Independent effects of arbuscular mycorrhiza and earthworms on plant diversity and newcomer plant establishment

Authors


Abstract

Questions: How do arbuscular mycorrhiza and earthworms affect the structure and diversity of a ruderal plant community? Is the establishment success of newcomer plants enhanced by these soil organisms and their interactions?

Methods: We grew a native ruderal plant community composed of different functional groups (grasses, legumes and forbs) in the presence and absence of arbuscular mycorrhizal fungi (AMF) and endogeic earthworms in mesocosms. We introduced seeds of five, mainly exotic, plant species from the same functional groups after a disturbance simulating mowing. The effects of the soil organisms on the native ruderal plant community and seedling establishment of the newcomer plants were assessed.

Results: After disturbance, the total above-ground regrowth of the native plant community was not affected by the soil organisms. However, AMF increased plant diversity and shoot biomass of forbs, but decreased shoot biomass of grasses of the native plant community. Earthworms led to a reduction in total root biomass. Establishment of the introduced newcomer plants increased in the presence of AMF and earthworms. Especially, seedling establishment of the introduced non-native legume Lupinus polyphyllus and the native forb Plantago lanceolata was promoted in the presence of AMF and earthworms, respectively. The endogeic earthworms gained more weight in the presence of AMF and led to increased extraradical AMF hyphal length in soil. However, earthworms did not seem to modify the effect of AMF on the plant community.

Conclusion: The present study shows the importance of mutualistic soil organisms in mediating the establishment success of newcomer plants in a native plant community. Mutualistic soil organisms lead to changes in the structure and diversity of the native plant community and might promote newcomer plants, including exotic species.

Introduction

Soil organisms have major impacts on ecosystem processes (Wardle et al. 2004; van der Heijden et al. 2008) and might play a crucial role in community assembly and in invasion processes of plants (e.g. van der Heijden 2004; Wolfe & Klironomos 2005; Reinhart & Callaway 2006; Pringle et al. 2009). Positive interactions between plants and soil organisms are widespread, including enhancement of nutrient uptake and systemic defences. In general, such positive interactions might play a greater role in ecology than acknowledged so far (Bruno et al. 2003). Soil organisms such as arbuscular mycorrhizal fungi (AMF) and earthworms are known to benefit plant growth, but their effects vary by plant species. Plant species differ in their dependency on AMF (Klironomos 2003) and both subordinate and dominant plant species can be promoted by AMF (Urcelay & Diaz 2003). The presence and diversity of AMF can impact plant diversity in European grasslands, mainly by promoting forb growth (Grime et al. 1987; van der Heijden et al. 1998a, b). By increasing the nutrient availability for plants, earthworms often promote plants that are good below-ground competitors for soil nitrogen, such as grass species (Wurst et al. 2003, 2005; Eisenhauer & Scheu 2008a), and can thus influence the structure of grassland plant communities (Partsch et al. 2006). Combined effects of AMF and earthworms might differ from their individual effects on plants and plant communities. However, previous studies on interactions between AMF and earthworms on individual plants did not find strong interactive effects on plant performance (Tuffen et al. 2002; Wurst et al. 2004; Eisenhauer et al. 2009).

For plant species, the successful establishment as newcomers in a plant community is a crucial first step for their invasion, i.e. the ability to increase when rare (Mc Crawley 1997). The success of a newcomer plant may depend on its specific relationship to mutualistic soil organisms. Studies on the effects of AMF on the individual performance of exotic invasive plant species (Philip et al. 2001; Fumanal et al. 2006; Harner et al. 2010) and their competition with a native plant species (Marler et al. 1999; Callaway et al. 2001; Bray et al. 2003) suggest that the presence of AMF can enhance the competitive strength of exotic plants (Pringle et al. 2009). However, as far as we are aware, only one study (Stampe & Daehler 2003) investigated the effect of AMF on the invasion success of an exotic plant species into a native plant community. In that study, certain AMF species increased both the diversity of the native plant community and the invasion success of the exotic forb Bidens pilosa and decreased the dominance of the native grass Melinis repens. Focussing on native newcomers, van der Heijden (2004) reported positive effects of AMF on the biomass and phosphorus acquisition of native seedlings after 9 months of growth in a calcareous grassland community in Switzerland that contained the same plant species as adults. For earthworms, observational and experimental data suggest that European earthworms invading boreal forests in North America promote the invasion of the exotic shrub Rhamnus cathartica (Frelich et al. 2006). In contrast, negative effects of the native earthworm Lumbricus terrestris on seedling establishment of native grassland plant species were reported (Eisenhauer & Scheu 2008b).

Until the present, however, no study has looked at the individual and combined effects of native AMF and earthworms on the establishment of newcomer plants in a native plant community. Here we investigated whether AMF and earthworms (alone and in combination) promote the establishment of newcomer plant species, including exotic and potentially invasive plant species. A ruderal plant community consisting of seven native species from different functional groups (grasses, forbs, legumes) was established in mesocosms in the greenhouse. We used a ruderal plant community because the establishment of exotic plant species often starts at disturbed roadsides (Hansen & Clevenger 2005). After the establishment of the ruderal plant community and a disturbance (simulated mowing treatment), seeds of mainly exotic species were introduced. We hypothesized that (1) AMF and earthworms enhance newcomer establishment, and that (2) the introduced plant species differ in their response to the soil organisms. We expected that (a) the tested AMF-dependent plant species are promoted by AMF presence; (b) the tested grass species is promoted by earthworm presence, since grasses often react positively to the presence of earthworms; and (c) the tested legume is not affected by earthworm presence, since the symbiosis with N2-fixing bacteria provides it with nitrogen. Further, (d) we hypothesized that we would find no strong interactive effects of AMF and earthworms on newcomer establishment, since their effects on plant performance were reported to be mainly independent (Tuffen et al. 2002; Wurst et al. 2004; Eisenhauer et al. 2009; Wurst & Rillig 2011).

Methods

We established a fully factorial experiment with the factors AMF (presence and absence) and earthworms (presence and absence), with ten replicates for each of the four treatment combinations. For this, a total of 40 clay pots (height=28 cm; diameter: top=32 cm, bottom=22 cm) were set up in a greenhouse (temperature 18/25 °C (day/night); ambient light conditions). The hole in the bottom of the pots was closed with a nylon mesh (glued in place with silicon) to prevent earthworm escape. The pots were filled with 6 kg of steamed (90 °C; overnight), sieved, sandy, loamy disturbed soil (N=0.25%, C=4.06%, C/N=16.24) from the Botanical Garden in Berlin/Dahlem, Germany. The applied soil heating procedure kills all AMF propagules (Endlweber & Scheu 2006). Four hundred grams of steamed soil were mixed with 100 g of AMF inoculum and added to half of the pots (N=20). The AMF inoculum consisted of a mixture of Glomus etunicatum (Becker & Gerdemann), Glomus intraradices (Schenck & Smith) and Glomus claroideum (Schenk & Smith) hyphae and spores in a peat substrate and was provided by INOQ GmbH (Schnega, Germany). The control pots (N=20) received 400 g steamed soil mixed with 100 g autoclaved (121 °C; 1 h) AMF inoculum. To correct for differences in microbial background communities introduced with the AMF inoculum and to provide the pots with naturally occurring microorganisms, a soil solution was prepared from 400 g unsterilized experimental soil mixed with 100 g AMF inoculum and 5 l demineralized water. After thoroughly mixing, the coarse particles were let to sink for 4 h, then the solution was sieved through 500-, 212-, 53- and 20-μm sieves to filter out AMF hyphae and spores. One hundred milliliters of the 20-μm filtrate + 200 ml tap water were added to every experimental pot.

Seedlings of seven common plant species co-existing in ruderal Tanaceto-Artemisietum societies in Central Europe (Dierßen 1996) were planted into the experimental pots according to a planting scheme with altering positions for the individual plant species between the replicates. Seeds of the grasses Arrhenatherum elatius (L.) J. & K. Presl, Agrostis capillaris L., Festuca rubra L. and Poa pratensis L., the forbs Tanacetum vulgare L. and Hypericum perforatum L. and the legume Trifolium repens L. (all provided by Conrad Appel, Darmstadt, Germany) were sown on glass beads and transplanted into the experimental pots after 2 to 3 weeks, when they started to develop the first leaves (after the cotyledons). Seedlings that did not establish after transplanting were exchanged with seedlings of the same age until 3 weeks after the transplantation; thereafter, no further plants were exchanged (day 1 of the experiment). Each pot contained one seedling per plant species, resulting in a seedling density of ca.150 seedlings m−2.

In week 3 of the experiment three individuals (juveniles and adults) of endogeic earthworms (Aporrectodea caliginosa Savigny) (total biomass=1.57 g, SE=0.05) were added to half of the AMF and the non-AMF pots. For the juveniles, species determination was not 100% certain, but given the high abundance of A. caliginosa at the collection site (Berlin/Dahlem, Germany), they very likely were of the same species. The earthworms were collected by hand-sorting in a meadow at Berlin/Dahlem and kept for approximately 2 weeks at 15 °C in sterile experimental soil, which was renewed every 5 days to minimize the risk of contamination with AMF from the field.

In week 11, the above-ground parts of the plants were cut to 3 cm above the ground to simulate a disturbance by mowing, which is a common management practice on ruderal roadsides. The next day, 10 seeds per species of the newcomer plants were added as an even seed shower (total: 50 seeds) to each experimental pot. The plant species belonged to different functional groups and consisted of the AMF-dependent forbs Plantago lanceolata L. (the only species native to Central Europe) and Solidago canadensis L. (native to North America), the non-AMF forb Bunias orientalis L. (native to Southeast Europe), the legume Lupinus polyphyllus L. (native to North America), and the grass Eragrostis minor Host (native to Southern Europe and Africa). The seeds were provided from Appels Wilde Samen (Conrad Appel), except for seeds of E. minor and L. polyphyllus, which were provided by the Botanical Garden (Berlin, Dahlem) and BohemiaSeed (Praha, Czech Republic), respectively.

In week 13, 15 and 18, the number of germinated seedlings of the five added plant species was counted. The pots were watered regularly with the same amount of tap water per pot. In week 19, the experiment was harvested. The above-ground parts of the species from the native plant community were cut at ground level and dried in an oven at 70 °C for at least 72 h. The newcomer plant seedlings were harvested by carefully excavating the whole plants; whole plants (shoots and roots) were then dried in an oven at 70 °C for at least 72 h. Two days after above-ground harvesting, the below-ground harvest of the mesocosms was performed. Roots and earthworms were collected from the soil. Earthworm numbers and fresh weights were recorded. Subsamples of the soil (approximately 150 g fresh weight) were taken and dried in an oven at 70 °C for further analyses, such as measurement of fungal hyphal lengths. Roots were collected by hand, washed free of soil and dried in an oven at 70 °C until constant weight.

Mycorrhizal colonization of the roots and soil

Root subsamples collected from the total root biomass of the resident species were placed in 10 ml 10% KOH, boiled for a few seconds in a microwave, and left overnight at room temperature. Afterwards, the KOH was decanted, the roots were rinsed with demineralized water and 10 ml 3.7% HCl and a few drops of Schaeffer ink added (after Vierheilig et al. 1998). After staining the roots for 2 h, they were rinsed in demineralized water and destained in a lactic acid:water solution (1:1). The roots were checked for percentage root colonization by AMF at × 200 magnification using the magnified intersection method, checking on average 90 intersections (McGonigle et al. 1990). Soil subsamples (4 g) of each pot were used to extract exraradical hyphae using an aqueous extraction and membrane filtration technique, with subsequent determination of the extraradical AMF hyphal length according to Rillig et al (1999).

Statistical analysis

The data were analysed with factorial analyses of variance (ANOVA) (Statistica 6.0, Statsoft) with ‘AMF’ and ‘earthworms’ as explanatory variables. Shoot biomass of the resident plant species was analysed with multivariate analysis of variance (MANOVA). Significant effects of the MANOVA can be followed up by separate factorial ANOVAs to determine which plant species are affected (Scheiner & Gurevitch 2001). Since the MANOVA on the shoot biomass of the single plant species revealed no significant species-dependent effects of AMF and earthworms, no follow-up ANOVAs on the individual plant species were conducted. The diversity of the ruderal plant community was calculated with Shannon's diversity index, H=−Σpi× log2pi, where pi represents the proportional contribution of the ith species to the community in terms of biomass. Shannon's evenness EH was calculated by dividing H by log2S, where S is the number of species growing in the ruderal plant community. The establishment and biomass of the newcomer plants was also analysed using the same factorial ANOVA. The residuals were tested for normality (Kolmogorov-Smirnov one-sample test) and homogeneity of variances (Levene test) and log-transformed when necessary.

Results

Ruderal plant community at Harvest 1

In week 11, the ruderal plant community was dominated by Arrhenatherum elatius, followed by Tanacetum vulgare (Fig. 1). Total shoot biomass of the plant community was increased by presence of AMF (Table 1; Fig. 2a); whereas the presence of earthworms did not affect the total shoot biomass of the plant community. No interaction effect between AMF and earthworms on total shoot biomass was detected. The biomass of grasses and legumes was not changed in the presence of the soil organisms, while the biomass of forbs was enhanced by AMF, independent of the presence of earthworms (Table 1; Fig. 2a). The diversity and evenness of the ruderal plant community, calculated on a biomass basis, was not affected by the presence of AMF or earthworms at the first harvest (Table 1; Fig. 2c; not shown for evenness).

Figure 1.

 Shoot biomass of the ruderal plant species across the experimental treatments at Harvest 1 (white bars) and Harvest 2 (grey bars). Means+SE.

Table 1.   ANOVA table on the effects of AMF and earthworms on the biomass (g dry weight), diversity (H) and evenness (E) of the ruderal plant community at Harvest 1 and Harvest 2. Significant F values are shown in bold (*P<0.05; **P<0.01).
Harvest 1dfBiomass Diversity
ShootsGrassesForbsLegumesHE
AMF15.94*2.875.00*0.04 0.000.01
Earthworms12.013.442.771.24 0.920.40
AMF × E11.251.170.100.19 0.030.00
Error36       
Harvest 2dfShootsGrassesForbsLegumesRootsHE
AMF12.677.12*4.14*0.240.049.54**8.02**
Earthworms10.931.163.310.394.42*0.400.19
AMF × E10.730.070.752.031.060.210.09
Error36       
Figure 2.

 Effects of mycorrhiza and earthworms [C=Control; AMF=with AMF; E=with earthworms; AMF+E=with AMF and earthworms] on shoot biomass of the ruderal plant community at (a) Harvest 1 and (b) Harvest 2, and (c) on Shannon's diversity (H) of the ruderal plant community at Harvest 1 (white bars) and Harvest 2 (grey bars). Means+SE.

Ruderal plant community at Harvest 2

The shoot regrowth of the ruderal plant community gave a similar rank order in the plant species biomass (Fig. 1). The soil organisms had no effects on total shoot regrowth, while earthworms decreased total root biomass of the plant community from 0.35 g (SE=0.04) to 0.24 g (SE=0.03) (Table 1). The shoot regrowth to root ratio was enhanced by 23% by the presence of earthworms (F1,36=5.89, P=0.020) and decreased by 17% by AMF presence (F1,36=4.97, P=0.032). The shoot regrowth of the grasses was reduced in the presence of AMF, while the shoot regrowth of the forbs was promoted by AMF (Table 1; Fig. 2b). The regrowth of the legume was poor and not affected by the soil organisms. The diversity and evenness of the ruderal plant community was enhanced by AMF (Table 1; Fig. 2c; not shown for evenness). No interaction effects of AMF and earthworms on the biomass and diversity of the ruderal plant community were detected (Table 1).

Newcomer establishment

The number of experimentally introduced newcomer plants that established in the ruderal plant community was enhanced by the presence of AMF (F1,36=6.51, P=0.015) and earthworms (F1,36=7.21, P=0.011), while no interaction between AMF and earthworms was detected (F1,36=0.72, P=0.401; Fig. 3a). From the plant species added as a seed shower, Plantago lanceolata germinated best [germination rate: 59.5% (SE=2.6)], followed by Lupinus polyphyllus [germination rate: 56.8% (SE=3.0)]. The seedling establishment of P. lanceolata was promoted by the presence of earthworms (Table 2; Fig. 3b), while the establishment of L. polyphyllus was enhanced by AMF presence (Table 2; Fig. 3c). The germination of Solidago canadensis [germination rate: 23.7% (SE=2.6)] and Eragrostis minor [germination rate: 10.3% (SE=1.5)] was poor, and no effects of the soil organisms could be detected (Table 2); Bunias orientalis did not germinate in sufficient amounts (<2%). In addition to the experimentally introduced seeds, incidental seeds originating from the surroundings also germinated in the experimental pots and comprised at harvest 7.5% (SE=1.2) of the total number of newcomer plants. The number of incidentally establishing newcomer seedlings was not affected by the presence of AMF and earthworms, while the total number of newcomer plants (including experimentally introduced and incidentally establishing seedlings) was enhanced by AMF (F1,36=6.69, P=0.014) and earthworms (F1,36=5.93, P=0.020), but not affected by an interaction between the soil organisms (F1,36=0.31, P=0.580). The mean weight of the P. lanceolata seedlings increased by 47% in the presence of earthworms [without earthworms: 6.0 mg (SE=0.9), with earthworms: 8.8 mg (SE=0.9)] (Table 2). The mean weight of L. polyphyllus [25.6 mg (SE=5.4)] was not affected by the soil organisms. The weight of the other newcomer plant species was very low (<1 mg) and no treatment effects could be detected on the mean weight of S. canadensis [0.4 mg (SE=0.01)], while the mean weight of E. minor was increased from 0.7 mg (SE=0.2) to 1.4 mg (SE=0.3) in the presence of AMF (Table 2).

Figure 3.

 Effects of mycorrhiza and earthworms on (a) the total number of seedlings of the introduced newcomer plants, (b) the number of established seedlings of Plantago lanceolata and (c) the number of established seedlings of Lupinus polyphyllus at the end of the experiment (Harvest 2). [C=Control; AMF=with AMF; E=with earthworms; AMF+E=with AMF and earthworms] Means+SE.

Table 2.   ANOVA table on the effects of AMF and earthworms on the number and weight (g dry weight) of the introduced newcomer plant species. Bunias is not included, because its germination was <2%. Significant F values are shown in bold (*P<0.05; **P<0.01).
NumbersdfPlantagodfLupinusdfSolidagodfEragrostis
AMF12.9015.19*10.0110.10
Earthworms110.36**10.7510.6910.93
AMF × E10.8910.7511.0210.10
Error36 36 36 36 
WeightsdfPlantagodfLupinusdfSolidagodfEragrostis
AMF10.1112.6310.30112.54*
Earthworms14.95*10.1210.1311.00
AMF × E12.6410.0310.3912.24
Error36 36 28 4 

Below-ground interactions between AMF and earthworms

The roots of the plant community were overall colonized by approximately 12% in the AMF treatment, whereas the control roots did not contain AMF structures. While no effect of earthworms on the root colonization by AMF was detected, the extraradical AMF hyphal lengths in the mycorrhizal treatments were increased by the presence of earthworms (F1,17=5.78, P=0.027; Fig. 4b). Of the added earthworms, 88% were recovered at the end of the experiment. AMF presence enhanced total earthworm biomass by 34% during the course of the experiment (F1,18=10.47, P=0.005; Fig. 4a).

Figure 4.

 Effects of (a) AMF on total earthworm biomass, and (b) effects of earthworms on extraradical AMF hyphal length in the AMF treatment at the end of the experiment. Means+SE.

Discussion

Consistent with our first hypothesis, the establishment success of newcomer plants into the ruderal plant community was enhanced by AMF and earthworms. AMF increased both the diversity and invasibility of the ruderal plant community, because the more subordinate forbs growing within the ruderal plant community and the newcomers could benefit from AMF. Positive effects of AMF on plant diversity (Grime et al. 1987; van der Heijden et al. 1998a, b) and invasibility (Stampe & Daehler 2003) have been documented previously. Our second hypothesis, that the introduced plant species differ in their response to AMF or earthworms, was only partly supported because only a few plant species were able to establish in the ruderal plant community, which differed in their responses to the soil organisms. Contrary to our hypotheses, the impact of the soil organisms on the establishment and growth of the plant species was not easy to predict. As expected, no interactive effects of the two groups of soil organisms on plant performance were found; nevertheless, AMF and earthworms seemed to influence each other. In general, our results tend to support the mutualist facilitation hypothesis (Richardson et al. 2000) and the suggestion of Levine & D'Antonio (1999) that invaders benefit from the same factors as established species within a community.

In the first growing period, AMF had a positive effect on total shoot biomass in the ruderal plant community via an increase in forb biomass. In the second growing period, after cutting of the above-ground biomass of the ruderal plant community and the introduction of seeds of the mainly exotic plant species, the total shoot regrowth of the ruderal plant community was not affected by the soil organisms. However, AMF inoculation reduced the shoot regrowth of grasses and enhanced, once more, the shoot regrowth of forbs. In a long-term field study, root colonization by AMF was highest in leguminous species, intermediate in forbs and lowest in grasses (Karanika et al. 2008). Although the roots of the individual plant species could not be separated and analysed for individual biomass and AMF root colonization in the present study, the shoot biomass results also indicate a greater effect of AMF on forbs than on grasses. The legume species T. repens did not grow well and no significant effect of AMF on its growth was detected. We suggest that AMF promoted the competitive strength of the forbs against the grasses. By decreasing grass dominance, species diversity of the ruderal plant community was overall increased.

Total root biomass was decreased in the presence of earthworms. Both increases and reductions of root biomasses in the presence of earthworms have been reported (reviewed by Scheu 2003). Probably the activity of earthworms affected the soil microbial community and enhanced the nutrient availability for the plants (Wurst et al. 2003), leading to reduced root foraging in the pots. Earthworms did not affect the total shoot biomass of the ruderal plant community at either harvest, but the shoot regrowth to root ratio was enhanced by earthworm presence, showing the capacity of the plants for shoot regrowth despite lower root biomass in the presence of earthworms.

Consistent with previous studies (Tuffen et al. 2002; Wurst et al. 2004; Eisenhauer et al. 2009; Wurst & Rillig 2011), no significant interactions between AMF and earthworms on plant performance were found, suggesting independent mechanisms by which the soil organisms promoted plant growth. AMF generally increase phosphorus availability for plants, while earthworms promote nitrogen availability (Wurst et al. 2004; Bardgett 2007). We suggest that AMF and earthworms, by mainly acting on different nutrients, create distinct niches for plants, leading to additive rather than interactive effects on the vegetation. However, we found some indications that earthworms and AMF might interact within the soil. Earthworms had no effect on root colonization by AMF of the plant community but increased the total amount of extraradical fungal hyphae in the pots. Previous studies reported effects of earthworms on AMF spore distribution (reviewed by van der Heijden & Sanders 2003), and our results might indicate that earthworms stimulate extraradical hyphal growth of AMF. This could be by direct interactions, such as disruption of the hyphae and subsequent compensatory growth (Hedlund et al. 1991), or by indirect interactions, such as changed chemical or physical soil conditions due to earthworm activity. Earthworm biomass was enhanced when AMF were present in the pots, pointing also to either direct (feeding/ingesting of hyphae) or indirect (mediated by the plants or soil conditions) interactions between earthworms and AMF (Wurst 2010). However, this occurred without a detectable impact on the function of the mycorrhizal symbiosis. We assume that the detected interactions between the soil organisms were not strong enough to change their respective function in regard to plant performance. This is consistent with results of an earlier study (Ladygina et al. 2010), where interactions between soil organisms at higher resolutions were common, but did not change their mainly additive effects on the overall productivity and nutrient uptake of a grassland plant community.

The establishment of newcomer plants in the ruderal plant community was generally enhanced by the presence of AMF and earthworms. However, of the newcomer plant species added as a seed shower, only some were able to establish. P. lanceolata germinated best, and the number and weight of establishing seedlings was enhanced by earthworms. Since P. lanceolata is known to be an AMF-responsive plant, the lack of effect of AMF contradicted our hypothesis that AMF-dependent forbs would profit from AMF presence in the soil. However, a previous study (Gange et al. 2002) found that seedlings of P. lanceolata start to be colonized by AMF only after more than 3 weeks of growth. Thus it is possible that P. lanceolata will be promoted by AMF at a later stage of its development but likely profits first from the increased soil nitrogen availability due to the activity of earthworms, as shown previously (Wurst et al. 2003). A second plant species in the seed shower, the exotic legume L. polyphyllus, was not affected by earthworm presence but was promoted by AMF presence in its seedling establishment phase. The lack of earthworm effect on the legume might be explained by a lower dependency on soil nitrogen availability due its symbiosis with N2-fixing bacteria (Wurst et al. 2003). The effect of AMF on L. polyphyllus was consistent with our expectation that an AMF-responsive plant species should profit from AMF presence, since L. polyphyllus is colonized by AMF in the field (Odell & Trappe 1992) and the growth of legumes is often promoted by AMF (Scheublin et al. 2007; Karanika et al. 2008). Contrary to our hypothesis that the newcomer grass would profit from earthworm presence, earthworms had no effects on E. minor, while AMF increased the weight of the few establishing seedlings of this exotic grass. Besides a direct promotion of newcomer plants by AMF, the negative effect of AMF on grass regrowth in the ruderal plant community might have alleviated competition by the dominant grasses within the community and thus indirectly promoted newcomers such as L. polyphyllus and E. minor. In general, there is a lack of studies on the effects of AMF on germination and seedling survival of newcomer plants in established plant communities. One exception is the study of van der Heijden (2004), which documented a positive effect of AMF on the biomass and phosphorus acquisition of native seedlings after 9 months of growth in a calcareous grassland community that contained the same plant species as adults. He argued that existing AMF hyphal networks are the likely cause for the promotion of seedling recruitment by AMF. In the case of earthworms, recent studies (Milcu et al. 2006; Eisenhauer & Scheu 2008b) reported strong impacts of anecic earthworms (Lumbricus terrestris) on seed survival and germination due to the burying of seeds of specific size classes. However, in the present study we used endogeic earthworms that build horizontal burrows and are less likely to bury seeds. For unknown reasons, the other exotic plant species added as a seed shower (Solidago candensis and Bunias orientalis) germinated poorly and we were unable to detect effects of the soil organisms on their establishment and performance within the ruderal plant community.

In conclusion, our results indicate that AMF can enhance the diversity and invasibility of a ruderal plant community. Since some ruderal plant communities are dominated by plants that do not depend on AMF (Rydlová & Vosátka 2001), the promotion of subordinate plants and newcomers by AMF could accelerate plant succession (Allen & Allen 1984) and invasibility. The observed positive effects of AMF and earthworms on newcomer establishment likely apply to sites where phosphorus or nitrogen is limiting, because the promotion effect of AMF and earthworms on establishing plants is likely related to increased nutrient uptake due to the mutualistic soil organisms. The present study documents that besides promoting specific native plant species and subsequently changing plant community structure and diversity, mutualistic soil organisms can promote the establishment of newcomers, including exotic plant species.

Acknowledgements

We thank Veruschka Vender for help with analysing the hyphal length in soil. We are also grateful to Francesco de Bello and the anonymous reviewers for their constructive comments on the manuscript. The study was funded by the Freie Universität Berlin.

Ancillary