Green River Floodplain Habitat Restoration for Endangered Fish Species

Historically, flooded bottomlands along the Green River were frequently flooded by spring peak flows. These floodplain habitats have an influential role on the river ecology (Bayley 1995, Ligon et. al 1995, Townshend 1996). Water resources development, encroachment of non-native vegetation and land use changes have resulted in most of the bottomlands being cut-off from interaction with the river channel at all but high flows.The Colorado River Recovery Program (CRRP) has developed a Habitat Restoration Program to restore or enhance natural floodplain functions to re-establish self-sustaining populations of four endangered fishes endemic to the Upper Colorado River Basin. The project goals were to identify the pre- and post-water resources development hydrology of the river basin, evaluate existing bottomland sites for potential restoration, provide designs to enhance the floodability of these floodplain habitats, implement the restoration designs, and monitor the success of the restoration. This discussion focuses primarily on the success of physically manipulating the bottomland sites to flood and the erosion and sedimentation monitoring of the sites post-restoration. References are provided that discuss the success of the restoration in terms of ecosystem response and fish utilization.The study was initiated in 1995 with the evaluation of six Green River bottomland sites along the Ouray National Wildlife Refuge, near Ouray, Utah. In the following years, more than thirty sites have been studied in designated critical habitat reaches along the Green River near Vernal, Utah, and the Colorado and Gunnison Rivers near Grand Junction Colorado. Initially, only public lands were investigated, but in 1997, the floodability assessments were extended to many privately owned sites. To date, eight of these bottomland sites have been restored providing increased flood frequency and seasonal fish access to over two thousand acres of bottomland habitat. Monitoring at these sites has been conducted since they were constructed in 1997 and 1998 to determine the success of seasonal flooding and fish usage.The floodability assessments involved the collection of hydrologic data including river cross section surveys, topographic mapping of the river banks and overbank areas, and discharge measurements to evaluate stage discharge relationships. A review of the river hydrology was also performed to determine pre- and post-water resources development peak flow frequencies and durations for specific reaches. The hydrologic restoration goals were to physically manipulate the site such that overbank flooding occurred with frequencies similar to pre-dam hydrology under the current hydrologic regime. For example, in the Green River near Vernal, Utah the average bankfull discharge of 20,300 cfs occurred two out of every three years before water resources development. Since water resources development in the basin occurred, this discharge is exceeded only once every three to four years, statistically. Therefore, these eight pilot project sites have been physically manipulated by levee removal and/or channel excavation so the restored flooding elevation corresponds to 13,000 cfs. This Green River flow corresponds to the pre-water resources development flood frequency using post-dam hydrology.Several different configurations were used to hydrologically connect bottomland sites to the river. Configurations ranged from a small twenty-foot wide channel at the lower end of the bottomland that allows backwater flows to connect the river to the bottomland to large-scale 1000′ long levee removals providing a long lateral connection between the river and bottomland. Most of the sites were connected at the lower end of the site to allow flows to backwater into the bottomland and minimize the amount of deposition experienced in the openings. Three years of erosion and sedimentation data have been collected at the restoration sites.The success of the different restoration configurations have been evaluated based on their ability to provide a sustained connection for fish passage between the river and bottomland, the influence of local hydraulics on larval drift, and the magnitude of erosion and sedimentation in the inlets and outlets. Preliminary results have shown that the sites have been relatively successful in maintaining their floodability. Deposition on the order of one to two feet per year was experienced at some of the larger openings. However, when the sites drained, scour channels developed that would maintain the original flooding elevation more or less. Unfortunately, the scoured channels only represent a small portion of the river-bottomland connection. Therefore, the as-built lateral surface connections at the larger openings have been reduced, in some cases up to 80%, thus reducing the area available for fish passage. However, significant deposition was anticipated in the design and full surface connections have been occurring, just at slightly higher flows than the original design flows.Preliminary results have also indicated that restoration activities have not had any negative impact on local morphology and that deposition in the bottomlands themselves has been minimal. There has been some significant sand bar development adjacent to some of the manipulated openings. However, none of the sand bar development or changes at adjacent river cross sections have been out the realm of typical changes to river morphology in a sand bed channel. Long-term monitoring may provide some trends that develop due to restoration activities.A conservative approach has been taken thus far to the floodplain restoration activities in the Green River basin in that the size of the excavated openings have been relatively small and the restored sites only represent a small fraction of the available floodplain acreage in the basin. The sites have proven successful in providing and maintaining river floodplain interaction and the ecosystem response has been encouraging in terms of the overall productivity of the river system (Levee Removal Evaluation Group, 1998). However, improvements can be made in terms of the success of larval drift into the sites. Proposed improvements include more flow-through sites.Flow-through sites provide openings at the upstream and downstream ends of the site. This type of configuration will allow for more larval drift into the floodplain habitat. However, these types of sites may be more prone to deposition at the upstream inlet and in the bottomland. Two of the restored sites have been re-configured during the spring of 2000 to investigate the success of this configuration.Several thousand acres of floodplain habitat exist in the Upper Colorado River Basin. The success of these floodplain restoration activities will have a significant influence on the future of floodplain habitat restoration in the Upper Colorado River Basin.