Poster Presentation Society for Freshwater Science 2025 Annual Meeting

An approach to using continuous sensor and macroinvertebrate biomonitoring data to assess the influence of dissolved oxygen on remediation and restoration efforts at Great Lakes AOC sites. (117430)

Roger Yeardley 1 , Marc Mills 1 , Adam Balz 1 , Cal Buelo 1 , Erica Yang 1
  1. U.S. Environmental Protection Agency, OH, United States

Since 2020, continuous dissolved oxygen (DO) has been monitored with deployed sensors by U.S. EPA’s Office of Research and Development (ORD) at Great Lakes Areas of Concern (AOC) sites, where tributaries interact with the world’s largest freshwater ecosystems. A comprehensive understanding of the relationship between DO and macroinvertebrate community structure is useful in assessing whether DO is a causal factor when community changes are seen at AOC remediation and restoration sites. To simultaneously characterize macroinvertebrate community and DO conditions, moorings fitted with Hester-Dendy samplers and continuous DO loggers are deployed for six-weeks across multiple stations per site. One question we address is what types of metrics, derived from the continuous DO data, correlate with macroinvertebrate community assemblages? Initial metrics evaluated were median DO, amplitude of diel DO cycles (as a measure of DO stability), and the number of days DO is below a threshold value. Initial analyses of data from 6 AOC sites using non-metric multidimensional scaling (NMS) plots, followed by Multi-Response Permutation Procedures (MRPP), indicated that median DO was the only metric to sufficiently differentiate macroinvertebrate communities (MRPP: p=0.0059) associated with 3 different levels/ categories of median DO. The Median DO metric values vector did not exactly parallel the multi-metric Lacustuary Invertebrate Community Index (LICI) or Hilsenhoff Biotic Index (HBI) vectors in NMS plots, but trended somewhat in the same direction, indicating some correspondence of DO levels with these indicators of water quality. MRPP analyses showed clear grouping of communities by LICI and HBI narrative conditions, similar to those seen between sites (p < 0. 0001). Examining grouping by environmental variables related to DO, there were clear community differences based on 3 flow regime categories (p < 0. 0001), but not based upon water depth (p=0.195).