SUCCESSIVE APPROXIMATION I N DESCRIPTIVE ECOLOGY
56
relative number of stands occurring a t various points on their axis is
taken into account.
But even were their data convincing, there would be no reason t o
assume, therefore, that all vegetation consists of continua, or even that
the upland forest communities of southern Wisconsin, if analysed in a
different way, might not have presented more marked discontinuities
(cf. Daubenmire, 1960).
The studies of Poore (1955a,b,c, 1956) in Breadalbane, Scotland
made use of subjective methods, but special attention was devoted t o
the problems arising from mosaics, heterogeneity of the stand and
ecotones. I n the final tabulation of stands the following categories were
omitted. (i) Communities which were obviously unstable; (ii) small
fragments ; (iii) ecotones; (iv) mosaics.
It was found that the remainder could be classified without difficulty
into abstract units (noda), with the exception of certain communities
which were unique in the region and others which appeared t o fall into
floristic series and not into discrete groups. The final conclusion from
this study was: “It may well be t,hat further research will show that
most, if not all, communities fall into series, and that the discontinuities
found are artifacts. But I have found that it is convenient, even in a
continuous series, t o establish arbitrary noda as points of reference. On
the other hand, it is logically preferable, until discontinuities are
actually established t o assume that all communities can be arranged
in series, rather than the reverse attitude which is held by the BraunBlanquet school, whose followers maintain the distinctness of the
associations as a doctrine only by the arbitrary rejection of descriptive
material.”
It must remain a t present an operi question whether or not the
variation in vegetation is continuous, and it is hard t o see how sufficiently extensive data could be obtained t o establish either alternative
as a general proposition. For such a study t o prove its point it would
not only have t o establish that communities can be found that form
continua along a number of axes of variation, but that all these communities occupy significantly comparable areas. I do not think that it
is either necessary or important to find an absolute answer t o this
question for the result would alter very little the fundamental logical
processes associated with the classification of vegetation, and would
add little t o our knowledge of the functioning of vegetation.
As we have seen this problem of continuity versus discontinuity
arises at two separate and distinct stages: in defining the stand or
concrete unit, and in defining the nodum or abstract unit. Discontinuities can be found between stands of vegetation due, for example,
t o sudden differences of soil, change of dominant species or t o different
56
relative number of stands occurring a t various points on their axis is
taken into account.
But even were their data convincing, there would be no reason t o
assume, therefore, that all vegetation consists of continua, or even that
the upland forest communities of southern Wisconsin, if analysed in a
different way, might not have presented more marked discontinuities
(cf. Daubenmire, 1960).
The studies of Poore (1955a,b,c, 1956) in Breadalbane, Scotland
made use of subjective methods, but special attention was devoted t o
the problems arising from mosaics, heterogeneity of the stand and
ecotones. I n the final tabulation of stands the following categories were
omitted. (i) Communities which were obviously unstable; (ii) small
fragments ; (iii) ecotones; (iv) mosaics.
It was found that the remainder could be classified without difficulty
into abstract units (noda), with the exception of certain communities
which were unique in the region and others which appeared t o fall into
floristic series and not into discrete groups. The final conclusion from
this study was: “It may well be t,hat further research will show that
most, if not all, communities fall into series, and that the discontinuities
found are artifacts. But I have found that it is convenient, even in a
continuous series, t o establish arbitrary noda as points of reference. On
the other hand, it is logically preferable, until discontinuities are
actually established t o assume that all communities can be arranged
in series, rather than the reverse attitude which is held by the BraunBlanquet school, whose followers maintain the distinctness of the
associations as a doctrine only by the arbitrary rejection of descriptive
material.”
It must remain a t present an operi question whether or not the
variation in vegetation is continuous, and it is hard t o see how sufficiently extensive data could be obtained t o establish either alternative
as a general proposition. For such a study t o prove its point it would
not only have t o establish that communities can be found that form
continua along a number of axes of variation, but that all these communities occupy significantly comparable areas. I do not think that it
is either necessary or important to find an absolute answer t o this
question for the result would alter very little the fundamental logical
processes associated with the classification of vegetation, and would
add little t o our knowledge of the functioning of vegetation.
As we have seen this problem of continuity versus discontinuity
arises at two separate and distinct stages: in defining the stand or
concrete unit, and in defining the nodum or abstract unit. Discontinuities can be found between stands of vegetation due, for example,
t o sudden differences of soil, change of dominant species or t o different