Premium
Scale‐dependent variation in visual estimates of grassland plant cover
Author(s) -
Klimeš Leoš
Publication year - 2003
Publication title -
journal of vegetation science
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.1
H-Index - 115
eISSN - 1654-1103
pISSN - 1100-9233
DOI - 10.1111/j.1654-1103.2003.tb02214.x
Subject(s) - grassland , plant cover , coefficient of variation , vegetation (pathology) , cover (algebra) , variation (astronomy) , scale (ratio) , ecology , physical geography , vegetation cover , agronomy , biology , environmental science , mathematics , geography , statistics , canopy , cartography , grazing , medicine , mechanical engineering , physics , pathology , astrophysics , engineering
. Plant cover was visually estimated by five observers, independent of each other, in a species‐rich grassland in the Bílé Karpaty Mts., southeastern Czech Republic, in seven plots ranging from 0.001 to 4 m 2 . Variation of total plant cover among the observers was high at small scales: 0.001–0.016 m 2 ; coefficient of variation, CV = 35 to 45%, but much lower at larger scales: 0.06–4 m 2 ; CV = 7 to 15%. Differences between visual estimates of plant cover of individual species made by different observers were affected by plot size, total cover and morphology of particular plants. CV of the cover of individual species ranged from 0 to 225% and decreased with increasing plot size. For abundant plants the CV attained ca. 50%, independent of plot size. In spite of a very high number of sterile plants with similar leaf morphology and colour, the observed variation in cover estimates in the studied grassland was comparable with results reported from other vegetation types. Differences between estimates by individual observers were often larger than usual year to year changes in undisturbed grasslands. Therefore, I suggest that to avoid difficulties in the interpretation of results based on plant cover data obtained from visual estimates, several observers should always work together, adjusting their extreme estimates.