Multi-scale hydroclimate reconstruction using co-located lake and bog records from Maine and comparison with other records from the Northeast US

Abstract:

Sedimentary lake-level records and ombrotrophic bog water-table depth records both document hydrologic variability over the Holocene. Lake level records have long temporal length (10,000+ years) and fidelity in preserving low-frequency trends and centennial to millennial length events. Hydrologic reconstructions based on peatland testate amoebae assemblage composition are sensitive to moisture variability at interannual to multidecadal time scales and precipitation on the bog surface is the sole moisture input. However, bog records are generally not as long as lake level records and bog development processes can confound centennial to millennial trends. In this study we present and combine new reconstructions from Giles Pond, Aurora, Maine, USA and Caribou Bog, Old Town, ME USA.

The lake-level record from Giles Pond extends a network of lake-level records from southern New England that show an orbitally driven long-term trend toward wetter conditions punctuated by low-water phases in the mid- to late-Holocene that each lasted 100 to 400+ years. Some of these low lake level events appear to be synchronous across multiple sites in New England (Newby, et al. 2014 GRL). Preliminary data from Giles Pond suggest that some of these events extended all the way to Maine. Thus, there were New England-wide dry periods within the last 5000 years that lasted more than 100 years. These long low stands are unlike anything observed during the historical period and the interannual to decadal variability during these low stands is poorly understood. This leads to challenges in understanding the modern and future implications of the lake-level record alone. The Caribou Bog record also builds on a network of peatland water-table reconstructions from the Northeast, and contributes higher-resolution hydroclimate information that adds interannual to multidecadal texture to the centennial to millennial variability of the Giles Pond record.

Our multiproxy approach allows us to use the strengths of each of these co-located proxy records. In doing so, we develop a multi-scale reconstruction of the hydroclimate history of Maine and the broader region. The combined record is more closely comparable to instrumental climate data and future projections and therefore more relevant for resource managers and policymakers.