Invasive earthworms can change understory plant community traits and reduce plant functional diversity

Summary Among the most important impacts of biological invasions on biodiversity is biotic homogenization, which may further compromise key ecosystem processes. However, the extent to which they homogenize functional diversity and shift dominant ecological strategies of invaded communities remains uncertain. Here, we investigated changes in plant communities in a northern North American forest in response to invasive earthworms, by examining the taxonomic and functional diversity of the plant community and soil ecosystem functions. We found that although plant taxonomic diversity did not change in response to invasive earthworms, they modified the dominance structure of plant functional groups. Invasive earthworms promoted the dominance of fast-growing plants at the expense of slow-growing ones. Moreover, earthworms decreased plant functional diversity, which coincided with changes in abiotic and biotic soil properties. Our study reveals that invasive earthworms erode multiple biodiversity facets of invaded forests, with potential cascading effects on ecosystem functioning.


INTRODUCTION
There is growing evidence that ecosystems and biodiversity face multiple threats. 1,24][5][6][7] Biotic homogenization is one of the most prominent impacts of biological invasions on biodiversity, which may further compromise key ecosystem processes. 8,9It consists of the increase of the genetic, taxonomic, and/or functional similarity between communities. 8,10While there is growing evidence that biological invasions homogenize native communities taxonomically, whether invaded communities become more functionally similar is uncertain, 8,11 which is in line with the poor use of the functional diversity indices in the context of biological invasions. 12iotic homogenization can result from the increase in the number of species shared between communities: when the invader of the same trophic-level as the community establishes at several locations, it increases community similarity. 8,13The homogenization process can also happen when the establishment of the invasive species (from the same trophic-level or not as the one of the communities considered) induces a similar species loss across the communities, resulting in greater compositional similarity. 8,13In this case, we could also expect biotic differentiation with an increase in dissimilarity, if different species are lost across communities for example. 14The ecological mechanisms that underlie biotic homogenization due to invasive species establishment and spread in native ecosystems can be diverse.Invasive species can, for example, directly affect native species via competitive 15,16 or feeding interactions. 17,18They also indirectly affect native species via other organisms [19][20][21] or by altering environmental conditions. 22Consequently, invasive species can shape the future of the invaded native communities by selecting for particular species of the native community thus giving to the community a novel combination of functional trait values, 12 and probably facilitating the invasion of other species. 23Shifts in species composition, dominance structure, and trait diversity and composition due to biological invasions are likely to impact ecosystem functioning, 24 following the mass ratio hypothesis. 25hile the impacts of aboveground invasive species are well known, those of belowground invasive invertebrates, such as earthworms, are underappreciated. 13However, they could be more significant when invasive species are ecosystem engineers. 26,27Invasive earthworms, for example, can be particularly impactful to native biodiversity and ecosystems that previously lacked native earthworms, such as in most northern North American forests [28][29][30][31] and Arctic regions. 32Recent studies showed that invasive earthworms affect soil microorganisms, soil ll OPEN ACCESS invertebrates, [33][34][35][36] as well as native plant communities, 37,38 and soil surface/vegetation-dwelling invertebrates. 31,36,39The ecological mechanisms behind their impacts are numerous, and the direction and magnitude of the effects vary with earthworm diversity and community composition. 35,38,401][42][43][44][45][46][47][48] Second, invasive earthworms can directly affect plant communities via multiple paths.][55][56][57][58] Moreover, invasive earthworms do not only affect the initial stages of plant community assembly (or re-assembly).Their effects on plant communities could also be explained by their impacts on plant traits and biomass. 59,60However, their impacts on trait diversity and composition, while hypothesized, have yet to be examined.Earthworms have been shown to impact traits linked to plant development and resource uptake, such as plant height, 61 nutrient content, 48 root growth, 32,62 and mycorrhizal association, 63 but also traits linked to plant vegetative reproduction, such as the number of culms. 64Plant trait responses to earthworm presence are context dependent, depending on plant species and functional group identity. 60,65,66Indeed, although the underlying ecological mechanisms are understudied, grasses seem to benefit more from earthworm activity than herbs, 60 and dominance of grass species is commonly observed in the presence of invasive earthworms. 38,67e, therefore, hypothesize that plant species that are more efficient in taking up resources under conditions of higher resource availability will dominate areas with invasive earthworms which, in turn, will alter habitat conditions.The abundance of fast, or 'acquisitive' plant species, 68,69 which typically have high specific leaf area, nitrogen content, and specific root length, as well as low leaf dry matter and carbon content (like many grass species), would be higher in invaded areas.Indeed, most grass species possess traits associated with fast growth and resource uptake. 70,71This selection process linked to the activity of invasive earthworms is expected to lead to a decrease in taxonomic diversity and an increase in plant species functional similarity in invaded areas.As the dominance of fast-growing plant species would increase with the presence of invasive earthworms, we expect to observe a shift in trait composition.Overall, the community-weighted mean of leaf nitrogen content and specific leaf area would be higher, while the community-weighted mean leaf dry matter content would be lower in plant communities in invaded areas than in non-invaded areas.This shift in trait composition would coincide with a decrease in the functional richness, dispersion, and evenness of the plant community.
Our study aims to give new insights into plant community responses to earthworm invasion by evaluating the effect of invasive earthworms on plant functional diversity, and multiple facets of biotic homogenization.
By studying the taxonomic and functional diversity responses of plant communities in a northern North American aspen forest, we disentangle the ecological mechanisms behind changes in understory plant communities and ecosystem functions linked to invasive earthworm presence.We expected invasive earthworms to lead to (1) a biotic homogenization of the plant community, with lower taxonomic and functional diversity in invaded area than in uninvaded area, and to (2) a shift in trait composition, with invaded area dominated by plant species with acquisitive strategies compared to uninvaded area.We also expect to observe (3) a change in ecosystem functions, such as litter decomposition, soil nutrient content, and soil microbial biomass and activity in the invaded area.

Effects of invasive earthworms on plant taxonomic diversity at the community and plant functional group levels
While we observed a homogeneous dispersion among areas, i.e., the average Bray-Curtis distance to the center of the invaded and control areas were similar (F = 0.19; p = 0.66), the PERMANOVA showed that plant community composition differed between the non-invaded and invaded earthworm areas (F = 4.87; p < 0.001, Figure 1A), with the earthworm invasion status explaining 11% of the variation in plant community composition (r 2 = 0.11).
In contrast, we observed changes in diversity and plant cover at the plant functional group level, with earthworms affecting mainly grasses positively and woody plants negatively (Figure 2; Tables S3 and S8).Indeed, relative cover varied across plant functional groups (F = 132.84,p < 0.001), earthworm invasion status (F = 4.34, p = 0.04), and their interaction (F = 22.20, p < 0.001), and marginally in response to canopy openness (F = 3.41, p = 0.07).We found that in the presence of earthworms the cover of woody plants was reduced by 71.5%, while that of grasses increased by 90% (Figure 2A).Thus, the relative cover of grasses and woody species, which were on average $16% and $20% cover (back-transformed estimated marginal means) in plots of the control area, respectively, shifted to $30% and $6% of cover on average per plots (back-transformed estimated marginal means), respectively, in invaded area.
Similarly, our study shows that the species richness (Figure 2B) and Shannon diversity (Figure 2C) differed between plant functional groups (F = 251.56,p < 0.001, and F = 170.15,p < 0.001 respectively), and varied across earthworm invasion status (F = 4.54, p = 0.03 and F = 2.84, p = 0.09, respectively), and their interaction (F = 10.45,p < 0.001 and F = 5.99, p < 0.001, respectively), but not in response to canopy openness (F = 0.33, p = 0.56 and F = 0.28, p = 0.60 respectively).In invaded area, species richness and Shannon diversity of woody plants declined by 49.8% and by 51.6%, respectively.While we observed an average woody plant richness of $3.2 species and a Shannon diversity of 0.92 in the control area, species richness was about $1.6 and the Shannon diversity of 0.44 in the invaded area.However, we did not observe any significant differences across plant functional groups (F = 1.71, p = 0.17), earthworm invasion status (F = 0.05, p = 0.81), their interaction (F = 1.67, p = 0.18), or canopy openness (F = 0.002, p = 0.96) for plant functional group evenness (Figure 2D).
We also found that plant community trait composition changed in the earthworm-invaded area (Figure 4).The CWM of height was $13 cm lower on average in the invaded area than in the control area (percentage of change: À29%, F = 84.6,p < 0.001, Figure 4A).Leaf trait values also shifted in response to earthworm invasion: the CWM of SLA increased by 10.3% (F = 62.69, p < 0.001, Figure 4C), and the CWM of SLA was 2.7 mm 2 mg À1 higher in the invaded area than in the control area.By contrast, the CWM of LDMC decreased by 8.2% (F = 8.96, p = 0.005, Figure 4D) in the invaded area: the CWM of LDMC was 21.7 mg g À1 lower in the invaded area than in the control area.Moreover, the CWM of leaf carbon was reduced by 2.3% (F = 40.42,p < 0.001, Figure 4B) and reached an averaged content of 45.2% of carbon in the leaves of the communities of the invaded area, while the CWM of leaf nitrogen increased by 8.9% (F = 24.94,p < 0.001, Figure 4B) to reach an average amount of 2.72% of nitrogen in leaves in the invaded area.Canopy openness increased the CWM of height (estimate = 0.23, CI 95% = [0.02,0.44], F = 4.79, p = 0.035), and decreased the CWM of SLA (estimate = À0.8,CI 95% = [-0.13,À0.03], F = 8.76, p = 0.005), but did not significantly affect the CWM of LDMC (F = 0.08, p = 0.78), leaf carbon (F = 0.73, p = 0.40), and nitrogen content (F = 0.23, p = 0.64).See also Table S2.
Moreover, the PCA reveals a shift in dominant plant strategies in response to earthworm invasion (Figure 5), from communities dominated by slow plant species with low nitrogen content, and SLA to those dominated by fast plant species with high nitrogen content and SLA, but low LDMC.

Effects of invasive earthworms on soil abiotic properties and ecosystem functions
Some soil properties and ecosystems functions like the soil water, nitrogen and carbon content, as well as microbial biomass, respiration, and activity shifted in the earthworm-invaded area, especially in the upper (0-5cm) soil layer (Figure S2; Table S11).Indeed, the pH of the soil was more acidic in the invaded than in the control area (pH of 5.9 in the invaded area vs. 6.1 in the control area; percentage of change: À3.6%, F = 4.45, p = 0.04, Figure 6A), and in the deeper soil layer than in the upper one (pH of 5.8 vs. 6.2, respectively; percentage of change: À7.1%, F = 17.81, p < 0.001).Moreover, soil contained 54.4% less carbon in the invaded area than in the control area (earthworm effect: F = 24.03,p < 0.001) and 75.4% less in the deeper soil layer than in the upper soil layer (soil depth effect: F = 75.98,p < 0.001).However, the interaction between soil depth and invasive earthworm status was not significant for these two soil parameters (pH: F = 0.73, p = 0.40, and soil carbon: F = 1.03, p = 0.32).Removing one outlier in the soil carbon data lead to a marginally significant effect of the interaction of earthworm invasion status and soil depth (F = 3.87, p = 0.06, Figure 6C; Tables S6 and S8), with a negative effect of invasive earthworms being more pronounced on soil carbon in the upper soil layer as compared to the deeper soil layer.Moreover, the soil water content and nitrogen content also decreased in invaded area (À27.3%,F = 20.08,p < 0.001, Figure 6B and À54.7%, F = 33.55,p < 0.001, Figure 6D, respectively), by soil depth (À41.8%,F = 56.42,p < 0.001, and À68.4% F = 67.11,p < 0.001, respectively), and were significantly affected by the interaction between the two factors (F = 4.39, p = 0.04 and F = 9.43, p = 0.004, respectively).More specifically, while the presence of invasive earthworms did not affect the nitrogen content in the deeper soil layer significantly, we observed a decrease by $61% in the upper soil layer in the earthworm-invaded area: the nitrogen content reached only 0.6% in the invaded area, while having an average value of 1.7% in the control area in this layer.In a similar way, the soil water content did not change in the deeper soil layer due to earthworms, but while the percentage of water was $55.6% in the control area, it decreased to 38.3% in the invaded area in the upper soil layer, which represents an approximate decrease of $31% of water content.
In addition, invasive earthworms marginally reduced total litter layer thickness by 28.3% (F = 4.12, p = 0.06, Figure 7A; Table S7): the thickness was on average $2 cm thinner in the invaded area than in the control area.This was probably largely driven by the negative effects of

Effects of invasive earthworms on plant functional group diversity and cover
Relative cover (A), richness (B) Shannon diversity (C) and evenness (D) of the different plant functional groups measured in the Barrier Lake North forest according to the earthworm invasion status (control (B) vs. invaded (d) area).Estimated marginal means and 95% confidence intervals are shown (after being backtransformed, when necessary), while data points are included in the background.The results of the post hoc tests performed when the interactions between earthworm invasion status and plant functional groups were significant in the linear models are displayed with stars that show significant differences between earthworm invasion status for the specific functional group.Significance codes: *** <0.001.See also Table S3.
invasive earthworm presence on the thickness of the L litter (À87.5%,Wilcoxon test; W = 96.5, p < 0.001, Figure 7B) and Of litter layers (À70%, Wilcoxon test; W = 89.5, p = 0.002, Figure 7C).Litter thickness in the control area was on average $1.2 cm (L layer) and $2 cm (Of layer), respectively, and were reduced to a thickness of $0.15 cm and $0.6 cm, respectively, in the invaded area.The thickness of the Oh layer did not change significantly between control and invaded areas (F = 0.27, p = 0.61, Figure 7D).Moreover, soil microbial biomass decreased significantly (À52%) in the earthworm-invaded area (F = 18.34, p < 0.001, Figure 8A; Table S6) and across soil depths (À82.4% from the top to the deeper soil layer, F = 102.51,p < 0.001), but did not change in response to the interaction between these factors (F = 1.04, p = 0.31).Soil basal respiration also decreased significantly by 33.3% in the presence of invasive earthworms (F = 30.47,p < 0.001, Figure 8B) and by 29.1% across soil depths (soil depth: F = 95.02,p < 0.001).The significant interaction effect between the presence of invasive earthworms and soil depth (F = 6.03, p = 0.02) was probably due to a stronger, negative effect of invasive earthworm presence on soil basal respiration in the upper soil layer than in the lower soil layer.Soil basal respiration was on average 28.6 mL O 2 h À1 g À1 in the control plots and reached on average 8 mL O 2 h À1 g À1 in the invaded area (À72.1%), while in the lower soil layer, it was on average 3.7 mL O 2 h À1 g À1 in the control area and reached 2 mL O 2 h À1 g À1 in the invaded area (À46.1%).The soil microbial specific respiratory quotient was significantly reduced (À16%) in the invaded area (F = 6.17, p = 0.018, Figure 8C), but this effect depended on soil depth, resulting in a significant interaction between invasive earthworms and soil depth (F = 6.45, p = 0.016).More specifically, the microbial specific respiratory quotient of the upper soil layer significantly decreased from 0.005 mL O 2 mg À1 Cmic h À1 in the control area to 0.0036 mL O 2 mg À1 Cmic h À1 in the invaded area, thus corresponding to a decrease of 30.7%, in the upper soil layer, but it did not change significantly in the lower soil layer.However, soil depth alone had no significant effect on the microbial specific respiratory quotient (F = 0.24, p = 0.63).

Invasive earthworm presence is associated with functional homogenization of the plant community
We expected a negative effect of invasive earthworm presence on taxonomic plant diversity, with a taxonomic simplification of the community in the earthworm-invaded area.This decline in taxonomic diversity would be mainly driven by a decrease in the diversity of plant functional groups, particularly woody and herbaceous species.However, counter to our expectations, our results did not show any significant decline in the taxonomic diversity in the presence of invasive earthworms nor an increase in plant taxonomic similarity, despite a change in plant community composition.While some previous studies have shown positive or no relationships between taxonomic plant diversity and invasive earthworms, 49,55 most studies showed a decrease in native plant species richness with an increasing biomass/abundance of invasive earthworms. 49,52,72Moreover, the meta-analysis by Craven et al. 38 reported an overall decline in plant Shannon diversity but also showed that nonsignificant plant taxonomic diversity responses could mask contrasting changes in native and non-native plant species. 38Similarly, we found that the absence of changes in the taxonomic diversity at the plant community level in the invaded area, does not reflect the response observed at the plant functional group level, which implies that assessing community changes in response to earthworm invasion by measuring diversity indices at the community level may mask responses of particular species or plant functional groups to earthworms.
Indeed, when we zoomed in on the responses of plant functional group diversity, we found that in the earthworm-invaded area, the cover of grasses increased, while the cover, species richness, and Shannon diversity of woody plants decreased.Similar results with changes in plant community richness due to particular species or plant functional groups have been reported in the literature. 51,52,73,74For example, Drouin et al. 52 and Alexander et al. 74 showed that grass abundance/cover was higher in the presence of a high density of invasive earthworms, while the cover of herbs from the Asteraceae or Violaceae families 74 or woody plants, 51 decreased with earthworm invasion.However, the changes we observed were not related to an increase in the dominance of a particular plant species or functional group, as evenness was not significantly affected at the community or functional group levels.Thus, our results show that invasive earthworms can have significant effects on plant community structure and functional group composition without leading to an increase in the taxonomic similarity of the community.Moreover, although we did not observe taxonomic homogenization of the community, we found a functional homogenization of the communities in the earthworm-invaded area, with lower functional diversity in invaded area than in the control area.More specifically, we observed a significant reduction in functional richness in the presence of invasive earthworms.This change could be related to the decrease in the richness of woody species, which could be due to multiple mechanisms.Invasive earthworms could either directly affect seed germination via their burrowing activities, 54 or they could indirectly affect plant establishment and development by altering the soil nutrient content of the soil, thus promoting the development/establishment of more resource-acquisitive plant species. 38,65The decrease in the richness of woody species and/or the increase in the cover of the resource acquisitive species in invaded communities would be consistent with the decrease in the functional evenness and Rao's entropy, which were also negatively affected and respectively slightly correlated to the CWM LDMC and CWM carbon in the invaded area (Table S12).The decrease in the functional evenness in the earthworm-invaded area suggests that plant species and their abundances were less regularly distributed in the functional space than in the non-invaded areas.6][77][78] This finding is further confirmed by the decrease in Rao's entropy in the presence of invasive earthworms, indicating that plant species are more functionally similar in the earthworm-invaded area than in the control area.Our results thus show an increase in the functional similarity of the plant species and, thus a functional homogenization of the plant communities in the presence of invasive earthworms.
Invasive species effects on the functional diversity of native communities and functional homogenization are still understudied, and even more so when the invader does not belong to the same trophic level.It is thus difficult to compare our results to existing literature.To our knowledge, this study is the first to investigate changes in understory plant functional diversity in the context of earthworm invasion, and the few examples of biological invasion effects on functional diversity show mixed results.For example, Allen et al. 79 found that the invasive plant species Sonchus arvensis and Cirsium arvense had a positive effect on the functional divergence and evenness of boreal wetland plant communities.In contrast, Hejda and de Bello 80 found that invasive plants decreased the functional richness and evenness of plant communities, and Chabrerie et al. 81 showed that Prunus serotina induced a decrease of Rao's quadratic entropy in the invaded understory plant communities.Moreover, Wong et al. 82 showed that the ant species (also considered as an ecosystem engineer 26 ) Solenopsis invicta, acts as an environmental filter on native ant communities of tropical grasslands, with a decrease of the functional divergence and Rao's entropy in invaded communities, while functional richness and evenness were not affected by the presence of the invasive ants.These mixed results can be due to diverse causes.4][85] It is thus necessary to further investigate the different responses of the native plant functional diversity when considering invasive earthworm abundance, community composition, or its functional diversity instead of the presence/absence of invasive earthworms only.

Invasive earthworms change plant trait composition with potential implications for ecosystem functioning
Shifts in trait composition might have contributed to the observed changes in ecosystem functions, with non-invaded areas with plant communities dominated by slow plant species (i.e., high LDMC, carbon content, and low SLA) being associated with the highest soil water and nutrient content, microbial composition and activity (Table S11; Figure S2).In the present study, changes in plant functional group diversity and composition coincided with overall changes in trait composition, with invaded area dominated by plant species with resource acquisitive strategies, i.e., traits that lead to a more efficient nutrient uptake from the soil, compared to the non-invaded area.Indeed, the presence of invasive earthworms increased the CWM of specific leaf area and leaf nitrogen content and simultaneously decreased the CWM of height, leaf dry matter content, and leaf carbon content, which could potentially have contributed to the slight decrease in the total litter layer thickness.7][88] Thus, litter decomposition might be faster following earthworm invasion, as understory plant litter will be more readily decomposed.However, the contribution of the understory litter to total litter thickness needs to be investigated, as we would expect litter thickness to be driven by tree leaves in forests, which would explain the general absence of correlations between trait CWMs and the litter thickness in control and earthworm-invaded areas (Table S13).For example, Holdsworth et al., 89 Hobbie et al., 90 as well as Co ˆte ´and Fyles 91 showed that the composition and/or the quality of the tree species litter affected the rate of litter decomposition, with a faster disappearance of litter with high calcium content, 90 and of high litter quality in the presence of invasive earthworms. 89oreover, in our study, trait CWMs were weakly correlated with soil properties and ecosystem functions (Tables S14 and S15), and, surprisingly, we observed more significant correlations between trait CWMs and ecosystem properties in the deeper soil layer (5-10 cm; Table S15) than in the upper (0-5 cm) soil layer.For example, the CWM of LDMC was negatively correlated with the pH of the upper soil layer and to the nutrient content (C and N) as well as the basal respiration of the deeper soil layer of the earthworm-invaded area, while it was positively correlated with soil nitrogen and water content, as well as basal respiration and microbial biomass of the deeper soil layer in the control area.The negative relationship between the CWM of LDMC and the nutrient and water content in soil in the earthworm-invaded area could be explained by an increase in their leaching, and/or by changes in the resource uptake strategies of the plant community due to the reduction in wood plant diversity or the increase in grass cover in the presence of earthworms.These relationships are speculative, and we acknowledge that a higher number of replicates would be required in future studies to better disentangle the links between plant community traits/diversity and the soil and ecosystem parameters via multivariate analyses, such as structural equation modeling.We thus stress the need of investigating these relationships in further studies, by taking into account, among others, tree litter quality and amount, the priming effect, as well as the plant cover and litter microclimate effects on soil microbes that could reveal the mechanisms behind shifts in ecosystem functions and properties due to invasive earthworms.
Moreover, in addition to plant community traits, a possible explanation for the differences in the ecosystem properties and functions across soil layers might be the different feeding strategies of different invasive earthworm species. 35,50Thus, earthworm community Community weighted mean (CWM) of the height, the leaf carbon and nitrogen content, the specific leaf area (SLA) and the leaf dry matter content (LDMC) of the plant community of the Barrier Lake North forest, displayed using a Principal Component Analysis (PCA) to characterize the earthworm invasion areas.CWMs were centered and scaled, and the percentage of variance explained for each axis is in brackets.Contributions of CWMs to each axis can be found in Table S10, and the fit of ecosystem functions to axes in Table S11 and Figure S2.
composition could have affected the soil microbial biomass and soil microbial activity, 35 which were reduced by invasive earthworm presence and changed across soil layers.The soil microbial carbon use efficiency increased in the upper soil layer in the presence of invasive earthworms, i.e., a decrease in the soil microbial specific respiratory quotient, while the basal soil microbial activity, i.e., soil basal respiration, was lower in the invaded area with the earthworm effect being stronger in the lower soil depth.Effects of invasive earthworms on soil microbial biomass could be mediated by their impacts on soil abiotic parameters due to the burrowing and feeding activities of earthworms, which redistribute nutrients in the soil. 33,40,92,93Here, we showed that in the earthworm-invaded area soil pH, water, carbon, and nitrogen content declined, with a stronger reduction in the upper soil layer for water and nitrogen content, which may explain the observed decrease in soil microbial biomass and activity as also was shown by the positive correlation between these parameters (Tables S13 and S14).These results are partly in line with the results of a recent meta-analysis showing that earthworm invasion overall decreases the soil water and nitrogen content. 40However, these authors also observed that invasive earthworms shifted nutrient distribution between soil layers by decreasing the nutrient content in the organic soil layer while increasing that of the mineral soil layer, with this effect depending on earthworm ecological group richness. 40Furthermore, earthworm impacts on the soil microbial community of the rhizosphere could be mediated by changes in plant root traits, e.g., fine root length/growth 32,60 or by changes in root exudates 94 due to changes in plant taxonomic and functional composition.This relationship between root exudates and/or CWMs of root traits and soil microbial community in the context of invasive earthworms could also be studied in future studies to investigate the mechanism behind soil microbial community and activity changes.

Limitations of the study
Our study suggests that invasive earthworms increase the functional similarity of the understory plant community and select for fast-growing plant species that are more efficient in nutrient uptake from the soil than slow-growing plant species.Furthermore, we show for the first time that invasive earthworms have an effect on plant functional diversity and composition, which might have cascading impacts on ecosystem functioning.However, our study represents a snapshot at a given point in time and space, which may not fully capture how the temporal and spatial dynamics of earthworm invasions affect plant communities. 85,95For a more comprehensive approach, and also to avoid any potential confounding effects between the physical distance between areas and the earthworm invasion status, we suggest further studies to cover larger areas in several forests by setting up several clusters of plots, widely separated, inside each of these areas, as for example performed in Fleri and Arcese. 96Moreover, controlled field experiments on invasive earthworm effects might be a promising approach to investigate causal relationships. 97We thus stress the need for standardized comparisons of invasive earthworm effects on plant (functional) diversity between multiple forests over time, 98 to be able to gain further insights into the mechanisms that underpin the temporal dynamics of invasive earthworm effects on plant communities and ecosystem functions.Furthermore, we found that the presence of invasive earthworms drives ecosystem functions like litter decomposition and soil microbial biomass, which could be mediated by earthworm effects on plant trait diversity and composition.These causal links need to be further investigated, and adding information on belowground plant functional traits (e.g., Freschet et al. 99 ) to aboveground plant traits studied could provide additional explanatory power.As some of our results may depend on earthworm community composition, we also recommend that further studies should investigate the taxonomic and functional composition of the invasive earthworm communities to better understand the mechanisms by which invasive earthworms affect plant communities and ecosystem functions.

STAR+METHODS
Detailed methods are provided in the online version of this paper and include the following: See also Tables S7 and S13.

Figure 1 .
Figure 1.Effects of invasive earthworms on plant taxonomic diversityPrincipal coordinate analysis showing differences in plant community composition (A) (control (yellow) vs. invaded (orange) area), with ellipses showing the 95% confidence intervals), and the effects of invasive earthworms on plant species richness (B) Shannon diversity (C), and plant evenness (D) measured in the Barrier Lake North forest (control (B) vs. invaded (d) area).Estimated marginal means and 95% confidence intervals are shown (after being back-transformed when necessary), while data points are included in the background.The p values are based on linear models, and the adjusted r 2 of the models are shown when at least one factor was (even marginally) significant.Significance codes: (*) <0.10.See also TableS2.

Figure 3 .
Figure 3. Effects of invasive earthworms on plant functional diversityEffects of invasive earthworms on the standardized functional richness (A), functional evenness (B), and Rao's entropy index (C) of the plant community of the Barrier Lake North forest according to the earthworm invasion status (control (B) vs. invaded (d) area).Estimated marginal means and 95% confidence intervals are shown, while data points are included in the background.The p values are based on linear models, and the adjusted r 2 of the models are shown when at least one factor was significant.The different letters show significant differences between earthworm invasion status.Significance codes: *** <0.001; ** <0.01; *<0.05.See also Tables S4, S9, and S12.

Figure 4 .
Figure 4. Effects of invasive earthworms on plant community traits Community weighted mean (CWM) of the height (A), the leaf carbon and nitrogen content (B), the specific leaf area (C) and the leaf dry matter content (D) of the plant community of the Barrier Lake North forest according to the earthworm invasion status (control (B) vs. invaded (d) area).Estimated marginal means and 95% confidence intervals are shown, while data points are included in the background.The p values are based on linear models, and the adjusted r 2 of the models are shown when at least one factor was significant.The different letters show significant differences between earthworm invasion status.Significance codes: *** <0.001; ** <0.01; *<0.05.See also Tables S5, S9, and S12-S15.

Figure 5 .
Figure 5. Community weighted mean traits distribution characterizing the earthworm invasion areasCommunity weighted mean (CWM) of the height, the leaf carbon and nitrogen content, the specific leaf area (SLA) and the leaf dry matter content (LDMC) of the plant community of the Barrier Lake North forest, displayed using a Principal Component Analysis (PCA) to characterize the earthworm invasion areas.CWMs were centered and scaled, and the percentage of variance explained for each axis is in brackets.Contributions of CWMs to each axis can be found in TableS10, and the fit of ecosystem functions to axes in TableS11and FigureS2.

Figure 6 .
Figure 6.Effects of invasive earthworms on soil abiotic parameters Soil pH (A), water content (B), carbon (C) and nitrogen (D) measured in the Barrier Lake North forest according to the earthworm invasion status and soil depth (0-5 cm (B) vs. 5-10 cm (d)).Estimated marginal means and 95% confidence intervals are shown (after being back-transformed when necessary), while data points are included in the background.The p values and r 2 are based on linear models.Letters correspond to the results of post hoc tests performed when the interaction between earthworm and soil depth were significant: different letters show significant differences between soil depth and earthworm invasion areas.Significance codes: *** <0.001; ** <0.01; *<0.05; (*) <0.10.See also Tables S6, S8, S14, and S15.

Figure 7 .
Figure 7. of invasive earthworms on litter Total litter thickness (A), L litter thickness (B), Of litter thickness (C) and Oh litter thickness (D) measured in the Barrier Lake North forest according to the earthworm invasion status (control (B) vs. invaded (d) area).Estimated marginal means or means (for the L and Of litter thickness) and 95% confidence intervals are shown (after being back-transformed when necessary), while data points are included in the background.The p values and r 2 are based on linear models, or on non-parametric Wilcoxon tests when necessary.The different letters show significant differences between earthworm invasion status.Significance codes: *** <0.001; ** <0.01; (*) <0.10.See also TablesS7 and S13.

Figure 8 .
Figure 8. Effects of invasive earthworms on microbial activity Soil microbial biomass (A), basal respiration (B) and microbial specific respiratory quotient (C) measured in the Barrier Lake North according to the earthworm invasion status and soil depth (0-5 cm (B) vs. 5-10 cm (d)).Estimated marginal means and 95% confidence intervals are shown (after being back-transformed when necessary), while data points are included in the background.The p values and r 2 are based on linear models.Letters correspond to the results of post hoc tests performed when the interaction between earthworm invasion status and soil depth were significant: different letters show significant differences between soil depth and earthworm invasion status.Significance codes: *** <0.001; *<0.05.See also Tables S6, S14, and S15.