The Pecos River between Sumner Dam and Brantley Reservoir (320 km), New Mexico, contains populations of several pelagic-spawning cyprinid fish species, including the federally threatened Pecos bluntnose shiner (Notropis simus pecosensis). Females of this reproductive guild release semi-buoyant eggs that are fertilized and drift, incubate, and hatch on increased flows associated with spring runoff, summer rainstorms, or irrigation releases from Sumner Reservoir. We used artificial eggs (beads) to estimate transport and retention of natural eggs during an irrigation release in five reaches of the Pecos River occupied by bluntnose shiner. Bead retention was highest (1.7-4.5%/km) in the broad complex reaches of upper critical habitat and the quality section (upper 200 km), and bead retention was lowest (0.6%/km) in the narrow, deep, channelized reach of lower critical habitat (60 km). Only 10% of beads were transported more than 50 km from their release location in the reach with highest retention. Modeled spawning events in all five reaches showed highest egg retention of 35% in the reach that also had the highest density of adult Pecos bluntnose shiner (37 km from Cortez Gasline to Acme). The model predicted that during an irrigation release, about 90% of bluntnose shiner eggs produced during a natural spawning event are retained in the upper 250 km of occupied habitat, with only 10% passing downstream into unsuitable habitat or into Brantley Reservoir. We conclude that retention of drifting eggs is greatest during high flows when flood wave attenuation and channel storage are greatest and in broad complex channel reaches, which coincide with existing population centers.

Long‐distance drift of eggs and larvae has been identified as a possible cause of downstream displacement and poor recruitment of the endangered Rio Grande silvery minnow (Hybognathus amarus; silvery minnow). Seven experiments were conducted using artificial eggs to estimate silvery minnow egg drift and retention in the Albuquerque and Isleta reaches of the regulated Middle Rio Grande, New Mexico, USA over a range of flows during expected spawning times. Bead retention varied by reach, discharge, and shape of the hydrograph. Highest retention (6.9 and 9.7% per km in the Albuquerque and Isleta reaches, respectively) occurred on the ascending limb of a high flow in areas where there was substantial floodplain inundation. Retention was maximized at different flows in each reach (97 and 140 m3/s, respectively), possibly associated with reach‐specific floodplain inundation thresholds. Lowest retention in each reach (2.1 and 1.7%, respectively) occurred on the descending limb of low and high flows, respectively. Of the silvery minnow eggs produced in the combined Albuquerque and Isleta reaches in 2005, 8–14% are predicted to have been retained in the Albuquerque Reach (67 km) and 49–83% in the Isleta Reach (86 km) based on the distribution of adult fish and measured bead retention rates. Although silvery minnow propagules are capable of drifting long distances, our study suggests that considerable retention occurs in the Middle Rio Grande. Habitat restoration to increase channel habitat complexity, and flow management to promote floodplain inundation should help to retain a greater proportion of propagules in upstream reaches. Copyright © 2010 John Wiley & Sons, Ltd.

We quantified the availability and utilization of habitat types by eight small-bodied cyprinid fish species, including the federally threatened Pecos bluntnose shiner Notropis simus pecosensis, in the Pecos River, New Mexico. The Pecos River is a medium-sized, sand-bed river with a highly variable hydrograph and some reaches characterized by historic and recent periods of flow intermittency. Fish habitat was described in four reaches at flows of 0.05-2.29 m 3 /s by means of a simultaneous, coordinated program of channel topographic surveys, mesohabitat mapping, fish sampling, and measurement of microhabitat and chemical parameters. We determined that fish habitat in this sand-bed river was effectively described at the mesohabitat scale. Furthermore, mesohabitats were visually distinguishable and separable by means of hydraulic and geomorphic parameters and the relationship between flow and depth, velocity, and velocity : depth ratios. The eight cyprinid species in the Pecos River primarily used four mesohabitat types that were consistently available at the full range of flows examined, including perpendicular and parallel plunges, backwaters, and slackwaters. In winter, five species (red shiner Cyprinella lutrensis, plains killifish Fundulus zebrinus, plains minnow Hybognathus placitus, Arkansas River shiner N. girardi, and sand shiner N. stramineus) shifted to backwaters, where water temperatures were relatively warmer than those of the main channel. Runs, flats, and riffles were generally the most available mesohabitat types, but their availability changed with flow; only four species (red shiner, plains killifish, speckled chub Macrhybopsis aestivalis, and sand shiner) selected these types, and they did so infrequently in summer. The Pecos bluntnose shiner was a habitat specialist that selected primarily rare plunge pools at sandbar faces, whereas species such as the red shiner were habitat generalists that used a variety of mesohabitat types. We found that a range of base flows sustained the preferred habitats of the eight cyprinid species and that no single flow maximized this preferred habitat because of a scaling effect of flow on sand bedforms.

Multiple‐pass removal by use of small‐mesh seines within enclosed areas was performed to estimate numbers of nine small‐bodied fish species at 17 sites in the Pecos River, New Mexico, during October 2007. Site‐level population estimates were most precise for age‐0 red shiners Cyprinella lutrensis (coefficient of variation [CV, calculated as SE/mean] = 0.02–0.06) and least precise for age‐1 and older plains minnow Hybognathus placitus (CV = 0.03–0.42). Site estimates were expanded to derive species‐ and age‐specific population estimates for 284 km of river, including the full range of the threatened Pecos bluntnose shiner Notropis simus pecosensis (101,131 fish; 95% confidence interval = 76,437–125,825; CV = 0.12). Population estimates and species detection from multiple‐pass removal were compared with a closed, comprehensive, single‐pass catch rate index obtained on the first pass of the removal estimate (comprehensive catch per effort [C‐CPE]) and with an open, single‐pass catch rate index (single‐pass catch per effort [S‐CPE]), where a set of representative mesohabitats was seined in proportion to their availability (mean = 12.6 hauls/site). Compared with removal estimates, C‐CPE provided a reliable index of population size for the nine small‐bodied species combined (r2 = 0.90, P = 0.010) but did not perform as reliably for adult Pecos bluntnose shiners (r2 = 0.51) or age‐0 speckled chub Macrhybopsis aestivalis (r2 = 0.70). On average, C‐CPE detected 89% of species at a site, missing predatory species most frequently. By contrast, S‐CPE was a poor index of population size for all species (r2 = 0.16, P = 0.010) and detected an average of 52% of species at a site. The S‐CPE failed to detect the Rio Grande shiner N. jemezanus at 28.5% of sites, the sand shiner N. stramineus at 41.2% of sites, the Pecos bluntnose shiner at 29.4% of sites, and the speckled chub at 23.5% of sites; all of these are small‐boded native species. For applications that require reliable species detection and precise abundance estimates of small‐bodied fishes, the multiple‐pass removal method is recommended.

BioOne Complete (complete.BioOne.org) is a full-text database of 200 subscribed and open-access titles in the biological, ecological, and environmental sciences published by nonprofit societies, associations, museums, institutions, and presses.