home

RAN: Redesigning the American Neighborhood Project

RAN fitzgerald default

Back to Project Components page >>

Stream Ecosystem Response to Watershed Development

Evan Fitzgerald, Master's thesis research project in Natural Resources at the University of Vermont, 2007

Urban and suburban stormwater runoff is a widely recognized threat to surface water quality. The Impervious Cover Model (ICM) was developed to explain the general response of stream ecosystems to stormwater runoff across a range of physiographic regions. Research has shown that a stream ecosystem responds when the total impervious area (TIA) is at or above 10% of the watershed area, but little is known about the impacts of urbanization on stream ecosystems at different spatial and temporal scales.

Evan's work explored these impacts by testing (1) the effect of TIA on geomorphic stability, physical habitat conditions, and biotic communities at three different spatial scales; (2) the differences between urban and rural downstream hydraulic geometry (DHG) regressions; and (3) the response of stream communities to different stages of urban channel evolution.

The physical and biological conditions of stream reaches from 16 small watersheds in northwestern Vermont were assessed and analyzed for a response to TIA at multiple spatial scales. Separate analyses were performed for high and low-gradient stream types. Reach selection criteria minimized the influence of human impacts on channel boundary conditions (e.g. bank armoring) to ensure a robust test of the ICM for upslope TIA alone.

The response of geomorphic stability and sensitive macroinvertebrates to TIA was nonlinear and significant (P<0.001), decreasing rapidly at >5% TIA. The impact of TIA on stream condition was affected significantly by drainage area and channel slope (P<0.05). DHG regressions developed for urban and rural watersheds showed significant scale-dependent responses (p=0.001) of channel width to urbanization.

Macroinvertebrate data from reaches in different stages of channel evolution indicated that stable reaches supported greater numbers of sensitive biota and total species richness than unstable reaches.

Results of ICM analyses and DHG regressions demonstrated that streams in Vermont may be more sensitive than those in other regions of the country in their response to urbanization, and that the response is scale-dependent. Results also indicated that some recovery of biotic communities may be possible following natural channel restabilization.

A separate analysis using urban and rural DHG regression parameters published for different physiographic regions of the United States was performed. DHG parameters alpha and beta described the intercept and regression slope, respectively, and non-parametric tests showed significant differences between urban and rural watershed types using both parameters from channel width equations (P<0.005).

These results showed a consistent response of channel width to urbanization that was dependent upon the size of the watershed up to areas of ~50 km2, a scale that includes first- to third-order headwater streams that are often directly impacted by urban and suburban development.

These results have important implications for land use planners in urbanizing watersheds and stream restoration professionals intending to use DHG regressions for channel restoration designs. Failure to recognize scale-dependent differences in the response of channel geometry to urbanization could lead to improper channel restoration designs and project failure.

Evan Fitzgerald earned his BS in Environmental Conservation from the University of New Hampshire. After completing his MS degree, he started his own consulting firm, Fitzgerald Environmental, Inc. in Colchester, VT.

For specific data generated during this study, visit the Vermont Agency of Natural Resources Geomorphic Assessment website using the following links:

https://anrnode.anr.state.vt.us/ssl/sga/security/frmLogin.cfm

http://maps.vermont.gov/imf/sites/ANR_SGAT_RiversDMS/jsp/launch.jsp?popup_blocked=true

Back to Project Components page >>

Last modified October 16 2009 12:32 PM

Contact UVM © 2014 The University of Vermont - Burlington, VT 05405 - (802) 656-3131