Testing peatland water-table depth transfer functions using high-resolution hydrological monitoring data

•We test peatland testate amoeba transfer functions using independent datasets.•Reconstructions based on testate amoebae reflect directional changes in water table.•They are poor representations of the real mean or median water-table magnitudes.

Transfer functions are now commonly used to reconstruct past environmental variability from palaeoecological data. However, such approaches need to be critically appraised. Testate amoeba-based transfer functions are an established method for the quantitative reconstruction of past water-table variations in peatlands, and have been applied to research questions in palaeoclimatology, peatland ecohydrology and archaeology. We analysed automatically-logged peatland water-table data from dipwells located in England, Wales and Finland and a suite of three year, one year and summer water-table statistics were calculated from each location. Surface moss samples were extracted from beside each dipwell and the testate amoebae community composition was determined. Two published transfer functions were applied to the testate-amoeba data for prediction of water-table depth (England and Europe). Our results show that estimated water-table depths based on the testate amoeba community reflect directional changes, but that they are poor representations of the real mean or median water-table magnitudes for the study sites. We suggest that although testate amoeba-based reconstructions can be used to identify past shifts in peat hydrology, they cannot currently be used to establish precise hydrological baselines such as those needed to inform management and restoration of peatlands. One approach to avoid confusion with contemporary water-table determinations is to use residuals or standardised values for peatland water-table reconstructions. We contend that our test of transfer functions against independent instrumental data sets may be more powerful than relying on statistical testing alone.