, Volume 39, Issue 2, pp 97-114

First online:

Transformation of cover-abundance values in phytosociology and its effects on community similarity

  • E. van der MaabelAffiliated withDivision of Geobotany

Rent the article at a discount

Rent now

* Final gross prices may vary according to local VAT.

Get Access


Various cover and cover-abundance scales are compared, together with some current transformations, including the angular transformation and logarithmic transformations. A new cover-abundance scale, being a fully numerical, extended Braun-Blanquet scale is introduced under the name ordinal scale. The transformation of cover-abundance data is discussed in terms of a general transformation function y=x wand it is shown that by applying different values of w to the ordinal scale all other scales can be approached, ranging from the presence-absence transformation for w=0 to to the Tüxen & Ellenberg cover scale for w=4.

The effect of the transformation of cover-abundance values on community similarity is shown in three examples: (1) 25 relevés of Arrhenatheretum grassland, being the exampletable used by Mueller-Dombois & Ellenberg; (2) 22 relevés of Spartinetea communities; and (3) 23 relevés of Glauco-Puccinellietalia communities, the latter two sets being taken from selections devised by the Working Group for Data-Processing. In all examples classification and ordination results obtained with intermediate transformations were superior to those obtained with either a strong weighting on mere presence, or an emphasis on dominance. This comparison was based on previous phytosociological experience.

Weighting and standardization are compared with each other. It is concluded (1) that standardization by species has effects similar to those of weighting of rare species with low cover-abundance values; (2) that standar-dization by relevé is similar to weighting the dominant species; and (3) that no standardization and intermediate weighting are hence similar in effect.

It is stated that the relations between weighting and standardization are insufficiently studied and need more attention. A general formula for applying both forms of transformation is presented: y=a x w.

It is supposed that the effect of the amount of weighting will depend on the heterogeneity level of the data-set under study. Indeed some indications were found that weighting the dominant species may help distinguishing lower syntaxonomical units, especially facies, and also higher units, even up to the class level if the communities are characterized by one or a few dominant species.

Some remarks are made on the possible ecological background of the effects of weighting. It is argued that biological (reproduction) processes responsible for the performance of a species are of an exponential nature. Two performance types may be distinguished, viz. the abundance type and the dominance type. Both types may be approached with logarithmic scales.


Arrhenatheretum Classification Combined estimation Cover-abundance scale Ordination Phytosociology Salt marsh Similarity Transformation