In 2010–11, the U.S. Geological Survey (USGS), in cooperation with the U.S. Fish and Wildlife Service, evaluated the physical characteristics and fish assemblage composition of mapped river mesohabitats at four sites on the Rio Grande/Rio Bravo del Norte (hereinafter Rio Grande) in and near Big Bend National Park, Texas. The four sites used for the river habitat study were colocated with sites where the U.S. Fish and Wildlife Service has implemented an experimental reintroduction of the Rio Grande silvery minnow (Hybognathus amarus), a federally listed endangered species, into part of the historical range of this species. The four sites from upstream to downstream are USGS station 08374340 Rio Grande at Contrabando Canyon near Lajitas, Tex. (hereinafter the Contrabando site), USGS station 290956103363600 Rio Grande at Santa Elena Canyon, Big Bend National Park, Tex. (hereinafter the Santa Elena site), USGS station 291046102573900 Rio Grande near Ranger Station at Rio Grande Village, Tex. (hereinafter the Rio Grande Village site), and USGS station 292354102491100 Rio Grande above Stillwell Crossing near Big Bend National Park, Tex. (hereinafter the Stillwell Crossing site).

In-channel river habitat was mapped at the mesohabitat scale over a range of seasonal streamflows. A late summer (August–September 2010) high-flow regime, an early spring (April–May 2010) intermediate flow regime, and a late spring (May 2011) low-flow regime were the seasonal flows used in the study. River habitat was mapped in the field by using a geographic information system and a Global Positioning System unit to characterize the sites at the mesohabitat scale. Physical characteristics of a subset of mesohabitats in a reach of the Rio Grande at each site were measured during each flow regime and included depth, velocity, type and size of the substrate, and percent embeddedness. Selected water-quality properties (dissolved oxygen, pH, specific conductance, and temperature) of a subset of mesohabitats were also measured. The fish assemblage composition at the four sites was determined during the three flow regimes, and fish were collected by seining in each mesohabitat where physical characteristic data were measured, except during some periods of high flow when electrofishing was done to supplement seining.

The total number and number of types of mesohabitats were larger during low flows compared to intermediate flows, and larger during intermediate flows compared to high flows. Decreases in streamflow typically led to increases in channel complexity in terms of the number of different types and total number of mesohabitats present. The total wetted area increased and the number of mesohabitat types generally decreased as streamflow increased. At all four sites, the smallest depths and velocities were generally measured during low flow and the largest depths and velocities at high flow. Specific conductance was relatively consistent between the Contrabando and Santa Elena sites, the two most upstream sites. Specific conductance decreased appreciably between the Santa Elena site and the Rio Grande Village, and decreased slightly between the Rio Grande Village site and the Stillwell Crossing site. Specific-conductance values within and among mesohabitat types at a given site were relatively consistent. The pH values measured within and among mesohabitat types also were relatively consistent at all four sites. Median dissolved oxygen concentrations were relatively consistent between the Contrabando and Santa Elena sites (8.34 and 8.54 milligrams per liter [mg/L], respectively) but decreased along the stretch of river between the Santa Elena and Rio Grande Village sites to 7.31 mg/L, possibly because of small dissolved oxygen concentrations associated with contributions from springs between the Santa Elena and Rio Grande Village sites. Dissolved oxygen concentrations increased substantially between the Rio Grande Village and Stillwell Crossing sites to 10.06 mg/L. Mesohabitat water temperatures were generally highest in mesohabitats commonly associated with shallow water depths and low velocities (forewaters, backwaters, and embayments).

Of the 21 species of fish collected during the three flow regimes, red shiner (Cyprinella lutrensis) was the most abundant species overall, accounting for about 35 percent of all fish collected. Another minnow, the endemic Tamaulipas shiner (Notropis braytoni), was second in overall abundance. A nonnative species, the common carp (Cyprinus carpio), was the third most abundant species overall. No statistically significant differences in fish-species richness were found among the different mesohabitat types. Median fish-species richness and maximum fish-species richness values were larger, and fish-species richness was more variable in runs, pools, forewaters, and backwaters during low flow compared to the fish-species richness values calculated for intermediate and high flows. Fish density in backwater mesohabitats was significantly different from fish densities in run mesohabitats, but fish densities were not significantly different among the other mesohabitat types.

Of the 39 Rio Grande silvery minnow individuals collected at the four study sites, 21 (more than half) were collected at the Santa Elena site, 12 at the Contrabando site, and 3 each at the Rio Grande Village and Stillwell Crossing sites. Rio Grande silvery minnow fish-species densities followed the same order as abundance of this species at the sites; fish-species densities ranged from 0.95 fish per 100 square meters (m²) at the Santa Elena site to 0.11–0.47 fish per 100 m² at the other three sites. The Rio Grande silvery minnow was most common in pools and runs during low- and intermediate-flow regimes. This species was less commonly collected in backwaters, embayments, and rapids, and none were collected in forewaters or submerged channel bars. The Tamaulipas shiner has similar life-history characteristics compared to the Rio Grande silvery minnow, including similar feeding habits and habitat use. Tamaulipas shiner was most common in backwater, run, and riffle mesohabitats (in decreasing order) during low and intermediate flow and was less common in submerged channel bar, pool, forewater, rapid, and embayment mesohabitats (in decreasing order) during the same flows. The overall relative percent density (composite of all three flow regimes) of Rio Grande silvery minnow was largest in rapid and pool mesohabitats and for Tamaulipas shiner was largest in backwater mesohabitats.

There were no statistically significant differences between the stream velocities associated with seine hauls of the Rio Grande silvery minnow and Tamaulipas shiner. Stream velocities associated with the seine hauls that included Rio Grande silvery minnow indicate that this species is predominantly found in low-velocity mesohabitats. Velocities associated with seine hauls that included the Tamaulipas shiner represented a much broader overall range of velocities than those associated with Rio Grande silvery minnow collections. No statistically significant differences were found between the depths for seine hauls that included Rio Grande silvery minnow or Tamaulipas shiner. The Rio Grande silvery minnow was more commonly collected in seine hauls from mesohabitats dominated by cobble substrates and less frequently collected in mesohabitats with substrates dominated by fine-sized silt and clay particles, gravels, and sands, in that order. In contrast, the Tamaulipas shiner was broadly distributed among mesohabitats characterized as having gravel, cobble, and silt and clay.