Advanced

Testing the effect of site selection and parameter setting on REVEALS-model estimates of plant abundance using the Czech Quaternary Palynological Database

Mazier, F.; Gaillard, M. -J.; Kunes, P.; Sugita, S.; Trondman, A. -K. and Broström, Anna LU (2012) In Review of Palaeobotany and Palynology 187. p.38-49
Abstract
REVEALS-based quantitative reconstruction of Holocene vegetation cover (expressed in plant functional types. PFTs) is used in the LANDCLIM project to assess the effect of human-induced land-cover change on past climate in NW Europe. Using the Czech Quaternary Pollen Database, this case study evaluates the extent to which selection of data and input parameters for the REVEALS model applications would affect reconstruction outcomes. The REVEALS estimates of PFTs (grid-cell based REVEALS PET estimates, GB REVEALS PFT-s) are calculated for five time windows of the Holocene using fossil pollen records available in each 1 degrees x1 degrees grid cell of the Czech Republic. The input data and parameters selected for testing are: basin type and... (More)
REVEALS-based quantitative reconstruction of Holocene vegetation cover (expressed in plant functional types. PFTs) is used in the LANDCLIM project to assess the effect of human-induced land-cover change on past climate in NW Europe. Using the Czech Quaternary Pollen Database, this case study evaluates the extent to which selection of data and input parameters for the REVEALS model applications would affect reconstruction outcomes. The REVEALS estimates of PFTs (grid-cell based REVEALS PET estimates, GB REVEALS PFT-s) are calculated for five time windows of the Holocene using fossil pollen records available in each 1 degrees x1 degrees grid cell of the Czech Republic. The input data and parameters selected for testing are: basin type and size, number of C-14 dates used to establish the chronology of the pollen records, number of taxa, and pollen productivity estimates (PPE). We used the Spearman correlation coefficient to test the hypothesis that there is no association between GB REVEALS PET-s using different data and parameter inputs. The results show that differences in the basin size and type, number of dates, number and type of taxa (entomophilous included or not), and PPE dataset do not affect the rank orders of the GB REVEALS PET-s significantly, except for the cases when entomophilous taxa are included. It implies that, given careful selection of data and parameter and interpretation of results, REVEALS applications can use pollen records from lakes and bogs of different sizes together for reconstruction of past land cover at the regional to sub-continental spatial scales for purposes such as the study of past land cover-climate interactions. Our study also provides useful criteria to set up protocols for data compilation REVEALS applications of this kind. (C) 2012 Elsevier B.V. All rights reserved. (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
quantitative reconstruction of past land-cover pollen data, Holocene, NW, Europe, REVEALS model
in
Review of Palaeobotany and Palynology
volume
187
pages
38 - 49
publisher
Elsevier
external identifiers
  • wos:000312520700004
  • scopus:84867121240
ISSN
0034-6667
DOI
10.1016/j.revpalbo.2012.07.017
project
MERGE
language
English
LU publication?
yes
id
8f947cd3-00b4-4cfb-a687-7e293766749b (old id 3366336)
date added to LUP
2013-01-23 14:37:58
date last changed
2017-11-12 03:35:40
@article{8f947cd3-00b4-4cfb-a687-7e293766749b,
  abstract     = {REVEALS-based quantitative reconstruction of Holocene vegetation cover (expressed in plant functional types. PFTs) is used in the LANDCLIM project to assess the effect of human-induced land-cover change on past climate in NW Europe. Using the Czech Quaternary Pollen Database, this case study evaluates the extent to which selection of data and input parameters for the REVEALS model applications would affect reconstruction outcomes. The REVEALS estimates of PFTs (grid-cell based REVEALS PET estimates, GB REVEALS PFT-s) are calculated for five time windows of the Holocene using fossil pollen records available in each 1 degrees x1 degrees grid cell of the Czech Republic. The input data and parameters selected for testing are: basin type and size, number of C-14 dates used to establish the chronology of the pollen records, number of taxa, and pollen productivity estimates (PPE). We used the Spearman correlation coefficient to test the hypothesis that there is no association between GB REVEALS PET-s using different data and parameter inputs. The results show that differences in the basin size and type, number of dates, number and type of taxa (entomophilous included or not), and PPE dataset do not affect the rank orders of the GB REVEALS PET-s significantly, except for the cases when entomophilous taxa are included. It implies that, given careful selection of data and parameter and interpretation of results, REVEALS applications can use pollen records from lakes and bogs of different sizes together for reconstruction of past land cover at the regional to sub-continental spatial scales for purposes such as the study of past land cover-climate interactions. Our study also provides useful criteria to set up protocols for data compilation REVEALS applications of this kind. (C) 2012 Elsevier B.V. All rights reserved.},
  author       = {Mazier, F. and Gaillard, M. -J. and Kunes, P. and Sugita, S. and Trondman, A. -K. and Broström, Anna},
  issn         = {0034-6667},
  keyword      = {quantitative reconstruction of past land-cover pollen data,Holocene,NW,Europe,REVEALS model},
  language     = {eng},
  pages        = {38--49},
  publisher    = {Elsevier},
  series       = {Review of Palaeobotany and Palynology},
  title        = {Testing the effect of site selection and parameter setting on REVEALS-model estimates of plant abundance using the Czech Quaternary Palynological Database},
  url          = {http://dx.doi.org/10.1016/j.revpalbo.2012.07.017},
  volume       = {187},
  year         = {2012},
}