Peatland carbon balance and climate change: from the past to the future

Storage of atmospheric carbon in northern peatlands imparts a cooling effect upon global climate. Climate change may alter peatland carbon cycling, accelerating both decomposition and plant productivity, potentially driving positive or negative climate feedbacks. Experimental and palaeoecological methods are commonly used to investigate peatland responses to climate change, but results are often in disagreement. Whether positive or negative climate feedbacks will dominate in the future is uncertain. This thesis links experimental and palaeoecological approaches on a raised bog in Wales (Cors Fochno), testing the effects of ten-years of warming and increased drought frequency upon ecosystem functioning, and comparing climate responses with those that have occurred in the bog during the past ~1500 years. In the experimental plots, warming reduced both carbon accumulation and methane emissions, and when combined with drought caused the bog to become a net carbon emitter. Shrub abundance increased with warming, which was also seen in the palaeoecological record. Sphagnum abundance did not respond to experimental manipulations. During the past ~1500 years, changes in carbon accumulation corresponded with vegetational succession. Sustained water table drawdown and wildfires resulted in more decay-resilient plant communities. Climate manipulation altered the stabilisation of organic matter with depth. Temperature increases similar to those in the past millennium did not affect decay rates in the plots. Decay during the last ~1500 years resulted from vegetation changes, rather than from temperature changes. Climate change may cause positive feedbacks to dominate in the short term, but long-term shifts towards more decay resistant plant species may offset emissions and contribute to climate cooling. The climatic sensitivity of peatland palaeoecological proxies should be tested in modern settings before they are used to test climate models. As a result of anthropogenic disturbance, palaeoecological records are no longer a good analogue for contemporary and future peatland functioning.