Microbial diversity and physicochemical properties in farmland soils amended by effective microorganisms and fulvic acid for cropping Asian ginseng

Demand for products made from the dry mass of Asian ginseng (Panax ginseng) is growing, but harvest is limited by fungal disease infection when ginseng is replanted in the same field. Rotated cropping with maize can cope with the replant limit, but it may take decades. We aimed to amend post-maize-cropping farmland soils for cultivating Asian ginseng, using effective microorganisms EMs and fulvic acid (FA) additives and detecting and comparing their effects on soil microbial diversity and physiochemical properties. Amendments promoted seedling survival and depressed disease-infection. Both EMs and FA increased the relative abundances of Pseudomonas, Flavobacterium, Duganella, and Massilia spp., but, decreased the relative abundances of Fusarium and Sistotrema. In addition, soil nutrient availability and properties that benefitted nutrient availabilities were promoted. In conclusion, amendments with EMs and FA improved the fertility of farmland soils, and the quality of Asian ginseng, and revealed the relationship between soil microbial diversity and physiochemical properties.

the allelopathic inhibitory effects from replant cropping except for employing rotation cycles for decades Jiao et al., 2019). An upgraded culturing protocol is needed for sustainable Asian ginseng cultivation with a shorter rotation cycle.
Soils for growing Asian ginseng require a desired soil quality, high microbial diversity ( Xiao et al., 2016;Zhao et al., 2017;Tong et al., 2021), and a high nutrition input (Beyaert, 2006;Kang et al., 2016) for significant production and expected seedling growth. Remnant chemicals such as fertilizers and pesticides are commonly seen in farmland soils (Li et al., 2015). This stimulates uncertainties in soil quality and microbial diversity and even cause failure of establishment, or impact growth and development of Asian ginseng stocks. One of the negative consequences on Asian ginseng is deteriorating diseases that occur in farmland soils from some combined effects of microbial communities and physicochemical properties Wang et al., 2016). To improve farmland soils for optimizing field performance of Asian ginseng, adding amendments is suggested (Sun et al., 2015). However, it is unclear how microbial communities and physicochemical properties respond to the amendment strategy for Asian ginseng cultivation. Information on the effects of different amendment strategies on the inner correlation between microbial diversity and physicochemical properties is scarce.
Deep composting is a general strategy to cope with the negative effects of substrates caused by unknown microbial communities and chemical remnants. Effective microorganisms (EMs) are an efficient compound microbial reagent that accelerates decomposition and transformation of organic matter in soils. Empirical practices reveal that additives with EMs can increase the number of soil microbes, activate soil enzymes, control pedosphere diseases, and promote plant growth (Borowiak et al., 2021). Additionally, they contribute to amended soils in organic composting (Park, 2011), fruit and vegetable cultivation (Tommonaro et al., 2021), and chemical degradation ( Xing et al., 2007;Kim et al., 2011). These together suggest the potential of using additives with EMs to optimize farmland soils for Asian ginseng planting. To our knowledge, however, trials on farmland soils that were amended by EM additives for Asian ginseng cultivation are highly rare. The role and efficacy of EMs amendment in improving soil properties, enzyme activities, microbial diversity, nutrient absorption, and chemical reducing in Asian ginseng soils need to be further studied.
The other strategy to cope with farmland soils that have been negatively impacted is the optimization of soil structure by adjusting physicochemical properties to a desired level. Fulvic acid (FA) widely exists in nature as kinds of combined humic acid and water-soluble polymer (Gong et al., 2020). FA contains more than 70 minerals and nucleic acids, soluble sugars, amino acids, peptides, phytochemical compounds, and vitamins (Malan, 2015). Most FA exist in an ionic form, suggesting a strong physiological activity with absorption, exchange, complexation and other functions (Marosz, 2009). Therefore, FA was widely used as an amendment for soils in agriculture, forestry, animal husbandry, and environmental protection (Aydin et al., 2017).
Amendment with FA can improve soil quality by promoting aggregate structure, nutrient utilization, and soil organic matter (SOM) accumulation. These will further provide a carbon (C) source for microbial growth and stabilizing soil temperature to enhance enzyme activity by adjusting soil pH and motivating microbial activities (Cheng et al., 2013;Kotroczó et al., 2014). Some bioorganic fertilizers were developed on the basis of FAsynthesized substrates, which were found to enhance hydrolyzation of alkali-hydrolysable nitrogen (N), available phosphorus (P), and soluble potassium (K) and activate enzymes, rhizosphere soil bacteria, and actinomycetes (Abdel-Baky et al., 2019;Li et al., 2021). Again, few studies have tested the application of FA amendment for Asian ginseng soils.
Increasing evidence indicates that soil microbes are an important and active component that closely relate to soil fertility and health (Hadar and Papadopoulou, 2012). The diversity and contents of microbial communities can directly reflect soil health and function (Visser and Parkinson, 1992). Soil microbes play an important role in soil formation (Pratscher et al., 2011), development, material transformation (Jackson et al., 2012), nutrient cycles (Kuzyakov and Xu, 2013), organic matter transformation, pollutant degradation, and environmental purification (Cupples, 2005). The increase of beneficial microbes can promote nutrient uptake (Guiñazú et al., 2010) and enhance resistance to disease (Cummings, 2009). Growth, development, and 3 productivity of plants interplay with soil microbial communities (Chaparro et al., 2012). Soil amendment with additives will break the trade-off between plant and soil and form a new relationship between microbial diversity and physicochemical properties. Thus, one approach to meet Asian ginseng' cultivation target is to determine, in amended soils, microbial diversity and its relationship with physiochemical properties so soil amending regimes can be compared and evaluated.
In this study, farmland soils following maize rotation were amended with EMs and FA for cultivation of Asian ginseng stocks. Soil microbial composition and dominant population were quantified by highthroughput sequencing technology. Hence the relationship between soil microbial diversity and physicochemical properties can be detected. We hypothesized that: (i) farmland soils amended with EMs and FA will show a contrasting difference in composition and dominant population, with (ii) soil physicochemical properties significantly changed, and (iii) all these responses will synchronize as correlations between at least five pairs of variables in microbial diversity and physicochemical properties. Accordingly, it will be of great interest to further study the practice of amendments with EMs and FA as candidate strategies to improve farmland soils for better Asian ginseng production.

Site description
The experiment was conducted in a field of farmland that was prepared to cultivate Asian ginseng. The site was located in the eastern montane area. It had an average altitude of 672 m (43°41′-43°58′ N, 128°10'-128°32′ E) in Jilin Province, Northeast China. It has a temperate continental monsoon climate and four distinct seasons, with an average annual temperature of 2.6 °C, rainfall of 618 mm, relative humidity of 68%, and wind velocity of 2.7 m·s -1 . The land was rotated with maize before the commencement of this experiment, and the farmland was left fallow for one year.

Experimental design and farmland soil amendment
The study began on 20 April 2019. The farmland used had a total area of 500 m 2 (100 m × 5.0 m). It was divided into three plots, each with an area of 150 m 2 (100 m × 1.5 m), for Asian ginseng cultivation. Each plot was further divided into three operational blocks as three random replicates of sampling units. Each block had an area of 45 m 2 (30 m × 1.5 m) and a buffer (5.0 m × 1.5 m). Deep tillage was employed for all plots to a depth of 25 cm by a rotary cultivator before soil amendment. Adjacent cultivating areas were buffered by a lined land without any cropping. The first plot was used as a control (CK) without amendment. The second plot was designed to be amended by EMs (C1). EMs contained more than 100 billion live bacteria per gram (Organic-Biotech Inc., Luoyang, Henan, China). The third plot was amended by FA (C2) (Macklin Inc., Shanghai, China). Three weeks after the first rotary operation, additives of EMs and FA evenly rose to surface soil at a rate of 1.5 g·m -2 . In the second rotation, additives were given to soil 25 cm belowground to mix amendments with the soil. These practices were repeated once every four weeks for a total of three times up to late September 2019.
Asian ginseng seedling cultivation Asian ginseng seeds were obtained from a seed source in Fusong County, Jilin Province, China. This was run as a ginseng orchard of National and Local Joint Engineering Research Center for Ginseng Breeding and Application (43°48' N, 125°24' E), Jilin Agricultural University, Changchun City, Northeast China. Seeds were collected on 10 August 2019, treated by the Hierarchic Optimization Method to facilitate seeding. Seeds were reserved in washed sands at room temperature for 30 days. On 25 March 2020, seeds were germinated at an average temperature of 24 °C with relative humidity ranging between 49% and 89%. On 5 May 2020, germinated embryos were transplanted to the experimental plots of CK, C1, and C2. Embryos were directly 4 sown to the blocks which were subsequently shaded by sunshades placed 1.5 m above the ground. Two-layer fibers were used to shade ginseng seedlings during the culture. The outer layer was black porous shading fiber, and the inner layer was yellow-rainproof-plastic fiber film. This manufacture can allow sunlight illumination at a transmittance of about 35%. During the experiment, temperature was 11.8/31.5 °C (night/day) and the relative humidity ranged between 49% and 89%. Asian ginseng seedlings were planted in a spacing of 15 cm × 15 cm for five months. Therefore, a total of 2,000 Asian ginseng seedlings were planted in a block and 6,000 in a plot.

Seedling sampling and determination
On 10 October 2020, the experiment was terminated, and seedlings were sampled from CK, C1, and C2 plots. Twenty-five seedlings were randomly sampled from a block, bulked as a sampling unit, and measured for survival rate, infected rate, stem growth (length and diameter), leaf variables (length, width, area, and chlorophyll concentration), and root growth (length, diameter, and weight).

Soil sampling
The soil type used in the experimental farmland was dark brown soil. On 25 April 2020, the soils were randomly sampled at a depth of 5-10 cm belowground from five points along an "s" shape across a block and mixed to a block unit. This way, three composite samples were collected in each experimental field for repetition. Consequently, nine samples were collected in three fields. Visible plant and animal residues were carefully removed from the fresh soil samples. An amount of 1.0 kg samples was left according to the quartering method, moved into the sterile sampling tubes, sealed, and transported to the laboratory. Some samples were air dried at indoor temperature then screened through 2 mm sieve. Half of the dried soil samples were used for determining physicochemical properties. The rest were sent to the laboratory at 0-4 °C and stored at ultralow temperature (-80 °C) for high-throughput sequencing analysis.

Soil physicochemical properties analyses
A total of 5.0 g soil sample was measured for the weight after air-drying and the dried weight after being oven-dried at 105 °C then measured for dried weight. Soil water content was calculated using the difference between dried and fresh soil weights. Soil organic matter was determined using the K2Cr2O7 titration method. Hydrolyzed N was measured by the alkali hydrolysable method. Soil pH value was measured at a soil/water ratio of 1:2.5 (w/w) with a pH meter (PHSJ-6 L, INESA Scientific Instrument Inc., China). Electric conductance (EC) was measured at a soil/water ratio of 1:5 (w/w) with an EC meter (DDSJ-319 L, INESA Scientific Instrument Inc., China). Concentrations of available N, P, K, NH4 + -N and NO3 --N was determined spectrophotometrically following the methods used in previous studies (Triplett Jr. and Van Doren Jr., 1969;Wei et al., 2014;Zhu et al., 2016;Wei et al., 2017;Wei et al., 2019). Total N and P concentrations were determined through colorimeter analysis following digestion with H2SO4-H2O2. Total K concentration was determined using inductively coupled argon plasma-optical emission spectroscopy. Available P concentrations were extracted with the HCl/NH4F method. Samples were determined colorimetrically using ascorbic acid molybdate on a continuous flow autoanalyzer (Autoanalyzer III, Bran and Luebbe, Germany). Available K concentration was determined through NH4OAc extraction. K chloride extraction indophenol blue colorimetry was used for ammonia acetate extraction and Inductively Coupled Plasma Plus Optical Emission Spectrometry (ICP-OES) was used for determination of metal ions.
Soil urease activity (Guan, 1986) was determined through phenol sodium colorimetry, and enzyme activity was expressed by the milligram of NH4 + -N in an amount of 1.0 g soils after reacting in 24 h. Sucrase activity was determined by 3,5-Dinitrosalicylic acid colorimetry, and activity was expressed by the milligram of glucose in 1.0 g soil after 24 h. Catalase activity was determined by potassium permanganate titration. Copper blue oxidase was determined through 2,2-diazo-di (3-ethyl-benzothiazole-6-sulfonic acid) diamine salt (ABTS) 5 colorimetry. Enzyme activity can produce free radicals through the decomposition of substrate ABTS by laccase, and activity can be calculated by the increase rate of free radicals.
The PCR gel product was recovered by 2% agarose gel, purified by AxyPrep DNA Gel Extraction Kit (AxygenBiosciences, UnionCity, CA, USA), eluted by elution, and detected by 2% agarose gel electrophoresis. Quantifluor TM-st (Promega, USA) was used for detection and quantification. According to the standard operating procedures of Illumina Miseq platform (Illumina, SanDiego, USA), the purified amplification fragments were constructed into the library of PE2×300. Library construction followed four steps: first, connect the "Y" shaped connector; second, use magnetic beads to screen and remove the self-linked segments of the connector; third, use PCR amplification to enrich the library templates; finally, produce single strand DNA fragments with sodium hydroxide denaturation.

Bioinformatics analysis
The original sequences were controlled by TRIMMOMATIC software and spliced by FLASH software: (i) set up a 50 bp window, if the average quality in the window was less than 20, cut off all the sequences at the back end of the base from the front end of the window, and remove the sequences with a length less than 50 bp after quality control; (ii) spliced the sequences at both ends according to the overlapping base, where the maximum overlap fell in a range between overlaps during splicing, the mismatch ratio was 0.2 and the length should be more than 10 bp; (iii) according to barcodes and primers at both ends, sequences were divided into each sample. However, barcodes needed to match accurately. The primers allowed the mismatch of two bases, and removed the sequences with fuzzy bases. The UPARSE software (version 7.1) (UPARSE software, 2021) was used to cluster the sequences with operational taxonomic units (OTUs) according to 97% similarity, and to remove single sequences and mosaics in the process of clustering (Edgar, 2013). The RDP classifier was used to annotate the species classification of each sequence (Wang et al., 2007). The representative sequence was assigned to the Silva database (SSU123) (Pruesse et al., 2007). The threshold of comparison was set to 70% to obtain the classification information of bacterial community. Bacteria were classified according to an order of phylum, class, order, family and genus.

Statistical analysis
One-way analysis of variance (ANOVA) was used for the comparison of difference among three types of amendment treatments of CK, C1, and C2. Physicochemical properties of amended soils, soil enzyme activity, and seedling variables were analyzed using one-way ANOVA. Data were tested for normality and homogeneous variances. All of our data passed these two tests and no data transformation was needed. IBM SPSS Statistics 24.0 software (SPSS Inc., Chicago, IL, USA) was used to compare the above-mentioned variables, microbial diversity indices, and the relative abundances of soil microbial communities. The microbial relative abundances of CK, C1 and C2 were different at the genus level. They were then confirmed by a principle coordinate analysis, which was conducted on the basis of weighted UniFrac distance. Analysis of 6 similarity was used to depict differences in the compositions of the microbial community. The heatmap was generated using the plots package in R (version 2.15.3) to compare the top 45 bacterial and fungal genera in soil samples of CK, C1 and C2. Statistical analysis of soil factors and the top 20 microbial genera was used to screen soil variables with P > 0.05. The selected environmental factors, such as pH, SOM, iron, soil catalase, laccase, urease, and invertase, were used for correlation analysis. These correlation analyses were also performed to assess the relationships among soil properties, plant growth, and soil microbial community. Biodiversity for soil microbial community was assessed by Shannon index (Wang et al., 2018).

Soil physicochemical properties
Significant difference was found as responses in variables of SOM, pH, NH4 + -N, available P, Ca, Mg, Fe, and Zn under different amendment treatments ( Table 1). The concentrations of NH4 + -N, available P, Ca, Mg, and Zn were higher in the soils amended by FA than that in CK, but soil pH results were reversed. Note: The letters indicate the tested with shortest significant ranges at the P = 0.05 of different treatments. Different lowercase letters indicate that there were significant differences at the P < 0.05 level Also, as it is shown in Table 1, concentrations of soil NH4 + -N in the C1 treatment was 4.181 mg·kg -1 , which was 5.65 times higher than that in CK. Concentration of available P was 1.94 times higher than that in CK. Concentrations of available Ca, Mg, Zn, and SOM in the C1 treatment increased by 2.413 g·kg -1 , 290.270 mg·kg -1 , 13.480 mg·kg -1 , and 27.930 g·kg -1 , respectively, from those in CK. Soil bulk density was 0.94 g·cm -3 in the C2 treatment, which had no difference from that in the C1 treatment and CK. Concentration of NH4 + -N increased significantly in the C2 treatment compared to the C1 and CK. The available P concentration in the C2 treatment was 178.63 mg·kg -1 , which was 2.74 times higher than that in CK. Concentrations of available Ca, Mg, Zn, and SOM in the C2 treatment were 1.75 g·kg -1 , 226.8 mg·kg -1 , 15.55 mg·kg -1 , 30.59 g·kg -1 , respectively, and soil bulk density was 0.97 g·cm -3 .
In Table 1, it was also found that amendments of EMs and FA had significant effects on soil enzyme activities compared to CK. Enzyme activities of soil urease, laccase, catalase, and invertase were activated in the 7 C1 treatment compared to CK. Soil urease and laccase activities, however, were significantly increased in the C2 treatment. Enzyme activities of soil catalase and invertase were lower in the C2 treatment than in CK.

Soil microbial compositions
After taking the same size from each sample to the same sequencing depth (61562 reads per sample) and clustering, a total of 298,014 OTUs with 97% identity were obtained, and the number of OTUs per sample fell in a range between 23,181 and 47,569 as illustrated in Figure 1. Subs index of OTUs increased to be near an upper limit, and the goodness of fit reached 99.46%. This suggests that the detection rate of the microbial communities in soil samples was increased with sampling numbers up to saturation. In addition, the current sequencing amount covered most species in these samples. The curves of Shannon index tended to be flat, and the amount of sequencing data was large enough to reflect the vast majority of microbial information in these soils.
As it is shown in Table 2, the diversity indices of soil bacterial community were different in CK, C1 and C2. Shannon index was lower in the C2 treatment than that in the C1 treatment and CK. In contrast, Simpson index was higher in the C2 treatment than that in the C1 treatment and CK. Both Ace and Chao indexes were lower in the C2 than that in C1 and CK. Results of these two different algorithms verified or corroborated each other. Table 2. Diversity index of soil bacterial community in the experimental farmland (CK, C1 and C2) Note: The letters indicate the tested with Shortest Significant ranges at the P = 0.05 of different treatments. Different letters denote a significant difference at the P < 0.05 level. Table 3, the diversity indices of soil fungal community were different among CK, C1 and C2. Compared to CK, Shannon index was lower in the C1 treatment, with no significant difference between C2 and CK. Simpson index was higher in C1 compared to that in CK and C2. The diversity of soil fungal community was highest in CK and lowest in C1. Similar findings were reflected by the two algorithms, which were verified by each other. The Shannon index was unchanged in C2 from that in CK, but both were higher than that in C1. In contrast, Simpson index was higher in C1 than in CK and C2. The indexes of Ace and Chao did not show significant difference among the three types of amendment treatments.

As was listed in
In this study, non-metric multidimensional scaling (NMDS) analysis method was used on the diversity of soil microbial communities. As seen in Figure 2, there was a significant difference in soil microbial communities between CK and the C1 and C2 treatments. In bacterial and fungal compositions, soil microbial community was affected mainly by EMs and FA additives.  As was described in Figure 3, the heatmap showed the composition analysis of bacteria community in CK, C1, and C2 treatments. A total of forty-five genera were detected, and the black colour in the diagram represents the relative abundance of a certain genus. The relative abundances of norank_p_Saccharibacteria, Sphingomonas, and norank_f_Acidobacteriaceae_Subgroup_1_ accounted for 10%, 9%, and 7%, respectively, while those of norank_o_Gaiellales, Rhodanobacter, Rhizomicrobium, norank_f_ODP1230B8.23, and Bradyrhizobium accounted for 4% in total. The relative abundances of Pseudomonas, Roseiflexus, norank_f_Anaerolineaceae, norank_o_Subgroup_7, RB41, Pedobacte, Massilia, Flavobacterium, Duganella, and other bacteria were lower in the bacteria community of CK. Compared to CK, the relative abundances of Mizugakiibacter, Mizugakiibacter, Pseudomonas, RB41, Duganella, and others were significantly higher in C1, with the increases of Massilia, norank_c__Acidobacteria, norank_o__Subgroup_7. The relative abundance of Duganella was significantly lower, and those of Rhodanobacter, Granulicella, norank_f__Ktedonobacteraceae were slightly lower. Compared with C1, the relative abundances of Acachidicoccus, Pseudomonas, Mizugakiibacter and Arthrobacter, as well as Rhodanobater, were significantly higher in C2.  The heatmap (Figure 4) showed the compositions of fungi community in CK, C1, and C2. A total of forty bacteria genera were detected, in which the color block represents the relative abundance of a certain genus. The relative abundance of Sistotrema in CK was 19%, while those of Humicola, unclassified_o__Trechisporales, Fusarium, Mortierella, Chaetomium, Guehomyces, and Conocybe were 16%, 13%, 11%, 8%, 6%, 4%, and 4%, respectively. The relative abundances of Coprinopsis, Sporobolomyces, Myrothecium, Articlesopora, Metarhizium, Monographella, Pseudoeurotium, and Mrakia were lower compared to CK. The relative abundances of Chaetomium, Knupia, Cladophialophora, Neopeckia, Talaromyces, Sistotrema, Fusarium, and Conocybe decreased significantly, while those of Humiola, Chaetomium and Pseudoeurotium increased. Compared with C1, the relative abundance of Pseudoeurotium did not change, except for an insignificant decrease. The relative abundances of Fusarium, Chaetomium, Knutia and other fungal community decreased more significantly, while Humiola and Chaetomium increased more. The cluster tree of species and environmental factors can be selected. The correlation R value and P value are obtained by calculation, the R value is shown in different colours in the figure, if the P value is less than 0.05, it is marked with * and the legend on the right is the colour range of different R values, e.g. the symbol * represents the P value at 0.01 < P ≤ 0.05 level, the symbol * * represents the P value at 0.001 <P ≤ 0.01 level, the symbol * * * represents the P value at P ≤ 0.001 level. SOM: soil organic matter, SU: soil urease, SS: soil sucrase, CAT: soil catalase, L: soil laccase Figure 5 showed the relationship between soil physicochemical properties and bacterial community compositions across amended farmland soils. SOM negatively correlated with the relative abundances of Bradyrhizobium, gemmatimonas, norank_f__gemmatimonadaceae; iron concentration negatively correlated with Massilia community abundance and positively correlated with norank_f__ODP1230B8.23 community abundance; pH value negatively correlated with Pseudonas community abundance and positively correlated with community abundances of Bradyrhizobium, Gemmatimonas, Sphingomonas, norank_o_Gaillales, and norank_f__Gemmatimonadaceae. However, soil urease negatively correlated with Bradyrhizobium and positively correlated with Massilia. Soil invertase positively correlated with Massilia, negatively correlated with Bradyrhizobium, and had a significant negative correlation with norank_f__ODP1230B8.23. In addition, a negative correlation was found between soil catalase and community abundances of Archidicoccus, Mizugakiibacter, Rhodanobacter, and norank_f__Chitinophagaceae. A positive correlation was found between soil catalase and community abundances of norank_c__Acidobacteria and norank_f__DA101. Soil laccase had a negative relationship with community abundances of Bradyrhizobium, Bryobacter, Rhizomicrobium, norank_o__Gaiellales, and norank_f__ODP1230B8.23. Soil laccase also had a positive correlation with community abundances of Massilia and Pseudomonas. The cluster tree of species and environmental factors can be selected. The correlation R value and P value are obtained by calculation, the R value is shown in different colours in the figure, if the P value is less than 0.05, it is marked with * and the legend on the right is the colour range of different R values, e.g. the symbol * represents the P value at 0.01 < P ≤ 0.05 level, the symbol * * represents the P value at 0.001 < P ≤ 0.01 level, the symbol * * * represents the P value at P ≤ 0.001 level. SOM: soil organic matter, SU: soil urease, SS: soil sucrase, CAT: soil catalase, L: soil laccase 12 As shown in Figure 6, SOM negatively correlated with community abundances of Chaetomium, Conocybe, Fusarium, Knufia, unclassified_f__Lasiosphaeriaceae, and unclassified_o__Trechisporales, and with community abundances of Cladophialophora and unclassified_p__Ascomycota. Iron concentration negatively correlated with community abundances of Sistotrema and Trichoderma. Soil pH value positively correlated with community abundances of Chaetomium, conocybe, Cladophialophora, Fusarium, Knufia, unclassified_f__Lasiosphaeriaceae, unclassified_o__Trechisporales, and unclassified_p__Ascomycota, and negatively correlated with Humicola community abundance. In addition, soil invertase positively correlated with Trichoderma community abundance and negatively correlated with community abundances of Chaetomium, Conocybe, Sistotrema, and unclassified_o__Trechisporales. Soil catalase positively correlated with community abundances of Cryptococcus and Mortierella. It negatively correlated with unclassified_f__Chaetomiaceae community abundance. Soil laccase positively correlated with community abundances of Cladophialophora, Humicolaand, and Fusarium and negatively correlated with Conocybe community abundance.

Performance of Asian ginseng seedlings
As it is shown in Table 4, variables of seedling growth, as well as morphologies in root, stem, leaves, and leaf chlorophyll concentration, all improved in the C1 and C2 treatments compared to CK. In addition, disease rate of Asian ginseng seedlings decreased in soils with added amendments compared to that in the control. Survival rate was increased by soil amendments. Note: The significance of differences between different treatments is indicated by lowercase letters (P<0.05). Values followed by the same letter are not statistically different according to the independent t-test at P < 0.05.

Discussion
Soil microbes are the most active part, and an important part of material metabolism cycles (Bending et al., 2000;Wang et al., 2018). They not only participate in various metabolism activities as executor to shape SOM transformation (Prescott and Vesterdal, 2021), but also transform and hydrolyse biological residues and organic matters. These together promote the increase of soil nutrients. Moreover, the metabolisms of soil microbes are the main source of soil enzyme, and the correlation of nutrient transformation in soil enzymes. Therefore, it was speculated that soil microbes played an important, indirect role in the increase of soil fertility and ecosystem, and their diversity and activities were closely related to soil physicochemical properties. In this study, the modifiers of FA and EMs were added to farmland soils in order to investigate the response of soil microbes and their relationships with physiochemical properties. Overall results showed that bacterial and 13 fungal communities obviously changed in amended farmland soils. Soil microbes and physiochemical properties interplayed to promote change in soil quality.
Soil enzymes are the products of the decomposition of animal and plant residues, plant root exudates, and soil microbial metabolism. They are a kind of substance with biochemical catalytic activities and participate in many important biochemical processes in soils. These processes closely relate to the release and storage of various nutrients, the formation and development of humus in soil, and the soil structure and physicochemical properties. They are involved in the whole process of soil genesis and development, as well as the formation and evolution of soil fertility. Therefore, soil enzyme activity directly reflects changes in soil structures and compositions. The larger the index of soil enzyme is, the stronger its activity is and the stronger the soil fertility is. These results were consistent with the significant increase of soil nutrients, which also indicated that EMs and fulvic acid had significant effects on soil structures and fertility improved farmland soil.
Amendment with EMs had different effects on soil nutrients and enzyme activities. This variation of changes resulted from increases in concentrations of Ca, P, and N, activities of soil urease and sucrase, and soil pH value. These results showed that the EMs additive can improve soil physicochemical properties, available N and P, and the activities of soil enzymes through the physiological activities of microbes due to EMs containing phosphate solubilizing bacteria and nitrogen fixing bacteria (Higa and Wididana, 1991). Adding FA to the farmland soils was more effective than EMs in promoting soil quality with increasing concentrations of available P, hydrolysed N, and SOM, and reducing pH value. No changes in soil moisture content and bulk density were observed in this study. However, it was found that soil pH value and concentrations of SOM, iron, sucrase, laccase and catalase activities were closely related to the diversity of soil bacterial and fungal communities. These results suggested that organic matters that were chelated with iron were decomposed by soil bacteria and fungi at higher pH, which did not depend on additional water absorption (Smith and Paul, 1990).
It was also found that diversity of beneficial bacterial community had increased after farmland soils were amended. Parts of original components in bacterial community had changed, such as Pseudomonas, Flavobacterium, Duganella and Massilia. Other beneficial bacterial community with extremely low original relative abundances had significantly increased after the amendment. This was extremely significance to the changes of soil physicochemical properties and the strengthening of biological control of plant diseases.
Pseudomonas is a kind of bacteria that plays an important role in biological control. The secondary metabolites of Pseudomonas can dissolve the soil Al (Saleh et al., 2019), which is one of the greatest factors that cause Asian ginseng red skin disease (Liu et al., 2014). The reduction of soil Al concentration possibly reduced the occurrence of Asian ginseng red skin disease. These results suggested that Pseudomonas had a great inhibitory effect on a variety of soil borne plant disease bacteria (Nagel et al., 2012;Iglesias et al., 2018;Aiello et al., 2019). The existence of Flavobacterium spp. plays an important role in soil microbes and plant growth (Soltani et al., 2010). Results in the study revealed that Flavobacterium spp. increased concentrations of C and N in soils amended with EMs and FA. These changes promote the proliferation of soil microbes and plant growth (Aiello et al., 2019;Herrera et al., 2019). Duganella spp. are an important genus of Fusarium resistant bacteria, which can effectively reduce plant diseases caused by a variety of Fusarium spp. (Haack et al., 2016). Massilia spp. is a kind of promising genus of bacteria in improving the soil farmland of cultivated Asian ginseng, and that also has an excellent ability of dissolving P fractions . This can increase the available P concentration, and produce dimethyl disulfide which is expected to replace methyl bromide fumigant (Feng et al., 2016). Dimethyldisulfide has a significant effect on controlling soil pathogenic fungi and nematodes, while reducing the levels of pathogenic bacteria, soil fumigant, and pollution. In this study, EMs and FA changed the diversity of soil bacterial community in different degree. After adding FA to the farmland soils, the relative abundances of communities, such as Achidicoccus, Pseudomonas, Mizugakiibacter, and Arthropcharacter spp., as well as the relative abundance of Rhodanobacter, all increased accordingly. The enrichment of Rhodanobater spp. can improve the denitrification ability of soil (Green et al., 2012), strengthen the supply of soil microbes and plant N element, and promote the growth of plant and the increment of soil microbes.
14 The analysis of soil fungal diversity showed that farmland soils amended with EMs and FA could reduce the diversity of harmful fungal community. After soil amendment, parts of the structures and compositions of the original microbes were changed significantly. Furthermore, the relative abundances of harmful fungi, such as Sistotrema spp. and Fusarium spp., reduced due to EMs additions. This accounted for the decline in soil borne diseases (Higa and Wididana, 1991). Fusarium spp. are a common plant pathogen, which can cause a variety of crop diseases (Chen et al., 1995;Haack et al., 2016). The increase in the relative abundance of Dunganella spp. was likely due to soil amendment. This inhibited the propagation of Fusarium spp. Meanwhile, Sistotrema spp. have a bacteriostatic effect, and mycelia have antagonistic effects on Bacillus Subtilis and six other kinds of microbes (Haack et al., 2016). The decrease of its relative abundance was beneficial to the growth of bacteria and the increase of soil fertility. In this study, the community abundances of Fusarium, Chaetomium, Knupia, and other fungal community decreased significantly, and those for Humicola and Chaetomium increased in amended soils compared to controlled soils. According to the correlation analysis of soil fungi and physiochemical properties, community abundances of Fusarium, Chaetomium and Knunia spp. had a significant positive correlation with soil pH, but, Humicola and Chaetomium had a significant negative correlation with soil pH. Accordingly, the changes of fungous community were caused by fulvic acid reducing the soil pH.
Soil pH is an important soil chemical factor affecting the diversity of soil bacteria (Landesman et al., 2014). When soil pH was lower than 6.5, a significant negative correlation between soil bacterial diversity and soil pH will occur . In the present study, the amended soils were slightly acidic, which were prone to enable the co-existence of multiple bacteria genera. The beneficial bacteria, such as Bradyrhizobium, Gemmatimonas, and Sphingomonas spp., had a significant positive correlation with pH values in amended soils. Soil pH may not directly change the structures and compositions of bacterial community, but it may impose a force that results in overall changes in the soil environment. Several soil properties were directly or indirectly affected by the change in pH value. These soil variables included, but were not limited to, concentrations of available nutrients, cation exchange, SOM, and factors that will also change the diversity of soil bacterial community (Lauber et al., 2009). Besides, soil compositions had a significant influence on microbial communities (Girvan et al., 2003), of which pH was thought to exert primary domination on the compositions of soil bacterial community (Landesman et al., 2014). Therefore, diversity of soil bacterial community can be regulated by adjusting pH in agricultural production.
Diversity of soil fungal community was affected by environmental factors. According to the results of correlation analysis, soil pH was an important factor affecting fungal diversity. There was a significant positive correlation between various fungi, such as Fusarium, and pH, which indicated that fungi were more suitable for survival in acidic soil, and acidic soil would increase fungi reproduction. The results also showed that Fusarium, Conocybe, Cladophilophora, and other fungi had pathogenic effects on plants with a positive correlation with SOM, soil pH, and laccase activity. These soil variables played the same role to depress infection by fungi on Asian ginseng. Some fungi having pathogenic effects may have some physiological connections to some extent and the same response to a certain factor. Consequently, some variables can be adjusted to control the occurrence of soil pathogenic fungi in agronomy production.

Conclusions
In this study, we found that amendments with EMs and FA have unique effects on microbial structures and physiochemical properties in farmland soils planting Asian ginseng. The relative abundances of several communities of beneficial bacteria were changed by amendments, such as Pseudomonas, Flavobacterium, Duganella, and Massilia spp. Meanwhile, the relative abundances of pathogenic fungi, such as Fusarium and Sistotrema spp., were decreased by amendments. Changes in soil physiochemical properties and microbial community structure did not vary between treatments of amendments by EMs vs FA additives. Hence, two types of amendments had a comparable impact on modifying microbial activities but varied in specific functions. EMs can increase beneficial bacteria diversity and inhibit the rate of fungal disease, while FA can enhance soil structure and availability. Overall, the use of EMs and FA as additives to amend maize-rotated soils can be an available approach for replanting Asian ginseng by improving biotic and abiotic properties in farmland soils. The combination of these two additives together may further result in a better outcome, but confirmative results need more work to be identified.
All authors read and approved the final manuscript.