the C-tube and CF 100 for the O-tube), regardless of the type of tubes ( Figure 2). Compared with the CF 100 treatment, the soil pH also decreased significantly (p < 0.05) in the CF 80 B 20 treatment for the C-tube.
volume of Milli-Q water and chloroform; Mtot is the total mass of soil, compound fertilizer, calcium superphosphate, and/or PSB.
The pH, total organic C, total N, and total P of soil (including soil mixtures with compound fertilizer, calcium superphosphate, and/or PSB) were analyzed [31]. The pH was determined by potentiometry, as described above. The total organic C was determined by volumetry using K2Cr2O7-H2SO4 oxidation followed by FeSO4 titration. The total N was determined by the Kjeldahl method using digestion-distillation followed by H2SO4 titration. The total P was determined by Mo-Sb colorimetry using the MAPADA V-1200 UV-Vis spectrophotometer after digestion with H2SO4-HClO4.

Data Processing and Statistical Analysis
WPS Office 2021 (Beijing Kingsoft Office Software Inc., Beijing, China) was used for data processing and PASW Statistics 18 (SPSS Inc., Chicago, IL, USA) was used for statistical analysis. To investigate the influence of different treatments on C:N:P stoichiometry in soil, the soil total organic C, total N, total P, NH4 + -N, NO3 − -N, mineral N, and Olsen-P concentrations (mg kg −1 ) were transformed to a unit of mmol kg −1 , and soil C:N, C:P, N:P, NH4 + -N:Olsen-P, NO3 − -N:Olsen-P, and mineral N:Olsen-P ratios for each treatment were calculated as the molar ratios in this study. Data normality was examined using the Shapiro-Wilk test, which showed that most dependent variables were normally distributed. Homogeneity of variances was checked using Levene's test. If the assumption of homogeneity of variances was met, multiple comparisons of means were carried out using one-way analysis of variance (ANOVA) followed by the Tukey test; if the homogeneity of variance assumption was violated, means across multiple groups were compared using Welch's ANOVA followed by the Games-Howell test. Pairwise comparisons of means were performed using the Student's t-test (normality was assumed) or Mann-Whitney Utest (normality was not assumed) for independent samples. Calculation of Pearson's correlation coefficient (r value) was performed using a bivariate correlation analysis followed by the two-tailed t-test of significance.

Soil pH
Compared with the non-fertilization treatment (Control), the soil pH decreased significantly (p < 0.05) in almost all fertilization treatments (except CF90B10 for the C-tube and CF100 for the O-tube), regardless of the type of tubes ( Figure 2). Compared with the CF100 treatment, the soil pH also decreased significantly (p < 0.05) in the CF80B20 treatment for the C-tube.

NH 4 + -N, NO 3 − -N, and Mineral N Concentrations in Soil
Compared with the CF 100 treatment, the soil NH 4 + -N concentration was decreased significantly (p < 0.05) in the non-fertilization (Control) and all PSB-based fertilization treatments (except CF 90 B 10 for the S-tube), regardless of the type of tubes (Table 4). In addition, the higher reduction in chemical fertilization resulted in a greater percentage decrease in NH 4 + -N. For the S-tube, the percentage decreases in NH 4 + -N in the CF 90 B 10 , CF 85 B 15 , and CF 80 B 20 treatments were −5%, −15%, and −20%, respectively-of which, the latter two were in good agreement with the chemical fertilizer reductions of 15% and 20%, respectively. Similar observations were obtained for the soil mineral N concentration and its percentage decrease in the CF 80 B 20 relative to CF 100 treatments.
The effect of the reduced chemical fertilization combined with the PSB application on the NO 3 − -N concentration in soil varied significantly with the type of tubes and with the reduction in chemical fertilization (Table 4). For the S-tube and O-tube, the PSB-based fertilization treatments did not lead to significant changes in the NO 3 − -N concentration; for the C-tube, the concentration of NO 3 − -N was significantly (p < 0.05) higher in the CF 85 B 15 and CF 80 B 20 than CF 100 treatments (except no such difference in CF 90 B 10 vs. CF 100 ). The soil NO 3 − -N concentration was gradually decreased for both the S-tube and O-tube, but it was fluctuant for the C-tube with the reduction in chemical fertilization. For the S-tube, the percentage changes in NO 3 − -N in the CF 90 B 10 , CF 85 B 15 , and CF 80 B 20 treatments were 14%, −6%, and −22%, respectively-of which, the last one corresponded to the chemical fertilizer reduction of 20%, whereas no such corresponding effect was observed for the C-tube and O-tube.

Olsen-P Concentration in Soil
Compared with the CF 100 treatment, the soil Olsen-P concentration decreased significantly (p < 0.05) in the non-fertilization (Control) and CF 80 B 20 treatments for both the S-tube and C-tube (Table 4). Furthermore, the soil Olsen-P concentration showed a decreasing trend with the reduction in chemical fertilization in most cases. For the S-tube, the percentage decrease in Olsen-P in the CF 80 B 20 treatment (−19%) was close to the chemical fertilizer reduction of 20%, whereas, for the C-tube and O-tube, its percentage decreases in the CF 85 B 15 and CF 80 B 20 treatments (−12% and −19%, respectively) fluctuated around the chemical fertilizer reduction of 15%.
3.3. Soil Net N and P Mineralization 3.3.1. Net Ammonification, Net Nitrification, and Net N Mineralization in Soil As shown in Figure 3a, net ammonification in soil generally decreased with the reduction in chemical fertilization. Accordingly, soil net ammonification ranked in the order of CF 100 > CF 90 B 10 and CF 85 B 15 > CF 80 B 20 > Control (p < 0.05). Differently, the CF 80 B 20 and CF 85 B 15 treatments had a significantly (p < 0.05) higher soil net nitrification than the other treatments. Hence, net nitrification in soil was increased dramatically when PSB was applied at 1840 or 2460 kg ha −1 (CF 80 B 20 or CF 85 B 15 , respectively). Overall, net N mineralization in soil followed the order of (CF 100 > CF 90 B 10 and CF 80 B 20 ) ≈ CF 85 B 15 > Control (p < 0.05). Unlike soil net ammonification and nitrification, no significant difference in net N mineralization was found among all PSB-based fertilization treatments. The relative nitrification index (RNI), which acts as an indicator of the prevalence of nitrification over ammonification during mineralization processes, generally showed increasing trends with the reduction in chemical fertilization (Figure 3b). The CF 80 B 20 and CF 85 B 15 treatments had a significantly (p < 0.05) higher RNI than the CF 90 B 10 and CF 100 treatments.

Net P Mineralization in Soil
Unlike soil net N mineralization, net P mineralization was positive in the non-fertilization (Control) and negative in all fertilization treatments (Figure 3a). In the latter case, soil net P mineralization did not change significantly, regardless of chemical fertilizer reduction.

Net P Mineralization in Soil
Unlike soil net N mineralization, net P mineralization was positive in the nonfertilization (Control) and negative in all fertilization treatments (Figure 3a). In the latter case, soil net P mineralization did not change significantly, regardless of chemical fertilizer reduction. Compared with the CF 100 treatment, the NH 4 + -N loss from soil generally showed decreasing trends with the reduction in chemical fertilizer application; reversely, the loss of soil NO 3 − -N increased significantly (p < 0.05) in almost all PSB-based fertilization treatments (except CF 90 B 10 ; Figure 3c). However, the loss of soil mineral N did not change significantly, regardless of chemical fertilizer reduction. The relative NO 3 − -N loss index (RNLI), which is based on the NO 3 − -N loss as a percentage of the mineral N loss from soil, showed a general increasing trend with the reduction in chemical fertilization (Figure 3d). Statistically speaking, the CF 80 B 20 treatment had the highest RNLI, followed by CF 85 B 15 , and the CF 90 B 10 and CF 100 treatments had the lowest (p < 0.05).

Olsen-P Loss from Soil
The Olsen-P loss from soil was negative in the non-fertilization (Control) and positive in all fertilization treatments, which was contrary to the results of net P mineralization in soil (Figure 3c). Compared with the CF 100 treatment, all PSB-based fertilization treatments did not significantly affect the loss of soil Olsen-P.

Effect of Reduced Chemical Fertilizer Combined with Biochar Application on Availability and Transformation of Soil N
During short-term in situ incubation, reduced chemical fertilization coupled with PSB supplementation generally reduced soil NH 4 + -N compared with chemical fertilization alone, regardless of the PSB substitution rate, incubation condition, and sampling time ( Table 4). Further laboratory experiments showed that, compared with the CF 90 treatment, a significant (p < 0.05) reduction in the Q mix value of NH 4 + -N was found in the CF 90 B 10 treatment (Figure 4a). This result strongly supports the conclusion drawn from a previous meta-analysis that the application of N fertilizer (e.g., urea, urea-containing, or NH 4 +based) combined with biochar decreases NH 4 + -N in soil, regardless of the experimental conditions [12]. Unlike soil NH 4 + -N, however, the coupled effects of reduced chemical fertilization and PSB supplementation on soil NO 3 − -N varied in comparison with chemical fertilization alone, depending on the PSB substitution rate, incubation condition, and sampling time (Table 4). For example, the soil NO 3 − -N concentration was found to exhibit a decreasing trend with an increasing PSB substitution rate after 1 day of in situ incubation  . Treatments: CF100, 100% conventional fertilization; CF90, 90% conventional fertilization; CF90B10, 90% conventional fertilization plus 10% PSB-based fertilization; CF85, 85% conventional fertilization; CF85B15, 85% conventional fertilization plus 15% PSB-based fertilization; CF80, 80% conventional fertilization; CF80B20, 80% conventional fertilization plus 20% PSB-based fertilization. Data are expressed as means ± SD (n = 3). For each parameter, bars labeled with different lowercase letters (e.g., a, b, c, etc.) indicate significant (p < 0.05) differences among the rates of chemical fertilization alone based on the Tukey test. The asterisks indicate significant (*, p < 0.05) or highly significant (**, p < 0.01) differences, and ns indicates a non-significant (α = 0.05) difference between the treatments of chemical fertilization with and without PSB amendment based on the Student's t-test or Mann-Whitney U-test.   There is no doubt that the input of N fertilizer (e.g., compound fertilizer) can affect the concentrations of NH 4 + -N and NO 3 − -N in soil, depending on its form and rate applied [12]. The compound fertilizer used in this study contained approximately 20% soluble N compounds, including a majority of amide-N (83%; e.g., urea) and a minority of NH 4 + -N (17%; e.g., monoammonium phosphate, diammonium phosphate, and ammonium sulfate) and NO 3 − -N ( Table 2). Batch extraction experiments showed that, as the rate of urea-containing and NH 4 + -based compound fertilization decreased, the Q sum and Q mix values of NH 4 + -N, rather than those of NO 3 − -N, were gradually reduced (Figure 4a-d), suggesting that reduced chemical fertilization resulted in a decreasing trend in NH 4 + -N, despite an increased PSB application rate. As a result, after 1 day of in situ incubation, 15% and 20% substitutions of chemical fertilizer with PSB (1840 and 2460 kg ha −1 , respectively) exerted a strong effect on decreasing the NH 4 + -N concentration (p < 0.05), whereas the concentration of NO 3 − -N was unaffected, regardless of the PSB substitution rate (Table 4).
In the previous studies, urea or urea-containing fertilizer can be rapidly hydrolyzed to NH 4 + through soil urease activity [35], and subsequently converted to NO 3 − due to the activities of autotrophic nitrifiers, such as ammonia-oxidizing bacteria and archaea [36,37]. Hangs et al. [26] indicated that adding biochar accelerated urea hydrolysis and stimulated nitrifier activity in soil, and thus promoted the conversion of NH 4 + -N to NO 3 − -N via nitrification. In biochar-amended soil, the NH 4 + supply (e.g., applying NH 4 + -based fertilizer) is considered a dominant driver of nitrification [38]. A recent study showed that the nitrification-promoting effect of biochar application was more pronounced in the short-term (fresh biochar) than long-term (aged biochar) treatments [39]. In our study, urea hydrolysis (ammonification) and nitrification processes could also be related to the availability of soil NH 4 + -N and NO 3 − -N when the urea-containing and NH 4 + -based compound fertilizer is co-applied with PSB. According to our results, net ammonification tended to be reduced considerably in the treatments of PSB-based fertilization, as compared with that obtained in the treatment of chemical fertilization alone; reversely, net nitrification was stimulated by an increased PSB substitution rate, which enhanced the relative contribution of net nitrification to net N mineralization, as indicated by the greater RNI value (Figure 3a,b). As shown in Figure 4a-d and Table 5, with chemical fertilizer addition, the Q mix value of NH 4 + -N was more than its Q sum value in a given treatment, whereas, for NO 3 − -N, the Q mix value was less than the corresponding Q sum value, suggesting that chloroform would not be able to stop the urease activity but may inhibit the activities of autotrophic nitrifiers (as supported by many previous studies, such as Klose and Tabatabai [40] and Nortcliff and Gregory [41]); without chemical fertilizer addition, the effect of PSB amendment on the extraction of NH 4 + -N and NO 3 − -N was not obvious (especially for the Q mix value of NH 4 + -N), which is consistent with the findings of Xu et al. [42]. In addition, after 1 day of in situ incubation, the NH 4 + -N concentration in soil was lower than its Q mix value, whereas the concentration of soil NO 3 − -N was higher than its Q mix value, regardless of the PSB amendment rate (Table 4, Figure 4a,c). At the very least, these results indicate that, in the presence of urea-containing and NH 4 + -based compound fertilizer, ammonification and nitrification were believed to be the two major sources of NH 4 + and NO 3 − -N production in soil.  10 1.46 ± 0.63 b 14.5 ± 0.6 a 10.9 ± 0.3 ab B 15 4.32 ± 0.63 a 14.7 ± 0.6 a 11.0 ± 0.3 ab B 20 3.89 ± 0.63 a 14.9 ± 0.6 a 11.2 ± 0.3 a Parameters: Q mix , NH 4 + -N, NO 3 − -N, or PO 4 3− -P concentration in soil (including soil mixtures with compound fertilizer, calcium superphosphate, and/or PSB); Q sum , summation of NH 4 + -N, NO 3 − -N, or PO 4 3− -P concentration in soil and other individual materials (including compound fertilizer, calcium superphosphate, and/or PSB). Treatments: Control, no fertilization; B 10 , 10% PSB-based fertilization; B 15 , 15% PSB-based fertilization; B 20 20% PSB-based fertilization. Data are expressed as means ± SD (n = 3). Per column in the same row, data followed by different lowercase letters (e.g., a, b, c, etc.) indicate significant (p < 0.05) differences among the rates of PSB application alone based on the Tukey test.
Nitrogen loss took place simultaneously with the mineralization of N following biochar application [20,26], which could play an important role in determining the nature and degree of changes in the concentrations of soil NH 4 + -N and NO 3 − -N during shortterm in situ incubation. Due to the fact that more NO 3 − -N than NH 4 + -N can be easily lost by leaching, the higher RNI value led to a larger ratio of NO 3 − -N loss to mineral N loss (i.e., RNLI value), and the lower net ammonification corresponded to a lower loss of NH 4 + -N (Figure 3c,d). Under such circumstances, likely owing mainly to chemical fertilizer reduction, nitrification, and leaching, the "loss" of NH 4 + -N may exceed its direct release from soil-chemical fertilizer-PSB mixtures and indirect production via ammonification, and, thus, the soil NH 4 + -N concentration decreased in the PSB-based fertilization treatments regardless of the PSB substitution rate, incubation condition, and sampling time (Table 4). It is noteworthy that a decrease in NH 4 + -N concentration, along with an increase in NO 3 − -N concentration, was observed after 60 days of in situ incubation under top-covered incubation conditions (C-tube) regardless of the PSB substitution rate. A similar observation was also obtained after 1 day of in situ incubation (S-tube) when PSB was applied at 1230 kg ha −1 (i.e., CF 90 B 10 ). To some extent, these results reflect that, when there is no influence of rainfall and/or irrigation on the leaching of N, nitrification might play a major role in the inverse relationship between the concentrations of NH 4 + -N and NO 3 − -N. However, after 60 days of in situ incubation under top-open incubation conditions (O-tube), the reduction in NO 3 − -N by the increased PSB substitution rate may be due to the rainfall and/or irrigation during the whole experiment. Despite the above discrepancies in leaching, Chau et al. [33] concluded that the leaching loss is controlled by a wide range of factors, such as the rainfall, fertilization pattern, and uptake of N by plants. Quantitative data on the eggplant uptake were not available (lodging damage), so the losses of soil NH 4 + -N and NO 3 − -N from this pathway were not included in the calculations. Our NH 4 + -N and NO 3 − -N losses may therefore be overestimated. In addition to the above-mentioned mechanisms, physicochemical adsorption and microbial immobilization could also be the two potential mechanisms involved in the decreased availability of soil NH 4 + -N during the whole incubation process. Firstly, biochar can adsorb NH 4 + -N from the soil solution [12,13]. Bargmann et al. [43] reported that, after biochar application to soil, NH 4 + -N adsorption to their surfaces was likely to occur due to the large BET surface areas of the biochar. Similarly, the PSB used in our experiments showed a large BET surface area of 217 m 2 g −1 ( Table 1), suggesting that adsorption was potentially responsible for reducing the NH 4 + in soil after PSB application. The cation exchange capacity (CEC) of biochar is a function of its surface area [44]. A boosted regression tree analysis demonstrated significant correlations between biochar CEC and BET surface area vs. NH 4 + -N adsorption [12]. According to our results, the CEC in the PSBamended soil could be enhanced, because the CEC of the PSB (12.2 cmol kg −1 ) was 1.4-fold higher than that of the soil (8.73 cmol kg −1 ; Table 1), presumably leading to the increase in the soil NH 4 + -N retention capacity. Nonetheless, several studies have indicated that the surface area is not the key factor determining the NH 4 + -N adsorption capacity of biochar. For example, Takaya et al. [45] reported a positive relationship between biochar acid functional groups and NH 4 + -N adsorption, which suggested that NH 4 + -N adsorption may have occurred mainly via chemical reactions with oxygen-containing functional groups rather than physisorption. Hu et al. [46] further indicated that surface complexation, cation exchange, and electrostatic attraction are the main mechanisms of NH 4 + -N adsorption by biochar. Secondly, biochar with a high C:N ratio (>20) triggers the immobilization of NH 4 + -N by soil microbes [12]. For instance, the application of high C:N ratio biochar in soil generally promoted short-term N immobilization in microbial biomass [43]. Due to its high C:N ratio (38.1; Table 1), the PSB used in our study was assumed to stimulate N immobilization in the PSB-based fertilization treatments.
In general, soil pH is an important factor regulating N cycling after biochar application [38,39]. In our study, the soil pH reduced significantly (p < 0.05) in almost all fertilization treatments in comparison with non-fertilization treatments, regardless of the PSB substitution rate, incubation condition, and sampling time (Figure 2). This result is consistent with the findings of Zhou et al. [47], who observed that the application of N fertilizer resulted in soil acidification through both the nitrification of NH 4 + -N and the leaching of NO 3 − -N. The soil pH corresponds well to urea hydrolysis and nitrification processes via the overall reaction, which is as follows: CO(NH 2 ) 2 + H 2 O + 4O 2 = 2OH − + 4H + + 2NO 3 − + CO 2 [48]. The urea-containing compound fertilizer used in our experiments could decrease the soil pH as the hydrolysis to NH 4 + releases two OH − but the subsequent nitrification to NO 3 − releases four H + , leading to the net release of protons.
Previous studies suggested that biochar increases the pH of acid soils, resulting from the release of alkaline substances from biochar per se [48,49]. In the present study, however, the soil pH did not significantly change or even reduce significantly (p < 0.05) in the PSB-based fertilization in comparison with CF 100 treatments ( Figure 2). Our results contradict the findings reported by Hangs et al. [26], but are in good agreement with the results of Li et al. [50]. Two possible explanations are as follows: (1) although the amendment of PSB alone significantly (p < 0.05) increased the soil pH by 0.11-0.29 units relative to the non-amended control ( Figure S1), the partial substitution of the chemical fertilizer with PSB was not enough to fully negate the effect of the fertilization-induced acidification of the soil, suggesting that the relatively low application rates of PSB (i.e., 1230-2460 kg ha −1 ) may have no ameliorative effect on soil pH when it is combined with chemical fertilizer; (2) the short-term application of biochar was supposed to consistently enhance soil nitrification [39], resulting in a decreased soil pH. As far as the relationship between the pH and NO 3 − -N concentration is concerned, under some circumstances, a decreased soil pH was accompanied by an increased concentration of soil NO 3 − -N ( Figure 2, Table 4). When all fertilization treatments with and without PSB application were evaluated together, a highly significant negative correlation was found between the pH and NO 3 − -N concentration (r = −0.817 **, n = 12), which is consistent with the findings of Bista et al. [51].

Effect of Reduced Chemical Fertilizer Combined with Biochar Application on Availability and Transformation of Soil P
Owing to the complex and variable characteristics of soil and biochar, the results of studies on whether biochar application can enhance soil P availability are controversial. Biochar application has reportedly increased soil P availability in a number of laboratory, greenhouse, and field trials. For example, Zhang et al. [52] found that the application of mallee biochar to an acid soil had the potential to improve soil available P through its retention of fertilizer P, since more than half of the P adsorbed on the surface of the biochar via chemisorption could be released into a soil solution. Wang et al. [53] suggested that, after the long-term successive co-application of chemical fertilizer and rice straw biochar to a neutral soil, P adsorption may have occurred mainly via physisorption rather than via a chemical reaction, and thus may improve soil available P due to weak adsorption. Different to the above studies, our study showed that, during short-term in situ incubation, reduced chemical fertilization combined with PSB application did not increase the P availability in the vegetable farmland soil compared with chemical fertilization alone, regardless of the incubation condition and sampling time (Table 4). Further laboratory experiments showed that, with chemical fertilizer addition, a significant (p < 0.05) reduction in the Q mix value of PO 4 3− -P was observed in the CF 90 B 10 in comparison with CF 90 treatments, whereas there were no significant differences in CF 85 B 15 vs. CF 85 , as well as CF 80 B 20 vs. CF 80 ; without chemical fertilizer addition, no significant difference in the Q mix value of PO 4 3− -P was found among the rates of PSB application alone (Figure 4e, Table 5). Therefore, with or without chemical fertilizer addition, PSB amendment may have either a negative effect or no effect on the soil P availability, which is consistent with some previous studies reviewed in Li et al. [14].
It has been argued that biochar improves the concentration of available P in soils through its direct release from biochar per se [42,52]. However, the form and concentration of available P in soils vary significantly with different biochar types and application rates [14,52]. In our laboratory experiments, the Q mix values of PO 4 3− -P in the PSB-based fertilization treatments were significantly (p < 0.05) reduced when compared with that in the CF 100 treatment (Figure 4e, Table 5). Similar to the Q mix of PO 4 3− -P, as the rate of chemical fertilization decreased, the concentration of Olsen-P in soil generally showed a gradual decrease, despite an increased PSB application rate (Table 4). It deserves to be mentioned that both 10% and 15% substitutions of chemical fertilizer with PSB (1230 and 1840 kg ha −1 , respectively) may not substantially reduce soil Olsen-P, as indicated by slight or moderate but not statistically significant changes (average percentage decreases of −2% for the CF 90 B 10 and −11% for the CF 85 B 15 treatments; Table 4). Our results imply that the release of P from chemical fertilizer, rather than from PSB, can play a decisive role in determining the soil available P concentration, especially after the application of PSB at a relatively high rate of 2460 kg ha −1 to the vegetable farmland soil. Moreover, the behavior of P release from biochar is partly dependent on the inherent level of soil available P [14]. In our experiments, the soil had a relatively high Olsen-P concentration (64.9 mg kg −1 ; Table 1), which may inhibit the release of P from PSB.
On the other hand, the differences among experiments may be ascribed (at least in part) to the reactions of biochar in soils, such as shifts in the soil pH that influence the ratio of soluble-to-insoluble P pools and the formation of organo-mineral complexes that increase P solubility, as well as alterations in enzyme efficiencies and the microbial community structure that regulate P mineralization [13,54], which will tend to result in a decreased or increased availability of P in soils via a variety of physicochemical and biochemical processes (e.g., adsorption-desorption, precipitation-dissolution, and mineralization-immobilization). For instance, Liu et al. [19] and Xu et al. [55] observed that, as the biochar application rate increased, the P adsorption capacity in various soils increased or decreased depending on the soil pH, thereby affecting the soil P availability. Interestingly, batch extraction experiments showed that, with chemical fertilizer addition, the Q sum value of PO 4 3− -P far exceeded its Q mix value; without chemical fertilizer addition, however, there was no significant difference between the Q mix and Q sum values of PO 4 3− -P (Figure 4e,f, Table 5). Our results are in partial agreement with the findings of Xu et al. [42], who suggested that the observed P availability is greatly lower than the predicted P availability. One possible explanation for the reduction in the Q mix value of PO 4 3− -P is that iron/aluminum (hydr)oxides are the major P sorbents in subtropical acid soils of southern China [19], which can induce PO 4 3− -P adsorption or precipitation. In addition to this, Xu et al. [42] concluded that the decrease in P availability after biochar amendment is attributed to the biochar-derived calcium (Ca)-induced adsorption or precipitation. In the present study, however, the release of water-soluble Ca from the calcium superphosphate (81,500 mg kg −1 ) was approximately 52-fold higher than that from the PSB (1579 mg kg −1 ; Tables 1 and 2). Therefore, it seemed that, after applying chemical fertilizer alone or in combination with PSB, the PO 4 3− -P adsorption or precipitation may have occurred mainly via the calcium superphosphate rather than PSB in the vegetable farmland soil. In addition, another possible reason for the decreased availability of soil P may be explained by the high C:P ratio of PSB (252 ; Table 1), a value well above the threshold (100) for P immobilization [56]. All in all, the relationship between the P availability and fertilization pattern is complex, especially when considering the impact of a reduced chemical fertilization combined with biochar application.
Phosphorus-rich biochar can partially substitute for fertilizer P input and be used as a slow-release fertilizer [57,58]. In the present study, although the differences in the net P mineralization and Olsen-P loss between the chemical fertilization alone and PSB-based fertilization treatments were found not to be statistically significant, more inorganic P from chemical fertilizer applied in combination with PSB tended to be immobilized to unavailable forms and thus contributed to the reduction in Olsen-P loss with the reduction in chemical fertilization during short-term in situ incubation (Figure 3a,c). Therefore, an advantage of the partial substitution of PSB for chemical fertilizer is to minimize soil P accumulation and subsequently the environmental risk of soil P loss.

Effect of Reduced Chemical Fertilization Combined with Biochar Application on Soil C:N:P Stoichiometry
Ecological stoichiometry is the study of the balance of multiple chemical elements in ecological interactions, particularly on C, N, and P [59]. Understanding the characteristics of soil ecological stoichiometry is important for predicting elemental cycling from molecular to ecosystem scales [9,60]. The soil C:N and C:P ratios show the relationships between C and N cycles, as well as between C and P cycles, in the biological decomposition processes of soil organic matter and the potential contributions to soil fertility, and the soil N:P ratio indicates nutrient constraints in terrestrial ecosystems [60,61]. As C, N, and P cycles within the soil/rhizosphere-plant-atmosphere continuum could lead to a relatively stable soil C:N:P ratio primarily through biological processes, balanced soil C:N:P stoichiometry might be one of the mechanisms behind the enhanced N and P transformations in agroecosystems [60,62]. According to our results, the soil C:N:P ratio was 52:2.9:1 in the non-fertilization treatment (Control; Figure 5a-c), which was higher (C:N ratio) or lower (C:P and N:P ratios) than the results from previous studies on global (186:13:1; [63]) and Chinese (136:9.3:1; [64]) surface soils (0-10 cm); furthermore, the soil C:N ratio was also within the range (4.6:1-19.6:1) obtained from the surface layer (0-20 cm) of some paddy soils in the Pearl River Delta, South China [62]. While biochar application in agriculture is widely advocated, the large-scale use of biochar, particularly as a soil amendment, has been constrained by its high cost [28]. Since the high rate of biochar application had often been in debt for cost-effectiveness [29], it is of great significance to apply the appropriate rate of biochar when using it as a partial substitute for chemical fertilizer. In the present study, the relatively low application rate of PSB (e.g., 1840 kg ha −1 , being equivalent to the biochar application rate in the study of Wang et al. [10]) was expected to negate (at least partly) the fertilization-induced imbalance between soil C, N, and P stoichiometry when it is combined with chemical fertilizer, which could contribute to the turnover of mineralization-immobilization (e.g., increased soil net nitrification), thus maintaining the supply of soil available NO3 − -N and Olsen-P for the plant. Treatments: Control, no fertilization; CF100, 100% conventional fertilization; CF90B10, 90% conventional fertilization plus 10% PSB-based fertilization; CF85B15, 85% conventional fertilization plus 15% PSB-based fertilization; CF80B20, 80% conventional fertilization plus 20% PSB-based fertilization. Data are expressed as means ± SD (n = 3). For each tube type, bars labeled with different lowercase letters (e.g., a, b, c, etc.) indicate significant (p < 0.05) differences among the treatments based on the Tukey test or Games-Howell test.

Conclusions
Using an in situ sequential soil coring method, we found that the input of chemical fertilizer, mineralization, immobilization, and loss played an important role in determining the concentrations of NH4 + -N, NO3 -N, and Olsen-P in vegetable farmland soil during a short-term field investigation, which was more pronounced in NH4 + -N and NO3 − -N than Olsen-P. The results from this study indicate that, despite chemical fertilizer reduction, the partial substitution of chemical fertilizer with PSB at a relatively low application rate (e.g., 1840 kg ha −1 ) may not substantially reduce the soil available NO3 − -N and Olsen-P for the plant to some degree, and could contribute to the sustainable availability of N and P  In the soil ecosystem, biochar is conducive to the sustainable development of agriculture via balanced soil C:N:P stoichiometry. For example, a recent three-year pot study has shown that chemical fertilizer increased the imbalance between soil C and N stoichiometry, whereas peanut shell biochar at the annual application rate of 1800 kg ha −1 reduced this imbalance [10], thereby contributing to the production of ecosystem services (e.g., nutrient cycling that supports soil fertility and boosts crop production). The soil C:N:P ratio can vary in response to the type of soil, chemical fertilization rate, biochar type and application rate, and residence time of biochar in soil [10,65,66]. Despite discrepancies among stoichiometric studies regarding the effects of biochar, chemical fertilizer, and their interactions on soil

Conclusions
Using an in situ sequential soil coring method, we found that the input of chemical fertilizer, mineralization, immobilization, and loss played an important role in determining the concentrations of NH 4 + -N, NO 3 − -N, and Olsen-P in vegetable farmland soil during a short-term field investigation, which was more pronounced in NH 4 + -N and NO 3 − -N than Olsen-P. The results from this study indicate that, despite chemical fertilizer reduction, the partial substitution of chemical fertilizer with PSB at a relatively low application rate (e.g., 1840 kg ha −1 ) may not substantially reduce the soil available NO 3 − -N and Olsen-P for the plant to some degree, and could contribute to the sustainable availability of N and P in vegetable farmland soil via a variety of transformation processes, such as mineralization, immobilization, and loss. It provides an effective approach to understanding the availability and transformations of N and P in subtropical acid soils. There is a need for long-term studies evaluating the potential of the partial substitution of chemical fertilizer with PSB, especially in fields under continuous application.

Supplementary Materials:
The following are available online at https://www.mdpi.com/article/10 .3390/agriculture11100979/s1, Table S1. Analytical methods used to determine the main physicochemical properties of the vegetable farmland soil and palm silk biochar (PSB) used in the present study. Table S2. Analytical methods used to determine the main physicochemical properties of the compound fertilizer and calcium superphosphate used in the present study. Table S3. Soil moisture content (dry basis) in 0-40 cm soil layer in three types of tubes. Figure S1. Effect of palm silk biochar (PSB) amendment on pH water in soil without chemical fertilizer addition. Treatments: Control, no fertilization; B 10 , 10% PSB-based fertilization; B 15 , 15% PSB-based fertilization; B 20 20% PSB-based fertilization. Data are expressed as means ± SD (n = 3). Bars labeled with different lowercase letters (e.g., a, b, c, etc.) indicate significant (p < 0.05) differences among the rates of PSB application alone based on the Games-Howell test. Data Availability Statement: All supporting data have been included in this study, and are available from the corresponding authors upon request.