Document Type

Thesis - Open Access

Award Date


Degree Name

Master of Science (MS)


Biology and Microbiology

First Advisor

Nels H. Troelstrup, Jr.


EPA and state water resource agencies are now placing greater emphasis on monitoring and managing headwater streams. Two EPA stream protocols are available for headwater stream assessment but little effort has been made to compare these two methods or their resulting habitat quality index scores. The objectives of this effort were to 1) compare data types of the two protocols; 2) compare how the two protocols assess intermittent streams using habitat quality index (HQI) scores; and 3) compare stream characteristic emphases (geomorphology, riparia, substrate, in-stream cover for biota and hydrology) between the two protocols and their effect on overall HQI scores. This study was conducted within the Northern Glaciated Plains Ecoregion (NGP) of South Dakota. Forty reference sites were chosen using EPA’s Analytical Tools Interface for Landscape Assessments (ATtILA). Twenty more sites were chosen to validate the reference sites condition. Ten of the validation sites were selected at random and the other ten were targeted sites selected through consultation with state officials. All sites were field validated using the “North Carolina Division of Water Quality’s Identification Methods for the origins of Intermittent and Perennial Streams” and the “Riparian, Channel, and Environmental Inventory for small streams in the agricultural landscape.” Habitat assessments of 60 total streams occurred monthly (April-August) during the summer of 2008 following EPA’s “Western Pilot Study: Field Operations Manual for Wadeable Streams” and “Field Operations Manual for Assessing the Hydrologic Permanence and Ecological Condition of Headwater Streams.” Headwater streams in the NGP can be summarized as low gradient ( X = 0.02%) streams showing little incision ( X = 0.4 m). Channel dimensions were variable (CV = 1306.1 width/depth ratio) with flat banks ( X = 27.4ºC) and homogenous thalwegs ( X CV = 48.9 %). Substrates consisted of mostly soft/small sediments with herbaceous vegetation as the most frequently occurring instream cover for biota. With the exception of the Prairie Coteau Escarpment Ecoregion (46l), riparian trees were rare. Peck’s protocol had 51 measurements with a mixture of ratio (n = 14), interval (n = 2), ordinal (n = 23) and nominal (n = 12) data types. Fritz’s protocol had 15 measurements yielding mostly ratio (n = 10) data types, and a few interval (n = 2) and nominal data (n = 3). Substrate type was assessed differently by the two protocols. Organic substrates occurred with a frequency of 65% using Peck’s protocol, while the substrate class “sand/silt/clay” occurred most frequently (89%) using Fritz’s protocol. HQI scores for both protocols were compared using a sign test and a Wilcoxon Rank Sum test, revealing that they were different (p < 0.01). Reference HQI scores generated from Fritz metrics ( X = 71%) were higher (p < 0.01) than Peck’s HQI’s ( X = 63%). Riparian metrics composed 51% of Peck’s measurements and 7% of Fritz’s measurements but Peck’s riparian HQI’s scored lower (p < 0.01) than Fritz’s riparian HQI’s. Hydrologic metrics composed 36% of Fritz’s protocol and 4% of Peck’s protocol and still the HQI’s compared favorably between the two protocols. Evaluation of stream assessments within either protocol revealed high variability in stream characteristics within the NGP ecoregion. Stream habitat scores exhibited greater similarity within level IV EPA ecoregions than between ecoregions. This supports that regionalization by level IV ecoregions may be necessary to account for regional differences in landscape features. The use of more measurements for Peck’s protocol increased the ability to detect the influence of human management practices. However, some metrics were similar within Peck’s protocol, leading to high redundancy. Fritz’s protocol contained fewer metrics with less focus on riparian metrics, reducing the sensitivity of this protocol to human management practices. Data types also differed between and within the two protocols, complicating integration and analysis. Peck’s protocol included a large number of ordinal and nominal measurements, which require training and consistency to remain unbiased. Thus, Peck’s assessments were more subjective, adding another source of disparity between protocol assessments. Substrate was the only parameter measured by both protocols, but assessments differed due to the use of different substrate classes and a different cross-sectional methodology. Results of HQI differences provide evidence that the two protocols do not respond similarly to physical habitat changes. This can be attributed to the divergence in stream characteristics emphasized by the two protocols. Differences in metric emphasis reflect a focus on hydrologic permanence by the Fritz protocol and riparian metrics by the Peck protocol. Riparian condition reflect the influence of human activities more successfully based on HQI scores than hydrologic condition. This helps to explain differences seen in HQI scores and provides incentive for the continued use of riparian metrics in stream habitat assessments. A new combined habitat metric set is proposed which places more balance between riparian and hydrologic stream characteristics.

Library of Congress Subject Headings

Aquatic habitats
Stream ecology


Includes bibliographical references



Number of Pages



South Dakota State University


In Copyright - Non-Commercial Use Permitted