ST’CCESSIVE A P P R O X I M A T I O N I N DESCRIPTIVE E C O L O G Y
43
Jevons (1913) pointed out another most important feature of classification : “A class must be defined by the invariable presence of certain
common properties.” This requirement is not met by a classification
of vegetation according to fGithfu1 species. Poore (1956) has quoted an
example of an association, the Fabronietum pusillae (Barkman, 1950).
The published table of this association contains 14 lists and 7 faithful
species; but of these lists, one contains 5 of the faithful species, one 3,
two 2, eight 1 and two 0. Faithful species may be very useful in the
diagnosis of an association, but they cannot be used as the sole basis of
definition. In fact these communities are probably correctly placed in
the Fabronietum pusillae, but the association is actually defined not by
the faithful species but by other unspecified characteristics such as
habitat and constant species.
The “type” system used in taxonomy may seem at first sight t o contradict this principle, but, as Jevons wrote, it is only “an abbreviated
mode of representing a complicated method of arrangement”. The
“type” is a reference point which may be returned to in the light of
new evidence, or t o make certain of the intentions of the author who
originally described a species.
If we apply these principles t o the classification of vegetation, we
must recognize that there are numerous ways in which it may be
classified for different purposes. A chorological classification will reveal
relationships and permit generalizations which would not be possible
with a classification according t o life form and vice versa. Which of the
various classifications advanced is the most natural?
I think that the majority of ecologistswould agree that the generalizations most valuable to them are those which establish relationships between vegetation and its habitat, and that the most natural classification, in the sense given above, would be that which faithfully
reflects these relationships. Several such have been proposed and will
be discussed at a later stage. It is recognized that a classification of this
kind will only partially serve the needs of the plant geographer, t o
whorn some other will be most appropriate. Nevertheless the discussion
in this paper will be confined to the various systems which try in varying degrees to reflect faithfully the relationships between Vegetation
and habitat.
111. THE PROBLEM OF PATTERN
There are two great difficulties in the classification of plant communities and in the use of descriptions t o establish correlations between
vegetation and habitat. One is the frequent lack of clear-cut boundaries
between communities - a difficulty which will be discussed later; the
other is the prevalence of complex spatial patterns in vegetation. In
43
Jevons (1913) pointed out another most important feature of classification : “A class must be defined by the invariable presence of certain
common properties.” This requirement is not met by a classification
of vegetation according to fGithfu1 species. Poore (1956) has quoted an
example of an association, the Fabronietum pusillae (Barkman, 1950).
The published table of this association contains 14 lists and 7 faithful
species; but of these lists, one contains 5 of the faithful species, one 3,
two 2, eight 1 and two 0. Faithful species may be very useful in the
diagnosis of an association, but they cannot be used as the sole basis of
definition. In fact these communities are probably correctly placed in
the Fabronietum pusillae, but the association is actually defined not by
the faithful species but by other unspecified characteristics such as
habitat and constant species.
The “type” system used in taxonomy may seem at first sight t o contradict this principle, but, as Jevons wrote, it is only “an abbreviated
mode of representing a complicated method of arrangement”. The
“type” is a reference point which may be returned to in the light of
new evidence, or t o make certain of the intentions of the author who
originally described a species.
If we apply these principles t o the classification of vegetation, we
must recognize that there are numerous ways in which it may be
classified for different purposes. A chorological classification will reveal
relationships and permit generalizations which would not be possible
with a classification according t o life form and vice versa. Which of the
various classifications advanced is the most natural?
I think that the majority of ecologistswould agree that the generalizations most valuable to them are those which establish relationships between vegetation and its habitat, and that the most natural classification, in the sense given above, would be that which faithfully
reflects these relationships. Several such have been proposed and will
be discussed at a later stage. It is recognized that a classification of this
kind will only partially serve the needs of the plant geographer, t o
whorn some other will be most appropriate. Nevertheless the discussion
in this paper will be confined to the various systems which try in varying degrees to reflect faithfully the relationships between Vegetation
and habitat.
111. THE PROBLEM OF PATTERN
There are two great difficulties in the classification of plant communities and in the use of descriptions t o establish correlations between
vegetation and habitat. One is the frequent lack of clear-cut boundaries
between communities - a difficulty which will be discussed later; the
other is the prevalence of complex spatial patterns in vegetation. In