Over 3,300 unscreened agricultural water diversion pipes line the levees and riverbanks of the Sacramento River (California) watershed, where the threatened Southern Distinct Population Segment of green sturgeon, Acipenser medirostris, spawn. The number of sturgeon drawn into (entrained) and killed by these pipes is greatly unknown. We examined avoidance behaviors and entrainment susceptibility of juvenile green sturgeon (35±0.6 cm mean fork length) to entrainment in a large (>500-kl) outdoor flume with a 0.46-m-diameter water-diversion pipe. Fish entrainment was generally high (range: 26–61%), likely due to a lack of avoidance behavior prior to entering inescapable inflow conditions. We estimated that up to 52% of green sturgeon could be entrained after passing within 1.5 m of an active water-diversion pipe three times. These data suggest that green sturgeon are vulnerable to unscreened water-diversion pipes, and that additional research is needed to determine the potential impacts of entrainment mortality on declining sturgeon populations. Data under various hydraulic conditions also suggest that entrainment-related mortality could be decreased by extracting water at lower diversion rates over longer periods of time, balancing agricultural needs with green sturgeon conservation.

Juvenile Chinook Salmon Oncorhynchus tshawytscha from California's Central Valley must pass thousands of unscreened water diversion pipes during their out‐migration to the Pacific Ocean. The number of fish that become entrained into (drawn through) these diversions at different hydraulic and environmental conditions is currently unknown. We tested the ability of juvenile Chinook Salmon to avoid entrainment into a 0.46‐m‐diameter unscreened water diversion pipe while swimming in a large‐river‐simulation flume. Fish swimming experiments were conducted at 0.15, 0.38, and 0.61 m/s sweeping velocities (simulating the river current) with 0.42 and 0.57 m3/s water diversion rates during the day and at 0.15 and 0.61 m/s with a diversion rate of 0.57 m3/s in turbid water and during the night. The number of fish entrained during day experiments ranged from 0.8% (SE, 0.3) to 8.5% (SE, 0.3). The percentage of pipe passage events resulting in fish entrainment nearly doubled at the 0.57 m3/s water diversion rate (1.7%) compared with that at 0.42 m3/s (0.9%). In clear water conditions during the day, more fish became entrained at the higher water diversion rate (0.57 m3/s) and slower sweeping velocity (0.15 m/s), with fish entrainments starting 38.6 cm (SE, 1.6) from the center of the pipe inlet, where fish experienced an increased velocity gradient and a mean resultant velocity of 0.74 m/s. Fish entrainment was strongly influenced by the number of pipe passages per experiment, rather than by swimming orientation or time spent in the flume. More fish were entrained at the faster sweeping velocity (0.61 m/s) in turbid water during the day and at night, indicating that juvenile Chinook Salmon may use nonvisual guidance (e.g., lateral line system) to avoid water diversions in slower currents. These results help to provide a scientific basis for protecting out‐migrating juvenile Chinook Salmon exposed to unscreened water diversions.

Most of the water diversions on the Sacramento and San Joaquin Rivers (California, United States) and their tributaries are currently unscreened. These unscreened diversions are commonly used for irrigation and are potentially harmful to migrating and resident fishes. A large flume (test section: 18.29 m long, 3.05 m wide and 3.20 m high) was used to investigate the hydraulic fields near an unscreened water diversion under ecologically and hydraulically relevant diversion rates and channel flow characteristics. We investigated all combinations of three diversion rates (0.28, 0.42, and 0.57 m3/s) and three sweeping velocities (0.15, 0.38, and 0.61 m/s), with one additional test at 0.71 m3/s and 0.15 m/s. We measured the three‐dimensional velocity field at seven cross sections near a diversion pipe and constructed regression equations of the observed maximum velocities near the pipe. Because the velocity components in three directions (longitudinal, transverse, and vertical) were significantly greater near the diversion pipe inlet compared with those farther from it, they cannot be neglected in the modeling and design of fish guidance and protection devices for diversion pipes. Our results should be of great value in quantifying the hydraulic fields that are formed around fish guidance devices to design more effective protection for fishes from entrainment into unscreened water‐diversion pipes.