Introduction

Terrestrial plants interact with the biotic and abiotic components of their environment, resulting in variation in plant performance and community dynamics. Key among these interactions is the relationship between plants and their associated soil microbes, which may influence both the fitness of individual plants and their abundance within communities (van der Putten et al. 2013). Plants may condition their soil through the leaching of chemicals via litter or root exudation (Kong et al. 2008; Heinen et al. 2017; Veen et al. 2019), producing distinct soil microbial communities (Kardol et al. 2006; Pendergast et al. 2013; Orozco-Aceves et al. 2015) that may affect plant performance and coexistence (van der Putten et al. 2013; Smith-Ramesh and Reynolds 2017; Xue et al. 2018). However, associated soil microbes vary in their functional roles, generating variation in the benefits and drawbacks they provide to plant hosts (Magnoli and Lau 2020; Veen et al. 2021). Changes in soil microbial composition can impact the fitness of the individual plant itself, others of the same species (intraspecific), or individuals of other species (interspecific), leading to positive, neutral, or negative effects (Klironomos 2002; van der Putten et al. 2013; Mack and Bever 2014; Kos et al. 2015; Smith-Ramesh and Reynolds 2017). Through impacts on plant performance, interspecific variation in plant–soil microbe interactions appears to be a major driving force for plant community composition, succession, invasion, and ecosystem function (Ehrenfeld et al. 2005; Kardol et al. 2007; Miki 2012; Pendergast et al. 2013; Mack and Bever 2014; Verbeek and Kotanen 2019).

Intraspecific trait variation has long been recognized as a critical driving force in evolution that often manifests itself in population or community processes (Bolnick et al. 2011). As functional ecology has provided a more mechanistic view of plant community organization, intraspecific variation has also become a critical component (Violle et al. 2012; Hahn and Maron 2016). Intraspecific variation in plant chemistry can impact the quality of anti-predator defenses (Glassmire et al. 2016; Benedek et al. 2019), responses to interspecific competition (Jilani et al. 2008), as well as resistance to pathogens and parasites (Biere et al. 2004; Shikano et al. 2017), linking above- and below-ground processes (Van Der Putten et al. 2001; Bezemer and Van Dam 2005). Similarly, variation in plant chemical composition may result in changes to the composition of soil microbial communities (Lankau 2010; Benedek et al. 2019), with microbial biomass dependant on chemical and litter inputs and the responses of individual microbes to them (Kong et al. 2008). Litter chemical composition and quality act as an environmental filter that results in variable competitive outcomes between soil microorganisms for resources (Veen et al. 2021). Locally produced leaf litter may also provide soil legacy effects in the absence of the original host plants, affecting the soil microbiota, soil functioning, and growth of newly establishing plants (Wurst and Ohgushi, 2015; Veen et al. 2019). We should therefore expect that individual plant genotypes may influence the diversity and composition of soil microbial communities as a whole (Schweitzer et al. 2008; Bukowski and Petermann 2014; Gehring et al. 2017), as well as their reciprocal impacts on plants.

Although the preponderance of plant–microbe interaction work has focused on interspecific differences (Bukowski et al. 2018), critical intraspecific variation has been identified in a variety of systems. Genotype-specific microbial communities have been documented for both mutualistic (Gehring et al. 2017; Wang et al. 2019) and antagonistic (Liu et al. 2015; Eck et al. 2019) soil microbes, yielding population-level impacts (Bukowski et al. 2018). Overall, there may be significant variation associated with individual plant genotypes (Bukowski and Petermann 2014; Bukowski et al. 2018; Wang et al. 2019) and available microbes (Magnoli and Lau 2020), leading to a range in the strength and direction of the plant–microbe interactions generated.

Intraspecific variation in plant–microbe interactions likely varies spatially at the scale of the individual plant’s root system (Peacher and Meiners 2020). However, the effects of intraspecific plant chemical variation on soil microbes may be spatially much larger in clonal plant species, where vegetative reproduction of an individual plant can lead to a single genotype covering large areas for long periods of time (De Witte and Stöcklin 2010). Clonal reproduction is considered a competitive trait (Grime 2001), which is quite important in the dynamics of many successional plant communities (Prach et al. 1997; Wright and Fridley 2010). In systems where clonal species dominate and vary in chemistry or other key traits, the spatial scale of variation in soil microbial communities may represent an ecologically important source of interaction heterogeneity.

To assess intraspecific variation in plant–microbe interactions, we used a common garden of 24 genotypes of Solidago altissima, an aggressive dominant in many plant communities worldwide. We used soils collected from each individual genotype as inocula in a greenhouse study to determine the strength of plant–soil microbe interactions on two native target species, S. altissima and the bunchgrass Schizachyrium scoparium. Using these data, we addressed the following research questions: (1) Does appreciable intraspecific variation exists in the direction and magnitude of plant–soil microbe interactions across S. altissima genotypes? (2) Is this variation consistent across target species? (3) Is variation in the strength of plant–soil microbe interactions related to variation in S. altissima foliar chemistry? If appreciable variation occurs across S. altissima genotypes, this poses a methodological issue in how to deal with variation in characterizing plant–microbe interactions. Therefore, we also address: (4) Does pooling of soil microbial inocula provide a reliable estimate of the mean interaction across Solidago genotypes for both target species?

Materials and methods

Study species

Solidago altissima (syn. S. canadensis), is a chemically diverse, allelopathic perennial herb that is common in old fields and other open habitats across its native range of Eastern North America (Werner et al. 1980; Meiners et al. 2017; Kalske et al. 2019; Uesugi et al. 2019; Yip et al. 2019). Solidago altissima reproduces clonally through rhizomes and is very diverse in its chemical composition, allowing this species to become a successful invader across Europe, Japan, and Australia (Weber 1998; Uesugi et al. 2019; Yip et al. 2019). Solidago altissima has been used to document ecological impacts of chemical variation (Uesugi and Kessler 2013; Uesugi et al. 2019), impacts on soil microbial communities (Zhang et al. 2009), and variation in plant–microbe interactions (Dong et al. 2015; Howard et al. 2020; Peacher and Meiners 2020), representing a good model system to test for intraspecific variation in plant–soil microbe interactions.

Solidago altissima colonizes areas with disturbed soils through seed dispersal (Werner et al. 1980) and then begins to spread clonally. Initial populations of S. altissima have large numbers of genotypes. However, as stem density increases, genotypes interact and displace competitively inferior individuals, leading to fewer but larger clones in older populations (Hartnett and Bazzaz 1985). Therefore, S. altissima populations are expected to be more variable before clonal sorting has occurred, providing an opportunity to explore the full range of intraspecific variation in this species.

We added an additional target species, Schizachyrium scoparium, to assess the importance of intraspecific variation in interspecific plant–microbe interactions. This species is a C4 bunchgrass native to most of North America as a major component of grasslands and disturbed habitats (Pickett 1983; Huff et al. 1998; Duell et al. 2016). This second target species was added because it commonly co-occurs with S. altissima, is the dominant native grass at the garden site, responds to soil microbial communities (Bauer et al. 2017), and is inhibited by the soil microbial community of S. altissima (Peacher and Meiners 2020). The inclusion of S. scoparium provides insight into the role of intraspecific microbial variation on a common competitor of S. altissima.

Common garden

In May 2014, ramets of each of 24 genotypes of S. altissima were collected as rhizome/stem segments from the Douglas-Hart Nature Center (Mattoon, IL; 39° 29′ N; 88° 17′ W) in an area of recently restored prairie. Five rhizome segments per genotype were excavated from distinct patches within close proximity to one another (within 0.5 m) and isolated (> 5 m) from other such patches to ensure collection from different genotypes. The area had been in row-crop agriculture three years prior and S. altissima was not a part of the initial restoration seeding, so all S. altissima represented colonists from the surrounding landscape. This site age should also represent the phase before sorting of genotypes would be expected and should still retain high genetic (Hartnett and Bazzaz 1985) and potentially chemical diversity (see below).

The common garden site was a level section of land in Clark County, IL (39° 19′ N, 87° 55′ W) that had been used to grow corn the previous year using annual tillage practices. For this reason, we expect initial soil microbes and abiotic properties to be relatively homogeneous across the 12.5 × 20 m garden (Robertson et al. 1993; Röver and Kaiser 1999; Constancias et al. 2015). The 5 ramets from each S. altissima genotype were planted in a regular pattern (center and toward each corner) within a single, 1.6 × 1.6 m enclosure (one genotype per enclosure) with aluminum flashing buried 15 cm deep to prevent rhizome spread across plots. Plots were separated by 2 m and the spaces between enclosures were maintained by mowing. After planting S. altissima, other plants were allowed to colonize naturally to assess variation in competitive ability across the genotypes in a related study (B. S. Foster, Unpublished thesis). During the first two years, all S. altissima flowering heads were removed prior to seed set to prevent colonization of the plots by new genetic individuals.

Chemical analysis

High-performance liquid chromatography (HPLC) analysis was used to characterize leaf chemistry and determine whether S. altissima genotypes within the common garden varied in the amount or type of chemicals produced (Meiners et al. 2017). In the summer of 2016, five fully expanded leaves with no visible herbivore damage were collected from the top of separate stems within each enclosure. Pooled leaf samples were freeze dried with liquid nitrogen, ground, and extracted using 1 mL of HPLC-grade methanol from 100 mg of ground tissue. After centrifugation, the supernatant was filtered through a 0.22 µm filter and analyzed using a Hitachi Chromaster HPLC with a 5430 Diode Array detector. Separation was achieved using a Hypersil Gold C18 column (5 µm × 4.6 mm × 250 mm; Thermo Fisher Scientific). The mobile phase was a mixture of acetonitrile:water (v/v) at 20:80 from 0 to 5 min, a linear gradient of 20:80 to 95:5 from 5 to 45 min, 95:5 from 45 to 55 min, a linear gradient of 95:5 to 20:80 for 55–60 min, and 20:80 for 60–70 min. The flow rate was constant at 0.7 ml/min and the sample loading volume was 10 µL. Chemical constituents were separated by time of elution and the area of the peak used as an estimate of the amount present. Only peaks that were discernable from the baseline (> 75 µV s) were retained for analysis.

Chemical composition for all genotypes in the common garden was run through a non-metric multidimensional scaling (NMDS) analysis using Sorensen distance to generate independent axes of chemical variation. Peaks that occurred in 5 or fewer genotypes were omitted from analysis as uninformative, resulting in the inclusion of 117 peaks. The optimum number of dimensions for the ordination was determined by comparison to randomized data in PC-ORD 6.0 (MjM Software, Gleneden Beach, OR; McCune and Grace 2002). The three resulting axes were used to relate chemical composition to the strength and direction of plant–microbe interactions generated by each genotype (3D stress = 0.083). While above-ground chemistry may be related to plant–microbe interactions via leaching and decomposition (Veen et al. 2021), root-secreted chemicals often undergo modification in the soil, making adequate quantification difficult (Wang et al. 2020), so here we only focus on above-ground chemistry.

Greenhouse bioassay

In September 2019, the sixth growing season after planting approximately 500 ml of soil was collected from the top 10 cm within each genotype’s enclosure. Soil collection was done where S. altissima was most dense to avoid areas influenced by the roots of other species. By this time, the vast majority of biomass in most plots was contributed by S. altissima (B. S. Foster, Unpublished thesis). Collection tools were washed and sterilized with a 5% bleach solution between genotypes. Soil from each genotype was sifted to remove stones and root fragments, generating a live inoculum for each S. altissima genotype. From equal parts of each genotype, we also generated a pooled inoculum. Half of the pooled inoculum was autoclaved at 20 min at 120 °C and 0.07 MPa to produce a single control inoculum to compare to all individual genotype’s live soil inoculum. Because of the small size of the planting area and its previous use for row crops, abiotic soil conditions should have been relatively consistent across the site, justifying a single control inoculum. The remaining pooled inoculum was used live to assess whether sample pooling across genotypes would be equivalent to the mean interaction strength of the individual genotype’s microbial communities (question 4). This inoculum collection technique would contain mutualistic and antagonistic bacteria, fungi, and other soil organisms—collectively referred to here as soil microbes.

Two native species were used as targets in this experiment, Solidago altissima to detect potential direct population feedbacks and Schizachyrium scoparium to assess potential variation in microbe-mediated interspecific interactions. Seeds of S. altissima were collected from multiple individuals from the original source location of S. altissima genotypes (Douglas-Hart Nature Center) and were moist stratified for 60 d at 5 °C prior to planting. While seed collection may have included individuals originally collected to establish the common garden, clonal expansion had generated uniform cover of S. altissima, preventing isolation of any surviving individuals. Seeds of S. scoparium were purchased from Prairie Moon Nursery (Winona, MN, USA) and did not require stratification. Seeds of both species were initially grown in the greenhouse for two weeks in flats filled with soil-less potting medium (Pro-Mix BX, Premier Horticulture, Quakertown, PA, USA). This medium is not inoculated with any microbes and, while not sterile, would not possess any plant-specific microbes and was consistent across all experimental units.

Experimental plantings were done within cone-tainers (164 ml, Stuewe & Sons, Inc., Tangent, OR) filled 2/3 with the same soil-less potting medium. On top of that medium, a 10 ml layer of inoculum was added (one of 24 genotypes, live pooled, or pooled autoclaved) and mixed into the top 2 cm of potting mix. This small amount of inocula relative to growing medium (approximately 6% of volume) should minimize abiotic contributions to plant performance due to variation in soil quality across genotypes or the effects of autoclaving, but adequately represent the diversity of the soil microbial community (Howard et al. 2017). This level of inoculation should also be under the threshold for any allelopathic effects of S. altissima (Pisula and Meiners, 2010). The inoculum layer was capped with an additional 2.5 cm of soil-less potting medium to minimize potential cross-contamination. A seedling of either S. altissima or S. scoparium was transplanted into the top layer September 5, 2019. Seedlings dying within the first week were replaced; mortality following that resulted in reduced sample sizes. Initial replication was 20 for each target species/genotype combination and 40 per species for the pooled live and autoclaved control inocula. The range of final sample sizes was 15–20 (mean 19.3) for S. altissima and 14–20 (mean 17.5) for S. scoparium.

Racks of cone-tainers were placed on a single greenhouse bench and watered to saturation as necessary. After 60 d, above-ground biomass was harvested, dried at 60 °C for 4 days, and weighed. Previous work with these target species in this experimental system showed identical above- and below-ground responses to soil microbes (Peacher and Meiners 2020), and so below-ground biomass was not collected.

Data analysis

Biomass response of each target species to the soil inocula of the 24 genotypes and the autoclaved control were assessed with one-way ANOVA on inoculum identity (question 1). Genotype differences from the autoclaved control were determined from t tests of parameter estimates within the ANOVA. The overall consistency of genotype-specific soil impacts on target plant biomass of target species to inocula (24 genotypes + autoclaved control) was determined with a Pearson correlation between the means of each target species–inoculum combination (question 2). Means are used here as an estimate of the plant response to each inoculum. Similarly, the role of above-ground chemistry in influencing intraspecific variation in plant–microbe interactions was determined by relating ordination scores of S. altissima chemistry (3 NMDS axes, described above) to the average biomass of S. altissima and S. scoparium seedlings grown in each inocula using a multiple regression for each target species (question 3).

To assess whether inoculum produced from soil microbes pooled across genotypes yielded results equivalent to the average of genotypes tested individually (question 4), we conducted a two-way ANOVA of target species (two levels—S. altissima and S. scoparium) and soil handling (two levels—pooled and individual genotypes) effects on target plant biomass. For this analysis, we used the mean biomass for each target species grown in each genotypes’ soil microbiomes. Using averaged biomass instead of raw data for individual genotypes ensured that each genotype was weighed equally in the analysis (accounting for unequal sample sizes due to mortality) and resulted in a much more balanced design. All analyses were conducted in R version 3.5.1 (R Core Team 2016).

Results

There was remarkable variation in the impacts of individual genotypes’ soil microbial community on growth in both target species (question 1; Fig. 1). This variation was significant for both S. altissima (ANOVA: F24,474 = 9.63, P < 0.0001, R2 = 0.33) and S. scoparium (ANOVA: F24,426 = 4.89, P < 0.0001, R2 = 0.22). More importantly, the magnitude of soil microbial impacts of individual genotypes ranged from strongly inhibitory to being equivalent to autoclaved control inocula (neutral), to being strongly beneficial. Growth responses of S. scoparium tended to be more negative (12 genotypes) than positive (one genotype), whereas responses of S. altissima were slightly more positive (6 genotypes) than negative (three genotypes).

Fig. 1
figure 1

Variation in plant growth of Solidago altissima (top) and Schizachyrium scoparium (bottom) in response to the soil microbial communities of the 24 S. altissima genotypes. For each target species, genotypes are ordered from the smallest to largest plant mass, with closed symbols representing each of the microbial communities of the 24 S. altissima genotypes. The biomass of the control (open symbol) is indicated by the horizontal line in each panel. *Indicate genotypes with masses different from autoclaved controls. Data plotted are means ± 1 SE

Full size image

The responses of S. altissima to each soil microbial community were positively related to the responses of S. scoparium (question 2; Fig. 2; r = 0.604, P = 0.0018), although there was much variation in this relationship. Part of the variation in responses between target species may be explained by differences in their relationship to foliar chemistry. Growth response to the microbial community was related to the leaf chemistry of the genotype that produced the microbial source for S. scoparium seedlings, but not for S. altissima (question 3; Table 1). In Schizachyrium scoparium, plant biomass was positively associated with two of the three axes of plant chemistry (NMDS axes 2 and 3), accounting for nearly half of its variation in growth. Foliar chemistry of genotypes with inhibitory effects on S. scoparium biomass largely overlapped with that of genotypes with neutral effects (Fig. 3). However, the one genotype’s soil microbial community that enhanced the growth of S. scoparium had a distinct foliar chemistry.

Fig. 2
figure 2

Relationship between the biomass of both target species in response to the soil microbial communities of the 24 S. altissima genotypes. The biomass of plants in the autoclaved control is plotted as an open symbol. Values plotted are means for each plant-inoculum combination

Full size image
Table 1 Regression analysis of the impact of chemical composition (derived from an NMDS analysis) on the growth of Schizachyrium scoparium and Solidago altissima seedlings
Full size table
Fig. 3
figure 3

Variation in the foliar chemistry of the 24 S. altissima genotypes as determined by HPLC. Data plotted are two dimensions from an NMDS ordination of chemical composition. As only axes two and three were related to plant performance (Table 2), only those two axes are presented. Symbols represent the effects of S. altissima genotypes on S. scoparium growth, including neutral (closed circles), inhibitory (open circles), and beneficial (triangles) effects

Full size image

The results of the pooled live inoculum did not reflect the average of the bioassays conducted on the genotypes individually (question 4; Table 2; Fig. 4). For both target species, pooled inocula resulted in lower plant growth than the average of each genotype’s soil microbes analyzed separately. The magnitude of the pooled inoculum’s effects on plant growth were in line with the more antagonistic soil microbial communities that were found when genotypes were tested individually.

Table 2 Effects of inoculum pooling on plant biomass
Full size table
Fig. 4
figure 4

Impact of sample handling (pooled vs. individual) on the growth of both target species. Individual data represent the average growth in each of the 24 S. altissima genotypes tested separately. Pooled data represent the average seedling masses for plants grown in the pooled live inoculum

Full size image

Discussion

Our results show an appreciable level of intraspecific variation in the strength of plant–soil microbe interactions across S. altissima genotypes. Our data strongly suggest that intraspecific variation within S. altissima chemical profiles generates variation in soil microbial communities that represent a prominent source of interaction heterogeneity within communities that this species dominates.

Variation in plant chemistry and associated soil microbes is likely generated by a combination of ecological and genetic factors. For example, soil conditioning via litter inputs generate changes in soil conditions, soil microbial communities, and ecosystem function (Veen et al. 2019). Soil microbial communities of S. altissima change during succession, leading to increased herbivore resistance (Howard et al. 2020). Defensive chemistry and competitive ability in S. altissima have a strong genetic component (Genung et al. 2012; Heath et al. 2014), which can directly influence the quantity and quality of litter and soil inputs (Schweitzer et al. 2008; Uesugi and Kessler 2013). As an added level of complexity, soil microbial communities alter plant chemistry in S. altissima (Meiners et al. 2017) in ways that may additionally alter plant–plant interactions. Regardless of the underlying mechanisms, soil microbial impacts clearly varied across genotypes (Schweitzer et al. 2008; Bukowski et al. 2018; Eck et al. 2019).

The implication of genotype-specific variation in plant–microbe is that it may influence plant persistence within the community (Reynolds et al. 2003; Kardol et al. 2006; Bauer et al. 2015). Approximately half of the genotypes produced soil microbial communities that were antagonistic to S. altissima, which should promote local turnover and diversity (Liu et al. 2015; Eck et al. 2019). The soil microbial communities of 3/24 Solidago genotypes improved S. scoparium growth and were effectively neutral for another five genotypes, suggesting these S. altissima genotypes would be more invasive by other species. Variation in the plant–microbe interactions of individual genotypes may, in combination with competition and herbivore pressure, generate the successional filtering of S. altissima genotypes over time (Hartnett and Bazzaz 1985). As the individual S. altissima plants that initiated the planting were all early colonists, many of these genotypes would be displaced over time, contributing to changes in plant chemistry and soil microbial communities (Howard et al. 2020).

The correlation between the impacts of genotype’s soil microbes on plant performance in both target species suggests there are general components of the soil microbial communities that consistently impact plant species (Harrison and Bardgett 2010). As S. altissima interacts strongly with soil microbial communities (Zhang et al. 2007; Pendergast et al. 2013; Howard et al. 2020) and our genotypes varied substantially in foliar chemistry (Fig. 3), we expected variation in the resulting plant–microbe interactions. The impacts of soil microbial communities on plant performance were related to genotype foliar chemistry only for S. scoparium seedlings. This relationship appears to be due to the culturing of soil biota by the original S. altissima genotypes that generated species-specific feedbacks on S. altissima growth (Liu et al. 2015; Gehring et al. 2017; Eck et al. 2019), independent from plant chemistry. As microbes specific to S. altissima may not interact as strongly with S. scoparium, variation in soil microbes associated with foliar chemistry resulted in strong impacts of soils from most S. altissima genotypes.

While we invoke genotype-specific soil microbes as the most likely driver of our experimental results, other potential mechanisms exist. As the common garden was a small portion of a row-crop field prior to plot establishment, we expect initial soil microbial communities and abiotic conditions to be largely uniform and naïve to S. altissima when the genotypes were planted. The small size of the inoculum used in the experiment (≈ 6% by volume) should have been sufficient to transfer the soil microbial community (Howard et al. 2017), but should have contributed minimal nutrient or allelochemical inputs, well below the threshold detected in bioassays (Pisula and Meiners 2010). Therefore, the most plausible explanation of variation in target plant growth relative to autoclaved controls is the generation of genotype-specific soil microbial communities. However, this must be verified with sequencing of soil fungi and bacteria in soils cultured with S. altissima only. As genotype collections were not genetically verified, we may have collected multiple individuals from the same source genotype. However, variation in foliar chemistry shows distinct differences among most genotypes.

The variation in the strength and direction of plant–microbe interactions documented in this study presents an experimental challenge to researchers. The pooling of individual soils is often utilized in an attempt to better approximate the properties of natural, highly heterogeneous soils (Robertson et al. 1999). However, this experimental approach only works when the magnitude and direction are unchanged by sample pooling. In contrast, we found that soil pooling failed to reproduce the mean of the individual soil microbial communities’ effects on either Solidago altissima or Schizachyrium scoparium. Clearly, inocula formed by soil pooling produced markedly different impacts via plant–microbe interactions (Reinhart and Rinella 2016; Peacher and Meiners 2020). The pooling effect may be due to soil microbial beta diversity, leading to turnover among S. altissima plots and greater diversity in pooled inocula, but this effect may vary in a species-specific manner (Allen et al. 2021), generating different plant responses.

Although plant–microbe interactions for both target species ranged from positive to negative, pooled inocula generated consistently antagonistic effects on plant growth. Soil pooling may make important soil microbes present in more samples, allowing them to more frequently impact plant performance (Schnitzer et al. 2011; Peacher and Meiners 2020). In this study, pooled samples appear to have been more heavily influenced by the more antagonistic components of S. altissima soil microbial communities, yielding biased estimates of microbial impacts on both target species. Overall, plant–microbe interactions tend to be predominantly inhibitory (Kulmatiski et al. 2008), consistent with our results. However, we found variation in the strength and direction of microbial effects, representing a heterogeneous interaction landscape with potentially important implications for succession and competitive outcomes (Schnitzer et al. 2011; Pendergast et al. 2013; Mack and Bever 2014).

Ecological heterogeneity is a critical component in many plant systems, although it is often viewed as unwanted noise (Kolasa and Rollo 1991). We document biotically generated interaction heterogeneity (Pickett et al. 2000), related to inter-clonal variation in soil microbes. Interspecific differences in plant–soil microbe interactions have important impacts on population persistence and community dynamics (Van Der Heijden et al. 2008; Schnitzer et al. 2011; Liu et al. 2015; Eck et al. 2019). However, implications for intraspecific variation in plant–soil microbe interactions are difficult to predict due to the diversity of interactions between plants and soil microbes (Bukowski et al. 2018; Magnoli and Lau 2020), but may be as important as other aspects of intraspecific trait variability (Violle et al. 2012). As clonal organisms are major components of many plant communities that interact with large volumes of soil, understanding inter-clonal variation represents a critical opportunity for exploring plant–soil microbe interactions.