Movement behaviors are central to ecology and conservation. Movement sensing technologies can monitor behaviors that are otherwise difficult to observe under field conditions and may enhance the ability to quantify behaviors at the population scale. We monitored steelhead trout (Oncorhynchus mykiss) spawning behaviors in a seminatural enclosure using accelerometer telemetry tags while simultaneously observing behaviors with underwater cameras. Behavioral assignments from visual observations were compared to acceleration histories to develop assignment criteria for acceleration data, including for a key behavior (oviposition). Behavioral events independently classified using acceleration data prior to reviewing video were compared to video scoring and 97% of holding behaviors, 93% of digging behaviors, and 86% of oviposition/covering behaviors were correctly assigned using acceleration data alone. We applied the method to at‐liberty steelhead in spawning tributaries. Acceleration records revealed putative spawning and oviposition in at‐liberty female steelhead, and time budgets for at‐liberty steelhead were similar to those monitored within enclosures. The use of similar movement sensing tags and classification approaches offers a method for monitoring movement behavior, activity budgets, and habitat use in a broad array of aquatic and terrestrial taxa, and may be especially useful when behaviors are cryptic.

Adult summer‐run steelhead Oncorhynchus mykiss overwinter in freshwater for several months prior to spawning. In systems supporting mixed populations of fisheries and conservation importance, understanding the distribution and survival of pre‐ and postspawn fish is necessary for informed management. The upper Columbia River supports hatchery‐origin components valued by anglers, natural‐origin components of conservation concern, and temporary strays from downstream populations. We used radiotelemetry and PIT telemetry to monitor the behavior, distribution, and survival of adult steelhead during fall, overwintering, spawning, and postspawn periods, with a focus on use of the main stem versus four major tributaries. Adult steelhead (N = 807) were tagged at Priest Rapids Dam in 2015 and 2016. One‐fifth of steelhead fell back below Priest Rapids Dam and did not reascend. A slight majority of tagged steelhead that overwintered upstream of Priest Rapids Dam did so in main‐stem reservoirs (54%; N = 548). Overwintering in the main‐stem Columbia River was more likely for later‐arriving steelhead and was concentrated in the upstream‐most reservoir. Winter tributary use was highest in the Wenatchee (26%; 2016) and Methow (18%; 2015) rivers, whereas no steelhead overwintered in the Entiat River. Harvest of hatchery‐origin steelhead was 18% in 2015 and was near zero in 2016, when the fishery was suspended due to low adult returns. After accounting for reported harvest, annual overwinter survival did not differ between main‐stem and tributary habitats, and relatively low adjusted survival of hatchery‐origin steelhead in 2015 suggested unreported harvest. In contrast to low iteroparity rates (<3%), the majority of postspawn steelhead (56.5%) exited tributaries as kelts; kelt survival to Bonneville Dam was 65% in 2016 and 23% in 2017. Collectively, the results highlight the importance of understanding patterns of habitat use and mortality in steelhead populations when managers are faced with balancing harvest and conservation goals.