Ecological niche models and the geography of biological invasions: a review and a novel application - Invasive Plants: Ecological and Agricultural Aspects - page 30

The validity of the predicted distribution that results when the niche model
is projected onto a new geographic region depends on the assumption that the
niche of the species, as modeled, will be conserved when the species invades
this region. In other words, the factors governing the distributional limits of a
species in its native range also will govern the distributional limits of the spe-
cies in its invaded range, and that these factors are included in the model.
Because species can adapt to new environments, its niche may not be con-
served when the species invades this region. It has been argued that niche evo-
lution (i.e., evolutionary change in niche dimensions such that ecological nich-
es are not conserved over time) is a primary limitation on the predictability of
the geography of invasions. However, in response to this criticism, Peterson [8]
argued that empirical and theoretical evidence demonstrates the conservation
of ecological niches over evolutionary time, and he concluded that evolution
of niche space was not a serious limitation for ENM.
Few ecologists would question that climate is a dominant factor in structur-
ing the distribution of species at macro-scales. Many published niche model-
ing studies that successfully simulated the current distributions of species sup-
port this assertion. Difficulty arises when niche models are used to predict dis-
tributions in new settings because biotic interactions and dispersal rates are
likely to be different in a new geographic and biological context. These issues
could be especially problematic for predicting the distributions of invasive spe-
cies because invaders may leave behind competitors, pathogens, and other nat-
ural enemies. Indeed, release from harmful biotic interactions (i.e., as in the
enemy-release hypothesis (ERH) for invasive plants [15]) is a commonly
invoked explanation for the observed increases in fecundity, competitive abil-
ity or abundance of invasive species in their newly invaded range [15-17].
However, only one study [18] has provided evidence that release from enemies
enables an invasive species to exploit a habitat within its invaded range that it
could not exploit in its native range. Mack et al. [16] cited several cases in
which invaders became established in climatic regions unlike those that would
have been predicted based on knowledge of their native range, but mechanisms
for the differences were not discussed. To our knowledge, none of these stud-
ies have been analyzed using ENM. Therefore, it is unclear if ENM would
have failed to predict accurately the potential invaded distributions of these
species. Yet, differences in performance and habitat use by invaders in their
new ranges suggest that factors other than environmental conditions alone
mediate their success and that particular constraints on the distribution of spe-
cies are not conserved when an invader spreads to a new region. In conclusion,
it is not whether ecological niches tend to be conserved, but rather, whether the
niche a species occupies in its invaded range generally differs from the niche
the species occupies in its native range.
One potential way to evaluate if an invasive species occupies a different
niche in its invaded range than its native range is to compare the predicted
invasive distribution with the actual extent of invasion (i.e., a posteriori assess-
ment of the accuracy of the prediction). Several such comparisons conducted