Introduction

The Kootenai River White Sturgeon Recovery Team (U.S. Fish and Wildlife Service, 1999; 2000), composed of scientists and engineers from the U.S. Fish and Wildlife Service (USFWS), Kootenai Tribe of Idaho (KTOI), Idaho Department of Fish and Game (IDFG), and several local, State, Federal, and Canadian agencies, is working to understand the physical factors affecting quality and quantity of spawning and rearing habitat used by the endangered white sturgeon (Acipenser transmontanus) and to assess the feasibility of various habitat-enhancement scenarios to re-establish recruitment of white sturgeon. To that end, the U.S. Geological Survey (USGS) developed a two-dimensional flow model of the Kootenai River near Bonners Ferry, Idaho (Barton and others, 2005). The reach used for this model extended from river kilometer (RKM) 245.9 in Bonners Ferry to near Shorty’s Island (RKM 228.4) (figs. 1 and 2). The 2005 model consisted of a straight reach and a meander reach and included all observed white sturgeon spawning sites. The recovery team are using the model simulation results to design and implement ways to improve that habitat. Model simulations have proved useful in linking streamflow velocities and depths to spawning location and other habitat data (Barton and others, 2006a, 2006b, and 2006c; McDonald and others, 2006a).

The KTOI requested that the USGS construct the two-dimensional flow model of a braided reach upstream of the present-day white sturgeon spawning reach (fig. 2). Extending the 2005 modeling effort upstream into the braided reach is important because many scientists consider the braided reach a suitable substrate with adequate streamflow velocities for re-establishing recruitment of white sturgeon. Furthermore, sturgeon have been caught during the spawning season as recently as 1978 at Crossport (fig. 1), near RKM 252 in the braided reach (J.P. Runyan, resident, Bonners Ferry, Idaho, oral commun., 2006). The model of the braided reach was used to quantify physical characteristics of the reach that white sturgeons seldom inhabit. White sturgeon egg collection (fig. 3) and telemetry data indicate that most spawning currently is within the meander reach between RKMs 228 and 240. During 2001 and 2006, spawning also was observed in the straight reach between RKMs 245.0 and 245.8 (Paragamian and others, 2001; 2002). The spawning reach is defined here to range from RKMS 228 to 245.8. However, during the 1999 though 2006 spawning seasons about 10–67 percent of the annual spawning white sturgeon population briefly inhabited the lowermost part of the braided reach (Pete Rust, Idaho Fish and Game, oral commun., 2006). For this reason, the IDFG requested that the USGS also extend the two-dimensional flow model of the spawning reach several kilometers downstream of the white sturgeon spawning reach. This modified model was used to quantify the physical characteristics of a reach that white sturgeon passed through as they swim upstream from Kootenay Lake to the spawning reach.

This report documents two overlapping two-dimensional flow models that include the braided reach, straight reach, and meander reach. When combined, these models form a contiguous 31.8 km model reach from RKM 222.2 to 254.0. The upstream model extends from RKM 242.9 to 254.0, and includes the braided reach, the 2.8 km straight reach, and a 
1.6 km segment of the meander reach. This model is called the braided-straight reach multidimensional model (fig. 2). The second model is a modified version of the two-dimensional flow model developed by Barton and others (2005), which included the meander reach and the straight reach. For this study, the previous model was extended 6.2 km downstream. The modified model extends from RKM 222.2 to 245.9 (fig. 2). In this report, the modified model is referred as the meander-straight reach multidimensional model. The straight reach was included in the meander-straight reach model to improve model simulation in the meander reach. Locating the model boundary upstream of the area of interest improves model behavior for that area.

This report presents the braided-straight reach multidimensional model and documents modifications to the meander-straight reach model not included in Barton and others (2005). This report also presents results of a few simple simulations that can be used to link stream depths and streamflow velocities to biological or other habitat data.

Description of Study Reach

A brief description of the study reach (figs. 1 and 2) and white sturgeon spawning habitat is provided below. Further details are provided in previous USGS reports: Barton (2004); Barton and others (2005); Berenbrock and Bennett (2005); and McDonald and others (2006a).

Kootenay Lake, British Columbia, Canada, creates backwater conditions in the Kootenai River to a reach that extends from the mouth of Deep Creek (RKM 240.25) to several kilometers upstream of Bonners Ferry (fig. 2). The extent of backwater is a function of river flows upstream of the backwater reach and Kootenay Lake levels. During periods of low streamflow, backwater conditions diminish and free-flowing water may extend a few kilometers downstream of RKM 245.9. A detailed analysis of the location of transition between the free-flowing river and backwater under a range of streamflow conditions is presented in Berenbrock (2005).

Based on historical data collected at the Copeland gage (12318500) 1966 to 1971, during the pre-Libby Dam era, the median annual peak spring-summer streamflow was about 2,240 m³/s. The annual peak spring-summer runoff streamflow during the beginning of the Libby Dam era, 1970s through 1980s, was several times less than during the pre-Libby Dam era (Barton, 2003). Studies on Kootenai River white sturgeon during the past few decades indicated a problem in sturgeon recruitment. Since the mid-1990s, the U.S. Fish and Wildlife Service requested that Kootenai River streamflows be increased during white sturgeon spawning in May and June in an attempt to help re-establish recruitment. Kootenai River streamflows are augmented with additional release of water from Libby Dam; however, the median streamflow during spawning season is about one-half that of the pre-Libby Dam era. Recruitment occurs when a spawning event produces juvenile fish that survive to create a new year-class of fish in sufficient numbers to maintain the fish population.

Prior to the 1960s, dikes were constructed on the river’s natural levee throughout the study area for flood protection. The upper part of the study area is a braided reach that courses over gravel and cobbles and some sand. This reach extends downstream to a straight reach near RKM 245.9. The straight reach (Tetra Tech, Inc., 2003) lies between RKMs 244.5 and 245.9 and forms a transition zone between the braided reach and the meander reach. Here, the riverbed consists of gravel, sand, and traces of cobble and rock crop out along the banks in a few areas. The meander reach, downstream of the straight reach, is dominated by sand dunes with minor amounts of lacustrine clay and silt in the thalweg and along the outside bank of meanders. Small, isolated gravel lenses buried by a thin layer of sand also were identified at RKMs 243.8 and 241.6. In a few cases, gravel and cobble are in the riverbed near Myrtle Creek, and at the mouth of Lost Creek and Ball Creek. The source of this coarse-grained riverbed substrate may be an alluvial fan at the base of Cascade Ridge or the sediment that forms the dike along the left bank. Some riprap has rolled down the toe of the dike into the channel (Barton, 2004).