Testing the mutualism disruption hypothesis : physiological mechanisms for invasion of intact perennial plant communities

Soil resources derived from mutualistic arbuscular mycorrhizal fungi (AMF) play a critical role in the physiological function of many native plant species. Allelopathic plant invasion studies have revealed declines in AMF inoculation potential of invaded soils, and lost opportunities for plants to form new AMF associations. Yet, if allelochemicals also kill AMF external hyphae already associated with plant roots, this mutualism disruption should result in physiological stress for native plants. We previously demonstrated that forest soils infested with garlic mustard (Alliaria petiolata), an allelopathic invader, exhibit reduced fungal hyphal abundance. Here, we demonstrate for the first time that treatment with garlic mustard tissue reduces soil respiration rates and diminishes physiological function of false Solomon’s seal (Maianthemum racemosum), an AMF-dependant forest understory native. Treated plants exhibited reduced stomatal conductance and photosynthesis relative to controls, consistent with the proposed loss of AMF function. Such physiological declines, if sustained over several growing seasons, could decrease native understory perennials’ growth rates and increase their susceptibility to environmental stresses. These data provide an explicit mechanism that can help explain the loss of established native perennials from invaded mature forests. We propose that the physiological costs of mutualism disruption may be a widespread but previously untested mechanism enhancing the invasion of undisturbed ecosystems by allelopathic species.


INTRODUCTION
Plants rely on mutualistic interactions for a number of services that are vital for reproduction, defense, dispersal, and nutrient acquisition.The most widespread plant mutualism is likely the interaction with arbuscular mycorrhizal fungi (AMF): it is estimated that up to 90% of all land plants participate in AMF mutualisms (Smith and Read 2008).The basis for this mutualism is a two-way exchange of resources.Up to 20% of a plant's carbon is shunted to the obligately symbiotic AMF, while AMF improve the supply of water, phosphorus and nitrogen to the plant (Smith and Read 2008).As a result of the numerous benefits that plants derive from AMF, these belowground mutualists can strongly influence the physiology (Fig. 1), overall carbon gain, and likely the competitive ability of their host plants.Disruption of this key plant mutualism is hypothesized to facilitate invasion (Mitchell et al. 2006, Reinhart andCallaway 2006).
The ''mutualism disruption'' hypothesis sug-gests that inhibition of native mutualists can provide invaders with a competitive advantage over mutualism-dependent native species.While this hypothesis has been tested extensively using reproductive mutualisms (reviewed in Traveset and Richardson 2006), it is increasingly recog-Fig.1. Plant physiology is dependent on both nutritional benefits (solid arrows) received from AMF and carbon costs (dashed arrows) delivered to AMF by the plant.In the soil, external hyphae of AMF (fine black lines) uptake soil nitrogen (N) and phosphorus (P) that is transported to the host plant.Arrow 1: Nitrogen is essential for the formation of chlorophyll and ribulose bisphosphate-carboxylase/oxygenase (RuBisCO), and as a result, photosynthetic rate (A n ) is highly correlated with total leaf nitrogen content (Evans 1989).Arrow 2: Phosphorus (P) is required to build ATP and other cofactors that play important roles in the Calvin cycle.P is also crucial for the transport of carbon assimilates (represented as pentagons) out of the chloroplast and P deficiencies can lead to a build-up of assimilates and down-regulation of A n (Sivak and Walker 1986).Arrow 3: AMF enhance the plant's ability to capture water (H 2 O) and increase water availability, which results in greater stomatal conductance (g s ) and transpiration (E; Auge ´2001).Arrow 4: Maintenance costs and rapid turnover of AMF (Staddon et al. 2003) create an additional carbon sink and plants up-regulate photosynthetic rates in response.Studies in which non-mycorrhizal and mycorrhizal plants are matched for foliar [N] and [P] demonstrate that mycorrhizal plants have higher A n than non-mycorrhizal plants (Wright et al. 1998, Miller et al. 2002).
nized that AMF mutualism disruption may also play an important role in invasions.For example, plant invaders can reduce AMF density (Roberts andAnderson 2001, Vogelsang andBever 2009) and the diversity and abundance of AMF external hyphae (Mummey and Rillig 2006) in the soil.The adverse effects of invaders on AMF mutualists can result from either negative feedback (Vogelsang and Bever 2009) or novel chemical weapons (Callaway and Ridenour 2004).Many invaders are suspected of employing the latter, including Alliaria petiolata (garlic mustard, Brassicaceae), a rampant forest understory invader.If allelochemicals disrupt the function of AMF mutualists, physiological impairment of the plant host could provide a competitive edge for invaders.
Garlic mustard has become a model study system, and as a result, its allelopathic chemicals are increasingly understood.Garlic mustard produces numerous secondary compounds, including glucosinolates (Vaughn and Berhow 1999), which are unique to the Brassicaceae.Upon hydrolysis in the soil, glucosinolates are mainly converted to isothiocyanates, a class of compounds toxic to AMF and other soil organisms (Brown and Morra 1997).Greenhouse experiments demonstrate that soils from garlic mustard-invaded sites or soils pre-conditioned with garlic mustard reduce AMF colonization and biomass in native tree seedlings (Stinson et al. 2006) and increase mortality of herbaceous seedlings (Callaway et al. 2008).Recently, we showed that allyl isothiocyanate (AITC)-an abundant compound exclusive to garlic mustard (Vaughn andBerhow 1999, Barto et al. 2010)-is present in garlic mustard-invaded soils (Cantor et al. 2011).Our companion bioassay revealed that even low AITC concentrations can reduce AMF spore germination by ;60%.Furthermore, garlic mustard-invaded areas at our study site showed reduced colonization by fungal hyphae compared to control areas (Cantor et al. 2011).These new results clearly demonstrate that garlic mustard's allelochemicals are present in field soil, and that even low AITC concentrations are capable of reducing AMF growth and abundance.Thus, the forest invader, garlic mustard, is an ideal species for testing a ''mutualism disruption'' hypothesis.
The native understory perennial herbs of deciduous temperate forests are a phylogenetically diverse group of plants (Gilliam 2007) that are highly dependent on AMF (Brundrett and Kendrick 1988).The AMF-forest herb mutualism is unique: unlike AMF crop species whose arbuscules are short-lived (4-5 days), arbuscules within forest herbs' roots function for several months (Brundrett and Kendrick 1990).Furthermore, the coarse roots of many understory herbs lack fine root hairs, and AMF external hyphae may function as root hairs for these species (Brundrett 1991).Together these factors indicate high AMF dependence in forest herbs.
In this study we integrate the fields of plant ecophysiology and invasion biology to provide a first step in testing for physiological consequences of mutualism disruption.We propose that if an allelopathic invader depresses the function of AMF external hyphae, the influx of critical resources from the AMF to its host will be reduced, resulting in physiological stress of the host.We hypothesize that this stress could be manifested through several interacting physiological pathways (Fig. 1).Decreases in nutrient supply rates can reduce photosynthetic capacity and demand for CO 2 through reductions in RuBisCO, indirectly causing partial stomatal closure (Fig. 1, arrows 1 and 2).Limited water availability can directly cause partial stomatal closure (arrow 3).The interplay of the conflicting influences of water stress and CO 2 demand for photosynthesis largely determine stomatal conductance.A common physiological measurement-leaf internal CO 2 concentration-can be used to reveal the extent to which water and/or nutrient limitation is driving stomatal closure (Wong et al. 1979, Farquhar andSharkey 1982).Under nutrient limitation, internal CO 2 concentrations remain unchanged as the partially closed stomata provide sufficient CO 2 to the diminished photosynthetic machinery.In contrast, reduced internal CO 2 concentration occurs under water limitation because photosynthetic CO 2 demand outstrips its diffusion into the leaf.Finally, we expect that loss of sink strength via allelochemical-induced reductions in AMF hyphal function can further lower photosynthetic capacity (arrow 4).Here we use ecophysiological measurements in field and common garden experiments to determine which of these conflicting influences, water stress or nutrient driven reductions in CO 2 demand, impact physiological function.We show that a native perennial and its AMF exhibit reduced soil respiration rates with garlic mustard treatment and implicate water limitation in plant physiological declines.

Focal species
Garlic mustard is a Eurasian biennial plant introduced to North America circa 1868 that has since spread throughout forest understory habitats (Rodgers et al. 2008).Garlic mustard releases powerful allelochemicals into the soil (Cantor et al. 2011) and can drive declines in native plant abundance and diversity in forests (reviewed in Rodgers et al. 2008).
We chose Maianthemum racemosum (false Solomon's seal; Liliaceae) as a focal understory species because it is common in both deciduous and coniferous forests throughout North America, often occurs in sites invaded by garlic mustard (Burke 2008; A. Hale, personal observation; Fig. 2A), and is highly dependent on AMF.Like other understory perennial herbs, false Solomon's seal roots are highly colonized by AMF (76-94%; Brundrett andKendrick 1988, Burke 2008), and lack fine root hairs (Brundrett and Kendrick 1988; Fig. 2B).These attributes led Brundrett and Kendrick (1988) to classify false Solomon's seal as obligately dependent on AMF for growth and survival.AMF external hyphae associated with understory herbs typically exhibit peak growth in mid to late summer (Brundrett and Kendrick 1990).This peak coincides with garlic mustard adults' senescence (Anderson et al. 1996) and the release of allelochemicals (Cantor et al. 2011), making the AMF mutualism in false Solomon's seal particularly susceptible to disruption by garlic mustard.Finally, false Solomon's seal's roots grow at shallow depths (2-4 cm below the soil surface; A. Hale, personal observation), which increases their probability of encountering allelochemicals leaching from garlic mustard leaf litter or root exudates.

Common garden experiment: garlic mustard allelochemicals' effect on belowground respiration
We established a 3 3 4 m plot at the University of Pittsburgh's Pymatuning Laboratory of Ecolo-gy (PLE) and created a grid of 30 cm deep holes using a post-hole digger.We collected false Solomon's seal plants (N ¼ 34) on 4 June 2009 from Tryon-Weber Woods, a 34-hectare beechmaple forest in northwestern Pennsylvania, USA that is not invaded by garlic mustard.To allow measurement of belowground respiration (described below), we employed a unique pot design.We built pots of 30 cm tall 3 10 cm diameter PVC pipe with a mesh bottom (pore size: 1 mm) that allowed air and water to flow through the pot.We potted plants in a 50:50 mixture of Fafard:Turface (Conrad Fafard, Agawam, MA, USA; Profile Products, Buffalo Grove, IL, USA) with inoculum of local field soil collected around false Solomon's seal roots and 3.5 g Nutricote fertilizer (100 day release formula, Florikan E.S.A. Corporation, Sarasota, FL, USA).We randomly assigned each plant to a grid location, making the top of each pot flush with the soil surface.Since AMF colonization begins in late spring for false Solomon's seal (Brundrett and Kendrick 1990), the AMF community was already established in the root systems at the time of collection.However, since we disturbed the external hyphae during transplanting, plants and AMF were allowed to re-establish for 6 weeks prior to treatment (Jakobsen et al. 1992) to allow hyphal regrowth.We enclosed the entire plot with a wire cage to exclude mammalian herbivores and attached a 60% shade cloth to the top of the cage to simulate forest understory light levels.
Treatments.-We randomly applied one of three treatments to each pot: fresh garlic mustard tissue (N ¼ 11), fresh dame's rocket (Hesperis matronalis, Brassicaceae) tissue (another exotic mustard (N ¼ 11)), or no plant tissue (N ¼ 12).We collected green adult garlic mustard leaves, stems and roots from Wallace Woods, a mature, second-growth forest owned by PLE that was invaded by garlic mustard within the last decade (T.-L.Ashman, personal communication), and also collected green dame's rocket tissue on site at PLE.We placed 100 g of fresh garlic mustard or dame's rocket tissue into 20 3 20 3 1.5 cm fiberglass screen bags (pore size: 1 mm) and transported the bags to the common garden for application to the false Solomon's seal pots.We predicted that allelochemicals leaching from the garlic mustard treatment would kill AMF exter-nal hyphae.However, we predicted that the dame's rocket treatment would have a negligible effect on AMF as this species can sustain AMF colonization in its roots (Demars and Boerner 1995) despite its glucosinolate production (Larsen et al. 1992).For the no plant tissue treatment, we left the screen bags empty.Thus, the dame's rocket treatment allows us to separate the effects of garlic mustard glucosinolates from leaf tissue effects, while the empty screen bag treatment allowed us to assess background levels of soil respiration.We placed the screen bags at the base of the plants and fastened them to the soil with stainless steel pins, ensuring that the bags were in direct contact with the soil surface.Treatments were imposed on 20 and 21 July 2010.
The application of garlic mustard tissue to the pots allowed us to closely simulate natural levels of allelochemicals, as decomposition is a major route of allelochemical release into the soil (Rice 1974), and removed the confounding factor of competition when garlic mustard and false Solomon's seal are grown together in pots.Because of the rapid decomposition of garlic mustard tissue (Rodgers et al. 2008) and sorption of isothiocyanates in soil (Matthiessen and Shackleton 2005), coupled with the quick turnover rate of AMF external hyphae (i.e., 5-6 days; Staddon et al. 2003), we determined that a one-week treatment would be sufficient for suppression of AMF external hyphal function.
Belowground respiration.-Wemeasured belowground respiration in each pot using a LI-COR 6400 infrared gas analyzer (IRGA; LI-COR Biosciences, Lincoln, NE, USA).We fabricated a sealed airflow path to pass CO 2 -free air through the mesh bottom of each PVC pot, forcing the CO 2 in the soil matrix to flow out of the top of the pot into the IRGA (Appendix A).We recorded the ambient air temperature and the CO 2 concentration (our estimate of belowground respiration) of this air stream at time zero and every two minutes, for a total of 10 minutes.We chose this sampling interval because 1-hour trial runs revealed that .75% of available CO 2 in the soil was captured in the first 10 minutes.To separate the effect of garlic mustard on microbial vs. root respiration, at the end of the experiment we harvested each plant and recorded its fresh root mass.To assess the potential for direct effects of garlic mustard's allelochemicals on false Solomon's seal, we compared the fresh root mass across treatments.
Statistical analysis.-Wefit a curve ([CO 2 ] ¼ time) to the data for each pot for a total of 34 regressions.The area under each curve, calculated using Mathematica 7.0 (Wolfram Research, Champaign, Illinois, USA), estimates the total CO 2 captured across the sampling duration.We compared belowground respiration in the garlic Fig. 2. False Solomon's seal is an ideal native species for studying the impacts of mutualism disruption by garlic mustard because (A) it is commonly found in forests invaded by garlic mustard and (B) it also has very coarse roots that lack fine root hairs, suggesting a high degree of mycorrhizal dependency.v www.esajournals.orgmustard vs. dame's rocket treatments using an analysis of covariance (ANCOVA) with root mass and ambient temperature as covariates.We corrected the final means from this analysis by subtracting the mean respiration value in the empty screen treatment, which represented background levels of soil respiration (all uncorrected values are shown in Appendix A).Since our a priori prediction was for lower respiration in the garlic mustard treatment relative to the dame's rocket treatment, we report one-tailed Pvalues.We compared root mass using a Kruskal-Wallis non-parametric test with the model: root mass ¼ treatment.

Field experiment: garlic mustard allelochemicals' effect on native plant physiology
Our experimental site was the Trillium Trail Reserve, a 16-hectare mixed mesophytic forest that is owned and managed by the Fox Chapel Borough, PA, USA.We estimate that 73% of the 79 herbaceous species at Trillium Trail associate with AMF (Appendix B).Because garlic mustard invaded Trillium Trail in 1992 (L. Smith, personal communication), the garlic mustard plants in this young population are likely to have high glucosinolate concentrations (sensu Lankau et al. 2009).We have demonstrated that the detected AITC levels in invaded soil at this site can significantly reduce both AMF spore germination and fungal hyphal abundance (Cantor et al. 2011).Thus, the potential for AMF disruption at this site is high.
Experimental design.-Wepaired false Solomon's seal plants (N ¼ 18, 9 pairs) based on two factors that can influence physiology: individual plant size and microhabitat (Lambers et al. 2008).We used height as our proxy for plant size because it is highly correlated with total leaf area in false Solomon's seal at Trillium Trail (R 2 ¼ 0.97, N ¼ 23; S. Kalisz, unpublished data).In matching plants for microhabitat, we ensured that paired plants were no more than 1 m apart and experienced similar tree canopy cover and moisture regimes.We again prepared screen bags filled with garlic mustard tissue collected on site and an empty screen bag served as the control.We cleared the natural leaf litter from the immediate area surrounding each focal false Solomon's seal, and then randomly assigned one plant within each pair to the control, while the other received the garlic mustard treatment.We applied these treatments between 23 and 27 June 2008 and left them in place for two weeks.
Leaf gas exchange measurements.-Wedetermined that 600 lmolÁm À2 Ás À1 is a saturating irradiance level for false Solomon's seal (Appendix C) and used this light level for all subsequent physiological measurements because this maximizes our ability to detect water and/or nutrients, rather than light, as limiting resources.Between 15 and 21 June 2008, prior to imposing the treatments, we took pre-treatment physiological measures during which mean daily temperatures ranged from 15-228C.For each plant we recorded net photosynthetic rate (A n ), stomatal conductance (g s ), transpiration rate (E), and leaf internal CO 2 concentration (C i ) every 15 seconds for 1 minute, yielding 5 measures for each physiological trait.Measurements were averaged to provide pre-treatment estimates for each plant.On 8 and 10 July 2008 mean daily temperatures ranged from 21-248C and we repeated all physiological measurements to obtain post-treatment estimates.
Statistical analysis.-Becauseall of the physiological response variables are highly correlated, we first used a multivariate analysis of covariance (MANCOVA; physiology (conductance, transpiration, photosynthesis) ¼ temperature þ humidity þ size þ treatment) to determine the overall effect of garlic mustard on false Solomon's seal physiology.Temperature, relative humidity and plant size were covariates in the analyses, as all of these variables are known to affect one or more of the measured physiological traits (Lambers et al. 2008).All of the F-statistics reported for the MANCOVA were identical, therefore, here we report only the F-statistic for Roy's greatest root, as it leads the most naturally to post hoc tests (Scheiner 2001).Upon obtaining a significant F-statistic, we conducted separate analysis of covariance (ANCOVA) tests for each physiological variable (conductance, transpiration, or photosynthesis ¼ temperature þ humidity þ size þ treatment).Because these were planned comparisons, Type I error correction is not necessary.
After obtaining the physiological response results from the ANCOVAs, we then performed a post hoc analysis on the intercellular CO 2 concentration (C i ) data from both treatments with a Bonferroni correction to account for multiple comparisons (P ¼ 0.05/4 ¼ 0.0125; Scheiner 2001).Statistical analyses for the field and common garden experiments were run using SAS version 9.2 (SAS Institute, Cary, North Carolina, USA).

Common garden experiment: belowground respiration
We found that belowground respiration was dramatically lower for garlic mustard-treated relative to dame's rocket-treated (control) pots (Fig. 3A; F 3,18 ¼ 2.42, P ¼ 0.07).Over the sampling period, CO 2 levels declined as the residual CO 2 in the soil matrix and current production of CO 2 was pushed through the pot (Fig. 3B).At each time point, CO 2 levels from the garlic mustardtreated pots were significantly lower than those in dame's rocket-treated pots (Fig. 3B).Ambient air temperature and root mass were not significant covariates in this analysis (temperature range ¼ 21-358C; root mass range ¼ 2.1-8.2g).Root mass did not differ across treatments (Kruskal-Wallis: df ¼ 2; P ¼ 0.51).

Field experiment: plant physiological responses
Prior to treatment, there was no significant difference in physiological rates among false Solomon's seal plants destined for control or garlic mustard treatments (MANOVA; Roy's greatest root, F ¼ 0.46, P ¼ 0.72).However, after two weeks of treatment, garlic mustard-treated plants displayed significantly lower physiological rates relative to control plants (MANOVA; Roy's greatest root, F ¼ 3.70, P ¼ 0.05; Fig. 4A-C).Specifically, g s had the strongest response to garlic mustard treatment: g s in garlic mustardtreated plants was 36% lower than control plants (Fig. 4B; F 4,13 ¼ 13.11, P ¼ 0.009).Similarly, E and A n were significantly reduced in garlic mustardtreated plants (F 4,13 ¼ 8.73, P ¼ 0.03 and F 4,13 ¼ 4.58, P ¼ 0.05, respectively).All covariates in the g s , E and A n individual analyses were significant (P , 0.05), except for the ANCOVA for A n , where plant size was not significant (P ¼ 0.47).

DISCUSSION
Our field experimental data clearly demonstrate that short-term exposure to garlic mustard tissue, an allelopathic invasive species, can significantly reduce the physiological function of a native understory herb, false Solomon's seal.Stomatal conductance (g s ) in garlic mustardtreated false Solomon's seal adults was reduced by 36%, with concomitant reductions in transpiration (E; 25%) and photosynthesis (A n ; 17%) compared to controls (Fig. 4A-C).This study is v www.esajournals.orgthe first to reveal an explicit physiological mechanism underlying an allelopathic species' invasion of an established native plant community.
We attribute the aboveground physiological suppression (Fig. 4A-C) by garlic mustard to allelopathic disruption of AMF external hyphal function.In our common garden experiment, total belowground respiration declined in garlic mustard-treated pots compared to dame's rocket control pots (Fig. 3).Indeed, respiration in the garlic mustard-treated pots did not differ significantly from that measured in the no plant tissue treatment (Appendix A).Thus, while the flush of nutrients from the decomposing dame's rocket tissue stimulated microbial activity above background levels, the garlic mustard tissue did not.These findings are consistent with the prediction that any nutritional benefits of litter decomposition on microbial metabolism are offset in the garlic mustard treatment by loss of AMF external hyphae.AMF are more sensitive to garlic mustard than other soil microbial groups (Lankau 2011) and AMF hyphae can have a disproportionately large impact on overall soil respiration (Johnson et al. 2002).Additionally, our previous work has shown that even low levels of allelochemicals from garlic mustard disrupt AMF spore germination and depress fungal hyphal abundance in field soils (Cantor et al. 2011).Furthermore, we found no direct effects of garlic mustard on false Solomon's seal root mass.Together, our observed reductions in belowground respiration are consistent with a reduction in the function of AMF external hyphae in the garlic mustard treatment.
Our leaf internal CO 2 concentration (C i ) data provide important insight into which soil resource limits false Solomon's seal physiology.We observed significant reductions in both g s and C i v www.esajournals.org in garlic mustard-treated plants (Fig. 4B and D , respectively).This result implicates water limitation as the primary cause of the physiological suppression of false Solomon's seal.Together our data suggest a potential causal chain: garlic mustard inhibits external AMF hyphal function, which limits water availability to false Solomon's seal, which in turn reduces stomatal conductance and lowers photosynthetic rate.
Under natural conditions, native plants in the understory likely experience prolonged periods of exposure to garlic mustard's allelochemicals.In garlic mustard-invaded sites, the continuous presence of garlic mustard rosettes and/or adults, its high seedling and rosette mortality throughout the year (Davis et al. 2006), and a 2-month period of intense leaf litter input as adults senesce (Anderson et al. 1996, Cantor et al. 2011) likely contribute to the year-round release of allelochemicals into the soil.Given the significant physiological suppression shown by our short-term pulse experiments, we anticipate that longer-term exposure of AMF-dependent plants to garlic mustard allelochemicals could affect carbon storage and resource allocation.Indeed, if carbohydrate storage of AMF-dependant species is impacted by season-long declines in physiology, then garlic mustard invasions could have long-term fitness effects on native forest perennial plants.Experiments currently under way in our lab are testing for long-term impacts of mutualism disruption.
Our results have broad implications for understanding how allelopathic invaders can lead to the collapse of established native plant communities in forest understories.Stressful environments diminish mutualism effectiveness (Bronstein 1994, Kiers et al. 2010).Through secretion of allelochemicals in the soil, garlic mustard effectively creates a physiological stress that removes the native plants' mutualistic interactions with AMF.While physical disturbance often facilitates forest invasion (Luken 2003), our data provide a link between novel weapons and native plant declines, implicating physiological disturbance as the intermediate step underpinning invasion.Interference with nutritional mutualisms and subsequent physiological declines in natives may provide both the opportunity for invaders to establish and spread throughout previously stable ecosystems and increase the vulnerability of native species to other stressors that accompany global change.

Fig. 3 .
Fig. 3. Belowground respiration from pots treated with garlic mustard (GM) is lower than in pots treated with dame's rocket tissue in our common garden experiment.(A) Total CO 2 captured in a ten-minute sampling period averaged across all plants within a treatment (least squares (LS) means 6 1SE).(B) The mean CO 2 captured per 2minute sampling interval averaged across all plants within a treatment (LS means 6 1SE).

Fig. 4 .
Fig. 4. Garlic mustard tissue negatively affects false Solomon's seal's physiological function.All physiological parameters (A-C), photosynthetic rate (A n ), stomatal conductance (g s ), and transpiration (E) were significantly reduced (*** P , 0.01; ** P 0.05) after two weeks of garlic mustard (GM) treatment relative to the controls.In a post-hoc analysis (D), intercellular CO 2 concentration (C i ) was also significantly lower (** P ¼ 0.007) for garlic mustard-treated plants compared to controls suggesting a potential role of stomatal limitation in the observed A n declines (least squares (LS) means 6 1SE).

Fig. A1 .
Fig. A1.The sealed air-flow path through the PVC pot and into the IRGA of the LI-COR 6400.CO 2 -free air was passed through the bottom of the PVC pot (inflow, solid arrow), forcing the CO 2 in the soil matrix to flow out of the top of the pot into the IRGA (outflow, dashed arrows).

Fig. C1 .
Fig. C1.Light saturation curves for false Solomon's seal.Low observed maximum photosynthetic rates (A n ) and light saturation at low irradiance levels for false Solomon's seal are typical of forest understory herbs.Data from Trillium Trail (TT; 1 plant), Pymatuning Laboratory of Ecology (PLE; 3 plants), and from Hull (2002; 2 plants).