We monitored 94 mottled sculpins ( Cottus bairdii ) marked with passive integrated transponder (PIT) tags to examine seasonal movement patterns. High detection efficiency (mean ± standard error = 0.76 ± 0.03) of PIT-tagged fish was obtained with a portable antenna. Nighttime searches were more effective than daytime relocations in summer, and smaller fish (<65 mm) were detected more readily than larger fish (≥65 mm). Sculpins used erosional habitats more than depositional habitats, particularly during summer and nighttime. Search attempts within a 150 m reach showed that displacement, mean movement distance, and home range did not differ significantly between summer and winter, suggesting that season did not strongly influence movement. Additionally, movement metrics were unrelated to sculpin size. However, supplementary searches of a larger reach (700 m) showed that 16% of PIT-tagged fish moved more than 100 m over 1 year, with a maximum displacement of 511 m, and mobile fish typically displayed upstream bias. Although most mottled sculpins in our system were relatively sedentary, a small proportion of individuals moved further than previously reported for other systems.

Knowledge of gear bias is critical for conducting valid population and community assessments. We studied the biases in fyke netting by investigating the individual effects of soak time (fyke nets were fished for 6, 24, or 48 h), fish density (fyke nets were stocked with 0, 30, or 60 fish/net), and predators (fyke nets were stocked with one or zero bowfin Amia calva) on the escape probability and number of individuals captured (i.e., catch) for three fish species. Overall, escape probabilities were consistently lower for round gobies Neogobius melanostomus than for bluntnose minnows Pimephales notatus and banded killifish Fundulus diaphanus. Both escape probability and catch increased with soak time. Escape probabilities were lower at high fish densities and in the presence of a predator, whereas catch appeared to be unaffected by both factors. We documented predation on fish stocked in fyke nets by free‐ranging bowfins and yellow bullheads Ameiurus natalis, which is a potential source of bias that will probably vary among systems. Of the factors we investigated, variation in soak time had the strongest effect on catch. Our results were consistent with catch being proportional to soak times of at least 24 to 48 h, although this relationship was highly variable. Thus, standardizing catch by soak time (e.g., net nights) can be appropriate when confronted with low variation in soak time (e.g., 1 d). Finally, our study highlights potential differences in escape probabilities among fish species, a factor that is probably important in determining which species are overrepresented and which are underrepresented in entrapment gear.

Dams can be operated to mimic components of the natural flow regime to minimise impacts on downstream ecosystems. However, infrastructure, societal needs, water management, and catchment runoff constrain which and when flow regime attributes can be mimicked. We compared fish assemblage responses, including native and non‐native species, over 2 decades of managed environmental flows to those in a river retaining a relatively unaltered flow regime. Both of these arid‐land rivers are within the overallocated Colorado River basin and have experienced declines in catchment runoff over the past 20 years. We predicted that fish–flow relationships would be conserved across time and between managed and unmanaged rivers. Declines in flow in both rivers coincided with declines in some native fishes, and more native and non‐native fish species exhibited declines in the managed river than in the unmanaged river. Our ability to detect previously documented native fish–flow relationships diminished in the managed river system because established environmental flow targets were not met due to water management, but we detected these fish–flow relationships in the unmanaged river. Our results suggest declining catchment runoff and increased consumptive water use could reduce the effectiveness of environmental flows that have lower priority in most years.

Bonytail Gila elegans, a large‐bodied cyprinid that is endemic to the Colorado River basin of the American Southwest, was historically widespread and abundant in large warmwater streams but is now critically endangered. To increase recovery prospects, over 500,000 Bonytails have been stocked in the upper Colorado River basin since 2000, but adult survival has been low and reproduction has not been detected. We provide the first documented evidence of successful reproduction by stocked Bonytails in the upper Colorado River basin. Adult Bonytails were stocked in the Green River and accessed Stewart Lake and Johnson Bottom (managed floodplain wetlands in the middle Green River, Utah) during high flows in May 2015 (Stewart Lake only) and 2016. Draining of Stewart Lake in September 2015 revealed 19 age‐0 individuals of Gila sp. (37–64 mm TL) among over 405,000 fish. Four preserved specimens (41–48 mm TL) were verified as Bonytails by using morphological and molecular techniques. Otolith daily increment analysis confirmed reproduction by Bonytails in Stewart Lake. Bonytail reproduction was also noted during 2016 in Stewart Lake (probable) and Johnson Bottom. Young Bonytails survived despite the presence of abundant nonnative fish predators. Use of floodplain wetlands for reproduction may enhance the recovery of critically endangered Bonytail in the upper Colorado River basin. Received September 1, 2016; accepted January 2, 2017 Published online March 15, 2017