T298 Lakes of the World (Posters)

Wednesday, 4 November 2015
9:00 AM - 6:30 PM
Exhibit Hall

NEEDLES IN THE HAYSTACK: PARTICLE CHARACTERIZATION AND THE HUNT FOR DIATOMS IN HOLOCENE LAKE SEDIMENT FROM UPSTATE NEW YORK

Paper number: T298-9
Program: Technical Sessions
Time: 9:00 AM – 6:30 PM

Authors

Abstract

Otsego Lake occupies a glacial trough in central upstate New York and is characterized as oligo-mesotrophic, with relatively low productivity. Sedimentation in the lake is due to terrigenous influx from tributaries and biologic activity in the lake. We initiated a sediment coring campaign to reconstruct environmental records for the lake. We extracted cores taken proximal and distal to a delta in shallow water (~1 m and 4 m depths, ~30 m and 200 m offshore respectively) to capture changes in terrigenous inputs, ostensibly due to flood events. Cores span the Holocene, based on five 14C AMS dates of terrestrial plant material. The base of the longest core (5 m) is > 9 kyr BP. We report here on methods used to capture sediment characteristics, and focus on an anomalous section of core from ~6 to 2 ka. Sedimentation rates during that time decrease from ~1 mm/a to 0.12-0.15 mm/a; sediment exhibits a minor decrease in particle size, an absence of diatoms, an abundance of aquatic gastropod fragments (Hydrobiidae family) and carbonate material, and scattered plant roots. During this time interval sediment becomes more magnetic, the magnetic grain-size increases, and quartz/calcite ratio increases. We hypothesize that from 6-2 ka lake level was lower, or a long-lived avulsion of the stream on the delta delivered less terrigenous material to our core location.

Diatoms are abundant in core taken proximal to the delta, but are rare in the anomalous section of core taken distal to the delta. We employed automated particle detection to gather particle characteristics and to attempt to count diatoms in sediment smear photomicrographs using ImageJ and with a particle flow analyzer, FlowCam. We found that automated particle detection has a difficult time identifying diatoms. ImageJ could not detect diatoms as distinct from background, and FlowCam diatom images lacked detail. Automated particle analyses are helpful however in revealing grain size variations in muddy lake sediments.