Effects of predation tags on growth and stress response in juvenile rainbow trout Oncorhynchus mykiss

Abstract


Introduction
Biotelemetry technology continues to advance, with a relatively new type of tag, the acid-sensitive predation sensor recently introduced. This tag is designed to report a predation event due to contact with acids in the stomach of a predator [1]. These predation tags give the advantage over other tags in knowing whether the sh has been consumed by a predator whereas traditional tags simply show that the tagged sh is still moving; giving it the illusion that it is still alive [2,3]. Using predation tags can help eliminate the observation bias of seeing a moving tag and assuming that it is the original sh [4].
The ability of acid-sensitive predation tags to detect a predation event has been studied with mixed results. Lennox et al. [3] had a 50% false predation detection rate in laboratory trials and 30% false predation in eld trials. In a eld trial, Daniels et al. [4] had 24 out of 41 tags detect a predation event but could only declare that 5 of the tags were exclusively detected as post-predated. In this same experiment, 9 tags were never detected and 63% of tags signaling predation were only detected at one receiver. In another eld trial, Weinz et al. [5] was able to positively assign 15 out of 19 sh as being predated. The remaining 4 sh had unclear fates. Halfyard et al. [1] had a 94 and 95% success rate of predation tags triggering a predation response in staged predation events Because of the impact of tagging data on sheries management decisions, and the labor and cost associated with tagging, it is essential that the behavior, growth, and physiology of tagged sh be similar to that of untagged conspeci cs [6][7][8]. Also, there is currently, no standard protocol for recovery time post-surgery before stocking into the desired waterbody. The USGS recommends holding salmonoids for 18-36 hours post-surgery to lessen stress levels before transport and stocking [9]. However, stress levels of sh can remain high for over a week post-stocking which increases the chances of mortalities and tag expulsion [10].
There is a large need for information on the potential impact predation sensors may be having on sh physiology and growth. Thus, the objective of this study was to determine how the surgical implantation of dummy predation tags affect stress, growth, and survival of juvenile rainbow trout Oncorhynchus mykiss.

Methods
This experiment occurred at McNenny State Fish Hatchery, Spear sh, South Dakota, U.S.A. using degassed and aerated well water (11°C; total dissolved hardness 360 mg/L CaCO 3 ; alkalinity as CaCO 3 , 210 mg/L; pH 7.6, total dissolved solids 390 mg/L). Five 190-L, ow-through, semi-square tanks were used, with each tank near-fully covered with black corrugated plastic [11]. Each tank contained 20 Shasta strain rainbow trout Oncorhynchus mykiss (mean initial length 105 ± 0.7 mm and weight 13.3 ± 0.2 g, n = 100), with 10 sh receiving a Innovasea V5 (Nova Scotia, Canada) dummy transmitter (0.65 g in air, 12.7 mm length, and 4.3 x 5.73 mm diameter) and 10 control sh. The V5 transmitter is designed so that it can either function as a traditional acoustic transmitter, depth sensor, or predation tag depending on the user's speci cations. This transmitter tag burden (± SE) on the sh was 4.8 ± 0.09%. All sh were fed 1.5mm oating feed (Protec FW Skretting; Tooele, Utah) daily to satiation from vibrating feeders (Sweeny AVF6; Cary, North Carolina). Just prior to the start of the experiment and approximately every week there-after, each sh was measured (total length) to the nearest millimeter and weighed to the nearest 0.1 gram (tagged sh had 0.65 g subtracted from their total weight to account for the tag). The experiment lasted for a total of 30 days.
To surgically implant the dummy tags, each sh was anesthetized to stage IV using tricane methane sulfonate (Tricane-S MS-222, Syndel; Ferndale, Washington) [12]. After anesthetization, control sh were handled and then placed in their respective tank. Experimental sh were placed ventral side upwards in a grooved sponge. A small incision, just large enough to insert the tag, was then made along the midventral line. Iodine-soaked dummy acid sensitive tags were then inserted into the peritoneal cavity. A singular suture (Securocryl Poliglecaprone 25 Mono lament, Riverpoint Medical; Portland, Oregon) was made to close the incision site. After suturing, the tagged sh were placed in their respective tank for recovery.
Glucose and hematocrit data were collected from the common pool of sh initially and then from one tagged and one untagged rainbow trout from each tank at 2, 24, 48, 96, 168 hours, and 30 days after the start of the experiment. To collect the blood sample, sh were euthanized using a lethal dose of tricane methane sulfonate and blood was collected by severing the caudal n. Glucose was recorded using a blood glucose monitor (Accu-Check Guide Me, Roche Diabetic Care; Indianapolis, Indiana). Hematocrit was measured by rst collecting a blood sample in a heparinized microhematocrit capillary tube (Fisher Scienti c; Pittsburg, Pennsylvania) sealed with Critoseal (Oxford Labware; St. Louis, Missouri) and placed in a centrifuge for 10 minutes at 11,500 rpm. The percentage of red blood cells in relation to total blood volume was then recorded. Tanks were checked daily for ejected tags or mortality.
Data were analyzed using the SPSS (24.0) statistical analysis program (IBM, Armonk, New York, USA) with signi cance predetermined at P < 0.05. A repeated measures ANOVA was used to determine if differences occurred over the course of the study for glucose, hematocrit, weight, and length. If a signi cant difference was detected, then a one-way ANOVA on each timepoint was ran as a post-hoc test.
Chi-square was used to determine if there was a signi cant difference in survival.

Results
Survival was signi cantly different between treatments at 100% in the untagged controls compared to 92% in the tagged sh (P = 0.041). All mortality occurred in the rst week. Tag retention was 76%, with two tags lost in second week, three tags lost in the third week, and one tag lost in the fourth week. Mean lengths were signi cantly different between the tagged and untagged sh over the course of the trial (F 2.38, 19.02 = 13.998, P = 0.0001, Fig. 1). Subsequent one-way ANOVA indicated that the control sh were signi cantly longer beginning in the second week and continuing to the end of the experiment. Similarly, untagged sh were also signi cantly heavier over the course of the trial (F 1.36, 10.91 = 19.365, P = 0.001, Fig. 2). Beginning at week 2, the control sh began gaining weight signi cantly faster. Over the course of the trial, the control sh grew 14 ± 1 mm and gained 4.8 ± 0.6 g while the tagged sh grew 6 ± 1 mm and gained 0.9 ± 0.9 g.
Hematocrit levels were signi cantly different between the tagged and untagged sh throughout the course of the trial (F 2.9, 23.19 = 5.360, P = 0.006, Fig. 3). Beginning at 24-hours and extending to the end of the study, control sh had similar hematocrit levels to basal. However, hematocrit decreased up to 50% in the tagged sh. By the end of the trial, tagged sh hematocrit began to increase slightly (30% reduction compared to control) but never attained levels close to basal. Glucose showed no signi cant difference over the course of the trial (F 2.16, 17.29 = 2.601, P = 0.1, Fig. 4).

Discussion
The results of this study displayed the negative short-term impacts of transmitter surgery to juvenile rainbow trout. Fish that underwent surgery grew signi cantly less, had reduced survival, and had signi cantly lower hematocrit levels compared to control sh. Growth and behavioral results from acoustic and predation transmitter surgeries have been inconsistent. Urbaniak et al. [13] found that acoustically tagged rainbow trout grew slower for 38 days compared to untagged rainbow trout. Similarly, acoustically tagged juvenile Atlantic salmon Salmo salar have been shown to have reduced growth performance [14,15]. Contrarily, Brown et al. [16] had no difference in growth of sockeye salmon Oncorhynchus nerka but tagged sh did have reduced swimming performance compared to control and sham sh. Smircich and Kelly [17] did not witness a difference in swimming performance but had reduced growth of brook trout Salvelinus fontinalis in their heaviest tag treatment. The inconclusive nature of the results of these studies shows the di culty in determining just how impactful these types of surgeries and tags are on the sh being studied.
Hematocrit is a measure of the capacity of red blood cells to carry oxygen though the body. Thus, a reduction in hematocrit means a reduced ability to effectively function at optimum levels [18]. The untagged sh in this study maintained a 40% hematocrit level throughout the trial, which is well within the normal range of 30-40% [19]. However, the tagged sh were clearly anemic. Hematocrit levels of tagged sh began lowering within 48 hrs after surgery and continued to decrease for 168 hrs (7 days), nally reaching a nearly 50% reduction. By the end of the trial (30 days), hematocrit levels of tagged sh were still approximately 30% lower than initial values and those of the control group. Reduced hematocrit (anemia) in sh is usually associated with infection [20], parasites [21], or toxins in the diet or water [22,23]. Rainbow trout with as little as a 22% reduction in hematocrit have been shown to have signi cant reductions in critical swimming velocity and maximal oxygen uptake [24].
High stress environments should trigger an increase in hematocrit to enhance the blood's ability to carry oxygen under the high energy demand of stress [25]. For example, Fazio et al. [26] found an increase in hematocrit in sea bream faced with multiple acute handling stresses. Smircich and Kelly [17] did not nd an increase in hematocrit with swimming trials in tagged brook trout but also did not have a control group that did not have surgery for comparison or a basal hematocrit reading prior to surgery. The fact that tagged sh in this study had reduced hematocrit levels that acted more like infection suggests that the body treated the tags and sutures as an infection. A long-term 30% reduction in hematocrit would likely impair the ability of the tagged sh to function after release into the wild. Thus, compared to the untagged sh without anemia, the tagged sh would likely be more susceptible to predation and hinder their ability to feed.
Increases in glucose, a secondary stress response, are not as immediate as primary stress markers [27]. As expected, both the tagged and control groups' glucose levels uctuated at similar rates despite the relatively high variance throughout the experiment. The glucose levels observed were relatively lower than the normative values of 108 ± 9.98 mg/dL reported for rainbow trout [28]. However, glucose levels in both the tagged and untagged control sh rose and fell along the same points throughout the experiment. The initial glucose stress response and return to basal levels found in this experiment were similar to other studies also using rainbow trout [29][30][31].
Tag retention, after the exclusion of euthanized sh for glucose and hematocrit data collection, was 19/25 sh, or 76%. This is similar to the 73 and 78% tag retention for hydroacoustic tags in rainbow trout reported by Urbaniak et al. [13] and Kientz et al. [32]. Tags were lost either through expulsion at the incision site or through the nearby skin, both of which involve proliferation of tissue at the site of least resistance [33]. Fish that expelled tags survived for the remainder of the study.
It is unknown if the 5% tag burden used in this study could have affected the results. While Winter [34] initially stated that the tag should not exceed 2% of the sh body weight, subsequent studies have successfully implanted tags well above that level. Lennox et al. [35] showed no effect on migration or behavior of Atlantic salmon subjected to a 5.2% predation tag burden. Brown et al. [36] showed no effect on swimming performance for rainbow trout with a tag burden of 6-to-12%. Salmonids in general have fared well with increased tag-to-body ratios [17,[37][38][39].

Conclusions
The results of this study show the negative impacts of predation tag implantation on juvenile rainbow trout physiology and growth. Of particular concern is the relatively long-term anemia associated with predation tags. While individual sh behavior was not evaluated in this study, based on the negative impacts associated with the implantation and retention of predation tags, assuming identical behavior to untagged conspeci cs may not be correct. Although there is currently no recommended post-surgery holding time, the results of this study indicate that surgical implantation of predation tags should occur at least 30 days, and likely much longer, prior to sh stocking into the wild. If surgeries are performed "in the eld" with sh being released shortly after surgery, it is likely that implanted sh are at a competitive disadvantage compared to un-tagged conspeci cs. Considerable additional controlled research in a closed environment is needed to determine post-surgery recovery times, particularly in relation to tag burden, and investigate techniques to minimize the negative effects of predation tag surgery and retention on sh physiology, growth, and survival.  Mean total length (mm) of rainbow trout Oncorhynchus mykiss subjected to surgical implantation of a dummy acid-sensitive acoustic tag and a control group over a four-week experimental trial. The control group had signi cantly longer mean lengths over the four-week period (F 2.38, 19.02 = 13.998, P = 0.0001).
Means in a week with different letters above are signi cantly different from each other (P = 0.05).

Figure 2
Mean total weight (g) of rainbow trout Oncorhynchus mykiss subjected to surgical implantation of a dummy acid-sensitive acoustic tag and a control group over a four-week experimental trial. The control group had signi cantly greater mean weights over the four-week period (F 1.36, 10.91 = 19.365, P = 0.0001) Means in a week with different letters above are signi cantly different from each other (P = 0.05).

Figure 3
Mean hematocrit levels of rainbow trout Oncorhynchus mykiss subjected to surgical implantation of a dummy acid-sensitive acoustic tag and a control group over a four-week experimental trial. The control group had signi cantly higher hematocrit levels over the four-week period (F 2.9, 23.19 = 5.360, P = 0.006).
Means in a week with different letters above are signi cantly different from each other (P = 0.05).