Doria R. Gordon
Environmental Defense Fund, Washington DC, USA and University of Florida, Department of Biology, Gainesville, Florida, USA
Increasing evidence suggests that many of the species predicted to have a low risk of invasion using trait-based weed risk assessment (WRA) tools will become invasive given sufficient time in the introduced range. However, if minimum residence time and propagule pressure are the ultimate predictors of invasion for casual species, why does WRA have high accuracy in differentiating invaders from non-invaders? I hypothesize that the risk predicted by the WRA may be the risk of more rapid than slower transition from a casual to invasive status. I will examine evidence for this hypothesis from multiple locations and explore the implications.
The WRA was developed for precautionary purposes: to preclude import of species with high probability of becoming harmful invaders. Climate and environmental constraints are included in most WRA tools, but much of the emphasis is on plant traits. While we have limited data on introductions that result in unsuccessful establishment, species are unlikely to naturalize where environmental conditions preclude survival without cultural treatments, regardless of propagule pressure. As the proportion of intentionally introduced species has increased over time, the probability of environmental and climate matching has increased. Repeated purposeful introduction and cultivation increases not only propagule pressure over larger areas, but also the probability of adaptation via greater genetic diversity. Species traits appear to have more influence on timing rather than the probability of invasion, with different types of traits more important at different stages in the invasion process. For example, reproductive traits appear more critical as species move from the casual to the naturalized stage and dispersal traits become more important for transition to the invasive stage. Life history, therefore, should also influence the rate of these transitions.
The relative speed of invasion may be critical for predicting potential impact, control cost and control success. Native species may have a lower probability of persisting with a non-native that becomes invasive rapidly after introduction, where the non-native species has high growth rates and develops high biomass. Area invaded is inversely correlated with eradication success and positively with management cost. As a result, WRA and other efforts developed to identify predictive traits of invaders appear as relevant for screening imports pre-border as well as for post-border regulation and management. Species with traits correlated with invasion, including high WRA scores, have a higher probability of costly environmental and economic impacts on the timeframe of decades, consistent with the regulatory timeframe. While perhaps not predicting the intended risk, pre- and post-border use of WRA for identifying species for noxious designation or management focus likely has the intended management result. As such, we can successfully protect native species and functions through deliberate focus on species likely to be early invaders and on those moving slowly but are likely to be system changers. Accurate identification of the latter may require additional assessment tools.