Patterns and Drivers of Recent Peatland Carbon Accumulation in Northeastern Canada

Abstract

Northern peatlands are an important component of the global carbon (C) cycle and have been a net sink of atmospheric C during the Holocene. Under current climate warming conditions, the future sink-source balance of these peatlands is uncertain. In particular, peatlands near the southern limit of permafrost are likely to be sensitive to changes in topography as well as climate. In order to predict how the sink-source balance may change, this thesis focuses on determining the generality of observed patterns of C accumulation in Northeastern Canada.

The methodological approach in this thesis is unique. A total of 30 cores were taken from 9 peatlands located in 3 ecoclimatic regions along the North Shore of the Gulf of St Lawrence. This replication of records allows for climate-scale (allogenic) signals to be separated from the internal or local factors (autogenic), and for statistical testing of differences between regions and within sites over time. Trends in carbon accumulation rates (CAR) were analysed on three levels: (1) within individual sites along a hydrological or microtopography gradient, (2) between overall regions located along a climatic or permafrost gradient, and (3) over time on a multi-centennial scale. Lead-210 (210Pb) dating was used throughout the analysis to increase temporal resolution for the last 150-200 years of C accumulation. The method was thoroughly tested from preparation to analysis and found to produce reliable results, comparable with other dating methods. These dates were then used to develop combined age-depth models for longer-term context. Replicated records of 210Pb inventories and fallout rates were also used to address questions of deposition patterns and post-depositional mobility in peat profiles. Total inventories decreased with water table depth, with lichen hummocks having significantly higher inventories. One site also received significantly higher 210Pb deposition than the other two, as it is more sheltered from the Gulf influence.

Recent carbon accumulation rates for the 150-year period for all microforms across all regions was 62.1 ± 4.4 g C m-2 a-1, and were highest for Sphagnum hummocks (79.9 ± 8.9 g C m-2 a-1) and lowest for dry lichen hummocks (42.7 ± 6.2 g C m-2 a-1). Patterns and trends at this scale were mainly driven by autogenic processes, including incomplete decomposition in the acrotelm peat. Models of peat accumulation related to acrotelm thickness were found to be overly simplistic, as carbon accumulation for intermediate microforms showed large natural variability driven by changing ecohydrological feedbacks, in part due to permafrost degradation at one of the sites. Over a multi-centennial scale, carbon accumulation rates were driven by a combination of climatic changes and ecohydrological feedbacks due to shifts in the microform configuration in response to permafrost degradation. Changes in carbon accumulation rates were detected and coincided with Little Ice Age temperature/solar minima (including the Spörer, Maunder and Dalton Minima), permafrost degradation since the 1950s, and recent climatic changes in the mid-1990s. Snow cover and exposure of sites and microforms were found to play an important role, rather than solely climatic variables. Rapid Sphagnum re-establishment in post-permafrost degraded features and increasing temperatures meant that carbon accumulation was highest for the northernmost site in the transect. Age-depth models using a combination of lead-210 and radiocarbon dates allowed for the calculation of carbon accumulation rates at a decadal resolution. While peat carbon sequestration is projected to increase in northern regions, the fate of peatland C near the southern limit of permafrost is complex. Future studies seeking to interpret recent changes should include multiple cores and consider both regional climatic and local ecohydrological drivers.