36
M. E. D. POORE
It is the common experience of plant ecologists that vegetation is
not a random assemblage of individuals of many species, but that
plants are associated in communities, which have a definite structure
and often a regular specific composition. Only the species which are
adapted t o a particular habitat are selected from the available propagules of the total flora of a region, and the number of these is further
restricted by considerations of space and by mutual compatibility in
requirements for nutrients and light. The resulting communities vary
in their degree of cohesion, which is least in transitory populations of
ephemerals with single-layer structure and greatest in the tropical
forest with its complex development of interrelated synusiae.
The community is one of the key concepts in the science of vegetation, and data in all branches of the subject should be treated in the
community context. For example the plant geographer cannot consider
the migration of taxa in isolation; for every species must migrate in a
chain of suitable environments and in communities where a niche for it
exists. No more can the autecologist consider the behaviour of a single
species out of the context of the community, for its physiological
optimum is frequently modified by competition and does not correspond
with its ecological optimum.
The exact description and characterization of the community bhus
becomes essential t o all who work in the science of vegetation ; for it is
important that ecological findings should be related to communities
which are well described. Not t o do so is tantamount t o carrying out a
physical experiment without stating the conditions. Many methods have
been developed, for both the description and the characterization of
communities, which vary from subjective and qualitative t o elaborately
quantitative. The choice of method is a personal matter influenced
partly by the purpose of the investigation and partly by the particular
view taken by the worker of the nature of vegetation. But it should be
remembered that, however exact the method, it cannot describe all the
attributes of even the simplest community a t one moment of time.
Every description is an abstraction from the available data.
The argument so far is that, to be fully useful, ecological observations
must be specified in terms of the community in which they are made.
Nevertheless, even if this is done, no inductive generalizations are possible without classification. Without it any observation may be only an
isolated case and consequently of very limited general value. This need
for classification is quite independent of the subject matter ; it, applies
equally to any objects of which we wish t o think, whether it be books,
musical instruments, the shape of clouds or human emotions.
Some phenomena are easier t o classify than others. This depends on
how distinct is the discontinuity between one class and those nearly
M. E. D. POORE
It is the common experience of plant ecologists that vegetation is
not a random assemblage of individuals of many species, but that
plants are associated in communities, which have a definite structure
and often a regular specific composition. Only the species which are
adapted t o a particular habitat are selected from the available propagules of the total flora of a region, and the number of these is further
restricted by considerations of space and by mutual compatibility in
requirements for nutrients and light. The resulting communities vary
in their degree of cohesion, which is least in transitory populations of
ephemerals with single-layer structure and greatest in the tropical
forest with its complex development of interrelated synusiae.
The community is one of the key concepts in the science of vegetation, and data in all branches of the subject should be treated in the
community context. For example the plant geographer cannot consider
the migration of taxa in isolation; for every species must migrate in a
chain of suitable environments and in communities where a niche for it
exists. No more can the autecologist consider the behaviour of a single
species out of the context of the community, for its physiological
optimum is frequently modified by competition and does not correspond
with its ecological optimum.
The exact description and characterization of the community bhus
becomes essential t o all who work in the science of vegetation ; for it is
important that ecological findings should be related to communities
which are well described. Not t o do so is tantamount t o carrying out a
physical experiment without stating the conditions. Many methods have
been developed, for both the description and the characterization of
communities, which vary from subjective and qualitative t o elaborately
quantitative. The choice of method is a personal matter influenced
partly by the purpose of the investigation and partly by the particular
view taken by the worker of the nature of vegetation. But it should be
remembered that, however exact the method, it cannot describe all the
attributes of even the simplest community a t one moment of time.
Every description is an abstraction from the available data.
The argument so far is that, to be fully useful, ecological observations
must be specified in terms of the community in which they are made.
Nevertheless, even if this is done, no inductive generalizations are possible without classification. Without it any observation may be only an
isolated case and consequently of very limited general value. This need
for classification is quite independent of the subject matter ; it, applies
equally to any objects of which we wish t o think, whether it be books,
musical instruments, the shape of clouds or human emotions.
Some phenomena are easier t o classify than others. This depends on
how distinct is the discontinuity between one class and those nearly