The spread of spruce in Scandinavia according to Victoria Hofbauer
Quantitative reconstructions of past vegetation change give unique insight into the drivers of vegetation change and allow the testing of models that can be used to predict future scenarios. The project main aims are (1) to improve and apply innovative methods of quantitative vegetation reconstruction and (2) to analyse and synthesise pollen data across Europe with a special focus on central Europe, in order to find new insight into drivers of past vegetation change. The project will also produce new pollen productivity estimates, which are a prerequisite for quantitative reconstructions. Quantitative reconstructions of vegetation cover shall be carried out across Europe and in detail on selected sites. These reconstructions will serve as comparisons for the results of the dynamic vegetation model LPJ-GUESS run for the past using climate scenarios from atmospheric general circulation models. Pollen data will be compiled into maps of past abundance and analysed for population dynamics of selected wide spread species. Results of this project will improve our ability to forecast changes in species distributions and biodiversity which are important for forest management and nature conservation.
The Holocene spread of trees
The Late Glacial and early Holocene race of temperate and boreal trees for new habitats in treeless areas reached incredibly high apparent spreading rates. Results of recent investigations suggest that many species occurred during the coldest stages of the last Ice age much further north than previously anticipated in so called “cryptic refugia”. This leads to new concepts for the mechanisms and dynamics of the early Holocene and late Glacial spread of species.
4 Models of tree spread; broken lines mark dispersal:
A) Continuous moving front: Dispersal takes place from a mature dense population to just beyond the distribution limit.
B) Advance as a discontinuous population: Discrete long distance dispersal events lead to founder populations, which in turn expand and finally merge with each other.
C) Spread at low population density: Frequent long distance dispersal leads to small scattered founder populations, which are new sources for long distance dispersal events. Populations subsequently expand in a wave like manner as most long distance events occur from the dense population.
D) Merging of disjoint populations: Small scattered refuge populations expand and are a source for dispersal events as well as large distant populations. Both populations merge eventually.
A) and B) after (Davis, M. B. 1987. Invasion of forest communities during the Holocene: beech and hemlock in the Great Lakes region. in A. J. Gray, M. J. Crawley, and P. J. Edwards, editors. Colonization, succession, and stability. Blackwell Scientific, Oxford)
Drivers for vegetation change
The question on why plant cover changed through time on local, regional and continental scales is as old as the subject of palaeoecology. Already early pollen analytical results showed that parallel patterns of change occurred over large regions (e.g. von Post, 1918; Rudolf, 1930) and many mechanisms were suggested to explain regional Holocene vegetation history: i) variable Holocene climate; ii) plant succession; iii) soil maturation and leaching; iv) migrational lag; v) human disturbance; and vi) population dynamics and species interaction. The hypothesis that plant distributions during the Holocene were in equilibrium with a changing climate (Webb, 1986) is most important. It is the base for climate reconstructions from pollen diagrams, evaluations of atmospheric general circulation models (AGCM) runs for the past (e.g. Bonfils, 2004) as well as for consequences of climate change scenarios (e.g. Sykes et al. 1996).
Quantitative reconstructions of past vegetation cover
While models readily compute vegetation cover in quantitative terms, reconstructions from pollen diagrams have so far mainly been descriptive or at the most semi-quantitative. In order to compare vegetation models with vegetation reconstructions it is important to express the latter in quantitative terms. Full quantitative reconstructions of past vegetation change will facilitate data/model comparisons, enable a better understanding of the drivers of past vegetation change and provide an important aid for archaeological investigations such as visualisations of likely environmental settings.
Over the last four decades palynologists aimed at relating pollen percentages and pollen accumulation rates to vegetation cover and biomass. By describing pollen dispersal with a numerical model Prentice (1985) laid the basis for the modern understanding of pollen spectra. The model was extended by Sugita (1994) and together with results from the POLLANDCAL – NorFA network (co-ordinated by M.-J. Gaillard www.geog.ucl.ac.uk/ecrc/pollandcal/ ) the REVEALS model (Sugita, 2007) was developed, which now enables full quantitative reconstruction of vegetation cover (Anderson et al., 2006).
The amount of pollen from a given species that is deposited on a unit surface per unit time should be proportional to the biomass of the parent plant. Early investigations that used results from sites with varying characteristics indicated this is true in general terms (Davis et al., 1973). Recently Giesecke and Fontana (2008) showed that pollen accumulation rate (PAR) values from different lakes may be compared in detail if the site characteristics are kept constant. Thus if PAR can be scaled to forest biomass than we would have a direct measure of changes in biomass over time.
