SUCCESS I V E -4 PPRO X I MA T I 0 N I N DESCRIPTIVE E C OLO a T 45
widely in nature. The dominant in its various phases has a reaction on
the habitat which determines its associates, but the effect on the habitat
is temporary and the cycle repeats itself. The mosaic theory of regeneration put forward by Aubrbville (quoted by Richards, 1952) envisages a
pattern similar in kind t o this, but much larger in scale, involving populations of dominants rather than individuals.
These are all patterns which can occur on an ideally uniform site.
But there are also many kinds of pattern in the physico-chemical
environment, particularly systematic variations in soil characteristics
or micro-climate associated with pattern in the topography, which are
superimposed on the pattern caused by the vegetatioii itself.
The plant sociologist is interested in establishing correlations between
the composition of vegetation and habitat conditions. He, therefore,
always tries to describe his stands from uniform sites. It is not always
easy t o distinguish those components of the pattern which are due t o
the reaction of the vegetation and are repetitive from those which are
caused by underlying variation in the habitat ; but some approximation
can usually be made, and obviously heterogeneous stands will be detected and can be discarded at a later stage in the analysis.
Even, however, when pattern due t o the environment has been excluded as far as possible, decisions must be made about ways of dealing
with the pattern due t o the vegetation itself. Any pattern caused by
the dominant which repeats itself several times on a uniform site is
usually considered by the phytosociologist as being within the permitted
range of variation; but the various phases of the pattern may be sufficiently important t o treat separately. It is necessary to strike a balance
between making categories too broad t o be useful and multiplyiiig
micro-communities to a ridiculous extent. “The answer seems t o be
that neither alternative is necessarily right or wrong ; each case should
be judged on its merits and different levels of abstraction niay be appropriate for different ends. But in a floristic survey of a region some considered and consistent standpoint is desirable. The most practicable
resolution of the problem seems t o be as follows: if one element of a
mosaic occurs in very similar or identical form as a pure community
independently of the other elements with which it is usually associated,
it should be described separately ; simiiarly, if one element is associated
with more than one other element t o form different mosaics, it should
again be treated as an independent entity” (Poore, 1955b). This also
provides a practioable way of dealing with the problem of synusiae.
In any community of complex structure there are smaller groups of
species which are closely related ecologically and behave t o a certain
extent independently of the community as a whole. Such are, for
example, communities of corticolous bryophytes, herb societies of the
widely in nature. The dominant in its various phases has a reaction on
the habitat which determines its associates, but the effect on the habitat
is temporary and the cycle repeats itself. The mosaic theory of regeneration put forward by Aubrbville (quoted by Richards, 1952) envisages a
pattern similar in kind t o this, but much larger in scale, involving populations of dominants rather than individuals.
These are all patterns which can occur on an ideally uniform site.
But there are also many kinds of pattern in the physico-chemical
environment, particularly systematic variations in soil characteristics
or micro-climate associated with pattern in the topography, which are
superimposed on the pattern caused by the vegetatioii itself.
The plant sociologist is interested in establishing correlations between
the composition of vegetation and habitat conditions. He, therefore,
always tries to describe his stands from uniform sites. It is not always
easy t o distinguish those components of the pattern which are due t o
the reaction of the vegetation and are repetitive from those which are
caused by underlying variation in the habitat ; but some approximation
can usually be made, and obviously heterogeneous stands will be detected and can be discarded at a later stage in the analysis.
Even, however, when pattern due t o the environment has been excluded as far as possible, decisions must be made about ways of dealing
with the pattern due t o the vegetation itself. Any pattern caused by
the dominant which repeats itself several times on a uniform site is
usually considered by the phytosociologist as being within the permitted
range of variation; but the various phases of the pattern may be sufficiently important t o treat separately. It is necessary to strike a balance
between making categories too broad t o be useful and multiplyiiig
micro-communities to a ridiculous extent. “The answer seems t o be
that neither alternative is necessarily right or wrong ; each case should
be judged on its merits and different levels of abstraction niay be appropriate for different ends. But in a floristic survey of a region some considered and consistent standpoint is desirable. The most practicable
resolution of the problem seems t o be as follows: if one element of a
mosaic occurs in very similar or identical form as a pure community
independently of the other elements with which it is usually associated,
it should be described separately ; simiiarly, if one element is associated
with more than one other element t o form different mosaics, it should
again be treated as an independent entity” (Poore, 1955b). This also
provides a practioable way of dealing with the problem of synusiae.
In any community of complex structure there are smaller groups of
species which are closely related ecologically and behave t o a certain
extent independently of the community as a whole. Such are, for
example, communities of corticolous bryophytes, herb societies of the