Brachypodium sylvaticum
(slender false brome)

B. sylvaticum is a bunchgrass naturally occurring in old world temperate forests and temperate zones of tropical Asian mountains. Its extensive native range includes most of Eurasia (e.g. Europe, Russia, China, Japan, India, Indonesia) as well as the Middle East (e.g. Lebanon, Syria, Iran) and North Africa (e.g. Algeria, Eritrea). It is invasive in North America (Piep, 2003), South America (Zuloaga et al., 1994), New Zealand (Edgar and Connor, 2000), and Australia (IBIS, 2009). It is shade tolerant (Murchie and Horton, 1998), spreads rapidly by seeds (Petersen and Philipp, 2001), has a persistent seed bank (Donelan and Thompson, 1980; Buckley et al., 1997) and is long-lived (Haeggström and Skytén, 1996). It forms monocultures and crowds out native plants and rare butterflies (Kaye and Blakeley-Smith, 2006; Severns and Warren, 2008). Furthermore, grasses significantly reduce recruitment of conifers (Powell et al., 1994; Lehmkuhl, 2002; Kruse et al., 2004). It is on noxious weed lists for three USA States: California, Oregon and Washington (CDFA, 2009; NWCB, 2009; ODA, 2009).

B. sylvaticum is native to the Old World. To date it is known to be invading five states in the USA, New Zealand and small areas in Argentina and Australia. The sources of the invasions in the Pacific Northwest of the USA are European (Rosenthal et al., 2008).

B. sylvaticum is a quarantined weed in three states in the USA: Washington (NWCB, 2009), Oregon (ODA, 2009), and California (CDFA, 2009). There is high risk of further spread because there are multiple pathways for it to do so (see the text section 'Means of Movement and Dispersal'). It can be carried by animals and humans, on machinery, by rivers and streams, it may also spread as a result of increasing popularity as a model organism, through intentional introductions and the horticultural trade.

Forests

- Temperate conifer forests (e.g. Scots pine, Pinussylvestris: Ninot et al., 2000; coast redwood, Sequoia sempervirens:Hrusa, 2003; and Douglas fir, Pseudotsugamenziesii:False Brome Working Group, 2009). Found in both disturbed and undisturbed forests, and under both open and closed canopies (Hrusa, 2003; Parks et al., 2005; BA Roy, University of Oregon, USA, personal communication, 2009).

- Temperate deciduous forests (including Quercus, Fagus, Fraxinus), in both forest and open areas (Szujkó-Lacza and Fekete, 1974; Petersen and Philipp, 2001; Safaian et al., 2005; Chaideftou et al., 2009; BA Roy, University of Oregon, USA, personal observation, 2009).

- Mediterranean forests (Safaian et al., 2005; Chaideftou et al., 2009).

- Boreal forests (Holten, 1980; Haeggström and Skytén, 1996; Aarrestad, 2000; Petersen and Philipp, 2001).

Meadows/Openings

- Stream edges and lakesides, including in arid regions (Hrusa, 2003; False Brome Working Group, 2009; BA Roy, University of Oregon, USA, personal communication, 2009: in Europe).

- High altitude meadows such as the Tibetan steppe and the tops of tropical mountains (Smith, 1975; Roder et al., 2007; Singh et al., 2008).

- Temperate meadows (Davies and Long, 1991; Stace, 1997; US Fish and Wildlife Service, 2006; Severns and Warren, 2008).

The Pacific Northwest is world renowned for its timber production from conifers. It has not been established through a controlled study that B. sylvaticum in particular competes with conifers; however, several lines of evidence suggest that this grass will reduce survival and growth of conifer seedlings. First, a number of studies have established that the germination, growth and survival of conifers are negatively affected by competition with grasses (e.g. Powell et al., 1994; Lehmkuhl, 2002; Kruse et al., 2004). Second, at least one timber company (Starker Forests Inc.) has noticed that dense patches of false-brome provide safe cover for voles, which girdle conifer seedlings (G Fitzpatrick, The Nature Conservancy, Oregon, USA, personal communication, 2009). Third, none of the native grasses affected form a solid carpet in the forest, whereas B. sylvaticum does. These dense carpets will compete with seedlings in both logged and unlogged forests, and the build up of thatch may increase fire risk (Anzinger and Radosevich, 2008; False Brome Working Group, 2009). On the other hand, because B. sylvaticum remains green throughout the summer it may decrease fire risk (Anzinger and Radosevich, 2008; False Brome Working Group, 2009). These questions about fire need to be addressed with further research because this grass has the potential to cause ecosystem change if it alters fire behaviour.

The native grasses and herbs that live in the habitats being invaded are likely to diminish in cover and may face local extinction as a result of competition. A recent study showed that under shady high nutrient conditions, B. sylvaticum is a superior competitor to a native prairie grass (Festuca roemeri), a native forest grass (Elymus glaucus) as well as to another aggressive invasive Schedonorus arundinaceus (previously known as Festuca arundinacea) (BA Roy, University of Oregon, USA, personal observation, 2009). Two other grasses that occur sporadically in the forest (Melica subulata and Bromus carinatus) are also quite likely to be negatively affected. In its native range, B. sylvaticum is known for its ability to out compete other species due to its rapid relative growth rate (RGR) and ability to form persistent leaf litter (Grime et al., 1988; Haeggström and Skytén, 1996; Alonso et al., 2001).

Genetics

B. sylvaticum is a perennial, self-compatible, wind-pollinated grass that can also outcross (Khan and Stace, 1999; Rosenthal et al., 2008). It is a diploid, but variable in chromosome number, typically 2n=14, 16 or 18 (Veldkamp and Vanscheindelen, 1989; Khan and Stace, 1999; Piep, 2003). It is known to form interspecific hybrids with Brachypodium pinnatum (Khan and Stace, 1999), but has not done so in the invaded range (Rosenthal et al., 2008). However, it has formed intraspecific hybrids among individuals from different European origins (Rosenthal et al., 2008).

Plants in the genus Brachypodium are rapidly becoming model organisms because of their close relationship with wheat, and to a lesser extent rice, and because they have a much smaller genome size than these grains (Opanowicz et al., 2008). Most of the research activity is focusing on the annual Brachypodium distachyon (Garvin, 2007; Opanowicz et al., 2008; Bakker et al., 2009), but B. sylvaticum is also receiving significant attention (e.g. Foote et al., 2004; Wolny and Hasterok, 2007; Faris et al., 2008).

Reproductive Biology

B. sylvaticum is a perennial species that typically flowers late in the season relative to other temperate grasses. For example, in the Willamette Valley of Oregon, USA all the other grasses flower in May and June, whereas B. sylvaticum does not flower until early July. It is also late flowering in the native range of Europe (BA Roy, University of Oregon, USA, personal observation, 2009). B. sylvaticum produces seeds from both sexually produced tillers (that germinate from seeds) and vegetative tillers (asexual). A recent demographic study monitored population growth rates at four sites in the native range (Switzerland) and four sites in the invaded range (Oregon) (BA Roy, University of Oregon, USA, personal observation, 2009). In both ranges, tillers from seeds generally flowered 1 year post germination, whereas vegetative tillers could flower in the year produced, provided they were produced early in the year (BA Roy, University of Oregon, USA, personal observation, 2009). Population growth rates were significantly higher in the invaded range than the native range because asexual tillering, seed production from all tillers (sexual and asexual) and seed germination were all higher in the introduced range.

Physiology and Phenology

There is considerable variation among localities in numerous traits in B. sylvaticum, from morphology to phenology, physiology and pathogen resistance. Some of this variation is the result of phenotypic plasticity related to climate or other abiotic conditions, and some is related to genetic differences among populations or ranges (Paszko, 2008).

In colder climates or at higher elevations, B. sylvaticum dies back to the ground and only begins to grow after snowmelt, which can vary depending on year (BA Roy, University of Oregon, USA, personal observation, 2009). Under more favourable conditions, such as in oak forests in Hungary (Szujkó-Lacza and Fekete, 1974) and Oregon (BA Roy, University of Oregon, USA, personal observation, 2009), some shoots are able to continue photosynthesizing even in the middle of winter. B. sylvaticum is capable of germinating and growing in soils drier than many species can tolerate, perhaps because they have relatively large seeds (Evans and Etherington, 1991). Seed production, on the other hand, is strongly dependent on climate, with spring droughts in temperate forests limiting seed set (BA Roy, University of Oregon, USA, personal observation, 2009).

Under common garden conditions, Blaser (2008) found consistent evidence for higher water use efficiency (WUE, estimated with d¹³C) in plants originating from North American seeds compared to those from Switzerland. Plants from Oregon populations had higher WUE, allocated more biomass to their roots, invested less biomass in leaf area and had thicker leaves with a lower specific leaf area (SLA) than Swiss plants. No differences between origins or among populations were found for leaf mass area (LMA), photosynthesis or transpiration.

A common garden in the invaded range was used to examine genetic variation in resistance to pests (BA Roy, University of Oregon, USA, personal observation, 2009). The garden included eight populations from the native European range (ranging from Spain to England) and eight from the invaded North American range. Populations differed in both the frequency of infection and the amount of damage by the pathogens Drechslera erythrospila and Alternaria sp. and by herbivores, indicating genetic variation in resistance. On average, North American plants had significantly more damage (were less resistant) than European plants. However, plants from invaded North America nonetheless had greater reproduction.

Associations

Besides pests, which are treated in another section, there are two important associations for B. sylvaticum: endophytic infections by Epichloësylvatica and vesicular-arbuscular mycorrhizae (VAM). Infection by E.sylvatica (asexual state=Neotyphodium) ranges in effects on the host plant from mutualistic to parasitic. Infection reduces herbivory (Brem and Leuchtmann, 2001), but some strains cause choke disease, and strongly reduce fitness as they keep the plant from flowering (Bucheli and Leuchtmann, 1996; Meijer and Leuchtmann, 2000; Brem and Leuchtmann, 2003). Epichloë/Neotyphodium infection rates are high; in Europe they are 100% infected (Bucheli and Leuchtmann, 1996; Brem and Leuchtmann, 2001; A Leuchtmann, Swiss Federal Institute of Technology, Zurich, Switzerland, personal communication, 2009), and preliminary data from the USA also suggests high rates of infection in the invaded range (BA Roy, University of Oregon, USA, personal observation, 2009).

Grass endophytes are well-known for producing toxins that are poisonous to some insects as well as to mammals such as sheep and cattle. Mammals that ingest these toxins have high abortion rates and can succumb to gruesome neurological disorders (Clay, 1996; Brem and Leuchtmann, 2001). B. sylvaticum is known to be toxic to army worm (Spodoptera) larvae in ways that are consistent with the presence of alkaloids (Brem and Leuchtmann, 2001). However, which alkaloids are present is not known (Leuchtmann et al., 2000).

As is true for many grasses, B. sylvaticum commonly forms associations with AM fungi (Abeyakoon and Pigott, 1975; BA Roy, University of Oregon, USA, personal observation, 2009), which likely aid in mineral nutrition (Smith and Read, 1997). Little is known about the particular associations with B. sylvaticum, except that they occur in both the native (Abeyakoon and Pigott, 1975) and invaded ranges (BA Roy, University of Oregon, USA, personal observation, 2009), and there is variation in infection rates among populations in the invaded range.

Environmental Requirements

The environmental requirements for this plant are broad. It ranges from arid regions such as the Middle East (Bor, 1970) to the boreal zone (Holten, 1980; Aarrestad, 2000) to high elevation tropical mountains (Hara, 1966), and from sea level to more than 4000 m (Sheehy et al., 2006; Roder et al., 2007). Based on the information from references cited in the distribution table, it is most common in areas receiving 40 cm or more of rain per year, but if rainfall levels are less that this, it will persist along river and lake margins, as for example in California, USA (Hrusa, 2003) and Eastern Oregon (False Brome Working Group, 2009). In the North temperate zone, it is most often found in shady forests, or in meadows adjacent to forests (Davies and Long, 1991; Stace, 1997; BA Roy, University of Oregon, USA, personal communication, 2009). However, it grows in open grasslands in the Northern part of the UK (Stace, 1997), in extensive grasslands in Tibet (Sheehy et al., 2006) and in tropical montane and alpine meadows (Hara, 1966; Roder et al., 2007; Singh et al., 2008).

In Europe, B. sylvaticum is often an indicator of prior disturbance such as logging or coppicing (Abeyakoon and Pigott, 1975; Rodwell, 1998; Ninot et al., 2000; Corney et al., 2008; Palo et al., 2008) and in invaded North America it is also often associated with logging (False Brome Working Group, 2009; Fletcher, 2009) and transportation corridors (BA Roy, University of Oregon, USA, personal communication, 2009). It is shade tolerant (Murchie and Horton, 1998), but tends to grow in areas where the canopy is more open and conditions are better lit (Corney et al., 2008). However, it can also be found in deep shade in undisturbed forest (Hrusa, 2003; Parks et al., 2005; Palo et al., 2008; BA Roy, University of Oregon, USA, personal observation, 2009). It has a broad tolerance of soil pH, and is even able to grow in limestone-rich, rocky areas (quarries and scree) and soils (Abeyakoon and Pigott, 1975), where it can be the first colonizer (Abeyakoon and Pigott, 1975).

There are mixed reports on how well B. sylvaticum tolerates fire. Two different studies in the native range (Canary Islands and Iran) compared the vegetation of recently (within 5 years) burned to unburned areas and found that there was no B. sylvaticum in the burned areas (Arévalo et al., 2001; Safaian et al., 2005). This information conflicts with anecdotal data on control in the invaded range, which suggested that fire was ineffective because the grass re-sprouted (False Brome Working Group, 2009). It is possible that a more intense fire was needed to kill the tussocks.

B. sylvaticum is attacked by a diversity of pathogens and herbivores in both the native and invaded ranges (BPI, 2009; Halbritter, 2009). A recent study compared attack rates in 10 populations in the native range (Switzerland) to 10 populations in the invaded range (Oregon), and found that although there was greater pathogen damage in the native range, the pathogens were mostly grass generalists and thus not useful as potential biocontrol agents (Halbritter, 2009). Common pathogen genera included: Ascochyta, Alternaria, Claviceps, Colletotrichum, Didymella, Drechslera, Phaeosphaeria, and Puccinia. It is not known whether these pathogens are native or invasive in either range. Given the long history of grass introductions to the USA, it is possible that the large degree of pathogen overlap between the native and invasive ranges is the result of accidental introductions; phylogenetic studies of molecular data and population genetic analyses are necessary to disentangle the origins of the pathogens.

Interestingly, although pathogen damage was higher in the native range, herbivore damage (percent area removed) was higher in the invaded range in the USA (Halbritter, 2009). Chewing damage (holes and bitten edges) is caused by non-specific grasshoppers and by caterpillars, which tend to be grass specialists, but are otherwise non-specific, for example: the chequered or arctic skipper, Carterocephaluspalaemon, which preferred B. sylvaticum to other grasses in England (Asher et al., 2001), Maniola jurtina (meadow brown), which doesn’t prefer B. sylvaticum, but will eat it (Asher et al., 2001), the skipperlings, Thymelicus acteon and Thymelicus sylvestris (Dennis, 1992) and the browns, Pararge aegeria and Pararge xiphia (Dennis, 1992). Leaf and stem mining has been reported by numerous species of Elachista moths, the moth Helcystogramma rufescens and by three agromyzids: Agromyza albipennis, Cerodonthapygmaea and Chromatomyianigra (Pitkin et al., 2009).

Animal feed, fodder, forage

• Forage
• Invertebrate food

General

• Botanical garden/zoo
• Laboratory use
• Ornamental
• Research model

Ornamental

• Seed trade

For a grass, B. sylvaticum is relatively easy for botanists to identify because it grows in the shade of the forest (and often along trails, streams and rivers) and is quite pretty: it stays a bright vibrant green all season long, and has broad lax leaves and nodding inflorescences. Nonetheless, for identification it is best to involve a trained botanist as there are a couple of native forest grasses in North America with which it could be confused (for example, Bromus vulgaris, and possibly Melica subulata, Elymus glaucus and Bromus carinatus). Excellent identification information can be found at Piep (2003) and Shouliang and Phillips (2006).

B. sylvaticum is closely related to Brachypodium pinnatum (Catalán and Olmstead, 2000). However, the taxa are easily differentiated because B. sylvaticum has longer awns (7-15 mm vs. 0-7 mm in B. pinnatum), has nodding racemes instead of erect ones, and it does not have spreading rhizomes (Piep, 2003). Paszko (2008) adds that B. sylvaticum always has pubescent abaxial palea surfaces whereas B. pinnatum does not.

Due to the variable regulations around (de)registration of pesticides, your national list of registered pesticides or relevant authority should be consulted to determine which products are legally allowed for use in your country when considering chemical control. Pesticides should always be used in a lawful manner, consistent with the product's label.

Prevention

SPS measures

The best prevention is to prohibit it, restrict its movement (including reducing sales over the internet), to never plant it, and to teach people how to identify and remove it. It has been designated as a quarantined “A & Q” weed in California (CDFA, 2009), it is a class ‘A’ quarantined weed in Washington State, USA (eradication is required), and a class ‘B’ quarantined weed (intensive control where needed) in Oregon (ODA, 2009).

Eradication

This grass has two weaknesses despite it’s overall strength. Firstly, it is susceptible to glyphosate (False Brome Working Group, 2009), and because it stays green late into the year this allows application of the herbicide with reduced consequences for the associated native herb layer, which typically dies back during the summer. Second, it is not rhizomatous, thus once the clumps have been removed, they stay removed.

Containment/Zoning

Containment is difficult to impossible with this species due to its ability to bank seeds (Donelan and Thompson, 1980; Buckley et al., 1997), and the fact that seeds are readily spread by wildlife (Heinken and Raudnitschka, 2002).

Control

Physical/Mechanical Control

If the invaded area is small, clumps can be removed by hand. However, it does have a persistent seed bank (Donelan and Thompson, 1980; Buckley et al., 1997), thus control measures need to be maintained for several years. Mowing can be used to reduce seed set, if timed properly. However, there are problems with mowing: mowing does not decrease competition with the native species and it is not very effective for seed control because timing is critical and many plants will be missed. If mowing happens too early, the plants will re-sprout to flower anyway, and if it is too late, the mowers will simply broadcast the seed. Hot foam ( (a system that uses a hot surfactant foam to deliver and trap superheated steam onto foliage) is effective, but is only practical on roadsides and is more expensive than herbicides (False Brome Working Group, 2009).

Movement Control

B. sylvaticum is known to be carried by animals in their fur (Heinken and Raudnitschka, 2002), and by humans in their socks and shoes. The strong correlation with roadsides (BA Roy, University of Oregon, USA, personal communication, 2009) and with logging (Corney et al., 2008; Palo et al., 2008; Fletcher, 2009) also strongly suggests that it is carried by vehicles. Implementing boot/shoe cleaning stations in public parks that are known to be infested is a good idea (False Brome Working Group, 2009), and all vehicles that have been in infested forest should be thoroughly cleaned before moving to new areas (False Brome Working Group, 2009; Fletcher, 2009).

Biological Control

Biological control is not likely to be possible given the relationship with species of agricultural importance, such as wheat [Triticum aestivum] (Catalán and Olmstead, 2000; Opanowicz et al., 2008; Rathore and Shekhawat, 2009), and the fact that most grass pathogens and herbivores are generalists that attack numerous species. Even if a specific agent were to be found, given the commonness of this family worldwide, the risk of host jumping at some point would be high.

There is one specialist that is absent from North America, and that is the sexual strain of the “mutualistic” endophyte Neotyphodium/Epichloë sylvatica. When this fungus reproduces asexually, it should be called Neotyphodium and it is likely to be of benefit to the host (mutualistic), but when it reproduces sexually it should be called Epichloë sylvatica and is a parasite. Plants of B. sylvaticum in Europe are always infected by the Neotyphodium/Epichloë fungus, and usually by the asexual, or Neotyphodium, strain. However, some European populations of false-brome endophyte, have infection by the sexual strain, which should be called Epichloë sylvatica in this case. The sexual strain causes ‘choke’ in false-brome. ‘Choke’ reduces and/or prevents seed production because the fungus reproduces instead of the host. Leuchtmann and colleagues have shown that the strain that sexually reproduces is genetically different from the one that asexually reproduces (Bucheli and Leuchtmann, 1996; Meijer and Leuchtmann, 2001). Furthermore, the sexual strain is quite rare in Europe (Bucheli and Leuchtmann, 1996; Meijer and Leuchtmann, 1999). North American populations are only infected by the asexual, or Neotyphodium strain, and are thus never choked by the fungus (Halbritter, 2009; BA Roy, University of Oregon, USA, personal observation, 2009). The sexual genotype, E. sylvaticum, causes choke and thus reduces host fitness, but we don’t advise introducing it to North America for biocontrol because the fungus might jump hosts, as it has done in the past (Brem and Leuchtmann, 2003). Furthermore, the fact that the sexual strain is rare in Europe suggests that it is not favoured in natural populations and thus may not sustain itself after release, nor be effective. Substantial screening of fungal genotypes and vulnerable native species, research, and planning are required before release would be appropriate or permitted.

Chemical Control

Chemical control is generally effective as the grass is susceptible to glyphosate. More information can be found at the False Brome Working Group website (False Brome Working Group, 2009).

Control by Utilization

Until more is known about the type and quantity of toxins produced by the endophyte that almost always infects this grass (Brem and Leuchtmann, 2001), grazing is not recommended. Grass endophytes are well known for producing toxins that are poisonous to some insects as well as to mammals such as sheep and cattle (Clay, 1990; Brem and Leuchtman, 2001). Mammals that ingest these toxins have high abortion rates and can succumb to gruesome neurological disorders (Clay, 1996; Brem and Leuchtmann, 2001). Probably due to its inedibility, animals, including sheep, generally avoid this grass. A study was done to determine whether sheep could be trained to eat B. sylvaticum; they found that when sheep were exposed early in life, they did eat more of it later (Scholz, 2007). Even if it is eventually found that grazing this grass is safe, grazing will not eliminate it, although it could reduce seed production if the timing of grazing was carefully controlled.

Monitoring and Surveillance (incl. remote sensing)

Monitoring in the invaded USA range is on going, including weed control personnel at the Federal State (e.g. US Forest Service and Bureau of Land Management) and local levels. Friends of Buford Park (non-profit stewards of a 956 hectare Lane County, Oregon park) have creatively engaged volunteers to adopt areas for patrolling and controlling this grass. Surveillance with remote sensing has not yet been developed.

Ecosystem Restoration

Restoration has been a priority with The Nature Conservancy (TNC) and with local parks in the Eugene, Oregon area (e.g. Elijah-Bristow State Park and the Howard Buford Recreation Area). For large areas, glyphosate is used (see False Brome Working Group, 2009, for recommended procedures) in the late-summer or autumn after most of the native vegetation has died back. Herbicide applications continue for at least 2 years, and the area is re-seeded with natives.

- The effects of B sylvaticum on fire behaviour in Pacific Northwest forests need to be evaluated because if this grass changes forest fire susceptibility, it has the potential to be an ecosystem changer (Anzinger and Radosevich, 2008). Furthermore, the potential for fire to control the grass in the invasive range needs to be re-evaluated in light of data from the native range showing a negative association with burned areas (Arévalo et al., 2001; Safaian et al., 2005).

- The effects of B. sylvaticum on germination, growth and survival of trees in the invaded range (e.g. Pseudotsuga menziesii, Quercus garryana, Sequoia sempervirens) needs to be evaluated. It is known that the presence of grasses in general decreases conifer fitness, but it is not known whether this grass has the same effect (it could even be worse given its shade tolerance).

- For detection in meadow areas, it would be extremely useful to determine whether B. sylvaticum has a unique signal that could be detected using remote sensing. The grass has a unique colour in the visible spectrum, but it is not yet known whether it can be screened for with remote sensing. This would facilitate tracking it in open portions of roadless areas and on private lands.

- The population dynamics within the invasive range need to be better understood to facilitate control. Current practice for controlling this grass is to focus on small satellite populations, which seem possible to control, rather than the large populations that seem like a lost cause. However, there may be reasons why the small populations are less worrisome (for example, if inbreeding depression depresses seed set in small populations). Work on population dynamics could address a number of useful questions, such as: Are there source and sink populations? Are there ways to identify source populations? Does endophyte infection affect herbivory and control of population growth rates by enemies? Does inbreeding depression depress seed set in small populations? Is disturbance necessary for invasion? Does invasion occur under all light levels?

- More work needs to be done on understanding the relationships between habitat and dispersal. In Europe, the grass is associated with logging and other disturbances and it has been assumed that this relationship is the result of opening the forest up to more light (Rodwell, 1998; Corney et al., 2008). There is evidence that light may be a limiting factor given the germination and growth patterns of the grass (Grime et al., 1988; Buckley et al., 1997). However, another hypothesis for which there is some evidence is that the logging trucks and equipment move seeds around. Two lines of evidence support the logging = moving around hypothesis. First, in the invaded range there is a strong association with B. sylvaticum and the proximity of roads and paths (BA Roy, University of Oregon, USA, personal communication, 2009). Second, genetic evidence suggests that the invasion near Sweet Home, Oregon is related to the one at McDonald-Dunn forest near Corvallis, and the most parsimonious idea for spread to Sweet Home was via seeds on logging equipment (M Cruzan, Portland State University, Oregon, USA, personal communication, 2009). A third hypothesis that deserves exploration is that disturbance of the litter is in someway necessary for the grass to get a foothold (M Cruzan, Portland State University, Oregon, USA, personal communication, 2009).

- The taxonomy of B. sylvaticum needs to be re-evaluated. Numerous varieties have been described, mostly on the basis of pubescence differences (Veldkamp and Vanscheindelen, 1989; Piep, 2003; Shouliang and Phillips, 2006). However, given that hairiness can vary even within populations (Paszko, 2008), these taxa may not be real. On the other hand, it seems possible that the high elevation grassland types found in tropical alpine Asia, may well be different from the B. sylvaticum found in low elevation North European forests. The best way to analyze the variation across the entire range of B. sylvaticum would be to combine data from a common garden (to hold the environment constant so that all the variation was genetic) with molecular data such as microsatellites.

18/08/09 Original text by:

Bitty Roy, University of Oregon, Biology Department, University of Oregon, Eugene, OR 97403-1210, USA