Geoscience Reference
In-Depth Information
Thompson 1990; Aspinall 1992; Flather et al. 1992; Aspinall and Veitch
1993), whereas at the most local scales, indigenous land use structures become
increasingly significant (Thomasma et al. 1991; Picozzi et al. 1992; Herr and
Queen 1993) to the extent that even an individual stand of timber (Pausas et
al. 1995) or a single pond (Genard and Lescourret 1992) can play a role. Gen-
erally speaking, the factors that are important vary according to scale, meaning
that factors that are important at one scale level can lose their importance
(Noss 1992), or at least much of it, at others.
As with any type of classification, the relationship between scale and envi-
ronmental variables that drive ecological processes should not be taken too
rigidly, and although most authors tend to agree that for broader scales climate
is the most important factor, the same cannot be said when trying to identify
the driving forces at finer scales. For instance, variables considered useful at
coarser scales are used in detailed studies, as in the cases of Pereira and Itami
(1991) and Ross et al. (1993), which use topography to explain species distri-
bution at a much finer scale than the regional one. The same consideration
applies to the studies of Aspinall and Matthews (1994), which use climatic
data on a regional scale. On the other hand, land use is often used in distribu-
tion models developed at regional scale (Livingston et al. 1990; Flather and
King 1992).
Finally, we must consider that distribution is the result of the interaction of
many different biological events and that an ecological event cannot be
described exhaustively on any single specific scale, but is the result of complex
interactions of phenomena happening at different scales (Levin 1992; Noss
1992). Thus the limit of the applicability of a given environmental variable to
describe distribution on any given scale may not be so sharp and the challenge
is toward the integration of different scales in the description of the species'
distributions. Buckland and Elston (1993) gave an example of the integration
of environmental variables stored at different resolutions within the same dis-
tribution model.
It is important to note that the concept of scale not only determines the
biological extent to which a distribution model can be applied but also affects
the use that can be made of such a model for conservation. Also, conservation
actions can be seen as having a hierarchical approach (Kolasa 1989). For
instance, Scott et al. (1987) identified six different levels of intervention: land-
scape, ecosystem, community, species, population, and individual. Not sur-
prisingly, conservation actions tend to become more effective and less expen-
sive when the assessment moves toward broader scales, that is, when one moves
from the individual to the landscape approach (Scott et al. 1987). Obviously
