Recent studies (e.g. Nielsen, 2004; Nielsen & Odgaard, 2005; Hellman 2005; Caseldine & Fyfe, 2006) have shown that it is possible to achieve quantitative reconstructions of the abundance of different plants in the past using models of pollen representation, dispersal and deposition (Prentice, 1985; Sugita, 1993, 1994).

The so called LRA (Landscape Reconstruction Algorithm) (Sugita, 2007a, b), can be used on a network of pollen diagrams from larger lakes and smaller sites, for reconstructing the extent of woodland, heathland, arable land etc. in earlier periods. This method builds on the Extended R-Value (ERV) model of the relationship between pollen sedimentation in lakes and bogs, and the plant abundance in the surrounding landscape (Parsons & Prentice, 1981; Prentice & Parsons, 1983; Sugita, 1994), which assumes a linear relationship between pollen loading and plant abundance within the pollen source area, provided the plant abundance is distance weighted according to the dispersal properties of the individual pollen types.
The most important parameters for reconstruction are the pollen productivity of different plants and the input of regional background pollen from areas outside the relevant pollen source area (Sugita, 1994). These are estimated from calibration datasets of pollen assemblages and plant abundance. Estimates of pollen productivity are available, or becoming available for many parts of Europe (Broström et al., 2008), but more work on the spatial variability in pollen productivity, and for example their relation to climate, is still needed.

Quantitative reconstructions of past vegetation and landscapes are urgently needed for the purposes of nature and biodiversity conservation, landscape management research and for climate research. For conservation and landscape management, quantitative reconstructions of past landscapes are needed to understand long-term ecological processes, and the impact of historical management, which are the basis of the current biodiversity. As much of the biodiversity of Europe today is linked to traditional cultural landscapes, historical models in modern landscape management can be very useful. Nature managers are also interested in the quantitative composition of the landscape before the introduction, as this gives new insights into the “present natural” landscape, for example on the role of grazing animals in the time before the introduction of agriculture.

For climate research, the interest are in comparing past tree species composition reconstructed from pollen data to predictions based on climate-vegetation models. Furthermore, past land cover has feedback mechanisms to the climate system through impacts on albedo and carbon cycling.


Cited literature

• Broström, A., Nielsen, A.B., Gaillard, M.-J., Hjelle, K., Mazier, F., Binney, H., Bunting, J., Fyfe, R., Meltsov, V., Poska, A., Räsänen, S., Soepboer, W., von Stedingk, H., Suutari, H and Sugita, S. (2008): Pollen productivity estimates of key European plant taxa for quantitative reconstruction of past vegetation: a review. - Vegetation History and Archaeobotany, 17: 461-478
• Caseldine, C. and Fyfe, R., 2006. A modelling approach to locating and characterising elm decline/landnam landscapes. - Quaternary Science Reviews, 25: 632-644
• Hellman, S.E.V., 2005. Quantitative Reconstruction of Past Cultural Landscapes in Southern Sweden, 3000-0 BP: Validation of a modelling approach to estimate regional plant abundance in Skåne and Småland. - The ESS Bulletin, 3(1)
• Nielsen, A.B. and Odgaard, B.V. (2005): Reconstructing land-cover from pollen assemblages from small lakes in Denmark. - Review of Palaeobotany and Palynology 133, 1-21