Fossil pollen and plant macrofossils collected from lake and wetland sediments are typically used to provide qualitative paleoclimate data since it is well known that spatial distributions of plant species will shift over time in response to changes in temperature and precipitation. Some researchers have further calibrated pollen data to provide actual paleotemperature and paleoprecipitation values. Regardless of whether paleoclimate reconstructions derived from paleobotanical data are qualitative or quantitative, they all rely upon the assumption that vegetation is responding primarily to centennial- to millennial-scale variations in climate. This assumption seems to hold true for the mid-Holocene to present, but not for earlier postglacial and late-glacial times when paleovegetation assemblages responded not just to climate changes, but also to the legacies of deglaciation.

Temporal and spatial patterns of plant communities in recently glaciated areas may be explained in relation to lags in meltwater drainage, soil development, plant dispersal, and other non-climatic factors that persisted for several millennia after the retreat of the Laurentide Ice Sheet, and operated in concert with climate changes. This interpretation is based upon the analysis of >130 pollen and >400 plant macrofossil samples and 28 associated radiocarbon ages from five paleolake sites located in the glaciated northeastern Great Plains of the United States and Canada. This work also demonstrates that by investigating paleobotanical sites representative of a diversity of depositional settings and geomorphic histories that the regional "climate signal" may be discerned from the background site-specific "environmental noise".