Modelling microtopography in boreal peatlands: hummocks and hollows

Abstract

Surface topography of boreal peatlands is frequently hummocky. The hummocks and hollows develop from the peat and remain stable over long periods of time, even while climate and other environmental conditions change. Processes leading to these stable surface forms have not been clearly understood. The HOllow–HUMmock (HOHUM) model provides an explanation of bog microtopography, resilience, and stability. It suggests that bog microtopography results from moss species properties interacting with the physical environment. HOHUM is a dynamic, process-based simulation model simulating biological and physical processes leading to hummocks, hollows, and pools in boreal peatlands. It consists of several interacting submodels that grow hummocks and hollows at a monthly time step. Climate is represented by long-term monthly means for a specific location. The hydrology submodel calculates seasonal water table fluctuations relative to the peat surface. The vegetation submodel grows Sphagnum mosses using climate and moisture information and generates litter input; and the peat submodel simulates species- and layer-specific peat decay and accumulation. The model tracks the total depth of the top peat layer (acrotelm) of bogs produced by three common species of Sphagnum mosses, one of which forms high hummocks, another forms lower or middle hummock layers, and the third produces hollows. Model results indicate that hummocks and hollows arise primarily from species differences expressed through different peat accumulation rates. Interactions among moisture, unique Sphagnum species characteristics, and decomposition rates are responsible for maintenance and resilience of hummocks and hollows. Feedbacks between Sphagnum growth, decay, and moisture restrict peat accumulation over time, creating a stable equilibrium with local climate within decades. These feedbacks appear responsible for the long-term maintenance of bog hummocks and hollows and their adaptation to changing environmental conditions. Model validation showed excellent agreement (within 2 cm) between predicted and measured heights of hummocks, hollows, and water table depths for North American and European sites, and successfully reproduced conditions not anticipated during model development. Sensitivity analyses indicate that the model is most sensitive to decomposition rates, Sphagnum production rates, peat bulk densities, and the effects of hummock ice retention into the spring. Simulations for 35 northern US and Canadian sites predict distinct bog surface microtopography: hummocks and hollows are well developed in the eastern and northeastern regions of North America with the highest hummocks in far northeastern Canada, north of which hummocks grade into low surface relief. In the high-latitude Canadian interior, extreme arctic conditions (cold and dry) prevent bog formation. South of the Great Lakes, wet hollow species are replaced by more drought-resistant species to form relatively flat lawns. These simulated geographic patterns agree with those reported for North American peatlands.