Advanced

Assessing spider diversity on the forest floor: expert knowledge beats systematic design

Sereda, Elvira; Blick, Theo; Dorow, Wolfgang H. O.; Wolters, Volkmar and Birkhofer, Klaus LU (2014) In The Journal of Arachnology 42(1). p.44-51
Abstract
The design of sampling schemes affects the results of biodiversity inventories. As an approach for quantifying the implications of such effects, we compared data on spider communities sampled in a beech-dominated forest floor habitat by 1) a regular grid of pitfall traps (systematic design) and 2) an expert-based distribution of traps (stratified design). We tested whether the two designs would lead to similar conclusions about the diversity and composition of ground-dwelling spider communities. Estimates of species richness, rarefied species richness and activity density calculated per trap were significantly higher in the stratified than in the systematic design. The community composition based on the presence or absence of sampled... (More)
The design of sampling schemes affects the results of biodiversity inventories. As an approach for quantifying the implications of such effects, we compared data on spider communities sampled in a beech-dominated forest floor habitat by 1) a regular grid of pitfall traps (systematic design) and 2) an expert-based distribution of traps (stratified design). We tested whether the two designs would lead to similar conclusions about the diversity and composition of ground-dwelling spider communities. Estimates of species richness, rarefied species richness and activity density calculated per trap were significantly higher in the stratified than in the systematic design. The community composition based on the presence or absence of sampled species or based on log-transformed activity densities differed significantly. Most of the dissimilarity between the community estimates of the two designs was attributable to three species, with Pardosa saltans Topfer-Hofmann 2000 being more common in traps of the stratified design and Tenuiphantes zimmermanni (Bertkau 1890) and Walckenaeria cuspidata Blackwall 1833 being more frequently observed in traps of the systematic design. Our study suggests that a stratified sampling design is better suited for inventory surveys of spider communities of forest-floor habitats, as trap locations of this design reflect specific habitat needs. It is important to note that inventories are a major field for the application of such designs and that greater care is needed for the application of inferential statistics. For example, the non-randomness that is caused by expert selection of sampling sites may violate fundamental assumptions of simple linear models. (Less)
Please use this url to cite or link to this publication:
author
organization
publishing date
type
Contribution to journal
publication status
published
subject
keywords
Araneae, biodiversity, inventory, expert-based sampling, regular, sampling, sampling design
in
The Journal of Arachnology
volume
42
issue
1
pages
44 - 51
publisher
American Arachnological Society
external identifiers
  • wos:000335125900005
  • scopus:84898448912
ISSN
1937-2396
DOI
10.1636/P13-16.1
language
English
LU publication?
yes
id
8663399e-03fa-406a-bf72-26517e560f71 (old id 4487730)
date added to LUP
2014-06-23 14:55:38
date last changed
2017-08-27 03:43:44
@article{8663399e-03fa-406a-bf72-26517e560f71,
  abstract     = {The design of sampling schemes affects the results of biodiversity inventories. As an approach for quantifying the implications of such effects, we compared data on spider communities sampled in a beech-dominated forest floor habitat by 1) a regular grid of pitfall traps (systematic design) and 2) an expert-based distribution of traps (stratified design). We tested whether the two designs would lead to similar conclusions about the diversity and composition of ground-dwelling spider communities. Estimates of species richness, rarefied species richness and activity density calculated per trap were significantly higher in the stratified than in the systematic design. The community composition based on the presence or absence of sampled species or based on log-transformed activity densities differed significantly. Most of the dissimilarity between the community estimates of the two designs was attributable to three species, with Pardosa saltans Topfer-Hofmann 2000 being more common in traps of the stratified design and Tenuiphantes zimmermanni (Bertkau 1890) and Walckenaeria cuspidata Blackwall 1833 being more frequently observed in traps of the systematic design. Our study suggests that a stratified sampling design is better suited for inventory surveys of spider communities of forest-floor habitats, as trap locations of this design reflect specific habitat needs. It is important to note that inventories are a major field for the application of such designs and that greater care is needed for the application of inferential statistics. For example, the non-randomness that is caused by expert selection of sampling sites may violate fundamental assumptions of simple linear models.},
  author       = {Sereda, Elvira and Blick, Theo and Dorow, Wolfgang H. O. and Wolters, Volkmar and Birkhofer, Klaus},
  issn         = {1937-2396},
  keyword      = {Araneae,biodiversity,inventory,expert-based sampling,regular,sampling,sampling design},
  language     = {eng},
  number       = {1},
  pages        = {44--51},
  publisher    = {American Arachnological Society},
  series       = {The Journal of Arachnology},
  title        = {Assessing spider diversity on the forest floor: expert knowledge beats systematic design},
  url          = {http://dx.doi.org/10.1636/P13-16.1},
  volume       = {42},
  year         = {2014},
}