The development of assays for stress in marine fishes is vital for studying the impacts of bycatch in fisheries and for determining the health of fish being cultured or used in research. This research developed behavioral and physiological assays for stress in juvenile sablefish, Anoplopoma fimbria, a species that comprises a valuable North Pacific fishery and is often a substantial part of incidental discard. The effects of conditions, intrinsic or extrinsic to the fish, on the variability of the stress response were also investigated. A moderate stress of 15 minutes in air was used to elicit an acute stress response. Behavioral responses and physiological values were evaluated at 1, 5, and 24 hours after the 15 minute air stress, and were compared with control fish that received only a minimal air stress. In the first series of experiments, behavioral patterns and changes in behavior over time of stressed and control fish were determined, and the protocols and time course for measuring behavioral effects of stress were established. In the second series of experiments, physiological assays were added to the behavioral protocols developed during the first series of experiments. The behavioral assays included activity levels, swimming speed at capture, and appetitive behavioral patterns in response to a chemical food stimulus (squid extract), and to a visual food stimulus (squid without odor). These behaviors are ecologically relevant in terms of performance and survival. All of these behaviors were sensitive to stress. In general, behavioral responses were depressed by stress at 1 hour, followed by recovery to control levels by 5 hours. However, the intensity of behavioral responses was affected by feeding history (1 or 5 days of food deprivation) and group influence (recovering alone or in visual contact with 3 conspecifics), which therefore affected the ability of the behavioral responses to assess stress. The behavioral assays were less capable of detecting differences between stressed and control fish when the responses of control fish were depressed as a consequence ofbeing fed the day before. Visual contact with conspecifics facilitated recovery of activity in stressed fish, but therefore also resulted in apparent activity responses to chemical food stimulus that were more likely attributable to activity increases of conspecifics than to appetitive behavior. The focus of attention of isolated fish on activity of conspecifics often interfered with visual detection of food. The physiological assays included plasma concentrations of cortisol, glucose, and lactate, all of which proved to be sensitive measures of stress in sablefish. These parameters were elevated by stress at 1 hour, followed by a decreasing trend to 5 and 24 hours. The physiological assays were affected by feeding history, and an effect of group influence was also indicated. Cortisol and lactate levels in stressed fish fed the day before recovered faster than for stressed fish that were deprived of food for 5 days. Glucose levels in stressed fish fed the day before were not elevated above controls. These results suggested an alleviating effect of feeding on the biochemical stress response. At 5 hours, cortisol and glucose were elevated above baseline levels in both solitary stressed fish and in stressed fish influenced by a group, but also for controls influenced by a group, suggesting an exacerbating effect of isolated fish being in visual contact with groups. There were critical cortisol, glucose and lactate thresholds (180 ng/ml, 140 mg/dl, and 175 mg/dl, respectively) above which no appetitive behavioral responses occurred. These clear demarcation levels are extremely valuable for linking behavioral and physiological responses. These results indicate that behavioral and physiological assays are sensitive indicators of stress in sablefish, although the magnitude, time courses, and correlation of responses may be affected by factors intrinsic and extrinsic to the fish that may vary before and during recovery. There was a correspondence between behavioral and physiological indications of stress shortly after the stressor had been removed and levels of stress were still severe. However, there was a temporal discrepancy after partial recovery had occurred, at which time recovery of physiological norms had not yet been established although behavioral responses had recovered. While behavioral patterns may readjust quickly, the persistence of an energetic load during recovery from stress, as indicated by continued physiological perturbations, may compromise ability to respond to additional stressors. / Graduation date: 2001