Grass species selection and regular mowing are essential for maintaining aesthetic and environmentally sound turfgrass systems. However, their impacts on the soil microbial community, the driving force for soil N cycle and thus the environmental fate of N, are largely unknown. Here, the high throughput sequencing of 16S rRNA gene and internal transcribed spacer (ITS) region was used to evaluate how long-term defoliation management and grass growth habits (propagation types and photosynthetic pathways) modulated the soil microbial community. The investigation included three cool-season C3 grasses (creeping bentgrass, Kentucky bluegrass, and tall fescue) and three warm-season C4 grasses (bermudagrass, St. Augustinegrass, and zoysiagrass). Creeping bentgrass and bermudagrass were managed as putting greens with a lower mowing height; tall fescue spread in a tussock manner via tiller production whereas other grasses propagated in a creeping manner via rhizomes and/or stolons. Ordination analysis showed that both bacterial and fungal communities were primarily separated between putting green and non-putting green systems; and so were N-cycle gene relative abundances, with the putting greens being greater in N mineralization but lower in nitrification. Compared to warm-season grasses, cool-season grasses slightly and yet significantly enhanced the relative abundances of Chloroflexi, Verrucomicrobia, and Glomeromycota. Tall fescue yielded significantly greater bacterial and fungal richness than non-tussock grasses. As the main explanatory soil property, pH only contributed to < 18% of community compositional variations among turfgrass systems. Our results indicate that defoliation management was the main factor in shaping the soil microbial community and grass growth habits was secondary in modulating microbial taxon distribution.
Citation: Xia Q, Chen H, Yang T, Miller G, Shi W (2019) Defoliation management and grass growth habits modulated the soil microbial community of turfgrass systems. PLoS ONE 14(6): e0218967. https://doi.org/10.1371/journal.pone.0218967
Editor: Cheng Gao, University of California Berkeley, UNITED STATES
Received: March 27, 2019; Accepted: June 12, 2019; Published: June 24, 2019
Copyright: © 2019 Xia 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: NCBI Sequence Archive with the BioProject accession number PRJNA484409.
Funding: This work was supported by Center for Environmental Turfgrass Research and Education, North Carolina, to WS; US-EPA ORD to HC; and China Scholarship Council and the National Natural Science Foundation of China #11305047 to TY. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Turfgrass, covering over 16 million hectares in the US, is one of the most important irrigated crops in the country and provides significant regulating (e.g., C sequestration, soil erosion control, and cooling), supporting (e.g., nutrient cycles), and cultural (e.g. spiritual and recreational benefits) services [1–3]. Characterized with intensive management, including fertilization, irrigation, and the use of pesticide, however, the mono-cultured turf (i.e., grass and the subtending soil) has long been criticized for the high rate of fertilization and thus feedforward effects on the environment due to nutrient loss. Over years, research emphasis has been on seeking management practices to improve fertilizer use efficiency and mitigate nutrient loss (e.g., N leaching and gas emissions), and also continuously introducing new cultivars to reduce management cost while increasing sustainability [4,5]. There are a number of ways to choose turfgrass species, but all depend on grass species characteristics, e.g., appearance, cultivation requirement, pest resistance, and stress tolerance. While knowledge has been considerably advanced on turfgrass physiology and ecology, information is still lacking on how grass species modulates the diversity, composition, and function of the soil microbial community, the key component for understanding N cycle and thus the fate of N in the environment.
A plant species may exert multiple selection pressures on soil microbes. Rhizodeposits, the root-derived C compounds that originate from sloughed-off of root cells and tissues, mucilages, volatiles, and soluble lysates and exudates , are a strong determinant of the soil microbial community. As a readily available and rich C source, rhizodeposits may facilitate the proliferation of copiotrophs over oligotrophs to enhance competition exclusion and therefore reduce microbial species richness and evenness [7–9]. Rhizodeposits may also target specific microbial taxa and thus shape the soil microbial community . Nonetheless, such effects are often spatially restricted to the rhizosphere. Plants also affect microbes in the root zone and bulk soil through controls on soil physical and chemical properties, including pH, nutrients, and pore size and distribution. For instance, roots stimulate soil aggregation and stability through physical entanglement and also through mucilage production to bind soil particles. Roots can alter soil pH via cation-anion exchange balance, organic anion release, root exudation and respiration, and redox-coupled processes . Further, root architecture (e.g., elongation rate, lateral root production, and root length density) may interfere with soil hydraulic conductivity to regulate water flow and nutrient movement . Such plant-driven selection on microbes has been increasingly recognized, and selective effects have been found to differ even at a plant species or genotype scale [12–15].
Turfgrasses vary largely in growth habits, despite that all possess narrow leaves and fibrous roots. Based on ways of new growth generation, turfgrasses can be classified as either tussock grass or non-tussock grass. A tussock grass (or bunch-type grass) produces new grasses from tillers in a cluster or bunch. In contrast, a non-tussock grass (or creeping-type grass) produces new grasses from above- and/or belowground lateral stems (i.e., stolons and rhizomes, respectively). As such, stolonferous and rhizomatous grasses have a greater capacity to spread laterally relative to tussock grasses and can quickly make a dense and uniform land cover. Turfgrasses can also be grouped into C3 cool-season and C4 warm-season grasses, with C4 being more efficient for photosynthesis at elevated temperature and more tolerant to drought, and thus being preferred in warm and arid regions. While some geographic areas are suitable for both warm- and cool-season grasses , most cool-season turfgrasses will suffer the loss of root system during mid-summer when soil temperature rises above 17°C . Because grow habits might affect belowground resource (e.g., C, N, and water) availability and distribution, we hypothesized that soil microbial communities would differ between warm- and cool-season grasses and between bunch-type and creeping-type grasses. The main objective of this work was to examine how soil microbial species richness, diversity, and N-cycle functional genes vary with turfgrass species identities and growth habits.
The investigation took advantage of an existing turfgrass site where grasses of different growth habits had been established on soils of similar texture and subjected to different defoliation practices (e.g., mowing intensity and frequency). Defoliation may affect root architecture, morphology, and biomass allocation [18,19] and therefore the assemblage of soil microbes. By including defoliation management, we could better evaluate the impacts of grass growth habits. We also examined relationships of microbial community structural metrics with soil properties, considering the influences of edaphic factors, e.g., pH, moisture, and soil texture on the soil microbial community [20–24].
Materials and methods
Field plots and soil sampling
Soil samples were taken from individual plots of six turfgrass species at the Lake Wheeler Turfgrass Field Laboratory, North Carolina State University, Raleigh NC, USA in August 2016. The six species included bermudagrass (Cynodon dactylon x C. transvaalensis cv. ‘Champion’), creeping bentgrass (Agrostis stolonifera cv. ‘Penncross’), Kentucky bluegrass (Poa pratensis), tall fescue (Festuca arundinacea), St. Augustinegrass (Stenotaphrum secundatum cv. ‘Raleigh’), and zoysiagrass (Zoysia japonica cv. ‘El Toro’) and had been established over 5 years (see Table 1 for grass growth habits and defoliation management). It is worth mentioning that Greens (Table 1) were mown more frequently and at a lower height than Non-greens. The plot sizes of these turfgrass species varied, being smallest, ~ 0.1 ha for St. Augustinegrass (ST), Kentucky bluegrass (KB) and zoysiagrass (ZG), and largest, ~ 0.4 ha for tall fescue (TF). All turfgrasses had been managed for mowing, irrigation, and fertilization according to professional standards. In brief, for the non-putting greens, fertilizers were applied two or three times a year with a cumulative rate of 100–195 kg N ha-1 yr-1 and mostly in summer for warm-season grasses and in spring and fall for cool season grasses. Common herbicide, either glyphosate (C3H8NO5P) or oxadiazon (C15H18Cl2N2O3) was applied yearly. The two putting greens, bermudagrass (BM) and creeping bentgrass (CB), had fertilizers applied more frequently throughout the year at a rate of 191 and 333 kg N ha-1 yr-1, respectively. Fungicides like triadimefon, triticonazole, and fluoxastrobin were also applied to BM and CB with a rate of 50 and 150 kg ha-1 yr-1, respectively. Irrigation was applied as needed to prevent stress. Soil in the plots is classified as fine sandy loam (fine, kaolinitic, thermic Typic Kanhapludults).
Defoliation was more intensive and frequent in green than in non-green turfgrass systems.
To make representative samples, each turf plot was further divided into three roughly equal-size subplots. Then, six to eight soil cores (2.5 cm dia. x 10 cm length) were taken randomly from spots of >1 m away from edges of each subplot and mixed, resulting in three composite samples for each grass species. For TF, three composite samples also represented three cultivars, ‘Fesnova’, ‘Raptor III’, and ‘Regenerook’, respectively. There were a total of 18 composite soil samples (i.e., 6 turfgrass species x 3 subplots as replicates). Soil was sieved (< 2 mm), one aliquot stored at -20°C prior to DNA extraction, and the other aliquot stored at 4°C until the analyses of soil and microbial properties.
Soil chemical and biological properties
Soil pH was determined with a soil (g)/water (ml) ratio of 1:2.5. Dry combustion method was used to analyze soil total C and N with a Perkin-Elmer 2400 CHN analyzer (Perkin-Elmer Corporation, Norwalk, CT, USA). Soil inorganic N (NH4+-N and NO3—N) was extracted using 0.5M K2SO4 at a ratio of 1:5 soil (g)/solution (ml), filtered through Whatman #42 filter paper, and determined with FIA QuikChem 8000 autoanalyzer (Lachat Instruments, Loveland, CO, USA). Extracted total C and N in solution were also analyzed with TOC analyzer (TOC-5000, Shimadzu Scientific Instruments, Japan). A chloroform fumigation-extraction method was used to determine soil microbial biomass C (MBC) and N (MBN) with extraction coefficients of 0.38 and 0.54 to biomass C and N, respectively [25,26].
DNA extraction, amplification and sequencing
Soil DNA was extracted using FastDNA Spin Kit for Soil (MP Bio, Solon, OH, USA) from 0.6 g of each soil sample. The extracted DNA was then column-purified using OneStep PCR Inhibitor Removal Kit (Zymo Research, Orang, CA, USA) and a concentration of more than 50 ng μL-1 was acquired for each sample. DNA purity of about 1.70–1.90 was assessed by the ratio of absorbance at 260 and 280 nm using a NanoDrop Spectrophotometer (Thermo Scientific, Wilmington, DE, USA).
Bacterial 16S rRNA gene and fungal ITS was PCR-amplified with primer pairs targeting V3-V4 (F319: 5’-ACTCCTACGGGAGGCAGCAG-3’ and R806: 5’- GGACTACHVGGGTWTCTAAT-3’) and ITS1-ITS2 (F_KYO2: 5’-TAGAGGAAGTAAAAGTCGTAA-3’ and R_KYO2: 5’- TTYRCTRCGTTCTTCATC-3’), respectively, and with Illumina MiSeq overhang adapters [27,28]. The PCR was a 50 μL reaction consisting of 25 μL 2x KAPA HiFi HotStart ReadyMix (KAPA Biosystems, Wilmington, MA, USA), 2.5 μL template DNA (4–20 ng μL-1), 2.5 μL 10 mM of each primer, and 17.5 μL nuclease-free water. A negative control with no DNA template was also included as a control of extraneous DNA contamination. The thermocycling condition for PCR of 16S rRNA genes was: initial denaturation at 95 oC for 3 min; 25 cycles of 98 oC for 30 sec, 55 oC for 15 sec, and 72 oC for 30 sec; final elongation at 72 oC for 5 min. For fungi, all the thermocycling condition was the same except for that annealing temperature was 51 oC. PCR products were cleaned up with AMPure XP beads (Beckman Coulter Genomics, Danvers, MA, USA) and then eluted in 10 mM Tris buffer (pH 8.5). Unique index (barcode) sequences were added to purified DNA fragments at both ends using the Nextera XT Index Kit (Illumina, San Diego, CA, USA) and a second clean-up was performed. The purified 16S rRNA gene and ITS fragments were mixed equimolarly and paired-end sequenced on Illumina Miseq platform (300×2 paired end, v3 chemistry) (Illumina, San Diego, CA, USA). Due to no detectable DNA amplification, negative controls were not sequenced. The Miseq sequences were deposited in GenBank with the BioProject accession number PRJNA484409.
Demultiplexed sequencing data were trimmed based on the expectation of amplicon size (430-470bp for 16S rRNA gene and 180-360bp for ITS), filtered by the maximum error rate, 0.5% using USEARCH v9.1.13 , and then chimeras of ~ 40% sequence reads on average identified and removed using a usearch61 method in QIIME 1.9.1 . Operational taxonomic units (OTUs) were classified with a threshold of 97% similarity and then assigned to taxa using the open reference algorithm for both 16S rRNA gene and ITS sequencing data by using methods of usearch61 against Greengenes database (13_8) and RDP (Ribosomal Database Project) against UNITE database (12_11) [31–33], respectively. Singletons were removed during OTUs pickup; and sequences in a range of ~ 30,000 to 300,000 across samples were rarefied to a depth of 27,000 and 25,000 for bacteria and fungi, respectively. Alpha diversity metrics, including observed OTUs based on rarefaction curves, chao1, and Shannon index were analyzed in QIIME. Matrices of weighted unifrac and Bray-Curtis distance were used for beta diversity analysis of bacterial and fungal communities, respectively, by principal coordinate analysis (PCoA).
Relative abundance of putative functional genes involved in N cycle was predicted using PICRUSt (Phylogenetic Investigation of Communities by Reconstruction of Unobserved States)  by matching bacterial OTUs predicted from 16S rRNA gene with reference genomes. Bacterial OTUs with genes involved in N cycle processes were collected using metagenome_contributions.py script based on Kyoto Encyclopedia of Genes and Genomes (KEGG) orthology (KO) database and were then taxonomically classified. Generally, an NSTI (i.e., weighted nearest sequenced taxon index) of around 0.17 or less was considered reliable for soil samples . The NSTI scores of our samples were 0.18 ± 0.01 (standard deviation).
One-way ANOVA, followed by post-hoc Tukey’s test (SAS 9.4, SAS Institute Inc. Cary, NC, USA) was performed for multiple comparisons of soil biochemical properties, microbial alpha diversity, and N-cycle gene relative abundances among grass species. Correlations between soil biochemical properties and microbial communities were analyzed using DistLM (distance-based linear models) in PRIMER (Plymouth Routines in Multivariate Ecological Research Statistical Software, v7.0.13, PRIMER-E Ltd, UK); forward selection was applied to add one soil property to the model at each step and the property increasing adjusted R2 most at each step was chosen. All the statistical significance was based on P ≤ 0.05 if not specified. Significance of sample grouping in PCoA was analyzed using Adonis method in QIIME. Significance of taxon relative abundance was determined using linear discriminant analysis effect size (LEfSe) (http://huttenhower.sph.harvard.edu/galaxy/) .
Bacterial and fungal communities
Both bacterial and fungal species richness (i.e. observed OTUs and chao1) were grass species-dependent, being greatest in tall fescue, following by zoysiagrass, St. Augustine, Kentucky bluegrass, and bermudagrass, and lowest in creeping bentgrass (Table 2). Shannon diversity index followed the similar pattern except for the index for the fungal community of creeping bentgrass. However, neither microbial richness nor Shannon diversity index differed between warm- and cool- season grass species. PCoA analysis showed differences in the soil microbial community structure among grass species, with bermudagrass and creeping bentgrass of more intensive defoliation management being similar (Fig 1). Minor and yet significant differences were also found between tall fescue and other grass species along the PCoA-3, and between warm- and cool-season grasses for the fungal community along the PCoA-2.
Soil bacterial (a, b) and fungal (c, d) communities in six turfgrass systems (BM, bermudagrass; CB, creeping bentgrass; KB, Kentucky bluegrass; TF, tall fescue; ST, St. Augustinegrass; ZG, zoysiagrass). Cool- and warm-season turfgrass systems are represented by empty and filled symbols, respectively.
Different letters within each row indicate significant differences at P < 0.05.
A total of 47 bacterial phyla were detected in turfgrass systems, while only 13 had a relative abundance > 1% (S1 Fig). The most abundant phylum was Proteobacteria, accounting for ~32%, followed by Acidobacteria (~15%), Actinobacteria (~14%) and Chloroflexi (~8%). Of 269 identified bacterial taxa from phyla to genera with > 0.1% relative abundance, ~77% differed significantly among turfgrass systems (Fig 2). Compared to other grass species, creeping bentgrass was most abundant in Acidobacteria and Alphaproteobacteria, and least in Actinobacteria, Bacteroidetes and Planctomycetes. In contrast, St. Augustinegrass was relatively more abundant in Actinobacteria; zoysiagrass in Bacteroidetes; and Kentucky bluegrass in Planctomycetes. There were also compositional differences among grass species in the sublevel taxa from class to genus. For example, [Chloracidobacteria] was relatively abundant in in tall fescue; Rhodospirillales in bermudagrass; and Acidobacteriales, an order of Acidobacteria, accounting for 10% in creeping bentgrass. Similarity between creeping bentgrass and bermudagrass in Fig 1 was in line with similarity in the relative abundances of taxonomic groups. They both showed greater abundance in Acidobacteria, Chlamydiae, Cyanobacteria, Hyphomicrobiaceae, and Methylocystaceae, but lower abundance in Bacteroidetes, Planctomycetes, and Betaproteobacteria, compared to other grasses. Minor and yet significant differences in microbial community compositions were also associated with cool- vs. warm-season grasses, with cool-season grasses being relatively more abundant in Chlorofexi and Verrucomicrobia (S2 Fig).
The heatmap contained bacterial taxa that differed significantly among six turfgrass systems (BM, bermudagrass; CB, creeping bentgrass; KB, Kentucky bluegrass; TF, tall fescue; ST, St. Augustinegrass; ZG, zoysiagrass). Only taxa with ≥ 2.5% relative abundance and assigned at least to the phylum level are included. The color scale indicates the relative abundance (%).
Unlike bacteria, a great portion of ITS sequences (~ 35%) could not be assigned into a phylum and lower taxonomic ranks, except into the fungal domain (S1 Fig), perhaps because of incompleteness of the UNITE database and/or non-target amplicons of other eukaryotes. This portion was considered to have little impact on fungal community comparisons since its proportion was similar among samples. Nonetheless, turfgrasses were dominant with Ascomycota (~ 44% on average across six turfgrass systems), followed by Basidiomycota (~ 8%). Of 111 identified taxa with > 0.1% abundance, ~ 78% showed significant differences among turfgrass systems. Basidiomycota and Zygomycota were most abundant in Kentucky bluegrass, accounting for 15.2% and 2.6%, respectively (Fig 3). Sordariomycetes, the class of Ascomycota was relatively more abundant in tall fescue, whereas Dothideomycetes, Eurotiomycetes, and Leotiomycetes were relatively more abundant in warm-season grasses (bermudagrass, St. Augustinegrass, and zoysiagrass). Some taxa presented solely in one turfgrass system but negligible in all the others, e.g., Incertae_sedis accounting for 3.4% in creeping bentgrass while < 0.1% in the other five turfgrass systems. Moderate differences in taxa were also found between cool- and warm- season grasses, with Chytridiomycota and Glomeromycota relatively more abundant in cool-season grasses (S2 Fig).
The heatmap contained fungal taxa that differed significantly among six turfgrass systems (BM, bermudagrass; CB, creeping bentgrass; KB, Kentucky bluegrass; TF, tall fescue; ST, St. Augustinegrass; ZG, zoysiagrass). Only taxa with ≥ 2.5% relative abundance and assigned at least to the phylum level are included. The color scale indicates the relative abundance (%).
Gene abundances involved in N cycling
Relative abundances of genes involved in N cycle varied significantly with specific pathways, being the greatest for mineralization, followed by assimilatory NO3- reduction, dissimilatory nitrate reduction to ammonium (DNRA), N fixation, denitrification, and the lowest for nitrification (Fig 4). However, there were significant variations in all N cycle processes among turfgrasses. Creeping bentgrass showed the lowest relative abundance of nitrification gene but the greatest relative abundances of genes in assimilatory NO3- reduction and N fixation. In addition, this turfgrass system was characterized by the lowest relative gene abundances in dissimilatory NO3- reduction to NO2- and N2O reduction to N2. As a result, the ratio of relative gene abundances between mineralization and nitrification in creeping bentgrass was nearly two-fold greater than that of other turfgrasses except for bermudagrass. However, creeping bentgrass had the lowest ratio of nitrification to assimilatory NO3- reduction and the lowest ratio of dissimilatory N2O reduction to dissimilatory NO reduction.
Genes for N transformations were predicted from the bacterial marker gene 16S rRNA using PICRUSt with a sequence depth of 19,890. Arrow thickness is positively related to gene abundances of individual N pathways, and bar height reflects relative gene abundances, which were nomalized to the highest values of indivisual processes among six turfgrass systems (BM, bermudagrass; CB, creeping bentgrass; KB, Kentucky bluegrass; TF, tall fescue; ST, St. Augustinegrass; ZG, zoysiagrass). Gene abundances were calculated as: K00260 + K00261 + K00262 for mineralization; ((K10944 + K10945 + K10946)/3 + K10535)/2 for nitrification; (((K00370 + K00371 + K00374 + K00373)/4 + (K02567 + K02568)/2) + (K00362 + K00363)/2 + K03385)/2 for dissimilatory NO3- reduction to NH4+ (DNRA); (K00367 + K00372 + K00360 + K00366)/2 for assimilatory NO3- reduction to NH4+; (K02588+K02586+K02591)/3+K00531 for N fixation; (K00370+K00371+K00374+K00373)/4 +(K02567+K02568)/2 for dissimilatory NO3- reduction to NO2-; K00368 for dissimilatory NO2- reduction to NO; (K04561+K02305)/2 for dissimilatory NO reduction to N2O; K00376 for dissimilatory N2O reduction to N2.
OTUs possessing these genes also diverged among turfgrasses, showing similar patterns to those of the total bacterial community. Dominant phyla involved in nosZ for dissimilatory N2O reduction to N2 were Proteobacteria and Chloroflexi, whereas dominant phyla in hao for hydroxylamine hydrolysis were Proteobacteria, Plantomycetes and Nitrospriae (S3 Fig). Compared to other grasses, creeping bentgrass had the lower relative abundance of hao in Plantomycetes but more in Proteobacteria. This turfgrass system also had lower nosZ abundance in Bacteriodetes but more in Chloroflexi.
Correlations of microbial community attributes with soil properties
Soil properties differed significantly among turfgrasses, but coefficients of variation were moderate, being lowest for pH, ~7.4% and highest for soil organic N ~ 44% (Table 3). All the properties did not show significant differences between cool- and warm-season grasses.
Distance based linear model analysis showed that only soil pH was significantly correlated with both bacterial and fungal community compositions, explaining ~ 18% and 12% of the total variations, respectively (Table 4). Several other soil properties were also significantly correlated with fungal community compositions, and yet all explained no more than 10% of the total variation. Soil pH, extractable organic C, and inorganic N together explained 43% of the total variation in the bacterial community; and soil pH, NH4+, and extractable organic C together explained ~34% of total variation in the fungal community (Table 5).
Augustinegrass; ZG, zoysiagrass). Different letters within each row indicate significant differences at P < 0.05.
Our initial hypothesis was that grass growth habits, both propagation types and photosynthetic pathways, could play important roles in modulating the diversity, composition, and functional gene abundances of the soil microbial community. However, our data showed that propagation types and photosynthetic pathways affected different metrics of the soil microbial community. Propagation types appeared to regulate microbial species richness, whereas photosynthetic pathways controlled the community composition despite moderate influences compared to defoliation management.
Propagation type affected the alpha diversity of the microbial community
Both microbial species richness and Shannon diversity index were greatest in tall fescue, but no single examined soil property could explain such variations. Instead, grass growth habit seemed to be the cause. Of the six grasses, tall fescue was the only tussock-type grass. Due to tiller production and no lateral stems, this grass grows as singular plants in tufts, leading to uneven soil coverage of shoots and roots. In contrast, non-tussock grasses generated new growth by aboveground stolons and/or belowground rhizomes and were likely to form a uniform lawn.
Resource translocation and information sharing among ramets of clonal plants have been well studied . Ramets can translocate water, carbohydrates, and minerals from individuals with high supply to those with low supply. Plants may also share information by translocating signal molecules [37,38] or secreting massive perfumes  when they are exposed to herbivore damage and defoliation. It is possible that chemical and information sharing among ramets promoted physiological synchronization among individuals and thus enhanced similarity in rhizodeposit biochemistry. Relatively ‘long-distant’ ramets of non-tussock grasses might distribute C, N, water, and other resources over a wide spatial scale and helped to increase resource homogeneity. In contrast, ramets of tussock grasses were clustered, resulting in resource distribution in a more confined area. As such, tussock grasses were more likely to introduce a fine-scale heterogeneity in soil properties.
Fine-scale heterogeneity in soil physicochemical properties (e.g. organic C, nutrients, water, pH, and aerobic conditions) has been considered as a key driver to promote biodiversity [40,41]. Often, fine-scale heterogeneity in soil is realized through aggregation [42,43], because not only does it help to create more divergent niches for species adaptation, but also it helps to separate competitive species and thereby limit competitive exclusion [44–46]. Accordingly, factors that influence aggregation are expected to impact soil biodiversity. Although soil organic matter could positively contribute to soil aggregation [47,48] and varied up to three fold among grass species in this work, it did not contribute to the divergence in microbial species richness. This concurred with another work where microbial species richness was found to be stable over a chronosequence of bermudagrass systems, despite ~ three-fold differences in soil organic C . Together, these suggest that root type rather than soil organic C content was the main driver for the divergence of microbial species richness among turfgrass systems.
Our data have two implications. First, growth habit-associated fibrous roots might contribute largely to divergence in soil aggregation. Second, non-tussock grasses might promote resource translocation and information sharing, and thereby improving the fine-scale homogeneity of soil properties. Hence, soil cultivated with stoloniferous or rhizomatous grass species was prone to harbor the microbial community of lower richness, compared with soil cultivated with the tussock-type tall fescue.
Growth habit and defoliation intensity shaped the soil microbial community
As hypothesized, growth habits did help to structure the soil microbial community. However, their effects were moderate, given that tussock-type tall fescue differed from other grasses mainly along the PCoA-3, and the warm-season grasses differed from the cool-season grasses only for fungi along the PCoA-2. Compared to the warm-season grasses, cool-season grasses enhanced the proliferation of Chloroflexi, Verrucomicrobia, Chytridiomycota, and Glomeromycota, the phyla that have been documented to be sensitive to soil nutrient status and/or play roles in plant nutrient uptake [50–52]. In summer, root dieback occurred for cool-season grasses when their photosynthetic activities were reduced by high temperature and yet they were still actively respiring. Because grass demanded more carbohydrates than photosynthesis could provide, root mass would decrease dramatically. As a consequence, grass ability of water and nutrient uptake declined. Our results suggest that to meet the challenge of root mass loss in summer, cool-season grasses might recruit beneficial microbes to help water and nutrient uptake.
Soil microbial communities differed mainly by defoliation intensity since bermudagrass and creeping bentgrass, the two grasses that were mowed at a lower height but greater frequency, clustered and were well separated from other grass species. Several studies examined the impacts of defoliation on the soil microbial community; however, results are mixed [51,53–56]. For example, mowing in a steppe ecosystem had little effect on the soil bacterial community , but mowing in a grassland of the Great Plains of North America was found to change the abundance of some bacteria and fungi, e.g., Actinobacteria, Bacteroidetes, Chloroflexi, Planctomycetes, and Ascomycota . Such inconsistency is perhaps because defoliation impacts varied not only with the intensity and frequency of defoliation, but also with the time when evaluations were made (e.g., evaluation right after one-time defoliation versus after years of defoliation). Nonetheless, research tends to support that long-term and cumulative effects of defoliation can be substantial [56,57]. Our results were aligned with those of Bartlett et al. (2008) that soil microbial communities differed between turfgrasses subjected to different intensities of long-term defoliation and management.
Decline in plant photosynthesis and therefore C supply to soil is one of the possible consequences of mowing. Compared to other grasses, more intensive and frequent mowing in bermudagrass and creeping bentgrass might result in a greater reduction of C flow from plant to soil. As such, long-term and intensive defoliation could lead to a resource-poor environment that favored the growth and proliferation of oligotrophic microbes. Indeed, the intensively mowed bermudagrass and creeping bentgrass harbored more Acidobaceria and Alphaproteobacteria but less Betaproteobacteria and Bacteroidetes than other grasses, the former two being documented as oligotrophs and the latter two as copiotrophs [58,59]. Guo et al. (2018) also showed that the relative abundance of Bacteroidetes was significantly reduced after long-term clipping. They also found that genes involved in the decomposition of complex compounds, e.g., starch, hemicellulose, pectin, cellulose, chitin and lignin, were much greater in mowed grassland than in the grassland without defoliation. Together, these results imply that defoliation affected the soil microbial community through controls on the quantity and biochemistry of carbon allocated from plant to soil.
It was reasonable to assume that defoliation management could affect the soil organic C content/biochemistry, but there was no correlation between the two. Neither total soil organic C nor extractable organic C could explain defoliation intensity-based groupings in the soil microbial community. Similar to other studies [21,60], soil pH was the most robust factor in explaining variations of the soil microbial community. However, the explanatory power of pH was small, only accounting for < 18% of total variations in bacteria or fungi. Perhaps, all the soil properties examined in this work were not the most and direct consequences of defoliation management.
Intensively mowed grasses generally demand more additional management practices, such as higher rates of fertilization and pesticide use. A bulk of publications [49,61–63] suggests that plant protection management only played minor roles in shaping the community structure. In this study, N fertilization rates were also unlikely the cause for divergences in the soil microbial community between intensively mowed bermudagrass and creeping bentgrass and other grasses. Nitrogen fertilization in bermudagrass was ~ 40% lower than that in creeping bentgrass but similar to the other grasses. However, microbial communities in bermudagrass and creeping bentgrass were similar and differed from the others. This further helps to infer that defoliation management was a dominant management practice in structuring the soil microbial community.
Grass species-specific characteristics in N cycle
Nitrogen cycle is the foundation of soil fertility and also crucial to understand the environmental fate of N. A survey of genes encoding enzymes for various N transformations may help to diagnose the potential of soil N supply and efflux at a given ecosystem. Similar to the community structure, N-cycle gene relative abundances drew a clear line between intensively mowed grass systems (creeping bentgrass and bermudagrass) and others. Creeping bentgrass and bermudagrass were characterized by greater relative gene abundances in mineralization, N fixation, and assimilatory nitrate reduction, but lower relative gene abundances in nitrification. Such a N-cycle pattern suggests that the two turfgrass systems were more N limited, and therefore they must mine N through mineralization and N fixation and, on the other hand, reduced the activity that could potentially lead to N loss. Defoliation is known to influence belowground microbial processes, but its magnitude depends on plant species as well as defoliation intensity [53,64]. Generally, defoliation is thought to enhance N mineralization by a pulse input of short-lived labile C [65,66]. Here, we proposed that ephemeral and yet frequent inputs of rhizodeposits in intensively mowed turfgrasses could stimulate microbial activity and turnover. The greater the microbial biomass turnover rate, the more N would be released from microbes for grass uptake. Overtime, a large fraction of N would be locked into grass biomass and removed from soil by defoliation. As a consequent, this would generate a limitation of N to soil microbes despite that N fertilization rates were much greater in creeping bentgrass than the other systems. To combat N limitation, the soil microbial community needed to not only enhance its capacity of N fixation and uptake but also minimize nitrification. Denitrification is another important microbial process that will lead to N loss via gas emissions. The relative abundance of gene involved in the first step (i.e., NO3- reduction to NO2-) of sequential reactions of denitrification was also lower in intensively mowed grasses than the other grasses. This further suggests that defoliation might reduce the processes that lead to soil N loss.
Similar variations in the soil microbial community structure and N cycle gene relative abundances suggest a tight linkage between the community structure and function in turfgrass systems. Nitrogen-cycle gene relative abundances also appeared to correlate with the alpha diversity of the soil microbial community; the lower taxon richness in intensively mowed grasses coincided with the less relative abundance of nitrification genes. Although linkage between microbial community structure and function is often found to be weak , loss in microbial diversity has been documented to affect N transformations in soil . Schimel  stated whether or not microbial community structure is an important control on ecological processes was the issue of scale; and linkage between microbial community structure and function became loose as the scale was moved up. Nonetheless, our data suggest that mono-cultured turfgrass systems seem to be at the scale that microbial community structure could be used to predict soil functions.
Grass growth habits (propagation types and photosynthetic pathways) significantly affected soil microbial communities. The tussock-type tall fescue was more beneficial to bacterial and fungal taxon richness than non-tussock grasses, likely due to promotion of soil heterogeneity. Cool-season grasses enhanced the relative abundance of Chloroflexi, Verrucomicrobia, and Glomeromycota, compared to the warm-season grasses, perhaps because as a compensation strategy, root dieback in summer triggered cool-season grasses to recruit microbes for helping nutrient acquisition. However, defoliation intensity was found to be most robust in modulating the soil microbial community and N-cycling gene abundances, with more intensively and frequently mowed turfgrass systems having the lower relative abundances of nitrification genes. This work is significant because it helps to better understand the consequences of the choice of grass species and defoliation management on soil N processes and thus the environmental fate of N.
S1 Fig. Stacked bar charts of microbial community composition.
S2 Fig. LEfSe of cool- and warm-season turfgrass systems.
S3 Fig. Relative abundances of soil bacterial phyla having nosZ and hao genes.
We thank the Environmental and Agricultural Testing Service Laboratory, NCSU for soil organic carbon and nitrogen analysis.
- 1. Beard JB, Green RL (1994) The Role of Turfgrasses in Environmental Protection and Their Benefits to Humans. Journal of Environment Quality 23 (3): 452–460.
- 2. Qian Y, Follett RF (2002) Assessing Soil Carbon Sequestration in Turfgrass Systems Using Long-Term Soil Testing Data. Agronomy Journal 94 (4): 930–935.
- 3. Simmons M, Bertelsen M, Windhager S, Zafian H (2011) The performance of native and non-native turfgrass monocultures and native turfgrass polycultures: An ecological approach to sustainable lawns. Ecological Engineering 37 (8): 1095–1103.
- 4. Moll RH, Kamprath EJ, Jackson WA (1982) Analysis and Interpretation of Factors Which Contribute to Efficiency of Nitrogen Utilization. Agronomy Journal 74 (3): 562–564.
- 5. Bowman DC, Cherney CT, Rufty TW (2002) Fate and Transport of Nitrogen Applied to Six Warm-Season Turfgrasses. Crop Science 42 (3): 833–841.
- 6. Dennis PG, Miller AJ, Hirsch PR (2010) Are root exudates more important than other sources of rhizodeposits in structuring rhizosphere bacterial communities. FEMS microbiology ecology 72 (3): 313–327. pmid:20370828
- 7. Marilley L, Vogt G, Blanc M, Aragno M (1998) Bacterial diversity in the bulk soil and rhizosphere fractions of Lolium perenne and Trifolium repens as revealed by PCR restriction analysis of 16S rDNA. Plant and Soil 198 (2): 219–224.
- 8. Costa R, Götz M, Mrotzek N, Lottmann J, Berg G et al. (2006) Effects of site and plant species on rhizosphere community structure as revealed by molecular analysis of microbial guilds. FEMS microbiology ecology 56 (2): 236–249. pmid:16629753
- 9. Shi S, Nuccio E, Herman DJ, Rijkers R, Estera K et al. (2015) Successional Trajectories of Rhizosphere Bacterial Communities over Consecutive Seasons. mBio 6 (4): e00746. pmid:26242625
- 10. Huang X-F, Chaparro JM, Reardon KF, Zhang R, Shen Q et al. (2014) Rhizosphere interactions: root exudates, microbes, and microbial communities. Botany 92 (4): 267–275.
- 11. Hinsinger P, Plassard C, Tang C, Jaillard B (2003) Origins of root-mediated pH changes in the rhizosphere and their responses to environmental constraints: A review. Plant and Soil 248 (1–2): 43–59.
- 12. Gould IJ, Quinton JN, Weigelt A, Deyn GB de, Bardgett RD (2016) Plant diversity and root traits benefit physical properties key to soil function in grasslands. Ecology Letters 19 (9): 1140–1149. pmid:27459206
- 13. Miethling R, Wieland G, Backhaus H, Tebbe CC (2000) Variation of Microbial Rhizosphere Communities in Response to Crop Species, Soil Origin, and Inoculation with Sinorhizobium meliloti L33. Microb Ecol 40 (1): 43–56. pmid:10977876
- 14. Singh BK, Millard P, Whiteley AS, Murrell JC (2004) Unravelling rhizosphere-microbial interactions: opportunities and limitations. Trends in microbiology 12 (8): 386–393. pmid:15276615
- 15. Garbeva P, van Elsas JD, van Veen JA (2008) Rhizosphere microbial community and its response to plant species and soil history. Plant Soil 302 (1–2): 19–32.
- 16. Wherley BG, Sinclair TR (2009) Differential sensitivity of C3 and C4 turfgrass species to increasing atmospheric vapor pressure deficit. Environmental and Experimental Botany 67 (2): 372–376.
- 17. Beard JB (2001) Turfgrass root basics. TURFAX 9 (3): 4–7.
- 18. Arredondo JT, Johnson DA (1998) Clipping Effects on Root Architecture and Morphology of 3 Range Grasses. Journal of Range Management 51 (2): 207–213.
- 19. Arredondo JT, Johnson DA (1999) Root architecture and biomass allocation of three range grasses in response to nonuniform supply of nutrients and shoot defoliation. The New Phytologist 143 (2): 373–385.
- 20. Griffiths BS, Bonkowski M, Roy J, Ritz K (2001) Functional stability, substrate utilisation and biological indicators of soils following environmental impacts. Applied Soil Ecology 16 (1): 49–61.
- 21. Lauber CL, Strickland MS, Bradford MA, Fierer N (2008) The influence of soil properties on the structure of bacterial and fungal communities across land-use types. Soil Biology and Biochemistry 40 (9): 2407–2415.
- 22. Cruz-Martínez K, Suttle KB, Brodie EL, Power ME, Andersen GL et al. (2009) Despite strong seasonal responses, soil microbial consortia are more resilient to long-term changes in rainfall than overlying grassland. The ISME journal 3 (6): 738–744. pmid:19279669
- 23. Rousk J, Bååth E, Brookes PC, Lauber CL, Lozupone C et al. (2010) Soil bacterial and fungal communities across a pH gradient in an arable soil. The ISME journal 4 (10): 1340–1351. pmid:20445636
- 24. Kaiser K, Wemheuer B, Korolkow V, Wemheuer F, Nacke H et al. (2016) Driving forces of soil bacterial community structure, diversity, and function in temperate grasslands and forests. Scientific Reports 6.
- 25. Brookes PC, Landman A, Pruden G, Jenkinson DS (1985) Chloroform fumigation and the release of soil nitrogen: A rapid direct extraction method to measure microbial biomass nitrogen in soil. Soil Biology and Biochemistry 17 (6): 837–842.
- 26. Vance ED, Brookes PC, Jenkinson DS (1987) An extraction method for measuring soil microbial biomass C. Soil Biology and Biochemistry 19 (6): 703–707.
- 27. Toju H, Tanabe AS, Yamamoto S, Sato H (2012) High-coverage ITS primers for the DNA-based identification of ascomycetes and basidiomycetes in environmental samples. PloS one 7 (7): e40863. pmid:22808280
- 28. Klindworth A, Pruesse E, Schweer T, Peplies J, Quast C et al. (2013) Evaluation of general 16S ribosomal RNA gene PCR primers for classical and next-generation sequencing-based diversity studies. Nucleic acids research 41 (1): e1. pmid:22933715
- 29. Edgar RC (2010) Search and clustering orders of magnitude faster than BLAST. Bioinformatics (Oxford, England) 26 (19): 2460–2461.
- 30. Caporaso JG, Kuczynski J, Stombaugh J, Bittinger K, Bushman FD et al. (2010) QIIME allows analysis of high-throughput community sequencing data. Nature Methods 7 (5): 335–336. pmid:20383131
- 31. Wang Q, Garrity GM, Tiedje JM, Cole JR (2007) Naive Bayesian classifier for rapid assignment of rRNA sequences into the new bacterial taxonomy. Applied and environmental microbiology 73 (16): 5261–5267. pmid:17586664
- 32. Kõljalg U, Nilsson RH, Abarenkov K, Tedersoo L, Taylor AFS et al. (2013) Towards a unified paradigm for sequence-based identification of fungi. Molecular ecology 22 (21): 5271–5277. pmid:24112409
- 33. Macdonald LM, Paterson E, Dawson LA, McDonald AJS (2004) Short-term effects of defoliation on the soil microbial community associated with two contrasting Lolium perenne cultivars. Soil Biology and Biochemistry 36 (3): 489–498.
- 34. Langille MGI, Zaneveld J, Caporaso JG, McDonald D, Knights D et al. (2013) Predictive functional profiling of microbial communities using 16S rRNA marker gene sequences. Nature Biotechnology 31 (9): 814–821. pmid:23975157
- 35. Segata N, Izard J, Waldron L, Gevers D, Miropolsky L et al. (2011) Metagenomic biomarker discovery and explanation. Genome biology 12 (6): R60. pmid:21702898
- 36. Stuefer JF, Gómez S, van Mölken T (2004) Clonal integration beyond resource sharing: implications for defence signalling and disease transmission in clonal plant networks. Evol Ecol 18 (5–6): 647–667.
- 37. Thaler JS, Karban R, Ullman DE, Boege K, Bostock RM (2002) Cross-talk between jasmonate and salicylate plant defense pathways: effects on several plant parasites. Oecologia 131 (2): 227–235. pmid:28547690
- 38. Stratmann J (2003) Long distance run in the wound response–jasmonic acid is pulling ahead. Trends in plant science 8 (6): 247–250. pmid:12818656
- 39. Pichersky E, Gershenzon J (2002) The formation and function of plant volatiles: perfumes for pollinator attraction and defense. Current Opinion in Plant Biology 5 (3): 237–243. pmid:11960742
- 40. Ettema C, Wardle D (2002) Spatial soil ecology. Trends in Ecology & Evolution 17 (4): 177–183.
- 41. Vos M, Wolf AB, Jennings SJ, Kowalchuk GA (2013) Micro-scale determinants of bacterial diversity in soil. FEMS microbiology reviews 37 (6): 936–954. pmid:23550883
- 42. Sexstone AJ, Revsbech NP, Parkin TB, Tiedje JM (1985) Direct Measurement of Oxygen Profiles and Denitrification Rates in Soil Aggregates. Soil Science Society of America Journal 49 (3): 645–651.
- 43. Elliott ET (1986) Aggregate Structure and Carbon, Nitrogen, and Phosphorus in Native and Cultivated Soils. Soil Science Society of America Journal 50 (3): 627–633.
Tilman D, Kareiva PM (1997) Spatial ecology. The role of space in population dynamics and interspecific interactions. Princeton, N.J., Chichester: Princeton University Press.
- 45. Amarasekare P (2003) Competitive coexistence in spatially structured environments: a synthesis. Ecology Letters 6 (12): 1109–1122.
- 46. Stein A, Gerstner K, Kreft H (2014) Environmental heterogeneity as a universal driver of species richness across taxa, biomes and spatial scales. Ecology Letters 17 (7): 866–880. pmid:24751205
- 47. Angers DA (1992) Changes in Soil Aggregation and Organic Carbon under Corn and Alfalfa. Soil Science Society of America Journal 56 (4): 1244–1249.
- 48. Cambardella CA, Elliott ET (1993) Carbon and Nitrogen Distribution in Aggregates from Cultivated and Native Grassland Soils. Soil Science Society of America Journal 57 (4): 1071–1076.
- 49. Chen H, Xia Q, Yang T, Shi W (2018) Eighteen-Year Farming Management Moderately Shapes the Soil Microbial Community Structure but Promotes Habitat-Specific Taxa. Frontiers in microbiology 9: 1776. pmid:30116234
- 50. Willis A, Rodrigues BF, Harris PJC (2013) The Ecology of Arbuscular Mycorrhizal Fungi. Critical Reviews in Plant Sciences 32 (1): 1–20.
- 51. Zhang X, Chen Q, Han X (2013) Soil bacterial communities respond to mowing and nutrient addition in a steppe ecosystem. PloS one 8 (12): e84210. pmid:24391915
- 52. Navarrete AA, Tsai SM, Mendes LW, Faust K, Hollander M de et al. (2015) Soil microbiome responses to the short-term effects of Amazonian deforestation. Molecular ecology 24 (10): 2433–2448. pmid:25809788
- 53. Guitian R, Bardgett RD (2000) Plant and soil microbial responses to defoliation in temperate semi-natural grassland. Plant and Soil 220 (1–2): 271.
- 54. McDonald D, Price MN, Goodrich J, Nawrocki EP, DeSantis TZ et al. (2012) An improved Greengenes taxonomy with explicit ranks for ecological and evolutionary analyses of bacteria and archaea. The ISME journal 6 (3): 610–618. pmid:22134646
- 55. Williams MC, Wardle GM (2007) Pine and eucalypt litterfall in a pine-invaded eucalypt woodland: The role of fire and canopy cover. Forest Ecology and Management 253 (1–3): 1–10.
- 56. Guo X, Zhou X, Hale L, Yuan M, Feng J et al. (2018) Taxonomic and Functional Responses of Soil Microbial Communities to Annual Removal of Aboveground Plant Biomass. Frontiers in microbiology 9: 954. pmid:29904372
- 57. Bartlett MD, James IT, Harris JA, Ritz K (2008) Size and phenotypic structure of microbial communities within soil profiles in relation to different playing areas on a UK golf course. European Journal of Soil Science 59 (5): 835–841.
- 58. Smit E, Leeflang P, Gommans S, van den Broek J, van Mil S et al. (2001) Diversity and seasonal fluctuations of the dominant members of the bacterial soil community in a wheat field as determined by cultivation and molecular methods. Applied and environmental microbiology 67 (5): 2284–2291. pmid:11319113
- 59. Fierer N, Bradford MA, Jackson RB (2007) Toward an ecological classification of soil bacteria. Ecology 88 (6): 1354–1364. pmid:17601128
- 60. Fierer N, Jackson RB (2006) The diversity and biogeography of soil bacterial communities. Proceedings of the National Academy of Sciences of the United States of America 103 (3): 626–631. pmid:16407148
- 61. Gevao B, Semple KT, Jones KC (2000) Bound pesticide residues in soils: a review. Environmental Pollution 108 (1): 3–14. pmid:15092962
- 62. Kalam A, Tah J, Mukherjee AK (2004) Pesticide effects on microbial population and soil enzyme activities during vermicomposting of agricultural waste. J Environ Biol 25 (2): 201–208. pmid:15529880
- 63. Hartman GL, Chang H-X, Leandro LF (2015) Research advances and management of soybean sudden death syndrome. Crop Protection 73: 60–66.
- 64. Fu S, Cheng W (2004) Defoliation affects rhizosphere respiration and rhizosphere priming effect on decomposition of soil organic matter under a sunflower species: Helianthus annuus. Plant and Soil 263 (1): 345–352.
- 65. Frank DA, Groffman PM (1998) Ungulate vs. Landscape control of soil C and N processes in grasslands of Yellowstone National Park. Ecology 79 (7): 2229–2241.
- 66. Hamilton EW, Frank DA (2001) Can plants stimulate soil microbes and their own nutrient supply? Evidence from a grazing tolerant grass. Ecology 82 (9): 2397–2402.
- 67. Nemergut DR, Schmidt SK, Fukami T, O’Neill SP, Bilinski TM et al. (2013) Patterns and processes of microbial community assembly. Microbiology and molecular biology reviews: MMBR 77 (3): 342–356. pmid:24006468
- 68. Philippot L, Spor A, Hénault C, Bru D, Bizouard F et al. (2013) Loss in microbial diversity affects nitrogen cycling in soil. The ISME journal 7 (8): 1609–1619. pmid:23466702
- 69. Schimel J (1995) Ecosystem Consequences of Microbial Diversity and Community Structure. Ecological Studies: Analysis and Synthesis 113: 239–254.