SUMMARY
Invasive species may benefit from introduction to new regions where they can escape their natural enemies. Here we examined whether geographic patterns of microbial community composition support a role for enemy escape in the invasion of California, USA by yellow starthistle, a highly invasive plant in western North America.
We used high-throughput sequencing of the 16S V4 region to characterize bacterial community composition in the phyllosphere, rhizosphere, leaves, and roots of plants from seven populations in California and eight populations in the native European range. We compared bacterial diversity between the native and invaded ranges, and with previously published estimates of plant genetic diversity within each population.
Bacterial communities differed significantly among plant compartments, and between native and invaded ranges within compartments, with consistently lower diversity in the invaded range. Plant genetic diversity did not explain this pattern in bacterial diversity, but a positive relationship was found within ranges between bacterial diversity in roots and plant genetic diversity within populations.
Our observation of lower bacterial diversity in the invaded relative to the native range of yellow starthistle is consistent with potential enemy escape, providing some of the first evidence for this scenario in plant microbiomes.
INTRODUCTION
Humans continue to transport plant species around the globe, and increasing numbers of these translocations result in the invasive expansion of non-native species into recipient communities (Lonsdale, 1999; Butchart et al., 2010; Essl et al., 2011; Ellis et al., 2012). There is a longstanding hypothesis that many species become invasive after escaping from enemies that reduce invader fitness and limit their populations in their native ranges (Darwin, 1859; Williamson, 1996). Known as the ‘Enemy Release’ hypothesis, this idea is highly intuitive and forms a basis for the biological control of invasive species (Keane & Crawley, 2002). Initial tests of enemy release focused on quantifying visible changes in above-ground herbivore damage (Keane & Crawley, 2002), but there has been increasing recognition that microbial enemies above- and below-ground can have large effects on plant fitness, and could thus determine whether invasive plants benefit from escaping negative species interactions (Callaway et al., 2004; Colautti et al., 2004; Torchin & Mitchell, 2004; Agrawal et al., 2005; Mitchell et al., 2006; Kulmatiski et al., 2008; van der Putten et al., 2013; Dawson & Schrama, 2016; Faillace et al., 2017).
In recent years, microbial communities have emerged as particularly likely candidates for generating enemy release. Although many interactions between plants and microbes can be beneficial, microbial communities often appear to have negative net effects on plant fitness which may become more negative over time, e.g., via plant-soil feedbacks (Bever, 2003; Reinhart & Callaway, 2006; Kulmatiski et al., 2008; Petermann et al., 2008). It is now apparent that interactions between plants and their microbiomes can vary over space and environment (Nemergut et al., 2013; van der Putten et al., 2013; terHorst & Zee, 2016), creating opportunities for introduced plants to escape the microbial communities that characterize their native ranges. Moreover, evidence is building that reductions in microbial diversity are occurring in response to environmental change and human disturbances, and these reductions in diversity may reduce the resistance of ecosystems to invasion (Schnitzer et al., 2011; Wagg et al., 2014; Dawson & Schrama, 2016; van der Putten et al., 2016).
Invasive plant species have provided some of the best evidence to date that microbial interactions can be locally evolved, and can vary considerably over geographic regions (Rout & Callaway, 2012). Invaders have been shown to vary in their response to soil communities from their native and invaded ranges, and there are now many examples of more favorable interactions between plants and soil from the invaded range, consistent with escape from enemies (Reinhart et al., 2003; Callaway et al., 2004; Mitchell et al., 2006; Engelkes et al., 2008; Kulmatiski et al., 2008; Maron et al., 2014; van der Putten et al., 2016). Plant-microbe interactions which provide net benefits to invasive species can be explained by reduced negative effects of key microbial pathogens, increased direct beneficial effects of mutualistic taxa, or increased indirect benefits from taxa that affect competitors more negatively than they do the invader (Dawson & Schrama, 2016). These mechanisms should manifest as differences in the microbial communities associated with invading vs. native plants, specifically as divergence in taxonomic composition, reduction in diversity, and/or the loss or gain of groups known to have pathogenic or mutualistic effects, where taxonomic resolution permits inference of function (Herrera Paredes & Lebeis, 2016).
Release from enemies is expected to be beneficial in and of itself, but it may further promote invasion by changing the pattern of natural selection on resource allocation by the invader (Sakai et al., 2001). Plants that require reduced defenses against negative enemy interactions have the potential to adapt to invest a larger proportion of resources in traits that increase competitiveness, reproduction, and/or spread. This idea, known as the Evolution of Increased Competitive Ability (EICA) hypothesis (Blossey & Notzold, 1995) has received a great deal of attention but mixed empirical support (Maron et al., 2004; Bossdorf et al., 2005; Mitchell et al., 2006; Felker-Quinn et al., 2013). Potential contributors to evolutionary responses to enemy release are likely to become better resolved as our understanding of microbial community interactions increases, particularly since adaptive responses to microbial disease are known to be among the most rapid evolutionary changes that occur in any organism (Tiffin & Moeller, 2006; Bomblies et al., 2007; Salvaudon et al., 2008).
Here we conduct one of the first comparisons of plant microbiomes between invading populations and their native source region, explicitly testing for patterns consistent with enemy release (see also Gundale et al., 2016). We ask whether changes in plant-associated microbial communities have the potential to generate enemy escape in the highly invasive plant yellow starthistle (Centaurea solstitialis). Yellow starthistle is native to a wide region of Eurasia and was introduced to South America in the 1600’s and North America in the 1800’s as a contaminant of alfalfa seed (Gerlach, 1997). This herbaceous annual is a colonizer of grassland ecosystems, and is often called one of the ‘10 Worst Weeds of the West’ in North America (DiTomaso & Healy, 2007). Its extensive invasion of California in the USA (>14 million acres; Pitcairn et al., 2006) is well-studied, and invading genotypes in this region have been shown to grow larger and produce more flowers than plants in the native range, suggesting an adaptive shift in resource allocation and an increase in invasiveness (Widmer et al., 2007; Eriksen et al., 2012; Dlugosch et al., 2015). Previous research has demonstrated that yellow starthistle throughout all of its native and invaded ranges experiences net fitness reductions when grown with its local soil communities (Andonian et al., 2011, 2012; Andonian & Hierro, 2011). However, these studies have also indicated that this negative interaction is weaker (more favorable) in California, raising the possibility that escape from microbial enemies has promoted this aggressive invasion.
We sample microbial communities associated with leaves (phyllosphere and endosphere) and roots (rhizosphere and endosphere) of yellow starthistle plants in both the California invasion and its source region in Europe. Previous experiments with fungicide treatments have shown that plant-soil interactions between yellow starthistle and fungi in California are more negative (less favorable) than those in the native range, inconsistent with a role for fungi in escape from microbial enemies (Hierro et al., 2016). Here, we focus on bacterial communities as candidates for a potential role in enemy escape in this system. We use high-throughput sequencing of prokaryotic ribosomal 16S sequences to quantify diversity and relative abundance of taxa in yellow starthistle microbiomes, designing a novel modified peptide nucleic acid clamp (Lundberg et al., 2013) to reduce non-target sequencing of host plastids. We ask whether there are patterns of reduced taxonomic diversity and/or potential loss of pathogens in the invaded range, and whether patterns of diversity in plant-associated bacteria can be explained by geographic patterns of plant genetic diversity. Loss of potential pathogens would be consistent with opportunities for enemy escape that could contribute to the evolution of increased invasiveness in yellow starthistle’s highly successful invasion of California.
MATERIALS AND METHODS
Study species
Yellow starthistle (Centaurea solstitialis L., Asteraceae) is an obligately outcrossing annual plant, diploid throughout its range (Heiser & Whitaker, 1948; Widmer et al., 2007; Öztürk et al., 2009). Plants form a taproot and grow as a rosette through mild winter and/or spring conditions, bolting and producing flowering heads (capitula) throughout the summer. The species is native to Eurasia, where distinct genetic subpopulations have been identified in Mediterranean western Europe, central-eastern Europe, Asia (including the Middle East), and the Balkan-Apennine peninsulas (Barker et al., 2017). The invasion of California as well as those in South America appear to be derived almost entirely from western European genotypes (Fig. 1; Barker et al., 2017).
Sample collection
Fifteen populations of yellow starthistle were sampled for microbial communities: seven populations across the invasion of California, six in western Europe, and two in eastern Europe (Fig. 1; Supporting Information Table S1). At each population, plants were sampled every meter (or to the nearest meter mark) along a 25 meter transect, to yield 25 individuals per population. Individuals in rosette or early bolting stages were preferentially selected. In one population (HU29), low plant density yielded 20 individuals along the 25 meter transect.
Using sterile technique, plants were manually pulled and each individual sampled for phyllosphere, rhizosphere, leaves, and roots using modified versions of protocols by Lundberg et al. (2013) and Lebeis et al. (2015) as described below. A control (blank) sample was collected for each population. Plants were pressed and dried after sampling, and submitted to the University of Arizona Herbarium (ARIZ; Supporting Information Table S1).
Phyllosphere and rhizosphere — one to three basal, non-senescent leaves were collected from each plant, as well as the upper 2-5 cm of the taproot, together with accompanying lateral roots (excess soil was brushed or shaken off). Leaf and root samples were placed in individual 50 ml tubes containing 25 ml of sterile wash solution (45.9 mM NaH2PO4, 61.6 mM Na2HPO4, 0.1% Tween 20). Tubes were shaken by hand for one minute (timed). Leaf and root samples were then removed and stored on ice in separate tubes (leaves in empty tubes, roots in tubes containing 10 ml of wash solution) until further processing. Wash samples were stored on ice during transport, then refrigerated at 4°C. Phyllosphere and rhizosphere washes were pooled per population, then centrifuged at 2,200 g at 4°C for 15 minutes. Supernatants were discarded, and pellets were air-dried and stored at -20°C until DNA extraction.
Leaf endosphere — leaves were surface sterilized by submerging in bleach solution (10% commercial bleach, 0.1% tween 20) for two minutes. Leaves were then rinsed in distilled water, patted dry using clean kimwipe, and sealed in individual sterile surgical envelopes (Fisherbrand #01-812-50). Envelopes were kept in silica gel desiccant until leaf tissue was completely dry, then stored at room temperature until DNA extraction.
Root surface and endosphere (hereafter ‘whole root’) — roots were further washed by shaking in 10 ml of wash solution until visible residual soil was removed. Washed roots were stored and dried as described above for leaves.
Controls — at each collection site, a tube of sterile wash solution was left uncapped while sampling plants. Disinfected tools were periodically swished in the blank wash tube before sterilization and use for the next sample collection. For each population, rinse water and wipes used to process tissue samples were represented in controls by rinsing and wiping flame-sterilized forceps, then swishing the forceps in the blank wash tube. Controls were stored and processed in the same manner as phyllosphere and rhizosphere samples.
DNA extraction
Extractions were carried out using sterile technique in a laminar flow hood. Leaf and root DNA was extracted as bulk samples from tissue pooled by population (15 total populations), and as individual samples from 8 plants from each of 10 populations (80 total individuals). For pooled tissue extractions, equal sections of leaf tissue (50 mm2) and root tissue (12.5 mm3 plus 10 mm of lateral roots) were collected from each individual sample per population and pooled prior to extraction. Control (blank) samples were collected for each batch of extractions by swabbing tools and surfaces, then extracting DNA from the swab head.
All DNA samples were extracted using the MO BIO PowerSoil kit (MO BIO Laboratories, Inc.). Phyllosphere and rhizosphere DNA was extracted from up to 0.25 g of wash pellets following the standard kit protocol. Leaf and root tissues were ground to powder or sawdust consistency in liquid nitrogen using sterile mortars and pestles. Leaf and root DNA was extracted from 20 mg (leaf) or 100 mg (root) of ground tissue with the following modification to the standard protocol: tissue was incubated at 65°C for 10 minutes in extraction buffer, then vortexed for 1 minute, followed by a second 10 minute incubation (as described under “alternative lysis methods” in the kit protocol). Control DNA was extracted by placing whole swab heads directly into extraction tubes. Extracted DNA was eluted in PCR-grade water and stored at -20°C pending library preparation.
Library preparation and sequencing
To remove secondary compounds inhibiting PCR, DNA extracted from root and leaf tissue (together with corresponding blanks) was purified using a ZR-96 genomic DNA clean-up kit (Zymo Research). All DNA concentrations were quantified using a Qubit fluorometer high-sensitivity assay for double-stranded DNA (Invitrogen), and standardized to equimolar amounts.
Library preparation followed a dual barcoded two-step PCR protocol. In the first step (target-specific PCR), the V4 region of the 16S rRNA gene was amplified using target specific primers (515F and 806R) appended with common sequence (CS) tags through a linker sequence which varied from two to five nucleotides in length. Target-specific PCR was carried out using Phusion Flash master mix (Thermo Scientific) in 25 μl reaction volume in a Mastercycler pro thermocycler (Eppendorf) under the following conditions: 25 cycles of 1 s at 98°C, 5 s at 78°C, 5 s at 57°C, 15 s at 72°C. Products were visualized on an agarose gel and diluted by up to 1:15 (depending on yield); 1 μl of diluted product was then used as template in the second step (barcode-adapter attachment PCR). Using reagents and equipment as described above, barcoded primer pairs incorporating Illumina P5 and P7 adapters were used to amplify products from target-specific PCR in 25 μl reaction volumes under the following conditions: 10 cycles of 1 s at 98°C, 5 s at 78°C, 5 s at 51°C, 15 s at 72°C. Barcoded amplicons were quantified by fluorometry, pooled in equimolar amounts, cleaned, and submitted to the University of Idaho’s IBEST Genomic Resources Core for QC and sequencing. Amplicons were multiplexed to use half the capacity of one 2 × 300 bp run on an Illumina MiSeq platform. Raw sequence data are deposited in the NCBI Short Read Archive under accession number XXXXXX [pending submission].
Peptide nucleic acid clamps (PNAs) were included in both PCR steps of library preparation to block amplification of plant chloroplast and mitochondrial 16S as recommended by Lundberg et al. (2013). Clamp sequences published by Lundberg et al. (2013) were compared with chloroplast and mitochondrial 16S sequences from yellow starthistle and three other species of Asteraceae with published organellar genomes (Centaurea diffusa, Helianthus annuus, Lactuca sativa). We found a single nucleotide mismatch between the Asteraceae chloroplast 16S and the plastid PNA sequences, and designed an alternative plastid PNA specific to the Asteraceae sequence (5’—GGCTCAACTCTGGACAG—3’). All samples for this study were amplified using the plastid PNA of our design, together with the mitochondrial PNA published by Lundberg et al. (2013). To gauge the effectiveness of our alternative PNA, two duplicate samples were processed using both PNAs published by Lundberg et al. (2013).
Identification of operational taxa
Demultiplexed paired reads were merged and quality filtered using tools from the USEARCH package version 9.0.2132 (Edgar & Flyvbjerg, 2015). Merged reads were truncated to uniform length and primer sequences were removed using a combination of the seqtk toolkit version 1.2 (github.com/lh3/seqtk) and a custom script. The UPARSE pipeline (Edgar, 2013) implemented in the USEARCH package was used for further data processing and analysis: unique sequences were identified, and those represented only once or twice in the processed read set were discarded as likely PCR or sequencing errors. Remaining sequences were clustered into operational taxonomic units (OTUs) at a 97% threshold, chimeras were filtered out, and per-sample OTU read counts were tabulated using the UPARSE-OTU algorithm. Assignment of OTUs to nearest taxonomic match in the Greengenes database (McDonald et al., 2012) was carried out using the UCLUST algorithm implemented in QIIME version 1.9.1 (Caporaso et al., 2010; Edgar, 2010). Data were further processed using tools from the QIIME package: reads mapping to chloroplast and mitochondrial OTUs were removed, and samples were rarefied by plant compartment. Rarefaction levels were chosen to reflect the distribution of read counts per sample within plant compartments, subsampling to the minimum number of reads necessary to include all samples except those that were outliers for low read count.
Microbial community analyses
All statistical analyses were performed in R (R Core Team, 2015). We evaluated overall differences in bacterial community composition between plant compartments, and between native and invaded ranges within plant compartments, by performing non-metric multidimensional scaling (NMDS) using the R packages vegan (Oksanen et al., 2016) and MASS (Venables & Ripley, 2002). Ordinations were based on Bray-Curtis distances, and were performed using a two-dimensional configuration to minimize stress, using Wisconsin double standardized and square root transformed data, with expanded weighted averages of species scores added to the final NMDS solution. Significant differences among plant compartments and between native and invaded samples were assessed using the envfit function in vegan. Ellipses were drawn on NMDS plots using the vegan function ordiellipse, representing 95% confidence limits of the standard error of the weighted average of scores.
We further explored the underlying correlates of bacterial community variation using Principal Components Analysis (PCA; using R function prcomp) for samples from native and invaded ranges within each plant compartment. We identified the OTUs with the highest loading on the dominant PC axis of variation by examining the matrix of variable loadings produced by prcomp. The OTU composition of samples pooled by population (phyllosphere, rhizosphere, and bulk root samples; hereafter ‘bulk samples’) was visualized using a heatmap generated in ggplot2 (Wickham, 2009). Bulk samples were hierarchically clustered by Bray-Curtis dissimilarity (hclust function in R) using McQuitty’s method (McQuitty, 1966).
We compared the diversity of OTUs between the native and invaded range for each plant compartment using both richness (R) and the Shannon diversity index (H’; Shannon, 1948), which reflects the contributions of both taxonomic richness and evenness to diversity. These values were calculated using the vegan package, and compared between native and invaded ranges using a nonparametric Kruskal-Wallis rank sum test on rarefied read counts. For plant tissue samples that included multiple individuals per site, we compared the diversity among sites using a Kruskal-Wallis test within regions.
Finally, we asked whether the geographic distribution of plant-associated bacterial diversity could be explained by the geographic distribution of genetic diversity in the plants. Measurements of plant genetic diversity at each of our sampling sites were obtained from previously published genome-wide marker analyses by Barker and colleagues (Barker et al., 2017), calculated as the average proportion of pairwise nucleotide differences between alleles (π) at variable sites across the yellow starthistle genome. Diversity estimates (H’) for each plant compartment were predicted using linear models that included fixed effects of plant genetic diversity, region (native vs. invaded), and the interaction between these two effects.
RESULTS
Sequencing and data processing
Sequencing yielded 9,672,898 read pairs, of which 6,217,852 remained after merging and quality control; these were 253 bp in length after removing artificial and primer sequences. The number of raw read counts per sample ranged from 16 to 306,200 with a median of 21,964.
Analysis of the merged and processed reads resulted in 4,014 OTUs, of which 60 were identified as plastid or mitochondrial. Sequences representing yellow starthistle chloroplast and mitochondrial 16S accounted for 40% and 1% of all reads, respectively. Amplification of host chloroplast in samples using the Asteraceae-specific plastid PNA was reduced by up to 51% compared with the Lundberg et al. (2013) PNA (Supporting Information Table S2). This is consistent with results from a broader comparison of the two PNAs, using five Asteraceae species, which also found that blocking of host chloroplast amplification was improved by using the Asteraceae-specific PNA (FitzPatrick et al., unpublished).
Despite PNA blocking activity, 83% of the total reads from leaf endosphere samples were yellow starthistle chloroplast sequences. This might potentially be attributable to high chloroplast DNA concentrations relative to endophyte DNA in leaf tissue total DNA extracts. After removal of chloroplast and mitochondrial reads, remaining read counts for most leaf endosphere samples were low (Fig. 2), so no further analysis of leaf endosphere bacterial communities was performed.
Rarefaction levels (chosen to reflect the minimum number of reads per sample by compartment, not including outliers) were 18,000 reads per sample for phyllosphere, 17,000 for rhizosphere, and 5,000 for whole root samples. These levels resulted in the exclusion of six samples which were outliers for low read count: one phyllosphere (DIA), one rhizosphere (SAZ), one bulk root (CAN), and three individual root samples (two from SAZ; one from SIE).
Microbial community analyses
Results from NMDS ordinations indicated that bacterial communities differed overall among the phyllosphere, rhizosphere, and whole root compartments (Fig. 3a; P = 0.001). Within compartments, NMDS further revealed significant differences between native and invaded range whole root samples (Fig. 3b; P = 0.001) and rhizosphere samples (P = 0.001). Native and invaded range phyllosphere samples differed with marginal significance (P = 0.05). The dominant phyla among bacterial communities were Proteobacteria, Actinobacteria, Bacteroidetes, and Firmicutes (Fig. 4), which is consistent with the findings of previous characterizations of plant-associated bacterial communities (reviewed by Bulgarelli et al., 2013). Principal component analyses suggested that the strongest contributions to changes in bacterial community composition between the native and invaded ranges were made by shifts in the representation of Pseudomonas, Erwinia, Chryseobacterium, Xanthomonadaceae, and Bacillus taxa (Supporting Information Fig. S1; Table S3). Clustering analyses within the phyllosphere and rhizosphere compartments consistently grouped invaded range samples together, as well as samples from the source region in western Europe (Supporting Information Fig. S2). Native range samples from eastern Europe (HU01 and HU29) clustered together in these compartments but were variable in their relationship to the other regions. Bulk root samples showed less consistent clustering by range.
Bacterial OTU diversity (H’: Fig. 5) was significantly lower in the invaded range in the phyllosphere (X21 = 5.36, P = 0.02) and rhizosphere (X21 = 6.21, P = 0.01), and marginally lower in bulk whole root samples (X21 = 3.01, P = 0.08). Diversity in whole root individual samples (Fig. 5d) did not vary significantly among populations within the native range (X24 = 1.82, P = 0.77), but did vary significantly within the invaded range (X24= 15.30, P = 0.004), such that the two most extreme populations (TRI, SIE) were significantly different from one another but not the remaining sites. Bacterial OTU richness (R) values showed patterns similar to H’ in general, but differences between native and invading regions were much weaker overall (Supporting Information Fig. S3), indicating that both richness and evenness of OTU representation contributed to differences in diversity between the ranges. For whole roots, our most extensively sampled plant compartment, an analysis of OTUs observed across all bulk and individual samples combined indicated that the native and invaded range shared 51% of observed OTUs, with 31% fewer unique OTUs observed in the invaded relative to the native range (Fig. 6a). These patterns were reflected across both major groups of Proteobacteria and Actinobacteria (Fig. 6b,c).
A linear model predicting microbial diversity (H’) from plant diversity was significant for bulk whole root samples (Fig. 7; F(2,12) = 4.99; P = 0.02; r2adj = 0.36), with significant main effects of both plant genetic diversity (P = 0.04) and region (native vs. invaded; P = 0.009). The interaction between these two effects was not significant (P = 0.69) and was removed from the final model. Similar linear models did not identify significant effects of plant diversity when predicting phyllosphere (P = 0.42) or rhizosphere (P = 0.11) diversity, nor the median diversity of individual whole root samples at a site (P = 0.95).
DISCUSSION
Our study revealed strikingly lower diversity of bacterial communities in yellow starthistle’s invasion of California (USA), relative to native European populations within and beyond the source region for the invasion. Reduced bacterial diversity was apparent across phyllosphere, rhizosphere, and whole root communities, and across dominant bacterial phyla. These patterns are consistent with opportunities for enemy escape in this invasion and could explain why soils have more favorable (less negative) effects on yellow starthistle fitness in California relative to other parts of its range (Andonian et al., 2011, 2012; Andonian & Hierro, 2011).
In line with other surveys of plant microbiomes, we found that differences in bacterial communities were greatest among plant compartments (i.e., phyllosphere, rhizosphere, and roots; (Bulgarelli et al., 2013; Vandenkoornhuyse et al., 2015). The numbers and diversity of taxa (OTUs) that we recovered in samples from each compartment were generally similar in magnitude to those reported in other studies of prokaryotic 16S sequences, from angiosperm groups as diverse as e.g. Agavaceae (Coleman-Derr et al., 2016), Brassicaceae (Bodenhausen et al., 2013), Cactaceae (Fonseca-García et al., 2016), and other Asteraceae (Leff et al., 2016). Notably, we found that diversity was approximately twice as high in the roots than the rhizosphere. Higher root endosphere diversity relative to the rhizosphere is also reported in some other recent studies (Fonseca-García et al., 2016; Leff et al., 2016), but previous reviews have concluded that root endosphere communities are typically less diverse than those in the rhizosphere (Bulgarelli et al., 2013; Vandenkoornhuyse et al., 2015). Our root collections were not surface sterilized and may represent some of the rhizoplane/rhizosphere in addition to the endosphere, elevating our estimates of diversity, though it is also possible that yellow starthistle deviates from previous trends. In contrast to phyllosphere, rhizosphere, and root compartments, we recovered few microbial sequences for leaf bacterial endophytes, even when compared to control (blank) samples, suggesting low sequence coverage due to persistent chloroplast contamination, and potentially low overall bacterial loads within yellow starthistle leaves.
Within compartments, differences in community composition between ranges were substantial and dominated by differences in OTU diversity, which ranged from 18-40% lower in the invaded range. This variation in diversity is similar in scale to other studies that have sampled distant geographic locations (e.g. across regions of North America: (Bodenhausen et al., 2013; Peiffer et al., 2013). A variety of factors may explain this pattern, including environmental differences across sites (Fierer & Jackson, 2006; Bulgarelli et al., 2013; Nemergut et al., 2013; Vandenkoornhuyse et al., 2015). Soil type appears to have a particularly strong influence on microbial communities (e.g. Bulgarelli et al., 2012; Lundberg et al., 2012), and is known to differ broadly across yellow starthistle’s range (Hierro et al., 2016). In addition, populations in California include temperature and precipitation environments that are on the warm and dry extreme of yellow starthistle’s climatic niche (Dlugosch et al., 2015), and our sampling was conducted at the end of a period of severe drought in California (Griffin & Anchukaitis, 2014; Diffenbaugh et al., 2015), which could have amplified microbial differences related to climate (Schrama & Bardgett, 2016). Interestingly, a recent study of grassland plants found that microbial diversity increased under drought, whereas we found reduced diversity in our drought-affected invaded range (terHorst et al., 2014).
We also observed an effect of plant genotypic diversity on microbial diversity in bulk samples of roots. We note that bacterial diversity at the level of the individual plant did not covary with plant population genetic diversity, indicating that it was only when samples from different plants were combined that an effect of plant genotype variation was apparent. Such within-species plant genotype effects have been observed in other studies and may interact with the effect of environment to shape microbial communities (Peiffer et al., 2013; terHorst & Zee, 2016). In many cases, it appears that genotype effects are minor relative to site effects (e.g. (Lundberg et al., 2012; Peiffer et al., 2013; Bodenhausen et al., 2014; Bulgarelli et al., 2015), and we also found that genotypic effects were not stronger than between-region variation in yellow starthistle. Moreover, plant genotype effects were only significant in the roots, the only endophytic compartment analyzed: consistent with plant genotype having the strongest influence on microbial taxa colonizing within the plant itself (Bulgarelli et al., 2012; Lundberg et al., 2012; Lebeis et al., 2015). The ability of plants to shape the composition and quantity of endophytic microbial taxa in particular has been suggested to contribute to a ‘core’ microbiome shared across environments (Lundberg et al., 2012; Shade & Handelsman, 2012; Reisberg et al., 2013), and indeed we found that the majority of OTUs observed in yellow starthistle roots were shared across invading and native populations on different continents.
Importantly, yellow starthistle’s invasion to high density could be a cause rather than an effect of low microbial diversity. Species invasions and range expansions have been shown to change microbial composition over short timescales (Collins et al., 2016; Gibbons et al., 2017). Yellow starthistle invasions are denser than populations surveyed in the native range by an order of magnitude or more (Uygur et al., 2004; Andonian et al., 2011). Invaded communities that are dominated by yellow starthistle include lower diversity of plant species overall (Seabloom et al., 2003; Zavaleta & Hulvey, 2004; D’Antonio et al., 2007), and low plant diversity may depress the diversity of plant-associated microbes in the environment (Garbeva et al., 2004; Schnitzer et al., 2011; Coleman-Derr et al., 2016). Such an effect of plant density could provide an explanation for a general pattern of weaker plant-soil interactions for invasive species in their introduced ranges (Kulmatiski et al., 2008; Dawson & Schrama, 2016). These results reinforce a growing need for explicit observational and experimental tests of the association between microbial diversity and the potential influences of plant density, plant community diversity, and environmental gradients (Dawson & Schrama, 2016).
This study is among the first to examine differences in microbial taxa between the native and introduced ranges of an invasive species. Gundale and colleagues (Gundale et al., 2014) also identified more favorable soil interactions in invasions of lodgepole pine (Pinus contorta), and explored potential enemy escape in its fungal endophyte community (Gundale et al., 2016). For lodgepole pine, microbial communities differed among regions, but there was no consistent pattern of loss of potential fungal pathogens or gain of mutualists in the invaded range, and it remains unclear what part of the soil community is responsible for observed differences in interactions across ranges (Gundale et al., 2016). Finkel and colleagues (Finkel et al., 2011; 2016) similarly explored the phyllosphere community of multiple species of Tamarix in native and introduced parts of their range, finding that communities are most strongly structured by geographic region. Our study reveals that this type of comparative microbiome approach can be fruitful for identifying changes in species interactions that might be contributing to invasion success.
One of the central challenges in testing the hypothesis that invaders are benefitting from enemy release is quantifying the impact of all types of enemies, with the microbial community being historically the hardest to observe (Keane & Crawley, 2002; Beckstead & Parker, 2003; Dawson & Schrama, 2016; Müller et al., 2016; van der Putten et al., 2016; Crawford & Knight, 2017). For yellow starthistle, a great deal of effort has gone into the identification of potential native herbivores/seed predators that could be used as biocontrol in California. Six specialist biocontrol insect species and one fungal foliar pathogen have been released into this area without resulting in effective control (DiTomaso et al., 2006; Swope & Parker, 2012), suggesting that escape from these species has not facilitated the invasion. Our finding that invaded populations have not only unique, but also less diverse bacterial communities, suggests particularly strong opportunities for pathogen escape in this system.
We have previously argued that yellow starthistle has invaded into a low competition environment in California, benefitting from the historical loss of plant competitors for water in this system (Dlugosch et al., 2015). Disturbance of the native community is critical for yellow starthistle establishment, and functionally similar native species compete well against it in experiments; however, key competitors have been lost from the ecosystem due to a variety of perturbations prior to the yellow starthistle invasion (Zavaleta & Hulvey, 2004; Hooper & Dukes, 2010; Hierro et al., 2011, 2016; Hulvey & Zavaleta, 2012). Any benefits to yellow starthistle of reduced bacterial diversity could be independent of these interactions with native plant species, but there are clear opportunities for these factors to be related. If a lack of competition allowed yellow starthistle to increase in density, then this could have reduced plant-associated microbial diversity in the environment, as noted above. However, while this scenario could explain lower diversity among bacteria, it appears that density is unlikely to explain reduced negative interactions with the soil community in California. Yellow starthistle experiences some of its strongest negative plant-soil feedbacks across generations in California soils (Andonian et al., 2011), suggesting that the build up of high plant densities is unlikely to explain patterns of enemy release. Alternatively, the historical loss of native species diversity in California (D’Antonio et al., 2007) could have resulted in the loss of associated microbial diversity, generating particularly strong opportunities for invasion into a system with both reduced competition and reduced pathogen diversity. Microbial surveys of remnant native communities, as well as across densities of yellow starthistle would help to clarify alternative interacting effects of plant and microbial diversity, and it may be particularly enlightening to explore microbial communities preserved on native plant specimens pre-dating the extensive invasion of yellow starthistle into this region.
Conclusions
To our knowledge, our study is the first to find evidence consistent with opportunities for release from microbial enemies during invasion. We find lower overall bacterial diversity in invading plant populations, similar in scale to geographic variation in bacterial diversity that has been observed in other studies. These patterns suggest that yellow starthistle may have benefitted from introduction into disturbed plant communities with relatively low microbial diversity. Microbial interactions appear to be important for plant fitness in this system, but may interact with other factors shaping invasiveness, including disturbance and lack of effective competition from native plant species. In particular, escape from both microbial enemies and plant competitors might have created an opportunity for adaptive allocation of resources away from defensive functions and towards reproduction and the evolution of increased invasiveness in yellow starthistle. Comparative surveys of the microbiome in invading and native populations, as presented here, can reveal important variation in the species interactions that are shaping patterns of invasion.
AUTHOR CONTRIBUTIONS
P.L-I., D.A.B., and K.M.D. designed the study. P.L-I. and J.H. collected the samples with assistance from H.S., S.M.S., and S.R.W. P.L-I conducted the microbial sequencing and bioinformatics. P.L-I, S.R.W., and K.M.D. analyzed the data. P.L-I and K.M.D. wrote the manuscript, which was edited by all authors.
ACKNOWLEDGMENTS
We thank D. Lundberg and S. Lebeis for helpful discussions regarding sampling design; G. Reardon and C. Chandeyson for assistance with field collections; E. Arnold, J. Aspinwall, J. Braasch, E. Carlson, K. Hockett, T. O’Connor, M. Schneider, J. U’Ren, and N. Zimmerman for 16S library preparation assistance and discussion; A. Gerritsen, D. New and staff at iBEST for assistance with sequencing; K. Andonian, J. Hierro, and # reviewers for helpful feedback on the manuscript. We are particularly indebted to C.E. Morris for hosting the European sample processing at INRA Station de Pathologie Végétale, Montfavet, France. Data collection and analyses performed by the IBEST Genomics Resources Core at the University of Idaho were supported in part by NIH COBRE grant P30GM103324. This study was supported by USDA grant 2015-67013-23000 to K.M.D., D.A.B., and S.M.S.