THE USE O F STATISTICS I N PHYTOSOCIOLOQY
93
The question as to the most appropriate stage at which to introduce
“quantitative methods” depends on what is meant by such a term.
Poore’s later text in fact suggests a lack of distinction between the use
of quantitative methods and the use of quantitative measures. Since we
have already indicated (p. 69) that more information resides in qualitative differences than in quantitative differences in the floristic composition of different sites, we would agree that quantitative memures
should be reserved for detailed work rather than used in primary survey.
However, if the term “quantitative methods” is equated with the use
of statistical techniques, then - provided the method selected is both
appropriate and efficient - we believe that such methods have as much
virtue in primary survey as in subsequent more detailed work. We shall
return to the question of the practicability of using such techniques in
large-scale survey in our final section. Meanwhile, since we are ultimately concerned with the abstraction of information as efficiently
as possible, we must examine phytosociological concepts involved in the
erection of abstract units and try to assess the general situation
regarding schemes of vegetational classification.
C. THE BASIC “VEQETATION-UNIT”
The value of classification as an investigational tool is that it brings
together common information to enable abstractions to be made
efficiently. If vegetation is regarded as a tripartite system of species/
site/habitat, but with no assumption that species and habitat factors
are coincidentally related, then a common abstraction must have due
regard to information from each side. Much of the argument which
constantly arises as to whether environmental features should be used
in vegetational classification seems in fact to stem from the fluctuating
concept of vegetation we have already mentioned (p. 88) ; it is frequently
not realized that different workers are quite legitimately using different
concepts. This point is recognized by Poore (1962, p. 61), who distingGshes “vegetation” from “vegetation-habitat complex”, and
suggests that no features of the habitat should be used in the definition
of vegetation-units. However, although we have suggested that in
general it is most efficient to deal initially with plant/site relationships
alone, we believe it is more informative to include also site/habitat
relationships in the ultimate definition of vegetation-units. Such doublydefined units we have elsewhere (p. 83) called “noda”, and this concept
now needs to be matched against pre-existing ideas.
The original concept of a nodum as an abstract unit in phytosociological work can be attributed to Poore himself, whose “noda”, however, were subjectively extracted and defined solely by reference to
floristic composition. The noda we are here concerned with possess a
D2
C.E.R.
93
The question as to the most appropriate stage at which to introduce
“quantitative methods” depends on what is meant by such a term.
Poore’s later text in fact suggests a lack of distinction between the use
of quantitative methods and the use of quantitative measures. Since we
have already indicated (p. 69) that more information resides in qualitative differences than in quantitative differences in the floristic composition of different sites, we would agree that quantitative memures
should be reserved for detailed work rather than used in primary survey.
However, if the term “quantitative methods” is equated with the use
of statistical techniques, then - provided the method selected is both
appropriate and efficient - we believe that such methods have as much
virtue in primary survey as in subsequent more detailed work. We shall
return to the question of the practicability of using such techniques in
large-scale survey in our final section. Meanwhile, since we are ultimately concerned with the abstraction of information as efficiently
as possible, we must examine phytosociological concepts involved in the
erection of abstract units and try to assess the general situation
regarding schemes of vegetational classification.
C. THE BASIC “VEQETATION-UNIT”
The value of classification as an investigational tool is that it brings
together common information to enable abstractions to be made
efficiently. If vegetation is regarded as a tripartite system of species/
site/habitat, but with no assumption that species and habitat factors
are coincidentally related, then a common abstraction must have due
regard to information from each side. Much of the argument which
constantly arises as to whether environmental features should be used
in vegetational classification seems in fact to stem from the fluctuating
concept of vegetation we have already mentioned (p. 88) ; it is frequently
not realized that different workers are quite legitimately using different
concepts. This point is recognized by Poore (1962, p. 61), who distingGshes “vegetation” from “vegetation-habitat complex”, and
suggests that no features of the habitat should be used in the definition
of vegetation-units. However, although we have suggested that in
general it is most efficient to deal initially with plant/site relationships
alone, we believe it is more informative to include also site/habitat
relationships in the ultimate definition of vegetation-units. Such doublydefined units we have elsewhere (p. 83) called “noda”, and this concept
now needs to be matched against pre-existing ideas.
The original concept of a nodum as an abstract unit in phytosociological work can be attributed to Poore himself, whose “noda”, however, were subjectively extracted and defined solely by reference to
floristic composition. The noda we are here concerned with possess a
D2
C.E.R.