Research from a variety of areas and approaches is converging upon the concept of peatlands as complex adaptive systems: self-regulating to some degree, but capable of rapid change and reorganization in response to internal developmental changes or to external forcing (4). It has long been known, for instance, that the surface of a peatland can rise and fall, sometimes dramatically, in response to rainfall or mild drought, while maintaining a fairly constant water level relative to the surface. This “Mooratmung” (“bog-breathing”) is related to the sponge-like nature of Sphagnum, which can adsorb water and trap gases. Recent studies have shown that the carbon balance of peatlands can in turn be surprisingly resilient to perturbations, even fairly severe ones. For example, subjecting peat cores (5) or a peatland field site (6) to a water table drawdown similar to a prolonged drought initially led to a respiration-driven loss of soil carbon. But both carbon loss and subsidence (6) lowered the peat surface, decreasing its height above the water table, and effectively shifted the system back toward its starting state. Conversely, a rising water table stimulated growth of Sphagnum and other vegetation, which increased carbon accumulation, raised the surface of the peat and, in effect, lowered the local water table (5). Thus, an environmental perturbation may trigger an initial gain or loss of carbon, but recovery in the direction of the initial state can moderate the impact.
Long-term global changes—particularly warming, drought, and elevated nitrogen deposition—are likely to ultimately induce shifts in some existing peat-forming areas to new ecosystems such as grassland or shrubland (10, 11), and the increase in biomass from vascular plants could in part compensate for carbon losses from soil oxidation during the transition. However, even if some net carbon accumulation returns, the gains are short-lived: The key peatland quality of slowly removing and storing carbon for hundreds or thousands of years is lost.
Considering peatlands as complex adaptive systems characterized by quasistable equilibrium states—resilient to change at some level of perturbation but shifting to new states at higher levels of disturbance—provides a meaningful framework for understanding and modeling their response to environmental change. Ignoring the strong feedbacks inherent in peatlands may lead to substantial under- or overestimates of their response to global change. The challenge is to forecast both the future environmental conditions that peatlands will experience and the internal feedbacks and state changes that may be triggered by these conditions. To meet this challenge it is vital to continue and expand long-term monitoring networks to characterize the present, paleo-environment research to reconstruct the past, and manipulation experiments in the field and laboratory to build our understanding of these unique and valuable ecosystems.