Representation and biases: Pollen–vegetation relationships and their contribution to the study of fossil pollen records in SW Iberia

In order to provide a better framework for the interpretation of fossil pollen records, a set of 49 modern pollen samples collected across different habitats of SW Iberia, categorised by context and sample type (soil vs moss), were selected to explore pollen–vegetation relationships by multivariate analysis and identify possible biases in pollen representativeness. The distribution of both samples and pollen taxa in the plots is successfully explained according to specific environmental variables in some clusters (A, B and E) in which plant communities depend on certain environmental conditions. On the other hand, other clusters (C, D and F) are better explained by opposition to some variables and suggest a wider range of adaptability of their floristic communities. The classification of samples by context consistently explains their differences in vegetal and environmental terms. However, some discrepancies seem to be better explained by sample type (soil vs moss), which raises some questions on how taphonomic factors may cause biases in both modern and fossil pollen samples.