Peat humification analysis has been used widely over the last three decades to reconstruct bog surface wetness (BSW) for use as a palaeoclimate proxy. The technique has the advantage that it is quick and relatively inexpensive to perform, allowing for high resolution and contiguous sampling of peat archives. However, some concerns have been raised over the quality of the resultant proxy-climate records because changes in the plant species composition of peat may contribute a 'species signal' to records, potentially confusing the relationship between bog water table position and the apparent degree of peat humification. This paper uses the k-values of fresh plant material (sensu Overbeck, 1947 - i.e. the absorption value of the alkali extracts of fresh plant material) to explore the impact of changing plant colouration in a Holocene peat humification-based palaeoclimate archive from Newfoundland. We calculate k-scores for peat samples, using plant macrofossil data and the k-values of individual species to provide a down-core visualisation of the plant species signal. Although, overall, the humification data are validated, comparison of the original humification data with a k-adjusted version shows that the species signal is sometimes sufficient to change the timing and number of decadal to centennial-scale events recorded in the data as well as millennial to multi-millennial-scale trends. (C) 2011 Elsevier Ltd and INQUA. All rights reserved.