Revisiting Propagule Pressure Theory: A Global Meta-Analysis of Seed Endophytes from Alien Plants on Plant Performance and Trait Variability

Revisiting Propagule Pressure Theory: A Global Meta-Analysis of Seed Endophytes from Alien Plants on Plant Performance and Trait Variability | Authorea try { document.documentElement.classList.add('js'); } catch (e) { } var _gaq = _gaq || []; _gaq.push(['_setAccount', 'G-8VDV14Y67G']); _gaq.push(['_trackPageview']); (function() { var ga = document.createElement('script'); ga.type = 'text/javascript'; ga.async = true; ga.src = ('https:' == document.location.protocol ? 'https://ssl' : 'http://www') + '.google-analytics.com/ga.js'; var s = document.getElementsByTagName('script')[0]; s.parentNode.insertBefore(ga, s); })(); Skip to main content Preprints Collections Wiley Open Research IET Open Research Ecological Society of Japan All Collections About About Authorea FAQs Contact Us Quick Search anywhere Search for preprint articles, keywords, etc. Search Search ADVANCED SEARCH SCROLL This is a preprint and has not been peer reviewed. Data may be preliminary. 30 January 2026 V1 Latest version Share on Revisiting Propagule Pressure Theory: A Global Meta-Analysis of Seed Endophytes from Alien Plants on Plant Performance and Trait Variability Authors : Zafar Reshi 0000-0001-9567-7484 [email protected] and Iflah Rafiq Authors Info & Affiliations https://doi.org/10.22541/au.176978530.04074018/v1 172 views 89 downloads Contents Abstract Effects of target species status on seed microbiome–mediated plant performance Information & Authors Metrics & Citations View Options References Figures Tables Media Share Abstract The seed microbiome, encompassing bacteria, fungi, archaea and viruses residing within or on seeds, represents an underexplored but potentially critical determinant of plant establishment and spread. In invasive plants, these co-dispersed microbial communities may function as portable adaptive toolkits that enhance nutrient uptake, pathogen defence, and host performance. This perspective broadens the concept of propagule pressure to include plant–microbe assemblages as functional units of invasion. To evaluate this perspective, we conducted a systematic review and meta-analysis following PRISMA-EcoEvo guidelines, screening Web of Science, Google Scholar, and Citation Chaser. Seventeen studies representing 287 effect sizes met the inclusion criteria. We calculated effect sizes as log response ratios (lnRR) for trait means and log coefficient of variation ratios (lnCVR) for trait variability using multilevel meta-analytic models and meta-regressions. Across studies, alien invasive seed endophyte inoculation significantly increased mean plant trait performance (lnRR = 0.30; 95% CI: 0.13–0.46; P < 0.001), though effects varied widely (prediction interval: –0.61 to 1.20). The trait variability responses were also significant (lnCVR = -0.28; 95% CI: -0.49 – -0.07; P = 0.01). Stronger positive effects emerged under field and laboratory conditions compared to greenhouse or growth chamber experiments, with both bacterial and fungal inocula enhancing means, and bacterial inoculation additionally reducing variability. These findings highlight seed-associated microbiomes as overlooked yet influential facilitators of invasion success. By enhancing various traits, these microbial assemblages can directly shape establishment success and competitive outcomes. Conceptually, this shift in perspective moves invasion biology from a plant-centric to a holobiont-centric perspective, where the seed and its microbiota together represent the operative unit of dispersal and establishment. Practically, it calls for expanding propagule pressure frameworks to include the diversity and functional roles of microbial symbionts, and for integrating seed microbiomes into risk assessment, biosecurity and management strategies 0. 1em 0.0 1em Introduction The global proliferation of invasive plant species is a major driver of ecological disruption, with profound consequences for biodiversity, ecosystem function, and economic stability (Essl et al. 2020; Pyšek et al. 2020; Roy et al. 2023).Classic invasion theory attributes the success of these species to mechanisms such as high propagule pressure, competitive dominance, and escape from natural enemies (Lockwood et al. 2005, Catford et al. 2009). These frameworks have been invaluable for understanding the mechanisms and processes that drive the establishment, spread, and impact of invasive species. However, these have largely emphasized plant traits and abiotic conditions in biological invasions, underestimating the role of biotic partners, particularly the microorganisms that plants carry with them across biogeographic boundaries (Van Der Putten et al. 2007, Rout and Callaway 2012, War et al. 2023, Quaglini et al. 2025, Kholostiakov et al. 2025). Among these partners, the seed microbiome , the consortium of bacteria, fungi, archaea, and viruses residing within or on seeds, has emerged as a potential but underexplored determinant of plant establishment and spread (Nelson 2017, Shade et al. 2017, Simonin et al. 2022). Seeds are not only dispersal units for plants but also vehicles for both vertically and horizontally transmitted microbes. Vertical transmission occurs when microorganisms are passed directly from the maternal plant to the seed during seed development, whereas horizontal transmission involves the acquisition of microbes from the surrounding environment, such as soil, air, or other organisms, after seed formation. These seed-associated microbes can influence germination, seedling vigour, stress resilience, and competitive interactions (Truyens et al. 2015, Berg and Raaijmakers 2018). In the context of invasion, these co-dispersed microbial assemblages may serve as a portable “adaptive toolkit,” enhancing nutrient acquisition (e.g., phosphate solubilization, nitrogen fixation), suppressing pathogens, and modulating host hormone signalling (Mendes et al. 2011, Compant et al. 2019). This perspective invites a re-examination of the well-established concept of propagule pressure , traditionally defined as the quantity, quality, and frequency of individuals introduced to a novel range (Simberloff 2009). While propagule pressure theory has emphasized the numerical aspect—more propagules mean higher establishment probability—it has rarely accounted for the biological quality of the propagules themselves. We propose that this framework be expanded to explicitly incorporate microbial propagule pressure —the diversity, abundance, and functional potential of microbial taxa co-introduced with plant hosts (Lekberg et al. 2018, Reinhart et al. 2021). This addition could bridge invasion biology with microbial ecology, shifting the focus from purely plant-centered processes to plant–microbe assemblages as the true units of invasion. Most research has concentrated on root- or soil-associated microbiomes (Reinhart and Callaway 2006a, Dawson and Schrama 2016), leaving seed-associated communities comparatively understudied. Even fewer studies have explicitly compared invasive species with model plants such as Arabidopsis thaliana , which is phylogenetically tractable due to its well-resolved evolutionary history, extensive genomic resources, and standardized reference datasets, and serves as an ecologically neutral benchmark in invasion studies because it is neither invasive nor dominant in most natural systems. This comparison allows invasion-specific microbial effects to be distinguished from general plant–microbiome associations (Bulgarelli et al. 2013, Escudero-Martinez and Bulgarelli 2019). Early evidence suggests that invasive species such as Centaurea stoebe and Alliaria petiolata harbour seed microbiomes enriched in taxa like Pseudomonas and Bacillus , groups associated with pathogen suppression and plant growth promotion (Reinhart et al. 2010, Rout and Callaway 2012). Yet, effects vary widely across systems: while seed endophytes have been shown to enhance germination and growth in Acacia longifolia (Condessa et al. 2024), they appear neutral in Bromus tectorum (Beckstead et al. 2010) and detrimental to growth in certain cases (Hu et al. 2025). This heterogeneity points to the need for a quantitative synthesis. Here, we present the first global meta-analysis of seed microbiome effects in invasive and model plants, drawing on studies that evaluated seed-associated endophytes from alien invasive species using two complementary experimental approaches: inoculation onto the same invasive host species from which the endophytes were isolated, and inoculation onto other plant species, including non-hosts and established model species such as Arabidopsis thaliana . These designs allowed explicit assessment of both host-specific and cross-species effects of invasive plant seed endophytes on plant performance. Using this framework, we tested whether: (1) seed microbiomes consistently enhance plant performance; (2) invasive plant species gain greater benefits from their seed microbiomes than non-alien or model species; (3) different microbial inoculum types (bacteria, fungi, or microbial consortia) differ in their effects on mean performance and trait variability; and (4) experimental conditions (field, greenhouse, growth chamber, and laboratory) systematically modify these effects. By integrating microbial ecology into invasion theory, our work explores how seed-associated microbiomes contribute to plant invasion in a context-dependent manner, ranging from passive co-dispersal to conditional influences on establishment and performance. We further assess whether incorporating microbial propagule pressure alongside plant traits can improve conceptual predictions of invasion risk and inform management strategies. 0. 1em 0.0 1em Material and methods We conducted a systematic, rigorous and comprehensive multi-database literature search to identify empirical studies quantifying the effects of seed-associated microbial communities on the performance of invasive alien plant species following the PRISMA-EcoEvo guidelines (O’Dea et al. 2021), an extension of the PRISMA framework (Page et al. 2021) designed for ecological and evolutionary syntheses (Fig. 1). Inclusion and exclusion criteria Studies were included in the meta-analysis if they met all of the following criteria: 1. Reported empirical, quantitative experiments involving seed-associated microbes (e.g. bacteria, fungi, or consortia) isolated from widespread alien or invasive plant species. 2. Experimentally tested the effects of seed-associated microbes on plant performance traits (e.g. germination, growth, biomass, root or shoot traits). 3. Included a non-inoculated control or equivalent comparator. 4. Reported sufficient statistical information to calculate effect sizes (means, standard deviations or errors, and sample sizes), or data that could be reliably extracted. Studies were excluded if they: 1. Did not explicitly involve seed-associated microbiomes. 2. Focused exclusively on native species. 3. Were observational, conceptual, or review articles. 4. Lacked extractable quantitative data. 5. Duplicated data already included from another source. Fig. 1. PRISMA-EcoEvo flow diagram outlining the systematic literature screening process for studies assessing the influence of seed-associated microbial inoculation on invasive plant performance (www.prisma-statement.org). A structured literature search was first performed in Web of Science (Clarivate Analytics), which enables comprehensive Boolean querying across indexed peer-reviewed journals. The following search string was applied to titles, abstracts, and author keywords: (“seed microbiome” OR “seed endophyte” OR “seed-associated microbiome” OR “seed bacteria” OR “seed fungi” OR “seed virome”) AND (“plant invasion” OR “invasive plants” OR “biological invasion” OR “invasive species” OR “non-native plants”) This search returned 653 records. After title and abstract screening, six duplicate records were removed, leaving 647 articles for full-text evaluation. Of these, 640 articles were excluded due to insufficient statistical information (n = 231), focus on non-invasive species (n = 107), narrative or conceptual reviews (n = 114), lack of relevance to seed-associated microbiomes (n = 169), or incomplete or inaccessible data (n = 19). This process yielded 12 eligible articles, from which six duplicates were subsequently removed, resulting in six unique studies retained from Web of Science. To complement the database-indexed search and ensure broader coverage, particularly of interdisciplinary and recently published studies, we conducted an additional search in Google Scholar. Given the platform’s limited support for complex Boolean logic and field-specific querying, a simplified search string was used: (“seed microbiome” OR “seed endophyte” OR “seed bacteria” OR “seed fungi”) AND (“invasive plant” OR “plant invasion” OR “invasive species”) Searches conducted using Google Scholar yielded 276 records, of which 264 articles remained after duplicate removal and were assessed at the full-text stage. Following exclusion based on predefined eligibility criteria, nine unique studies were retained. Citation Chaser identified an additional 232 records; after duplicate removal and screening, two unique studies met the inclusion criteria (Fig. 1). Data compilation and preparation Following the systematic screening process outlined in the PRISMA-EcoEvo framework, this multi-database, multi-step screening ensured a rigorous and comprehensive collection of empirical studies on the influence of seed microbiomes on invasive plant performance, resulting in seventeen (17) eligible studies retained for meta-analysis after duplicate removal. The compiled dataset is provided in Appendix 1, with extracted data presented in Sheet 1 and the full citations of the 17 included studies listed in Sheet 2. While most retained studies focused on plant species explicitly described as invasive or alien by the original authors, two species that are better characterized as introduced or widespread rather than strictly invasive (e.g. Crotalaria pumila , Poa alsodes ) were retained because seed-associated microbes were used to test mechanistic and cross-species effects of their seed endophytes on plant performance. In this meta-analysis, we adopted the invasion or alien status as defined in the original studies, where “alien” generally refers to plant species occurring outside their native range due to human-mediated introduction, and “invasive” denotes alien species reported by the original authors to have established, spread, or exert ecological effects in the recipient region (Richardson and Pyšek 2012), rather than reclassifying species independently. This approach is widely used in quantitative syntheses, particularly when drawing on experimental studies conducted under controlled or semi-controlled conditions and allows consistency with the ecological and biogeographic context intended by the original authors. From the selected studies, we compiled a comprehensive dataset on the effects of seed endophyte inoculation on various growth and performance parameters of plants. The dataset included mean values, standard deviations, and sample sizes for both inoculated (treated) and non-inoculated (control) groups. To enable moderator analyses, categorical variables such as experimental condition, inoculum type, target plant status (native or invasive), and the specific performance traits assessed (e.g., seed germination, root and shoot length) were also extracted. Data were obtained from text, tables, and figures, with values digitized from graphs using WebPlotDigitizer (https://automeris.io/WebPlotDigitizer/) when necessary. Effect size calculation and mean–variance relationship To select appropriate effect size metrics, we first examined the mean–variance relationship in the dataset (Ahmad et al. 2024). In ecological and biological data, non-constant variance is common, such that variability increases with the mean. Under these conditions, effect sizes based on raw mean differences can be inappropriate, whereas log-transformed ratios provide scale-independent and biologically interpretable measures. We, therefore, assessed the relationship between the natural logarithm of the mean and the natural logarithm of the standard deviation for both seed endophyte–inoculated and control groups. These relationships were visualized by plotting ln(mean) against ln(SD) and quantified using Pearson correlation coefficients. Strong positive correlations were observed in both treatments (inoculated: r = 0.886; control: r = 0.894), indicating pronounced mean–variance scaling (Fig. S1). This pattern justified the use of ratio-based, log-transformed effect sizes to quantify both mean responses and variability. Accordingly, we used the log response ratio (lnRR) to estimate proportional changes in mean plant performance and the log coefficient of variation ratio (lnCVR) to quantify treatment effects on variability (Hedges et al. 1999). Effect sizes and their sampling variances were calculated using the escalc function in the metafor package (Viechtbauer 2010). The lnRR was computed as the natural logarithm of the ratio of the treated to control means, while lnCVR captured proportional differences in the coefficient of variation between treatment and control groups. No observations exhibited a zero standard deviation in the control group; therefore, no effect sizes were excluded on this basis. Cohorts and Variance–Covariance Structure Many studies contributed multiple, non-independent effect sizes derived from the same study, often sharing a common control group. To account for non-independence among effect sizes, we grouped observations into cohorts defined as unique combinations of study identity, target plant species, and microbial inoculum. Each cohort represents a distinct plant–microbe interaction evaluated within a study, from which multiple effect sizes may arise due to the measurement of different traits or response variables. Effect sizes within a cohort are, therefore, expected to be correlated due to shared host material, microbial identity, and the experimental conditions, and were not treated as independent replicates. Microbial inoculum was defined operationally as the inoculation unit used in each study (strain, species, or consortium). We also constructed variance–covariance (VCV) matrices using the make_VCV_matrix function in metafor (Nakagawa et al. 2015). Effect sizes originating from the same experimental cohort were assumed to be correlated, and a correlation coefficient of ρ = 0.5 was specified to represent moderate dependence among outcomes derived from the same experiment. This value is commonly used in ecological meta-analyses (Ahmad et al. 2024) when multiple related response variables are measured on the same experimental units and provides a conservative balance between assuming independence (ρ = 0) and perfect correlation (ρ = 1). Meta-analytic models and heterogeneity estimation We conducted multilevel meta-analyses using the rma.mv() function in the metafor package to account for hierarchical data structure and multiple sources of non-independence in effect sizes across studies. Analyses were performed separately for lnRR and lnCVR. Random effects were specified as ~ 1 | Study_ID / Cohort / Effect_size_ID, partitioning variance among studies, plant–microbe cohorts within studies, and individual effect sizes within cohorts, thereby capturing between-study heterogeneity, biologically meaningful within-study variation, and residual variation among multiple outcomes from the same cohort. We first fitted models without moderators to estimate overall mean effects of seed microbiome inoculation on plant performance and variability. From these models, we extracted pooled effect size estimates along with 95% confidence intervals (CIs) and 95% prediction intervals (PIs), the latter representing the expected range of effects in future comparable studies (Higgins and Thompson 2002). Between-study heterogeneity was assessed using the I² statistic , which quantifies the proportion of total variance in effect sizes attributable to true heterogeneity rather than sampling error. I² is calculated as the ratio of random-effects variance to total variance and is expressed as a percentage. In the multilevel meta-analytic model, I² was decomposed across study-, cohort-, and effect-size-level random effects to estimate their relative contributions to overall heterogeneity (Ahmad et al. 2024). Moderator analyses To identify factors explaining variation in effect sizes, we conducted meta-regression analyses using biologically and experimentally relevant moderators (Rafiq and Reshi 2025). Specifically, we examined the effects of (i) experimental condition (field, nursery, greenhouse, growth chamber, laboratory), (ii) biogeographic status of the target plant (alien vs. other species), (iii) type of microbial inoculum (bacteria, fungi, or microbial consortia), and (iv) target plant species. It needs to be mentioned that across the studies retained for meta-analysis, seed-associated endophytes originating from alien invasive plant species were evaluated using two main experimental approaches: (i) inoculation onto the same invasive host species from which the endophytes were isolated, and (ii) inoculation onto other plant species, including both non-host plants and established model species such as Arabidopsis thaliana . These designs allowed assessment of both host-specific and cross-species effects of invasive plant seed endophytes on plant performance. For each moderator, models were fitted using category-specific intercepts (i.e., excluding a common intercept), allowing direct estimation of effect sizes for each level of the moderator. Model results were visualized using orchard plots, which display estimated mean effects alongside their confidence and prediction intervals, facilitating comparison among categories while illustrating the underlying heterogeneity. The explanatory power of moderators was assessed using marginal R² values, representing the proportion of variance explained by fixed effects. All moderator-specific effect estimates, CIs, and PIs are reported in tabular form to support comprehensive interpretation. Publication bias To assess the potential influence of publication bias in our meta-analysis, we applied a combination of visual and statistical approaches. We first constructed funnel plots to visually examine the relationship between effect sizes and their associated sampling precision, expressed as standard errors. Deviations from funnel plot symmetry may suggest the presence of publication bias or other small‐study effects. To formally test for such asymmetry, we fitted a meta‐regression model incorporating the sampling standard error (i.e., the square root of the sampling variance) as a moderator, following a modified version of Egger’s regression framework (Nakagawa et al. 2022). In addition, we conducted a separate meta‐regression including publication year as a moderator to detect any systematic changes in effect sizes over time, such as a decline in reported effects in more recent studies (Sánchez-Tójar et al. 2020, Atkinson et al. 2022). Together, these complementary approaches provided a robust assessment of potential biases and strengthened the reliability of our meta‐analytic inferences. All statistical analyses were conducted in R (version 4.5.1; R Core Team, 2025), with meta-analytic and meta-regression models fitted using the metafor package. 0. 1em 0.0 1em Results Overall effects of the seed microbiome on plant performance Across all studies and experimental contexts, seed microbiome inoculation had a significant positive effect on mean plant performance (Fig. 2). Across traits, seed endophyte inoculation was associated with predominantly positive effects on mean plant performance, as indicated by lnRR values clustering above zero for most growth and biomass-related traits (Fig. 2a). The strongest and most consistent increases were observed for fresh mass, shoot length, root length, and plant height, whereas reproductive growth and seed germination showed weaker and more variable responses. In contrast, effects on trait variability were comparatively small and inconsistent. lnCVR estimates (Fig. 2b) were generally centred around zero, indicating no uniform tendency for endophyte inoculation to either stabilize or increase phenotypic variation, although individual traits exhibited substantial dispersion in both directions. The overall meta-analytic estimate for lnRR was 0.30 (95% CI: 0.13 to 0.46; p < 0.001), indicating that, on average, plants associated with seed endophytes exhibited higher performance than uninoculated controls (Table 1). However, prediction intervals were wide (−0.61 to 1.20), suggesting substantial variability in outcomes among studies and experimental systems. Fig. 2. Effects of seed-associated endophyte inoculation on plant performance traits across studies. Panel (a) shows effect sizes expressed as the log response ratio (lnRR), representing proportional changes in mean trait values following inoculation, while panel (b) shows effect sizes expressed as the log coefficient of variation ratio (lnCVR), representing changes in trait variability. Each coloured point represents an individual effect size, grouped by plant trait. Point size is proportional to study precision (inverse of the standard error, 1/SE), with larger points indicating more precise estimates. Table 1: Overall mean effect sizes with 95% confidence intervals (CIs) and 95% prediction intervals (PIs). lnRR = log response ratio; lnCVR = log coefficient of variation ratio. lnRR 0.30 0.13 0.46 0.00 -0.61 1.20 lnCVR -0.28 -0.49 -0.07 0.01 -1.41 0.85 Table 2: Decomposition of random-effects variance across model levels. lnRR Between study (Study_ID) 0.14 66.90 lnRR Within study (Effect_size_ID) 0.00 0.00 lnRR Cohort 0.07 33.10 lnCVR Between study (Study_ID) 0.22 67.49 lnCVR Within study (Effect_size_ID) 0.00 0.00 lnCVR Cohort 0.10 32.51 0. 1em 0.0 1em Table 3: Decomposition of heterogeneity (I²) across random effects. 0. 1em 0.0 1em lnRR 99.73 66.71 0 lnCVR 80.07 54.04 0 0. 1em 0.0 1em In contrast, variability responses showed a significant overall reduction in trait variability following inoculation. The pooled lnCVR estimate was −0.28 (95% CI: −0.49 to −0.07; p = 0.01), indicating that seed microbiomes tended to reduce inter-individual variability in plant traits. As with mean effects, prediction intervals were broad (−1.41 to 0.85), highlighting strong context dependence. Variance decomposition indicated that heterogeneity in both lnRR and lnCVR was primarily structured across hierarchical levels of the meta-analytic model, reflecting differences among studies and among biologically defined cohorts within studies. For lnRR, between-study variation (Study_ID; variance = 0.14) accounted for 66.90% of the total heterogeneity, while cohort-level variation—representing distinct plant–microbe interaction combinations within studies—explained 33.10% (Table 2). No detectable variance was attributed to within-study differences among individual effect sizes (Effect_size_ID), likely due to the limited number of replicated effect sizes within cohorts, which constrained estimation of lower-level variance components. A comparable pattern was observed for lnCVR, where between-study heterogeneity explained 67.49% of the total variance and cohort-level effects accounted for the remaining 32.51%, again with no detectable within-study variance (Table 2) Consistent with these results, overall heterogeneity was high for both lnRR (I² = 99.73%) and lnCVR (I² = 80.07%), indicating that most observed variability represents true differences among studies rather than sampling error (Table 3). For lnRR, between‐study heterogeneity accounted for 66.71% of total I², whereas cohort‐level heterogeneity was negligible. For lnCVR, between‐study differences explained 54.04% of total heterogeneity, with no measurable contribution from cohort‐level effects. Together, these findings indicate that heterogeneity is dominated by variation among studies, while the absence of detectable within‐study variance is likely attributable to limited replication within cohorts rather than a true lack of variability. Moderator analyses Effects of experimental conditions on seed microbiome–mediated plant performance Experimental conditions significantly influenced plant performance responses mediated by the seed microbiome, with clear differences observed between mean effects (lnRR) and variability effects (lnCVR) across experimental settings (Table 4; Fig. S2). For lnRR, positive and statistically significant effects were detected under field conditions (estimate = 0.202, 95% CI: 0.038–0.367, P = 0.016) and laboratory conditions (estimate = 0.532, 95% CI: 0.286–0.778, P < 0.001), indicating enhanced plant performance in these environments. In contrast, effects observed in greenhouse and growth chamber experiments were positive but not statistically significant, with confidence intervals overlapping zero, suggesting weaker or more variable responses under these controlled conditions. Table 4: Effects of experimental conditions on plant performance mediated by the seed microbiome. Estimates are shown with 95% confidence intervals. lnRR = log response ratio; lnCVR = log coefficient of variation ratio. lnRR Field 0.202 0.038 0.367 0.016 lnRR Greenhouse 0.075 -0.047 0.197 0.228 lnRR Growth Chamber 0.094 -0.043 0.231 0.179 lnRR Laboratory 0.532 0.286 0.778 0.000 lnCVR Field -0.186 -0.373 0.001 0.051 lnCVR Greenhouse 0.306 0.167 0.445 0.000 lnCVR Growth Chamber 0.005 -0.154 0.164 0.949 lnCVR Laboratory -0.371 -0.716 -0.026 0.035 Table 5: Decomposition of random-effects variance across model levels for experimental condition. lnRR Between study (Study_ID) 0.05 26.77 lnRR Within study (Effect_size_ID) 0.05 26.77 lnRR Cohort 0.08 46.45 lnCVR Between study (Study_ID) 0.04 21.41 lnCVR Within study (Effect_size_ID) 0.04 21.41 lnCVR Cohort 0.10 57.18 0. 1em 0.0 1em Table 6: Decomposition of heterogeneity (I²) across random effects for experimental condition. 0. 1em 0.0 1em lnRR I 2 _total 99.70 lnRR I 2 _Study_ID 26.69 lnRR I 2 _Study_ID/Cohort 26.69 lnRR I 2 _Study_ID/Cohort/Effect_size_ID 46.31 lnCVR I 2 _total 68.22 lnCVR I 2 _Study_ID 14.60 lnCVR I 2 _Study_ID/Cohort 14.60 lnCVR I 2 _Study_ID/Cohort/Effect_size_ID 39.01 0. 1em 0.0 1em Patterns for lnCVR revealed contrasting effects of experimental condition on variability in plant performance. Variability tended to decrease under field conditions (estimate = −0.186), although this effect was marginally non-significant (P = 0.051), while a significant increase in variability was observed in greenhouse experiments (estimate = 0.306, 95% CI: 0.167–0.445, P < 0.001). Growth chamber experiments showed no detectable effect on variability, whereas laboratory conditions were associated with a significant reduction in variability (estimate = −0.371, 95% CI: −0.716 to −0.026, P = 0.035). Together, these results indicate that experimental context not only affects the magnitude of plant responses but also strongly influences the consistency of those responses. Variance decomposition further revealed that heterogeneity in responses differed across experimental conditions and was largely structured at higher hierarchical levels (Table 5). For lnRR, cohort-level variation accounted for the largest proportion of heterogeneity (46.45%), while between-study and within-study components contributed equally (26.77% each). A similar pattern was observed for lnCVR, where cohort-level effects explained the majority of variance (57.18%), with smaller and comparable contributions from between-study and within-study sources (21.41% each). In general, the results highlight the importance of cohort-specific factors in shaping responses across experimental settings. Consistent with this partitioning, overall heterogeneity remained high for lnRR (I² = 99.70%) and moderate for lnCVR (I² = 68.22%), indicating substantial real variation beyond sampling error (Table 6). For lnRR, heterogeneity was primarily distributed across nested model levels, with 46.31% attributed to the combined cohort, study, and effect-size levels, and 26.69% each associated with Study_ID and cohort-level effects. For lnCVR, 39.01% of the heterogeneity was explained by variation across all hierarchical levels, while 14.60% was attributable Study_ID and cohort-level differences. The identical I² values for certain components (Tables 5 and 6) arise from the scaling of variance components in the multilevel framework, particularly under limited within-study replication, and should not be interpreted as indicating identical biological processes. Effects of target species status on seed microbiome–mediated plant performance Meta-analytic results indicated that target species status significantly influenced mean plant performance responses (lnRR) mediated by the seed microbiome (Table 7; Fig. S3). Inoculation with endophytes originating from alien invasive species resulted in a significant positive effect on plant performance when tested on alien hosts (estimate = 0.224, 95% CI: 0.054–0.394, P = 0.010). An even stronger positive effect was observed when these endophytes were inoculated onto other plant species, including non-host and model plants (estimate = 0.394, 95% CI: 0.211–0.577, P < 0.001). In contrast, effects on variability in plant performance (lnCVR) differed between target species categories. For alien hosts, lnCVR did not differ significantly from zero, indicating no detectable change in performance variability. However, when endophytes were applied to other plant species, lnCVR was significantly negative (estimate = −0.560, 95% CI: −0.782 to −0.339, P < 0.001), suggesting a substantial reduction in variability and more consistent performance outcomes. Variance decomposition revealed that heterogeneity in lnRR and lnCVR was predominantly structured at higher hierarchical levels (Table 8). For lnRR, between-study variation accounted for 64.32% of the total heterogeneity, while cohort-level effects explained the remaining 35.68%. No detectable variance was attributed to within-study differences among effect sizes, likely reflecting limited replication within cohorts. A similar pattern was observed for lnCVR, where between-study heterogeneity explained 57.63% of the variance and cohort-level effects accounted for 42.37%, again with no measurable within-study variance. Consistent with these results, overall heterogeneity was high for lnRR (I² = 99.70%) and substantial for lnCVR (I² = 73.88%), indicating that most observed variability reflects real differences among studies rather than sampling error (Table 9). For lnRR, the majority of heterogeneity was attributable to between-study differences (64.13%), with the remainder associated with variation across nested cohort and effect-size levels (35.57%). For lnCVR, between-study heterogeneity accounted for 42.58% of the total I², while 31.30% was associated with combined cohort- and effect-size-level variation. Collectively, these findings demonstrate that the effects of seed-associated endophytes from alien invasive plants on plant performance are context-dependent, varying with target species identity and study-level characteristics, and that heterogeneity is driven primarily by differences among studies and experimental cohorts. Table 7: Effects of status of target species on plant performance mediated by the seed microbiome. Estimates are shown with 95% confidence intervals. lnRR = log response ratio; lnCVR = log coefficient of variation ratio. lnRR Alien 0.224 0.054 0.394 0.010 lnRR Others 0.394 0.211 0.577 0.000 lnCVR Alien -0.014 -0.214 0.187 0.893 lnCVR Others -0.560 -0.782 -0.339 0.000 0. 1em 0.0 1em Table 8: Decomposition of random-effects variance across model levels for experimental condition. 0. 1em 0.0 1em lnRR Between study (Study_ID) 0.12 64.32 lnRR Within study (Effect_size_ID) 0.00 0.00 lnRR Cohort 0.07 35.68 lnCVR Between study (Study_ID) 0.13 57.63 lnCVR Within study (Effect_size_ID) 0.00 0.00 lnCVR Cohort 0.10 42.37 0. 1em 0.0 1em Table 9: Decomposition of heterogeneity (I²) across random effects for experimental condition. 0. 1em 0.0 1em lnRR I 2 _total 99.70 lnRR I 2 _Study_ID 64.13 lnRR I 2 _Study_ID/Cohort 0.00 lnRR I 2 _Study_ID/Cohort/Effect_size_ID 35.57 lnCVR I 2 _total 73.88 lnCVR I 2 _Study_ID 42.58 lnCVR I 2 _Study_ID/Cohort 0.00 lnCVR I 2 _Study_ID/Cohort/Effect_size_ID 31.30 Effects of inoculum type on seed microbiome–mediated plant performance Inoculum type had a significant influence on plant performance mediated by the seed microbiome, with contrasting effects observed across microbial groups (Table 10; Fig. S4). Inoculation with bacterial isolates resulted in a significant increase in mean plant performance (lnRR estimate = 0.293, 95% CI: 0.085–0.501, P = 0.006) and was also associated with a significant reduction in performance variability (lnCVR estimate = −0.333, 95% CI: −0.601 to −0.065, P = 0.015), indicating both enhanced and more consistent plant responses. Fungal inocula likewise produced a significant positive effect on mean performance (lnRR estimate = 0.359, 95% CI: 0.086–0.632, P = 0.010), although their effects on variability were not statistically significant. In contrast, microbial consortia did not significantly affect either mean performance or variability, with confidence intervals overlapping zero for both lnRR and lnCVR, suggesting highly context-dependent or inconsistent outcomes for mixed inocula. Table 10: Effects of inoculum type on plant performance mediated by the seed microbiome. Estimates are shown with 95% confidence intervals. lnRR = log response ratio; lnCVR = log coefficient of variation ratio. 0. 1em 0.0 1em lnRR Bacteria 0.293 0.085 0.501 0.006 lnRR Consortium -0.044 -0.409 0.322 0.814 lnRR Fungi 0.359 0.086 0.632 0.010 lnCVR Bacteria -0.333 -0.601 -0.065 0.015 lnCVR Consortium -0.273 -0.842 0.295 0.346 lnCVR Fungi -0.192 -0.569 0.185 0.318 0. 1em 0.0 1em Table 11: Decomposition of random-effects variance across model levels for experimental condition. 0. 1em 0.0 1em lnRR Between study (Study_ID) 0.14 67.14 lnRR Within study (Effect_size_ID) 0.00 0.00 lnRR Cohort 0.07 32.85 lnCVR Between study (Study_ID) 0.23 68.75 lnCVR Within study (Effect_size_ID) 0.00 0.00 lnCVR Cohort 0.10 31.25 0. 1em 0.0 1em Table 12: Decomposition of heterogeneity (I²) across random effects for experimental condition. lnRR I 2 _total 99.73 lnRR I 2 _Study_ID 66.96 lnRR I 2 _Study_ID/Cohort 0.00 lnRR I 2 _Study_ID/Cohort/Effect_size_ID 32.77 lnCVR I 2 _total 80.71 lnCVR I 2 _Study_ID 55.49 lnCVR I 2 _Study_ID/Cohort 0.00 lnCVR I 2 _Study_ID/Cohort/Effect_size_ID 25.22 0. 1em 0.0 1em Variance decomposition revealed that heterogeneity associated with inoculum type was predominantly structured at higher hierarchical levels (Table 11). For lnRR, between-study differences accounted for 67.14% of the total heterogeneity, while cohort-level variation explained the remaining 32.85%, with no detectable contribution from within-study differences among effect sizes. A similar pattern was observed for lnCVR, where between-study heterogeneity explained 68.75% of the variance and cohort-level effects accounted for 31.25%. Consistent with this variance partitioning, overall heterogeneity was high for lnRR (I² = 99.73%) and substantial for lnCVR (I² = 80.71%), indicating that most observed variability reflects real differences among studies rather than sampling error (Table 12). For lnRR, 66.96% of the total heterogeneity was attributable to between-study variation, with the remaining 32.77% associated with combined cohort- and effect-size-level differences. For lnCVR, between-study heterogeneity accounted for 55.49% of total I², while 25.22% was attributable to variation across nested cohort and effect-size levels. Overall, these results indicate that plant responses to seed microbiome inoculation vary strongly with inoculum type and that heterogeneity is driven primarily by differences among studies and experimental cohorts. Assessment of publication bias and small-study effects Evidence for publication bias and small-study effects was evaluated using meta-regression models with the square root of the sampling variance included as a moderator. For lnRR, the intercept was significantly different from zero (estimate = 0.385, 95% CI: 0.182 to 0.588, p < 0.001), indicating a positive overall effect size. However, the effect of study precision (√sampling variance) was not statistically significant (estimate = −0.352, 95% CI: −0.779 to 0.075, p = 0.106). This suggests no strong evidence of small-study effects or publication bias influencing lnRR estimates. For lnCVR, the intercept was marginally significant and negative (estimate = −0.338, 95% CI: −0.676 to −0.001, p = 0.049), indicating a modest overall reduction in variability. In contrast, the coefficient for study precision was non-significant (estimate = 0.079, 95% CI: −0.468 to 0.625, p = 0.777), providing no indication of small-study effects or asymmetry in lnCVR. Overall, the non-significant relationships between effect sizes and their sampling variances for both lnRR and lnCVR indicate little evidence for publication bias or small-study effect s in the dataset. . Fig. 3 . Funnel plots assessing publication bias in the studies. Each point represents an individual study’s effect size (residual value) plotted against its precision (inverse standard error). The vertical dashed line at 0 indicates no effect. Symmetrical distribution around the midline suggests minimal bias, while asymmetry may indicate missing studies. Points are color-coded by p-value ranges, with darker shades representing more statistically significant results (p ≤ 0.01). The inverted funnel shape reflects expected precision-effect relationships in unbiased meta-analyses. Temporal trends in effect sizes and precision Precision-weighted temporal meta-regressions revealed no strong evidence for systematic changes in mean effect sizes over time. For lnRR, effect sizes were distributed around a small positive mean across publication years, with the precision-weighted regression line remaining largely flat (Fig. S5a). Larger, more precise studies clustered closely around the overall mean, whereas greater dispersion was observed among smaller and less precise studies, particularly in more recent years. Despite this variability, the absence of a pronounced temporal slope indicates that reported lnRR estimates have remained broadly consistent over time, providing no indication of time-lag bias or a decline effect. Prediction intervals remained wide throughout the study period, highlighting substantial context dependence among individual studies. In contrast, lnCVR exhibited a weak negative temporal trend (Fig. S5b), suggesting a modest reduction in reported variability effects over time. Early studies showed greater scatter and more extreme lnCVR values, while later studies—especially those with higher precision—converged toward values close to zero. This pattern indicates increasing stabilization of variability estimates in more recent literature, likely reflecting improved experimental design or larger sample sizes. Nonetheless, the wide prediction intervals observed across all years underscore persistent heterogeneity among studies. Overall, these temporal patterns suggest that neither mean effects nor variability estimates are strongly influenced by publication year, and that the meta-analytic conclusions are robust to potential temporal and small-study biases. 0. 1em 0.0 1em Discussion Our global meta-analysis revisits a cornerstone of invasion biology, namely propagule pressure, by explicitly incorporating its long-overlooked microbial dimension. We show that seed-borne endophytes associated with alien plants consistently enhance mean plant trait performance, while their effects on trait variability are weak, inconsistent, and strongly context dependent. Specifically, our results reveal that seed endophyte inoculation from alien plant species exerts a consistent, positive overall effect on mean plant trait performance, whereas effects on trait variability are weaker, context-dependent, and largely non-significant. This asymmetry necessitates a reconceptualization of the propagule (seed) of alien plant species: instead of viewing it as only a dormant embryo, it should be understood as a functional holobiont (Vandenkoornhuyse et al. 2015)—a biological unit in which associated microbes play a key role in determining establishment success. The growth advantage conferred by these endophytes further suggests that classical formulations of propagule pressure, which emphasize seed number and introduction frequency (Simberloff 2009), must be expanded to include a qualitative dimension of inoculation potential—the capacity of each seed to introduce a beneficial microbiome into the recipient environment. The robust positive effect on mean trait performance (lnRR ≈ 0.18, corresponding to an ~20% increase) represents a biologically meaningful enhancement of propagule quality. This result extends the Enhanced Mutualist Hypothesis by demonstrating that the source of enhancement can be intrinsic to the propagule itself, rather than acquired post-dispersal from the invaded soil community (Reinhart and Callaway 2006). By producing phytohormones, mobilizing nutrients, and priming host defence pathways, this vertically transmitted “symbiotic toolkit” reduces early dependence on the novel soil microbiome and increases the probability of successful establishment per arriving seed (Hardoim et al. 2015). Importantly, this microbial advantage operates during the most vulnerable life-history stages, directly alleviating a key demographic bottleneck in the invasion process. At the same time, the extreme heterogeneity surrounding this mean effect (I² ≈ 99%) is not merely statistical noise, but an ecological signal of contingency. The expression of microbial benefits is filtered through a triad of interacting factors: host genotype compatibility, functional identity of the inoculum, and environmental context (Compant et al., 2010). Consequently, the microbial contribution to propagule pressure should be understood not as a fixed trait, but as a realized potential that varies across invasion landscapes (Bonthond et al. 2021, Zobel et al. 2024). This framework helps explain why identical seed inputs can yield contrasting invasion outcomes and reconciles previously inconsistent findings in the literature: while the microbial toolkit is potent, its deployment is inherently context specific. In contrast, the influence of seed endophytes on trait variability—a potential proxy for phenotypic buffering against environmental stochasticity—was weak and inconsistent across studies. This finding challenges the assumption that seed-associated microbes universally function as insurance mechanisms under variable establishment conditions. Although individual endophytes can mitigate specific stresses, e.g. drought tolerance (Zarraga-Barco et al. 2024), our synthesis indicates that broad-spectrum stabilization is not a conserved property of alien seed endophyte communities (Rodriguez and Redman 2008). Conceptually, these microbes function more reliably as generalist “Jacks-of-All-Trades” (Xu et al. 2021) that promote growth than as ubiquitous “Masters-of-Some” that consistently buffer stress (Reynolds et al. 2003). From an invasion-dynamics perspective, this distinction is critical: while seed endophytes elevate average cohort fitness, they do not necessarily reduce variance in individual success, leaving demographic stochasticity as a persistent filter on invasion outcomes (Palamara et al. 2016). Our moderator analyses highlight the ecological contexts that modulate the expression of this microbial toolkit. The contrasting responses observed between laboratory and greenhouse studies underscore how different experimental conditions alter endophyte function (Rafiq and Reshi 2025). Reduced trait variability emerged primarily under simplified, controlled conditions, whereas more complex environments amplified variability, likely through interacting biotic and abiotic factors (Zhang et al. 2025). Inoculum composition also proved decisive: single-strain inoculations—particularly bacterial isolates—produced more consistent benefits than multi-species consortia, where inter-microbial competition may dilute functional effects (Liu et al. 2023). Notably, non-alien host plants derived greater and more stable benefits from alien endophytes than alien hosts themselves (Mei et al. 2022, Bard et al. 2024). This pattern raises the provocative possibility that the microbial assemblage of seeds may exert substantial spillover effects, potentially reshaping native plant, thereby impacting microbial feedbacks in invaded communities. Synthesizing these insights, we propose an expanded conceptual framework— Holobiont Propagule Pressure —that integrates numerical, microbial, and contextual dimensions of invasion. In this framework, propagule pressure is not defined solely by the quantity and frequency of seed arrival, but also by the microbial quality of the propagule, namely the composition and functional capacity of the vertically transmitted seed microbiome, together with the probability that these microbial benefits are realized under local environmental and biotic conditions. The microbial consortium associated with each propagule thus functions as a force multiplier on numerical pressure by elevating the mean probability of establishment per seed. However, because the expression of these microbial benefits is strongly context dependent, the magnitude and consistency of this multiplier vary across invasion scenarios, being shaped by interactions among host identity, microbial function, and the receiving environment. Within this framework, the microbial consortium acts as a force multiplier on numerical pressure by elevating mean establishment probability, while its context dependence determines the magnitude and consistency of this effect across invasion scenarios. 0. 1em 0.0 1em Conclusion In conclusion, this meta-analysis elevates the seed microbiome from a peripheral curiosity to a central component of invasion theory. The evidence compels a revision of propagule pressure to incorporate the qualitative, microbial attributes of arriving diaspores. Future research must move beyond documenting associations to uncovering mechanistic causation by integrating functional genomics of seed endophyte communities with manipulative experiments conducted across realistic environmental gradients. Such efforts will allow invasion biology to progress from counting seeds to predicting invasion outcomes—ultimately yielding a more mechanistic and powerful framework for understanding and forecasting plant invasions. 0. 1em 0.0 1em Implications and future directions Collectively, our results indicate that seed endophytes of invasive alien plant species possess real but conditional potential to enhance plant performance. Their effects are best understood not as universal drivers of invasiveness, but as context-dependent modifiers of plant traits that may tip the balance under favourable conditions. Future research should move beyond simple inoculation trials and integrate environmental gradients, host functional traits, and microbial functional profiles into predictive frameworks. Field-based experiments and comparative studies across native and invaded ranges will be particularly important for determining when seed endophytes meaningfully contribute to invasion success. Ultimately, understanding seed microbiomes as context-sensitive ecological partners, rather than universal mutualists, will be key to resolving their role in plant invasions and harnessing their potential in applied settings. 0. 1em 0.0 1em Conflict of interest statement The authors declare that they have no competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. 0. 1em 0.0 1em Data availability statement All data used in this meta-analysis are provided in Supplementary Appendix 1. Sheet 1 contains the complete dataset used for the analyses, and Sheet 2 provides the full bibliographic details of all studies included in the synthesis. In addition, the R code used for all analyses is also shared. 0. 1em 0.0 1em References Ahmad, R., Lone, S. A., Rashid, I. and Khuroo, A. A. 2024. Ecological impacts of a global plant invader: synthesizing mean and variance effects using meta-analysis. - Oikos 2024: e10102. Atkinson, J., Brudvig, L. A., Mallen-Cooper, M., Nakagawa, S., Moles, A. T. and Bonser, S. P. 2022. Terrestrial ecosystem restoration increases biodiversity and reduces its variability, but not to reference levels: A global meta-analysis. - Ecol. Lett. 25: 1725–1737. Bard, N. W., Cronk, Q. C. B. and Davies, T. J. 2024. Fungal endophytes can modulate plant invasion. - Biological Reviews 99: 1652–1671. Beckstead, J., Meyer, S. E., Street, L. E. and Allen, P. S. 2010. Effect of fire on a seed bank pathogen and on seeds of its host Bromus tectorum. - In: Pendleton, Rosemary; Meyer, Susan; Schultz, Bitsy, eds. Conference Proceedings: Seed Ecology III - The Third International Society for Seed Science Meeting on Seeds and the Environment - “Seeds and Change”; June 20-June 24, 2010; Salt Lake City, Utah…: 10–11. Berg, G. and Raaijmakers, J. M. 2018. Saving seed microbiomes. - ISME J. 12: 1167–1170. Bonthond, G., Bayer, T., Krueger-Hadfield, S. A., Stärck, N., Wang, G., Nakaoka, M., Künzel, S. and Weinberger, F. 2021. The role of host promiscuity in the invasion process of a seaweed holobiont. - ISME J. 15: 1668. Bulgarelli, D., Schlaeppi, K., Spaepen, S., Van Themaat, E. V. L. and Schulze-Lefert, P. 2013. Structure and functions of the bacterial microbiota of plants. - Annu. Rev. Plant Biol. 64: 807–838. Catford, J. A., Jansson, R. and Nilsson, C. 2009. Reducing redundancy in invasion ecology by integrating hypotheses into a single theoretical framework. - Divers. Distrib. 15: 22–40. Compant, S., Samad, A., Faist, H. and Sessitsch, A. 2019. A review on the plant microbiome: Ecology, functions, and emerging trends in microbial application. - J. Adv. Res. 19: 29–37. Condessa, M., Jesus, J. G., Máguas, C., Le Roux, J. J. and Trindade, H. 2024. The seeds of invasion: a comparison of endophytic seed bacteria of Acacia longifolia between its native and invasive ranges. - Symbiosis 2024 93:1 93: 29–42. Dawson, W. and Schrama, M. 2016. Identifying the role of soil microbes in plant invasions. - Journal of Ecology 104: 1211–1218. Escudero-Martinez, C. and Bulgarelli, D. 2019. Tracing the evolutionary routes of plant–microbiota interactions. - Curr. Opin. Microbiol. 49: 34–40. Essl, F., Lenzner, B., Bacher, S., Bailey, S., Capinha, C., Daehler, C., Dullinger, S., Genovesi, P., Hui, C., Hulme, P. E., Jeschke, J. M., Katsanevakis, S., Kühn, I., Leung, B., Liebhold, A., Liu, C., MacIsaac, H. J., Meyerson, L. A., Nuñez, M. A., Pauchard, A., Pyšek, P., Rabitsch, W., Richardson, D. M., Roy, H. E., Ruiz, G. M., Russell, J. C., Sanders, N. J., Sax, D. F., Scalera, R., Seebens, H., Springborn, M., Turbelin, A., van Kleunen, M., von Holle, B., Winter, M., Zenni, R. D., Mattsson, B. J. and Roura-Pascual, N. 2020. Drivers of future alien species impacts: An expert-based assessment. - Glob. Chang. Biol. 26: 4880–4893. Hardoim, P. R., van Overbeek, L. S., Berg, G., Pirttilä, A. M., Compant, S., Campisano, A., Döring, M. and Sessitsch, A. 2015. The Hidden World within Plants: Ecological and Evolutionary Considerations for Defining Functioning of Microbial Endophytes. - Microbiology and Molecular Biology Reviews 79: 293–320. Hedges, L. V., Gurevitch, J. and Curtis, P. S. 1999. The Meta-Analysis of Response Ratios in Experimental Ecology. - Ecology 80: 1150. Higgins, J. P. T. and Thompson, S. G. 2002. Quantifying heterogeneity in a meta-analysis. - Stat. Med. 21: 1539–1558. Hu, H., Geng, S., Zhu, Y., He, X., Pan, X., Yang, M., Hu, H., Geng, S., Zhu, Y., He, X., Pan, X. and Yang, M. 2025. Seed-Borne Endophytes and Their Host Effects. - Microorganisms 2025, Vol. 13, in press. Kholostiakov, V., Burns, B., Ridgway, H. and Padamsee, M. 2025. Variation in seed-borne microbial communities of Metrosideros excelsa Sol. ex Gaertn. with consequences for germination success. - N. Z. J. Bot. 63: 1981–2004. Lekberg, Y., Bever, J. D., Bunn, R. A., Callaway, R. M., Hart, M. M., Kivlin, S. N., Klironomos, J., Larkin, B. G., Maron, J. L., Reinhart, K. O., Remke, M. and van der Putten, W. H. 2018. Relative importance of competition and plant–soil feedback, their synergy, context dependency and implications for coexistence. - Ecol. Lett. 21: 1268–1281. Liu, X., Mei, S. and Salles, J. F. 2023. Inoculated microbial consortia perform better than single strains in living soil: A meta-analysis. - Applied Soil Ecology 190: 105011. Lockwood, J. L., Cassey, P. and Blackburn, T. 2005. The role of propagule pressure in explaining species invasions. - Trends Ecol. Evol. 20: 223–228. Mei, Y. H., Li, X., Zhou, J. Y., Kong, F. L., Qi, S. S., Zhu, B., Naz, M., Dai, Z. C. and Du, D. L. 2022. Both Adaptability and Endophytic Bacteria Are Linked to the Functional Traits in the Invasive Clonal Plant Wedelia trilobata. - Plants 11: 3369. Mendes, R., Kruijt, M., De Bruijn, I., Dekkers, E., Van Der Voort, M., Schneider, J. H. M., Piceno, Y. M., DeSantis, T. Z., Andersen, G. L., Bakker, P. A. H. M. and Raaijmakers, J. M. 2011. Deciphering the rhizosphere microbiome for disease-suppressive bacteria. - Science (1979). 332: 1097–1100. Nakagawa, S., Poulin, R., Mengersen, K., Reinhold, K., Engqvist, L., Lagisz, M. and Senior, A. M. 2015. Meta-analysis of variation: Ecological and evolutionary applications and beyond. - Methods Ecol. Evol. 6: 143–152. Nakagawa, S., Lagisz, M., Jennions, M. D., Koricheva, J., Noble, D. W. A., Parker, T. H., Sánchez-Tójar, A., Yang, Y. and O’Dea, R. E. 2022. Methods for testing publication bias in ecological and evolutionary meta-analyses. - Methods Ecol. Evol. 13: 4–21. Nelson, E. B. 2017. The seed microbiome: Origins, interactions, and impacts. - Plant and Soil 2017 422:1 422: 7–34. O’Dea, R. E., Lagisz, M., Jennions, M. D., Koricheva, J., Noble, D. W. A., Parker, T. H., Gurevitch, J., Page, M. J., Stewart, G., Moher, D. and Nakagawa, S. 2021. Preferred reporting items for systematic reviews and meta-analyses in ecology and evolutionary biology: a PRISMA extension. - Biological Reviews 96: 1695–1722. Palamara, G. M., Carrara, F., Smith, M. J. and Petchey, O. L. 2016. The effects of demographic stochasticity and parameter uncertainty on predicting the establishment of introduced species. - Ecol. Evol. 6: 8440–8451. Pyšek, P., Hulme, P. E., Simberloff, D., Bacher, S., Blackburn, T. M., Carlton, J. T., Dawson, W., Essl, F., Foxcroft, L. C., Genovesi, P., Jeschke, J. M., Kühn, I., Liebhold, A. M., Mandrak, N. E., Meyerson, L. A., Pauchard, A., Pergl, J., Roy, H. E., Seebens, H., van Kleunen, M., Vilà, M., Wingfield, M. J. and Richardson, D. M. 2020. Scientists’ warning on invasive alien species. - Biol. Rev. Camb. Philos. Soc. 95: 1511–1534. Quaglini, L. A., Yannelli, F. A., Fasano, F., Montagnani, C., Caronni, S., Citterio, S. and Gentili, R. 2025. Abiotic and biotic factors shape the invasion success of the alien plant species Senecio inaequidens (Asteraceae) in northern Italy. - Journal of Plant Ecology in press. Rafiq, I. and Reshi, Z. A. 2025. Global meta-analysis of endophytic inoculation effects on seed germination, seedling growth, and nutrient uptake of trees. - Rhizosphere 34: 101079. Reinhart, K. O. and Callaway, R. M. 2006a. Soil biota and invasive plants. - New Phytologist 170: 445–457. Reinhart, K. O. and Callaway, R. M. 2006b. Soil biota and invasive plants. - New Phytologist 170: 445–457. Reinhart, K. O., Tytgat, T., Van der Putten, W. H. and Clay, K. 2010. Virulence of soil-borne pathogens and invasion by Prunus serotina. - New Phytologist 186: 484–495. Reinhart, K. O., Bauer, J. T., McCarthy-Neumann, S., MacDougall, A. S., Hierro, J. L., Chiuffo, M. C., Mangan, S. A., Heinze, J., Bergmann, J., Joshi, J., Duncan, R. P., Diez, J. M., Kardol, P., Rutten, G., Fischer, M., van der Putten, W. H., Bezemer, T. M. and Klironomos, J. 2021. Globally, plant-soil feedbacks are weak predictors of plant abundance. - Ecol. Evol. 11: 1756–1768. Reynolds, H. L., Packer, A., Bever, J. D. and Clay, K. 2003. GRASSROOTS ECOLOGY: PLANT–MICROBE–SOIL INTERACTIONS AS DRIVERS OF PLANT COMMUNITY STRUCTURE AND DYNAMICS. - Ecology 84: 2281–2291. Richardson, D. M. and Pyšek, P. 2012. Naturalization of introduced plants: Ecological drivers of biogeographical patterns. - New Phytologist 196: 383–396. Rodriguez, R. and Redman, R. 2008. More than 400 million years of evolution and some plants still can’t make it on their own: plant stress tolerance via fungal symbiosis. - J. Exp. Bot. 59: 1109–1114. Rout, M. E. and Callaway, R. M. 2012. Interactions between exotic invasive plants and soil microbes in the rhizosphere suggest that ‘everything is not everywhere.’ - Ann. Bot. 110: 213–222. Roy, H. E., Pauchard, A., Stoett, P., Truong, T. R. and Lipinskaya, T. 2023. IPBES invasive alien species assessment: Chapter 1. Introducing biological invasions and the IPBES thematic assessment of invasive alien species and their control. - In: Thematic Assessment Report on Invasive Alien Species and thier Control of the Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services. in press. Sánchez-Tójar, A., Moran, N. P., O’Dea, R. E., Reinhold, K. and Nakagawa, S. 2020. Illustrating the importance of meta-analysing variances alongside means in ecology and evolution. - J. Evol. Biol. 33: 1216–1223. Shade, A., Jacques, M. A. and Barret, M. 2017. Ecological patterns of seed microbiome diversity, transmission, and assembly. - Curr. Opin. Microbiol. 37: 15–22. Simberloff, D. 2009a. The role of propagule pressure in biological invasions. - Annu. Rev. Ecol. Evol. Syst. 40: 81–102. Simberloff, D. 2009b. The role of propagule pressure in biological invasions. - Annu. Rev. Ecol. Evol. Syst. 40: 81–102. Simonin, M., Briand, M., Chesneau, G., Rochefort, A., Marais, C., Sarniguet, A. and Barret, M. 2022. Seed microbiota revealed by a large-scale meta-analysis including 50 plant species. - New Phytologist 234: 1448–1463. Truyens, S., Weyens, N., Cuypers, A. and Vangronsveld, J. 2015. Bacterial seed endophytes: Genera, vertical transmission and interaction with plants. - Environ. Microbiol. Rep. 7: 40–50. Van Der Putten, W. H., Klironomos, J. N. and Wardle, D. A. 2007. Microbial ecology of biological invasions. - ISME J. 1: 28–37. Vandenkoornhuyse, P., Quaiser, A., Duhamel, M., Le Van, A. and Dufresne, A. 2015. The importance of the microbiome of the plant holobiont. - New Phytologist 206: 1196–1206. Viechtbauer, W. 2010. Conducting Meta-Analyses in R with the metafor Package. - J. Stat. Softw. in press. War, A. F., Bashir, I., Reshi, Z. A., Kardol, P. and Rashid, I. 2023. Insights into the seed microbiome and its ecological significance in plant life. - Microbiol. Res. 269: 127318. Xu, Q., Vandenkoornhuyse, P., Li, L., Guo, J., Zhu, C., Guo, S., Ling, N. and Shen, Q. 2021. Microbial generalists and specialists differently contribute to the community diversity in farmland soils. - J. Adv. Res. 40: 17. Zarraga-Barco, F., Ueno, A. C., Casabella, M. P., Casas, C., Molina Montenegro, M. A., Ramos, P., Schnyder, H. and Gundel, P. E. 2024. A seed-borne endophyte mediates plant drought responses and intergenerational effects on seed characteristics. - Environ. Exp. Bot. 221: 105719. Zhang, Y. Y., Li, Y. Z. and Shi, Z. J. 2025. Host-Specific and Environment-Dependent Effects of Endophyte Alternaria oxytropis on Three Locoweed Oxytropis Species in China. - Journal of Fungi in press. Zobel, M., Koorem, K., Moora, M., Semchenko, M. and Davison, J. 2024. Symbiont plasticity as a driver of plant success. - New Phytologist 241: 2340–2352. Information & Authors Information Version history V1 Version 1 30 January 2026 Copyright This work is licensed under a Non Exclusive No Reuse License. Keywords establishment holobiont invasion log coefficient of variation ratio log response ratio microbiome Authors Affiliations Zafar Reshi 0000-0001-9567-7484 [email protected] University of Kashmir View all articles by this author Iflah Rafiq University of Kashmir View all articles by this author Metrics & Citations Metrics Article Usage 172 views 89 downloads .FvxKWukQNSOunydq8rnd { width: 100px; } Citations Download citation Zafar Reshi, Iflah Rafiq. Revisiting Propagule Pressure Theory: A Global Meta-Analysis of Seed Endophytes from Alien Plants on Plant Performance and Trait Variability. Authorea . 30 January 2026. DOI: https://doi.org/10.22541/au.176978530.04074018/v1 If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download. For more information or tips please see 'Downloading to a citation manager' in the Help menu . Format Please select one from the list RIS (ProCite, Reference Manager) EndNote BibTex Medlars RefWorks Direct import Tips for downloading citations document.getElementById('citMgrHelpLink').addEventListener('click', function() { popupHelp(this.href); return false; }); $(".js__slcInclude").on("change", function(e){ if ($(this).val() == 'refworks') $('#direct').prop("checked", false); $('#direct').prop("disabled", ($(this).val() == 'refworks')); }); View Options View options PDF View PDF Figures Tables Media Share Share Share article link Copy Link Copied! Copying failed. Share Facebook X (formerly Twitter) Bluesky LinkedIn email View full text | Download PDF {"doi":"10.22541/au.176978530.04074018/v1","type":"Article"} Now Reading: Share Figures Tables Close figure viewer Back to article Figure title goes here Change zoom level Go to figure location within the article Download figure Toggle share panel Toggle share panel Share Toggle information panel Toggle information panel Go to previous graphic Go to next graphic Go to previous table Go to next table All figures All tables View all material View all material xrefBack.goTo xrefBack.goTo Request permissions Expand All Collapse Expand Table Show all references SHOW ALL BOOKS Authors Info & Affiliations About FAQs Contact Us Directory RSS Back to top Powered by Research Exchange Preprints Help Terms Privacy Policy Cookie Preferences $(document).ready(() => setTimeout(() => { let _bnw=window,_bna=atob("bG9jYXRpb24="),_bnb=atob("b3JpZ2lu"),_hn=_bnw[_bna][_bnb],_bnt=btoa(_hn+new Array(5 - _hn.length % 4).join(" ")); $.get("/resource/lodash?t="+_bnt); },4000)); (function(){function c(){var b=a.contentDocument||a.contentWindow.document;if(b){var d=b.createElement('script');d.innerHTML="window.__CF$cv$params={r:'9fdf93f4ee421640',t:'MTc3OTE1Njc1OQ=='};var a=document.createElement('script');a.src='/cdn-cgi/challenge-platform/scripts/jsd/main.js';document.getElementsByTagName('head')[0].appendChild(a);";b.getElementsByTagName('head')[0].appendChild(d)}}if(document.body){var a=document.createElement('iframe');a.height=1;a.width=1;a.style.position='absolute';a.style.top=0;a.style.left=0;a.style.border='none';a.style.visibility='hidden';document.body.appendChild(a);if('loading'!==document.readyState)c();else if(window.addEventListener)document.addEventListener('DOMContentLoaded',c);else{var e=document.onreadystatechange||function(){};document.onreadystatechange=function(b){e(b);'loading'!==document.readyState&&(document.onreadystatechange=e,c())}}}})();