The Use of Statistics in Phytosociology
J . M . LAMBERT and M . B . DALE
Botany Department. University of Southampton. England
I . Introduction .......................................................... 59
A . The Aim of the Contribution ......................................... 59
B . DefinitionofTerms ................................................. 60
C . The Function of Statistics in Phytosociology ...........................
62
I1 . The Nature of Phytosociological Data ....................................
64
A . The Nature of Vegetation ............................................
64
B . The Nature of the Variables Concerned ................................ 66
C . The Nature of the Sites ............................................. 67
D . The Nature of the Measures .......................................... 68
I11 . Methods of Analysis ................................................... 70
A . The Statistical Properties of Phytosociological Data ..................... 70
B . Ordination and Classification ......................................... 72
C . MethodsofClassification ............................................ 77
D . Inverse Analysis ................................................... 81
E . Nodal Analys is ..................................................... 83
F . Relationship with the Environment ..................................
85
IV . Comparison of Phytosociological Concepts ................................ 87
A . TheBasic Assumptions ............................................. 87
B . TheBctsic Approach ................................................
90
C . The Basic "Vegetation-Unit" ........................................ 93
V . The Future of Statistics in Phytosociology ................................ 95
Acknowledgments ........................................................ 97
References ................................................................
98
I . INTRODUCTION
8 . T H E A I M O F THE CONTRIBUTION
The title of this paper may at first sight seem misleading . We shall
not concern ourselves with an enumeration of the various applications
of mathematics which have progressively invaded the study of vegetation within the last 30 years; nor shall we consider every facet of
ecological work which could conceivably be included under the heading
of phytosociology in its widest sense; and. even within the restricted
field with which we shall be dealing. we shall not attempt a detailed
comparison of all the various techniques and parameters which have
been proposed a t one time or another for the assessment of a given
situation . Certain of these aspects have already been covered excellently
in recent publications by Goodall (1952. 1954a. 1962) and Greig-Smith