Integrating landscape connectivity and habitat suitability to guide offensive and defensive invasive species management.

Published online
01 Apr 2015
Content type
Journal article
Journal title
Journal of Applied Ecology

Stewart-Koster, B. & Olden, J. D. & Johnson, P. T. J.
Contact email(s)

Publication language
USA & Michigan & Wisconsin


Preventing the arrival of invasive species is the most effective way of controlling their impact. Preventative strategies may be 'offensive' aimed at preventing the invader leaving colonised locations or 'defensive' aimed at preventing its arrival at uninvaded locations. The limited resources for invasive species control must be prioritized, particularly for numerous vulnerable locations or uncertainty about which sites are already invaded. We developed an integrative modelling framework to prioritise locations for either strategy by incorporating connectivity and habitat suitability. We applied this framework to a data set comprising 5189 water bodies in Wisconsin and Michigan, U.S.A, for zebra mussels Dreissena polymorpha and Eurasian watermilfoil Myriophyllum spicatum. We developed the framework with a spatial graph based on recreational boater movement and habitat suitability models. An historical graph comprised 3105 natural lakes connected in one of 18 components, whereas a total of 3944 water bodies (lakes and reservoirs) were connected in one of 13 separate components in a graph of the contemporary system. Habitat suitability models accounted for around half of the deviance in the distribution data for each species. There was a distinct spatial pattern in the levels of risk and subsequent recommended allocation of management interventions across several levels of investment. Higher risk water bodies were generally found in the largest component of the spatial graph. At comparatively low levels of investment, where managers target 5% of all locales to control D. polymorpha, the results suggested that 71% and 27% of this effort should be committed to defensive and offensive strategies, respectively, in the largest component. For M. spicatum, 92% and 8% of this effort should be allocated in this component to defensive and offensive strategies, respectively. It is only with much greater investment that water bodies in other components should be targeted. Synthesis and applications. Allocating limited resources to prevent the spread of invasive species is a challenge that transcends ecosystems and geography. We successfully identified a reduced number of locations to target for offensive and defensive intervention strategies for two species. This framework is readily applicable to other aquatic and terrestrial ecosystems vulnerable to invasive species.

Key words