Equifinality and uncertainty in the interpretation of pollen data: the Multiple Scenario Approach to reconstruction of past vegetation mosaics

The long-term goal of pollen analysis has always been the reconstruction of past vegetation mosaics. However, the pollen signal is spatially integrated, and ecologically distinct vegetation structures can produce identical pollen signals. This paper proposes an iterative approach to reconstruction, the Multiple Scenario Approach, which involves creating a large number of hypothetical landscapes using a combination of deterministic and probabilistic 'rules' for plant placement coupled with environmental parameters such as topography, simulating the pollen assemblage produced by each, and comparing them statistically with a sedimentary pollen signal to identify possible reconstructions. The method expressly aims to find multiple possible reconstructions, and also gives some insight into the spatial aspects of the vegetation being reconstructed in addition to its composition.