Geoscience Reference
In-Depth Information
and time independence, states that observations reflect distribution. In other
words, information on absence can be derived from observation data (Rexstad
et al. 1988; Clark et al. 1993), which is obviously seldom the case. In fact, any
time we have a record for a species we can be sure that the species (at least occa-
sionally) occurs at that location. In contrast, if there is no observation for a
species, we can only assume that we have a record of absence if there is no bias
in our sampling scheme and that we have conducted our observations over a
sufficiently long period. Even then we have no way of evaluating the random
effects that are intrinsic in observing animals.
These assumptions can have statistical relevance in dealing with induc-
tive-analytical approaches, but must hold true also for the deductive models.
If there is a constant bias in the visibility of a species' individuals, for instance
because part of their range is less accessible than others to researchers and thus
cannot be as carefully investigated, the species-environment relationship re-
flects this bias. For instance, observation data are often gathered through sight-
ings carried out by volunteers (Stoms et al. 1992; Hausser 1995), which do not
follow a predefined (e.g., random) sampling scheme. Habitat cover may limit
observations to areas where the species is visible (Agee et al. 1989). This may
create an artificial response curve that associates a positive relationship to the
values of the environmental variables measured in the locations where the
species is more visible and a negative one in the ones measured in areas were
the species has been less investigated. In such cases, we would end up mapping
the areas where the species and the observers are most likely to meet, not the
true distribution of the species.
This example is tailored to inductive-analytical models but can easily be
extended to deductive ones, both descriptive and analytical, considering that
the deductive approach is based on the a priori knowledge of specialists who
rely on series of observations to gain experience and define the species-envi-
ronment relationship. Again, these observations can suffer from accessibility
or visibility biases.
A further assumption is that observations reflect the environmental selec-
tion of the species. Obviously this is not always true; for example, occurrences
of migrant or vagrant individuals whose presence in a given location is occa-
sional may be considered among observations. An extreme case is represented
by locust swarms blown into the middle of the desert by strong winds. Clearly,
their presence does not reflect any ecological preference. Nevertheless, if we
consider only the observation per se, we would conclude that high densities of
locusts are found in the desert and that locusts do prefer (with all the limita-
tions that this term carries along in such an analysis) desert environments.
