A comparison of two sampling schemes in phytosociology - preferential and stratified random

MICHALCOVÁ, Dana. A comparison of two sampling schemes in phytosociology - preferential and stratified random. In European Vegetation Survey, 19th Workshop, Pécs (Hungary), 29 April - 2 May 2010. 2010.

Základní údaje

• Originální název: A comparison of two sampling schemes in phytosociology - preferential and stratified random
• Autoři: MICHALCOVÁ, Dana.
• Vydání: European Vegetation Survey, 19th Workshop, Pécs (Hungary), 29 April - 2 May 2010, 2010.

Další údaje

• Typ výsledku: Konferenční abstrakt
• Utajení: není předmětem státního či obchodního tajemství
• WWW: URL
• Organizační jednotka: Přírodovědecká fakulta

Anotace

Sampling design is an important issue in phytosociological methodology. Preferentially positioned relevés, which prevail in phytosociological databases, can provide biased results due to diff erent author’s experience or judgment. Th erefore, the stratifi ed-random sampling design is considered as a promising alternative. Th e aim of this study is to compare eff ects of preferential and random stratifi ed sampling on the results of vegetation studies. Two parallel analyses of forest vegetation were performed in two diff erent areas of the Czech Republic, comparing random- stratifi ed data sets with a selection of relevés of forest vegetation from the same areas, obtained from the Czech National Phytosociological Database. Th e following properties of these datasets were compared: 1) number and percentage cover of vascular plant species per relevé, 2) number, percentage cover and proportional number of endangered species per relevé, 3) number, percentage cover and proportional number of invasive species per relevé, 4) number, percentage cover and proportional number of neophytes per relevé, 5) sample rarefaction curves for all species, endangered species, invasive species and neophytes 6) beta diversity, calculated as mean Jaccard dissimilarity and 7) patterns in ordination diagrams. Th e analysis showed that in the preferentially sampled relevés the number, percentage cover and proportional number of endangered species per relevé, as well as beta diversity, were higher than in the relevés sampled following the stratifi ed-random design. However, the properties regarding the vascular plant species, invasive species and neophytes provided inconsistent results (i.e. differences between the two datasets were signifi cant only in one study area), partly depending on the purpose of original studies and local idiosyncrasies of the two studied areas.