The Use of Subjective Estimates of Cover-Abundance as the Basis for Ordination

Abstract

The main disadvantage of ordination techniques (e.g. Curtis & McIntosh 1951; Bray & Curtis 1957)* as at present practised would seem to lie in the amount of time needed to collect sufficient quantitative field data. Even simple measures such as frequency may take a considerable amount of time, especially in species-rich areas. However, once the data are assembled, it is possible to develop quite a complex ordination without the aid of an electronic computer. This is in contrast to classificatory techniques, such as the associationanalysis of Williams & Lambert (1959, 1960), which require the simplest of field data but involve a considerable amount of computation, so that only extremely species-poor communities are amenable to hand-computation. It would, therefore, seem that a technique combining the ease of collection of field data found in association-analysis and the lighter computational load of ordination techniques might have wider use in ecological teaching and research. When subjective estimates of cover-abundance are used in the field, the time needed to collect data is often little more than that needed to make records of present and absence, and the saving on the time needed to make full quantitative records (e.g. frequency and cover) is of the order of 80%. A common scale of cover-abundance, frequently used in phytosociological studies in Britain, is the Domin scale. Many classifications deriving from data using Domin values are based largely on species' presence and absence rather than on the information contained in the Domin values. Thus if the Domin scale could be used as a measure of coverabundance for the purposes of ordination, it would both allow new phytosociological data to be compared with previous work and in addition make already assembled data (e.g. McVean & Ratcliffe 1962) amenable to ordination analyses.