Long term landscape dynamics in a Caledonian pine forest
An understanding of long-term landscape dynamics provides a useful foundation for present and future management planning. Most long-term processes operate outside human timescales, so this thesis uses a combination of palaeoecological records and pollen dispersal modelling on an important threatened ecosystem, the Caledonian pine forest, to i) reconstruct feasible past landscapes using novel methods ii) study the long-term dynamics leading to community transitions within the landscape. The overall aim is to increase understanding of the role of processes such as fire and human activity on landscape dynamics in the second half of the Holocene and the development of the present day mosaic. Firstly, the performance of 'the Prentice model of pollen dispersal is assessed against modern vegetation and pollen assemblage data at Abernethy Forest, Invemess-shire. The model performed well for the major landscape forming taxa (Pinus, Calluna, Betula and Poaceae/Cyperaceae), but less well for taxa occurring infrequently in the modem vegetation such as Alnus, Salix and Corylus. Secondly, stratigraphic, pollen, charcoal and tephra data from the analysis of seven peat cores are presented. The data imply that once Pinus sylvestris became dominant in the region c. 7000 BP, little change in forest composition, openness and fire regime occurred until c. 4000BP, when human activity began to affect the region. Cyclical heathland-forest transitions become established in some locations from around 3000 BP. Hypotheses are then advanced linking major processes to vegetation dynamics, and used as the basis for simulation of possible past vegetation mosaics at two key time slices, defined by the Hekla-4 (3830 BP) and Glen Garry (2100 BP) tephra isochrones. Thirdly, the Prentice model of pollen dispersal is applied to these possible mosaics to simulate pollen assemblages at the coring locations, which are statistically compared with the fossil assemblages to identify the most feasible landscape mosaic for each time slice. These landscape mosaics are in themselves valuable contributions to debates on the 'natural' state of the studied area and the effects of different management and environmental regimes. They also permit refinement of the hypotheses connecting process and vegetation dynamics for Abernethy Forest over the long term (10² - 10³ years).Finally, suggestions for how these techniques can be best applied to informing conservation management are made, together with discussion of strengths and limitations of the approach and future directions such work could take.