Pollen, spores, and dinoflagellate cysts are used with the modern analog technique (MAT) to provide paleoclimate reconstructions for terrestrial and marine environments in northeastern North America. Multivariate analysis of marine and nearby terrestrial pollen sequences from Hudson Bay, Labrador and the St. Lawrence, differentiate tundra, boreal and deciduous forest assemblages in time and space. These three regions had differing climate histories with respect to deglaciation and air mass boundaries. Prior to 6000 14C yr BP, cooler temperatures reconstructed along the Labrador margins agree with climate simulations indicating a persistent anticyclone over the Quebec-Labrador ice sheet. A late Holocene cooling at forest-tundra sites suggests a recent southern movement in the mean position of the polar front. The degree to which those critical thresholds of dissimilarity, that are used to identify non-analog pollen assemblages, are due to limitations of the modern pollen database or critical decisions within the MAT are explored by means of stochastic simulation, spatial statistics and graphical techniques. Critical thresholds of SQD, as derived by the expected value under randomization, become greater as the number of taxa in the pollen set increases. Larger pollen sets, with continentally infrequent but regionally abundant taxa, better distinguish between continental vegetation zones. Global and local spatial autocorrelation within climate anomalies indicate where the modern sample network induces biases in the climate reconstruction using the MAT. The spatial scale of terrestrial climate or vegetation reconstructions from pollen in lake-sediments is investigated through the study of pollen source area in southern Quebec. Despite the different regional vegetation, estimated pollen source areas and relevant pollen productivity for Pinus, Picea, Abies, Fagus, Quercus and Tsuga are consistent with studies from Michigan, Wisconsin and Sweden. These estimates are robust with respect to various plant abundance distance-weighting schemes and imply that the same inferences can be made regarding plant abundance from pollen throughout a lake-derived fossil pollen sequence. Stochastic simulations illustrate that the definition of relevant pollen source area requires consistent within-site vegetation heterogeneity within a network of pollen sites. Underutilized proxy-climate data from wetland taxa are demonstrated to contain climate signals at the continental scale and have the potential to further our climatic and biogeographic picture of North America over the past 21,000 years. Pollen and spores from modern wetland taxa conform to their geographic ranges and allow interpretations of their past range changes. The climatic tolerances that govern their geographic distributions are used to interpret past range changes in climatic terns. Sphagnum spore distributions suggest major peatland developments after 9 ka and 5 ka. Sphagnum, Potamogeton, Isoetes, Myriophyllum Typha/Sparganium, and Menyanthes trifoliata were in Alaska during the last glacial maximum (LGM) and moved into the ice-free corridor by 13 ka. Since the LGM, four migration routes for aquatic taxa are identified in response to the climate changes of the late Quaternary.
Identifer | oai:union.ndltd.org:uottawa.ca/oai:ruor.uottawa.ca:10393/9031 |
Date | January 2001 |
Creators | Sawada, Michael Charles. |
Contributors | Gajewski, Konrad, |
Publisher | University of Ottawa (Canada) |
Source Sets | Université d’Ottawa |
Detected Language | English |
Type | Thesis |
Format | 254 p. |
Page generated in 0.0278 seconds