Sarah Reichard
University of Washington, Center for Urban Horticulture

Dr. Reichard's research involves a study of woody invasive plants in North America. She has identified 236 species in 125 genera and 53 families that are successful invaders living in North America. Successful invaders are defined as species introduced far outside their native ranges which are able to establish themselves and reproduce outside of cultivation, as if they were native. Unsuccessful invaders are defined as species introduced to North America prior to 1930, that have never been documented as existing outside of cultivation; Reichard has identified 114 species, 80 genera, and 48 families of unsuccessful invaders.

Reichard began by identifying a list of characteristics that "conventional wisdom" says make species successful invaders. Preliminary observations and anecdotal evidence are the sources for this list. Not all of these generalizations have been proven; indeed, some may well turn out to be false. This list is not based on large scale surveys or statistical analysis, so we shouldn't accept this as "truth." With that caveat:

Successful invaders, in general:

1. have widespread distributions and are abundant in their native range (habitat generalists; possibly higher genetic variation; higher probability of being introduced, especially accidentally)

2. have high genetic variability (high heterozygousity provides more adaptability to varying conditions)

3. have a close correspondence of climatic regions and habitat between their native and invaded ranges (having evolved under certain conditions, they won't thrive under radically different conditions)

4. are feeding generalists (animals) or tolerant of varied soil conditions (plants) (more adaptable to unpredictable conditions in the invaded range; there is very little support for this generalization, little predictive ability)

5. are successful because they are released from the predators in their native range (removal from predators allows more individuals to reach reproductive stage; Reichard says there is very little support for this in the plant world)

6. are associated with humans for dispersal and/or resource enhancement (all of the 236 species Reichard found to be successful invaders were introduced by humans; in addition, humans degrade site conditions, facilitating invasions)

7. have highly effective dispersal mechanisms, in addition to humans (for invader to expand, it must be able to disperse and build populations away from the mother population; successful woody plant invaders are 60% biotically dispersed and 40% abiotically dispersed)

8. have short juvenile periods (allowing them to reach reproductive age quickly, accelerating the population size; successful woody plant invaders mean juvenile period is 4.03 years, while that of unsuccessful invaders is 6.9 years)

9. are self-compatible or able to colonize from a single fertilized female (effective dispersal may place individuals far from others and reproduction must continue for invasion to succeed; successful woody plant invaders are 44% self-compatible compared to 25% self- compatibility among unsuccessful invaders)

10. have high reproductive output and are therefore able to build populations quickly (invasions progress faster and lead to greater success if populations increase quickly; examples include high output of zebra mussels, fire ants)

We should stop thinking of these things as truths; they are really hypotheses and require more research to test them. We also need to link research to action and use this type of information to perform triage ‹ to decide which invaders we can and should seek to control or eliminate. There is some discussion around questions such as: Should we use this approach to predict which species should be excluded? And once we make such a prediction, can we use a blanket approach to exclude whole classes of species, and place the burden of proof on those who wish to bring in species that fit a set of characteristics?

Note: The "missing predator" characteristic is the one most often sited in the popular press (because it is easy to understand), but it is not, in Reichard's opinion, the most significant factor in a successful invasion. There is a suite of factors that make an invasion successful.

The information that follows is based on Dr. Reichard's research and analysis, as opposed to the preceding material, which is anecdotal in nature.

Can we predict invasions of native plants?

Reichard conducted a discriminant analysis of 236 successful invaders and 114 non-invasive woody plant species introduced to North America in order to identify variables that distinguish the two groups. These variables are largely related to increased reproductive output and stress-tolerance, although the knowledge that a species invades elsewhere in the world is the best single indicator.

From this discriminant analysis, as well as characteristic comparisons, and classification and regression trees, a predictive flow chart was developed as a practical method of evaluating the risk of invasiveness (see Figure 1). The discriminant analysis model was created using 149 species (75%) selected at random, and validated on the remaining 58 species (25%). Based on this model, Dr. Reichard would have denied admission to the US to 86% of the invasive species and would have admitted 70.8% of non-invasive species. The predictive flow chart, tested on all of the successful and unsuccessful invaders, admitted no invasive species (although 11% were recommended for possible admission following further analysis) and admitted 38% of the non-invasive species (with another 39% recommended for possible admission following further analysis).