Carbon Cycle Responses to Experimental Drought and Warming in a Welsh Ombrotrophic Peatland in the Context of Late Holocene Carbon Accumulation

Peatlands are important sinks and stores of terrestrial carbon, but their future role in the global climate system under climate change is uncertain. Palaeoecological and experimental climate manipulation studies are commonly used to address these uncertainties, although their results often contradict each other. Furthermore, carbon accumulation rates estimated from both methods frequently fail to converge.We examine the effects of warming and episodic drought on carbon cycling from a long-term experimental site in an ombrotrophic peatland in Wales, simulating expected changes in temperature and drought frequency for the next c.70 years. We compare carbon accumulation rates modelled from the 10-year experiment (2010-2020) with those derived from a peat core covering the past c. 1500 years. We partition components of the annual carbon budget for each experimental treatment and compare them with core-derived carbon accumulation rates.Significant differences in carbon accumulation occurred between experimental treatments, resulting from cumulative non-significant changes in decay rates and productivity, driven by differences in air temperature and water table depth. We identify a positive time-lagged relationship between carbon accumulation and reconstructed temperature anomalies in the long-term carbon accumulation record. Changes in carbon accumulation also correspond to vegetation changes. We suggest that palaeoecological studies should not consider changes in plant productivity to be the sole determinant controlling carbon accumulation rates and should acknowledge the role of changing decay rates following plant succession upon long-term carbon accumulation. Carbon accumulation rates estimated from both methods may converge, but often differ due to unquantifiable losses such as by reduced productivity, decay, or anthropogenic disturbances like fire. We conclude that both methods are useful for discerning future peatland responses to climate change; however, core derived rates should be interpreted with consideration of all the contemporary mechanisms governing carbon accumulation. Based on these findings, we posit that the carbon sink function of existing peatlands will be weakened by increased warming and drought