| This thesis examines the developmental history of two large peatland complexes in boreal Alberta, and uses paleoecological analysis and simulation modelling to examine the applicability of established models of peatland development to continental peatlands.; At both Athabasca (55°03'N, 113°15'W) and Rainbow Lake (58°17'N, 119°22'W) the earliest records of peat formation coincide with the end of early or mid-Holocene aridity, and early peat-forming communities were wet, monocot-dominated marshes or fens that occupied topographic depressions. At Athabasca, the speed and direction of lateral peat expansion were largely controlled by topography, although autogenically induced changes in drainage were important in promoting paludification. At both sites, basal paludified communities include monocot-dominated or woody rich fens and Sphagnum -dominated poor fens or bogs. Local successional pathways are diverse and point towards a continued influence of allogenic effects in peatland development. In peat plateau areas at Rainbow Lake, community change in the late Holocene reflects the effect of periodic fires and permafrost dynamics. At Athabasca, there is an overall tendency for succession of wet, minerotrophic communities to Sphagnum. However, topographically controlled differences in water flow have played an important role in controlling local community development.; Long-term (apparent) rates of peat accumulation are highly variable within each site. At Rainbow Lake, decreased accumulation during the late Holocene most likely reflects regional permafrost development. At Athabasca, peat buildup has been fastest in wet, minerotrophic fens and slowest in Sphagnum -dominated communities. A simulation model that examines effects of litter quality and environment on peat accumulation shows changes in the importance of individual parameters over time and suggests that dominant controls over peat accumulation differ between community types. Sudden changes in peat accumulation dynamics that follow environmental change are not necessarily detectable from cumulative mass/age profiles, and simple peat accumulation models may fit core profiles well even if several of their basic assumptions are violated. In order to increase the utility of paleoecological records in predicting peatland responses to climatic change, we need a more dynamic approach to the interpretation of paleoecological data. |
