Biology Reference
In-Depth Information
attributes or their relationship to native species (Jarnevich et al. 2006; Stohlgren
et al. 2006; Stohlgren et al. 2003). Additionally, modeling approaches commonly
attempt to predict the potential distribution of individual invasive species using
snap-shot-in-time datasets, meaning a dataset collected over a short time for a spe-
cific location (Elith et al. 2006). Although useful, these models generally provide
no estimate of when a species may arrive at a particular location - only that it may
at some unknown future point. Predictions for individual species' potential abun-
dance or rates of spread are less common, often due to lack of data. As technology
has advanced over the past couple of decades, spatial models have become more
sophisticated and accurate. These models have proven valuable in the management
of invasive species. The increased availability of geographic positioning system
(GPS) technology has also helped data collection. However, these forecasting tools
are still limited by gaps in the field data being collected and synthesized to calibrate
or independently validate their predictions. There are also still limitations in our
ability to model natural systems where nonnative species establish and spread.
In a recent survey of existing invasive species databases in the United States,
Crall et al. (2006) found that 38% of the 254 databases discovered contained data
covering 10 or more years. This survey covered databases, so groups collecting
field data stored in a less technologically advanced system were not included. Thus,
this number ignores many collections, but is probably somewhat reflective of real-
ity, with less than half of the datasets holding long-term data. Additionally, 82 of
the datasets covered an area equal to a county or smaller. Even the datasets that do
exist may not always be readily available and in the same format. Data integration
would help solve some of these limitations. To effectively manage invasions, we
need information across broad spatial extents and over long time periods as inva-
sions occur over these scales.
Partly due to the lack of these data, predictions including a specific temporal
component are much rarer in the literature than spatial predictions (i.e., a species'
potential habitat). A literature search in Web of Science including the terms spatial,
modeling, and invasive revealed many articles, while one including temporal or
time, modeling, and invasive revealed a dearth of articles published in peer reviewed
journals.
Invasive species management involves many concepts and careful consideration
of analysis techniques. There are several important points to keep in mind when
creating predictions of species spread in addition to specific individual species
traits. Range expansion of a species will be a function of the number and spatial
arrangements of introductions, time since invasion, propagule pressure (frequency
of propagules), a vector for dispersal, seeds being dispersed to a favorable location,
hybridization, and many other factors. Forecasting invasions including richness,
distribution, and abundance of invasive species with a temporal component as
opposed to species distribution models that predict potential distribution regardless
of time can be accomplished with several different methods. The most appropriate
analysis method for forecasting a particular invasion may vary depending on the
spatial resolution, the species, and the stage of invasion. Additionally, managing
species effectively through time involves assessing the long-term potential
