Fróðskaparrit - 01.01.1998, Síða 15

EIN OYGGJALÍVLANDAFRØÐILIG GREINING AV FLORUNI í FØROYUM
21
thumb i.e. that a tenfold increase of the area
results in a doubling of the number of
species, corresponds to a z value of 0.30.
In areas composed by parts of continents
or lying close to same, values of 0.12-0.19
may well be expected according to
MacArthur and Wilson (1967). The reason
is that such non-isolated areas will have a
number of uncommon species which are
not adjusted to the area, but nevertheless
present because they may be found well-
established not too far away. These species
in transit will increase the immigration
speed and thus raise the number of species.
As the number of species of the small areas
is increased relatively more than that of the
bigger areas, the result will become a re-
uced z. On the other hand, the more iso-
lated the islands are, the greater the z value.
MacArthur and Wilson (1967) further
maintain that if the islands are located
closely to other islands in the group, lower
z values must be anticipated. The islands
will influence each other in terms of an in-
creased immigration rate from one island to
another. The closer the islands are located
to each other, the higher the immigration
rate and the number of species will thus in-
crease on the small islands.
The Faroese z value of 0.20 is quite con-
sistent with the above. As a whole, the
Faroes are isolated islands, which would
tend to give a high z value. On the other
hand, the group of islands is very compact,
which would bring down z.
Criticism of the depiction method
Some criticism has been directed against
the way in which materials have been de-
picted in Figure 1. Connor and McCoy
(1979) have examined 100 sets of data cor-
responding to the present in the literature
and they find that only 43 are best de-
scribed via the logS-logA model. In the re-
maining cases, the set of points depicts a
straight line just as well or better (or the de-
viation is less pronounced) via an untrans-
formed (S-A), a depiction via S-logA or a
logS-A.
In the case of the Faroese material, the
fact, however, is that it is the logS-logA
method which best transposes the points
into a straight line.
But it is a rough depiction inethod: The
method will tend to depict any set of points
as something which appears linearly if only
the dependent variable (here: number of
species) varies considerably less than the
independent variable (here: area).
There are problems in both ends of the
graph, i.e. for very large and very small ar-
eas. Even insignificant fluctuations in
terms of number of species for small areas
bring about dramatic changes in the line’s
equation. If Lítla Dímun is included in the
calculations, it means that the equation will
be changed to S = 45 A0-314. As already
mentioned, we have omitted Lítla Dímun,
on the grounds that there are doubtlessly
more than 12 species on the island.
At the other end, the graph is stable way
beyond what must be considered reason-
able. An alteration of e.g. the number of
species on Streymoy to the double amount
will cause only a minor alteration of the
line’s equation compared with the scenario
in which Lítla Dímun would be included.