Scientific Investigations Report 2006-5111

Scientific Investigations Report 2006-5111

Back to Table of Contents

Data Compilation and Analysis

Subbasins, Sampling Reaches, and Water-Quality Sampling Locations

The lower Boise River Basin was divided into four subbasins to assess associations in land use, habitat, and fish community. The divisions were selected using both geomorphic channel features and water-quality aspects of the river. These divisions are described in more detail in MacCoy and Blew (2005). These subbasins include: upstream of Eagle Island to Lucky Peak Dam (upstream of Eagle Island); upstream of the south channel of the Boise River at Eagle Island downstream of West Boise WTF at Eagle Road (upstream of Eagle Road); upstream of Middleton (Middleton); and upstream of mouth (Mouth) (fig. 1). Each subbasin consists of one to three fish sampling reaches and one to two water-quality sampling locations (table 2). Beginning in 1994, the USGS sampled water quality at four lower Boise River sites: Boise River below Diversion Dam 13203510 (Diversion), Boise River above Glenwood Bridge 1320600 (Glenwood), Boise River near Middleton 13210050 (Middleton), and Boise River near Parma 13213000 (Parma) (fig. 1). Diversion and Glenwood are in the upstream of Eagle Island subbasin, Middleton is in the Middleton subbasin, and Parma is in the upstream of Mouth subbasin.

Data-Collection Methods

Habitat, Hydrology, and Water Quality

Data pertaining to recent land and channel features were obtained primarily from biological surveys done between 1994 and 2002 by the USGS (Mullins, 1999a). With the exception of measurements of channel width, most of the data collected were qualitative. Recent measurements of bankfull width were obtained from cross sections of the lower Boise River surveyed in 1997 and 1998 (Hortness and Werner, 1999) and from surveys at biological sampling sites (Mullins, 1999a).

Historic geomorphic data (1867 and 1868) are used to compare fish habitat prior to and following hydrologic modification in the lower Boise River Basin. Land-cover data from cadastral survey notes obtained from the Bureau of Land Management (BLM) for the lower Boise Valley were used to re-create the “historical” lower Boise River. These data were summarized by the Idaho Department of Water Resources (David Blew, written commun., 2003) and in the methods published by MacCoy and Blew (2005). In their notes, surveyors documented slough and meander widths and azimuths, which typically were measured at the mean high-water mark. Therefore, these measurements were used to indicate bankfull width of the river and tributaries. The surveyors also noted land features such as gravel or sand bars.

Flow records were obtained from the National Water Information System web site (NWIS). Historic and recent flow conditions were compared by analyzing discharge data from the USGS gaging station “Boise River at Boise,” (13202000), which has the longest record on the lower Boise River from 1895 to 2002 (U.S. Geological Survey National Water Information System web site, accessed October 1, 2005, at The Indicators of Hydrologic Alteration (IHA) program was used to evaluate the magnitude of change in the natural flow regime following dam construction (The Nature Conservancy, 2001). The magnitude and variation of mean monthly discharges and the average monthly discharges for December and August were summarized for the Boise River at the Boise streamflow gaging station.

Water-quality collection methods and data on the lower Boise River from 1994 to 2002 presented in this report are published in MacCoy (2004). Temperature, dissolved oxygen, pH, and specific conductance data were based on instantaneous readings. Data on suspended sediment and nutrients were based on depth- and width-integrated water samples.

Fish Community

Fish-community data were compiled from studies conducted by the USGS and IDFG between 1974 and 2004 (table 1). U.S. Geological Survey fish sampling reaches were usually located near water-quality sampling locations in the lower Boise River (MacCoy, 2004). Fish communities were assessed by electrofishing a representative reach of river using protocols developed by the USGS National Water Quality Assessment (NAWQA) Program (Meador and others, 1993). Shallow riffle areas were sampled using backpack electrofishing equipment (Smith-Root models 12 and 12A), and deep-water areas were sampled using a drift boat or a pontoon boat carrying a Smith-Root model VI-A and a 5,000-watt, 240-volt generator with either multiple handheld or two bow-mounted electrodes. Netting crews consisted of four to six people and included personnel from IDFG, USGS, and the City of Boise. Usually two electrofishing passes were made through each reach, and an effort was made to sample all representative habitat types. Captured fish were held in livewells until they were processed and released. Fish were identified, counted, measured, weighed, and examined for types and numbers of anomalies. Fish were identified onsite by Dorene MacCoy and Terry Maret, USGS; Don Zoroban, IDEQ; and Dale Allen, IDFG using taxonomic names described in Nelson and others (2004). Voucher samples were taken of selected species, and those samples are in the collection of the Orma J. Smith Museum of Natural History, Albertson College, Caldwell, Idaho. The taxonomy of sculpin (Cottus sp.) and dace (Rhinichthys sp.) was verified by Dr. Carl E. Bond and Dr. Douglas F. Markel, Oregon State University, Corvallis, Oregon, and by Dr. Gordon Haas, University of British Columbia, Vancouver, Canada.

Sampling techniques in November 2003 and August 2004 changed slightly with the use of a raft-mounted electrofisher and techniques described in the U.S. Environmental Protection Agency’s Environmental Monitoring and Assessment Program (EMAP) fish sampling protocols (Peck and others, 2002). Each sampling reach was 40 times the mean channel width, or about 1 mi. A raft-mounted electrofisher was used to collect fish from near-shore habitats while floating downstream through the entire sampling reach. In addition, 100 m of riffle within each reach were sampled using a backpack electrofisher to capture small benthic species often missed by boat electrofishing. Sample data for these studies are summarized on the USGS Web site, accessed June 15, 2006, at Increasing the reach length provided a larger sample of the population but the percentage of composition of each species would be similar.

General locations of IDFG sampling areas were predetermined by either the senior IDFG fishery research biologist or cooperators such as the USGS and the City of Boise. The IDFG used boat-mounted or shore-based electrofishing equipment for all fish community sampling. For the 1974 sampling, boats were powered upstream against the current while electrofishing (Idaho Department of Fish and Game, 1975). Boat-mounted electrofishing equipment using a variable voltage pulsator (0-600 watts DC) was powered by a 2,000 watt portable generator. This equipment was mounted on either an aluminum jet boat or on a smaller boat where maneuverability was limited. All species were counted, game fish were weighed and measured, and a summary of findings was produced (Idaho Department of Fish and Game, 1975).

In 1988, the City of Boise and the USGS worked with the IDFG to choose sampling locations in reaches 2 through 5 (fig.  1; table 2). A three-pass population depletion electrofishing technique (Zippin, 1958) was used at six 200‑m reaches from Diversion Dam to Star. Crews in two drift electrofishing boats sampled from upstream to downstream in each reach using Coffelt model VVP 2-E electrofishers with direct current (600 volt) on a pulse frequency of 120 and a pulse width of 5 (Idaho Department of Fish and Game, 1988). Both game and nongame species were placed in livewells and counted separately. The game fish were weighed and measured for their total length. Nongame fish were counted, and all fish were returned to the river following the third pass. Attempts were made to use block nets at the upstream and downstream ends of each reach, but organic debris and higher streamflows made it difficult to keep the nets in place (Idaho Department of Fish and Game, 1988).

In 1992 and 1994, estimation techniques similar to the 1988 sampling were used (J. Dillon, Idaho Department of Fish and Game, written commun., 2004). In 1994, electrofishing efficiency was improved with new equipment. A new Coffelt VVP-15 electrofisher with a 5,000-watt, shore-based generator and five anodes was used (Idaho Department of Fish and Game, 2000). In 1994, the focus of the sampling was to better characterize the trout and mountain whitefish populations in the river. No nongame species were identified. The 1994 data are mentioned, but not used in statistical summaries in this report due to the lack of nongame species.

Analytical Methods

As a result of the Clean Water Act’s objective to “restore and maintain the physical, chemical, and biological integrity of the Nation’s waters,” there has been a growing focus on the development of biocriteria in State water-quality standards. Increasingly, biological monitoring programs and biocriteria development have expanded to include large rivers. In the United States, the IBI is used by the EPA and many State agencies to assess fish assemblage structure because it serves as an indicator of the history and current health or condition of a stream system (Karr, 1991). The IDEQ has recently published monitoring protocols and an IBI to evaluate large rivers of Idaho using aquatic organisms and habitat measures (Grafe, 2002; Mebane and others, 2003). Zaroban and others (1999) classified Northwest fish species into various attributes to facilitate the evaluation of surface-water resource conditions.

The fish community was evaluated using an IBI (Mebane and others, 2003) that consists of: (1) number of cold water native species; (2) percentage of abundance of sculpin; (3) percentage of cold water species; (4) percentage of sensitive native individuals; (5) percentage of tolerant individuals; (6) number of nonindigenous species; (7) catch per unit effort (CPUE) of cold water fish; (8) percentage of fish with anomalies (deformities, eroded fins, lesions, or tumors); (9) number of trout age classes (determined by length distribution); and (10) percentage of individual species of common carp. Hatchery fish were not included in the IBI calculations. Each of these 10 metrics was standardized and weighted to produce a score ranging from 0 to 100. Within this range, three classifications of biotic integrity can be identified. According to Mebane and others (2003), sites with scores between 75 and 100 exhibit high biotic integrity with minimal disturbance, and they possess an abundant and diverse community of native cold water species (classification = high biotic integrity). Sites with scores between 50 and 74 are of somewhat lower quality. Nonindigenous species occur more frequently, and the community is dominated by cold water, native species (classification = intermediate biotic integrity). Finally, sites with scores less than 50 have poor biotic integrity. In these sites, cold water and sensitive species are rare or absent, and tolerant fish predominate (classification = poor biotic integrity). The relative abundance of each species by site and year, origin (native or introduced), tolerance to pollutants (tolerant, intermediate, or sensitive), and trophic guilds (percentage of invertivores and piscivores, and percentage of omnivores and herbivores) also are summarized.

Selected metrics and IBI scores were summarized for fish community data collected during low-flow periods (November through March) only. The sample effort was similar for each study: two or more netters using at least a 2,000 watt generator to sample at least 0.16 mi of the river (about six times the channel width). Maret and Ott (2003) found that a sample size of greater than 100 represented 85 percent of the species in a reach; therefore, IBI scores were calculated for samples of at least 120 individuals. Individual fish community metrics and IBI scores were compared spatially and temporally in the lower Boise River. They were then compared to one upstream site near Twin Springs (13185000) assumed to be unaffected by urban and agricultural activities, as well as to three least-disturbed sampling sites in southern Idaho (Maret and others, 2001; Terry Maret, U.S. Geological Survey, oral commun., April 2005). The least-disturbed sites were the Henry’s Fork River near St. Anthony (13050500), South Fork Snake River near Heise (13037500), and South Fork Payette River near Lowman (13235000). These sites were sampled as part of the Statewide Water Quality Network. For more information on this network, see the web page at Land-use and water-quality parameters for these sites have been published in Clark (1994), Maret (1997), and Maret and others (2001). The least-disturbed sites were sampled during normal flow years but not always during the same years as the lower Boise River (mean monthly flows and long-term flows for the least-disturbed sites can be accessed at These sites provide a comparison of the best available data from least-disturbed streams in Idaho.

Land-use derived metrics that include area of developed land, area of impervious surface, and number of major diversions calculated for each subbasin were compared with IBI scores. Subbasins were delineated upstream of the downstream end of a fish-sampling reach, and land-use metrics were derived from Geographic Information System (GIS) spatial datasets of the basin. Subbasin boundaries were delineated from 10-m digital elevation model (DEM) spatial data (accessed June 15, 2006, at and visually compared with digital raster graphic (DRG) datasets to detect any delineation errors. Points of diversions within each subbasin were obtained from the Idaho Department of Water Resources (IDWR) web site, accessed June 16, 2006, at, and impervious surface data were obtained from the NOAA-NESDIS National Geophysical Data Center’s web site, accessed June 15, 2006, at Habitat features were summarized from both historic (1867 to 1868) and recent (1994 to 2002) qualitative and quantitative measurements. Spearman rank correlation coefficients (Zar, 1974) were used to determine significant correlation between land use and select water-quality parameters and IBI scores. A Spearman Rank coefficient is considered significant if it is greater than 0.5.

Condition indices were used to determine the ‘health’ or ‘robustness’ of individual fish by comparing length to weight. Mountain whitefish, a native, and the most abundant salmonid in the lower Boise River, was used to compare relative condition of this species to the North American standard weight equation (Rogers and others, 1996) and measurements of mountain whitefish from least-disturbed sites in southern Idaho. Because only a few mountain whitefish were sampled in the downstream reaches of the lower Boise River, mountain whitefish at sites upstream of Eagle Road of a length between 150 and 350 mm were used for comparison. To determine a linear relation between fish length and weight, data were log10 transformed prior to regression analysis. Exponential equations for the length and weight relation used the following equation described in Armour and others (1983):

Equation , (1)



is weight, and


is length.

The equation is a transformation of the log-linear equations where “Equation” is the antilog of the Y-intercept and “Equation” is the slope of the regression line.

Back to Table of Contents

AccessibilityFOIAPrivacyPolicies and Notices

Take Pride in America home page.FirstGov buttonU.S. Department of the Interior | U.S. Geological Survey
Persistent URL:
Page Contact Information: Publications Team
Page Last Modified: Thursday, 01-Dec-2016 19:08:34 EST