Figures
Abstract
Soil microbial communities are affected by interactions between agricultural management (e.g., fertilizer) and soil compartment, but few studies have considered combinations of these factors. We compared the microbial abundance, diversity and community structure in two fertilizer dose (high vs. low NPK) and soil compartment (rhizosphere vs. bulk soils) under 6-year fertilization regimes in a continuous garlic cropping system in China. The soil contents of NO3– and available K were significantly higher in bulk soil in the high-NPK. The 16S rRNA gene-based bacterial and archaeal abundances were positively affected by both the fertilizer dose and soil compartment, and were higher in the high-NPK fertilization and rhizosphere samples. High-NPK fertilization increased the Shannon index and decreased bacterial and archaeal richness, whereas the evenness was marginally positively affected by both the fertilizer dose and soil compartment. Soil compartment exerted a greater effect on the bacterial and archaeal community structure than did the fertilization dose, as demonstrated by both the nonmetric multidimensional scaling and redundancy analysis results. We found that rhizosphere effects significantly distinguished 12 dominant classes of bacterial and archaeal communities, whereas the fertilizer dose significantly identified four dominant classes. In particular, a Linear Effect Size analysis showed that some taxa, including Alphaproteobacteria, Rhizobiales, Xanthomonadaceae and Flavobacterium, were enriched in the garlic rhizosphere of the high-NPK fertilizer samples. Overall, the fertilizer dose interacted with soil compartment to shape the bacterial and archaeal community composition, abundance, and biodiversity in the garlic rhizosphere. These results provide an important basis for further understanding adaptive garlic-microbe feedback, reframing roots as a significant moderating influence in agricultural management and shaping the microbial community.
Citation: Zhou J, Kong Y, Zhao W, Wei G, Wang Q, Ma L, et al. (2021) Comparison of bacterial and archaeal communities in two fertilizer doses and soil compartments under continuous cultivation system of garlic. PLoS ONE 16(5): e0250571. https://doi.org/10.1371/journal.pone.0250571
Editor: Mehdi Rahimi, Graduate University of Advanced Technology, Kerman Iran, ISLAMIC REPUBLIC OF IRAN
Received: July 7, 2020; Accepted: April 12, 2021; Published: May 14, 2021
Copyright: © 2021 Zhou et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the paper and its Supporting Information files.
Funding: This work was supported by the National Natural Science Foundation of China in the form of a grant awarded to JZ (No.41807053), the Young Talents Invitation Program of Shandong Provincial Colleges and Universities in the form of a grant awarded to JZ (No.2019-6-1), the National Natural Science Foundation of China in the form of a grant awarded to FS (No. 31672255) and the Key Research and Development Project of Shandong Province (Public Welfare Science and Technology Research) in the form of a grant awarded to LM (2019GNC106092).
Competing interests: The authors have declared that no competing interests exist.
Introduction
Many studies have confirmed that the rhizosphere microbiota is a subset of the entire bulk soil community [1]. However, the continuous cropping of Dihuang (Rehmannia glutinosa) [2], strawberries (Fragaria ananassa) [3], and maize (Zea mays L.) [4] has different effects on the bacterial communities in different soil compartment (bulk and rhizosphere) due to the additional complexity resulting from roots. Furthermore, changes in nutrient availability resulting from fertilization regimes can affect the rhizosphere microbiome by altering the root morphology and root exudation [5]. Previous studies have focused on the effects of fertilization regimes on various microbial groups in soil, including bacteria [6], fungi [7] and functional microbiota (e.g., ammonia-oxidizing archaeal communities) [8]. However, the combined influence of fertilization regime and rhizosphere effects on bacterial communities is not well understood.
Garlic (Allium sativum L), which is used worldwide as a flavouring agent and cover crop, is an important vegetable and medicinal plant [9]. China ranks first worldwide in terms of cultivated area (7,914,000 hectares) and output (2,005.84 million tonnes) of garlic (Food and Agriculture Organization, 2017). Garlic root exudates transfer allelochemicals to the soil environment [10] and selectively re-establish microbial communities in the rhizosphere, which results in the creation of taxonomically distinct microbial communities from bulk soil [11]. Relevant data have shown that the number of beneficial microorganisms and the level of enzyme activity in the soil initially increase and then decrease with continuous garlic cropping years [12]. After 15–20 years of continuous garlic cropping, various factors in the soil system are disordered, resulting in an imbalance in the soil microbial population structure [12]. In addition to an increase in soil fungi, other positive indicators of the growth and development of garlic, such as soil microbial species and soil ammoniated bacteria, show downward trends, which results in serious obstacles to continuous cropping and annual reductions in the garlic yield [13]. It is necessary to explore how fertilization regimes and plant selection affect rhizosphere bacterial and archaeal communities and to consider how they interact in combination rather than individually. However, few studies have focused on the effects of the rhizosphere and fertilizer regimes on bacterial and archaeal communities in continuous garlic cropping systems.
The aims of this study were to (i) compare the changes in the compositions of bacterial and archaeal communities under different doses of chemical fertilizer; (ii) compare the changes in the bacterial and archaeal community compositions of the bulk soil and garlic rhizosphere; and (iii) determine the main factors shaping the bacterial and archaeal community structure.
Methods
Ethics statement
This study was carried out on private land and the owner of the land gave permission to conduct the study on this site. The study was observational, involving no cruelty to animals, no damage to habitats and no harm to endangered plants, and thus no review from the ethnic committee was required in China. All the work was carried out under the Wildlife Protection Law of the People’s Republic of China.
Sample collection
The site of the garlic test field was in Gaohe town, Jinxiang County, Shandong Province, China (35°05′11″N, 116°22′38″E) and was established in 2010. The soil is a typical calcareous fluvo-aquic soil (aquic inceptisol) of the North China Plain, consisting of a calcareous, fluvo-aquic sandy loam [14]. The initial soil nutrient status before the experiment was analysed at the Shandong Engineering and Technology Research Center for Garlic (Jining, Shandong Province, China), and the results are shown in Table 1. One garlic cultivar, ‘Caijiapo Red Skin’ (G026), was grown in the field starting in early September 2015 and ended in late July 2016. Considering it is impossible to grow garlic locally in the region without applying fertilizer, five experimental replicates were conducted using two fertilization treatments in a randomized complete block design. There is a recommended dose of Nitrogen (N), phosphorus (P) and Potassium (K)fertilizers of garlic for towns in Jinxiang County (such as Huayu, Yushan and Mamiao): N2 = 450 kg/ha, P2O5 = 240 kg/ha, K2O = 360 kg/ha [15]. However, there is no recommended dose in Gaohe town, and High and Low doses were selected according to the amount used by local farmers. The 36-m2 (9 m × 4 m) test field received two fertilizer treatments refer to the local fertilization amount of NPK fertilizer: a low dose of NPK fertilizer (kg/ha) [135 N (urea), 76.5 P2O5 (Calcium phosphate) and 135 K2O (potassium sulphate)] and a high dose of NPK fertilizer (kg/ha) (270 N, 135 P2O5, and 270 K2O). Full does of calcium phosphate (to meet the phosphorus needs of garlic cloves during germination) andcalcium sulphate (to help garlic have a well-developed root system, strong stems and control against pythium root rot and soft rot [16]), 1/3 urea and potassium sulphate were applied as the basal fertilizer five days before planting, and the remaining 2/3 of urea and potassium sulphate were applied as a top dressing during the garlic bolting period. All other garlic management measures followed local practices.
In May 2016 (at the garlic bolting stage), six garlic root samples in each plot were randomly collected and pooled together as the rhizosphere soil sample, according to Mcpherson et al. [17]. The bulk soil was collected according to Zhou et al. [6]. The bulk soil samples from the high- and low-NPK treatments were denoted HO and LO, respectively, while the corresponding rhizosphere soil samples were named HR and LR. All 20 samples were placed into individually-labelled zipper storage bag and put in cooling container with ice (2–8°C) and transported to the laboratory. All soil samples were sieved (2 mm) to remove garlic roots and were evenly divided into two subsamples; one of the subsamples was air-dried, and the other was stored at −80°C for DNA extraction.
Soil chemical properties and garlic yield
The soil pH was determined by a pH electrode at a soil-to-water ratio of 1:2.5 [18]. The organic matter (OM) was determined from weight loss according to Parker [19], and the total nitrogen (total N) was determined according to Sims [18]. The soil KCl extractable nitrate (NO3−) and ammonium (NH4+) concentrations were determined according to Sparks et al. [20]. The soil available phosphorus (Avail P) was determined at a soil-to-sodium bicarbonate ratio of a 1:10 by Olsen’s procedure, and the available potassium (Avail K) was analysed after extraction at a soil-to-ammonium acetate ratio of 1:10 by an ion-exchange resin procedure [21]. The garlic yield was determined by weighing the collected bulbs, and this value was used to estimate the yield per hectare [22].
High-throughput sequencing analysis
The total genomic DNA from the soil samples was extracted using the PowerSoil DNA Extraction Kit (MO BIO Laboratories, Carlsbad, CA) according to the manufacturer’s recommended protocol [23]. The DNA was amplified with the primers 515F and 806R [24]. The reverse primer contained a unique 6-bp error-correction barcode for each sample. After library construction of the 16S rRNA gene fragments, high-throughput sequencing was performed on the Illumina MiSeq 2 x 250 platform at Shanghai Majorbio Bio-pharm Technology Co., Ltd., China. The original sequences are stored in the NCBI sequence read archive (accession number PRJNA531809).
Paired end reads were processed using the Quantitative Insights into Microbial Ecology (QIIME) software pipeline according to standard protocols [25]. Presumptive chimeric sequences were screened and discarded using UCHIME [26]. Relevant non-chimeric sequence groups were assigned to operational taxonomic units (OTUs) with a maximum classification distance of 3%. The α-diversity in each sample was calculated using the UPARSE pipeline [27]. The representative sequences of each subsampled OTU were classified and assigned using SINA aligner (version 1.1) [28] and the SILVA 16S rRNA database [29].
qPCR analysis
The abundance of bacterial and archaeal 16S rRNA was determined by ABI 7500 real-time PCR according to Zhou et al. [6]. Briefly, the PCR conditions were as follows: the enzyme was activated for 1 min at 95°C followed by 40 cycles of 15 s at 94°C, 34 s at 55°C and 15 s at 72°C. All PCR assays were performed using Maxima® SYBR Green/ROX qPCR Master Mix (Qiagen, Germantown, MD, USA), as specified by the manufacturer. Interesting gene fragments were cloned from the amplified 16S rRNA gene using the 515f/806r primers. The linearized plasmid was diluted 10 times to produce a standard template. The standard curve exhibited an R2 > 0.99.
Statistical analysis
We used the Kolmogorov-Smirnov test to determine whether the sample data were normally distributed before other analyses. The differences in the chemical properties, α-diversity indices and relative abundances of the dominant groups were tested using a one-way analysis of variance by using the "aov" function in R v. 3.6.1. Pearson correlation coefficients between the soil properties and bacterial and archaeal diversity and abundance were calculated using SPSS 19.1. In all trials, P < 0.05 was considered significant. The bacterial and archaeal genera that were differentially represented between the four soil sample types were measured by coupling the linear discriminant analysis (LDA) and with the effect size (LEfSe) [30, 31]. For analysis of the β-diversity, a nonmetric multidimensional scaling (NMDS) ranking of Hellinger distances was performed using the cmdscale function [32]. Permutational multivariate (999) analysis of variance (PERMANOVA) was performed to assess the differences in the pairwise combinations (HO-HR, HO-LO, HO-LR, HR-LO, HR-LR and LO-LR) to complement the ANOSIM results at the genus level. A redundancy analysis (RDA) was performed to explore the possible linkages between bacterial and archaeal communities at the genus level and soil properties using CANOCO 5.0 and a Monte Carlo test with 499 permutations.
Results
Soil chemical properties
The soil pH value ranged from 7.65 to 8.08 among the four experimental groups (Table 2), and there was no significant difference in soil pH among the experimental groups. However, the soil pH was lower at the end of the experiment (Tables 1 and 2); this phenomenon may be caused by the long-term accumulation of organic acids (i.e. acetic acid and maleic acid) and acidic amino acids (i.e. aspartic acid) in garlic root exudates [33], and this effect is greater than the alkalinity brought by calcium addition. Similar results were observed for the soil contents of Avail P and organic matter (Tables 1 and 2). Soil contents of NO3–, Avail K and total N were significantly higher in HO and HR than in LO and LR, respectively (Table 2). Furthermore, higher levels of NO3– and Avail K but lower NH4+ were observed in HO compared to HR, which could be caused garlic roots having a strong ability to absorb NO3–, instead of NH4+ [34]. No significant difference in soil properties was detected between LO and LR (Table 2). However, the six years of high-NPK fertilizer application significantly increased the garlic yield from 48.2% to 64.0% compared with the yield obtained with the low-NPK fertilizer samples during 2011–2016 (S1 Fig).
Bacterial and archaeal 16S rRNA gene copy number
We observed the 16S rRNA copy number in four treatments ranged from 1.5 × 108 to 4.3 × 109 copies in 1 g of wet soil (Fig 1). The bacterial and archaeal abundances in the rhizosphere were significantly higher than those in the bulk soil in both the high- (HR: HO = 1.57) and low- (LR:LO = 7.90) NPK samples. Furthermore, higher 16S rRNA copy numbers were observed in the high-NPK treatment than in the low-NPK treatment in both the rhizosphere (HR:LR = 5.65) and bulk (HO:LO = 28.43) soils (Fig 1). The bacterial and archaeal abundances were positively correlated with the concentrations of Avail K (r = 0.895, P < 0.01), total N (F = 0.955, P < 0.01) and organic matter (r = 0.664, P < 0.05) (S1 Table).
Bacterial and archaeal α-diversity analysis
After the selection and chimaera analysis of the OTUs, 1,296,155 high-quality sequences were assigned to 6,484 non-singleton OTUs, resulting in the classification of 612 taxa at the genus level. Based on a similarity cutoff of 97%, Good’s coverages ranged from 96.5% to 99.0% (Table 3), which indicated that the number of reads were sufficient to represent the bacterial and archaeal diversity in all the samples.
No obvious differences (P > 0.05) in ACE were found among the four soil samples. However, the Shannon and Chao indices in low NPK treatment were significantly different, compared to high NPK fertilizer soils. The Shannon index of both the rhizosphere and bulk samples was significantly lower under the low-NPK treatment compared to the high-NPK treatment, whereas the Chao index showed the opposite trend (P < 0.05, Table 3). The evenness (Shannoneven) was marginally positively affected by both soil compartment and fertilizer dose. Compared with the Shannoneven in the bulk soil (HO and LO), the values in the respective rhizosphere samples (HR and LR) were slightly higher (Table 3). Similarly, the Shannoneven values obtained for the high-NPK fertilizer samples (HO and HR) were slightly higher than those found for their respective low-NPK fertilizer samples (LO and LR) (Table 3).
The Shannon index was positively and significantly correlated with the soil NH4+ content, and ACE was negatively and significantly (F = 0.447, P < 0.05, S2 Table) correlated with the total N. However, no obvious correlation was found between the Chao index and the measured soil properties (S2 Table).
Bacterial and archaeal composition
Proteobacteria, Bacteroidetes, Thaumarchaeota, Planctomycetes and Acidobacteria were the top five dominant phyla and accounted for 76.4–78.2% of the bacterial sequences obtained from the soil samples (Fig 2A). The top dominant classes were the Soil_Crenarchaeotic_Group (13.2–16.3%), Acidobacteria (10.8–13.0%), Sphingobacteriia (6.9–7.8%) and Cytophagia (6.9–7.8%) (Fig 2B). The top five most abundant genera were Blastocatella (1.9–2.4%), Terrimonas (1.3–19%), Bryobacter (1.5–1.8%), Pirellula (1.7–2.2%) and Chryseolinea (1.6–1.7%) (Fig 2C).
A. Phyla (relative abundance >1%), B. Classes (relative abundance >0.1%), C. Genera (relative abundance >0.01%).
Soil compartment significantly distinguished 12 classes of bacteria and archaea, as shown in Fig 3. The classes Flavobacteriia, Alphaproteobacteria, Betaproteobacteria, Gammaproteobacteria, Verrucomicrobiae and Erysipelotrichia were significantly enriched (P < 0.05) in the rhizosphere relative to bulk soil (Fig 3A–3F and Table 4), whereas the classes Acidimicrobiia, Chthonomonadetes, Ardenticatenia, Caldilineae, Deltaproteobacteria and Nitrospira were more abundant (P<0.05) in the bulk soil than in the rhizosphere (Fig 3G–3L, Table 4).
The error bars indicate the standard deviations of the relative abundances between five replicate samples.
We also observed obvious difference in four classes of bacteria and archaea between low and high NPK soils, as shown in Fig 4. The proportion of the class Anaerolineae was higher under the high-NPK treatment than under the low-NPK treatment (Fig 4A, Table 4). In contrast, the percentage of the classes Ktedonobacteria, Fibrobacteria and Thermomicrobia were significantly (P < 0.05) higher in the low-NPK soils than in the high-NPK soils (Fig 4B–4D, respectively and Table 4).
The error bars indicate the standard deviations of the relative abundances between five replicate samples.
Furthermore, LEfSe analyses with an LDA threshold of 3.5 identified 43 bacterial taxa exhibiting significant differences among all the soil samples (Fig 5). Different number of taxa were enriched in the different soils (21 taxa in HO, 4 in HR, 6 in LO, 12 in HR) (Fig 5B). These groups primarily belonged to five phyla: Proteobacteria, Planctomycetes, Nitrospirae, Bacteroidetes and Actinobacteria (Fig 5A). In particular, bacterial taxa including the class Alphaproteobacteria, order Rhizobiales, family Xanthomonadaceae and genus Flavobacterium (S2A–S2D Fig, respectively) were identified as biomarkers in the HR soil samples.
A. LEfSe taxonomic cladograms. Significantly discriminant taxon nodes are coloured. Branches are shaded according to the highest ranked group for the corresponding taxon. The yellow nodes represent the taxa with no significant differences among the sampled habitats. B. Histogram of LDA scores showing the discriminant taxa among four soil samples.
NMDS and RDA
A graphic representation of the NMDS results allowed comparison of the samples based on weighted phylogenetic criteria and sequence alignment (Fig 6A). The LR samples were strongly separated from the three other soil samples along axis 1, and the HR samples were separated from the other samples along axis 2. The ANOSIM results showed that the bacterial communities in the four groups were significantly different and that the differences between groups were significantly greater than the differences within each group (Global R = 0.878, P = 0.001) (Fig 6A). The pairwise PERMANOVA comparisons were significantly different (P < 0.05, Table 5). We observed significant difference (P < 0.05) between HO and LR (R = 0.615) and between HR and LO (R = 0.367), which showed that the strength of the soil compartment on bacterial and archaeal recruitment differed among fertilization regimes. Greater differences were observed between the bulk and rhizosphere communities under the low-NPK treatment (LO/LR, R = 0.607) compared with the high-NPK treatment (HO/HR, R = 0.413) (Table 5). We also found a greater difference between the high- and low-NPK communities in the rhizosphere samples (HR/LR, R = 0.594) than in the bulk samples (HO/LO, R = 0.340) (Table 5).
The RDA results shown in Fig 6B revealed that axis 1 explained 12.99% of the total variance separating the rhizosphere (LR and HR) from the bulk (LO and HO) samples. Along axis 2, which explained 10.8% of the total variance, HR was strongly separated from the other three samples. The three most important factors affecting the changes in bacterial and archaeal communities were the soil concentrations of NH4 + (F = 2.1, P = 0.002), Avail K (F = 1.6, P = 0.012) and NO3- (F = 1.5, P = 0.022), explaining 11.7%, 8.5% and 7.9% of the observed variance, respectively (S3 Table). All environmental variables together accounted for 50.8% of the microbial community changes between samples.
Discussion
Positive fertilizer and rhizosphere effects on bacterial and archaeal abundance
The relatively higher bacterial and archaeal abundance in high NPK soils (Fig 1) may be due to the higher content of nutrients required for bacterial and archaeal growth [35] in the high-NPK fertilizer treatments. Among these nutrients, the correlation analysis also showed that the concentrations of Avail K, total N and organic matter were directly proportional to the bacterial and archaeal abundances (S1 Table). These results may be related to the dominant Proteobacteria, which are heterotrophic and were the major microbiota component obtained from root exudates [36]. Alphaproteobacteria, Betaproteobacteria and Gammaproteobacteria in Proteobacteria are often characterized as rapidly growing r-strategists that respond positively to low-molecular-weight substrates [37], which were more abundant in the garlic rhizosphere than in the bulk soils in this study. In addition, Thaumarchaeota was the dominant (13.2–15.2%, S2A Fig) archaeal taxon in 20 samples, and studies of Thaumarchaeota have shown that these species play an important role in the biogeochemical cycling of nitrogen and carbon [38].
Arnault’s [39] research showed that the degradation of garlic tissue (leaves, stems and roots) that occurs during harvesting or straw return releases sulphurous volatiles such as thiosulphinates and zwiebelanes. These volatiles are converted into disulphides that show biocidal activities against fungi, nematodes and arthropods. This effect may lead to an increase in the number of bacteria competing with the above organisms for nutrients, which may explain the positive rhiosphere effects on the bacterial and archaeal abundance (Fig 1) found in this study. Moreover, the garlic rhizosphere exhibits a high content of soil organic carbon [40]. For example, dibutyl phthalate, which provides a carbon source for Arthrobacter sp. [41], may lead to the recruitment of specific soil microorganisms and affect root-knot nematode egg hatching [42], and thus increase the bacterial and archaeal abundance.
Comparison of the bacterial and archaeal richness, evenness and diversity
The one way ANOVA result of Chao index indicated that high-NPK fertilization decreased the richness of bacteria. Although no soil properties were correlated with the Chao index, the total N was negatively and significantly correlated with ACE, which may be due to higher total N reducing the viability of certain microorganisms (Table 3). For example, some microorganisms in the classes Ktedonobacteria [43], Thermomicrobia [44] and Fibrobacteria [45] (Fig 4B, 4C and 4D, respectively) may show weakened growth and struggle to survive or even die when they do not adapt to high-nutrition conditions, resulting in a reduction in their abundance and a decrease in their richness in HO and HR.
In the high-NPK samples compared with the low-NPK samples, the richness decreased but the Shannon index increased (Table 3), possibly due to an increased evenness, which has been identified as a key factor in preserving the functional stability of an ecosystem [46].
Bacterial and archaeal community composition
The examination of how the fertilizer doses and garlic roots shape the bacterial and archaeal community composition revealed differences between the high/low-NPK and bulk/rhizosphere samples at different classification levels.
The species of class Alphaproteobacteria are organotrophic, can thrive in nutrient-rich soils and utilize reduced forms of inorganic N (ammonia, nitrite or nitrate) as energy sources [47]. In this context, the high garlic yield in the high-NPK fertilizer treatments from 2011–2016 (S1 Fig) may benefit from the enrichment of Alphaproteobacteria in HR (Fig 5B and S2A Fig) due to their contribution to accelerating the N cycle [48]. Rhizobia naturally infect legumes as host plants, whereas some Rhizobium strains in the order Rhizobiales can form symbiotic relationships with nonlegume species such as Parasponia [49], radishes [50], Arabidopsis [51] and rice [52]. The enrichment of the order Rhizobiales in HR (Fig 5B and S2B Fig) and the higher proportion of the genus Rhizobium in HR and LR (S2C Fig) may indicate that garlic has the potential to form a symbiotic relationship with Rhizobia and thereby promote garlic growth.
We observed positive rhizosphere effects on the class Gammaproteobacteria (Fig 3D), and the family Xanthomonadaceae (S2C Fig) belonging to this class was dominant in HR. Xanthomonadaceae is a widespread family of bacteria in soil, including the plant-pathogenic genera Xanthomonas, Xylella and Stenotrophomonas [53], isolates that are recognized as important plant pathogens [54]. Fortunately, Xanthomonas and Xylella did not appear in all soil samples, but we found a higher Stenotrophomonas abundance in HR than in LR, LO and LR (Fig 2C). This finding indicates the potential disease danger (i.e. leaf scorch disease) [55] in garlic with the long-term application of large amounts of inorganic NPK.
The class Deltaproteobacteria was more abundant in bulk soil compared to the garlic rhizosphere, as also found in soybeans [56]. These changes in the percentage of specific taxa might result from changes in the composition of garlic root exudates during growth and development because garlic roots exhibit a reciprocal relationship with adjacent bacterial and archaeal groups in the rhizosphere [57]. Doolotkelvieva et al. [58] found that the class Flavobacteriia dominated the soybean rhizosphere, as also found in the garlic rhizosphere (Fig 3A). Among Flavobacteriia, several species of the genus Flavobacterium are associated with plant protection and growth promotion [59] and with bioremediation in soils [60] and marine sediments [61]. The LEfSe analysis revealed that Flavobacterium was the dominant group in HR (Fig 5B), which indicates the potential of these species in promoting the growth and increasing the garlic yield from 2011–2016 (S1 Fig). This result may guide the isolation and development of microbial agents such as plant growth-promoting rhizobacteria.
The classes Ardenticatenia and Caldilineae were more abundant in bulk soil (Fig 3C and 3D and Table 4). A previous study also indicated that microorganisms in these two classes could use oxygen as an electron acceptor and that the aerobic absorption of the matrix is a common feature [62]. The percentage of Thermomicrobia, many of which are thermophilic and have a broad chemoorganotrophic substrate specificity [63], decreased in the high-NPK group compared with the low-NPK group (Fig 4D and Table 4), indicating the negative influence of a high concentration on these groups. This finding was consistent with our study of the effects of 34 years of nitrogen fertilization on intensive black soil bacterial and archaeal communities in northeast China [6]. In contrast, Anaerolineae was dominant in the high-NPK samples compared with the low-NPK samples (Fig 4A). This result is consistent with the results of Wang et al. [64], who indicated that high NPK significantly stimulates the growth and production of Anaerolineae and that its abundance presents a positive relationship with the Avail P content. However, no significant difference in Avail P was found between the high- and low-NPK treatments (Table 2) in this study.
Factors shaping the bacterial and archaeal community structure
The results from the NMDS analysis (Fig 6A) were consistent with the RDA results (Fig 6B), which indicated that bacterial communities in bulk soil (HO and LO) was more similar and significantly separated from those in the rhizosphere (HR and LR). The effect of soil compartment on the bacterial community structure were stronger than those of fertilization regime, indicating that microbial species respond more strongly to root exudates, as reported by Wang et al. [7]. The soil compartment effect on bacteria under the continuous cropping of garlic may be related to the root exudates. The accumulation of carbohydrates and amino acids in root exudates provides the required energy for certain species and promotes the growth and reproduction of rhizosphere bacteria [13]. In contrast, phenolic acids secreted by garlic roots inhibit the growth of beneficial soil microorganisms and exert their own toxic effects, disrupting the balance of the original microbial community and its structure in the rhizosphere of garlic [13].
However, the bacterial communities differed between HR and LR, which showed that fertilizer dose also significantly affected the bacterial communities, and many researchers agree on this point [6, 7, 65, 66]. The concentration of NH4+ in soil (contribution of 23%, P = 0.002) was the most important factor driving the changes in the bacterial community. This is consistent with the findings of Wang et al. [67], who concluded that NH4+ is the key factor affecting the population size of bacterial ammonia oxidizers. Our previous research [6] showed that the composition of the bacterial community was closely correlated with the NO3− concentration in cultivated black soil, and we obtained the same result in this study. Similar to the results of the current study, Li et al. [68] found that the Avail K content was significantly associated with microbial changes in the rhizosphere under maize–peanut intercropping.
Conclusion
This study systematically analysed the interactive effects of two important factors in a continuous cropping garlic system (fertilizer dose and soil compartment) on soil microbial communities. We found that the fertilizer dose impacted the bacterial and archaeal communities in rhizosphere differently than those in bulk soil, and these findings could be used to guide research priorities and management decisions. The changes in the soil concentrations of NH4+, Avail K and NO3− appeared to change the community composition. The long-term use of high NPK fertilizer reduced the diversity and richness of bacteria. Significant differences in bacterial and archaeal community structure were found between the rhizosphere and bulk soils under different fertilization doses. Our results provide evidence that the influences of fertilizer dose and the soil compartment interact to shape the bacterial and archaeal communities in a continuous garlic cropping system, and we highlight the need to reframe the fertilizer and rhizosphere interaction as a dynamic process. However, whether the rules identified in this study are applicable to other continuous cropping systems and which changes in microbial functions correspond to changes in these microbial indicators affecting the yield and garlic health are issues to be further explored in the future.
Supporting information
S1 Fig. Garlic yield in the high (H) and low (L) NPK fertilizer treatments from 2011 to 2016.
https://doi.org/10.1371/journal.pone.0250571.s001
(TIF)
S2 Fig. Differently enriched taxa in soil groups based on abundance within phylogenetic lineages (LEfSe analysis).
https://doi.org/10.1371/journal.pone.0250571.s002
(TIF)
S1 Table. Pearson’s correlation coefficients between soil chemical properties and number of 16S rRNA gene copies.
https://doi.org/10.1371/journal.pone.0250571.s003
(DOCX)
S2 Table. Pearson’s corrections between soil properties and ɑ-diversity.
https://doi.org/10.1371/journal.pone.0250571.s004
(DOCX)
S3 Table. Forward selection results of RDA analysis.
https://doi.org/10.1371/journal.pone.0250571.s005
(DOCX)
Acknowledgments
We thank the land owers Fang Xinjun and Chang Xinan gave permission to conduct the study on this site, all participating members of this research and Jonathan J. Fong in Lingnan University in Hongkong, China for language editing.
References
- 1. Wang Q, Xin J, Guan D, Dan W, Zhao B, Ma M, et al. Long-term fertilization changes bacterial diversity and bacterial communities in the maize rhizosphere of Chinese Mollisols. Applied Soil Ecology. 2017; 125: 88–96. https://doi.org10.1016/j.apsoil.2017.12.007
- 2.
Wang M. Effects of Rehmannia glutinosa Libosch. continuous cropping on microbial communities. Henan Agriculture University. 2008.
- 3. Li WH, Liu QZ, Chen P. Effect of long-term continuous cropping of strawberry on soil bacterial community structure and diversity. Journal of Integrative Agriculture. 2018; 17: 2570–2582.
- 4. Zhang Y, Li X, Gregorich EG, Mclaughlin NB, Zhang X, Guo Y, et al. No-tillage with continuous maize cropping enhances soil aggregation and organic carbon storage in Northeast China. Geoderma, 2018; 330:204–211. pmid:32476338
- 5. Miransari M. Soil microbes and the availability of soil nutrients. Acta Physiologiae Plantarum. 2013; 35(11), 3075–3084.
- 6. Zhou J, Guan D, Zhou B, Zhao B, Ma M, Qin J, et al. Influence of 34-years of fertilization on bacterial communities in an intensively cultivated black soil in northeast China. Soil Biology and Biochemistry. 2015; 90, 42–51.
- 7. Wang Q, Ma M, Jiang X, Zhou BK, Guan DW, Cao FM, et al. Long-term N fertilization altered 13C-labeled fungal community composition but not diversity in wheat rhizosphere of Chinese black soil. Soil Biology and Biochemistry. 2019; 135: 117–126.
- 8. Chang Y., Fan F, Song A, Cui P, Li T, Liang Y. Denitrification potential under different fertilization regimes is closely coupled with changes in the denitrifying community in a black soil. Applied Microbiology and Biotechnology. 2015; 99: 5719–5729. pmid:25715781
- 9. Wang H, Li X, Liu S, Jin S. Quantitative determination of allicin in Allium sativum L. bulbs by UPLC. Chromatographia. 2009; 71: 159–161.
- 10. Khan MA, Cheng ZH. Influence of garlic root exudates on cyto-morphological alteration of the hyphae of Phytophthora capsici, the cause of phytophthora blight in Pepper. Pakistan Journal of Botany. 2010; 42: 4353–4361.
- 11. Schmidt JE, Kent AD, Brisson V, Gaudin A. Agricultural management and plant selection interactively affect rhizosphere microbial community structure and nitrogen cycling. Microbiome. 2019; 1: 1–18. pmid:31699148
- 12. Liu S, Liu S, Zhang Z, Wei H, Qi J, Duan J, Influence of garlic continuous cropping on rhizosphere soil microorganisms and enzyme activities. Scientia Agricultura Sinica. 2010; 43: 1000–1006 (in Chinese).
- 13. Liu S, Liu S, Zhang Z, Wei H, Huang Z, Zhang Y. Inhibition effect of garlic root exudates on the genus allium. Scientia Agricultura Sinica. 2011; 44:2625–2632 (in Chinese).
- 14. Cai G, Chen D, White R, Fan X, Pacholski A, Zhu Z,. Gaseous nitrogen losses from urea applied to maize on a calcareous fluvo-aquic soil in the North China Plain. Australian Journal of Soil Research. 2002; 40: 737–748.
- 15. Wang N, Qiu J, Zhu Q. The research of garlic’s nutrient requirements and fertilization model. Jilin Vegatables. 2014; 3: 2–4. (in Chinese)
- 16. Wang Z, Wang C, Yang A, Wang C, Zhang J, Wang X, Effects of calcium fertilizer on garlic diseases control. China Cucurbits and Vegetables. 2016; 29 (7): 33–35.
- 17. Mcpherson M, Peng W, Marsh EL, Mitchell RB. Isolation and analysis of microbial communities in soil, rhizosphere, and roots in perennial grass experiments. Journal of Visualized Experiments. 2018; (137) e57932
- 18. Sims JT, Soil pH Effects on the distribution and plant availability of manganese, Copper, and Zinc. Soil Science Society of America Journal. 1986; 50: 367–373.
- 19. Parker JG, A comparison of methods used for the measurement of organic matter in marine sediment. Chemistry and Ecology. 1983; 1(3), 201–209.
- 20. Sparks DL, Page AL, Helmke PA, Loeppert RH, Soltanpour PN, Tabatabai MA. Methods of soil analysis. 2009; Part 3—chemical methods.
- 21. Zhao HB, Wang ZH, Xue C, Huang DL, Malhi SS. Increased dryland winter wheat yields by nitrogen fertilizer topdressing and its relationship to soil moisture, available N, P and K in Northwestern China. Communications in Soil Science & Plant Analysis. 2016; 1141917. https://doi.org10.1080/00103624.2016.1141917.
- 22. Kaur A, Manoj S, Jatinder M, Krishi B. Effect of method of sowing on bulb size and yield in garlic. International Journal of Current Microbiology and Applied Sciences. 2017; 6(11): 5354–5357.
- 23. Jones RT, Robeson MS, Lauber CL, Hamady M, Knight R, Fierer N. A comprehensive survey of soil acidobacterial diversity using pyrosequencing and clone library analyses. The ISME Journal. 2009; 3(4):442–453. pmid:19129864
- 24. Caporaso JG, Lauber CL, Walters WA, Berg-Lyons D, Huntley J, Fierer N, et al. Ultra-high-throughput microbial community analysis on the Illumina HiSeq and MiSeq platforms. The ISME Journal. 2012; 6, 1621–1624. pmid:22402401
- 25. Caporaso JG, Kuczynski J, Stombaugh J, Bittinger K, Bushman FD, Costello EK, et al. QIIME allows analysis of high-throughput community sequencing data. Nature Methods. 2010; 7: 335–336. pmid:20383131
- 26. Edgar RC, Haas BJ, Clemente JC, Quince C, Knight R. UCHIME improves sensitivity and speed of chimera detection. Bioinformatics. 2011; 27: 2194–2200. pmid:21700674
- 27. Edgar RC. UPARSE: highly accurate OTU sequences from microbial amplicon reads. Nature Methods. 2013; 10: 996. pmid:23955772
- 28. Pruesse E, Peplies J, Glöckner FO. SINA: accurate high-throughput multiple sequence alignment of ribosomal RNA genes. Bioinformatics. 2012; 28: 1823–1829. pmid:22556368
- 29. Pruesse E, Quast C, Knittel K, Fuchs BM, Ludwig W, Peplies J, et al. SILVA: a comprehensive online resource for quality checked and aligned ribosomal RNA sequence data compatible with ARB. Nucleic Acids Research. 2007; 35: 7188–7196. pmid:17947321
- 30. Segata N, Izard J, Waldron L, Gevers D, Miropolsky L, Garrett WS, et al. Metagenomic biomarker discovery and explanation. Genome Biology. 2011; 12(6): R60. https://doi.org/10.1186/gb-2011-12-6-r60 pmid:21702898
- 31. Zhou H, Zhang D, Jiang Z, Sun P, Xiao H, Yuxin W, et al. Changes in the soil microbial communities of alpine steppe at Qinghai-Tibetan Plateau under different degradation levels. Science of the Total Environment. 2019; 651, 2281–2291. pmid:30326458
- 32. Saad RF, Kobaissi A, Echevarria G, Kidd P, Calusinska M, Goux X, et al. Influence of new agromining cropping systems on soil bacterial diversity and the physico-chemical characteristics of an ultramafic soil. Science of the Total Environment. 2018; 645: 380–392. pmid:30029117
- 33. Zhang L, Qiu DY, Wei P, Liu BY, Effect of different polarity composition of garlic root exudates on angelica sinensis germination and seedlings. Modern Chinese Medicine. 2015; 17(8): 815–820 (in Chinese).
- 34. Zhao XC, Liu WB, Wang QH, Zhang M, Dong HW, Li XH. Influence of companion planting on rhizosphere soil nutrients of greenhouse cucumber. Shandong Agricultural Sciences. 2016; 48 (5): 75–79.
- 35. Yuan HZ, Wu H, Ge TD, Li K, Wu JS, & Wang JR. Effects of long-term fertilization on bacterial and archaeal diversity and community structure within subtropical red paddy soils. The Journal of Applied Ecology. 2015; 26 (6), 1807. pmid:26572036
- 36. Haichar FZ, Marol C, Berge O, Rangel-Castro JI, Prosser JI, Balesdent J, et al. Plant host habitat and root exudates shape soil bacterial community structure. The ISME Journal. 2008; 2: 1221–1230. pmid:18754043
- 37. Goldfarb KC, Karaoz U, Hanson CA, Santee CA, Bradford MA, Treseder KK, et al. Differential growth responses of soil bacterial taxa to carbon substrates of varying chemical recalcitrance. Frontiers Microbiology. 2011; 2: 94. pmid:21833332
- 38. Zhang LM, He JZ. A novel archaeal phylum: Thaumarchaeota—a review. Acta Microbiology Sinica. 2012; 52(4):411–421 (in Chinese). pmid:22799205
- 39. Arnault I, Mondy N, Diwo S, Auger J, Soil behaviour of sulfur natural fumigants used as methyl bromide substitutes. International Journal of Environmental Analytical Chemistry. 2004; 84(1–3): 75–82.
- 40. Xiao XM, Cheng ZH, Meng HW, Liu LH, Li HZ, Dong YX. Intercropping of green garlic (allium sativum l.) induces nutrient concentration changes in the soil and plants in continuously cropped cucumber (cucumis sativus l.) in a plastic tunnel. Plos One. 2013; 8(4), e62173 pmid:23637994
- 41. Chu JY, Liu TF, Zhang HY, Qiu LQ. Screening and characterization of a dibutyl phthalate-degrading bacterium. Microbiology China. 2017; 44(7): 1555–1562
- 42. Yang G, Zhou B, Zhang X, Zhang Z, Wu Y, & Zhang Y, et al. Effects of tomato root exudates on meloidogyne incognita. Plos One. 2016; 11(4), e0154675. pmid:27128659
- 43. Fan Z, Lu S, Liu S, Liu S, Guo H, Peng X. Changes in plant rhizosphere microbial communities under different vegetation restoration patterns in karst and non-karst ecosystems. Scientific Reports. 2019; 9, 8761. pmid:31217455
- 44. Jiménez-Bueno N, Valenzuela-Encinas C, Marsch R, Ortiz-Gutiérrez D, Verhulst N, Govaerts B, et al. Bacterial indicator taxa in soils under different long‐term agricultural management. Journal of Applied Microbiology. 2016; 120: 921–933. pmid:26808352
- 45. Neves V, Oliveira D, Oliveira DC, Perucci LO, Santos T, Fernandes I, et al. High‐sugar diet intake, physical activity, and gut microbiota crosstalk: implications for obesity in rats. Food Science and Nutrition. 2020; 8:5683–5695. pmid:33133570
- 46. Wittebolle L, Marzorati M, Clement L, Balloi A, Daffonchio D, Heylen K, et al, Initial community evenness favours functionality under selective stress. Nature. 2009; 458(7238): 623–626. pmid:19270679
- 47. Ahn JH, Song J, Kim BY, Kim MS, Joa JH, Weon HY. Characterization of the bacterial and archaeal communities in rice field soils subjected to long-term fertilization practices. Journal of Microbiology. 2012; 50, 754–765 pmid:23124742
- 48.
OpenStax., 2021. Proteobacteria. Available from: https://chem.libretexts.org/@go/page/42487.
- 49. Trinick MJ. Structures of nitrogen fixing root nodules formed on Parasponia andersonii (Planch). Canadian Journal of Microbiology. 1979; 25:565–578. pmid:476539
- 50. Antoun H, Beauchamp CJ, Goussard N, Chabot R, Lalandi R. Potential of Rhizobium and Bradyrhizobium as plant growth promoting rhizobacteria on non-legumes: Effect on radishes (Raphanus sativus L.). Plant and Soil. 1998; 204:57–68.
- 51. Berge O, Lodhi A, Brandelet G, Santaella C, Roncato MA, Christen R, et al. Rhizobium alamii sp. nov., an exopolysaccharide-producing species isolated from legume and non-legume rhizospheres. International Journal of Systematic and Evolutionary Microbiology. 2009; 59(2):367–72. pmid:19196780
- 52. Reddy PM, Ladha JK, So RB, Hernandez RJ, Ramos MC, Angeles OR, et al. Rhizobial communication with rice roots: induction of phenotypic changes, mode of invasion and extend of colonization. Plant and Soil 1997; 194(1–2):81–98.
- 53. Nadia MH, Marie-Agnès J, Ralf K. Adhesion mechanisms of plant-pathogenic Xanthomonadaceae. Oxygen Transport to Tissue XXXIII. 2011; 715:71–89. pmid:21557058
- 54. Naushad HS, Gupta RS. Phylogenomics and molecular signatures for species from the plant pathogen-containing order Xanthomonadales. PLoS one. 2013; 8(2):e55216. pmid:23408961
- 55. Bian DR, Xue H, Piao CG, Li Y. Stenotrophomonas cyclobalanopsidis sp. nov., isolated from the leaf spot disease of Cyclobalanopsis patelliformis. Antonie van Leeuwenhoek. 2020; 113(10):1447–1454. pmid:32748076
- 56. Mendes LW, Kuramae EE, Navarrete AA, Veen JA, Tsai SM. Taxonomical and functional microbial community selection in soybean rhizosphere. The ISME Journal. 2014; 8: 1577–1587. pmid:24553468
- 57. Shi S, Nuccio E, Herman DJ, Rijkers R, Estera K, Li J, et al. Successional trajectories of rhizosphere bacterial communities over consecutive seasons. mBio. 2015; 6: e00746. pmid:26242625
- 58. Doolotkelvieva T, Bobusheva S, Konurbaeva M. Impact of biocontrol agent streptomyces fumanus on bacterial communities in the rhizosphere of wheat and soybean in newly cultivated soil. Global Advanced Research Journal of Agricultural Science. 2015; 4. 211–221.
- 59. Sang MK, Kim KD,. The volatile-producing Flavobacterium johnsoniae strain GSE09 shows biocontrol activity against Phytophthora capsici in pepper. Journal of Applied Microbiology. 2012; 113: 383–398. pmid:22563881
- 60. Jit S, Dadhwal M, Prakash O, Lal R. Flavobacterium lindanitolerans sp nov., isolated from hexachlorocyclohexane contaminated soil. International Journal of Systematic and Evolutionary Microbiology. 2008; 58: 1665–1669. pmid:18599713
- 61. Fu Y, Tang X, Lai Q, Zhang C, Zhong H, Li W, et al. (). Flavobacterium beibuense sp. nov., isolated from marine sediment. International Journal of Systematic and Evolutionary Microbiology. 2011; 61, 205–209. pmid:20190020
- 62. McIlroy SJ, Karst SM, Nierychlo M, Dueholm MS, Albertsen M, Kirkegaard RH, et al. Genomic and in situ investigations of the novel uncultured Chloroflexi associated with 0092 morphotype filamentous bulking in activated sludge. The ISME Journal. 2016; 10: 2223–2234. pmid:26905629
- 63. Gaylarde C, Baptista-Neto JA, Ogawa A, Kowalski M, Celikkol-Aydin S, Beech I. Epilithic and endolithic microorganisms and deterioration on stone church facades subject to urban pollution in a sub-tropical climate. Biofouling. 2017; 33(2):113–127. pmid:28054493
- 64. Wang Y, Zhao X, Guo Z, Jia Z, Wang S, Ding K. Response of soil microbes to a reduction in phosphorus fertilizer in rice-wheat rotation paddy soils with varying soil P levels. Soil and Tillage Research. 2018; 181: 127–135.
- 65. Ding JL, Jiang X, Ma MC, Zhou BK, Guan DW. Zhao BS, et al. Effect of 35 years inorganic fertilizer and manure amendment on structure of bacterial and archaeal communities in black soil of northeast China. Applied Soil Ecology. 2016; 105: 187–195.
- 66. Ma MC, Zhou J, Ongena M, Liu W, Wei D, Zhao BS, et al. Effect of long-term fertilization strategies on bacterial community composition in a 35-year field experiment of Chinese Mollisols. AMB Express. 2018; 8: 20. pmid:29442257
- 67. Wang Y, Zhu G, Song L, Wang S, Yin C. Manure fertilization alters the population of ammonia-oxidizing bacteria rather than ammonia-oxidizing archaea in a paddy soil. Journal of Basic Microbiology. 2014; 54: 190–197. pmid:23686819
- 68. Li QS, Wu LK, Chen J, Khan MA, Luo XM, Lin WX. Biochemical and microbial properties of rhizospheres under maize/peanut intercropping. Journal of Integrative Agriculture. 2016; 15: 101–110.