Plant breeders have played an essential role in improving agricultural crops, and their efforts will be critical to meet the increasing demand for cellulosic bioenergy feedstocks. However, a major concern is the potential development of novel invasive species that result from breeders' efforts to improve agronomic traits in a crop. We use reed canarygrass as a case study to evaluate the potential of plant breeding to give rise to invasive species. Reed canarygrass has been improved by breeders for use as a forage crop, but it is unclear whether breeding efforts have given rise to more vigorous populations of the species. We evaluated cultivars, European wild, and North American invader populations in upland and wetland environments to identify differences in vigor between the groups of populations. While cultivars were among the most vigorous populations in an agricultural environment (upland soils with nitrogen addition), there were no differences in above- or below-ground production between any populations in wetland environments. These results suggest that breeding has only marginally increased vigor in upland environments and that these gains are not maintained in wetland environments. Breeding focuses on selection for improvements of a specific target population of environments, and stability across a wide range of environments has proved elusive for even the most intensively bred crops. We conclude that breeding efforts are not responsible for wetland invasion by reed canarygrass and offer guidelines that will help reduce the possibility of breeding programs releasing cultivars that will become invasive.
Citation: Jakubowski AR, Casler MD, Jackson RD (2011) Has Selection for Improved Agronomic Traits Made Reed Canarygrass Invasive? PLoS ONE 6(10): e25757. https://doi.org/10.1371/journal.pone.0025757
Editor: Dorian Q. Fuller, University College London, United Kingdom
Received: June 22, 2011; Accepted: September 11, 2011; Published: October 3, 2011
This is an open-access article, free of all copyright, and may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose. The work is made available under the Creative Commons CC0 public domain dedication.
Funding: This work was funded in part by Grazing Lands Conservation Initiative grant 941-3 and by the United States Department of Energy-Great Lakes Bioenergy Research Center (DOE Office of Science BER DE-FC02-07ER64494). No additional external funding was received for this study. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Plant breeding has played a critical role in increasing food and fiber production throughout the world . Continued crop improvement will be necessary to increase food production and meet the increasing demand of fiber for animal feed, manufacturing and as a cellulosic bioenergy feedstock , , . However, a major concern is the potential development of novel invasive species through breeding efforts and the introduction of exotic crops , , . Many of the traits associated with invasion potential, such as rapid growth rates, plasticity across a range of environments, high yield, and cold and drought tolerance, are targets for improvement in crops , , , , , . In addition, crops are more likely to become naturalized or invasive because their introduction is likely to occur at a scale that will establish large founder populations with sustained establishment efforts .
Recent efforts to assess invasion by crops has focused on potential biofuel feedstock crops by using weed risk assessment modeling , . However, risk assessment is difficult when unknown environmental (climate change) and evolutionary (breeding and hybridization) changes interact to facilitate invasion by a species . A complementary approach to weed risk assessment modeling is to evaluate the history of perennial crop breeding to determine whether improvements in agronomic traits have fostered development of invasive species in the past. There is a long history of perennial crop breeding with a particular focus on grasses and legumes for agriculture . Herbaceous perennial crops that have undergone significant selection efforts include Phalaris arundinacea (reed canarygrass), Schedonorus phoenix (tall fescue), Schedonorus pratensis (meadow fescue), Lolium perenne (perennial ryegrass), Dactylis glomerata (orchardgrass), Bromus inermis (smooth bromegrass), Poa pratensis (Kentucky bluegrass), Phleum pratense (timothy), Medicago sativa (alfalfa), Trifolium pratense (red clover), and Trifolium repens (white clover) . Many of these crops are now ubiquitous in temperate regions across North America and Eurasia. Most have been identified both as beneficial agricultural species and weedy invaders depending on the context , . Using reed canarygrass as our case study, we evaluate the history of introduction and breeding efforts of the species to evaluate the risks associated with breeding perennial crops.
Reed canarygrass is a circumboreal cool-season grass native to North America, Europe, and Asia . The species has been planted for forage since the early 19th Century in North America, actively harvested in Europe since at least 1806 , and recently identified as a potential cellulosic biofuel feedstock , , . Because of the importance of the species to the grazing community, active breeding programs have existed since the early 20th Century with a focus on improving yield, quality, and palatability . However, the species is considered one of the most noxious wetland invaders in the northern United States because it displaces native species and forms monocultures in disturbed wetlands . Hypotheses for the development of reed canarygrass invasiveness in North America include; introduced wild European and Asian populations outcompeted native North American populations , formerly distinct populations crossed to create populations with increased genetic variability and hybrid vigor , and more aggressive cultivars released by plant breeding programs outcompeted or introgressed with wild populations . Many in the invasion science and management community have called for the banning of breeding programs and the sale of seed, although the hypotheses listed above have not been evaluated . For these reasons, reed canarygrass serves as a useful model for evaluating the effects of breeding on the invasiveness of a species.
Here, we address four objectives. First, we evaluate the genetic similarity between European wild populations, cultivars, and North American invaders. Second, we compare the production and fecundity of cultivars, European wild populations, and North American invader populations of reed canarygrass. Third, we determine whether cultivars selected for improved yield in an agricultural setting (upland soils with nitrogen addition) also show improved yield in non-agricultural settings (unfertilized uplands and wetlands). Fourth, we offer guidelines that will help reduce the possibility of breeding programs releasing cultivars that will become invasive.
In the summer of 2008, we established common gardens testing four groups of reed canarygrass populations. The four groups, each consisting of three populations were North American cultivars, European cultivars, European wild, and North American invader populations (Table 1). The populations chosen within the European groups included at least one population from the Scandinavian and central European refugia suggested by previous research . The seed of eight of the twelve populations was obtained from the USDA-ARS Germplasm Resources Information Network (http://www.ars-grin.gov), while the three North American invader populations were collected from invaded wetlands in Iowa, Wisconsin, and New York, USA. Seed of Bamse, a European cultivar, was obtained through a commercial seed dealer. Plants were grown as plugs from seed in the greenhouse and transplanted into each of the common gardens. Differences in germination and establishment vigor were considered population-specific effects and were not normalized prior to planting in the common gardens. The upland common garden was established at the University of Wisconsin OJ Noer Turfgrass Research Facility in Verona, WI in June 2008. The design was a randomized complete block with split-plot restrictions comparing two nitrogen treatments (no nitrogen addition and 8 g/N/m2 applied twice annually in June and September of each year as ammonium nitrate) with 12 replicates of each treatment. Plants were space planted at 1-m intervals, with a 2-m buffer space between nitrogen treatments. Two years of morphological and production data were collected.
A second garden experiment was established during the summer of 2008 in a simulated wetland in a greenhouse at the University of Wisconsin West Madison Agricultural Research Station. The populations were fully randomized within the simulated wetland with six replicates per population. Plugs were planted into 19-L pots of nitrogen-rich homogenized field soil and placed in a pool with 30 cm of standing water. Plants were grown for 122 days before harvest in the wetland garden and 134 days before harvest in the upland garden.
Additional seed of each population was germinated in the greenhouse in 2008. Fresh tissue (0.1 – 0.2 g) was collected approximately three weeks following germination and frozen for future use. Total genomic DNA was obtained using standard methods  and normalized to approximately 10 ng/ul. Between two and six samples per population were evaluated.
One hundred and thirty SSR primer pairs developed for use with Phalaris canariensis were synthesized and evaluated for amplification and allelic polymorphisms . The forward primer of each pair was synthesized with the universal M13 tail (CACGACGTTGTAAAACGAC) at the 5′ end to facilitate fluorescent labeling . The M13 tail was labeled either with carboxyfluorescein (FAM) or hexachlorofluorescein (HEX) fluorescent tags. In addition, to reduce polyadenylation and improve genotyping, the PIG sequence (GTTTCTT) was appended to the 5′ end of the reverse primer . Of these primer pairs, only 11 yielded polymorphic and easily scored alleles without extraneous peaks (Table S1). Four additional primers were included that were developed from a project identifying conserved primers within the Poaceae by downloading 2,340 SSR primers from Maize GDB  and finding perfectly conserved primer sequences in the Sorghum bicolor (sbi1) genome . Polymerase chain reactions (PCR) were performed in 8 ul total volume using 3.5 ul 1X JumpStart REDTaq ReadyMix (Sigma, St. Louis, MO, USA), 10 ng genomic DNA, 1.25 ul of H2O, 0.5 ul 5 M M13-FAM/HEX primer, 0.5 ul 5 M reverse/0.5 M forward primer, 0.125 ul 5 M betaine (Sigma, St. Louis, MO, USA), and 0.125 ul 50 mg/ml BSA (CHIMERx, Milwaukee, WI, USA). Thermocycling conditions consisted of an initial melting step (94°C for 3 min), followed by 35 cycles of 94°C for 15 s, 54°C for 90 s, and 72°C for 2 min, and an elongation step (72°C for 20 min), followed by an indefinite soak at 4°C. PCR products (2 ul) using different fluorescent labels (i.e., FAM and HEX) were pooled and combined with 15 ul Hi-Di formamide (Applied Biosystems, Foster City, CA, USA) and 0.15 ul of carboxy-X-rhodamine (ROX) custom size standard (Custom MapMarker, BioVentures, Murfreesboro, TN, USA). SSR allele genotyping was performed using an ABI 3730 fluorescent sequencer (POP-6 and a 50-cm array; Applied Biosystems, Foster City, CA, USA). Alleles were scored using GeneMarker Software version 1.7 (SoftGenetics, State College, PA, USA).
Data obtained with the SSR loci were scored in a binary format as presence (1) or absence (0) of bands. A pairwise individual-by-individual Euclidean distance matrix was calculated in GenAlEx 6.4  for the binary data and was used to perform the following analyses. Principal component analysis (PCA) was performed based on the genetic distance of all samples. An analysis of molecular variance (AMOVA) was calculated to test for significance of genetic variation between groups and populations, and to confirm that the North American invader and European wild populations were distinct from the cultivars. Significance was determined by comparing actual values to a null distribution generated from 9,999 permutation of the data.
Aboveground biomass production was measured using allometric equations developed from the harvest of additional plants. Unique allometric equations were developed with multiple regression for each year to account for changes in plant growth using canopy height, canopy diameter, and basal perimeter to predict aboveground biomass. Aboveground biomass was clipped in early November of each season, dried at 65°C for four days, and weighed. Canopy height was measured as the distance from the soil surface to the height of the highest leaf. Canopy diameter was measured as the distance between the outermost leaves in the canopy. Basal perimeter was measured as the distance around the outermost tillers of the plant at the soil surface. Because of the strong relationship between these vegetative characteristics and aboveground biomass, we use aboveground biomass production as our estimate of vegetative vigor.
Belowground production allometric equations were developed for the upland plants over a 2-year period in which 85 additional plants were harvested to determine total belowground production. The plants were harvested by digging all soil within approximately 15 cm of the outermost tillers and to a depth of approximately 35 cm. The root mass of each plant was hand washed, dried at 65°C for four days, and weighed. End-of-season aboveground biomass was used as a predictor of belowground production , . The roots of all plants in the wetland garden were washed, dried, and weighed.
An index of fecundity of each plant was developed using PCA to integrate values of the number of panicles, average panicle length and average panicle height of each plant into a single value. Because plants did not flower in the upland garden during the first season, the index was calculated using first year data in the wetland garden and second year data in the upland garden. Average panicle length was measured for three panicles on each plant. Average panicle height was measured as the distance from the soil surface to the highest point of three panicles per plant. The number of panicles of each plant was counted in late July in the uplands and prior to final harvest in the wetland. The first principal component generated from this analysis was normalized to a range of zero to one . The positive relationship between each trait and the index was confirmed using Pearson correlation coefficients.
Maximum likelihood mixed-effects models were developed using the NLME package in R 2.10.1 to test the significance of fixed effects. Linear mixed models were compared to a null model using a likelihood ratio test to determine whether fixed effects were significant at p = 0.05 . A likelihood ratio test was used to compare two models for each response variable: group as a fixed effect vs. group and populations nested within group as fixed effects. Group, population, year, and environment (upland with nitrogen addition, upland with no nitrogen addition, and wetland) were treated as fixed effects, while block was treated as a random effect. Year was used as a fixed-effect repeated measure to account for expected differences in morphology between the establishment and persistence years, allowing estimation of the autoregressive correlation structure between years. Variables were log transformed as needed to meet normality assumptions.
The PCA based on genetic data from the 39 samples did not identify any completely distinct clusters, and emphasized the diversity represented by both North American and European cultivars (Figure 1). European wild populations were the most distinct from all other groups, but all four groups had some degree of overlap. The first two axes of this PCA accounted for 41% of the total genetic variation. The AMOVA results determined that the populations were significantly different genetically (p<0.0001), accounting for 15% of the total genetic variation (Table 2). The group classification did not explain a significant amount of variation, explaining only 2% of the total genetic variation (p = 0.16). The remaining 84% of variation was present within populations indicating the high level of genetic diversity present within each cultivar and population.
Principal component analysis based on Euclidean genetic distance of 39 samples from four groups made up of three populations each based on 15 SSR markers.
The R2 value of the allometric equations for predicting aboveground biomass using morphological measurements was 0.81 in year one and 0.89 in year two. The R2 value of the equation predicting belowground production using aboveground production was 0.69 across both years.
For all response variables, the use of a separate estimate for each population significantly improved models using an estimator for each a priori group (p<0.05), suggesting that there was significant variability between the populations within a group. In year one, both group x environment and population x environment interactions were significant (p<0.001 for both, Figure 2). In both upland environments, the European wild group was less productive than all other groups (no nitrogen, p = 0.002 and nitrogen addition, p = 0.01). Neither group nor population were significant in the wetland environment (p = 0.11 and p = 0.36, respectively).
The aboveground production (Means ± S.E.) of the 12 populations compared across the three environments in year one. The group x environment interaction was significant (p = 0.001). The European wild group was less productive than all other groups in both upland environments (p = 0.002 and p = .01, respectively). There were no significant differences in production in the wetland environment (p = 0.28).
Similar patterns were found for belowground production in the first season. In both upland environments, the European wild group produced significantly less biomass (p<0.001). Neither group nor population were significant predictors of belowground biomass in the wetland environment (p = 0.11, and p = 0.12).
We were unable to maintain populations in the simulated wetland under realistic conditions for two seasons because of the inability to simulate overwintering and a lack of space due to plant size, so we were unable to test group x year interactions in this environment. However, upland experiments were evaluated for two growing seasons. Both group x year and population x year interactions were significant (p<0.001), suggesting differences in vigor in the establishment and persistence years (Figure 3). The nutrient treatment x year interaction was also significant, likely due to an increased nitrogen limitation in unfertilized plots in year two (p<0.001). The group and population x upland nutrient treatment interaction were not significant (p = 0.33 and p = 0.72), indicating that groups and populations were not responding differently to nitrogen addition. On average, the addition of nitrogen increased production by 115.3 (±13.0) g/plant. In both environments, the European wild group was significantly less productive than all other groups (p<0.001). Because of the strong correlation between aboveground and belowground production in the upland treatments, differences in belowground production between populations in year two are similar to aboveground differences and are not reported.
Aboveground production (Means ± S.E.) of the twelve populations compared across the two upland environments in year two. The group x year interaction was significant (p<0.0001), as was the environment x year interaction (p<0.001), but the group x environment interaction was not significant (p = 0.72). The European wild group was significantly different from all other groups (p<0.001).
The first principal component of the fecundity index extracted 60% of the variation of the three traits used. The first principal component was positively correlated with each of the three traits used to generate it (p<0.001) and had an average Pearson correlation coefficient with the three traits of 0.74 (±0.20 S.D.). Both the group and population x environment interactions were significant for the fecundity index (p = 0.009 and p = 0.01, Figure 4). In uplands with nitrogen addition, the North American invader group was significantly more fecund than all other groups (p = 0.007). In the wetland environment, the European wild group was significantly less fecund than all other groups (p = 0.005). The fecundity index value was correlated with end of season production (year one in wetland, year two in upland environments) with a Pearson correlation coefficient of 0.65.
The fecundity index (Means ± S.E.) compared across the different environments. The population x environment interaction was significant (p = 0.01). In uplands without nitrogen addition, the North American cultivar group was significantly less fecund than all other populations (p = 0.01). In uplands with nitrogen addition, the invader group was significantly more fecund than all other groups (p = 0.007). In wetlands, the European wild group was significantly less fecund than all other groups (p = 0.005).
Understanding the origins of reed canarygrass invasion
Breeding efforts do not appear to be the primary reason for invasion by reed canarygrass, but our data does not completely rule out the possibility. The genetic data offers conflicting evidence regarding the genetic origin of invasive populations. The PCA clusters the three European wild populations as somewhat distinct from cultivars, but among-group variability was not significant in the AMOVA. The invasive populations overlap with both groups of cultivars and have only minor overlap with the European wild populations. Two hypotheses could explain these results. First, cultivars are thought to be derived from wild European germplasm and should represent significantly less diversity than wild European germplasm due to selection and genetic bottlenecks . If this hypothesis is correct, the three European wild populations used in this study are not representative of the total diversity of the species in Europe. Alternatively, our results could be accurate in suggesting that invasive populations in North America are more similar to cultivars than to wild European germplasm. Introductions by early European settlers of agronomic varieties were likely the founding populations of present-day invasive populations in North America. The progeny of these early introductions may make up the majority of invasive populations in North America, regardless of whether breeding efforts made the species more invasive. Additional sampling of European and North American populations is necessary to evaluate these hypotheses further.
Previous research suggested that increased genetic variability in North American reed canarygrass populations resulted in increased vegetative vigor when compared to European wild populations . Our results do not support this hypothesis of higher vigor and phenotypic variability in North American invader populations over European wild populations. There was significant phenotypic variability among populations within each continent, suggesting that a plant's provenance is not a good predictor of its vigor. One of the European wild populations, Mountain Swiss, collected in an Alpine meadow in Switzerland was as productive and fecund as the most vigorous and fecund population in all environments in year one. The European wild population D-1827, collected in Russia, was consistently the least productive and fecund in upland environments, was as productive as all other populations in the wetland environment, but was significantly less fecund in the wetland environment. However, the fecundity index from the wetland environment may not be appropriate because plants did not undergo vernalization, which is generally necessary for flowering in the field . Our results suggest that the least vigorous European populations may have been eliminated in North America following their introduction due to evolutionary pressure, but our inference here is limited by our small sampling of populations from both continents. Alternatively, the planting of early introductions and cultivars could have made the species more common in North America, even though improvement of traits via breeding have had a minimal effect on the invasiveness of the species. Additional support for this hypothesis would come from a genetic study of a large number of European samples, cultivars, and North American invasive populations to determine the genetic similarity of European and North American populations. The fecundity of cultivars and North American invader population was no higher than some European wild populations. While increased fecundity may have been a goal of early breeding programs , the goal of improving forage quality may have unintentionally selected against highly fecund individuals as increased flowering can reduce forage quality .
Early cultivars of reed canarygrass frequently exhibited issues of low palatability and resulted in suppressed intake and weight gain by ruminant livestock due to the presence of indole alkaloids . As a result, breeding programs focused on improving palatability by altering alkaloid profiles during the past 30+ years . This selection for improved palatability likely has reduced the fitness of cultivars by making them more susceptible to insect herbivory . Further, the selection for improved palatability has limited the ability to select for increased yield by constraining the available germplasm from which selections can be made. A recent evaluation of 72 wild accessions collected in North America and eight cultivars found that 39 of the wild accessions ranked higher in biomass yield than all eight cultivars . A similar limitation is likely to occur in cellulosic biomass crops due to efforts to alter lignin and cellulose production to improve the efficiency of conversion to ethanol .
The hypothesis that breeding for improved agronomic traits is responsible for creating invasive populations of reed canarygrass is not supported by our results. While cultivars are consistently among the highest yielding populations in the environment for which they were selected (uplands with nitrogen addition), these populations are not significantly more productive than the highest producing European wild population in uplands, and there were no differences in above- or below-ground production in the wetland environment. This suggests that breeding efforts in reed canarygrass have had little effect on biomass production in upland environments and no significant effect on production in wetland environments. This is not surprising, as it has proven difficult in many crops to achieve improvements that are consistent across a wide range of environments ,  If breeding were the primary cause of the invasive traits of the species, we would expect cultivars and North American invasive populations to be far more productive than all European wild populations in wetland environments. While possible, it is unlikely that marginal increases in production or fecundity are responsible for a species becoming invasive. In our study, the most fecund population in wetlands, Vantage, a North American cultivar, was 10% more fecund than the most fecund European wild population. Previous research comparing seed yield of eight cultivars, 14 breeding populations, and 53 European wild populations found wide variation among populations within each group, with four wild populations more fecund than the most fecund cultivar, Vantage. The wild population with the highest seed yield had a seed production index 17% higher than Vantage, lending further support to the hypothesis that breeding has not led to the development of levels of seed production not already found in wild populations . While reed canarygrass is considered an extremely plastic species due to its ability to flourish in upland and lowland environments , , our data suggests that improvements from breeding have not been maintained across all environments. Breeding and the intentional planting of cultivars may have resulted in a higher abundance of the species on the landscape, but breeding does not appear to be responsible for biomass production and fecundity levels not already found in wild European populations.
Additional research is necessary to understand how this native species has become invasive, but previous research has concluded that the species is successful in eutrophic, highly disturbed wetlands , , . This niche has expanded greatly in North America following the expansion and intensification of agriculture during the 20th century . Rather than changes in aggressiveness of the species among European populations being responsible for invasion, an increase in the ideal environment of the species may be the driver of invasion by reed canarygrass .
Applying knowledge to the risks of perennial breeding
Using reed canarygrass as an analogue for the effects of breeding on perennial crops is most appropriate for species with a similar life history (e.g. switchgrass); however, our study offers useful insights to minimize the risks associated with breeding perennial crops. The risks of invasion by crops in the near future will be dominated by the introduction of novel bioenergy crops, a legitimate concern given a recent estimate of $26.4 billion for the costs of management and lost production due to invasive weeds in the US agricultural economy . The simplest response to this concern is to restrict the use of any exotic or improved crops deemed potentially invasive as a precautionary measure. While developing systems to screen out the most likely invaders will be important , preventing the use of all possible improved crops eliminates the possibility of the many beneficial ecosystem services that accompany crop improvements and may accompany a shift to a bioenergy economy , .
There are alternatives to allowing all species to be planted or highly restricting any potential invaders. In reed canarygrass, improvements in yield were only present in upland environments, and were not maintained in wetland environments. Breeding often selects improvements for a very specific target population of environments, and stability across a wide range of environments has proved elusive for even the most highly bred crops . Selecting cultivars for specific environments or requiring specific management (i.e. high nitrogen environments or cultivation) reduces the probability that a crop will escape cultivation. Breeders generally strive for cultivar stability across a range of environments, as these cultivars are more marketable and simplify selection for a producer . In contrast to typical cultivar development, a range of cultivars that are each designed for a specific environment may be preferable for biofuel feedstock crops. In addition, the evaluation of crops in environments in which they have the potential to be invasive may be appropriate to select against the populations that are highly productive in these environments. This will require collaboration between breeders and weed scientists to determine which environments are most at risk.
Another safeguard to reduce the chance of creating a novel invader is to select for reduced fecundity of perennial crops. Efforts to select for slowed maturity and decreased flowering to improve forage quality have been successful in several perennial pasture species , . In addition, research to reduce or eliminate flowering by altering the genetic mechanisms involved in the flowering pathway has shown promise , . This technique not only reduces the fecundity of the crop but can also increase biomass production due to a reallocation of resources. Selecting for reduced fecundity in biomass crops has the potential to reduce the risks of introgression of improved genes or traits into native populations and the escape of improved varieties into undesirable environments.
Our reed canarygrass study suggests that breeding efforts do not appear to be the source for the invasive traits of the species. Cultivars were among the most productive populations in the environment for which they were selected (uplands with nitrogen addition), but were not more productive or fecund than all wild populations in the environment in which they are considered invasive. While early introductions of the species for agriculture likely acted as the initial founding populations prior to invasion, there do not appear to be major differences in biomass production or fecundity between modern cultivars and wild populations in wetland environments. Additional research is required to evaluate the risks associated with the introduction of novel crops and cultivars. To achieve this goal, an understanding of the interaction of landscape changes and human-influenced evolutionary change (either direct as in breeding or indirect through the introduction of exotic species) is necessary. While every crop is different, the study of historical perennial crop breeding can help to identify issues and solutions before novel crops or cultivars are released on the landscape.
Thanks to the University of Wisconsin OJ Noer Turfgrass Research and Education Facility and to Nick Baker for their assistance in establishing and managing the field trials. Additional thanks to USDA-GRIN for access to germplasm.
Conceived and designed the experiments: ARJ MDC RDJ. Performed the experiments: ARJ. Analyzed the data: ARJ. Contributed reagents/materials/analysis tools: MDC RDJ. Wrote the paper: ARJ MDC RDJ.
- 1. Borlaug NE (1983) Contributions of conventional plant breeding to food production. Science 219: 689.NE Borlaug1983Contributions of conventional plant breeding to food production.Science219689
- 2. Koonin SE (2006) Getting Serious About Biofuels. Science 311: 435.SE Koonin2006Getting Serious About Biofuels.Science311435
- 3. Perlack RD, Wright LL, Turhollow AF, Graham RL, Stokes BJ, et al. (2005) Biomass as Feedstock for a Bioenergy and Bioproducts Industry: The Technical Feasibility of a Billion-Ton Annual Supply. Oak Ridge, TN. RD PerlackLL WrightAF TurhollowRL GrahamBJ Stokes2005Biomass as Feedstock for a Bioenergy and Bioproducts Industry: The Technical Feasibility of a Billion-Ton Annual Supply.Oak Ridge, TN
- 4. Edgerton MD (2009) Increasing Crop Productivity to Meet Global Needs for Feed, Food, and Fuel. Plant Physiology 149: 7–13.MD Edgerton2009Increasing Crop Productivity to Meet Global Needs for Feed, Food, and Fuel.Plant Physiology149713
- 5. Barney JN, DiTomaso JM (2008) Nonnative Species and Bioenergy: Are We Cultivating the Next Invader? BioScience 58: 64–70.JN BarneyJM DiTomaso2008Nonnative Species and Bioenergy: Are We Cultivating the Next Invader?BioScience586470
- 6. Raghu S, Anderson RC, Daehler CC, Davis AS, Wiedenmann RN, et al. (2006) Adding biofuels to the invasive species fire. Science 313: 1742.S. RaghuRC AndersonCC DaehlerAS DavisRN Wiedenmann2006Adding biofuels to the invasive species fire.Science3131742
- 7. Davis AS, Cousens RD, Hill J, Mack RN, Simberloff D, et al. (2010) Screening bioenergy feedstock crops to mitigate invasion risk. Frontiers in Ecology and the Environment 8: 533–539.AS DavisRD CousensJ. HillRN MackD. Simberloff2010Screening bioenergy feedstock crops to mitigate invasion risk.Frontiers in Ecology and the Environment8533539
- 8. Theoharides Ka, Dukes JS (2007) Plant Invasion Across Space and Time: Factors Affecting Nonindigenous Species Success During Four Stages of Invasion. New Phytologist 176: 256–273.Ka TheoharidesJS Dukes2007Plant Invasion Across Space and Time: Factors Affecting Nonindigenous Species Success During Four Stages of Invasion.New Phytologist176256273
- 9. Heaton EA, Flavell RB, Mascia PN, Thomas SR, Dohleman FG, et al. (2008) Herbaceous energy crop development: recent progress and future prospects. Current Opinion in Biotechnology 19: 202–209.EA HeatonRB FlavellPN MasciaSR ThomasFG Dohleman2008Herbaceous energy crop development: recent progress and future prospects.Current Opinion in Biotechnology19202209
- 10. Sakai AK, Allendorf FW, Holt JS, Lodge DM, Molofsky J, et al. (2001) The population biology of invasive species. Annual Review of Ecology and Systematics 32: 305–332.AK SakaiFW AllendorfJS HoltDM LodgeJ. Molofsky2001The population biology of invasive species.Annual Review of Ecology and Systematics32305332
- 11. Van Kleunen M, Weber E, Fischer M (2010) A meta analysis of trait differences between invasive and non invasive plant species. Ecology Letters 13: 235–245.M. Van KleunenE. WeberM. Fischer2010A meta analysis of trait differences between invasive and non invasive plant species.Ecology Letters13235245
- 12. McLaughlin SB, Kszos L (2005) Development of switchgrass (Panicum virgatum) as a bioenergy feedstock in the United States. Biomass and Bioenergy 28: 515–535.SB McLaughlinL. Kszos2005Development of switchgrass (Panicum virgatum) as a bioenergy feedstock in the United States.Biomass and Bioenergy28515535
- 13. Kolar CS, Lodge DM (2001) Progress in invasion biology: predicting invaders. Trends in Ecology & Evolution 16: 199–204.CS KolarDM Lodge2001Progress in invasion biology: predicting invaders.Trends in Ecology & Evolution16199204
- 14. Clements DR, Ditommaso A (2011) Climate change and weed adaptation: can evolution of invasive plants lead to greater range expansion than forecasted? Weed Research 51: 227–240.DR ClementsA. Ditommaso2011Climate change and weed adaptation: can evolution of invasive plants lead to greater range expansion than forecasted?Weed Research51227240
- 15. Casler MD, Pedersen JF, Eizenga GC, Stratton SD (1996) Germplasm and cultivar development. In: Moser LE, Buxton DR, Casler MD, editors. Cool-Season Forage Grasses. Madison, WI: Agronomy Society of America Monographs No. 34. pp. 413–469.MD CaslerJF PedersenGC EizengaSD Stratton1996Germplasm and cultivar development.LE MoserDR BuxtonMD CaslerCool-Season Forage GrassesMadison, WIAgronomy Society of America Monographs No. 34413469
- 16. Wilkins PW, Humphreys MO (2003) Progress in breeding perennial forage grasses for temperate agriculture. The Journal of Agricultural Science 140: 129–150.PW WilkinsMO Humphreys2003Progress in breeding perennial forage grasses for temperate agriculture.The Journal of Agricultural Science140129150
- 17. Keeney D, Sanderson MA (2006) Forages and the environment. In: Moore KJ, Barnes RF, Nelson CJ, Collins M, editors. Forages: The Science of Grassland Agriculture. Sixth Edition ed. Ames, IA: Blackwell. pp. 167–176.D. KeeneyMA Sanderson2006Forages and the environment.KJ MooreRF BarnesCJ NelsonM. CollinsForages: The Science of Grassland Agriculture. Sixth Edition edAmes, IABlackwell167176
- 18. Cully AC, Cully JF, Hiebert RD (2003) Invasion of Exotic Plant Species in Tallgrass Prairie Fragments. Conservation Biology 17: 990–998.AC CullyJF CullyRD Hiebert2003Invasion of Exotic Plant Species in Tallgrass Prairie Fragments.Conservation Biology17990998
- 19. Merigliano MF, Lesica P (1998) The native status of reed canarygrass (Phalaris arundinacea L.) in the Inland Northwest, USA. Natural Areas Journal 18: 223–230.MF MeriglianoP. Lesica1998The native status of reed canarygrass (Phalaris arundinacea L.) in the Inland Northwest, USA.Natural Areas Journal18223230
- 20. Alway F (1931) Early trials and use of reed canary grass as a forage plant. Journal of the American Society of Agronomy 23: 64–66.F. Alway1931Early trials and use of reed canary grass as a forage plant.Journal of the American Society of Agronomy236466
- 21. Burvall J (1997) Influence of harvest time and soil type on fuel quality in reed canary grass (Phalaris arundinacea L.). Biomass and Bioenergy 12: 149–154.J. Burvall1997Influence of harvest time and soil type on fuel quality in reed canary grass (Phalaris arundinacea L.).Biomass and Bioenergy12149154
- 22. Adler PR, Del Grosso SJ, Parton WJ (2007) Life-cycle assessment of net greenhouse-gas flux for bioenergy cropping systems. Ecological Applications 17: 675–691.PR AdlerSJ Del GrossoWJ Parton2007Life-cycle assessment of net greenhouse-gas flux for bioenergy cropping systems.Ecological Applications17675691
- 23. Jakubowski AR, Casler MD, Jackson RD (2010) The Benefits of Harvesting Wetland Invaders for Cellulosic Biofuel: An Ecosystem Services Perspective. Restoration Ecology 18: 789–795.AR JakubowskiMD CaslerRD Jackson2010The Benefits of Harvesting Wetland Invaders for Cellulosic Biofuel: An Ecosystem Services Perspective.Restoration Ecology18789795
- 24. Carlson IT, Oram RN, Surpenant J (1996) Reed Canarygrass and Other Phalaris Species. In: Moser LE, Buxton DR, Casler MD, editors. Cool-Season Forage Grasses. Madison, WI: American Society of Agronomy, Crop Science Society of America, Soil Science Society of America. pp. 569–604.IT CarlsonRN OramJ. Surpenant1996Reed Canarygrass and Other Phalaris Species.LE MoserDR BuxtonMD CaslerCool-Season Forage GrassesMadison, WIAmerican Society of Agronomy, Crop Science Society of America, Soil Science Society of America569604
- 25. Lavergne S, Molofsky J (2004) Reed canary grass (Phalaris arundinacea) as a biological model in the study of plant invasions. Critical Reviews in Plant Sciences 23: 415–429.S. LavergneJ. Molofsky2004Reed canary grass (Phalaris arundinacea) as a biological model in the study of plant invasions.Critical Reviews in Plant Sciences23415429
- 26. Dore W, McNeill J (1980) Grasses of Ontario: Agriculture and Agri-Food Canada. W. DoreJ. McNeill1980Grasses of Ontario: Agriculture and Agri-Food Canada
- 27. Lavergne S, Molofsky J (2007) Increased genetic variation and evolutionary potential drive the success of an invasive grass. Proceedings of the National Academy of Sciences of the United States of America 104: 3883–3888.S. LavergneJ. Molofsky2007Increased genetic variation and evolutionary potential drive the success of an invasive grass.Proceedings of the National Academy of Sciences of the United States of America10438833888
- 28. Casler MD, Phillips MM, Krohn AL (2009) DNA Polymorphisms Reveal Geographic Races of Reed Canarygrass. Crop Science 49: 2139–2148.MD CaslerMM PhillipsAL Krohn2009DNA Polymorphisms Reveal Geographic Races of Reed Canarygrass.Crop Science4921392148
- 29. Storchova H, Hrdlickova R, Chrtek J , Tetera M, Fitze D, et al. (2000) An improved method of DNA isolation from plants collected in the field and conserved in NACL/CTAB solution. Taxon 49: 79–84.H. StorchovaR. HrdlickovaChrtek J JrM. TeteraD. Fitze2000An improved method of DNA isolation from plants collected in the field and conserved in NACL/CTAB solution.Taxon497984
- 30. Li J, Båga M, Hucl P, Chibbar R (2010) Development of microsatellite markers in canary seed (Phalaris canariensis L.). Molecular Breeding 1–11.J. LiM. BågaP. HuclR. Chibbar2010Development of microsatellite markers in canary seed (Phalaris canariensis L.).Molecular Breeding111DOI:10.1007/s11032-010-9513-2. DOI:10.1007/s11032-010-9513-2.
- 31. Schuelke M (2000) An economic method for the fluorescent labeling of PCR fragments. Nat Biotech 18: 233–234.M. Schuelke2000An economic method for the fluorescent labeling of PCR fragments.Nat Biotech18233234
- 32. Brownstein M, Carpten J, Smith J (1996) Modulation of Non-Templated Nucleotide Addition by TaqDNA Polymerase: Primer Modifications that Facilitate Genotyping. Biotechniques 20: 1004–1010.M. BrownsteinJ. CarptenJ. Smith1996Modulation of Non-Templated Nucleotide Addition by TaqDNA Polymerase: Primer Modifications that Facilitate Genotyping.Biotechniques2010041010
- 33. Lawrence CJ, Dong Q, Polacco ML, Seigfried TE, Brendel V (2004) MaizeGDB, the community database for maize genetics and genomics. Nucleic Acids Research 32: D393–D397.CJ LawrenceQ. DongML PolaccoTE SeigfriedV. Brendel2004MaizeGDB, the community database for maize genetics and genomics.Nucleic Acids Research32D393D397
- 34. Paterson AH, Bowers JE, Bruggmann R, Dubchak I, Grimwood J, et al. (2009) The Sorghum bicolor genome and the diversification of grasses. Nature 457: 551–556.AH PatersonJE BowersR. BruggmannI. DubchakJ. Grimwood2009The Sorghum bicolor genome and the diversification of grasses.Nature457551556
- 35. Peakall ROD, Smouse PE (2006) genalex 6: genetic analysis in Excel. Population genetic software for teaching and research. Molecular Ecology Notes 6: 288–295.ROD PeakallPE Smouse2006genalex 6: genetic analysis in Excel. Population genetic software for teaching and research.Molecular Ecology Notes6288295
- 36. Kucharik CJ, Fayram NJ, Cahill KN (2006) A paired study of prairie carbon stocks, fluxes, and phenology: comparing the world's oldest prairie restoration with an adjacent remnant. Global change biology 12: 122–139.CJ KucharikNJ FayramKN Cahill2006A paired study of prairie carbon stocks, fluxes, and phenology: comparing the world's oldest prairie restoration with an adjacent remnant.Global change biology12122139
- 37. Gross MF, Hardisky MA, Wolf PL, Klemas V (1991) Relationship between Aboveground and Belowground Biomass of Spartina alterniflora (Smooth Cordgrass). Estuaries 14: 180–191.MF GrossMA HardiskyPL WolfV. Klemas1991Relationship between Aboveground and Belowground Biomass of Spartina alterniflora (Smooth Cordgrass).Estuaries14180191
- 38. Baeten L, Vanhellemont M, De Frenne P, De Schrijver A, Hermy M, et al. (2010) Plasticity in response to phosphorus and light availability in four forest herbs. Oecologia 163: 1021–1032.L. BaetenM. VanhellemontP. De FrenneA. De SchrijverM. Hermy2010Plasticity in response to phosphorus and light availability in four forest herbs.Oecologia16310211032
- 39. Crawley MJ (2007) The R Book. Hoboken: Wiley. MJ Crawley2007The R Book.HobokenWiley
- 40. Casler MD (2010) Genetics, Breeding, and Ecology of Reed Canarygrass. International Journal of Plant Breeding 4: 30–36.MD Casler2010Genetics, Breeding, and Ecology of Reed Canarygrass.International Journal of Plant Breeding43036
- 41. Heichel GH, Hovin AW, Henjum KI (1980) Seedling age and cold treatment effects on induction of panicle production in reed canarygrass. Crop Science 20: 683–687.GH HeichelAW HovinKI Henjum1980Seedling age and cold treatment effects on induction of panicle production in reed canarygrass.Crop Science20683687
- 42. Alderson J, Sharp WC (1994) Grass varieties in the United States; USDA-SCS, editor. Washington D.C.: U.S. Government Printing Office. J. AldersonWC Sharp1994Grass varieties in the United States; USDA-SCS, editor.Washington D.C.U.S. Government Printing Office
- 43. Marten GC (1989) Breeding forage grasses to maximize animal performance. In: Sleper DA, Asay KH, Pedersen JF, editors. Contributions from breeding forage and turf grasses CSSA Special Publication No 15. Madison, WI: Crop Science Society of America. pp. 71–104.GC Marten1989Breeding forage grasses to maximize animal performance.DA SleperKH AsayJF PedersenContributions from breeding forage and turf grasses CSSA Special Publication No 15Madison, WICrop Science Society of America71104
- 44. Byers RA, Sherwood RT (1979) Differential reaction of clones of Phalaris arundinacea to Oscinella frit. Environmental Entomology 8: 408–411.RA ByersRT Sherwood1979Differential reaction of clones of Phalaris arundinacea to Oscinella frit.Environmental Entomology8408411
- 45. Casler MD, Cherney JH, Brummer EC (2009) Biomass yield of naturalized populations and cultivars of reed canary grass. BioEnergy Research 2: 165–173.MD CaslerJH CherneyEC Brummer2009Biomass yield of naturalized populations and cultivars of reed canary grass.BioEnergy Research2165173
- 46. Casler MD, Buxton DR, Vogel KP (2002) Genetic modification of lignin concentration affects fitness of perennial herbaceous plants. Theoretical and Applied Genetics 104: 127–131.MD CaslerDR BuxtonKP Vogel2002Genetic modification of lignin concentration affects fitness of perennial herbaceous plants.Theoretical and Applied Genetics104127131
- 47. Eberhart SA, Russell WA (1966) Stability parameters for comparing varieties. Crop Science 6: 36–40.SA EberhartWA Russell1966Stability parameters for comparing varieties.Crop Science63640
- 48. Simmonds NW (1991) Selection for local adaptation in a plant-breeding program. Theoretical and Applied Genetics 82: 363–367.NW Simmonds1991Selection for local adaptation in a plant-breeding program.Theoretical and Applied Genetics82363367
- 49. Sahramaa M (2003) Evaluation of reed canary grass for different end-uses and in breeding. Agricultural and food science in Finland 12: 227–241.M. Sahramaa2003Evaluation of reed canary grass for different end-uses and in breeding.Agricultural and food science in Finland12227241
- 50. Miller RC, Zedler JB (2003) Responses of native and invasive wetland plants to hydroperiod and water depth. Plant Ecology. pp. 57–69.RC MillerJB Zedler2003Responses of native and invasive wetland plants to hydroperiod and water depth.Plant Ecology5769
- 51. Coops H, van den Brink FWB, van der Velde G (1996) Growth and morphological responses of four helophyte species in an experimental water-depth gradient. Aquatic Botany 54: 11–24.H. CoopsFWB van den BrinkG. van der Velde1996Growth and morphological responses of four helophyte species in an experimental water-depth gradient.Aquatic Botany541124
- 52. Maurer DA, Zedler JB (2002) Differential invasion of a wetland grass explained by tests of nutrients and light availability on establishment and clonal growth. Oecologia 131: 279–288.DA MaurerJB Zedler2002Differential invasion of a wetland grass explained by tests of nutrients and light availability on establishment and clonal growth.Oecologia131279288
- 53. Green EK, Galatowitsch SM (2001) Differences in wetland plant community establishment with additions of nitrate-N and invasive species (Phalaris arundinacea and Typha X glauca). Canadian Journal of Botany 79: 170–178.EK GreenSM Galatowitsch2001Differences in wetland plant community establishment with additions of nitrate-N and invasive species (Phalaris arundinacea and Typha X glauca).Canadian Journal of Botany79170178
- 54. Green EK, Galatowitsch SM (2002) Effects of Phalaris arundinacea and nitrate-N addition on the establishment of wetland plant communties. Journal of Applied Ecology. pp. 134–144.EK GreenSM Galatowitsch2002Effects of Phalaris arundinacea and nitrate-N addition on the establishment of wetland plant communties.Journal of Applied Ecology134144
- 55. Pimentel D, Harvey C, Resosudarmo P, Sinclair K, Kurz D, et al. (1995) Environmental and Economic Costs of Soil Erosion and Conservation Benefits. Science 267: 1117–1123.D. PimentelC. HarveyP. ResosudarmoK. SinclairD. Kurz1995Environmental and Economic Costs of Soil Erosion and Conservation Benefits.Science26711171123
- 56. Jakubowski AR, Casler MD, Jackson RD (2010) Landscape Context Predicts Reed Canarygrass Invasion: Implications for Management. Wetlands 30: 685–692.AR JakubowskiMD CaslerRD Jackson2010Landscape Context Predicts Reed Canarygrass Invasion: Implications for Management.Wetlands30685692
- 57. Pimentel D, Lach L, Zuniga R, Morrison D (2000) Environmental and economic costs of nonindigenous species in the United States. Bioscience 50: 53–65.D. PimentelL. LachR. ZunigaD. Morrison2000Environmental and economic costs of nonindigenous species in the United States.Bioscience505365
- 58. Conway G, Toenniessen G (1999) Feeding the world in the twenty-first century. Nature 402: 55.G. ConwayG. Toenniessen1999Feeding the world in the twenty-first century.Nature40255
- 59. Robertson GP, Dale VH, Doering OC, Hamburg SP, Melillo JM, et al. (2008) Sustainable Biofuels Redux. Science 322: 49–50.GP RobertsonVH DaleOC DoeringSP HamburgJM Melillo2008Sustainable Biofuels Redux.Science3224950
- 60. Buxton DR (1996) Quality-related characteristics of forages as influenced by plant environment and agronomic factors. Animal Feed Science and Technology 59: 37–49.DR Buxton1996Quality-related characteristics of forages as influenced by plant environment and agronomic factors.Animal Feed Science and Technology593749
- 61. Casler MD, Barker RE, Cherney JH, Papadopolous YA (2004) Stability of Non-Flowering Orchardgrass. Crop Science 44: 1601–1607.MD CaslerRE BarkerJH CherneyYA Papadopolous2004Stability of Non-Flowering Orchardgrass.Crop Science4416011607
- 62. Jung C, Müller AE (2009) Flowering time control and applications in plant breeding. Trends in Plant Science 14: 563–573.C. JungAE Müller2009Flowering time control and applications in plant breeding.Trends in Plant Science14563573
- 63. Salehi H, Ransom CB, Oraby HF, Seddighi Z, Sticklen MB (2005) Delay in flowering and increase in biomass of transgenic tobacco expressing the Arabidopsis floral repressor gene FLOWERING LOCUS C. Journal of Plant Physiology 162: 711–717.H. SalehiCB RansomHF OrabyZ. SeddighiMB Sticklen2005Delay in flowering and increase in biomass of transgenic tobacco expressing the Arabidopsis floral repressor gene FLOWERING LOCUS C.Journal of Plant Physiology162711717