MEPS 332:41-51 (2007)  -  doi:10.3354/meps332041

ABSTRACT: Despite mounting evidence of invasive species’ impacts on the environment and society, our ability to predict invasion establishment, spread, and impact are inadequate. Efforts to explain and predict invasion outcomes have been limited primarily to terrestrial and freshwater ecosystems. Invasions are also common in coastal marine ecosystems, yet to date predictive marine invasion models are absent. Here we present a model based on biological attributes associated with invasion success (establishment) of marine molluscs that compares successful and failed invasions from a group of 93 species introduced to San Francisco Bay (SFB) in association with commercial oyster transfers from eastern North America (ca. 1869 to 1940). A multiple logistic regression model correctly classified 83% of successful and 80% of failed invaders according to their source region abundance at the time of oyster transfers, tolerance of low salinity, and developmental mode. We tested the generality of the SFB invasion model by applying it to 3 coastal locations (2 in North America and 1 in Europe) that received oyster transfers from the same source and during the same time as SFB. The model correctly predicted 100, 75, and 86% of successful invaders in these locations, indicating that abundance, environmental tolerance (ability to withstand low salinity), and developmental mode not only explain patterns of invasion success in SFB, but more importantly, predict invasion success in geographically disparate marine ecosystems. Finally, we demonstrate that the proportion of marine molluscs that succeeded in the latter stages of invasion (i.e. that establish self-sustaining populations, spread and become pests) is much greater than has been previously predicted or shown for other animals and plants.

KEY WORDS: Invasion · Bivalve · Gastropod · Mollusc · Marine · Oyster · Vector · Risk assessment