54
M. E. D. POORE
of field descriptions depend on the initial selection of stand. By including or omitting certain communities it is perfectly possible t o
select data which can be grouped into discrete associations or which
fall into continuous series. The attitude of a particular worker t o the
description of phytocoenoses containing a number of different synusiae,
or his conception of what constitutes a mosaic or an ecotone, or the
rigidity of his definition of uniformity are all factors which easily transform the kind of result which he finally obtains.
The data of Bray and Curtis (1957) from the upland forests of Wisconsin provide a useful demonstration of the way in which vegetation
can be treated as a continuum, and the value in certain circumstances
of analysing it as such; but it in no way proves that this vegetation is a
continuum. A more detailed discussion of these results niay illustrate
points in this argument.
The apparently continuous variation of vegetation is probably enhanced by the method of sampling used.
(i) The areas which were used as samples were all more than fifteen
acres in extent and, where possible, more than forty acres. It is extremely unlikely that such large surfaces would meet the exacting
demands for site uniformity made by Dahl in his work. If even a small
range of variation were included in each sample, these might overlap
and the series might well show continuous variation.
(ii) Fourteen of the twenty-six species chosen for the ordination
were herbs and shrubs whose occurrence is likely to be governed by
site factors of a different amplitude to those of the tree synusia. This
problem has been discussed above.
The first and principal axis of variation extracted from the data (the
x axis) shows “mainly the relationship of the community to major past
disturbance factors and t o its own developmental recovery from those
factors”. It is probable that species composition and importance does
change continuously during a succession ; though there is still doubt
whether some successional stages may not be passed through more
rapidly than others, thus introducing discontinuities into the sere. By
contrast the situation which Dahl investigated is one in which the
vegetation shows distinct zonation and in which no succession of importance can take place between the different zones.
If one considers the arrangement of stands along this x axis, some
stands certainly appear all along the axis; but the majority of stands
are conspicuously grouped. This hardly represents a continuous distribution.
Thus in spite of the method of sampling, which would be likely t o
iron out discontinuities, the data of Bray and Curtis do not provide very
convincing evidence for the continuous nature of vegetation if the
M. E. D. POORE
of field descriptions depend on the initial selection of stand. By including or omitting certain communities it is perfectly possible t o
select data which can be grouped into discrete associations or which
fall into continuous series. The attitude of a particular worker t o the
description of phytocoenoses containing a number of different synusiae,
or his conception of what constitutes a mosaic or an ecotone, or the
rigidity of his definition of uniformity are all factors which easily transform the kind of result which he finally obtains.
The data of Bray and Curtis (1957) from the upland forests of Wisconsin provide a useful demonstration of the way in which vegetation
can be treated as a continuum, and the value in certain circumstances
of analysing it as such; but it in no way proves that this vegetation is a
continuum. A more detailed discussion of these results niay illustrate
points in this argument.
The apparently continuous variation of vegetation is probably enhanced by the method of sampling used.
(i) The areas which were used as samples were all more than fifteen
acres in extent and, where possible, more than forty acres. It is extremely unlikely that such large surfaces would meet the exacting
demands for site uniformity made by Dahl in his work. If even a small
range of variation were included in each sample, these might overlap
and the series might well show continuous variation.
(ii) Fourteen of the twenty-six species chosen for the ordination
were herbs and shrubs whose occurrence is likely to be governed by
site factors of a different amplitude to those of the tree synusia. This
problem has been discussed above.
The first and principal axis of variation extracted from the data (the
x axis) shows “mainly the relationship of the community to major past
disturbance factors and t o its own developmental recovery from those
factors”. It is probable that species composition and importance does
change continuously during a succession ; though there is still doubt
whether some successional stages may not be passed through more
rapidly than others, thus introducing discontinuities into the sere. By
contrast the situation which Dahl investigated is one in which the
vegetation shows distinct zonation and in which no succession of importance can take place between the different zones.
If one considers the arrangement of stands along this x axis, some
stands certainly appear all along the axis; but the majority of stands
are conspicuously grouped. This hardly represents a continuous distribution.
Thus in spite of the method of sampling, which would be likely t o
iron out discontinuities, the data of Bray and Curtis do not provide very
convincing evidence for the continuous nature of vegetation if the