Exercise and Structure Improve Juvenile Chinook Salmon Rearing Performance

This experiment evaluated the use of an exercise routine and vertically-sus-pended structure during juvenile landlocked fall Chinook salmon (Oncorhyn-chus tshawytscha; mean ± SE, initial weight 1.47 ± 0.03 g, total length 56.4 ± 0.4 mm) rearing. Four treatments were used: 1) no exercise routine nor verti-cally-suspended structure, 2) exercise and structure, 3) exercise and no structure, and 4) no exercise and structure. Water velocities in tanks without exercise were 12 cm/s, where-as the exercise routine consisted of seven days at 12 cm/s followed by seven days at 18 cm/s. The structure was an array consisting of four vertically-suspended aluminum angles. Total tank gain and percent gain were significantly greater after 50 days in the tanks of salmon subjected to the exercise routine and structure compared to the three other treatments. Gain and percent gain were also significantly greater in the tanks receiving structure without exercise compared to tanks with exercise and no structure or tanks with neither exercise nor structure. Feed conversion ratio was significantly improved in the salmon tanks with structure and without exercise compared to all other treatments. There were no significant differences for individual fish weight, total length, specific growth rate, or condition factor among any of the treatments at the end of the experiment. These results indicate that the use of both an exercise routine and vertically-suspended structure may be beneficial during the rearing of juvenile landlocked fall Chinook salmon.

To retain the inherent self-cleaning of circular tanks and also provide in-tank enrichment, Kientz and Barnes [35] developed vertically-suspended structure.
Kientz and Barnes [35] observed hatchery rearing improvements associated with the vertical-suspension of an aluminum rod array. Subsequent experiments with suspended strings of spheres, aluminum angles, and plastic conduit produced similar benefits during the rearing of salmonids [15] [35]- [41]. However, positive results from vertically-suspended structure were not observed by White et al. [42], Huysman et al. [43], and Jones et al. [44]. Because these structures are vertically-suspended they do not interfere with the self-cleaning nature of circular tanks like the other structures discussed do.
Little research has been done evaluating the potential impacts of using both exercise and structure. These experiments have all used juvenile rainbow trout (Oncorhynchus mykiss) and the results have been inconsistent. Voorhees et al. [45] noted a significant interaction between exercise and vertically-suspended structure on trout rearing performance. In contrast, Voorhees et al. [46] reported no such interaction, stating that each form of environmental enrichment is likely acting independently. This study was undertaken to provide more clarity into the possible combined effects of both an exercise regime and vertically-suspended structure. In addition, juvenile fall Chinook salmon (Oncorhynchus tshawytscha) were used to expand the knowledge base beyond rainbow trout. Four treatments (n = 3) were used: 1) no routine exercise nor vertically-suspended structure (control), 2) both an exercise routine and structure, 3) exercise routine and no structure, and 4) no exercise routine and structure. Water velocities in tanks without exercise were 12 cm/s, where-as the exercise routine consisted of alternating seven days at 12 cm/s followed by seven days at 18 cm/s. Velocities were measured directly behind the spray bar at a depth of 0.2 m (halfway in water column) using a flowmeter (JDC Electronics Flowatch Flowmeter, JDC, Yverdon-les-Bains, Switzerland). Tank velocities were adjusted by adjusting the angle of the spray bars, and water flows were kept constant throughout the study. All tanks, regardless of treatment, were almost fully covered with corrugated plastic as described by Walker et al. [47]. The vertically-suspended structure used was an array of four aluminum angles suspended through the corrugated-plastic covers as described by Krebs et al. [36].

Methods
Feed was projected using the hatchery constant method [48] with an expected feed conversion ratio of 1.1. The feeding rates used were at or slightly above satiation and varied between the exercise and non-exercise treatments [49]. Tanks of fish not subjected to the exercise routine were fed for a projected growth rate of 0.065 cm/day, while exercised fish were fed for a projected growth rate of 0.75 cm/day.
At the end of the experiment ten fish from every tank were individually weighed to the nearest 1.0 g and measured (total length) to the nearest 1.0 mm, and the total biomass of each tank was weighed to the nearest 0.05 kg. The following equations were used: Condition Factor K 10 fish weight fish length = * One-way Analysis of Variance (ANOVA) was used to analyze the data with SPSS (24.0) statistical program (IBM, Armonk, New York, USA). Because the tanks were the experimental units, and not the individual fish, nested ANOVA was conducted on individual fish data. If the ANOVA indicated significant differences, Fisher's Protected Least Significant Difference procedure was used for pair-wise comparisons. Significance was pre-determined at P < 0.05.

Results
Final tank weight, gain, and percent gain were all significantly greater in the tanks of fish receiving both an exercise routine and structure compared to all other treatments (Table 1). Gain and percent gain were also significantly greater in the tanks receiving structure without exercise compared to tanks with exercise and no structure or tanks with neither exercise nor structure. Feed conversion ratio was significantly better in the unexercised tanks with structure (mean ± SE; 1.02 ± 0.01), compared to unexercised and unstructured tanks (1.15 ± 0.03) which was similar to routine exercise with structure (1.18 ± 0.03). The tanks on routine exercise without structure were significantly poorer than all other treatments (1.45 ± 0.01). Percent mortality was relatively minor and not significantly different among the treatments.
There were no significant differences in individual fish length, weight, specific growth rate, or condition factor among any of the treatments ( Table 2). Specific growth rates were similar among the treatments, ranging from 2.90 ± 0.08 (mean ± SE) in the fish receiving the exercise routine without structure to 3.04 ± 0.04 in the fish receiving both exercise and structure.

Discussion
The results of this study indicate that using two forms of environmental enrichment in combination can have positive effects during fish rearing. These results with Chinook salmon support those obtained by Voorhees et al. [45] with rainbow trout subjected to continual exercise. However, Voorhees et al. [46] reported no benefits of combining both exercise and structure during juvenile Table 1. Mean ± SE final tank weight, gain, percent gain, food fed, feed conversion ratio (FCR a ), and percent mortality of Chinook salmon reared with or without vertically-suspended structure and with or without routine exercise. Means with different letters in the same row differ significantly (p < 0.05; n = 3).  Table 2. Individual mean ± SE total length, weight, specific growth rate (SGR a ), and condition factor (K b ) of Chinook salmon reared with or without vertically-suspended structure and with or without routine exercise (n = 3). rainbow trout rearing. The Voorhees et al. [46] study lasted 109 days, and final tank weights were likely excessive. As suggested by Huysman et al. [43], density-dependent growth rates may have hindered the ability to detect significant differences in the Voorhees et al. [46] experiment. Of course, there may be speciesspecific differences in response to environmental enrichment [14]. The intermittent exercise routine used in this study was designed to prevent exercise fatigue [50] [51]. Although a combination of both exercise and structure led to rearing improvements in this study, the lack of effect of exercise alone on Chinook salmon growth is not unexpected. Davison [52] indicated that Chinook salmon do not benefit from exercise and Parker and Barnes [53] reported similar growth between exercised and non-exercised Chinook salmon. Kiessling et al. [54] and Gallaugher et al. [12] observed decreased growth in exercised Chinook salmon. In other fish species, exercise has been shown to generally increase growth [ [63]. The velocities used in this experiment correspond to approximately 2 to 3 body lengths/sec at the start and 1.5 to 2 body lengths/sec at the end, both of which are slightly faster than the 0.5 to 1.7 body lengths/sec suggested by McKenzie et al. [64]. However, these velocities were taken prior to structure and fish being added to tanks. Vertically-suspended structure radically changes and typically decreases circular tank water velocities [65] [66] [67], and the presence of fish alter velocity profiles as well [67] [68]. It is possible that the positive effect on gain observed in the tanks receiving both exercise and structure, in comparison to the lack of effect of just exercise, could be due to the water velocity changes caused by the aluminum angle array [35].
Similar to the results observed in this study, salmonid growth has been improved using a variety of vertically-suspended structures in numerous experiments [15] [35]- [41]. Rosburg et al. [41] also reported increased growth in similarly-sized landlocked fall Chinook salmon reared with vertically-suspended structure. However, no effects on salmonid growth from suspended structure were observed by White et al. [13], Huysman et al. [53], and Jones et al. [44].
The relatively poor feed conversion ratios observed in the exercised tanks was likely the result of over-feeding. Feeding rates were elevated in the exercise treatments because Parker and Barnes [49] indicated exercise only improved growth if food rations were increased in comparison to unexercised fish. Overall, the feed conversion ratios seen in this study are similar to those reported in other salmonid experiments evaluating exercise [46]  Huysman et al. [39] reported poorer feed conversion ratios in their exercise and structure studies using landlocked fall Chinook salmon.
The lack of significant differences in individual fish lengths and weights is not surprising given the small sample sizes [74]. In addition, as is typically observed during the rearing of feral landlocked Chinook salmon at McNenny Hatchery, considerable variation in fish size was also evident and likely affected the ability to detect any significant differences in individual fish metrics [75]. The specific growth rate in this study is higher than that observed with exercised Chinook salmon by Kiessling et al. [54]. However, it is lower than the 4.5% reported by Huysman et al. [39] for the same strain of landlocked fall Chinook salmon. The salmon used in the Huysman et al. [39] study were smaller than those used in this study, and specific growth rate is affected by the size of the fish [76]. Huysman et al. [39] also reported a condition factor 20% lower than that reported in this study and by other authors examining Chinook salmon during rearing with exercise [8] [53] [54] or structure [40] [42] [70].

Conclusion
In conclusion, this study indicates that the use of both an exercise routine and vertically-suspended structure may be beneficial during the rearing of juvenile landlocked fall Chinook salmon. The exact mechanism where-by both exercise and structure are beneficial, while exercise alone is not, has yet to be determined.
Additional research should be conducted on other possible benefits resulting from the combination of exercise and structure during the hatchery rearing of Chinook salmon, particularly with the positive effects of environmental enrichment on Chinook salmon stress reduction and post-stocking survival reported by Cogliati et al. [31] and Fast et al. [77]. There is also still a need to determine the ideal swimming speeds and exercise routines for the different sizes of nearly all fish species.