Biological invasions alter environmental microbiomes: A meta-analysis.

Biological invasions impact both agricultural and natural systems. The damage can be quantified in terms of both economic loss and reduction of biodiversity. Although the literature is quite rich about the impact of invasive species on plant and animal communities, their impact on environmental microbiomes is underexplored. Here, we re-analyze publicly available data using a common framework to create a global synthesis of the effects of biological invasions on environmental microbial communities. Our findings suggest that non-native species are responsible for the loss of microbial diversity and shifts in the structure of microbial populations. Therefore, the impact of biological invasions on native ecosystems might be more pervasive than previously thought, influencing both macro- and micro-biomes. We also identified gaps in the literature which encourage research on a wider variety of environments and invaders, and the influence of invaders across seasons and geographical ranges.

Introduction

Biological invasions have severe impacts on biodiversity, community composition and ecosystem functions [1-5]. Invasive plants can alter many important ecosystem functions including the nitrogen cycle [6], carbon cycle, and decomposition. For example, invasion by the plant Amur Honeysuckle altered the decomposition rate in the invaded environment likely through changes in litter quality [7]. Exotic snails have been found to alter carbon and nitrogen fluxes in freshwater systems through their consumption/excretion activity [8]. These functions are provided by environmental microbiomes. Yet, despite the implications for ecosystem functioning, we are still learning the consequences of biological invasions on environmental microbiomes.

Previous studies have shown biological invasions can impact the diversity and taxonomical structure of environmental microbiomes. For example, we often see a shift in soil microbiota following invasion by non-native plant species [9-19]. Removal of feral pigs increased the diversity of soil bacterial communities and shifted their structure [20], and invasive crustaceans [21], mussels [22] and jellyfish [23] produce changes in the structure of water microbiomes. However, shifts in environmental microbiome as consequence of biological invasions do not always occur. For example, invasion by the plants Robinia pseudoacacia [24], Eucalyptus sp. [25], and Vincetoxicum rossicum [26] did not alter the structure of soil microbial communities. Also, some microcosms exposed to the simultaneous invasion of multiple plant species [27-29] did not alter soil microbiomes. Similarly, soil microbiome structure in microcosms did not change with the addition of the invasive earthworm Aporrectodea trapezoides [30]. Several of these studies used techniques (e.g. DGGE, PLFA, t-RFLP) that limit fine scale investigations of biological invasions on environmental microbiome diversity and taxonomical composition. Among the studies using high-throughput amplicon-sequencing techniques, most did not find changes in microbiome diversity [13, 15, 18, 19, 21, 25, 27, 28, 30], few reported a decrease of microbial diversity in response to invasion [16, 20, 29], and fewer still reported an increase [11, 14]. Thus, there is little consensus on the effects of biological invasions on the diversity and taxonomical structure of the environmental microbiomes, both tied to the stability and function of microbial communities [31, 32].

Our ability to draw broad conclusions from published studies is limited, because individual studies have occurred within a limited geographical range or with a limited group of species. Meta-analyses of published biological means have long enabled more robust conclusions than individual studies [33-36]. However, the meta-analytic approach has less frequently been applied to amplicon-sequencing data that represent environmental microbiome community composition. The majority of meta-analytic metabarcoding studies have occurred in the medical sciences [37-45]. This approach can be successfully used to address ecological questions. For example, meta-analytic metabarcoding studies have found common patterns in the structure of indoor microbiomes [46] and freshwater eukaryotes [47]. Shade et al. [48] also used a meta-analysis of metabarcoding datasets from different environments highlighting a time-dependent structure of microbiomes. A meta-analytic approach has also been used to test the effects of stressors (e.g. water availability, temperature, heavy metals) on environmental microbiomes [49]. Thus meta-analyses on microbiome data have a striking potential to address global-scale questions, generate new hypotheses and model common patterns [50], because they provide across study comparisons [39, 51, 52].

Here, we aim to test whether the effect of biological invasion on environmental microbiomes can be generalized or is idiosyncratic. To do so, we collected publicly available data and re-analyzed this data under a common framework. We tested the effect of invasive species on the diversity and structure of environmental microbiomes, with the hypothesis that the presence of invasive species will decrease microbial diversity and alter the composition of the environmental microbiome. We then investigated whether certain taxonomical groups are more responsive to biological invasions.

Methods

Data collection

We searched for metabarcoding studies that evaluated the effect of biological invasions on environmental microbiomes, and compared invaded and non-invaded habitats. Our literature search for this study was conducted using Web of Science Core Collection (accessed on March 6th, 2020) using the keywords “Invasive speci*” and “microbio*” published between 2010–2020, and found 1,471 studies. Two additional studies were added by searching the same keywords on Google Scholar (S1 Fig). Records were manually filtered based on the study design appropriate for our research question. This step yielded 22 studies, and we further filtered these studies based on data availability in public repositories. When data was not available, we attempted to contact the corresponding author. Finally we selected only studies that used the 16S rRNA marker gene, primer pair 515F/806R [53] or 341F/785R [54], and Illumina MiSeq sequencing platform. After discarding studies that failed quality checks (see below), we were able to include a total of five studies (Table 1), summing up to a total of 356 samples. The study by Gibbons et al. [28] tested the impact of five invasive plant species (Agropyron cristatum, Bromus tectorum, Sisymbrium altissimum, Erodium cicutarium and Poa bulbosa) on soil microbiome using microcosms, comparing monocultures of each one of them towards a mixture of eight native plant species. A similar question was tested in Rodrigues et al. [19] in field condition. They identified three locations invaded by three different exotic plant species (Microstegium vimineum, Rhamnus davurica and Ailanthus altissima) and, within each location, they sampled soil form an invaded area and a non-invaded area for comparisons. Similarly, Collins et al. [11] compared the soil microbial community of field sites invaded by Artemisia rothrockii to non-invaded sites. The study by Wehr et al. [20] focused on the effects of feral pig (Sus scrofa) invasion on soil microbiome comparing invaded areas to those where pigs were removed over a ~25 year chronosequence. Finally, the only study performed in an aquatic environment [22] compared water samples collected in lake areas invaded by the exotic mussel Dreissena bugensis to non-invaded sites. Three studies focused on invasive plants, and the remaining studies focused on a mammal and a mussel (Table 1).

Table 1

Summary of studies included in the meta-analysis.

Study ID	Invasive organism	Species	Invaded environment	Reference
MPG13011	Plant	Agropyron cristatum, Bromus tectorum, Sisymbrium altissimum, Erodium cicutarium and Poa bulbosa	Soil	Gibbons et al. [28]
MPG87547	Mammal	Sus scrofa	Soil	Wehr et al. [20]
PRJNA296487	Plant	Microstegium vimineum, Rhamnus davurica and Ailanthus altissima	Soil	Rodrigues et al. [19]
PRJNA320310	Plant	Artemisia rothrockii	Soil	Collins et al. [11]
PRJNA385848	Mussel	Dreissena bugensis	Water	Denef et al. [22]

We took the following steps to alleviate some of the potential sources of bias due to studies performed in different labs, using different protocols and sequenced on different instruments. First, all studies included were performed using the Illumina MiSeq platform, in order to reduce the potential bias that might be generated by directly comparing data obtained from different platforms. Second, all studies targeted the same region of the 16S rRNA, as several primer pairs targeting different regions are currently published and widely used. Three out of five papers we considered in our analysis used the 515F/806R primer pair [53], while two used the 341F/785R [54]. Because these primer pairs overlap in the V4 region of 16S rRNA we feel confident that the chance of including spurious OTUs in our analysis is quite negligible. To account for study-specific variances due to small differences in sampling procedures and lab protocols, we also included the study itself, the environment where the study was performed (i.e., soil or water) and the identity of the invasive species as stratification variables in the PERMANOVA and as random factors in our linear model. This allowed us to ensure that our results are not biased by study-specific features.

Once the papers were selected, we assigned each a “Study ID” and collected meta-data from each sample in each study (invasive species, type of organism, invaded environment). We then downloaded data from repositories using SRA Toolkit 2.10.4 for data on the SRA databases, or by manually downloading files from the MG-RAST database.

Data processing and analysis

Paired-end reads were merged using FLASH 1.2.11 [55] and data were processed using QIIME 1.9.1 [56]. Quality-filtering of reads was performed using default parameters, binning OTUs and discarding chimeric sequences identified with VSEARCH 2.14.2 [57]. Taxonomy for representative sequences was determined by querying against the SILVA database v132 [58] using the BLAST method. A phylogeny was obtained by aligning representative sequences using MAFFT v7.464 [59] and reconstructing a phylogenetic tree using FastTree [60].

Data analysis was performed using R statistical software 3.5 [61] with the packages phyloseq [62] and vegan [63]. Read counts were normalized using DESeq2 v1.22.2 [64] prior to data analysis. Singletons and sequences classified as chloroplast were excluded, as well as samples which had less than 5000 sequence counts. Shannon diversity was fit to a linear mixed-effects model specifying sample type (invaded or control), organism (plant, mammal, mussel), and their interactions as fixed factors. We included studyID or both studyID and environment (soil or water) as random factors, and both models reported similar results (S2 Table). We focused on the one with only studyID as random effect due to the lower AIC value. Models were fit using the lmer() function under the lme4 package [65] and the package emmeans was used to infer pairwise contrasts (corrected using False Discovery Rate, FDR). Furthermore, we explored the effects of sample type and organism on the structure of the microbial communities using a multivariate approach. Distances between pairs of samples, in terms of community composition, were calculated using a Unifrac matrix, and then visualized using an RDA procedure. Differences between sample groups were inferred through PERMANOVA multivariate analysis (999 permutations). We ran two different PERMANOVA models: in one we stratified permutations at level of studyID and identity of invasive species, and in the other we stratified permutations at level of studyID, environment and identity of invasive species, obtaining similar results (S3 Table). Pairwise contrasts from PERMANOVA were subjected to FDR correction. Finally, the relative abundance of each bacterial family was fit using the lmer() function to test the effects of sample type (invaded or control) on individual taxa. We ran two different linear mixed-effects models, one including studyID, organism (plant, mammal, mussel) and environment as random factors, and another with studyID and organism as random factors, obtaining similar results (S4 Table).

Results

Our search yielded 5 studies with an appropriate experimental design and available data, for a total of 356 samples. A few samples failed quality checks and we further considered 335 samples for downstream analyses. Sequences clustered into 22831 OTUs, after quality checks, removal of singletons and “chloroplast” reads, with an average of 61776.92 reads per sample. This high OTU count is likely a result of increased richness from analyzing samples across multiple environments (soil and water) and from different geographical regions.

Biological invasions led to a reduction in Shannon diversity (Control = 6.92±0.06, Invaded = 6.73±0.07, χ2 = 3.85, df = 1, P = 0.04). We also found biological invasions altered microbiome community composition in the invaded environment compared to the control (Table 2 and Fig 1). The type of invasive organism (plant, mammal, or mussel) produced a different community structure (pairwise P<0.01, FDR corrected). A deeper analysis of bacterial families (S5 Table) revealed that some taxonomic groups are significantly more abundant in invaded environments (Blastocatellaceae, Chitinophagaceae, Nitrosomonadaceae, Pirellulaceae, Sphingomonadaceae), while others are more abundant in non-invaded samples (Acetobacteraceae, Beijerinckiaceae, Gemmataceae, Micromonosporaceae, Pedosphaeraceae, Solibacteraceae, Solirubrobacteraceae).

Table 2

Results from PERMANOVA analysis testing the effects of sample type (invaded/control), organism group (plant, mammal, mussel) and their interaction on microbial community composition.

The factors studyID (unique for each study) was used as strata to constrain permutations.

Factor	df	R2	F	P
Sample type (Invaded/Control)	1	0.011	6.68	<0.001
Organism group (plant, mammal, mussel)	3	0.411	118.86	<0.001
Sample type × Organism group	3	0.007	2.1	0.01

Fig 1

RDA ordination using a Bray-Curtis distance matrix of samples.

Discussion

Here we show biological invasions decrease the diversity of environmental microbiomes. While several studies have investigated the effects of species invasions on environmental microbiomes, we still lack a generalized consensus across different environmental microbiomes and systems. Previous studies have found that invasive species increased environmental microbial diversity [11, 14], while others reported a decrease [16, 20, 29]. However, the majority of studies did not analyze the microbial diversity, as they used techniques that did not allow for such analysis, or reported no changes [9, 10, 12, 13, 15, 17–19, 21–28, 30]. Within the studies included in our analysis, invasion by feral pigs decreased soil microbial diversity, while invasion by Artemisia rothrockii increased soil microbial diversity. The remaining three studies in our analysis reported no effects of biological invasions on environmental microbiome diversity. Our analysis was constrained in terms of sampled environment (soil) and invasive organism (plants), and an expanded dataset would be beneficial to generalize our results. Microbial diversity is tied to the function of microbiomes, and changes in diversity can reflect changes in function [66-68]. Changes in microbial diversity and function do not always have the same direction [69], which might explain the discrepancy between our results and other studies. While we observed a relatively small reduction of the Shannon diversity index in invaded environments compared to non-invaded environment, this matches with the changes we observed in terms of community composition.

Indeed, our report of changes in community composition was relatively consistent with the published literature and the individual results of the studies we analyzed. Most studies of the influence of biological invasions on environmental microbiomes found that biological invasions alter environmental microbial community composition. However, some previous reports did not report changes [24-30], including the study by Gibbons et al. [28] considered in our analysis. This variation may be due to individual effects of organisms on the environment. For example, invasive plants may alter soil microbiome composition through root exudates [5], and invasive mussels may alter water microbiome composition via bacterial removal through their feeding activity [22]. Thus, reported influences on community composition are more consistent. Alternatively, changes in community composition might be due to the response of some bacterial groups to environmental disturbance. The bacterial families that we found to be differentially abundant between the invaded and control environments have diverse ecological functions ranging from nitrogen fixation and carbohydrate metabolism to antimicrobial properties. Although the differences in relative abundance we found might be relatively small, they can have an important impact on the functions of the environmental microbiome [70]. Many of the families that showed a significant difference in abundance between invaded and control environments include taxa that play important roles at various points during nitrogen and carbon cycling (i.e. Nitrosomonadaceae, Acetobacteraceae, Chitinophagaceae, Micromonosporaceae, Gemmataceae, Beijerinckiaceae, Pirellulaceae) [71-81]. However, nitrogen-fixing and carbohydrate-degrading bacteria did not have a unified response to invaded environments as some increased and others decreased in abundance in invaded environments. Many nitrogen-fixing bacteria have been shown to respond to environmental disturbance, such as Acidobacteria abundances during forest to pasture conversions or Pirellulaceae’s response to the presence of microplastics [82, 83]. Thus, changes in environmental microbiome community composition appear to be linked to changes in ecosystem functions, although this pattern is not yet predictable across functions and taxa.

Few previous studies on biological invasions have reported details on the differential abundance of taxa, and among these we found limited general consensus. For example, some studies report a decrease in abundance of bacteria associated with nitrogen cycling (e.g. Nitrosphaeria, Nitrospira, Nitrosomonadales) [13, 14, 19], while others report an increase of Nitrosomonadaceae following invasion [30]. In our study some groups associated with the nitrogen cycle were positively associated with biological invasions (i.e. Nitrosomonadaceae, Pirellulaceae, Chitinophagaceae) while others were negatively associated (Beijerinckiaceae, Micromonosporaceae). Unfortunately, amplicon-based sequencing has a limited power to infer changes in the functions of microbiomes. Future metagenomic and metatranscriptomic studies are needed to investigate whether biological invasions alter gene content or gene expression of environmental microbiomes, and whether this reflects changes in biogeochemical cycling.

Meta-analyses are also useful to highlight gaps in the literature, and here we highlight some aspects that warrant further investigation. First, we identified a large gap in the availability of sequencing data from multiple types of environments and types of invasive species. For our analysis almost all available data came from two environments: four sets of data came from soil and one came from freshwater. Greater effort is needed for sample collection from invasions in both freshwater and marine environments. Without sufficient diversity of sample environments, it is impossible to tell whether microbial shifts following an invasion are unique to an invaded environment. Second, in our analysis the majority of data came from one type of invasive species: plants. Noticeably absent from our dataset were invasions by insects, fish, and amphibians. Sequencing data is needed from a larger number of invasive species to allow us to broadly assess shifts in microbial community structure. A third gap we identified was the lack of spatial and temporal resolution. Almost all of the initially identified 22 studies we assessed were also restricted to one season of sampling and were conducted in the Northern Hemisphere. Thus, it is infeasible with existing datasets to validate the influence of latitude or explore how seasonality and biological invasions interact to modulate microbial communities. Thus, there are a number of opportunities for future research on how biological invasions alter environmental microbial communities.

Here we analyzed 16S amplicon sequencing data from five studies and show that biological invasions influence both the diversity and the structure of environmental microbiomes. Understanding the impact of biological invasions on environmental microbiomes is of high priority to preserve ecosystem functions [84]. We identified a number of gaps in our knowledge, including the need to assess a wider range of environments, invasive species, temporal variation, and latitudinal variation. We also demonstrate the power of re-analysis of publicly available datasets using a common pipeline which benefited from open-data initiatives.

PRISMA workflow.

From: Moher D, Liberati A, Tetzlaff J, Altman DG, The PRISMA Group (2009). Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement. PLoS Med 6(7): e1000097. doi:10.1371/journal.pmed1000097.

(PDF)

Click here for additional data file.

PRIMA checklist.

From: Moher D, Liberati A, Tetzlaff J, Altman DG, The PRISMA Group (2009). Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement. PLoS Med 6(7): e1000097. doi:10.1371/journal.pmed1000097.

(PDF)

Click here for additional data file.

Comparison of two different linear mixed-effects models testing the effect of biological sample type (invaded or control), organism (plant, mammal, mussel), and their interactions, on Shannon diversity index of the environmental microbiome.

In Model 1 we included studyID as random factor, while in Model 2 we included both studyID and environment (soil or water) as random factors.

(PDF)

Click here for additional data file.

Comparison of two different PERMANOVA models testing the effect of biological sample type (invaded or control), organism (plant, mammal, mussel), and their interactions, on the structure of the environmental microbiome.

In Model 1 we included studyID and speciesID as stratification factor, while in Model 2 we included studyID, speciesID and environment (soil or water) as stratification factors.

(PDF)

Click here for additional data file.

Results from two different the linear mixed-effects model testing the abundance of each bacterial family against sample type (invaded or control), organism (plant, mammal, mussel), and their interactions.

In Model 1 we included studyID as random factor, while in Model 2 we included both studyID and environment (soil or water) as random factors.

(PDF)

Click here for additional data file.

Comparison of the relative proportion of each bacterial family between control and invaded environments.

Differences are assessed using a linear mixed-effects model testing the normalized proportions of each bacterial family against sample type (Model 1 in S4 Table).

(PDF)

Click here for additional data file.

3 Aug 2020

PONE-D-20-18642

Biological invasions alter environmental microbiomes: a meta-analysis

PLOS ONE

Dear Dr. Malacrinò,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Dear Author, as you can see, the reviewers raise many doubts on the soundness of your data and conclusions drawn from them. I expect you to submit a complete and thorough rebuttal to their comments and a revised version of your manuscript that will undergo an additional round of reviews.

Please submit your revised manuscript by Sep 17 2020 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

We look forward to receiving your revised manuscript.

Kind regards,

Raffaella Casotti

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Please include a caption for figure 1.

3. Please include captions for your Supporting Information files at the end of your manuscript, and update any in-text citations to match accordingly. Please see our Supporting Information guidelines for more information: http://journals.plos.org/plosone/s/supporting-information.

Additional Editor Comments (if provided):

Dear authora, as you will see, the reviewers raise many concern about the soundness of the data presented and on the suitability for publication. I suggest to submit a thorough rebuttal to their criticisms and to submit a revised version of your manuscript, which will undergo an additional round of revision.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

Reviewer #2: No

Reviewer #3: Partly

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: No

Reviewer #3: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: I believe this mansucript addresses an important question, and that the authors have done what they can with very limited numbers of studies available for analysis. I am left wondering if 5 published studies is an adequate sample number, particulalry considering that 2 are from very different systems to the other three, to draw broad conculions. The work seems very preliminary based on such a small sample size.

I have attached a PDF file with some minor comments.

Reviewer #2: This paper is not a meta-analysis. It is an analysis of combined data from five studies, but meta-analysis procedures were not used. Unfortunately, I cannot comment on the microbial data methods and analysis in this paper because it is beyond my area of expertise, but for reasons outlined below, I do feel the paper would be better framed as a review + additional analysis, rather than a meta-analysis (which it is not). And the conclusions need to be tempered and put in a greater context. The paper makes broad statements such as, “We show biological invasions decrease the diversity of environmental microbiomes,” which I do not feel are at all justified by the analyses. First, the study includes data from only five papers, three on plants (affecting soil), one on a mammal (affecting soil), and one on an aquatic snail (affecting water). This data set is far too small and disparate to use for sweeping generalizations. Second, it is not at all clear that the results support the statements. For example, a quick look at Figures 2A and 2C would indicate that there are absolutely no differences in either diversity or abundance between control and invaded sites, yet statistical analyses report differences. This makes me question the statistical models used. For example, the mixed model assessing diversity using environment as a random effect is inappropriate because you need more levels (at least 5) for an effect to be random. The unbalanced groups (3 plant, 1 mammal and 1 aquatic snail) could also cause model problems. At the very least, the model diagnostics need to be discussed to show that the model was appropriate. Finally, I think more needs to be known about individual studies to make any firm conclusions. For example, in study [28] natives were planted in polycultures of up to 7 species and non-natives were planted in monocultures. Does the difference in microbial diversity in such a study represent a realistic shift caused by an invader, or is it a reflection of the experimental design? I think more generalizations could be made if each of the five studies were discussed in some detail and the results put in first a specific context, and then if patters support it, into a more general context.

Reviewer #3: Dear editor,

This study is a meta-analysis of previously published metabarcoding data that compares environmental microbiomes in invaded and uninvaded sites in order to investigate the impacts of invasive species on microbiome diversity and community composition. Individual studies have had inconsistent results, and this is the first study that attempts to quantitatively draw broad patterns in this way. After filtering to control for quality and methodology, the authors were left with only five studies in their analysis. Their data showed an overall significant effect of invasive species on microbiomes: Shannon diversity decreased, and community composition shifted in taxa-specific ways at the family level.

Although this study’s methodology was well thought-out and executed rigorously, I am not sure that all of the authors’ claims are robustly supported. In the Discussion, the authors explain the limitations and caveats of this study very well. For example, the five studies are mostly about plants, and mostly concern soil microbiomes. Therefore, although I am not refuting the results themselves, I think that some of the conclusions that the authors draw are overly broad. In some places, their claims are well-qualified (line 23: “Our findings suggest that non-native species are responsible…”; line 232: “Without sufficient diversity of sample environments, it is impossible to tell…”), but others seem to overreach. The first line of the Discussion (line 117: “We show biological invasions decrease the diversity of environmental microbiomes.”) feels too broad to me. Although the results did indeed show a decrease in diversity, I think this sentence would benefit from a clause acknowledging that the environmental sampling is limited and biased. The rest of the Discussion does exactly this—pointing out all the ways this sample is biased (soils, invasive plants, northern hemisphere, summer)—so it seems odd to me that this claim is so broad. Alongside these caveats, the Discussion also excellently highlights the overall knowledge gap revealed by this meta-analysis.

I also thought that the discussion of taxa-level shifts was very interesting. Connecting microbial family shifts to possible shifts in community function is intriguing, although I am very glad that the authors acknowledge that amplicon sequencing cannot definitively show changes in function. In this section, as in the rest of the Discussion, the authors combine the results of their meta-analysis with results of previously published studies smoothly and effectively.

Overall, this is an interesting manuscript with a compelling question. However, given the dearth of compatible studies in the final meta-analysis, I think that this manuscript raises many more questions than it answers—this manuscript’s strength is in its ‘call to arms’ for more research rather than in the results themselves.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

Reviewer #3: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

Submitted filename: PONE-D-20-18642_reviewer.pdf

Click here for additional data file.

24 Aug 2020

Response to reviewers’ comments (MS PONE-D-20-18642– REV1)

Answers to reviewers are reported after each comment

Reviewer #1

Comment #1 - I believe this manuscript addresses an important question, and that the authors have done what they can with very limited numbers of studies available for analysis. I am left wondering if 5 published studies is an adequate sample number, particularly considering that 2 are from very different systems to the other three, to draw broad conclusions. The work seems very preliminary based on such a small sample size.

Answer: Thank you! Yes, we agree that our analysis would benefit from a higher number of studies. In principle, a meta-analysis is defined as “the statistical combination of results from two or more separate studies” . Also, looking at the published literature (references 37-49 in our study), previous meta-analyses focused on a wide range of studies spanning from 2 (Bisanz et al.) to 30 (Duvallet et al.), with several of them in the range between 2 and 9 (Bisanz et al, Walters et al, Krych et al., Jiao et al., Wirbel et al., Rocca et al.). In addition, we took several steps to guarantee that data analysis would consider the heterogeneity between studies into account. Therefore, we believe that the number of studies and the approach we took guarantee a low bias in our results. We acknowledge these limitations in our discussions (L199-201, clean copy).

Comment #2 - I have attached a PDF file with some minor comments.

Answer: We integrated your suggestions in our manuscript, including a brief description of violin plots in the caption of Fig. 1A (L184-185, clean copy).

Reviewer #2

Comment #1 - This paper is not a meta-analysis. It is an analysis of combined data from five studies, but meta-analysis procedures were not used. Unfortunately, I cannot comment on the microbial data methods and analysis in this paper because it is beyond my area of expertise, but for reasons outlined below, I do feel the paper would be better framed as a review + additional analysis, rather than a meta-analysis (which it is not).

Answer: We thank the reviewer for this comment; however, we respectfully disagree with their point of view. A meta-analysis is defined as “the statistical combination of results from two or more separate studies”1. The NIH NCI defines the meta-analysis as “A process that analyzes data from different studies done about the same subject …” . Most of the published literature that used our approach define their analysis as “meta-analysis” (refs 37-49 in our paper). We acknowledge that using data for single points , rather than aggregated data, is quite unusual. Literature is rich of examples where a meta-analytical approach has been used on individual “samples” . Therefore, we believe that meta-analysis is the most appropriate term to use in our case.

Comment #2 - And the conclusions need to be tempered and put in a greater context.

Answer: Thanks for this suggestion. We added a sentence to the last paragraph to put our results in a greater context (L259-260, clean copy).

Comment #3 - The paper makes broad statements such as, “We show biological invasions decrease the diversity of environmental microbiomes,” which I do not feel are at all justified by the analyses.

Answer: We included a statement (L199-201, clean copy) where we acknowledge the limits of our analysis.

Comment #4 - First, the study includes data from only five papers, three on plants (affecting soil), one on a mammal (affecting soil), and one on an aquatic snail (affecting water). This data set is far too small and disparate to use for sweeping generalizations.

Answer: As we mentioned in our answer to comment #1, a meta-analysis can be done with even 2 studies. Furthermore, our analysis considers a total of 335 samples and we took a series of steps (L111-124, clean copy) to alleviate potential sources of bias. We acknowledge the limitations of our study throughout the discussions.

Comment #5 - Second, it is not at all clear that the results support the statements. For example, a quick look at Figures 2A and 2C would indicate that there are absolutely no differences in either diversity or abundance between control and invaded sites, yet statistical analyses report differences. This makes me question the statistical models used. For example, the mixed model assessing diversity using environment as a random effect is inappropriate because you need more levels (at least 5) for an effect to be random. The unbalanced groups (3 plant, 1 mammal and 1 aquatic snail) could also cause model problems. At the very least, the model diagnostics need to be discussed to show that the model was appropriate.

Answer: Thanks for this suggestion. We ran all our models including or excluding “environment” as random effect, and we reported the comparison in the supplementary material. Overall, we found no differences between the two modelling strategies.

Comment #6 - Finally, I think more needs to be known about individual studies to make any firm conclusions. For example, in study [28] natives were planted in polycultures of up to 7 species and non-natives were planted in monocultures. Does the difference in microbial diversity in such a study represent a realistic shift caused by an invader, or is it a reflection of the experimental design? I think more generalizations could be made if each of the five studies were discussed in some detail and the results put in first a specific context, and then if patters support it, into a more general context.

Answer: Thanks for this suggestion. We agree with this comment and we provided a description of the individual studies (L94-106, clean copy). Discussion of our results in the context of individual studies was already provided (L196-201 and L205-215, clean copy).

Reviewer #3

Comment #1 - In the Discussion, the authors explain the limitations and caveats of this study very well. For example, the five studies are mostly about plants, and mostly concern soil microbiomes. Therefore, although I am not refuting the results themselves, I think that some of the conclusions that the authors draw are overly broad. In some places, their claims are well-qualified (line 23: “Our findings suggest that non-native species are responsible…”; line 232: “Without sufficient diversity of sample environments, it is impossible to tell…”), but others seem to overreach. The first line of the Discussion (line 117: “We show biological invasions decrease the diversity of environmental microbiomes.”) feels too broad to me. Although the results did indeed show a decrease in diversity, I think this sentence would benefit from a clause acknowledging that the environmental sampling is limited and biased. The rest of the Discussion does exactly this—pointing out all the ways this sample is biased (soils, invasive plants, northern hemisphere, summer)—so it seems odd to me that this claim is so broad. Alongside these caveats, the Discussion also excellently highlights the overall knowledge gap revealed by this meta-analysis.

Answer: Thanks for your suggestion. We included it in the manuscript (L199-201, clean copy).

Comment #2 - I also thought that the discussion of taxa-level shifts was very interesting. Connecting microbial family shifts to possible shifts in community function is intriguing, although I am very glad that the authors acknowledge that amplicon sequencing cannot definitively show changes in function. In this section, as in the rest of the Discussion, the authors combine the results of their meta-analysis with results of previously published studies smoothly and effectively.

Answer: Thank you!

Comment #3 - Overall, this is an interesting manuscript with a compelling question. However, given the dearth of compatible studies in the final meta-analysis, I think that this manuscript raises many more questions than it answers—this manuscript’s strength is in its ‘call to arms’ for more research rather than in the results themselves.

Answer: Thank you!

Submitted filename: Response to Reviewers.docx

Click here for additional data file.

21 Sep 2020

PONE-D-20-18642R1

Biological invasions alter environmental microbiomes: a meta-analysis

PLOS ONE

Dear Dr. Malacrinò,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please submit your revised manuscript by Nov 05 2020 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

We look forward to receiving your revised manuscript.

Kind regards,

Raffaella Casotti

Academic Editor

PLOS ONE

Additional Editor Comments (if provided):

Dear authors, the reviewers recognize the value of your work but raise some additional concern about conclusions drawn from statistical tests. I kindly ask you to address the point raised by acknowldging it in the etxt and commenting further the limitations of the analyses and of the limited dataset analyzed. Please, do NOT simply rebutt the comment but acknowledge it and comment it appropriately supporting your statement with solid arguments.

Thanks for submitting you best work to this journal

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: (No Response)

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: (No Response)

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: I Don't Know

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: The authors have addressed the concerns I had with the paper to what appears to be a reasonable extent. They argue that their data set are adequate to address their questions, and provide an explanation why they believe this is the case. I find this hard to judge, and I expect the other reviewers will comment further on the matter. My feeling that the work is quite preliminary stands, but I am happy to be convinced otherwise.

I was concerned with the response to reviewer #2, who commented on the statistical analyses. I do agree that many of the comparisons declared statistically different certainly do not appear to be, from the data shown in the figure. E.g. Fig 1c, it seems very unlikely that Sphingomonadaceae were different in control vs invaded – the medians are essentially identical; yet Gemmatimonadaceae are not different? There are a number of such instances in the figures, and the authors need to explain this clearly. I suspect the asterisks indicating significance are incorrectly annotated, maybe? Their assertion that including as a random effect in their analysis and finding it made no difference to the conclusions, does not go a long way to explaining the apparent lack of differences shown in the figures, despite the pairs being labelled as significantly different. Fig 1a likewise seems to show two very similar data distributions, with the medians essentially the same.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

1 Oct 2020

Reviewer #1

Comment #1 - I was concerned with the response to reviewer #2, who commented on the statistical analyses. I do agree that many of the comparisons declared statistically different certainly do not appear to be, from the data shown in the figure. E.g. Fig 1c, it seems very unlikely that Sphingomonadaceae were different in control vs invaded – the medians are essentially identical; yet Gemmatimonadaceae are not different? There are a number of such instances in the figures, and the authors need to explain this clearly. I suspect the asterisks indicating significance are incorrectly annotated, maybe? Their assertion that including as a random effect in their analysis and finding it made no difference to the conclusions, does not go a long way to explaining the apparent lack of differences shown in the figures, despite the pairs being labelled as significantly different. Fig 1a likewise seems to show two very similar data distributions, with the medians essentially the same.

Answer: We agree with the reviewer that our graphs do not properly show the differences we claim in our text. Thus, we changed our Fig 1: panel A has been moved to the text (L167-168), panel B is now the main figure, and panel C has been moved to S5 table. Furthermore, we integrated our discussion to reflect your comments (L200-203 and L217-218).

Submitted filename: Response to Reviewers.docx

Click here for additional data file.

7 Oct 2020

Biological invasions alter environmental microbiomes: a meta-analysis

PONE-D-20-18642R2

Dear Dr. Malacrinò,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Raffaella Casotti

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Dear authors, I think that your manuscript is now suitable for publication

Reviewers' comments:

14 Oct 2020

PONE-D-20-18642R2

Biological invasions alter environmental microbiomes: a meta-analysis

Dear Dr. Malacrinò:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Raffaella Casotti

Academic Editor

PLOS ONE