The ecology of biological invasions: past, present and future - Invasive Plants: Ecological and Agricultural Aspects - page 17

change in adventive populations has been repeatedly observed [115, 116].
Maron et al. [117] found evidence for adaptation among introduced popula-
tions of St. John's wort, Hypericum perforatum . Evolution over ecological
timescales has also been well demonstrated for Drosophila invasions [118].
Studies like these demonstrate that rapid evolution can occur in populations
of NIS, but more work is needed to show that adaptation has lead to increased
invasiveness.
Invasiveness as an evolutionary strategy (IES) hypothesis
Much of Herbert Baker's career was spent comparing 'weedy' and 'non-
weedy' congeners [102] with his eye trained towards developing a synthetic
list of the traits influencing invasiveness. The summary of this work [107, 119]
includes a popular list of characteristics of the 'ideal weed', which is used in
many regulatory frameworks: “the Baker list”. However, as studies grew in
scale and began comparing hundreds, instead of tens of species, it became
apparent that the traits Baker thought to be associated with invasiveness were
not independent of phylogeny. A very consistent finding of studies that look
for traits associated with large groups of invasive plants is that there is
increased invasion risk from plants that are closely related to an invader or that
the distribution of invaders is phylogenetically non-random. Scott and Panetta
[120] found that species in the same genus as species described as 'weeds'
were much more likely to be weedy themselves. Further, Reichard and
Hamilton's [121] criteria for rejecting potential plant invaders included having
a familial or generic invasive relative.
Along the same vein, many studies have found that nonindigenous invaders
are over-represented in relatively few taxonomic groups [122-126]. These
types of patterns are not explanations in themselves but reveal that some
underlying non-random trait distribution is likely to influence invasiveness.
For example, Daehler [125] examined lists of global natural area invaders and
found that plant families were over-represented by invasive species when they
had high proportions of woody species, contained abiotically-dispersed spe-
cies, or included climbing species.
The importance of uncovering these types of phylogenetic patterns is
twofold. First, the information gained in the above and similar studies has been
used to develop methods to predict potential invaders, which have proven
invaluable for informing management decisions where a precise mechanistic
understanding of the underlying ecology is lacking [121, 127, 128]. However,
the accuracy of such methods needs to be very high (e.g., >85%) for general
usefulness [129]. Secondly, these large-scale studies allow researchers to
uncover patterns, which the various mechanisms outlined in this chapter must
ultimately explain.