Biologically Based Urban Response Models for the South Atlantic Gulf and Tennessee River Basins
By T.F. Cuffney1, E.M.P. Giddings1, and M.B. Gregory2
1 U.S. Geological Survey, North Carolina Water Science Center, 3916 Sunset Ridge Road, Raleigh, North Carolina 27607
2 U.S. Geological Survey, Georgia Water Science Center, Peachtree Business Center, Suite 130, Amwiler Road, Atlanta, GA 30360-2824
Abstract
U.S. Geological Survey National Water-Quality Assessment (NAWQA) Program personnel are
investigating the effects of urbanization on streams across the United States by examining biological,
chemical, and physical changes along gradients of urban intensity defined by a multimetric urban
intensity index (UII). Biological-response models were derived by using invertebrate assemblages to
quantify ecological distances among sites and then relating these distances to the UII using linear
regression. Response models were derived for Birmingham, AL (BIR); Atlanta, GA (ATL); and the
urban corridor along I-40 from Raleigh to Winston-Salem, NC (RAL) based on richest habitat (RTH),
qualitative multi-habitat (QMH), or QQ (RTH + QMH) invertebrate samples.
Models based on quantitative single habitat (RTH) samples did not perform as well as models based on
qualitative multi-habitat (QMH and QQ) samples. In ATL, the RTH model explained only 31% of the
invertebrate response whereas the QQ model explained 69%. In all, the urban response models based on
QQ samples were able to explain 69% (ATL), 78% (RAL), and 83% (BIR) of the variation in biological
responses.
Response models (ATL, BIR, and RAL) were used to predict responses at 294 NAWQA Program sites in
the South Atlantic gulf and Tennessee River basins. The applicability of each model was assessed based
on the deviation of predicted from actual values for site score and UII. Sites with widely different natural
environmental characteristics tended to have fewer taxa in common with the underlying response model
and deviated more from expected results. For example, response models developed from small basins
performed poorly when applied to large rivers. There was, however, considerable correspondence among
the three models derived from small streams. The applicability of these biologically based models of
urbanization to small streams indicates that a general model of urban effects can be developed for this
area.
Citation:
Cuffney, T.F., Giddings, E.M.P., and Gregory, M.B., 2006, Biologically-based urbanization models in the South Atlantic-Gulf and Tennessee River Basins: Proceedings of the National Water-Quality Monitoring Council National Monitoring Conference, May 8-12, 2006, San Jose CA.
|