Environmental Engineering Reference
In-Depth Information
Iterative sampling for invasive species
Initial phase
Second phase
Iterative model
refinement phase
Subjectively sample
known locations
(location, cover, area)
Add stratified-random
sampling component
to assess mean conditions
within envelope
Add more stratified-
random samples
Add opportunistic
samples of locations
Note presence/absence,
cover, area
Add gradient
sampling component
to assess extreme
gradients within
suitable envelope and
to assess unsuitable
boundary
and/or
Add more
gradient samples
Iterative models
Validate all previous data
Continue to refine previous
models as new information
becomes available
Second approximation
model
Validate subjective data and
refine the first model
First approximation model
Environmental envelope of
presence/absence
Logical strata, based on suitable
habitats, major environmental
gradients, or TM heterogeneity
classes
Identify information gaps (soils,
other data)
Outputs: (1) current and potential distribution models;
(2) priority sites for control and restoration; (3) potential early
detection sites; and (4) probability and uncertainty analyses.
Assess model improvements
Fig. 2.3
An iterative sampling approach for documenting, mapping, and
predicting the abundance, distribution, and spread of invasive species.
h e second phase of surveys integrates unbiased stratifi ed-random sampling
with gradient sampling designs for robust spatial statistical models (Fig. 2.3). h is
provides much needed information on the probability of occurrence in diff erent
habitat types, and preliminary information on the environmental tolerances of the
target species. Note that the actual current and potential distributions of a species
are very di
cult to determine from limited surveys—but it's a start. Statistical and
spatial interpolation models based on stratifi ed random and gradient analysis tech-
niques allow for a second approximation of species' distribution and abundance.
New survey data from opportunistic sampling, stratifi ed random sampling, and
gradient sampling further improve and validate the distribution maps over time.
Since species migrate, adapt, hybridize, expand, and contract in population size,
risk analysis surveys for invasive species must be an iterative process (Fig. 2.3).
Another critical feature of risk analysis for invasive species is an understanding
of the taxonomic, geographic, and temporal completeness of data in the region
