Citation:

Fritz, K., B. Washington, G. Pond, J. Christensen, L. Alexander, R. Kashuba, AND B. Johnson. Identifying Invertebrate Indicators for Streamflow Duration Assessment Methods. 2021 Society for Freshwater Science Annual Meeting, Virtual, May 23 - 27, 2021.

Impact/Purpose:

The presence of surface flow is a fundamental basis for stream classification used in the management of water resources. However, because direct measurement of flow duration is too resource intensive there is a need for rapid approaches for accurately characterize flow duration class. Streamflow duration assessment methods (SDAMs) are rapid, indicator-based tools for classifying streamflow duration at the reach scale (40 - 200 m). This study explores the utility of density, biomass, and presence/absence of invertebrates identified at different taxonomic levels to accurate predict whether or not the associated stream reach has year-round flow (perennial) or dries for some period during the year (intermittent). Findings from this study support the use of invertebrates as indicators for SDAMs developed for small forested streams and the degree of detail needed to optimize classification accuracy. More information on SDAMs and how they are being developed by USEPA's Office of Water and Office of Research and Development can be found here: https://www.epa.gov/streamflow-duration-assessment

Description:

Streamflow duration assessment methods (SDAMs) are rapid, indicator-based tools for classifying streamflow duration at the reach scale. SDAMs use indicators as surrogates because direct measurement of flow duration is too resource intensive for many reaches. Invertebrates are commonly used as SDAM indicators because many are not highly mobile and have life stages that require flow for extended periods. Aquatic and terrestrial invertebrate data (presence/absence, density, and biomass) at the family- and genus-levels were analyzed from 370 samples across 36 intermittent and 53 perennial reaches distributed along 31 forested headwater streams. Random forest models for family- and genus-level datasets had classification accuracy ranging from 89.7% to 91.9%, with slightly higher accuracy for density and biomass than for presence/absence datasets. To reduce geographic bias, the family-level presence/absence dataset was weighted to balance bootstrap selection by ecoregion, and the global model had accuracy of 89.2%. Both aquatic and terrestrial taxa were among the top predictors and included positive (presence indicative of perennial) and negative (absence indicative of perennial) indicators. Our findings highlight that aquatic and terrestrial invertebrates can be effective indicators of flow class for forested reaches, supporting the data-driven development of SDAMs.

URLs/Downloads:

Record Details: