Fungal root endophytes influence plants in a species‐specific manner that depends on plant's growth stage

The mycobiome (fungal microbiome) influences plants—from seed germination to full maturation. While many studies on fungal‐plant interaction studies have focused on known mutualistic and pathogenic fungi, the functional role of ubiquitous endophytic fungi remains little explored. We examined how root‐inhabiting fungi (endophytes) influence range‐expanding plant species. We isolated endophytes from three European intra‐continental range‐expanders and three congenerics that are native both in the range expander's original (southern Europe) and new (northern Europe) range. To standardize our collection, endophytes were obtained from all six plant species growing under controlled conditions in northern (new range of the range expander) and southern (native range of the range expander) soils. We cultivated, molecularly identified and tested the effects of all isolates on seed germination, and growth of seedlings and older plants. Most of the 34 isolates could not be functionally characterized based on their taxonomic identity and literature information on functions. Endophytes affected plant growth in a plant species–endophyte‐specific manner, but overall differed between range‐expanders and natives. While endophytes reduced germination and growth of range‐expanders compared to natives, they reduced seedling growth of natives more than of range‐expanders. Synthesis. We conclude that endophytic fungi have a direct effect on plant growth in a plant growth stage‐dependent manner. While these effects differed between range expanders and natives, the effect strength and significance varied among the plant genera included in the present study. Nevertheless, endophytes likely influence the establishment of newly arriving plants and influence vegetation dynamics.


| INTRODUC TI ON
Many studies have examined the ecology of alien plant species that have been introduced from other continents, as a small proportion of them is highly invasive, changing local biodiversity and ecosystem functioning, and causing significant economic costs (Pimentel et al., 2001;Vilà et al., 2015;Vitousek et al., 1996). However, ecological consequences of climate warming-induced range-expanding plant species have received relatively little attention (De Frenne et al., 2014;van der Putten, 2012), despite the increasing incidents of intra-continental migration of native plant species to higher altitude and latitude due to climate change (Alexander et al., 2015;van der Putten, 2012).
As the growth of almost plants is influenced by the below-ground microbiome (Edwards et al., 2018;Mendes et al., 2013;Shi et al., 2016), micro-organisms may influence the success of plant range expansion as well. Especially those micro-organisms that directly penetrate plant roots have well-acknowledged positive (e.g. mycorrhizal fungi, rhizobial bacteria) and negative (e.g. bacterial, fungal and fungal-like oomycete pathogens) effects on plant growth (Raaijmakers et al., 2009).
Indeed, mutualistic micro-organisms are able to promote the success of invasive alien plants (Bever, 2002;Pringle et al., 2008) by enhancing plant abundance. Invasive alien plants even may change community composition of fungal mutualists to their own favour (Mummey & Rillig, 2006;Stinson et al., 2006;Zhang et al., 2010). In contrast to mutualists, many pathogens are limited in their host range (Gilbert & Webb, 2007) and often have a limited biogeographical distribution (Rout & Callaway, 2012). This has likely consequences for plant range shifts, as native plants can become released from their specialized pathogens, a hypothesis predicted by the enemy release hypothesis for introduced alien plants ( Keane & Crawley, 2002;Mitchell & Power, 2003;Reinhart et al., 2003;Wolfe, 2002). However, pathogen effects may depend on plant growth stage, with mature plants generally suffering less from pathogens than seeds or seedlings, with profound importance especially on short-lived herbaceous plant species (Bagchi et al., 2014;Blaney & Kotanen, 2001;Gilbert, 2002;Jarosz & Davelos, 1995;Mordecai, 2015;Packer & Clay, 2000). Similar to invaders from inter-continental origin, plants that expand their range within a continent may be less exposed to negative feedback from their soil microbiome, and host lower numbers of enemies in the new range (Dostálek et al., 2016;Morriën et al., 2010;van Grunsven et al., 2007van Grunsven et al., , 2010. While experiments to decipher interactions between range-shifting plant species and soil microbiomes have focused on overall plant-soil feedback patterns (Alexander et al., 2015;De Frenne et al., 2014;Dostálek et al., 2016;Engelkes et al., 2008;van Grunsven et al., 2007), the roles of specific pathogens and mutualists have received less attention.
Yet, under some conditions endophytes can shift, from plantneutral to mutualistic (Arnold et al., 2003;Clay & Holah, 1999;Clay & Schardl, 2002) or pathogenic (Busby et al., 2016;Hyde & Soytong, 2008;Kia et al., 2017;Rodriguez et al., 2009). Translating this ability to the symbiosis with invasive plants, endophytes may promote the spread of invasive plant species (Molina-Montenegro et al., 2015;Shearin et al., 2018). While endophytes have been studied in the context of inter-continental plant invasions (Klironomos, 2002;Knapp et al., 2012), only one study considered endophytes in the context of climate warming-induced range-expanding plant species. This study showed that range-expanders host different endophyte communities in the new than in the native range, while endophyte communities in congeneric natives did not differ between both ranges (Geisen et al., 2017). Yet, the functional role of endophytes in plant growth during this type of range shift remains unknown (Busby et al., 2016).
In order to study how endophytes may influence range-shifting plant species during various stages of their life history, we cultivated fungal and oomycete endophytes from roots of three range-expanding plant species and three congeneric natives. All plants were grown in northern soils from their new range (the Netherlands) and southern soil from their native range (Slovenia). We taxonomically identified all endophytes by sequencing the ITS region. Finally, we tested the effects of all cultivated root endophytes on the germination rate, seedling and plant growth of all six plant species. We tested the following general hypotheses: (1) Range-expanders will have higher germination rates and produce more biomass when inoculated with endophytes isolated from the expanded northern range soil than from the native southern range soil; we did not expect differences for native plant species, as endophytes with negative effects might not have expanded together with the range-expanders but have developed with natives. (2) The effects of endophytes are irrespective of plant growth stages, because we expected a functional conservatism of endophytes.

| Plant species
Two plant species in each of the three genera Centaurea (C. stoebe and C. jacea; family Asteraceae), Geranium (G. pyrenaicum and G. molle, family Geraniaceae) and Tragopogon (T. dubius and T. pratensis, family Asteraceae) were selected for this study. The first of each plant species (C. stoebe, G. pyrenaicum and T. dubius) in the genera originated from south-eastern Europe and expanded its range northwards in the 20th century (named: range-expanders, tRE, Sparrius, 2014).
The second of each plant species in the genera is native in Europe in both the original and the expanded new range (named: natives, tNA).
More details are found in Table 1. All six herbaceous plant species co-occur in riverine habitats along the river Waal in the Netherlands, which is the southernmost branch of the river Rhine. Seeds of most plant species were obtained from a seed supplier (Cruydt-Hoeck) that collects seeds from wild plant populations, with the exception of G. molle that we collected ourselves (Table 1).

| Endophyte culturing experiment
To culture root-inhabiting fungi and fungal-like oomycetes (in the present study collectively named 'endophytes'), we performed a greenhouse experiment using soil collected from three independent sites in Slovenia (southern soil) and three independent sites in the Netherlands (northern soil), where all six plant species commonly occur. We decided to isolate endophytes from roots of greenhousegrown plants rather than from the field to ensure that root endophytes were collected from all soils under the same environmental circumstances. Soil was collected from the top 3-15 cm of two sublocations in each site in the Netherlands and Slovenia. Soils from the two sublocations from each site were homogenized and sieved using a 4-mm mesh size to create three independent soil samples from both Slovenia and the Netherlands. Ten per cent of the resulting six independent soil mixes was stored in the dark at 4°C until further use. The remaining soil from the Dutch sites (Nl1-3) was combined in equal parts (1:1:1), mixed with sand (2:1) and was gamma-sterilized (20 kGy; Syngenta bv) to be used as sterile background soil. Further details on soil sampling and soil properties are described in the study by Koorem et al. (2018). The procedure of inoculating 10% of alive to 90% background soils was done to ensure that abiotic differences among soil samples were reduced to a minimum (Wilschut et al., 2019;Zhang et al., 2016).
Prior to germination, all plant seeds were surface-sterilized by washing seeds in 0.4% sodium hypochlorite solution for 3 min followed by rinsing with sterile distilled water (H 2 O dest ). Seeds of all plant species were germinated on sterile glass beads in a growth cabinet at 20/10°C (day/night temperature); 16 hr light/8 hr dark conditions. Seedlings were planted individually in 0.8 L pots containing 675 g of the sterilized sandy loam soil supplemented with 75 g of one of the live soil samples. This resulted in a total of 36 pots (2 geographic ranges × 3 sites as true soil replicates from each range × 6 plant species). Individual pots were weighed two times per week and watered with sterile H 2 O dest to a weight of 750 g, which corresponded to a dry weight-based moisture content of ~60%. The pots were placed on a cart in a greenhouse at 16 hr light, 8 hr dark; 20°C, 15°C and 60% relative humidity. The position of the pots on the carts was randomized weekly. To increase the diversity of potentially plant life stage-dependent endophytes, another seedling of the same species and the same age was added to the same pots 2 and 4 weeks after initiating the experiment. This resulted in three plants of different age per pot.
Six weeks after initiating the experiment, shoots of all three plants per pot were cut, combined and dried at 60°C for 3 days before determining the dry weight. Roots were carefully isolated from the soil and thoroughly washed under running water before cutting into pieces of ~0.5-cm lengths. Fifty randomly selected root pieces per plant individual were placed in 2-ml centrifuge tubes filled with sterile water and stored at 4°C for 1-2 days before endophyte isolation (see below), while the remaining roots were combined and dried at 60°C for 3 days before determining the dry weight.

| Isolation and molecular characterization of endophytes
The root pieces stored in the centrifuge tubes were thoroughly washed by transferring three times to new sterile demineralized H 2 O (H 2 O dest ) in order to minimize the number of root-attached spores.
Subsequently, roots were transferred to centrifuge tubes filled with 70% ethanol and incubated for 7 min under occasional mixing before final transfer and washing in a centrifuge tube containing sterile H 2 O dest . Root pieces were placed on sterile tissue paper to dry the surface under sterile conditions in a flow cabinet. Three individual root pieces per plant and pot were placed apart from each other in ten 10-cm Petri dish filled water agar (WA; 1.6% agar, pH 6.7, ampicillin 50 mg/L), resulting in a total of 30 root pieces per pot and a total of 360 Petri dishes. Plates were stored at 20°C in the dark.
Remaining roots per pot were combined and divided into three parts. Each part was placed on a 6-cm diameter Petri dish filled with a 1:1 mix of sterile pond water and sterile H 2 O dest containing dry and sterile grass leaves (Agrostis capillaris, 2-3 cm) for baiting zoospore forming oomycetes (Pettitt et al., 2002 However, 13% of those cultures with high similarity to sequences present in GenBank only matched with sequences not affiliated to described fungal or oomycete species but with unknown environmental sequences. For those isolates, described taxa showed a sequence similarity ≤96%. This suggests that more than 20% of all of our cultures represent species or even genera that are currently unknown or represent species with so far missing sequences in databases.
We further did an a priori functional investigation based on sequence match with the best BLASTn hits. This analysis was meant to evaluate if the endophytes were known pathogens or mutual- ists. We obtained this information using the best BLASTn hits and literature search on functioning of that or related species and genera. This allowed us to identify most cultures (27) as endophytes that are likely not plant pathogenic. From the remaining cultures, two most closely resembled pathogens, two potential mammal pathogens and one a saprophyte, while two could not be assigned reliably a function as they were phylogenetically too divergent from known taxa.

| Seed germination
Spores and hyphae of endophytes were extracted from well-grown cultures on malt extract agar by adding sterile H 2 O dest and carefully suspending endophyte material using a cell scraper. Spore suspensions were equilibrated to 1 × 10 6 spores or hyphal pieces per ml as counted under an inverted microscope at 400× magnification. Seeds of all targeted plants were surface-sterilized immediately before use by washing seeds in 10% sodium hypochlorite solution for 3 min (Centaurea and Geranium spp.). We applied this much harsher procedure than the one described above to ensure a complete elimination of all non-endophytic micro-organisms. Yet, tests revealed that this sterilization procedure was not eliminating all seed-attached fungi from Tragopogon species, so we used a harsher sterilization by pre-treating those seeds in 3% HCl for 1 min before washing seeds in 10% sodium hypochlorite solution for 5 min. Surface-sterilized seeds were rinsed with sterile H 2 O dest .
The germination experiment was conducted in 10-cm diame-

| Data analysis
All statistical analyses were carried out in R statistical language, version 3.6.1 (R Core Team, 2019).

| Endophyte culturing experiment
The effects of soil origin (south or north) and plant status (range expander: tRE or native: tNA), on plant biomass were analysed using ANOVA with plant genus (Centaurea, Geranium and Tragopogon), soil origin (north or south), status of the tested plant (tRE or tNA) and all their interactions as fixed factors.

| Inoculation experiments
The percentage of germinated seeds was calculated as p = 100 × (number of germinated seeds + 0.5)/(total number of seeds + 1) to guard against 0 or 100% values. We corrected the obtained values of percentage of germination or plant biomass using the mean values of the corresponding controls. We call this corrected values 'ratio to control'. For example, for the % seed germination, the ratio was calculated as % germination with endophyte/mean % germination in control treatment. If the resulting value is below 1, the endophytes had a negative effect; a value above 1 indicates a positive effect of the endophyte; and a value of 1 indicates that there was no effect of endophytes. As each pair of species was inoculated with a different set of endophytes, the effects of endophytes for each pair of congeners were analysed separately. The ratios were analysed using mixed-effects models. Prior to analysis, the ratios were log-transformed as ln(ratio + 0.5) to meet the assumptions of homogeneity of vari- The assumption of the homogeneity of variance and normality were checked graphically by inspecting the residuals plotted against fitted values, and against each explanatory variable in or outside the model. The Cook's distance values (Cook & Weisberg, 1982) were used to detect any influential observations. Two observations from two different datasets had much smaller residuals compared to the rest of the data points. After inspecting the raw data and confirming that these data points were largely differing from the rest in the same group, the two points were deleted from the analyses.
The mixed models were fitted using lmer function (r lme4 package; (Bates et al., 2015)

| Endophyte culturing experiment
In total, 34 unique fungal and oomycete endophyte cultures were obtained from roots of all six plant species grown in both northern and southern soils ( Table 2; Table S1). Plants grown in northern soil produced on average more total biomass than in southern soil (F 1,24 = 4.45, p = 0.045; Figure S1a).

| Germination
Seed germination of range-expanding Geranium was on average more negatively affected by the inoculated endophytes than its native congener (F 1,36 = 27.94; p < 0.001; Figure 2A). Seed germination of range-expanding and native Centaurea species was not affected by the endophytes (p > 0.05 in all cases, Table S2). Seed germination of range-expanding Tragopogon species was on average more negatively affected by the endophytes than its native congener, but the latter effect depended on whether the endophyte originated from the northern or southern soil ( Figure 2B). Specifically, seed germination of the native Tragopogon was positively affected by endophytes F I G U R E 1 Scheme illustrating the experimental set-up. We grew three pairs of native and range-expanding plant species in soils from the range-expanders' origin (South Europe) and the expanded range (Central Europe). We then cultivated root endophytes by placing root pieces on fungal-and oomycete-specific media. We obtained 34 unique cultures that we molecularly identified and used for functional experiments to test their effect on seed germination, seedling growth and plant growth [Colour figure can be viewed at wileyonlinelibrary.com] retrieved from the northern soil when compared to the southern soil (F 1,54 = 6.53; p = 0.014; Figure 2B).

| Seedling biomass
Range-expanding G. pyrenaicum seedlings produced more root biomass when exposed to the endophytes than the native G. molle, but this effect depended on the endophyte's soil origin (F 1,32 = 4.17; p = 0.049, Figure 3A). Endophytes originating from southern soil promoted the root biomass of the range-expanding G. pyrenaicum seedlings more than the endophytes from the northern soil, whereas native G. molle root biomass was not affected by origin of the endophyte ( Figure 3A).
A similar interactive effect of endophyte soil origin and status of the tested plant was found for the shoot biomass of Geranium species, but this effect was marginally significant (F 1,32 = 4.04; p = 0.053, F I G U R E 2 Ratio of percentage seed germination inoculated with endophytes to the averaged seed germination in control treatment.
Means of the ratios ± standard errors are shown. Difference between native (tNA) and range-expanding (tRE) Geranium (A), Centaurea (B) and Tragopogon (C) species exposed to the endophytes originating in northern (eNorth) and southern soils (eSouth). Black solid points depicture average responses of seeds exposed to the same fungal culture. The number of endophyte cultures is given in Table S1. Different letters indicate significant differences based on linear model analysis F I G U R E 3 Ratio of biomass of the seedlings inoculated with endophytes to the averaged biomass in control treatment. Means of the ratios ± standard errors are shown. Difference in seedlings root (top A-C) and shoot (bottom D-F) biomass between native (tNA) and range-expanding (tRE) Geranium (left A and D), Centaurea (middle B and E) and Tragopogon (right C and F) species exposed to endophytes originating in northern (eNorth) and southern soils (eSouth). Black solid points depicture average responses of seedlings exposed to the same fungal culture. The number of endophyte cultures is given in Table S1. Black-dashed line indicates no effect of endophytes. Different letters indicate significant differences based on linear model analysis F I G U R E 4 Ratio of root biomass of the plants inoculated with endophytes to the averaged root biomass in control treatment. Means of the ratios ± standard errors are shown. Top (A-C): Difference in plant root biomass of native (tNA) and range-expanding (tRE) Geranium (left A and D), Centaurea (middle B and E) and Tragopogon (right C and F) species exposed to endophytes originating in northern (eNorth) and southern soils (eSouth). Bottom (D-F): Difference in root biomass based on whether endophytes were isolated from native (eNA) or range-expanding (eRE) plant species. Black solid points depicture average responses of plants exposed to the same fungal culture. The number of endophyte cultures is given in Table S1. Black-dashed line indicates no effect of endophytes. Different letters indicate significant differences based on linear model analysis F I G U R E 5 Ratio of shoot biomass of the plants inoculated with endophytes to the averaged shoot biomass in control treatment. Means of the ratios ± standard errors are shown. Top (A-C): Difference in plant shoot biomass of native (tNA) and range-expanding (tRE) Geranium (left A and D), Centaurea (middle B and E) and Tragopogon (right C and F) species exposed to endophytes originating in northern (eNorth) and southern soils (eSouth). Bottom (D-F): Difference in shoot biomass based on whether endophytes were isolated from native (eNA) or range-expanding (eRE) plant species. Black solid points depicture average responses of plants exposed to the same fungal culture. The number of endophyte cultures is given in Table S1. Black-dashed line indicates no effect of endophytes. Different letters indicate significant differences based on linear model analysis Figure 3B). The root biomass of Centaurea seedlings was not affected by inoculated endophytes (p > 0.05 in all cases, Table S2), whereas shoot biomass of native C. jacea was negatively affected by the endophytes compared to the range-expanding congener (F 1,12 = 68.55; p < 0.001, Figure 3C). The root biomass of native T. pratensis was negatively affected by the endophytes compared to the range-expanding congener (F 1,12 = 18.05; p = 0.001, Figure 3D), whereas the shoot biomass of Tragopogon species was not affected by the endophytes (p > 0.05 in all cases, Table S2). Overall, we conclude that endophytes had variable effects on seedling growth that depended on the source of the endophyte and plan genus studied.

| Plant biomass
The root biomass of Geranium species was affected by the endophyte plant origin (F 1,35 = 4.72; p = 0.037; Figure 4). Here, endophytes isolated from native Geranium molle plants more negatively affected root biomass of Geranium species than those isolated from range-expanding G.
pyrenaicum. In contrast, the shoot biomass of range-expanding Geranium pyrenaicum was more negatively affected by the cultivated endophytes than the native congeners (F 1,246 = 10.22; p = 0.0016; Figure 5). In particular, the root biomass of Geranium plants (both Geranium species combined) was more negatively affected by endophytes that were iso-  Table S2). The root biomass of the range-expanding Tragopogon species was positively affected by endophytes retrieved from northern soil compared to endophytes from the southern soil, whereas root biomass of the native Tragopogon was not affected by origin of the endophytes (F 1,104 = 5.56; p = 0.02; Figure 4). The shoot biomass of range-expanding and native Tragopogon species was not affected by the endophytes (p > 0.05 in all cases, Figure 5; Table S2).

| D ISCUSS I ON
Here we show that effects on plants of fungal and oomycete root endophytes isolated from range-expanding and congeneric native plant species are variable depending on plant life stage.

| Endophytes with unknown functions dominate root cultures
We found a wide taxonomic diversity of endophytes that include fungal and oomycete species in roots of the six plant species. Many species of oomycetes are known to be notorious plant pathogens of numerous plant species and are commonly found in soils and the plants rhizosphere (Arcate et al., 2006;Geisen et al., 2015). confirm this pattern as this assumption is based on the comparably low number of 34 cultures and assumes an equal cultivability which selects against obligate rather than facultative plant pathogens. Interestingly, those presumable pathogens based on a priori assignments often did not negatively affect plant growth. This is partly due to the fact that there is a lack of functional knowledge of endophytes in non-crop plant species suggesting that a priori functional assignment in the little studied endophytes can be misleading especially in natural plant species (Lofgren et al., 2018;Malcolm et al., 2013). A limited cultivation efficiency of endophytes prevented a thorough investigation of differences in endophyte infection between plant species or locations-a pattern that exists for some plant species Glynou et al., 2016Glynou et al., , 2017). Yet, we believe that our rigorous cultivation approach resulting in few cultivated endophyte species is biased in the same way across treatments. Thus, the functional tests, as discussed next, are reliable but might miss some patterns that exist along the ones tested here.

| Endophytes differentially affect rangeexpanding and native plant species in a life stagedependent manner
Range-expanding and congeneric native plant species differed in their responses to endophytes, but the responses strongly depended on plant growth stages. This result opposes Hypothesis 2 (endophytes affect plants in the same way throughout plant growth).
The observed patterns were also more complex than assumed in Hypothesis 1 (Range-expanders perform best with endophytes from northern soils while natives do not show such a pattern), as there were no differences between range-expanders and congeneric natives in their responses to cultures obtained from northern than southern soils. All observed differences depended on plant genus, as well as on soil and plant origin of the endophytes.
The effect of endophytes on plant growth was the most pronounced in the earliest life stages: seed germination and seedling growth. These results support the idea that the impact of plant hostassociated organisms depends on plant life stages, being most strong in early plant growth stages (Bagchi et al., 2014;Blaney & Kotanen, 2001;Gilbert, 2002;Jarosz & Davelos, 1995;Mordecai, 2015;Packer & Clay, 2000). Nevertheless, in many experimental studies plant performance is measured as an integration of growth from seedlings to mature plants. Indeed, differences in germination success might explain the success of invasive exotic plant species when they germinate faster than related native plant species (Hirsch et al., 2012).
Our experimental test revealed that endophytes may play a key role in affecting plants especially during early growth stages such as during germination and seedling establishment and that these effects differ between native and range-expanding plant species. However, the effects of the endophytes on plant growth in this study are not caused by known plant pathogens, but by endophytes with an unknown a priori functioning (Arnold et al., 2000;Rodriguez et al., 2009).
Functions of most of these endophytes remain largely unknown and previous studies showed contradictory effects on plant growth ranging from positive (Newsham, 2011) to negative Mayerhofer et al., 2012). Our results support those observations using endophyte culture-dependent analyses. The positive effects of endophytes on plant growth might be attributed to stimulated nutrient exchange comparable to that of mycorrhizal fungi (Arnold et al., 2003;Rodriguez et al., 2009). Positive endophyte effects, however, have been suggested to occur predominantly under stress, for example through secondary metabolite production that confers resistance against herbivores (Cosme et al., 2016), while negative effects can be more common under non-stress conditions . Our results suggest that both positive and negative effects of endophytes on plant growth are common and depend on the partners that were interacting. As endophytes influenced plant growth during germination and seedling establishment, we conclude that endophytes are not functionally neutral, but affect plant species growth at various growth stages. Furthermore, the plant species-specific effects ranging from positive to negative may provide a mechanism of endophytes influencing plant community dynamics as shown as well for plant pathogens (Benítez et al., 2013;Sarmiento et al., 2017

| Implications for plant range shifts and outlook
Overall, using diverse cultures of root-inhabiting fungi and oomycetes, we provide evidence that root endophytes can affect plant growth, with potential consequences for the success of some range-expanding plant species. However, a limitation of the present study is the relatively low number of pairs tested. In order to determine the generality of these results, further studies are needed with additional plant species and from a variety of ecosystems. Furthermore, despite the notion that endophytes might hardly be affected by variation among plant genotypes, future studies should include plant seeds originating from a variation of populations rather than from one as was done in the present study.
Another important challenge will also be to perform studies under semi-natural and natural conditions, including the stress conditions that may occur in the field. Putten, 2014). Unlike Klironomos (2002), who showed that root-associated fungi affected seedlings of exotic invasive plant species less negatively than seedlings of rare native plants, we show that this trend is less pronounced among range-expanding plant species and closely related congeneric natives. Furthermore, we show that resulting effects of endophyte-plant interactions depend on plant growth stage highlighting the importance of the timing of this interaction in determining plant growth (Sikes et al., 2016). Studies that determine plant growth after a single inoculation or single sampling point might therefore uncover only part of the ecologically relevant interactions between plants and (soil) biota. However, it has to be noted that we conducted the seed germination and seedling growth analyses under artificial conditions on agar plates, while plant growth was determined in sterilized soils. Therefore, patterns observed might deviate under more complex conditions in soils.
Nevertheless, we aimed at uncovering potential interactions and therefore believe that our approach provides a valid model system for this purpose (Crowther et al., 2018). Yet, the generality of our findings that individual endophytes have differential effects on range-expanders and related natives may require further testing with other plant species and environmental conditions before conclusions may be generalized.
In conclusion, our findings suggest that a wide range of endophytes can directly impact plant growth, in addition to the known, often positive effects of AMF and negative effects of pathogenic fungi. Particularly, the impact on seed germination and seedling establishment was profound and may need to be investigated as well in agricultural settings, both to improve plant growth directly, as well as to enhance plant growth under environmental stress from drought, pathogens and other factors. While the present study shows the potential for an increased availability of biotic resources for bioengineering purposes, further steps are needed in order to explore the full consequences and possible solutions offered by fungal or fungal-like endophytes.

ACK N OWLED G EM ENTS
This work was conducted with support from ERC-ADV 323020

PEER R E V I E W
The peer review history for this article is available at https://publo ns.