A new perspective on Quaternary land cover in central Alaska

Abstract

In the high northern latitudes vegetation is already responding to increasing global temperatures, with shrubs expanding poleward in a process called “greening”. In light of increasing global temperatures, past climate-driven shifts in vegetation can be used as analogues for future warming scenarios. This study addresses vegetation change in Alaska throughout the late glacial and Holocene, specifically how the accuracy of vegetation composition can be improved over current reconstructions.

The Landscape Reconstruction Algorithm (LRA, Sugita 2007a, b) is a two-step framework for quantitative reconstruction of land cover across various spatial scales. The model REVEALS estimates regional vegetation abundance from pollen records from large lakes, which is then incorporated into the model LOVE to arrive at local estimates of vegetation (using pollen records from small lakes). Pollen productivity estimates (PPEs) are one of the major parameters used in the LRA. The first pollen productivity estimates for the dominant forest and tundra taxa in Interior Alaska are presented in this study. The taxonomic ranking of high and low pollen producers are comparable with Europe. PPEs were used with the REVEALS model to obtain quantitative reconstructions of regional vegetation in north-central Alaska. The effectiveness of the LRA approach was then assessed in the landscape of the south Brooks Range at two small lakes, Ruppert and Lake 3. Results indicate that small lakes in Alaska produce the same pollen signal as large lakes and that REVEALS provides similar results as LOVE as the local vegetation is very similar to the regional mean vegetation.

Several key insights and prospects arise from the application of REVEALS: i) after its late glacial expansion, Betula may not have been as dominant as previously thought, but rather co-dominant with Salix; ii) deciduous, Populus woodland may have been extensive in the early Holocene; and iii) The LRA in Alaska needs further testing and validation in the tundra with 3 small lakes and 1 large lake, and would provide an opportunity to further assess the reliability of the new Alaskan PPEs.

More information

Identifiers

ORCID for Peter Langdon: orcid.org/0000-0003-2724-2643

ORCID for Mary Edwards: orcid.org/0000-0002-3490-6682

Catalogue record

Export record