Scots pine litter decomposition along drainage succession and soil nutrient gradients in peatland forests, and the effects of inter-annual weather variation
Peatlands form a large carbon (C) pool but their C sink is labile and susceptible to changes in climate and land-use. Some pristine peatlands are forested, and others have the potential: the amount of arboreal vegetation is likely to increase if soil water levels are lowered as a consequence of climate change. On those sites tree litter dynamics may be crucial for the C balance. We studied the decomposition of Scots pine (Pinus sylvestris L.) needle and root litter in boreal peatland sites representing gradients in drainage succession (succession following water level drawdown caused by forest drainage) and soil nutrient level during several years of varying weather conditions. Neither gradient had an unambiguous effect on litter mass loss. Mass loss over 2 years was faster in undrained versus drained sites for both needle litter, incubated in the moss layer, and fine root litter, incubated in 0-2010 cm peat layer, suggesting moisture stress in the surface layers of the drained sites limited decomposition. Differences among the drained sites were not consistent. Among years, mass loss correlated positively with precipitation variables, and mostly negatively or not at all with temperature sum. We concluded that a long-term water level drawdown In peatlands does not necessarily enhance decay of fresh organic matter. Instead, the drained site may turn into a 'large hummock-system' where several factors, including litter quality, relative moisture deficiency, higher acidity, lower substrate temperature, and in deeper layers also oxygen deficiency, may interact to constrain organic matter decomposition. Further, the decomposition rates may not vary systematically among sites of different soil nutrient levels following water level drawdown. Our results emphasize the importance of annual weather variations on decomposition rates, and demonstrate that single-period incubation studies incorporate an indeterminable amount of temporal variation.