Northeastern Naturalist 372 R.F. Dougherty, L.D. Quinn, T.B. Voigt, and J.N. Barney 22001155 NORTHEASTERN NATURALIST 2V2(o2l). :2327,2 N–3o8. 62 Response of Naturalized and Ornamental Biotypes of Miscanthus sinensis to Soil-Moisture and Shade Stress Ryan F. Dougherty1, Lauren D. Quinn2, Thomas B. Voigt3, and Jacob N. Barney1,* Abstract - A recent trend in bioenergy-feedstock development includes the use of largestatured perennial grasses whose rapid growth and biomass-accumulation rates in lowfertility conditions make them highly desirable; however, these species tend to have much in common with many invasive plant species. Miscanthus sinensis (Chinese Silvergrass), an extremely popular ornamental grass and candidate bioenergy crop, has naturalized in over half of US states, yet little is known about its environmental-stress tolerance, which is a characteristic important for bioenergy development and invasiveness. Previous studies of Chinese Silvergrass have suggested that the species’ enhanced tolerance to shade and drought conditions may be contributing to its invasion success in the US. To test this hypothesis, we conducted a greenhouse study to compare shade and soil-moisture stress tolerance among phenotypically diverse ornamental cultivars and naturalized biotypes of Chinese Silvergrass. We found enhanced plant growth and vigor in naturalized biotypes compared to ornamental biotypes across light levels from 5% to 100% of full sun. We also found that both the naturalized and the ornamental cultivars were not significantly affected by soil-moisture stress, and thus exhibited significant drought tolerance. Greater vigor and performance of naturalized biotypes in low light conditions compared to ornamental biotypes suggest that naturalized biotypes have enhanced shade tolerance, possibly due to hybridization. Our results provide direction for additional evaluations and weed-risk assessments of Chinese Silvergrass that will be critical in preventing future invasions and guide breeding for horticulture and bioenergy. Introduction Invasive plants cause impacts to ecosystem function and native species composition (Vilà et al. 2011) and bring excessive economic costs from management efforts (Leung et al. 2012). Ironically, the majority of these species have been introduced intentionally through the horticulture and landscaping industries in the US (Dehnen-Schmutz et al. 2007, Reichard and White 2001). Over 80% of woody invasive species in the US are horticultural in origin (Reichard and White 2001), and ornamental/horticultural species comprise approximately 60% of the Florida- Invasive Plant Council (IPC) and California-IPC noxious/invasive plant lists (available at http://www.fleppc.org/ and http://www.cal-ipc.org/, respectively). Breeding and selection of horticultural species often results in traits that may later confer escape and invasion potential, including broad environmental tolerance, pest resistance, and shade tolerance (Culley and Hardiman 2007, Kitajima et al. 2006). 1Department of Plant Pathology, Physiology, and Weed Science, Virginia Tech, Blacksburg, VA 24061. 2Energy Biosciences Institute, University of Illinois, Urbana, IL 61801. 3Department of Crop Sciences, University of Illinois, Urbana, IL 61801. *Corresponding author - jnbarney@vt.edu. Manuscript Editor: Douglas DeBerry Northeastern Naturalist Vol. 22, No. 2 R.F. Dougherty, L.D. Quinn, T.B. Voigt, and J.N. Barney 2015 373 For example, Miscanthus sinensis Anderss. (Chinese Silvergrass), is a perennial grass native to East Asia that is one of the most popular ornamental species in the US (Quinn et al. 2012). Originally introduced to the US in the 19th century, it has since naturalized in over 25 US states, primarily along the Appalachian corridor (Dougherty et al. 2014). Chinese Silvergrass is a large-statured grass that currently has more than 100 commercially available ornamental cultivars (Grounds 1998), and annual retail sales of Chinese Silvergrass totaled nearly $40 million by 2009 in North Carolina alone (Trueblood 2009). Not only is Chinese Silvergrass a major ornamental species, but it is also under evaluation as a candidate bioenergy crop (as is its sterile hybrid with M. sacchariflorus (Maxim.) Hack. (Amur Silvergrass), Miscanthus × giganteus J.M. Greef and Deuter ex Hodk. and Renvoize [Giant Miscanthus]) due to its broad environmental tolerance and aboveground-biomass yield potentials (Quinn et al. 2012). Chinese Silvergrass is currently classified as an invasive species by the US Forest Service (2006) as well as by several regional invasive plant councils (SE-EPPS 2010). Classification as an invasive species by these organizations includes all varieties and ornamental cultivars of Chinese Silvergrass, although little is known about ecological requirements and invasive potential among cultivars and naturalized populations. Previous studies have found significant biological and ecological variation among ornamental cultivars of several species, including Hydrangea macrophylla Thun. (Hortensia; Reed 2002), Ruellia tweediana Griseb. (Mexican Petunia; Wilson and Mecca 2003), and Berberis thunbergii DC. (Japanese Barberry; Lehrer et al. 2006). Several of these studies found significant differences in seed production and biomass in as few as 3 varieties. Cultivars of Chinese Silvergrass are bred or selected to exhibit wide variation in phenotypic characteristics, including tiller height, basal diameter, flowering time, infloresence color, leaf width, variegation, and color. A recent survey of several naturalized Chinese Silvergrass populations across the eastern US also found tremendous phenotypic variation among naturalized biotypes (Dougherty et al. 2014). For example, we commonly encountered individuals with variegated striping like “Zebrinus” and narrow leaves like “Gracillimus”. It is therefore likely that important differences may exist in the ecology, invasive potential, and bioenergyfeedstock potential among commercially available ornamental cultivars of Chinese Silvergrass that contribute to the success of this species outside cultivation. The exact mechanisms by which Chinese Silvergrass successfully naturalizes outside cultivation are unknown. The most common habitats of naturalized Chinese Silvergrass populations are open areas, such as roadsides and forest edges, where light and water availability are rarely limiting (Dougherty et al. 2014). Some of these habitats are characterized by frequent disturbance and low soil-nutrient availability, which suggests that Chinese Silvergrass may employ several strategies to establish and survive. There is no evidence that naturalized biotypes have enhanced drought tolerance (Matlaga et al. 2012), but Chinese Silvergrass may possess shade tolerance as suggested by the many populations that have encroached into forest understories where there is reduced light availability and potentially reduced soil Northeastern Naturalist 374 R.F. Dougherty, L.D. Quinn, T.B. Voigt, and J.N. Barney 2015 Vol. 22, No. 2 moisture (Dougherty et al. 2014). Thus, tolerance to soil-moisture stress and shade may be important traits for establishment and naturalization of Chinese Silvergrass. Shade tolerance is the ability of a plant to thrive and survive under low-light conditions (Valladares and Niinemets 2008). Horton et al. (2010) found that Chinese Silvergrass individuals were capable of maintaining high photosynthetic rates within the natural light gradient (5–100% relative transmittance) of a forest understory. More recently, Matlaga et al. (2012) directly compared the morphology and light response of Chinese Silvergrass seedlings from its native and introduced ranges with mixed results; however, they found that seeds could germinate and seedlings could grow in as little as 30% of full light. Chinese Silvergrass has also been described as a drought-tolerant species (Clifton-Brown and Lewandowski 2000, Quinn et al. 2012), although empirical studies of this tolerance have generally used varieties bred specifically for bioenergy, rather than ornamental or naturalized biotypes (Zub and Brancourt-Hulmel 2010). A better understanding of the ecology and environmental tolerance of M. sinensis could add valuable insight to invasion ecology, the horticulture trade, and the bioenergy industry. By identifying traits and characteristics that may confer invasiveness, we can develop risk assessments and management protocols to mitigate further naturalization and spread of Chinese Silvergrass. In this study, we sought to evaluate shade- and drought-tolerance of Chinese Silvergrass in its introduced range, and identify ornamental cultivars that may contribute to its invasive potential. We conducted a greenhouse study to compare naturalized and ornamental biotypes that shared common habitats and genetic backgrounds. Because the naturalized biotypes likely experience more limiting and variable environmental conditions compared to planted ornamental cultivars, we predicted that (1) naturalized biotypes would have greater shade- and drought-tolerance than ornamental cultivars, and (2) there would be greater variation in shade- and drought-tolerance among the ornamental cultivars than the naturalized biotypes, suggesting that certain cultivars may be more likely to naturalize than others. Materials and Methods Cultivar selection Due to greenhouse space limitations, we could only evaluate 7 ornamental cultivars and 3 naturalized populations (Table 1). Although this sample did not fully represent the entirety of variation of Chinese Silvergrass, we chose them to capture variation in several important life-history traits. Our previous observations in the field (Dougherty et al. 2014) showed that naturalized biotypes are quite variable, and likely composed of several ornamental cultivars, perhaps hybridizing with each other. We selected some of the most readily available ornamental cultivars based on common phenotypic observations in naturalized biotypes (e.g., leaves that resembled “Zebrinus” and “Gracillimus”; Table 1; R.F. Dougherty et al., pers. observ.). We chose more upright and robust cultivars such as “Graziella” and “Gracillimus”, as well as shorter, bushier cultivars such as “Adagio”. We also included several variegated cultivars such as “Dixieland”, “Variegatus”, and Northeastern Naturalist Vol. 22, No. 2 R.F. Dougherty, L.D. Quinn, T.B. Voigt, and J.N. Barney 2015 375 Table 1. Summary and descriptions of ornamental and naturalized biotypes. Ornamental Mean seed setA cultivar Code (per individual) DescriptionB “Adagio” AD 27,078 1 m, blooms September–November, thin silver-gray foliage, pink inflorescence turning white “Autumn Light” AL 157,936 2.1–3 m, blooms in September, hardy variety, inflorescence bronze-red turning to silver “Dixieland” DX 785 1–1.2 m, blooms in September, wide green leaves with white stripes, strong reddish inflorescence (dwarf form of M. variegatus) “Gracillimus” GC 3146 1.5–1.8 m, blooms in October, slender foliage, inflorescence bronze-red turning to silver “Graziella” GZ 90,569 1.5–1.8 m, blooms in August, slender foliage, large white inflorescence “Variegatus” VR 211 1.8–2.1 m, blooms in late September, white-striped foliage, strong reddish inflorescence “Zebrinus” ZB 16,621 1.8–2.4m, blooms reddish in September–October, light green foliage, has horizontal yellow zebra-like bands Naturalized Population Code Latitude Longitude Habitat New York NY 40.7093 -73.1489 Open field and forest edge in conservation area Maryland MD 39.5598 -76.3825 Roadside and forest edge North Carolina NC 35.2690 -82.4102 Roadside and open field along railroad right-of-way AMadeja et al. (2012). BAvailable online at http://www.kurtbluemel.com/. Table 2. ANOVA of morphological responses and logistic regression of mortality. Biotype refers to ornamental or naturalized sources, and populations are the individual ornamental cultivars or naturalized biotypes from Table 1. Δ Height Δ Tiller number Δ Basal diameter Leaf diameter Mortality Variable F P F P F P F P χ2 P Population (biotype) 3.93 <0.001 2.30 0.0203 5.05 <0.001 5.23 <0.001 111.80 <0.001 Biotype (B) 164.66 <0.001 0.06 0.8083 19.63 <0.001 23.94 <0.001 28.86 <0.001 Shade (ST) 46.36 <0.001 8.96 <0.001 7.75 <0.001 8.62 <0.001 1.40e-4 0.9999 Moisture (MT) 1.17 0.3215 1.38 0.2479 3.35 0.0191 1.54 0.2049 1.50e-5 1.000 ST*MT 1.40 0.2139 2.62 0.0167 2.33 0.0321 1.50 0.1773 2.74 0.8402 B*ST 16.60 <0.001 4.21 0.0156 3.19 0.0422 3.68 0.0261 8.32e-6 1.000 B*MT 0.13 0.9426 0.31 0.8150 0.07 0.9975 1.71 0.1636 8.67e-6 1.000 B*ST*MT 1.11 0.3526 0.84 0.5384 0.48 0.8236 1.41 0.2087 2.62 0.8548 Northeastern Naturalist 376 R.F. Dougherty, L.D. Quinn, T.B. Voigt, and J.N. Barney 2015 Vol. 22, No. 2 “Zebrinus” because we frequently found variegated individuals within naturalized populations (R.F. Dougherty et al., pers. observ.). We also selected ornamental cultivars based on varying reproductive output, which ranged from an average of 211 seeds per individual (“Variegatus”) to 157,936 seeds per individual (“Autumn Light”) (Madeja et al. 2012). We purchased all 7 ornamental cultivars from Tidwell Nurseries (Greenville, GA) as 5” plugs. We selected naturalized biotypes from 18 naturalized populations of Chinese Silvergrass that we surveyed in the summer of 2011 and propagated the experimental plants from seed. We chose populations that were distributed across the entire latitudinal gradient of Chinese Silvergrass in the eastern US and that occurred in areas of varying light availability (see Dougherty et al. 2014 for more detail on these populations): Heckscher State Park, NY (code = NY), Loch Raven, MD (MD), and Henderson, NC (NC) (Table 1). Experimental design Immediately after purchase, we transplanted ornamental Chinese Silvergrass cultivars into 12.5 cm x 12.5 cm x 14.5 cm pots with Metro-Mix 510 media (Sun Gro Horticulture, Bellevue, WA) and allowed them to acclimate under greenhouse conditions (day/night = 28/20 °C) for 2 weeks. We propagated individuals from seeds collected from naturalized populations in November 2011 that were individually sown in several 128-cell trays (with 3 cm x 3 cm x 3 cm cells) with Metro-Mix 510 media and grown for 8 weeks under greenhouse conditions. This 8-week period allowed the naturalized biotypes to reach the approximate size of the commercially obtained ornamental cultivars. After the propagation period, we randomly selected naturalized individuals and transplanted them into 12.5 cm x 12.5 cm x 14.5 cm pots as above to ensure that all Chinese Silvergrass individuals were approximately the same size at the beginning of the study. We chose 4 soil-moisture-availability treatments for the experiment: high (40% v/v or field capacity), medium–high (30%), medium–low (20%), and low (10%). We used scheduled drip-irrigation to provide the conditions for the high, medium– high, and medium–low treatments and maintain soil moisture ± 5% of the treatment target at all times. The low treatment (10% v/v) was a simulated acute drought in which individual pots received 1L water biweekly. At the beginning of the study, we watered each pot to field capacity, and began soil-moisture treatments after ~7 days. We monitored soil-moisture levels weekly with a TH300 soil-moisture probe (Dynamax Inc., Houston, TX) and obtained water-potential values (MPa) for each treatment with a WP4 Dewpoint Potentiameter (Decagon Devices, Inc., Pullman, WA). The average water-potential values were -0.02 MPa ± 0.02 (high), -0.12 MPa ± 0.04 (medium high) -0.50 MPa ± 0.08 (medium low), and -4.05 MPa ± 0.56 (low). In addition to the soil-moisture-availability treatments, we imposed a series of light-availability treatments: 100% (high), 40% (medium), and 5% (low) relative transmittance; and used single or multiple layers of 60% shade cloth (International Greenhouse Company, Danville, IL) to achieve desired experimental levels. Using an AccuPAR LP-80 PAR ceptometer (Decagon Devices), we recorded photosynthetically active radiation (PAR) levels between 12:00 PM and 3:00 PM ~5 times Northeastern Naturalist Vol. 22, No. 2 R.F. Dougherty, L.D. Quinn, T.B. Voigt, and J.N. Barney 2015 377 over the first 2 weeks of the study in each light-availability treatment to confirm the levels of relative light transmittance. The high-, medium-, and low-light treatments had PAR levels of 1209 ± 120 μmol m-2 s-1, 488 ± 27 μmol m-2 s-1, and 60 ± 10 μmol m-2 s-1, respectively. Due to the irrigation infrastructure of the greenhouse, we had to spread plants in each soil-moisture treatment across 2 benches, except the 10% soil-moisture plants, which we hand-watered and located among the plants receiving the other treatments (i.e., 2 benches of 20, 30, and 40% soil moisture for 6 total benches). We divided each of the 6 benches in half, and for each moisture treatment randomly assigned 1 of the 3 light-availability treatments per half bench (leaving half of one bench empty). For example, the half of the bench assigned to 40% sun was covered with one layer of shade cloth supported by a PVC frame. There were 5 replicates of each population in each treatment combination, except for the 10% soil-moisture treatments, where only 3 individuals were included due to space limitations. Our results should be viewed within the limitations of the design imposed by the irrigation structure of the greenhouse. Data collection We collected plant-morphology data, including tiller height, tiller number, leaf width, and basal diameter prior to treatment initiation to account for starting-size variation among the populations. We recorded individual survival biweekly until the termination of the experiment at 16 weeks. We re-randomized the locations of individuals within their respective light treatments at each data collection to reduce location effects within the greenhouse. After 6 weeks of treatment, we collected photosynthetic data from 3 individuals of each population in each treatment using a LI-COR XT6400 gas exchange system (LICOR, Lincoln, NE). We subjected 2–3 of the youngest, fully expanded leaves from each individual to varying levels of light from 0 to 1500 μmol m -2 s-1. We generated steady-state light curves from photosynthetic data in SigmaPlot 11 (Systat Software Inc., San Jose, CA) from 3 individuals of each population in each treatment combination. Photosynthetic data was fitted to the Von Bertalanffy growth equation: p = a + Lθ(1 - e- kt) where p is photosynthetic rate, a is dark respiration, Lθ is the maximum photosynthetic capacity, k is quantum yield and t is time (Horton and Neufield 1998). Data analysis We analyzed plant-morphology data from surviving individuals, including tiller height, basal diameter, tiller number, and leaf width with ANOVA. Because ornamental and naturalized biotypes were initially transplanted at different stages of maturity and size (i.e., plugs and seeds), we analyzed plant-morphology data as a total-percent change over the entire 16-week trial rather than as raw values. All morphological responses were arcsin square-root transformed before analysis to meet the assumptions of ANOVA. We considered biotype Northeastern Naturalist 378 R.F. Dougherty, L.D. Quinn, T.B. Voigt, and J.N. Barney 2015 Vol. 22, No. 2 (i.e., ornamental, naturalized), population (nested within biotype), shade, and soil-moisture treatments as fixed effects with all interactions. Treatment means were compared with Tukey’s HSD (P < 0.05). At the end of the trial, individuals were assigned a binary mortality value, which was analyzed using logistic regression using the variable structure as above. We compared steady-state light-curve growth-constants a, Lθ, and k from each population with ANOVA to identify physiological differences between populations and their responses under varying treatment levels as above. All statistical analyses were conducted with JMP v10. Results Survival varied among population, biotype, light availability, and soil moisture (Fig. 1, Table 2). At the conclusion of the trial, 7 populations, 6 of which were ornamental, had no surviving individuals in at least 1 treatment combination. “Graziella”, MD, and NY populations had 1 or more surviving individuals in all treatment combinations, while “Autumn Light”, “Variegatus”, and “Zebrinus” all had zero surviving individuals in the low-light treatment, regardless of soil-moisture level. In total, 153 of 540 individuals (28%) did not survive. Light availability had a much stronger effect on mortality than soil moisture, with only 20% of individuals surviving the low-light treatment. Mortality under soil-moisture treatments was somewhat evenly spread among light treatments, at 18–31%. As expected, tiller height, tiller number, basal diameter, and leaf width all decreased as light availability decreased (Fig. 2, Table 2). These morphological responses were generally lower in the low-light treatment compared to the medium- and high-light levels, which performed similarly (P > 0.05). Naturalized biotypes had greater gains in tiller height and leaf width than ornamental cultivars under each soil-moisture treatment; however, the within-cultivar response did not differ among soil-moisture treatments. For example, gains in tiller height of “Gracillimus” were equal across soil-moisture treatments, but were lower than the MD, NY, or NC populations. Naturalized biotypes outperformed ornamental cultivars in tiller height in all treatments (Fig. 2a, Table 3). Changes in tiller height of naturalized biotypes in the low-light treatment were equal to changes in tiller height of ornamental cultivars in the high-light treatment (Fig. 2a). We detected subtle differences between populations of the same biotype, whether ornamental or naturalized; however, biotype was clearly a more significant driver of morphological response to stress than population alone (Table 2). Ornamental cultivars generally had higher basal diameters than naturalized biotypes (Fig. 2b). Naturalized biotypes were more tolerant to all levels of stress, accounting for less than 15% of all mortalities (n = 153). There were equivocal responses among ornamental and naturalized biotypes for culm number (Fig. 2c), while naturalized biotypes outperformed ornamental cultivars in total aboveground biomass (Fig. 2d) In addition to morphological responses, analyses of the steady-state light curves indicated significant physiological stress-response differences among cultivars and Northeastern Naturalist Vol. 22, No. 2 R.F. Dougherty, L.D. Quinn, T.B. Voigt, and J.N. Barney 2015 379 Figure 1. Average survival of each population (a) as a function of light availability, and (b) soil moisture (volume/volume, %). Northeastern Naturalist 380 R.F. Dougherty, L.D. Quinn, T.B. Voigt, and J.N. Barney 2015 Vol. 22, No. 2 biotypes (Table 3). Respiration rate, maximum photosynthetic rate, and quantum yield all varied among populations and light treatments (P < 0.05). Overall, naturalized biotypes had higher dark-respiration rates (-1.3 vs. -0.97 μmol CO2 m-2 s-1) and quantum yield (0.007 vs. 0.004) than ornamental cultivars, but ornamental cultivars had higher maximum photosynthetic rates (11.7 vs. 6.8 μmol m-2 s-1). Dark-respiration rates were higher under the low-light treatment than both the medium- and high-light treatments (5% relative transmittance = -1.18, 40% = -0.91, Figure 2. Total change in (a) tiller height (b) basal diameter, (c) culm number per individual, and (d) aboveground biomass of each biotype under all light treatments. Table 3. ANOVA of steady-state light-curve parameters (a is dark respiration, Lθ is the maximum photosynthetic capacity, and k is quantum yield) from all populations. Biotype refers to ornamental or naturalized sources and populations are the individual ornamental cultivars or naturalized biotypes from Table 1. a Lθ k (μmol CO2 m-2 s-1) (μmol CO2 m-2 s-1) (Slope) Variable F P F P F P Population (Biotype) 2.87 0.0050 7.45 less than 0.001 3.61 less than 0.001 Biotype (B) 3.06 0.0818 18.72 less than 0.001 12.30 less than 0.001 Shade (ST) 5.29 0.0058 13.37 less than 0.001 24.45 less than 0.001 Moisture (MT) 2.57 0.0555 3.27 0.0226 0.41 0.7459 B*ST 1.87 0.1565 1.91 0.1514 7.58 less than 0.001 B*MT 0.07 0.9781 1.74 0.1599 0.40 0.7505 Northeastern Naturalist Vol. 22, No. 2 R.F. Dougherty, L.D. Quinn, T.B. Voigt, and J.N. Barney 2015 381 100% = -1.34 μmol CO2 m-2 s-1; P = 0.0058); however, as expected, maximum photosynthetic rate (5% relative transmittance = 9.5, 40% = 7.0, 100% = 11.2 μmol CO2 m-2 s-1; P < 0.001) and quantum yield (5% relative transmittance = 0.008, 40% = 0.005, 100% = 0.003; P < 0.001) were greater in higher-light treatments. Maximum photosynthetic capacity of individuals exposed to the low soil-moisture-level treatment (6.76 μmol CO2 m-2 s-1) were significantly lower (P = 0.0226) than that of individuals at the medium-low (11.6 μmol CO2 m-2 s-1), medium-high (9.56 μmol CO2 m-2 s-1), and high treatments (9.50 μmol CO 2 m-2 s-1). Discussion Repeated introductions and breeding for traits that increase horticultural value have given rise to many invasive species of ornamental origin, such as Pyrus calleryana Decne. (Callery Pear), Japanese Barberry, Ardisia crenata Vent. (Coralberry), and Chinese Silvergrass (Culley and Hardiman 2007, Kitajima et al. 2006, Lehrer et al. 2006). Traits thought to increase invasiness include rapid growth, early flowering, increased flower number, broad environmental tolerance, and shade tol - erance (Culley and Hardiman 2007, Kitajima et al. 2006). The tremendous genetic and phenotypic variation among cultivars of ornamental species can lead to variation in the expression and magnitude of these traits (Conklin and Sellmer 2008, Kitajima et al. 2006, Lehrer et al. 2006). Empirical studies have found dramatic variation in the reproductive output and environmental tolerances between as few as 3 ornamental cultivars of a single species (Lehrer et al. 2006), which suggests that stress tolerance likely exists among the more than 100 phenotypically diverse ornamental types of Chinese Silvergrass. We tested the hypotheses that invasive, naturalized biotypes of Chinese Silvergrass exhibit greater tolerance to low light-availability and soil-moisture stress than ornamental cultivars, and that ornamental cultivars vary more in their degree of stress tolerance. We measured performance/ability to survive, physiological responses to light availability, physiological responses to soil-moisture availability, and the continuum of responses between naturalized and ornamental biotypes. We detected significant differences in plant morphology and survival between naturalized and ornamental biotypes of Chinese Silvergrass in response to shade and soil-moisture stress. We found that naturalized biotypes can tolerate extremely low light levels (60 ± 10 μmol m-2 s-1) and soil-moisture availability (-4.05 ± 0.56 MPa) to a greater degree than even the most-tolerant ornamental cultivars in terms of survival (Table 2), plant performance (Fig. 2), and ecophysiology (Table 3). Although we sampled only a fraction of all possible ornamental and naturalized Chinese Silvergrass populations, our results suggest that naturalized biotypes have evolved enhanced shade tolerance in the US, and that certain ornamental cultivars have greater stress tolerance than others. However, we found tolerance to low soil-moisture availability to be a trait universal to both naturalized and ornamental populations of Chinese Silvergrass. Results from our study should be viewed in the context of the limited genetic sampling of each population, and of Chinese Silvergrass as a species overall. Northeastern Naturalist 382 R.F. Dougherty, L.D. Quinn, T.B. Voigt, and J.N. Barney 2015 Vol. 22, No. 2 In the US, Chinese Silvergrass generally naturalizes in high-light areas such as roadsides and open fields, but individuals have also been found in low-light areas such forest edges, and to a lesser extent, understories (Dougherty et al. 2014). Horton and Neufeld (1998) suggested that successful invaders of forest understories tend to be shade tolerant, and express this tolerance through several morphological and physiological adaptations, such as increased leaf-area ratio, survival, and ability to maintain photosynthetic rates. Although we did not directly measure leaf-area ratio, leaf width increased in lower-light treatments; a trend associated with increased leaf-area ratios. In our study, under each level of light availability, naturalized biotypes grew taller and produced wider leaves than ornamental cultivars (Table 2, Fig. 2); however, all populations performed significantly better in the high- and medium-light treatments than in the lowlight treatment. Our results are consistent with the trait responses associated with shade tolerance, especially the ability to survive (Horton et al. 2010, McAlpine and Jesson 2007, Spencer 2012). We hypothesized that ornamental cultivars would have far greater mortality under shade stress than naturalized biotypes that had likely undergone natural selection over many generations under a range of canopy covers and light levels. Our results supported that prediction (Table 2) because naturalized biotypes accounted for less than 15% of all mortality. Not surprisingly, all mortality of naturalized biotypes occurred under the low-light treatment. In contrast, ornamental cultivars suffered greater mortality under all light treatments; however, approximately 75% of ornamental mortality was under the low-light treatment, compared to 16% and 9% in the medium- and high-light treatments, respectively. We could not, however, determine the mechanism of this enhanced performance of naturalized over ornamental cultivars. We also found significant differences in physiological responses to light availability between populations and biotypes. Plants grown in low-light environments often adapt by decreasing respiration rate (a) and maximum-photosynthetic rate (Lθ) while increasing quantum yield (k) as a way to conserve energy and maximize photosynthetic efficiency (Horton and Neufield 1998). Our results show that both biotypes followed these trends under low-light availability, although ornamental cultivars exhibited greater maximum photosynthetic rates (Table 3). Even with higher maximum photosynthetic rates, ornamental types performed very poorly in terms of morphological gains and survival, which suggests that maximum photosynthetic rate is not an appropriate indicator of shade tolerance. Quantum yield was greater in naturalized biotypes, which means they can reach maximum photosynthetic capacity more efficiently and in less time. This trait is likely the most important physiological adaptation in naturalized biotypes, and a vital reason for greater survival and performance relative to ornamental cultivars. Our results are consistent with the physiological responses reported for other invasive grasses such as Microstegium vimineum (Trin.) (Japanese Stiltgrass; Horton and Neufield 1998, Spencer 2012), and for naturalized Chinese Silvergrass (Horton et al. 2010). These exotic C4 grasses have formed invasive populations in the eastern US, and may exhibit enhanced shade tolerance. Generally, C4 plants do not adapt well to low-light Northeastern Naturalist Vol. 22, No. 2 R.F. Dougherty, L.D. Quinn, T.B. Voigt, and J.N. Barney 2015 383 environments relative to C3 species, but recent evidence suggests that C4 species such as Japanese Stiltgrass may actually have a competitive advantage in temperate understory habitats (Horton and Neufield 1998). Tolerance to low soil-moisture availability and drought conditions, as in Panicum virgatum L. (Switchgrass), can enhance naturalization potential across a broad range of habitats (Barney et al. 2009). Differences in tolerance to soil-moisture availability have been found between native and invasive species (McAlpine et al. 2008, Schumacher et al. 2008), as well as between cultivars of the same species (Prunty 1981), and we expected to identify similar differences in performance in Chinese Silvergrass. Chinese Silvergrass is anecdotally considered a drought-tolerant species (Quinn et al. 2012, Stewart et al. 2009), although little empirical evidence exists to support this notion. Others have found that Chinese Silvergrass is the most drought-tolerant of the Miscanthus genus (Clifton-Brown and Lewandowski 2000, Clifton-Brown et al. 2002), but we questioned how the species has responded to artificial selection and whether or not naturalized biotypes have an increased drought tolerance relative to ornamental cultivars. Our results show that soil moisture generally did not have a significant effect on morphological or physiological responses for naturalized biotypes (Table 3). Basal diameter decreased under the low- and medium-low soil-moisture treatments, although a significant interaction effect between light and soil-moisture treatments (P < 0.0001) suggested that this significant decrease was more likely linked to the combination of treatments rather than soil moisture alone. Differences in basal diameter between cultivars were possibly a result of artificially selected traits such as growth habit. Mortality in the low-moisture treatment (32%) was only slightly higher than in the high-moisture treatment (19%). Overall, naturalized biotypes were more tolerant to soil-moisture stress than ornamental cultivars, but regardless of biotype, there were no significant differences in mortality among the soil-moisture-availability treatments. This finding is in stark contrast to Barney et al.’s (2009) report of decreases in Switchgrass biomass, tiller height, specific leaf area, and survival across a similar water potential gradient of 0.0 to -4.0 MPa. Low soil-moisture availability has been shown to decrease growth and survival of other invasive species under low light-availability (Schumacher et al. 2008); a trend that our results did not support for Chinese Silvergrass. Mortality and changes in morphology did not vary between soil-moisture treatments (Table 2). These results support our prediction that Chinese Silvergrass is not only drought tolerant, but significantly more drought tolerant in naturalized than ornamental biotypes. Finally, we also predicted that the response of ornamental Chinese Silvergrass cultivars to light and water stresses would fall along a continuum—some ornamental cultivars would be more shade and drought tolerant than others, and subsequently be of higher risk for invasion. Our results show the “Graziella” cultivar possessed the greatest tolerance to shade and soil-moisture stress and performed most like the naturalized biotypes. We also found that “Variegatus” and “Zebrinus” were the least tolerant to these stresses and subsequently pose the lowest risk of Northeastern Naturalist 384 R.F. Dougherty, L.D. Quinn, T.B. Voigt, and J.N. Barney 2015 Vol. 22, No. 2 invasiveness. Though the ornamental cultivars we chose are some of the most commonly available, it is important to note that our results only represent 7 of the more than 100 ornamental cultivars of Chinese Silvergrass. Because significant differences in shade tolerance exist among a small portion of the total number of Chinese Silvergrass cultivars, it is logical to assume that other ornamental cultivars may be even more, or less, shade tolerant. Future studies of ornamental Chinese Silvergrass should include evaluations of a much more diverse selection of cultivars. Previous studies have also found that shade tolerance does not always pass through to progeny and some offspring of shade-tolerant parents are often shade intolerant (McAlpine and Jesson 2007). Chinese Silvergrass is also an obligate outcrosser, which means that hybridization within and between naturalized biotypes may select for or against enhanced shade tolerance. It is also possible that naturalized biotypes may hybridize with certain ornamental cultivars under the right circumstances. We suggest that future studies examine the reproductive output of Chinese Silvergrass under low-light stress and identify the long-term consequences, including seed set, germination, and inheritance of shade tolerance. In conclusion, Chinese Silvergrass expresses shade-tolerant traits such as increased leaf width, high photosynthetic efficiency, and most importantly, survival, in low-light environments. These characteristics, combined with its broad tolerance to drought stress, enhance the ability of Chinese Silvergrass to establish and naturalize in the eastern US in habitats of varying light and water availability. Artificial selection and breeding of ornamental cultivars result in phenotypic variation and differences in response to environmental stress and potential invasiveness, including shade and drought tolerance. The vast majority of naturalized biotypes of Chinese Silvergrass are found in high-light areas such as roadsides and forest edges, but many populations are also found in forest understories and other habitats with low light and water availability. Identifying ornamental cultivars of high and low risk for potential invasion is an essential step in management and control of Chinese Silvergrass. Repeated introduction of the most shade- and drought-tolerant ornamental cultivars could add genetic variation to the existing naturalized biotypes, which may accelerate the expansion of Chinese Silvergrass populations into forest understories and other low-light areas. Acknowledgments This work was supported by the Energy Biosciences Institute. Literature Cited Barney, J.N., J.J. Mann, G.B. Kyser, E. Blumwald, A. Van Deynze, and J.M. DiTomaso. 2009. Tolerance of Switchgrass to extreme soil-moisture stress: Ecological implications. Plant Science 177:724–732. Clifton-Brown, J.C., and I. Lewandowski. 2000. Water-use efficiency and biomass partitioning of three different Miscanthus genotypes with limited and unlimited water supply. Annals of Botany 86:191–200. Northeastern Naturalist Vol. 22, No. 2 R.F. Dougherty, L.D. Quinn, T.B. Voigt, and J.N. Barney 2015 385 Clifton-Brown, J.C., I. Lewandowski, F. Bangerth, and M.B. Jones. 2002. Comparative responses to water stress in stay-green, rapid- and slow-senescing genotypes of the biomass crop, Miscanthus. New Phytologist 154:335–345. Conklin, J.R., and J.C. Sellmer. 2008. Flower and seed production of Norway Maple cultivars. HortTechnology 19:91–95. Culley, T.M., and N.A. Hardiman. 2007. The beginning of a new invasive plant: A history of the ornamental Callery Pear in the United States. BioScience 57:956–964. Dehnen-Schmutz, J.T.K., C. Perrings, and M. Williamson. 2007. The horticultural trade and ornamental plant invasions in Britain. Conservation Biology 21:224–231. Dougherty, R.F., L. Quinn, A. Endres, T. Voigt, and J.N. Barney. 2014. Natural history survey of the ornamental grass Miscanthus sinensis in the invaded range. Invasive Plant Science and Management 7:113–120. Grounds, R. 1998. The Plantfinder’s Guide to Ornamental Grasses. Timber Press, Portland, OR. [#PP]. Horton, J.L., and H. Neufield. 1998. Photosynthetic responses of Microstegium vimineum (Trin.) A. Camus, a shade-tolerant, C4 grass, to variable light environments. Oecologia 114:11–19. Horton, J.L., R. Fortner, and M. Goklany. 2010. Photosynthetic characteristics of the C4 invasive exotic grass Miscanthus sinensis Andersson growing along gradients of light intensity in the Southeastern United States. Castanea 75:52–66. Kitajima, K., A.M. Fox, T. Sato, and D. Nagamatsu. 2006. Cultivar selection prior to introduction may increase invasiveness: Evidence from Ardisia crenata. Biological Invasions 8:1471–1482. Lehrer, J.M., M.H. Brand, and J.D. Lubell. 2006. Four cultivars of Japanese Barberry demonstrate differential reproductive potential under landscape conditions. HortScience 41:762–767. Leung, B., D.M. Lodge, D. Finnoff, J.F. Shogren, M.A. Lewis, and G. Lamberti. 2012. An ounce of prevention or a pound of cure: Bioeconomic-risk analysis of invasive species. Proceedings of the Royal Society of London B Biological Sciences 269:2407–2413. Madeja, G., L. Umek, and K. Havens. 2012. Differences in seed set and fill of cultivars of Miscanthus grown in USDA cold hardiness zone 5 and their potential for invasiveness source. Journal of Environmental Horticulture 30:42–50. Matlaga, D.P., L.D. Quinn, A.S. Davis, and J.R. Stewart. 2012. Light response of native and introduced Miscanthus sinensis seedlings. Invasive Plant Science and Management 5:363–374. McAlpine, K.G., and L.K. Jesson. 2007. Biomass allocation, shade tolerance, and seedling survival of the invasive species Berberis darwinii (Darwin’s barberry). New Zealand Journal of Ecology 31:1–12. McAlpine, K.G., L.K. Jesson, and D.S. Kubien. 2008. Photosynthesis and water-use efficiency: A comparison between invasive (exotic) and non-invasive (native) species. Austral Ecology 33:10–19. Prunty, L. 1981. Sunflower cultivar performance as influenced by soil water and plant population. Agronomy Journal 73:257–260. Quinn, L.D., J.R. Stewart, T. Yamada, Y. Toma, M. Saito, K. Shimoda, and F.G. Fernández. 2012. Environmental tolerances of Miscanthus sinensis in invasive and native populations. BioEnergy Research 5:139–148. Reed, S. 2002. Flowering performance of 21 Hydrangea macrophylla cultivars. Journal of Environmental Horticulture 20:155–160. Northeastern Naturalist 386 R.F. Dougherty, L.D. Quinn, T.B. Voigt, and J.N. Barney 2015 Vol. 22, No. 2 Reichard, S.H., and P. White. 2001. Horticulture as a pathway of invasive plant introductions in the United States. BioScience 51:103–113. Schumacher, E., C. Kueffer, M. Tobler, V. Gmür, P.J. Edwards, and H. Dietz. 2008. Influence of drought and shade on seedling growth of native and invasive trees in the Seychelles. Biotropica 40:543–549. Spencer, D. 2012. Response of Giant Reed (Arundo donax) to intermittent shading. Invasive Plant Science and Management 5:317–322. Stewart, J., Y.O. Toma, F.G. Fernández, A.Y.A. Nishiwaki, T. Yamada, and G. Bollero. 2009. The ecology and agronomy of Miscanthus sinensis, a species important to bioenergy crop development, in its native range in Japan: A review. GCB Bioenergy 1:126–153. Trueblood, C. 2009. An invasive-species assessment system for the North Carolina horticultural industry. North Carolina State University. Available online at http://www4.ncsu. edu/~jcneal/Website/index.html. Accessed November 2012. US Forest Service. 2006. Weed of the week: Chinese Silvergrass (Miscanthus sinensis Anderss.). Available online at http://www.na.fs.fed.us/fhp/invasive_plants/weeds/chinesesilvergrass. pdf. Accessed November 2012. Valladares, F., and U. Niinemets. 2008. Shade tolerance: A key plant feature of complex nature and consequences. Annual Review of Ecology and Systematics 39:237–257. Vilà, M., J.L. Espinar, M. Hejda, P.E. Hulme, V. Jarošík, J.L. Maron, J. Pergl, U. Schaffner, Y. Sun, and P. Pyšek. 2011. Ecological impacts of invasive alien plants: A meta-analysis of their effects on species, communities, and ecosystems. Ecology Letters 14:702–708. Wilson, S., and L. Mecca. 2003. Seed production and germination of eight cultivars and the wild type of Ruellia tweediana: A potentially invasive ornamental. Journal of Environmental Horticulture 21:137–143. Zub, H.W., and M. Brancourt-Hulmel. 2010. Agronomic and physiological performances of different species of Miscanthus, a major energy crop: A review. Agronomy of Sustainable Development 30:201–214.