Intraspecific variation among clones of a naïve rare grass affects competition with a nonnative, invasive forb


  • David J. Gibson,

    Corresponding author
    1. Department of Plant Biology, Center for Ecology, 1125 Lincoln Avenue, Southern Illinois University Carbondale, Carbondale, Illinois
    • Correspondence

      David J. Gibson, Department of Plant Biology, Center for Ecology, 1125 Lincoln Avenue, Southern Illinois University Carbondale, Illinois 62901-6509.

      Tel: 618 453 3231; Fax: 618 453 3441;


    Search for more papers by this author
  • Justin Dewey,

    1. Department of Plant Biology, Center for Ecology, 1125 Lincoln Avenue, Southern Illinois University Carbondale, Carbondale, Illinois
    Search for more papers by this author
  • Hélène Goossens,

    1. Department of Plant Biology, Center for Ecology, 1125 Lincoln Avenue, Southern Illinois University Carbondale, Carbondale, Illinois
    Search for more papers by this author
  • Misty M. Dodd

    1. Department of Plant Biology, Center for Ecology, 1125 Lincoln Avenue, Southern Illinois University Carbondale, Carbondale, Illinois
    Search for more papers by this author


Intraspecific variation can have a major impact on plant community composition yet there is little information available on the extent that such variation by an already established species affects interspecific interactions of an invading species. The current research examined the competitiveness of clones of a globally rare but locally common native grass, Calamagrostis porteri ssp. insperata to invasion by Alliaria petiolata, a non-native invasive species. A greenhouse experiment was conducted twice over consecutive years in which 15 clones from three populations of Calamagrostis were paired with rosettes of Alliaria in pots containing native forest soil previously uninvaded by Alliaria. Both species showed a negative response to the presence of the other species, although Alliaria more so than Calamagrostis. Moreover, the effect of Calamagrostis depended upon population, and, to a lesser extent, the individual clone paired with Alliaria. Competitive effects were stronger in the first experiment compared with when the experiment was repeated in the second year. The influence of Calamagrostis clones on the outcome of the experiment varied among populations and among clones, but also between years. Clones from one of the three populations were more influential than clones from the other two populations. Only one of 15 clones, both from the same population, was influential in both experiments. This research supports a growing literature indicating that intraspecific variability among clones of a dominant species can affect interspecific interactions and that such variability in a native species can affect performance of an invading species.


Understanding the importance of individual variation on population and community dynamics, and the upscaling of processes at the level of individuals to population scale patterns are key ecological challenges (Sutherland et al. 2013). In this vein, intraspecific variation among competing neighbors is known to affect species diversity in plant communities (Booth and Grime 2003; Vellend 2006; Gibson et al. 2012). These effects arise as intraspecific variability mediates local-scale interspecific interactions (Turkington 1996). A consequence of these interactions is that potential niche space availability may vary depending upon the presence or number of certain genotypes of dominant species in the community. These dominant species are those that make a substantial contribution to plant biomass in a community by virtue of their large relative size and high frequency of occurrence. The mass ratio hypothesis (MRH) states that the traits and functional diversity of dominant species in a community play a primary role in development of community and ecosystem function (Grime 1998). Thus, if a dominant species in the community exhibits ecologically relevant intraspecific variability, then the opportunity for establishment of a new species in a community may depend upon the particular genotypes of the dominant species that are present (Crutsinger et al. 2010; Adams et al. 2011).

It has been suggested that genetic diversity and the occurrence of species modulates invasion resistance (Crutsinger et al. 2008; Vellend et al. 2010) and can be considered a development of Elton's diversity-resistance hypothesis (Elton 1958). Infact, recent plant community models place primacy on the importance of genetic variability affecting the competitive ability of dominant species as determinants of invasion resistance (Vellend 2006; Gibson et al. 2012). While these models consider the outcome of these interactions on native species diversity (Fridley et al. 2007; Whitlock et al. 2007, 2011; Fridley and Grime 2010), less is known about how intraspecific variability of a dominant species affects competitive response and effect (sensu Goldberg and Landa 1991) with invasive species. The expectation from these conceptual models and the well-established principles of natural selection (Darwin 1859) is that some individuals in a native species population would be better competitors against an invasive non-native species than others.

In this study, we challenged individuals of the exotic, non-native (to North America) invasive species Alliaria petiolata (Bieb.) Cavara and Grande (hereafter Alliaria) with clones of a globally rare, perennial grass, Calamagrostis porteri ssp. insperata (Swallen) C. Greene (hereafter Calamagrostis) collected from three separate populations in Illinois, USA. We tested the hypothesis that the competitive effect and response (sensu Goldberg and Landa 1991) of Calamagrostis on Alliaria would be affected by differences among Calamagrostis populations or clones within each population, or both. Alliaria is highly competitive in mixtures against some native North American species (Meekins and McCarthy 1999), but has yet to invade Calamagrostis populations. Calamagrostis has not been exposed to Alliaria and is thus a naïve competitor with respect to Alliaria thereby removing the potential bias of competition between an exotic and a native species after the opportunity for either to have been subject to selection imposed by one on the other (Strauss et al. 2006). Calamagrostis populations have a low genetic diversity (Esselman et al. 1999) making the potential importance of significant effects of individual clones on competitive outcome between these two species more relevant than if we had ‘stacked the deck’ by carrying out this experiment with a highly diverse native species. When the genetic diversity of one of an interacting species pair is low, then genotype effects should be correspondingly low in comparison with interactions among species with high genetic diversity.

Materials and Methods

Study species

Calamagrostis porteri ssp. insperata (Poaceae) is a cool-season, loosely clumped, perennial grass limited in its global distribution to approximately 80 populations in five US states where it is considered endangered in Arkansas, Illinois, Kentucky, Maryland, Ohio, and Tennessee, threatened in Indiana. Calamagrostis occurs in forest openings and upland oak woodland throughout its range. Where Calamagrostis populations occur, the plant occurs as common scattered clumps in the forest understory. Vegetative growth is generally vigorous and determined by light, humidity, and soil temperature, with the rare flowering events requiring high early-mid growing season light and soil moisture (Bittner and Gibson 1998; Gibson et al. 2009). When Calamagrostis plants do flower, less than 1% of florets produce viable seed (Havens and Holland 1998). Molecular analysis of the Illinois populations of Calamagrostis based on ISSR markers showed high within-population genetic similarity, and low among-population genetic similarity (Marriage 2002). Despite low within-population genetic diversity, Calamagrostis clones 1 m or more apart are genetically unique likely reflecting relictual diversity from more frequent sexual reproduction in the past or more recent somatic mutation.

Alliaria petiolata (Brassicaceae) is a highly competitive, self-compatible, invasive non-native biennial herb widespread in deciduous forest understories and forest edges throughout the eastern North America. Native to Europe, molecular studies indicate that Alliaria appears to have been introduced multiple times to the US in the early 19th century (Durka et al. 2005), and it was first observed in southern Illinois in 1988 (Gibson et al. 2006), but has yet to invade any of the three Calamagrostis populations in the state despite being recorded from one of the sites (i.e., Lusk Creek Canyon Natural Area). Germinating in early spring, first year plants overwinter as a basal rosette before bolting, flowering and setting seed in late-spring before dying. Alliaria competes against native plants through early season flowering and seed set ahead of most native plants (Anderson et al. 1996), and the production of allelochemicals which can disrupt mycorrhizal associations (Stinson et al. 2006; Wolfe et al. 2008) and suppress native species recruitment (McCarthy 1997; Prati and Bossdorf 2004; Stinson et al. 2007).

Greenhouse experiment

Fifteen to seventeen Calamagrostis clones from each of the three Illinois populations at Bell Smith Springs Ecological Area (BSS: 37°31'16”N, 88°39'42”W), Lusk Creek Canyon Natural Area (LC: 37°30'57”N, 88°32'26”W), and Hays Creek Canyon (HC: 37°29'12”N, 88°36' 6”W) were haphazardly selected and collected in mid-summer 2002. Each clone that was collected consisted of a clump of physically connected rhizomes, roots, and attached vegetative tillers. Clumps were separated by at least 5 m and prior genetic profiling of the populations (Marriage 2002) and subsequent reaction norms on greenhouse grown clones (Gibson et al. 2009) indicated that the clumps were most likely separate genotypes. Thereafter, the clones were maintained and subdivided as necessary in individual 17.7-cm by 15.2-cm pots in a greenhouse to obtain genetically identical clonal replicates. The plants were periodically clipped to 9 cm above the soil surface and repotted so that individuals did not become root bound in the pots. Samples of these clones were used in a greenhouse and field experiment to test the effects of light availability and soil moisture (Gibson et al. 2009). There were no observable negative effects on the clones maintained under these conditions from the time of collection (2002) until the experiment was established (2008–2009). In October 2008, shortly before establishment of this experiment, the Calamagrostis plants were treated for leaf spot with the systemic benzimidazole fungicide Benomyl; the presence of pathogens was not observed thereafter.

Vegetative first year rosettes of Alliaria were collected from a large population (several hundred individuals) on the Southern Illinois University Carbondale campus, Jackson County, Illinois in early autumn 2008 and 2009. These plants were placed within 24 hr into experimental pots.

The experiment was established in 15-cm-diameter plastic pots filled with a 50:50 mixture of forest soil and sterilized masonry sand. The soil:sand mixture had pH 5.3, CEC 13.2 meq/100 g, 1.2% organic matter, 90 μg L−1 P, 132 μg L−1 K, 1493 μg L−1 Ca, and 376 μg L−1 Mg (A&L Analytical Laboratories, Inc., Memphis, Tn). The forest soil was collected from a Quercus velutina/Q. alba dominated area on the SIUC campus >100 m from the nearest Alliaria population. A pairwise competition experimental design was established (Gibson 2002) with monocultures of single Calamagrostis clones (7–23 tillers each when planted) or three Alliaria rosettes (1–8 leaves per plant) randomly assigned to each pot and mixtures of a central Calamagrostis clone surrounded by three Alliaria rosettes (Fig. 1). This design has proven to be useful for investigating invasive-native species interactions (Vilà and Weiner 2004) as well as pairwise competitive interactions in a variety of controlled settings (Gibson et al. 1999; Connolly et al. 2001). Experimental treatments were Calamagrostis source population (n = 3), Calamagrostis clone (n = 5 per source population), and competition (monoculture of Calamagrostis or Alliaria, or mixture). The particular Calamagrostis clones chosen for this experiment from the collection maintained in the greenhouse were randomly chosen from those with sufficient replicates of approximately the same size. There were 3 replicates per treatment combination giving a total of 135 experimental units randomly located by replicate on one of three greenhouse benches. Pots were bottom watered daily with an automated watering system with supplemental top watering by hand as necessary to main adequate soil moisture and avoid drought stress. The experiment was conducted twice to investigate consistency of clonal effects on competition with Alliaria. Calamagrostis plants were given a 2-week establishment period in the pots before Alliaria plants were planted. In year 1, Calamagrostis was planted November 10 to November 11, 2008 with Alliaria planted into pots on November 17, 2008. This experiment ran through May 29, 2009 when a final harvest was conducted. In year 2, Calamagrostis was planted September 29 through October 2, 2009, with Alliaria planted on October 13, 2009. Final harvest for the year 2 experiment was April 22 to April 30, 2010. Eight Alliaria plants died by February 16 in year 1 and these were not replaced. However, mortality was higher in year 2 and 77 dead plants were replaced through January 6, 2010. Occasional outbreaks of aphids and white flies on the Alliaria were controlled by spraying with the insect growth regulator Enstar® (active ingredient (S)-kinoprene) and the pyrethroid insecticide Talstar® (active ingredient bifenthrin). The experiment was conducted in the SIUC Tree Improvement Center Greenhouse. Temperature and supplemental greenhouse lighting were controlled to mimic late autumn sunlight and temperature conditions, which ranged 13–30 °C over the course of the experiments.

Figure 1.

An experimental unit showing a single clone of Calamagrostis porteri ssp. insperata growing in the center of a 15-cm-diameter pot surrounded by three plants of Alliaria petiolata.

Several response variables were measured to capture the dynamics of competitive interactions (Gibson et al. 1999; Trinder et al. 2013): On Calamagrostis clones, tiller counts and leaf number per plant were recorded at approximately three weekly intervals five times over the course of the experiment each year. For Alliaria, leaf number, leaf width (maximum width of largest leaf), plant height per plant, and the number of surviving plants per pot (of three planted) were recorded five times. At final harvest, individual plants were washed free of soil, separated into above- and belowground tissues, dried at 55 °C for >48 h, and weighed.

Data analysis

Growth of Calamagrostis and Alliaria was investigated separately on the mean value of response variables per pot with a mixed model analysis using proc MIXED in SAS, version 9.2 (SAS Institute Inc. 2002-2008) testing the main effects and the interaction of Calamagrostis source population (n = 3) and competition treatment (n = 2), with the main effect of Calamagrostis clone nested within source population. A repeated measures mixed model including the effect of time of measurement (days or weeks since start of the experiment) was conducted on variables measured over time. Model degrees of freedom were estimated using the Satterthwaite approximation, or the Kenward-Roger correction for repeated measures, and the appropriate covariance structure were incorporated in the analysis following Littell et al. (2006). The three greenhouse benches that pots were placed on were included as random factors. To run a full mixed model, it was necessary to assign pots without Calamagrostis, that is., Alliaria monoculture controls to a Calamagrostis source population and clone. This assignment was made randomly ten separate times with the mixed model being run following each assignment. Average F statistics for each treatment combination obtained from each of the ten random assignments and model runs was retained and tested for significance at α = 0.05. In addition, a restricted analysis of the effect of Calamagrostis clone (nested in source population) on mixtures where Alliaria was paired in a pot with a specific Calamagrostis clone (i.e., without monocultures) was undertaken to more precisely investigate the clone effect. Results of both the full model and the restricted model are reported. Separate analyses were undertaken for the experiments in each year (hereafter year 1 and year 2).

Cook's D influence statistics were calculated using the influence option of proc MIXED in SAS, version 9.2 (SAS Institute Inc. 2002-2008) to determine the Calamagrostis clone that was most influential on plant performance (i.e., a response variable) that the mixed models identified as showing a significant Calamagrostis clone effect on either Calamagrostis or Alliaria. Cook's D statistic (Cook 1977) is used in the detection of influential observations in linear regression and linear mixed models (e.g., Schowalter et al. 1991) and is calculated as the distance between original log likelihoods based on all observations and on removing, in this case, one Calamagrostis clone at a time from the dataset. D statistics of Calamagrostis clones exceeding the 90th percentile of all calculated values were considered to indicate that a clone had a significant effect on the analysis based upon the particular response variable being tested.


Competitive response of Calamagrostis to Alliaria

Regardless of whether plants were competing with Alliaria or not, Calamagrostis clones and populations grew significantly different to each other in both years of the experiment (Table 1). Recognition of interclonal or interpopulation differences depended upon the growth parameter and was not consistent from the first to the second year (Figs 2, 3). For example, BSS clones did not differ in terms of aboveground biomass in year 1, but showed clear differences among clones in year 2. Similarly, no populations showed differences among clones in terms of belowground biomass in year 1, whereas clones from all three populations exhibited differences in year 2. Clone H2 from HC was generally the largest clone regardless of response variable measured over the two years. There was not a clearly identifiable low performing clone. Overall, plants from BSS were larger than plants from HC, which were generally larger than plants from LC (Figs 2, 3). However, in year 2, HC plants had more tillers (21.0 ± 1.1) and leaves (66.3 ± 3.7) per plant than plants from the other populations (tillers: BSS 16.2 ± 1.1, LC 15.5 ± 1.0; leaves: BSS 46.8 ± 2.7, LC 46.8 ± 2.67).

Table 1. Competitive response of Calamagrostis to Alliaria. F statistics from mixed model analysis testing the effects of Calamagrostis source population (P), clone (nested within population), and competition (C) with or without Alliaria through time (T) on final biomass and leaf and tiller number of Calamagrostis.
 NumDF/DenDFBiomassLeaf numberTiller number
  1. Num DF, numerator degrees of freedom. Denominator degrees of freedom (Den DF) = 65 for analysis without repeated measures, and as shown following NumDF for leaf number and tiller number analyzed with repeated measures analysis, respectively.

  2. a

    P < 0.1.

  3. b

    P < 0.05.

  4. c

    P < 0.01.

  5. d

    P < 0.001.

Year 1
Population (P)2/70.7, 52.17.92d3.843.46b3.32b
Competition (C)1/70.7, 52.113.73d1.390.880.18
P * C2/70.7, 52.11.351.990.650.60
Time (T)4/333, 333  919.89d556.47d
P * T8/333, 333  0.871.09
C * T4/333, 333  1.453.60c
P * C * T8/333, 333  0.750.99
Clone12/70.7, 52.11.92b1.351.85a1.21
Year 2
Population (P)2/59.6, 62.28.84d0.197.18c7.31c
Competition (C)1/59.2, 59.02.92a0.580.520.07
P * C2/59.1, 58.90.562.220.040.51
Time (T)3/221, 229  128.65d205.72d
P * T6/221, 229  3.50c1.62
C * T3/221, 229  0.940.83
P * C * T6/221, 229  1.601.44
Clone12/58.8, 59.73.25d3.97d2.31b2.04b
Figure 2.

Mean (± SE) Calamagrostis response by population and clone, year 1, (A) Bell Smith Springs, (B) Hayes Creek Canyon, and (C) Lusk Creek Canyon. Bars sharing the same letter, or no letters, among clones and within populations are not significantly different (= 0.05).

Figure 3.

Mean (± SE) Calamagrostis response by population and clone, year 2, (A) Bell Smith Springs, (B) Hayes Creek Canyon, and (C) Lusk Creek Canyon. Bars sharing the same letter, or no letters, among clones and within populations are not significantly different (= 0.05).

A response of Calamagrostis to competition with Alliaria occurred in both years affecting aboveground biomass and number of tillers per plant in year 1 and marginally affecting aboveground biomass in year 2 (Table 1). In all cases, where a response was observed, the presence of Alliaria reduced growth of Calamagrostis (Figs 4, 5), although the response illustrated by tiller number was only evident at the end of the experiment in year 1 (Fig. 5).

Figure 4.

Mean (± SE) Calamagrostis response to competition treatments (Control = Calamagrostis monoculture, Plus Alliaria = Calamagrostis-Alliaria mixture; (A, B) aboveground biomass, and (C, D) belowground biomass in years 1 and 2, respectively. Symbols above pairs of bars indicate a significant difference between the means (+< 0.1, *< 0.05); on bars with no symbols there were no differences among treatments.

Figure 5.

Mean (± SE) Calamagrostis tiller number in response to competition treatments (Calamagrostis monoculture and in mixture including Alliaria) through time. There was a significant difference in tiller number per Calamagrostis plant in monoculture versus in mixture with Alliaria 25 days after planting in year (*< 0.05).

Competitive effect of Calamagrostis on Alliaria

The full mixed model suggested that neither Calamagrostis population nor clones within populations affected Alliaria growth in either year of the experiment (Table 2). However, the restricted model analysis showed that Calamagrostis clone nested within population affected Alliaria leaf width and height in year 1 and number of leaves per plant in year 2 (Table 2, and see Appendices S1 and S2). Competition with Calamagrostis regardless of population or clone reduced growth of Alliaria in year 1 affecting all measured variables (Fig. 6). In year 2, Alliaria plants grown in competition with Calamagrostis had fewer, larger leaves, and showed higher survivorship (Fig. 7) compared with Alliaria plants grown in the absence of Calamagrostis. Final biomass of Alliaria was unaffected by Calamagrostis (Fig. 6). Overall, Alliaria plants in year 2 were smaller than plants in year 1. Biomass of Alliaria in year 1 was twice that of Alliaria in year 2 and was reduced by competition in year 1 but not year 2.

Table 2. Competitive effect of Calamagrostis on Alliaria. F statistics from repeated measures mixed model analysis testing the effects of Calamagrostis population source (P), clone (nested within population), and competition (C) on leaf number, leaf width, height, and number of living plants through time (T), and final above- and belowground biomass of Alliaria. The restricted Calamagrostis clone analysis tested the subset of the experimental pots in which both Alliaria and Calamagrostis were planted together (i.e., not including the Calamagrostis or Alliaria monocultures).
 Num DF/Den DFBiomassLeaf numberLeaf widthHeightNumber of living Alliaria plants
  1. Num DF, numerator degrees of freedom. Denominator degrees of freedom (Den DF) = 68 for analysis without repeated measures, and the range of degrees of freedom values estimated using the Kenward-Roger correction for leaf number, leaf width, height, and number of living plants analyzed with repeated measures analysis are shown following Num DF.

  2. a

    P < 0.05.

  3. b

    P < 0.01.

  4. c

    P < 0.001.

Year 1
Population (P)2/69–1240.720.790.201.211.371.03
Competition (C)1/69–12469.66c15.23c55.89c62.66c29.30c2.42
P * C2/69–1240.841.530.191.611.241.12
Time (T)4/301–309  67.84c98.93c85.37c43.75c
P * T8/69–124  0.460.780.590.92
C * T4/301–309  8.11c4.95c3.99b2.55a
P * C * T8/312–334  0.620.810.711.12
Calamagrostis clone12/72–1280.710.620.461.160.990.73
Calamagrostis clone restricted analysis12/30.9–55.92.14a0.660.762.20a2.10a1.15
Year 2
Population (P)2/73–1890.821.990.650.250.610.36
Competition (C)1/72–1890.420.592.0110.36b1.921.00
P * C2/72–1891.
Time (T)5/361–395  29.06c86.00c89.51c103.38c
P * T10/385–405  1.421.551.231.05
C * T5/343–396  17.59c19.26c26.06c1.66
P * C * T10/370–405  1.261.571.350.85
Calamagrostis clone12/71–1981.
Calamagrostis clone restricted analysis12/30.6–54.91.460.941.93a0.910.560.46
Figure 6.

Effect of Calamagrostis competition (Control = Alliaria monoculture, Plus Calamagrostis = mixture including Calamagrostis) on Alliaria mean (± SE) (A, B) aboveground biomass, and (C, D) belowground biomass in years 1 (A, B) and 2 (B, D). Symbols above pairs of bars indicate a significant difference between the means (*< 0.05); on bars with no symbols there were no differences among treatments.

Figure 7.

Effect of Calamagrostis competition (Control = Alliaria monoculture, Plus Calamagrostis = mixture including Calamagrostis) on Alliaria [mean (± SE)] in years 1 and 2, (A, E) number of leaves per plant; (B, F) leaf width; (C, G) height, and (D, H) number of surviving Alliaria plants per pot. A * symbol above pairs of points indicate a significant difference in mean values at the number of weeks since planting shown (*< 0.05).

Competitive Influence of Calamagrostis clones

Cook's D statistics identified the most influential Calamagrostis clones (highest Cook's D statistics) either increasing or decreasing performance in the experiment more so than other clones. Calamagrostis clones from BSS were more influential in affecting growth of Calamagrostis in intraspecific mixture or Alliaria in interspecific mixtures with Calamagrostis than were clones from HC or LC (Table 3). Five of the fifteen Calamagrostis clones were identified as being influential (with Cook's D statistics exceeding the 90th percentile of values; Table 3). Four of these clones were from BSS (i.e., clones B6, B10, B13, and B14), one from HC (clone H6), and none from LC. However, of these, only clones B6, B10, and B14 from BSS affected more than one growth parameter. Alliaria was affected by three BSS clones (B10, B13, and B14), one HC clone (H6), but no LC clones. The Calamagrostis clones identified as having a significant influence did not necessarily reduce growth. For example, the Calamagrostis clone with the largest influence on belowground biomass of Calamagrostis (clone B14) had the lowest biomass in mixture compared with other clones from the BSS population (Fig. 3). There was little evidence of year-to-year consistency among the Calamagrostis clones identified as being the most influential in affecting growth with only one of fifteen clones (B10) being influential in affecting performance of Calamagrostis in both years. No Calamagrostis clones were influential in affecting growth of Alliaria in both years.

Table 3. Cook's D statistics indicating influence of Calamagrostis clones on Calamagrostis and Alliaria response variables that affected significantly performance (Calamagrostis clone effect in Tables 1 and 2). Cook's D statistics exceeding the 90th percentile (i.e., D ≥ 0.20) of all values are in bold.
Clone Calamagrostis Alliaria
Year 1Year 2Year 1Year 2
Aboveground biomassLeaf numberAboveground biomassBelowground biomassLeaf numberTiller numberAboveground biomassLeaf widthHeightLeaf number
Bell Smith Springs
B60.060.14 0.20 0.12 0.40 0.27
B10 0.23 0.22 0.28 0.14 0.26 0.29 0.13
B130. 0.54
B140.070.030.01 0.26 0.180.08< 0.28
Hayes Creek
H60. 0.24
Lusk Creek

Year-to-year ranking in growth

Regardless of competitive environment, there was a positive correlation between aboveground biomass of Calamagrostis clones from year 1 to year 2 across all populations (Spearman's paired rank correlation = 0.24, = 0.02, Fig. 8A), and for clones from HS in terms of both aboveground and belowground biomass (= 0.44, 0.37, respectively, < 0.05, Fig. 8A,B). In other words, the largest Calamagrostis clones in year 1 were also the largest clones in year 2, and vice versa. Similarly, belowground, Alliaria grew best in the presence of the same set of Calamgrostis clones in both years. Competitive effect of Calamagrostis clones on belowground, but not aboveground biomass of Alliaria was positively correlated from year 1 to year 2 (= 0.35, = 0.02, Fig. 8C,D) across all populations. Within Calamagrostis populations, the effect of Calamagrostis clones on belowground biomass of Alliaria was restricted to clones from LC (= 0.62, = 0.02, Fig. 8D).

Figure 8.

Relationship between aboveground (A, C) biomass and belowground (B, D) biomass of (A, B) Calamagrostis and (C, D) Alliaria between year 1 and year 2 by source population of Calamagrostis (BSS = Bell Smith Springs, Hays = Hays Creek Canyon, Lusk = Lusk Creek Canyon). The solid line shows the 1:1 biomass line between year 1 and year 2. Plants on the line had the same biomass in year 1 and year 2. r and P-values shown for significant correlations only (< 0.05) for all populations of Calamagrostis considered together (“all pops”) or by source population.


Intraspecific variability is increasingly recognized as an important internal biotic filter structuring communities (Violle et al. 2011). As part of an emerging community genetics paradigm (Hersch-Green et al. 2011), intraspecific variability adds to niche complexity and can allow coexistence of more species in a community than under strict models based upon concepts of limiting similarity (Abrams 1983; Gibson et al. 2012). In terms of invasion biology, intraspecific variability in members of the native plant community can show differential response to and effects on individual invader species, including Alliaria as we show here. Indeed, evolution of resistance by native species to invaders occurs at the population level (Rowe and Leger 2011), including for Alliaria (Cipollini and Hurley 2008). Ultimately, intraspecific variability in native species can affect resistance to invasion both in terms of the individual invader species and the resulting community (Crutsinger et al. 2008).

Intraspecific variability can affect neighboring species through variation in ecologically relevant morphological or physiological traits that contribute to competitive interactions (Johnson et al. 2008; Kotowska et al. 2010; Whitlock et al. 2010). In the current study, using a random subset of the clones used in Gibson et al. (2009), we show significant differences among clones of Calamagrostis regardless of whether they were competing with Alliaria or not. The ranking of growth of Calamagrostis clones was consistent in both years of the experiment (Fig. 8) as was the effect of Calamagrostis clonal identity on belowground biomass of Alliaria. When Alliaria was present in mixture, growth of Calamagrostis was reduced consistent with Alliaria's known role as a highly competitive invasive species (Meekins and McCarthy 1999). In addition, Calamagrostis generally reduced Alliaria growth in mixture, although this effect was not consistent from 1 year to the next and could not be consistently related to a particular Calamagrostis clone. A stronger effect of Calamagrostis was among source populations where clones from the Bell Smith Springs population had a larger effect on Alliaria than clones from the other two populations. These results suggest that while there is phenotypic intraspecific variation in Calamagrostis, its effect on potentially invading Alliaria may be greater among than within populations. Thus, we would predict that of the Illinois populations of Calamagrostis, the Bell Smith Springs population would show greater resistance to invasion by Alliaria than the Lusk Creek or Hayes Creek Canyon populations. This result is consistent with the proposition that operational importance of intraspecific variance is scale related (Violle et al. 2011).

Repeating the experiment a second year was informative because it showed that Calamagrostis clones that were most influential in affecting competitive response and effect were not consistent from 1 year to the next. This inconsistency and differential response among the clones could be due to uncontrolled differences in the greenhouse conditions from 1 year to the next, or differences among the genotypes of Alliaria used in the experiment from 1 year to the next. Certainly, invasive species, including Alliaria, are reported to evolve rapidly as they invade new habitat (Müller-Schärer et al. 2004). Although Alliaria plants were taken from the same source population each year, as a biennial plant we would have been collecting from a new and different generation of plants in the second year. However, our source Alliaria population would not have been subject to selection from Calamagrostis as the nearest population was > 69 km away, but they could nevertheless, have evolved in response to other selection pressures (Cipollini 2002; Bossdorf et al. 2005). Such a response of Alliaria over a 2-year time period may seem unlikely but would reflect rapid evolution in populations of this invading species. Regardless, the implication is that competitive response of an invasive plant to variability in native plants in the habitat being invaded may be contingent upon response to current environmental conditions (Goldberg 1996).

Mechanistically, we can speculate why some clones of Calamagrostis had more influence than others on the outcome of competition against Alliaria. Both above- and belowground biomass of Calamagrostis was reduced in the presence of Alliaria. However, Alliaria grows as a basal rosette that, except when the flower stem bolts, only minimally overtops and shades the tussock of upright leafy tillers of Calamagrostis (Fig. 1). In year 2, there was reduced aboveground biomass of Alliaria associated with the two Calamagrostis clones that had the largest aboveground biomass (Figs 3, 7, and see Appendix S1). Interspecific competition with Alliaria may occur principally belowground as it can compete against neighbors through allelopathic disruption of the mycorrhizal network (Cantor et al. 2011; Hale et al. 2011) through the action of various flavonoids and glycosides (Cipollini et al. 2008). The susceptibility of Calamagrostis porteri to allelopathically mediated competitive effects is unknown but could vary among clones. The congener C. canadensis is susceptible to allelopathic compounds (Winder 1997). Thus, some Calamagrostis clones may be better able to respond to competition against Alliaria than other clones within the context of allelopathic interactions. An alternative mechanism might simply be that the largest clones of Calamagrostis demand the largest amount of soil moisture in a pot effectively drawing down moisture below the level required by Alliaria for it to compete effectively (Meekins and McCarthy 2000). The converse mechanism is that intraspecific allopathic variation in the invader, Alliaria in this case, differentially affects invasive species success (Lankau 2011).

Regardless of the possible mechanisms of competitive interactions involved in the experiment presented here, we show that in a species of low genetic variability (Calamagrostis), there is nonetheless sufficient intraspecific genotypic variation to elicit differences in competitive effect and response. Even though the genetic variation effects were limited and varied between 1 year of the experiment and the next, the occurrence of variation both within and among populations suggests that invasion success of an exotic (Alliaria here) at the plant-to-plant neighborhood scale is likely to vary depending upon the competitiveness of individual genotypes that are encountered, and when they are encountered. As Alliaria continues to spread and if it reaches Calamagrostis populations, invasion success may be determined by success against particular genotypes of Calamagrostis through alteration of plant–soil feedbacks (Suding et al. 2013). Moreover, the genetic variability in competitiveness among individuals of Calamagrostis that we demonstrate may also be more affective in interactions with co-occurring native species than we found against the exotic Alliaria. In addition, the importance of intraspecific variation in competitiveness may be more relevant in widespread, more genetically diverse species than in the rare Calamagrostis studied here. Regardless, as noted by Wiens (1977), competition may well be temporarily sporadic, but as shown here, genotypically dependent as well.


We thank the USDA Forest Service Shawnee National Forest for allowing access to Calamagrostis sites. J. D. and M. McE. were supported by the SIUC Undergraduate Assistantship and REACH Programs. Rich Cole, Karen Frailey, and John Miller, all helped with the growth of the plants in the greenhouse. JF Cahill, Jr., provided comments on a draft of the manuscript.

Data Accessibility

Plant data for Calamagrostis and Alliaria for both years of the experiment have all been deposited in Dryad ( (DOI: 10.5061/dryad.k68n1).

Conflict of Interest

None declared.