Subfossil peatland trees as proxies for Holocene palaeohydrology and palaeoclimate

Due to the scarcity of reliable and highly resolved moisture proxies covering much of the Holocene, there has been increased interest in the study of living and subfossil peatland trees sensitive to gradual and extreme changes in hydrology, precipitation, and related environmental processes. Peatland development and the associated carbon accumulation, which are strongly influenced by hydrological fluctuations, are also of prime importance as peatlands represent long-term sinks of atmospheric carbon. Improved knowledge of peatland development and soil moisture variability during the Holocene is therefore essential to our understanding of long-term hydroclimate changes, the terrestrial carbon cycle, and to enable more robust predictions of peatland response to future climate changes. Here, we review the existing mid- to late Holocene peatland tree-ring chronologies that have been used to study climate variability on (sub-)annual to centennial scales with a primary focus on northern Europe. Since the 1970s, absolutely dated tree-ring chronologies covering substantial parts of the Holocene have been developed from excavated remains of oak (Quercus spp.) and pine (Pinus sylvestris L). The annual tree-ring patterns of these trees are often characterized by periods of depressed growth reflecting annual to decadal hydroclimatic changes. In addition, changes in the spatio-temporal distribution of trees throughout the Holocene are often found to reflect decadal to centennial climate and hydrological changes. Moreover, synchronicity between tree-ring chronologies and tree-population dynamics over larger geographical areas show periods of coherent regional climate forcing, especially during the mid-Holocene. This review (i) provides an overview of pioneering and recent studies presenting tree-ring chronologies developed from subfossil peatland trees, and (ii) presents recent developments in the fields of dendroecology (i.e. the response of tree growth and changes in vitality as a result of changes in climatic variables) and dendroclimatology (i.e. the reconstruction of climate fluctuations based on tree-ring analyses) in peatland regions. Moreover, we (iii) use long-term climate reconstructions based on alternative proxies for comparison, and (iv) present different ways to analyse tree-ring records to generate novel information on annual to centennial timescales. This analysis is based on an unprecedented network of tree-ring chronologies from Denmark, Finland, Germany, Great Britain, Ireland, Lithuania, the Netherlands, Poland, Sweden, and Canada, as well as a wealth of old and previously (un) published literature from Scandinavia and Germany, which has not been accessible to a wider audience in the past due to inaccessibility or linguistic barriers. Finally, a map of possible hotspots for the assessment of continuous peatlandtree studies is presented, along with suggestions for new research directions in the field. (C) 2016 Elsevier B.V. All rights reserved.