Potential implications of differential preservation of testate amoeba shells for paleoenvironmental reconstruction in peatlands

Testate amoebae are now commonly used in paleoenvironmental studies but little is known of their taphonomy. There is some experimental evidence for differential preservation of some testate amoeba shell types over others, but it is unclear what, if any impact this has on palaeoenvironmental reconstruction. To investigate this issue we looked at palaeoecological evidence for the preservation of different shell types. We then investigated the possible impact of selective preservation on quantitative palaeoenvironmental inference. We first used existing palaeoecological data sets to assess the vertical patterns of relative abundance in four testate amoeba shell types: (1) shells made of secreted biosilica plates (idiosomes, e.g. <i>Euglypha</i>), (2) idiosomes with thick organic coating (<i>Assulina</i>), (3) proteinaceous shells (e.g. <i>Hyalosphenia</i>), (4) shells built from recycled organic or mineral particles (xenosomes) (e.g. <i>Difflugia</i>, <i>Centropyxis</i>). In three diagrams a clear pattern of decay was only observed for the idiosome type. In order to assess the implications of differential preservation of testate amoeba taxa for paleoenvironmental reconstruction we then carried out simulations using three existing transfer functions and a wide range of scenarios, downweighting different test categories to represent the impact of selective test decomposition. Simulation results showed that downweighting generally reduced overall model performance. However downweighting a shell type only produced a consistent directional bias in inferred water table depth where that shell type is both dominant and shows a clear preference along the ecological gradient. Applying a scenario derived from previous experimental work did not lead to significant difference in inferred water table. Our results show that differential shell preservation has little impact on paleohydrological reconstruction from <i>Sphagnum</i>-dominated peatlands. By contrast, for the minerotrophic peatlands data-set loss of idiosome tests leads to consistent underestimation of water table depth. However there are few studies from fens and it is possible that idiosome tests are not always dominant, and/or that differential decomposition is less marked than in <i>Sphagnum</i> peatlands. Further work is clearly needed to assess the potential of testate amoebae for paleoecological studies of minerotrophic peatlands.