Biology Reference
In-Depth Information
models such as GARP and Bioclim now predict the potential new range of an inva-
sive species by identifying regions that are climatically similar to the species'
native range (Sutherst and Maywald 1985; Peterson and Vieglais 2001; Peterson
2003; Thuiller et al. 2005). This approach is most powerful when combined with
analysis of other factors, such as soil type, that can influence invasive plant distribu-
tion. Finally,
experiments
can be conducted to test empirically the predictions made
from the four approaches described earlier (Rejmánek 2000). However, the time lag
that is inherent in most invasive episodes (Kowarik 1995) usually makes such
experiments unappealing for land managers (Radosevich et al. 2005).
17.4.1
Need for Surveys and Monitoring
Rew et al. (2005) indicate that, historically, surveys have been linked to invasive
plant management, while monitoring may not occur until after management activity
has happened, if at all. Rew et al. also indicate that prior information should be used
to develop any survey scheme in order to maximize finding invasive plants. For
example, since invasive plants are most often introduced by humans, the vectors of
human transport such as roads and rails should be examined. Although some excep-
tions exist, occupancy by invasive plants usually declines as plant cover increases
such that a gradient of decreasing occupancy from areas of low to high cover should
be expected. Rew et al. (2005) offer an approach (Fig. 17.5) where monitoring is
used to improve the reliability of management tactics in areas inhabited by invasive
plants. They suggest that land managers (1) develop monitoring plans, (2) select
methods that will quickly meet monitoring objectives, and (3) clearly link monitor-
ing output to management decisions. Rew and Pokorny (2006) describe both on-
the-ground and remote sensing survey and monitoring approaches used in western
North America to assess invasive plant occurrence.
17.4.2
Risk Assessment Models
Although several models of range expansion from source populations have been
developed for invasive plants, this approach has not been incorporated widely into
management decisions. Only recently have simulation models of invasion been
developed that incorporate management options and outcomes (Goslee et al. 2006;
Kriticos et al. 2003). Endress (in Radosevich et al. 2005) provides an example of
how to construct a risk assessment model based on the susceptibility of native plant
communities to invasion, disturbance history of sites, and proximity to current
infestations. Although models of risk assessment can be valuable tools for land
managers, they require good information on species biology, site characteristics,
and reliable position coordinates for an area, watershed, or region. Heger and Trepl
(2003) and Prather (2006, in press) also describe computer programs that predict
