Identifying Sites for Riparian Wetland Restoration: Application of a Model to the Upper Arkansas River Basin Michael P. O'NeiU' John C. Schmidt' James P. Dohrowolski* Charles P. Hawkins' Christopher M. U. Neale' Abstract We present a conceptual model for identifying resto- ration sites for riparian wetlands and discuss its ap- plication to reaches within the Upper Arkansas River basin in Colorado. The model utilizes a Geographic Information System (GIS) to analyze a variety of spa- tial data useful in characterizing geomorphology, hy- drology, and vegetation of riparian wetland sites. The model focuses on three basic properties of riparian wetland sites: relative soil moisture, disturbance re- gime, and vegetative characteristics. A relative wet- ness index is used to define nominal soil moisture classes within the watershed. These classes generally coincide with uplands (low), channel margins (moder- ate), and channels or open water (high). Vegetative conditions are characterized using color infrared aerial photographs. Land cover types are grouped into five major land cover classes: riparian, moist herbaceous, bare ground, upland, and open water. Disturbance re- gime is characterized by a reach-based index of spe- cific power (w). Preliminary results indicate that reaches within the Upper Arkansas River basin can be classi- fied as high energy (to :s 8 W/m^) or low energy (co ^ 3W/m^), using discharge estimates that reflect the 10- year flood event. Field surveys of channel and flood- ' Watershed Science Unit, Utah State University, Logan, UT 84322-5250, U.S.A. © 1997 Society for Ecological R^^^ plain conditions show that high-energy reaches (w ^ 8 W/m^) are characterized by sites where the channel oc- cupies a large proportion of the valley bottom. By con- trast, low-energy reaches (to ^ 3 W/m^) are character- ized by meandering channels with wide alluvial valleys. Restoration potential is evaluated as a combination of nominal scores from wetness, land cover, and distur- bance indices. Application of these methods to field sites within the Upper Arkansas River basin identifies a wide range of riparian wetland sites for preservation or restoration. Potential sites within identified reaches are prioritized using size and proximity criteria. Introduction T he goal of riparian wetland creation or restoration is to restore wetland function or values (Kuster & Kentula 1990; Kentula et al. 1992). To achieve proper function, however, it is necessary to consider wetland environments within their formative spatial-temporal context {Orme 1990; Kentula et al. 1992). Riparian wet- lands in the west commonly are described as "narrow, linear features on the landscape, often lining streams with steep gradients and narrow floodpiains" (Mitch & Gosselink 1993). These areas are prone to disturbance by natural and anthropogenic influences (Friedman 1993). Flood hydrology (Richter et al. 1996; Scott et al. 1996), sedimentation (Sigafoos 1964; Fveritt 1968), and ef- fects of woody debris (Hickin 1984) all are considered critical elements of riparian zone ecology. As a result, re- cent restoration efforts consider geomorphology and hy- drology key elements to successful restoration (Carothers et al. 1990; Jensen & Platts 1990). Formative conditions for riparian wetlands involve complex interactions between stream channels, flood- plains, soils, and plant communities within the riparian corridor (Stromberg et al. 1991; Scott et al. 1996). Flood disturbance is recognized as a key element in the devel- opment and maintenance of riparian corridors. Stromberg et al. (1993) found that a 10-year flood on the Hassayampa River produced widespread erosion and deposition that led to creation of new recruitment sites for desirable riparian species. Friedman (1993) also demonstrated a strong link between disturbance and riparian condition. He indicated that without disturbance, riparian forest communities would be replaced by grassland or shrub- land along Plum Creek, Colorado. Furthermore, he showed that native plants were better at colonizing dis- turbance sites than exotics present at these same sites. In a geomorphic context, floods at or near the bank- full level commonly are believed to perform the great- est amount of work, particularly in alluvial stream channels (Wolman & Miller 1960; Leopold et al. 1964). However, larger magnitude floods may be responsible for performing more work in smaller systems and less DECEMBER 1997 Restoration Ecology Vol. 5 No, 4S, pp. 85-102