Risk assessment of invasive species - Invasive Species Management - page 21

discomforting that no county in the United States is free of invasive species
(Stohlgren et al . 2006), and more are arriving all the time (see http://www.
invasivespecies.gov). h us, a formal approach to risk assessment is essential
to detect, evaluate the spread and eff ects, respond to, and monitor harmful
invasive species.

2.1.2 Current state of risk assessment for biological invaders
We conducted a cursory review of recent literature that is not exhaustive, but
rather symptomatic of the types of risk analysis approaches that readers are likely
to fi nd. Many current 'case studies' of risk analysis for invasive species fall into three
categories:

1) Species-specifi c risk assessments;

2) Habitat-specifi c risk assessments; or

3)
One of many possible examples of species-specifi c risk assessments includes pre-
dicting the introduction of West Nile virus to the Galápagos Islands (Kilpatrick
et al . 2006). h e authors devised a predictive model for the virus by evaluating
the likely 'pathways' such as avian migration, transportation of day-old chickens,
infected humans, mosquitoes in cargo containers, etc., to assess risk. h e probabil-
ities of spreading the disease by various pathways were estimated to develop pre-
vention strategies. In another example, Pemberton and Cordo (2001) evaluated
the risk of biological control on Cactoblastis cactorum in an attempt to control the
escaped cactus moth that decimates Opuntia cacti. h e release of biological con-
trol organisms also assumes elements of risk.
In other cases, several species are evaluated simultaneously to rank species for
prevention, screening, or early detection and control. For example, Tassin et al .
(2006) ranked 26 of 318 introduced woody species on Réunion Island in the
Indian Ocean as more serious invaders, based on historical records.
Habitat risk assessments attempt to describe and map the suitability or vulner-
ability of various habitats to invasion. In this way, early detection eff orts might
be guided to a subset of potential habitats. For example, riparian zones and mesic
habitats appear more prone to invasion than xeric habitats in arid landscapes
(Stohlgren et al . 1998).
Risk analysis also must consider limits to, and the connectivity of, the potential
habitat of organisms. For example, Bossenbroek et al . (2001) developed a deter-
ministic model to estimate zebra mussel ( Dreissena polymorpha ) distributions
using a distance coe cient (i.e. connectivity), Great Lakes boat-ramp attractive-
ness, and colonization potential. h is 'gravity model' constrained the risk assess-
ment to successfully forecast zebra mussel dispersal into inland lakes of Illinois,
Indiana, Michigan, and Wisconsin. h us, species and habitat risk assessments can
be combined to set priorities for management based on species traits and habi-
tat characteristics (Chong et al . 2006), greatly narrowing the number of species

Species and habitat risk assessments.