Multiproxy analysis of peatland permafrost initiation in northern Norway

Palsa mires are located at the southern border of the permafrost zone making them highly sensitive to climate warming. When annual temperature and precipitation conditions are optimal, palsas form and collapse in a natural cycle. However, due to current climate warming, permafrost is widely thawing and palsas are degrading rapidly. In addition to being unique and valuable habitats for many species, palsas store significant amount of ancient carbon (C), which is released by reactivated decomposition processes when permafrost thaws, potentially changing the peatlands temporarily from C sinks to C sources. However, the age and extent of the C released remain uncertain.  

To predict how palsas will respond to ongoing climate change, understanding of the past dynamics is vital. In this study, we determined the permafrost aggradation date and reconstructed the past dynamics of a palsa in Karlebotn, northern Norway. A peat profile was radiocarbon (¹⁴C) dated, and oribatid mite analysis together with a peat type analysis and peat properties (C, nitrogen and bulk density) were performed. Oribatid mites are a diverse group of soil-living arthropods that have shown potential as environmental indicators. Recent findings suggest they can indicate the past initiation of permafrost.

Our results suggest that gradual permafrost aggradation at the Karlebotn palsa began after 1500 calibrated years Before Present (cal yr BP; present = 1950 AD) and that at 700 cal yr BP the permafrost conditions had stabilized. Using multiproxy analysis, we identified three phases in the palsa history. The first phase was characterized by moist fen conditions, the second phase was a transition phase with wet and dry condition species occurring together and the last phase was dominated by species adapted to dry conditions, and which are typical in permafrost environments. Our data also indicate that during permafrost conditions, the C accumulation rate was lower than in the early non-permafrost fen stage. While permafrost thaw will temporarily increase C emissions, the C sink capacity may ultimately increase again as the peatland shifts back to a fen stage following ground subsidence.

Only few palaeoecological studies exist from Fennoscandia where age of permafrost formation is determined. Most studies have used vegetation succession and peat properties to infer past permafrost presence, however, these methods are associated with uncertainties such as the absence of permafrost indicator plant species. This study provides additional data on historical palsa dynamics with a relatively robust chronology, based on multiple proxies, including oribatid mite community analysis. These findings contribute to our understanding of how palsas are responding to the ongoing and future climate change.