SITCC’ESSIVE APPROXIMATION I N DESCRIPTIVE ECOLOGY
47
IV. DISCUSSION OF T H E STEPS USED IN THE
PHYTO-SOCIOLOGICAL EXAMINATION O F VEGETATION
A. T H E CHOICE O F S T A N D A N D T H E CONCEPT O F UNIFORMITY
It has frequently been emphasized that the problems of the synecologist begin with his choice of stand.
Whatever analysis of the community he intends t o carry out and
however exact the techniques he intends to apply, they will be deeply
influenced by this initial choice. Much stress has been laid on the “uniformity” or “homogeneity” of the stand which should be chosen for
phytosociological sampling. Most field ecologists recognize that, in any
region, there are stands which are more uniform than others and their
choice of stand for sampling is confined to those “more uniform” areas
and especially to the communities which, after careful reconnaissance,
they consider t o occur frequently with more or less the same species
composition. Whatever the ecological basis of this uniformity and frequency of occurrence - and there are several possibilities - the
characterization of these types does provide a satisfactory and logically
sound basis for an empirical classification of the vegetation. This is a
Rubjective procedure ; its justification lies in its proven usefulness and
the fact that the judgements of experienced ecologists tend t o coincide.
The discussion of homogeneity in Goodall (1954b) is very valuable
in illustrating the approach of the quantitative ecologist. He poses the
central question in the introduction t o his discussion : “Plant sociology
is at a disadvantage compared with the other observational sciences in
that the units with which it deals are not naturally delimited. The
zoologist is in no doubt where one animal ends and an individual of
another species begins ; and the plant systematist rarely has difficulty
in detecting the spatial limits of the tissue of his systematic units; but
the plant sociologist must draw artificial lines separating portions of
vegetation from contiguous more or less similar vegetation before he has
the wherewithal for c1asaificat)ion.”
He admits that absolute homogeneity is foreign to vegetation and
suggests three ways of attempting to assess homogeneity. The first is
based on the random distribution of individuals of all species in the
community, a requirement which he suggests would be considered
too rigorous by most plant sociologists. That it certainly would be so
considered is shown by the inclusion of an index of sociability in the
standard description of a releve by the Zurich-Montpellier school; and
by the fact that Nordhagen, for example, takes ten samples of uniform
size in each stand t o obtain an estimate of heterogeneity.
The second method is to consider the relation between the variance
between sample areas and their spacing. If this variance is constant
47
IV. DISCUSSION OF T H E STEPS USED IN THE
PHYTO-SOCIOLOGICAL EXAMINATION O F VEGETATION
A. T H E CHOICE O F S T A N D A N D T H E CONCEPT O F UNIFORMITY
It has frequently been emphasized that the problems of the synecologist begin with his choice of stand.
Whatever analysis of the community he intends t o carry out and
however exact the techniques he intends to apply, they will be deeply
influenced by this initial choice. Much stress has been laid on the “uniformity” or “homogeneity” of the stand which should be chosen for
phytosociological sampling. Most field ecologists recognize that, in any
region, there are stands which are more uniform than others and their
choice of stand for sampling is confined to those “more uniform” areas
and especially to the communities which, after careful reconnaissance,
they consider t o occur frequently with more or less the same species
composition. Whatever the ecological basis of this uniformity and frequency of occurrence - and there are several possibilities - the
characterization of these types does provide a satisfactory and logically
sound basis for an empirical classification of the vegetation. This is a
Rubjective procedure ; its justification lies in its proven usefulness and
the fact that the judgements of experienced ecologists tend t o coincide.
The discussion of homogeneity in Goodall (1954b) is very valuable
in illustrating the approach of the quantitative ecologist. He poses the
central question in the introduction t o his discussion : “Plant sociology
is at a disadvantage compared with the other observational sciences in
that the units with which it deals are not naturally delimited. The
zoologist is in no doubt where one animal ends and an individual of
another species begins ; and the plant systematist rarely has difficulty
in detecting the spatial limits of the tissue of his systematic units; but
the plant sociologist must draw artificial lines separating portions of
vegetation from contiguous more or less similar vegetation before he has
the wherewithal for c1asaificat)ion.”
He admits that absolute homogeneity is foreign to vegetation and
suggests three ways of attempting to assess homogeneity. The first is
based on the random distribution of individuals of all species in the
community, a requirement which he suggests would be considered
too rigorous by most plant sociologists. That it certainly would be so
considered is shown by the inclusion of an index of sociability in the
standard description of a releve by the Zurich-Montpellier school; and
by the fact that Nordhagen, for example, takes ten samples of uniform
size in each stand t o obtain an estimate of heterogeneity.
The second method is to consider the relation between the variance
between sample areas and their spacing. If this variance is constant