S U C C E S S I V E A P P R O X I M A T I O N I N DESCRIPTIVE ECOLOGY
57
It is extremely difficult, even in limited studies, to grasp directions
of variation without reference t o particular fixed points. Moreover, in
extensive treatments of vegetation the data become incomprehensible
unless they can be arranged in classes. These two propositions may be
illustrated by the works respectively of Bray and Curtis (1957) and
Curtis (1959). In the former the axes of variation are constructed
between reference points. In the construction of the x axis, for example,
these consisted of two sets of three stands each. These two sets were
widely separated from each other, but the index values of the three
stands within each set were highly significantly correlated. These sets
correspond exactly t o the concept of “noda” and have been used by
these authors as points of reference. In the latter work, which is an
extensive treatment of the veget’ation of Wisconsin, arbitrary classes
are erected, “Southern Forests”, “Northern Forests”, “Prairie”,
“Sand barrens and bracken grassland” etc., and these are sub-divided
into, for example, “Southern Forests mesic”, “Southern Forests
xeric” and “Southern Forests lowland”.
This erection of points of reference or of classes cannot be avoided
if the data are to be presented in a comprehensible and useful form,
irrespective of whether the classes are arbitrary or natural. The nature
of variation within and between these classes may then be analysed as
the individual worker wishes, by ordination, by factor analysis or by
successive approximation”.
< <
V. SYSTEMS O F CLASSIFICATION A N D CRITERIA
FOR DEFINING NODA
A number of different systems of classification have been proposed
which variously illuminate the relationships between vegetation and
habitat, and thus fulfil the requirements which were earlier suggested
for a natural classification. These fall into two principal groups, the
physiognomic-ecological and the floristic ; but the distinctions between
them are frequently blurred and they often lead to the same major units.
There is not only the question of the principles of classification t o
be considered, but also the status of the units t o be classified. It was
argued above that synusiae and phytocoenoses are different in kind and
cannot be classified together. In some instances the examination and
classification of synusiae can have great value; indeed it is the only
practicable way of studying dependent societies and results of importance have been obtained by itt as, for example in the study of epiphytic
bryophyte communities in Japan (Hosokawa, 1956; Omura et al.,
1955; Omura and Hosokawa, 1959). These cannot be classified on the
same level as the forests in which they occur; neither can all the small
changes which occur in the make-up of the dependent societies be taken
57
It is extremely difficult, even in limited studies, to grasp directions
of variation without reference t o particular fixed points. Moreover, in
extensive treatments of vegetation the data become incomprehensible
unless they can be arranged in classes. These two propositions may be
illustrated by the works respectively of Bray and Curtis (1957) and
Curtis (1959). In the former the axes of variation are constructed
between reference points. In the construction of the x axis, for example,
these consisted of two sets of three stands each. These two sets were
widely separated from each other, but the index values of the three
stands within each set were highly significantly correlated. These sets
correspond exactly t o the concept of “noda” and have been used by
these authors as points of reference. In the latter work, which is an
extensive treatment of the veget’ation of Wisconsin, arbitrary classes
are erected, “Southern Forests”, “Northern Forests”, “Prairie”,
“Sand barrens and bracken grassland” etc., and these are sub-divided
into, for example, “Southern Forests mesic”, “Southern Forests
xeric” and “Southern Forests lowland”.
This erection of points of reference or of classes cannot be avoided
if the data are to be presented in a comprehensible and useful form,
irrespective of whether the classes are arbitrary or natural. The nature
of variation within and between these classes may then be analysed as
the individual worker wishes, by ordination, by factor analysis or by
successive approximation”.
< <
V. SYSTEMS O F CLASSIFICATION A N D CRITERIA
FOR DEFINING NODA
A number of different systems of classification have been proposed
which variously illuminate the relationships between vegetation and
habitat, and thus fulfil the requirements which were earlier suggested
for a natural classification. These fall into two principal groups, the
physiognomic-ecological and the floristic ; but the distinctions between
them are frequently blurred and they often lead to the same major units.
There is not only the question of the principles of classification t o
be considered, but also the status of the units t o be classified. It was
argued above that synusiae and phytocoenoses are different in kind and
cannot be classified together. In some instances the examination and
classification of synusiae can have great value; indeed it is the only
practicable way of studying dependent societies and results of importance have been obtained by itt as, for example in the study of epiphytic
bryophyte communities in Japan (Hosokawa, 1956; Omura et al.,
1955; Omura and Hosokawa, 1959). These cannot be classified on the
same level as the forests in which they occur; neither can all the small
changes which occur in the make-up of the dependent societies be taken