163we looked for species that are listed as a weed in one or more of three categories:
crop weed (weeds of crops), ruderal weed (weeds of waste and disturbed areas), and
environmental weed (weeds of natural and seminatural habitats). We defi ned crop
weed values as crop weed or not a crop weed , ruderal weed values as ruderal weed
or not a ruderal weed , and environmental weed values as environmental weed or not
an environmental weed. Species classifi ed both as widely introduced and as an envi-
ronmental weed were considered to be invasive.
We used 14 independent variables in the statistical analyses. These are section, life
span, year, maximum awn length, mean seed mass, ploidy level, human use, cultivar
availability, minimum temperature, maximum temperature, temperature range, mini-
mum precipitation, maximum precipitation, and precipitation range (Table 6.2 ).
For Section, we used values Genea , Bromus , Ceratochloa , and Bromopsis. The
three species in small sections Neobromus and Nevskiella were eliminated along
with Bromus andringitrensis (Camus) about which little is known (Camus 1956 ).
We used three values for life span : short for annuals, medium for biennials and
short-lived perennials, and long for perennials. Species sometimes described as
annual and sometimes as perennial were classifi ed as having a medium life span.
Year is the year in which a species was described and named. Information for this
variable came from the Missouri Botanic Garden Tropicos database ( 2013 ).
Seeds of Bromus grasses have lemma awns of various lengths from zero to over
40 mm. For maximum awn length, we used the high value of the awn length range
in the species descriptions in the GrassBase database (Clayton et al. 2006+ ). Seed
mass data were found for 64 species and entered into the database as the mass of
1000 seeds in grams. Seed mass values came from a variety of sources, including
the USDA ARS Germplasm Resource Information Network (GRIN) database
(USDA National Plant Information System 2011 ) and the Kew Gardens Seed
Information Database SID (Royal Botanical Gardens Kew 2008 ). Seed mass values
from all sources were averaged for each species to determine mean mass in mg.
For ploidy level, species were categorized as diploid , polyploid , or both , based
on evaluation of chromosome count and ploidy level data and also on information
in the literature. These categories are non-overlapping. Species were categorized as
diploid if all or nearly all records of ploidy level found were 2N (having 14 chromo-
somes) and if the literature indicated that the species is normally considered to be
diploid. Other species were categorized as polyploid if all or nearly all records of
ploidy level found were 4N or higher and if the literature indicated that the species
is normally considered to be polyploid. All Ceratochloa species were categorized as
polyploid, as this is a characteristic of the section.
We defi ned human use with values forage / hay / revegetation and no forage / hay /
revegetation based on information about Bromus species that are currently used for
these purposes or have been used in the past. Other human uses of Bromus species
were not included in the analyses. We defi ned cultivar availability as cultivars
and no cultivars based on whether or not we found information stating that culti-
vars had been developed (named varieties intentionally bred or selected for
cultivation).
6 Attributes That Confer Invasiveness and Impacts Across the Large Genus Bromus...