TY - JOUR
T1 - A step further towards quantitative reconstruction of past vegetation in Fennoscandian boreal forests: Pollen productivity estimates for six dominant taxa
AU - Räsänen, Satu
AU - Suutari, Henna
AU - Nielsen, Anne Birgitte
N1 - Funding Information:
This paper is a contribution to the POLLANDCAL (POLlen-LANDscape CALibration) network sponsored by NorFA (Nordic Council of Advanced Studies) and co-ordinated by M.-J. Gaillard (University of Kalmar, Sweden) (network website: www.geog.ucl.ac.uk/ecrc/pollandcal/ ). We are very thankful to all POLLANDCAL members for useful and inspiring discussions during the numerous network workshops held between 2002 and 2005, and in particular to Shinya Sugita (University of Minnesota, USA) for providing us with the basic research approaches of POLLANDCAL, and for allowing access to his numerous computer programs, and to Jane Bunting and Dick Middleton (University of Hull, U.K.) for developing excellent user-friendly software (the HUMPOL suite of programs) for use in network research activities. Anna Broström deserves our thanks for the co-operation during the beginning of the process of data handling, and Shinya Sugita for giving valuable comments on the data structure. Sheila Hicks has helped in many ways during the study design, collection and handling the data and writing the paper, from which she is warmly thanked. Anne Birgitte Nielsen's contribution to this research was funded by the Carlsberg Foundation, grant no. 04-0303/20. We thank two anonymous reviewers whose comments improved the manuscript.
PY - 2007/9
Y1 - 2007/9
N2 - Pollen productivity estimates (PPE) are needed for producing quantitative reconstructions of the past landscapes by means of modelling. This study aims to provide estimates of relative pollen productivity for the two main tree species (Pinus sylvestris L. and Betula sp.), for Cyperaceae and Gramineae, and the dominant dwarf shrubs (Empetrum, Vaccinium sp.) of northern boreal forest vegetation in Finnish Lapland. A set of 24 moss samples were collected and their pollen content analysed, and detailed vegetation analyses were carried out within a 10.5 m radius around the samples. Vegetation further away, up to 10 km from the sites, was analysed from the Finnish Land Survey data. Vegetation proportions were calculated from the remotely sensed data using the program PolFlow, and distance weighting was applied using the program ERV_v6. The distance weighted vegetation data from the field survey and the Land Survey data were combined, and analysed with the pollen data using the ERV model. The resulting PPEs were evaluated by simulating pollen loadings with the POLLSCAPE model for an independent set of eight moss samples from the same area, using the obtained PPEs. The estimated pollen loadings and the observed pollen loadings were compared using paired samples t-tests. The PPEs obtained for the two major tree species are in keeping with expectations: PPE for Pinus is 8.4 (reference taxon Gramineae = 1) and for Betula 4.6. For Cyperaceae (0.002), Empetrum (0.07), and Vaccinium (0.01) the PPEs are very low and, according to paired samples t-test of the predicted and actual pollen loading, not reliable for Cyperaceae and Gramineae. The radius of the relevant source area of pollen (RSAP) is estimated to be ca. 1000 m. The pollen productivity estimates of the species, especially Pinus and Betula, are different in the northern boreal forest than in previously studied areas, such as southern Sweden. This has great implications for quantitative interpretations of fossil pollen records from the region.
AB - Pollen productivity estimates (PPE) are needed for producing quantitative reconstructions of the past landscapes by means of modelling. This study aims to provide estimates of relative pollen productivity for the two main tree species (Pinus sylvestris L. and Betula sp.), for Cyperaceae and Gramineae, and the dominant dwarf shrubs (Empetrum, Vaccinium sp.) of northern boreal forest vegetation in Finnish Lapland. A set of 24 moss samples were collected and their pollen content analysed, and detailed vegetation analyses were carried out within a 10.5 m radius around the samples. Vegetation further away, up to 10 km from the sites, was analysed from the Finnish Land Survey data. Vegetation proportions were calculated from the remotely sensed data using the program PolFlow, and distance weighting was applied using the program ERV_v6. The distance weighted vegetation data from the field survey and the Land Survey data were combined, and analysed with the pollen data using the ERV model. The resulting PPEs were evaluated by simulating pollen loadings with the POLLSCAPE model for an independent set of eight moss samples from the same area, using the obtained PPEs. The estimated pollen loadings and the observed pollen loadings were compared using paired samples t-tests. The PPEs obtained for the two major tree species are in keeping with expectations: PPE for Pinus is 8.4 (reference taxon Gramineae = 1) and for Betula 4.6. For Cyperaceae (0.002), Empetrum (0.07), and Vaccinium (0.01) the PPEs are very low and, according to paired samples t-test of the predicted and actual pollen loading, not reliable for Cyperaceae and Gramineae. The radius of the relevant source area of pollen (RSAP) is estimated to be ca. 1000 m. The pollen productivity estimates of the species, especially Pinus and Betula, are different in the northern boreal forest than in previously studied areas, such as southern Sweden. This has great implications for quantitative interpretations of fossil pollen records from the region.
KW - boreal forest
KW - ERV model
KW - Fennoscandia
KW - pollen loading
KW - pollen productivity estimates
KW - relevant source area of pollen
UR - http://www.scopus.com/inward/record.url?scp=34548228539&partnerID=8YFLogxK
U2 - 10.1016/j.revpalbo.2007.04.004
DO - 10.1016/j.revpalbo.2007.04.004
M3 - Article
SN - 0034-6667
VL - 146
SP - 208
EP - 220
JO - Review of Palaeobotany and Palynology
JF - Review of Palaeobotany and Palynology
IS - 1-4
ER -
