Taxonomic, morphological and structural surrogates for the rapid assessment of vegetation.

摘要

Question: To what extent do surrogate approaches to vegetation description provide equivalent data and estimates of the patterns of variability in species composition to that which would be obtained using species-level abundance data? Locations: Picos de Europa in Northern Spain and South Uist, Outer Hebrides, Scotland, UK. Methods: Percentage cover data were collected from 117 plots in montane and sub-alpine grassland and woodland mosaics in the Picos de Europa and 217 quadrats across a gradient from temperate coastal dunes to upland grassland and heath on South Uist. From these data, 14 surrogate datasets were derived using supra-species (genus, family) levels of description, presence-absence, life form, growth form and plant functional types (PFTs; derived from seven easily measurable traits). The descriptive power of each of the 14 surrogate datasets in each area, relative to the original cover abundance dataset, was assessed by 'second stage' analysis, based on Spearman's rank correlation coefficient calculated between pairs of similarity matrices. Results: In both areas, presence-absence data showed the closest similarities to the original cover abundance data, followed by data generalized at supra-species level. PFT, life form, growth form and leaf trait data were less similar. Conclusions: Many situations exist where it may be more efficient or desirable to collect neither species nor abundance data and to use appropriate surrogates. The results demonstrate clearly that the information content of surrogate datasets in our study may be an acceptable substitute. These ideas merit further investigation in a wider range of habitats and contexts..