Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You and Your Team.

Learn More →

Heart rate and swimming activity as indicators of post-surgical recovery time of Atlantic salmon (Salmo salar)

Heart rate and swimming activity as indicators of post-surgical recovery time of Atlantic salmon... Background: Fish telemetry using electronic transmitter or data storage tags has become a common method for studying free-swimming fish both in the wild and in aquaculture. However, fish used in telemetry studies must be handled, anaesthetised and often subjected to surgical procedures to be equipped with tags, processes that will shift the fish from their normal physiological and behavioural states. In many projects, information is needed on when the fish has recovered after handling and tagging so that only the data recorded after the fish has fully recovered are used in analyses. We aimed to establish recovery times of adult Atlantic salmon (Salmo salar) after an intraperitoneal tag- ging procedure featuring handling, anaesthesia and surgery. Results: Based on ECG and accelerometer data collected with telemetry from nine individual Atlantic salmon during the first period after tagging, we found that heart rate was initially elevated in all fish and that it took an average of ≈ 4 days and a maximum of 6 days for heart rate to return to an assumed baseline level. One activity tag showed no consistent decline in activity, and two others did not show strong evidence of complete recovery by the end of the experiment: baseline levels of the remaining tags were on average reached after ≈ 3.3 days. Conclusion: Our findings showed that the Atlantic salmon used in this study required an average of ≈ 4 days, with a maximum interval of 6 days, of recovery after tagging before tag data could be considered valid. Moreover, the differences between recovery times for heart rate and activity imply that recovery time recommendations should be developed based on a combination of indicators and not just on e.g. behavioural observations. Keywords: Fish telemetry/biologging, Atlantic salmon, Post-tagging recovery, Heart rate, Swimming activity Background tags (DSTs) that store data in internal storage mediums Fish telemetry/biologging is a method of monitoring free- accessible only after the fish (and tag) has been recap - swimming fish where individual animals are equipped tured [2]. Irrespective of tag type, most studies using such with electronic tags that often contain sensors for collect- methods aim to assess the status of wild fish in ecologi - ing data on the conditions within or near the fish [1, 2]. cal settings (e.g. [4, 5]), to evaluate how fish communities Such tags may either be transmitter tags transferring data respond to man-made structures (e.g. [6]), or as a tool to wirelessly to the user (see [3] for details on the structure provide knowledge for fisheries management (reviewed of an electronic transmitter tag) or data storage/archival by [7]). The interest in using this approach in aquaculture is also increasing, both because ongoing technological advances are rapidly expanding the possibilities [8], and because new production philosophies such as Precision *Correspondence: martin.fore@ntnu.no Department of Engineering Cybernetics, Norwegian University Fish Farming promote monitoring at an individual level of Science and Technology, 7491 Trondheim, Norway [9]. Example uses of telemetry/biologging in aquaculture Full list of author information is available at the end of the article © The Author(s) 2021. This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creat iveco mmons .org/licen ses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creat iveco mmons .org/publi cdoma in/ zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Føre et al. Anim Biotelemetry (2021) 9:3 Page 2 of 13 include studies to assess fish responses during welfare- common indicators of metabolic effects due to stress is critical operations such as crowding (e.g. [10]) and the increase in plasma glucose concentration [22]. Such transport (e.g. [11]), and responses to environmental var- changes have recently been shown to lead to increased iability such as temperature variations (e.g. [12]). heart rates also in fish [27]. Other studies have aimed In animal monitoring, it is essential to ensure that to evaluate post-surgery recovery by comparing the the observed animals are representative of the targeted behaviour of the tagged fish to their behaviour before population. When using telemetry, the fish selected for surgery or in untagged cohabitant fish. This method tagging must therefore be representative both before has for instance been applied in laboratory experiments and after the tags are deployed. Ideally, this means that with tilapia (Tilapia sp.) who appeared fully recovered the selection of fish should be truly random and repre - 24  h post-surgery after displaying loss of equilibrium sentative and that the tags do not influence physiology or and reduced swimming activity and feeding just after behaviour in such a way that the tagged fish differ signifi - tagging [17]. Swimming activity was then assessed by cantly from untagged fish (e.g. [13]). In addition, tagging measuring the posture of the fish, and presented as the procedures include several steps (e.g. handling, anaes- percentage of the time the fish was resting (assuming thesia and surgical procedures) that may induce stress, an oblique angle with the snout towards the surface) that in turn may lead to physiological and/or behav- or actively swimming (horizontal orientation or snout ioural changes in the fish [2, 14–17]. Acute (short term) pointing toward the bottom). followed by chronic (long term) stress in farmed fish Recovery after tagging may also be studied with sen- may lead to undesirable effects such as reduced disease sor telemetry. The information conveyed by the tag must resistance, reduced growth rates, impaired health, and then reflect the state of the fish, and typical sensor val - increased mortality [18–21]. Stress responses in fish are ues for unstressed fish should be available as a baseline described by primary responses that include the release for comparison. Previous studies using this approach of stress hormones such as catecholamines and corti- include using heart rate tags to compare tagging methods sol into the circulation system, followed by secondary for black cod (Paranotothenia angustata, [14]), and more responses such as changes in glucose levels, electrolyte recently to study post-surgery stress-responses [28] and balance and heart rate and, finally tertiary (whole ani - potential effects of antibiotics on post-surgical recovery mal) responses. If the fish is unable to acclimate to the [29] in rainbow trout (O. mykiss). While Brijs et  al. [28] stressor at this stage, effects such as behavioural changes, implied a recovery from surgical implantation > 72  h, decreased reproductive capacity and growth may occur, Hjelmstedt et  al. [29] demonstrated a decrease in heart sometimes even resulting in that the animal dies (see [22] rate to within baseline levels 72–96  h after anaesthe- and references therein). If such changes are chronic, the sia and surgery. Other sensor measurements that could fish cannot be considered representative of the popula - potentially be used in this way include tri-axial accel- tion and should be excluded from further analyses [1, 23]. erometers, as previous studies have identified links Conversely, if the changes are transient, the fish may be between accelerometer-based activity proxies that are considered fully recovered once the response patterns particularly sensitive to tail beat frequency and amplitude return to those expected from an untagged fish. This and orientation changes, and stress in salmon [10, 30]. means that tagged fish can be used in analyses if the data Although Atlantic salmon (Salmo salar) has been fre- from the period of recovery are excluded. However, this quently studied using telemetry, there is still a lack of also raises the question: how can we define when a fish is detailed quantitative information on the post-surgery properly recovered after a tagging procedure? recovery of this species. We, therefore, sought to identify Jepsen et  al. [16] sought to identify the duration of the recovery time of Atlantic salmon after intraperitoneal post-surgery recovery for Chinook salmon (Oncorhyn- tagging. This was done using heart rate and acceleration chus tshawytscha) by studying changes in commonly data collected using intraperitoneally implanted elec- used blood indicators of the primary (cortisol) and tronic tags, meaning that data could be collected with- secondary (glucose and lactate) stress responses in tel- out introducing the additional handling stress that would eosts. The authors found that all measured parameters accompany other methods such as blood sampling. The decreased from initially elevated levels to within nor- parameters were chosen because they have previously mal ranges within 7  days post-surgery, with glucose been found to be linked with stress (e.g. [10, 28, 31]) and and lactate (substrate and by-product, respectively, of welfare [32] in salmonids and are commercially available elevated anaerobic metabolism) normalising during in archival and telemetry tags. The data were collected the first 24  h, a recovery time resembling that seen in in a controlled experiment in tanks studying how stress several studies (e.g. [24, 25]). Coping with stress is also responses in Atlantic salmon can be measured using an energy-demanding process [26] and one of the most state-of-the-art technology. The stress response part of F øre et al. Anim Biotelemetry (2021) 9:3 Page 3 of 13 this experiment is described in greater detail by Svendsen 600-60A 7  mm pellets). The fish were not subjected to et al. [27]. any fasting during the experiment period. Biotelemetry/logging systems and surgical procedures Materials and methods All 12 tagged fish (Table  1) were equipped with one Experimental site and fish of three different types of heart rate monitoring Data The experiments were conducted at the NINA Ims Storage Tags (DSTs, Star Oddi Ltd.): 4 × DST milli- Research Station near Stavanger, Norway, between HRT (39.5 × 13  mm, 11.8  g in air); 4 × DST centi- January and March 2019, using 60 hatchery-reared HRT (46 × 15  mm, 19  g); 4 × DST centi-HRT ACT adult Atlantic salmon of the Aqua Gen strain (mean (46 × 15 mm, 19 g). Using different DST types rather than 55.5 ± stdev 5.7  cm fork length, mean weight 2100  g). equipping all fish with the same tag types allowed us to The experiment started on January 28th by stocking also investigate whether all three tag varieties would be four square tanks (tank 1–4, 215 cm side, 122 cm depth, suitable for experiments with Atlantic salmon, which is 5600 L) with seven fish each. The fish were then allowed relevant because this is one of the first applications of this to habituate to the tanks for a period of 21  days until technology on this species. Furthermore, since all three February 18th when three fish in each of tanks 1–4 were tag types were from the same provider, contained the selected at random and equipped with tags, resulting in same type of heart rate sensor and comparable sampling 12 tagged fish in total (Table 1). frequencies (80  Hz over 7.5  s per HR sample point for The tanks were set up with flow-through configura - the centi tags and 100 Hz over 15 s per HR sample point tion, with filtered fresh water from the nearby Imsa river for the milli tags), and applied the same post-processing mixed with small amounts (3–6  ppt, average 5  ppt) of methods to the resulting data, they provided heart rate seawater supplied from seawater inlets at 30  m depth data sets that were comparable among tags. The milli- to ensure a stable and homogeneous water quality and HRT type was set with a higher sample storage interval avoid the introduction of parasites and pathogens to the (10  min) than the others (5  min) as they used more of tanks. Consequently, tank water properties followed the their internal storage medium for raw ECG traces. All ambient conditions in the river, temperatures increas- data were timestamped using the tag internal clocks to ing from 3.9 to 5.0 °C and with DO varying between 93.8 facilitate comparison, and eventual clock drift between and 101.2% between the start and end of the experiment individual clocks was negligible compared to the time (March 15th). Oxygen sensors and oxygenation were also scale of the experiment. One tag type (DST centi-HRT used to prevent unfavourable DO levels. The fish were ACT) also measured activity using an embedded tri-axial fed once per day between 08:00 and 10:00 in the morn- accelerometer (1 Hz sampling rate). ing throughout the entire experimental period, with each In addition to the DSTs that were applied, a total of 4 −1 meal consisting of 2  dL  tank (Skretting Røye Vitalis tagged fish (two fish each from tanks 1 and 2, Table  1) Table 1 Information about the tagged individuals used in the study Fish # Body length Est. body Tank DST type Acoustic tag Time Time surgery Sex Mature (cm) weight (kg) anaesthesia (mm:ss) (mm:ss) F1 59.5 2.3 1 centi HRT X 08:00 08:03 F F2 52 1.6 1 centi HRT X 08:20 08:12 F F3 57 2.1 1 milli HRT 08:45 06:31 F F4 63 2.8 2 centi HRT ACT X 08:06 06:30 M x F5 53 1.7 2 centi HRT ACT X 08:30 06:50 F x F6 53 1.7 2 milli HRT 07:30 06:00 M F7 74 4.5 4 centi HRT ACT 07:40 07:00 F F8 67 3.3 4 centi HRT ACT 08:00 06:30 M x F9 55 1.9 4 milli HRT 08:10 08:00 M F10 62.5 2.7 3 centi HRT 05:30 07:00 M x F11 61.5 2.6 3 centi HRT 07:00 06:30 F F12 55 1.9 3 milli HRT 07:00 06:50 F x Individual body weight was not measured and was hence estimated using an allometric model (W = aL ) with a = 0.0142 and b = 2.9401 Føre et al. Anim Biotelemetry (2021) 9:3 Page 4 of 13 experiment produced 12 data sets on heart rate, and 8 were fitted with acoustic tags (A MP-9, 24.4 × 9  mm, data sets on swimming activity. With mean fish weight 3.6 g; Thelma Biotel AS) that contained tri-axial acceler - being 2100 g and a maximum total tag weight carried by ometers (5  Hz sampling rate) and transmitted an activ- an individual at 22.6  g (DST centi-HRT + A MP-9) the ity proxy derived from the accelerometer measurements tag vs. fish weight ratio of all fish were well within the every 40  s. These tags compute the proxy by first high informal rule of thumb of 2% for maximum tag mass rela- pass filtering the accelerations from all three axes using tive to fish mass [2]. a cutoff frequency of 0.2  Hz to remove low-frequency Each tag implantation was started by capturing a ran- acceleration components due to gravity and body orien- dom fish from an experiment tank using a knotless dip tation. The remaining high-frequency components then net and immediately transferring it to an anaesthetic bath mainly contain accelerations caused by features related (Benzoak Vet, 70  mg/L) where the fish was kept until it to bodily movement that are of interest when evaluating lost its equilibrium and stage III anaesthesia [33] was activity levels, such as tail beats (frequency and ampli- reached (average time 7.7  min). The fish was then care - tude) and rapid changes in attitude/orientation. The fully placed with its ventral side up on a specialised sur- Euclidian norm of the three high pass filtered acceler - gical table with a v-shaped mid-section designed such ometer axes is then computed to yield the magnitude of that the head of the fish was immersed in water through - the total high pass filtered 3D acceleration sensed by the out the whole procedure. A hose circulating anaesthetic accelerometer. Although Føre et  al. [10] used the same −2 (Benzoak Vet, 35  mg/L) through the orobranchial cavity activity proxy with a maximum value of 3.465 m s , we −2 of the fish was inserted into its mouth and the head was chose to limit the proxy to 0–2.1  m  s in our study as covered by a moist cloth (Fig. 1). this gave us a higher resolution and hence precision for A 2–3  cm incision was made along the sagittal plane the activity measures. Moreover, Føre et al. [10] observed −2 starting slightly more than one tag length (i.e. the length very few activity values above 2 m s in Atlantic salmon of the tag to be implanted) posterior from the transverse during stressing, implying that using a lower range would pericardial septum. not compromise the ability to capture the dynamics A finger was inserted through the incision to locate associated with salmon swimming activity. To be com- the transverse pericardial septum. While retaining the parable with the data from the acoustic tags, the activ- finger inside the peritoneal cavity for support, a needle ity data from the centi-HRT ACT DSTs were analysed was positioned in the skin just posterior to the trans- similarly by applying filtering and computing the Euclid - verse septum and slightly laterally from the sagittal plane. ian norm as explained for the acoustic tags (see [27] for The finger was withdrawn, and a smooth plastic spoon more details). Adding the acoustic tags thus allowed us inserted through the incision until it was just below the to compare their activity proxies with those based on the needle insertion point. The needle was then pushed acceleration data from the DSTs and resulted in that the Fig. 1 Fish in the surgical table with anaesthetic circulation tube and head cover, indicating approximate locations of DSTs (white tag) and acoustic tags (black tag) after implantation F øre et al. Anim Biotelemetry (2021) 9:3 Page 5 of 13 through the peritoneal wall while simultaneously with- above which samples are rejected. The MAD decision drawing the spoon to extract the needle out through the criterion typically ranges from 2 (poorly conservative) incision while protecting the viscera. One end of a suture to 3 (very conservative). In this study, the choice of 3 is threaded through the end of the tag was inserted into the justified by the measured heart ranges compared to typ - tip of the needle. The needle was then withdrawn to pull ical heart rates published in the literature (15 < HR < 80) the suture out through the needle’s entry point. This pro - for Atlantic salmon and comparable species [28, 36]. cedure was then repeated on the other side of the sagit- Activity data from the DST centi-HRT ACT tags were tal plane. The tag was then inserted through the incision downloaded as raw acceleration values along all three and anchored anteriorly in the peritoneal cavity using the axes and then subjected to similar post processing as suture and an (external) surgical knot. For the four fish that used to compute the activity proxy in the A MP-9 also equipped with separate acoustic tags, the second acoustic transmitter tags to yield a comparable measure tag was inserted into the peritoneal cavity through the of activity between the two tag types. same incision. Finally, the incision was closed using inter- In a non-decomposed time-series, circadian vari- rupted sutures. The fish was then transferred to a recov - ation (that between day and night) and irregular vari- ery tank with circulating seawater where it was kept until ation (that other than circadian of long-term) had the it regained consciousness, upon which it was transferred potential to obscure long-term trends in heart rate and back into the tank it was collected from. See Table  1 for activity. Time-series of heart rate and activity were anaesthesia bath and surgery durations for all tagged fish. therefore first decomposed into circadian, long-term trend, and irregular components. Decomposition, and subsequent removal of the circadian and irregular com- Timeline and experimental design ponents of the time-series, leaving a long-term com- Since the present study focused on investigating the post- ponent (that showed the long-term growth or decline tagging recovery, the analyses only included data from of the time-series values over the temporal extent of the 2  weeks following tagging. To avoid inducing other the series), allowed for the examination of the form of stress effects that could disturb their recovery, the fish the long-term trends towards recovery. To decompose were sheltered from all potential stress factors except each time-series, it was first binned into 15  min inter - those necessary to feed and provide for the fish in this vals (each 15  min interval showing a mean heart rate period. or intensity over that interval) and then converted into None of the fish exhibited signs of adverse health after a time-series object [R function ts {stats}; Becker et  al. tagging or during the trials, and all fish were euthanised [37]]. Time-series objects were then decomposed using after the conclusion of the experiment. Posthumous the Seasonal Decomposition of Time Series by Loess R pathology of all remaining experimental fish at the end function stl {stats} (B. D. Ripley; Fortran code by Cleve- of the experiment (19 female, 23 male) revealed that land et  al. [38] from “netlib”). Long-term trend com- about one-third of these fish (14 in total, 8 F, 6 M) exhib - ponents were then analysed for a systematic change ited signs of sexual maturation through the experimen- in heart rate or activity that could be indicative of a tal period, including 5 of the tagged individuals (Table 1). post-surgery recovery by first modelling the temporal Although this appeared to have a little direct impact on relationship and then compartmentalising this into pre- the fish in three of the tanks, the data from the fish in one and post-recovery phases. of the tanks (tank 3) were excluded from the statistical The relationship between the long-term trend compo - data analyses due to perpetual inter-individual aggression nent of heart rate or activity (y) and time post-tagging between two matured males in that tank throughout the (t) was modelled using an exponential decay model: experimental period. This left nine fish tagged with DSTs measuring heart rate, six of which also measured activ- −αt y(t) = y + y − y e , p 0 p ity. Since two of these fish contained both a DST and an acoustic tag measuring activity, this resulted in a total of where α defines the decay constant from y (at eight time-series of activity. time zero) to y , the model plateau. Models were fit - ted with the nls {stats} R function (D. M. Bates and S. Data processing and statistics DebRoy: D. M. Gay for the Fortran code used by algo- Heart rate data were used as downloaded from the rithm = "port"), using the self-starting asymptotic DSTs. Outliers were removed using the Median Abso- regression function SSasymp {stats} (J. Pinheiro and D. lute Deviation (MAD) approach [34], using a MAD M. Bates). Most trend components followed an expo- decision criterion of 3, which is a conservative value nentially decaying pattern, ensuring model conver- (see [35]). The MAD decision criterion denotes the gence, but some included parts that were inconsistent standard deviation from the dataset’s sample average Føre et al. Anim Biotelemetry (2021) 9:3 Page 6 of 13 with an exponential decay. First, some tags (three asymptotes in the final 3  days, so it was reasonable to heart-rate tags and four activity tags) showed a short assume that values from these days represented post- initial post-surgery increase in registered values at the recovery signature. Thresholds were established on an beginning of the experiment. Secondly, some tags (one individual basis to allow for post-recovery heart rate or heart rate and two activity tags) showed an increase in activity to change according to individuals. registered values after ≈ 5–6 days. This late increase in activity or heart-rate was likely a result of a separate, Results post-recovery change in behaviour of these individuals. Post‑surgery recovery To ensure model convergence, these parts of the long- Daily heart rate significantly declined from a mean of term trend components were removed prior to model 36.0  bpm (range = 24.6–45.6, SD = 5.6, n = 9) on the day fitting. That is, the exponential model was only fitted of surgery to a mean of 22.3  bpm (range = 17.5–26.6, to parts of the long-term trend component that were SD = 2.6, n = 9) 13 days later (one-sided Wilcoxon’s rank consistent with a post-surgery exponential decline. One test, V = 45, p = 0.002) (Fig. 2a; raw data for all tags shown activity tag (fish F4 in tank 8) did not show an expo - in). Daily activity significantly declined from a mean of −2 nential decline with time and was thus not fitted with 0.57  m  s (range = 0.26–0.92, SD = 0.23, n = 8) on the −2 a model. day of surgery to a mean of 0.33  m  s (range = 0.27– Identification of breakpoints between pre- and post- 0.41, SD = 0.06, n = 8) 13  days later (one-sided Wil- recovery phases was done on an individual basis. The coxon’s rank test, V = 45, p = 0.008) (Fig.  2b). However, breakpoint between pre- and post-recovery for each tag individual variation in activity was high (Fig.  2b). Both was set where the heart rate or activity reached a recov- heart rate and activity displayed circadian variation. ery threshold, defined as the heart rate or activity level Heart rate was greater during daytime (mean = 25.8 bpm, delimiting those pre- and post-recovery. A recovery range = 22.2–26.7, SD = 1.9, n = 9) than during night threshold was defined for each tag as the mean + 2SD of (mean = 22.7  bpm, range = 19.6–24.9, SD = 1.9, n = 9) the long-term trend component values calculated from (Fig.  2a; raw data for all tags shown in Additional file  1: the final 3  days of the fitted series. Inspection of the Figures  S1, S2). In contrast, activity was greater during −2 tags showed that trend components were approaching night (mean = 0.47  m  s , range = 0.32–0.60, SD = 0.11, Fig. 2 Average a heart rate and b activity values (blue line) from all fish in tanks 1, 2 and 4 (tank 3 was excluded because of aggressive behaviour between two males) for the first 2 weeks post tagging. The light-blue envelope shows the range in heart rate values of all individuals F øre et al. Anim Biotelemetry (2021) 9:3 Page 7 of 13 −2 circadian components from the time series (Additional n = 8) than during daytime (mean = 0.34  m  s , file  1: Figures  S5, S6). There was generally greater activ - range = 0.25–0.43, SD = 0.06, n = 8) (Fig.  2b; raw data ity during the evening than during morning (Additional for all tags shown in Additional file  1: Figures  S3, S4). file 1: Figure S6). Circadian differences were present throughout the The heart rate trend component showed a decline experiment. However, the circadian difference in activ - that could be modelled with an exponential decay ity declined throughout the experiment. This circadian function (Fig.  3). However, the trend component still variation was shown by all individuals, as revealed by the Fig. 3 Long term trend in heart rate. Each panel represents data from one individual fish. The continuous black line shows the long-term trend component; the continuous blue line shows the fitted exponential decay model. The dashed horizontal line shows the threshold used to define a structural change; the dashed vertical line shows the breakpoint indicating when the structural change occurs between classified pre- and post-recovery phases Føre et al. Anim Biotelemetry (2021) 9:3 Page 8 of 13 showed considerable temporal variation, depending with an exponential decay (Fig.  4), except for one fish on the tagged individual. For example, the trend com- (fish F8) where an exponential decay model could not ponent for fish F4 showed a sharp decline during the be fitted due to the activity trend component peaking first day after tagging, but this then fluctuated for the ≈ 7 days after tagging. Two fish (fish F1 and F2) showed remainder of the 2-week post-tagging period. The activ - an exponential decline in activity but did not reach a ity trend component also showed a pattern consistent Fig. 4 Long term trend in activity. Each panel represents data from one individual fish. The continuous black line shows the long-term trend component; the continuous blue line shows the fitted exponential decay model. The dashed horizontal line shows the threshold used to define a structural change; the dashed vertical line shows the breakpoint indicating when the structural change occurs between classified pre- and post-recovery phases. It was not possible to fit an exponential decay model to the long-term trend component of fish F8. Thresholds and breakpoints were not established for fish F1 and F2 where the exponential model did not plateau. Tag type (DST = Data Storage Tag; Aco = Acoustic tag) is shown in parentheses following the fish ID F øre et al. Anim Biotelemetry (2021) 9:3 Page 9 of 13 plateau during the study period, suggesting that these individual (F8, Fig. 4), there was a similar trend for activ- −2 fish has not fully recovered in terms of activity. ity: day of anaesthesia and surgery, mean = 0.64  m  s , Time to recovery (as defined by the location of the range = 0.39–0.92, SD = 0.20, n = 7; recovery threshold, −2 breakpoint between pre- and post-recovery phases) var- mean = 0.36  m  s , range = 0.28–0.43 SD = 0.07, n = 7. ied between individuals, and the metric used (heart rate For both heart rate and activity, raw values pre-recovery or activity, Figs.  3, 4, Table  2). The mean threshold value were significantly greater than those post-recovery (one- for heart rate in a ‘recovered’ individual was 23.8  bpm sided Wilcoxon’s signed-rank test: heart rate, V = 45, (range = 21.2–26.0, SD = 1.18, n = 9). The mean time to p = 0.002, n = 9; activity, V = 28, p = 0.008, n = 7). reach this threshold (i.e. breakpoint between pre-recov- ery and post-recovery) was 4.1  days (range = 1.3–5.8, Discussion SD = 1.7, n = 9). The threshold for activity recovery was The current study showed plateauing of most time-series, −2 greater for the acoustic tags (mean = 0.44  m  s , n = 2) indicative of recovery, within the 14  days of the experi- −2 than the DSTs (mean = 0.29  m  s , n = 3), reflecting the ment. Two activity tags, F1(Aco) and F2(Aco), however, higher activity values registered by the acoustic tags. For did not show plateauing, suggesting that the tagged fish the activity tags where there was evidence of recovery, had not fully recovered in terms of activity during this the mean time taken to reach the threshold was similar to period. Other time-series showed gentle gradients even that for the heart rate tags (mean 3.3  days, range = 2.1– after the recovery breakpoint (for example, the F6 heart 5.7, SD = 0.09). For the two individuals that were each rate tag) so the definition of the point of recovery of some tagged with two activity tags, the identified breakpoints individuals as having fully recovered is less robust. How- between the parts of the time series classified as pre- and ever, identified breakpoints generally corresponded with post-recovery depended on the tag: in both individuals, systematic changes in the time-series. For example, the the threshold to reach post-recovery occurred later for breakpoint on the F6 heart rate tag occurred in a trough the acoustic tag than the DST. separating the sharp initial decline over the first 5.75 days Although raw values of mean heart rate on the day of with the gentle gradient afterwards, so it is reasonable anaesthesia and surgery (mean = 36.0 bpm, range = 24.6– to infer that the identified breakpoint corresponded to 45.6, SD = 5.6, n = 9) varied more than the recov- the transition to post-recovery. The modelling approach ery threshold (mean = 23.8  bpm, range = 21.2–26.0, used here allowed for a consistent method for establish- SD = 1.8, n = 9, Table 2), there was a clear declining trend ing the time until recovery among a group of time-series. for all tagged individuals. With the exception of one It should be noted, however, that estimated times until recovery is dependent on modelling approach used. For instance, fitting an exponential model to raw—rather than detrended timeseries, or using a different method Table 2 Recovery based on heart rate and activity sensors to establish a breakpoint between pre- and post-recovery Tag ID Heart rate recovery Activity recovery parts of the time-series, would yield different estimates. Threshold Time (days) Threshold Time (days) The exponential model used in this study is a well-vali - −2 (bpm) (m s ) dated method for modelling physiological recovery [39] but alternative approaches may also be considered (e.g. F1 24.90 4.33 No rec No rec [27]). The sample size of fish in this study was small F2 25.83 5.27 No rec No rec (N = 9); a larger sample size would allow a better quan- F3 23.17 4.42 tification of the range of behaviour during recovery and F4 21.43 2.92 0.31 3.59 allow better selection of the modelling approach. 0.45* 5.08* The heart rate data suggest that the tagged Atlantic F5 25.63 1.89 0.28 2.10 salmon in our study could only be considered fully recov- 0.43* 3.30* ered from the anaesthesia and surgical procedure of F6 21.24 5.75 intraperitoneal tag implantation after an average of ≈ 4 F7 25.97 5.41 0.28 2.60 and up to a maximum of 6 days post-surgery. While some F8 23.68 1.27 No fit No fit studies have indicated longer recovery times post-tagging F9 22.71 5.29 [32], our observations concur with several previous stud- Mean 23.84 4.06 0.35 3.3 ies that have reported similar lengths of recovery post- Activity sensors with a * suffix indicate acoustic tags. “No fit” indicates that the tagging as our study [11, 16, 24, 25, 28]. Although some long-term component of the time-series did not follow an exponential decline and that an exponential model could not be fitted; “No rec” indicates that it data series from the tagged fish in our study may visually was possible to fit an exponential model to the time-series but that recovery appear to continue declining after fulfilling the recovery thresholds and times were not assigned because the fitted exponential model threshold criteria, these changes were not found to be did plateau Føre et al. Anim Biotelemetry (2021) 9:3 Page 10 of 13 statistically significant. Recovery results based on activity accelerometers were found to be comparable to those data varied more in the recovery threshold criteria and measured by the acoustic tags (see [27] for details on this time to recovery than heart rate, suggesting it might be comparison). The lower absolute amplitude of the activ - a less consistent indicator of recovery between individu- ity proxies computed from the DST data was probably als. Moreover, both the temporal patterns and absolute caused by them sampling at a lower frequency (1  Hz) values changed less for activity than heart rate between than the acoustic tags (5  Hz), thereby capturing fewer post-tagging and post-recovery periods, implying a lower high-frequency components. The surgical procedure ratio between the baseline pattern (i.e. circadian varia- used to implant the heart rate tags was much simpler tions) and the changes in activity caused by the tagging than the procedure needed for multivariate implants procedure. Together, these factors suggest that activity recently used in rainbow trout by Brijs et al. [31] but was may be a less consistent indicator of post-tagging recov- more comprehensive and invasive than that used for con- ery than heart rate, and that heart rate might be a gener- ventional intraperitoneal tag placement. It is likely that ally more sensitive indicator than activity, especially for less complex surgical procedures would lead to shorter post-tagging recovery. recovery times in Atlantic salmon, as previously found for It is also important to note that there were individual rainbow trout [42, 43]. However, it is probably reasonable variations in the recovery time assessed from heart rate. to be conservative with respect to recovery times, espe- Although inter-individual variation in recovery time cially if the data are to be used e.g. as a management tool might be an inherent effect one should expect when tag - in aquaculture applications or to evaluate stress effects ging A. salmon, we did find that mature fish had a lower on fish in conjunction with ecological studies. Using data heart-rate recovery time than immature fish. However, from fish that are still recovering from post-anaesthesia/ the low sample size did not provide enough statistical surgery effects in such applications could result in sub- power to robustly test influences on recovery time, so we optimal management decisions or erroneous conclusions recommend further studies with larger sample sizes to that could have ramifications beyond the study itself. increase power in analyses of potential influences. The fish included in the analyses exhibited heart rates Based on these results, we urge caution on using that gradually stabilised at daily means between 21 and telemetry data collected after anaesthesia and surgery 26  bpm (daily variations between 15 and 30  bpm, simi- without first ensuring that the fish are fully recovered lar to that observed by for adult A. salmon of mean fork [1, 23]. Biosensors that measure heart rate and/or activ- length of 62.3 cm at 4 °C by Lucas [36]). Due to the simi- ity can be potent tools in such evaluations, as they pro- larities across tanks and individuals, this range in heart vide quantitative, high-resolution data that will be both rate may be typical for Atlantic salmon of this size and more consistent, precise and objective in capturing the with the prevailing temperatures. Moreover, all indi- full post-anaesthesia/surgery effects than e.g. comparing viduals in tanks 1, 2 and 4 had similar circadian rhythms behavioural observations of tagged vs. untagged fish. (higher heart rates during daytime than at night) and Alternative parameters that could be used to assess gradual post-surgery declines in mean daily heart rate post-tagging recovery in individual fish include blood (from more than 30 bpm after surgery to 21–26 bpm after glucose, lactate or pulse oximetry/ppg. These could pro - up to 6 days). This implies a regularity across individuals vide a more direct assessment of stress levels in salmon, that increases the likelihood that heart rate may func- but we are not aware of any commercial electronic tags tion as a consistent stress indicator in Atlantic salmon able to sense such parameters in live fish. Other tech - that may be used to assess fish recovery after tagging. The niques based on measuring cortisol in faecal matter [40] tagged fish in tank 3 were excluded from the study due or bioelectric field monitoring akin to that used by sharks to inter-individual aggression. These individuals demon - [41] could potentially result in future solutions that could strated measured heart rates that differed from the others be used evaluate recovery in a less invasive and independ- both in individual and aggregate values. Although these ent manner, where the fish are monitored before, during fish also showed signs of circadian variation in heart and after the procedure. However, these methods are still rate, the mean value did not appear to decline over the to be developed to a stage where they can be applied to days following tagging, an effect that was attributed to free-swimming fish, at least in large groups under com - inter-individual aggression, all else being held equal. This mercial production conditions, and would only be able to may indicate that the stress induced by the aggression provide information on a group level. between the two males in this tank overrode the stress All three DST types tested in this experiment appeared response due to recovery. A potential interpretation of to be suitable for applications on Atlantic salmon as all this is that the aggressive encounters caused chronically tagged fish provided valid heart rate data. Moreover, elevated stress levels that masked the recovery stress the activity proxy computed from the DSTs containing caused by handling, anaesthesia and surgery. This could F øre et al. Anim Biotelemetry (2021) 9:3 Page 11 of 13 further mean that recovery stress can be difficult to 2% rule”, [2]). Since we worked with adult salmon with a monitor if the fish are simultaneously influenced by inde - mean weight of 2100 g, and the maximum tag weight car- pendent external events, such as individual interactions ried by the fish was 22.6  g (around 1% of the fish body due to dominance hierarchies [44, 45]. mass) this was not a challenge in our study. Based on established knowledge on how salmon swim- ming speeds are affected by variations in light intensity Future research and potential technological improvements [46], as well as previous telemetry studies applying simi- Since this study only focused on Atlantic salmon exposed lar activity proxies on salmon in sea-cages (e.g. [10]), to one set of environmental conditions, it is difficult to we expected to see a circadian rhythm in activity where assess if these concerns are also relevant for other spe- activity was higher during the day than at night in the cies, and/or fish under different conditions. Similar stud - present study. In contrast to these expectations, the cir- ies on rainbow trout using the same tag type found that cadian trends in the activity of our fish were on average they recovered 72–96  h after surgery [28], which was higher during night-time than during the day. A simi- shorter than the Atlantic salmon in the present study. lar “inverse circadian” rhythm was observed in salmon Moreover, wounds in Atlantic salmon are known to heal reared in fish tanks during the period after tagging by faster in warmer temperatures than in cold water [51], Kolarevic et  al. [30] and could imply that a “normal cir- suggesting that the low water temperatures in the pre- cadian” activity rhythm may arise only after the salmon sent study may have contributed to longer recovery peri- have recovered after tagging in tanks. Conversely, the ods. These elements suggest that species-specific effects circadian rhythm in heart rate was more like expected or differences in external environmental conditions are (higher during daytime), meaning that the fish displayed important to consider when studying recovery times. generally higher heart rates when measured activity was Future studies on the relationship between heart rate and low than when activity was high. This may seem counter- post anaesthesia/surgery recovery time should therefore intuitive as one would expect more active fish to display be conducted for other species of interest, across relevant higher heart rates since salmon tend to display increased temperature ranges, to obtain a more complete picture of heart rates with increased swimming activity [47]. How- this relationship. ever, it is possible that the higher heart rates during day- In the present experiment, the fish were kept in groups time were caused by effects such as feeding activity [48, in small tanks. To investigate how recovery time is 49] or perceived increased predation risk due to higher affected by eventual scaling effects and social/inter-indi - light levels [50]. These results are unexpected and very vidual effects arising due to group dynamics, future stud - interesting, but further extrapolations and discussions on ies addressing post-tagging effects should be done with this matter would probably require further experiments a larger number of tagged fish at larger spatial scales. with more data. This would also enable a deeper scrutiny into individual Although this study underlines the importance of criti- variations in recovery, as a higher number of tagged fish cal evaluation with regards to recovery from anaesthesia would provide a good foundation for finding statistical and surgery when using telemetry, the data collected also relationships on the individual level. Although our pre- highlight the importance of telemetry as a method for sent results imply that inter-individual variations are a studying free-swimming fish. The heart rate and activity prominent feature in the recovery time of tagged salmon, values for all tagged fish eventually plateaued, possibly a larger sample number will be necessary to properly indicating that they all recovered from the anaesthe- conclude upon the nature of such variations. To increase sia/surgery, and posthumous pathology revealed no the relevance of a larger follow-up study, it could be done inflammations or other apparent morphological signs in fish cages in the marine environment, perhaps first by of reduced welfare due to the surgical procedures. Even using meso-scale size cages containing fewer fish than a though the low water temperatures during the experi- commercial cage but at similar densities, and then mov- ment may have led to handling and surgery having less ing to full-scale studies to cover all steps in the transition impact on the fish, the tagging procedure used here was from lab to industrial scale. more complex than conventional intraperitoneal tag- ging. It is thus reasonable to conclude that fish carrying Conclusion telemetry tags can be considered representative members The main conclusion from this study is that the Atlan - of the group they were selected from once they are fully tic salmon in these experiments required an average of recovered from anaesthesia and surgery, provided that ≈ 4 and up to a maximum interval of 6 days of recovery they were a representative selection to begin with. How- after anaesthesia and surgery before their heart rates ever, this also requires that the recommendations on ratio returned to assumed baseline routine values. Moreover, between tag size and fish size are not exceeded (e.g. “the although observation of behaviour and/or activity may Føre et al. Anim Biotelemetry (2021) 9:3 Page 12 of 13 Funding alone be insufficient to assess that the fish has physi - This study was funded by the Research Council of Norway (NFR Project ologically recovered, activity measurements indicated Number 280864). similar recovery periods to those based on heart rate, Availability of data and materials although there was a longer maximum period of 10 days. The datasets used and/or analysed during the current study are available from We, therefore, urge caution when using data collected the corresponding author on reasonable request. after surgery and anaesthesia in studies using biologging/ Ethics approval and consent to participate telemetry tags. Assuming that we want all individuals to All fish handling and surgery were made in compliance with the Norwegian be recovered, our study thus implies that only data col- animal welfare act and were approved by the Norwegian Animal Research lected after 6 days recovery time should be used for fur- Authority (Permit No. 18/18431). ther analyses. However, this recommendation would only Consent for publication be applicable to studies featuring Atlantic salmon reared Not applicable. in similar experimental conditions as we used. Since Competing interests recovery time will vary with factors such as fish species, The authors declare that they have no competing interests. water temperature, invasiveness of the surgery, anaesthe- sia time, fish density and physical scale, it is difficult to Author details Department of Engineering Cybernetics, Norwegian University of Science make general recommendations on when one can assume and Technology, 7491 Trondheim, Norway. SINTEF Ocean, 7465 Trondheim, the fish to be recovered from tagging, and the data to Norway. Norwegian Institute for Nature Research, 7485 Trondheim, Norway. be safe for use in biological analyses. However, by con- Department of Animal Environment and Health, Swedish University of Agri- cultural Sciences, 532 31 Skara, Sweden. Department of Biology, Norwegian ducting experiments similar to the present study where University of Science and Technology, 7491 Trondheim, Norway. these parameters are varied, a more complete picture of how we need to account for fish recovery after tagging in Received: 16 April 2020 Accepted: 9 December 2020 telemetry studies may be obtained. Supplementary Information References The online version contains supplementary material available at https ://doi. 1. Cooke SJ, Woodley CM, Eppard MB, Brown RS, Nielsen JL. Advancing org/10.1186/s4031 7-020-00226 -8. the surgical implantation of electronic tags in fish: a gap analysis and research agenda based on a review of trends in intracoelomic tagging Additional file 1: Figure S1. Measured heart rate for Tag IDs F1–F4. The effects studies. Rev Fish Biol Fish. 2011;21(1):127–51. y-axis upper limit has been constrained to 50 bpm. Figure S2. Measured 2. Thorstad EB, Rikardsen AH, Alp A, Økland F. The use of electronic tags in heart rate for Tag IDs F5–F9. The y-axis upper limit has been constrained fish research—an overview of fish telemetry methods. Turk J Fish Aquat to 50 bpm. Figure S3. Measured activity for acoustic tags. Figure S4. Sci. 2013;13(5):881–96. Measured activity for DSTs. The y-axis upper limit has been constrained 3. Føre M, Alfredsen JA, Gronningsater A. Development of two telemetry- −2 to 5 m s . Figure S5. Circadian component of heart rate. Each panel based systems for monitoring the feeding behaviour of Atlantic salmon represents data from one individual fish. Figure S6. Circadian component (Salmo salar L.) in aquaculture sea-cages. Comput Electron Agric. of the activity. Each panel represents data from one individual fish. 2011;76(2):240–51. 4. Taylor MD, Babcock RC, Simpfendorfer CA, Crook DA. Where technology meets ecology: acoustic telemetry in contemporary Australian aquatic Acknowledgements research and management. Mar Freshw Res. 2017;68(8):1397–402. In addition to thanking the Research Council of Norway for financial support 5. Welsh JQ, Bellwood DR. Spatial ecology of the steephead parrotfish (see “Funding”), we would like to thank the crew at NINA’s research station at (Chlorurus microrhinos): an evaluation using acoustic telemetry. Coral Ims, particularly Knut Bergesen, for their effort in the experiments. We would Reefs. 2012;31(1):55–65. also like to thank Asgeir Bjarnason at Star Oddi LTD and Erik Høy at Thelma 6. Cooke SJ, Bunt CM, Schreer JF. Understanding fish behavior, distribution, Biotel AS for collaboration in specifying the loggers and transmitter tags used and survival in thermal effluents using fixed telemetry arrays: a case study in the study. Bengt Finstad acknowledges the financial support obtained from of smallmouth bass in a discharge canal during winter. Environ Manag. DNV GL, while Martin Føre acknowledges the financial support from Salmar 2004;33(1):140–50. ASA/Salmar Ocean. 7. Crossin GT, Heupel MR, Holbrook CM, Hussey NE, Lowerre-Barbieri SK, Nguyen VM, Raby GD, Cooke SJ. Acoustic telemetry and fisheries man- Authors’ contributions agement. Ecol Appl. 2017;27(4):1031–49. MF manages the funding project, participated in the planning and execution 8. Hussey NE, Kessel ST, Aarestrup K, Cooke SJ, Cowley PD, Fisk AT, Harcourt of the experiments, and did the main job in preparing the manuscript. ES RG, Holland KN, Iverson SJ, Kocik JF, Flemming JEM. Aquatic animal participated in the experiments, prepared the data, and was central in analys- telemetry: a panoramic window into the underwater world. Science. ing the data. FØ was central in planning the experiments and conducted the 2015;348(6240):1255642. surgery to equip the fish with the tags (together with AG). AG programmed 9. Føre M, Frank K, Norton T, Svendsen E, Alfredsen JA, Dempster T, Eguiraun the heart rate loggers, conducted the surgery to equip the fish with the tags H, Watson W, Stahl A, Sunde LM, Schellewald C, Skøien KR, Alver MO, (together with FØ) and contributed in the data analyses. JAA participated in Berckmans D. Precision fish farming: a new framework to improve pro - the planning of the experiments and data analyses. BF was the main person duction in aquaculture. Biosyst Eng. 2018;173:176–93. behind the planning and execution of the experiment and participated in the 10. Føre M, Svendsen E, Alfredsen JA, Uglem I, Bloecher N, Sveier H, Sunde experiment and data analyses. RH was responsible for the statistical analyses LM, Frank K. Using acoustic telemetry to monitor the effects of crowding and contributed to the general processing of data. IU had a key role in plan- and delousing procedures on farmed Atlantic salmon (Salmo salar). ning the experiments and had the original idea behind the study resulting in Aquaculture. 2018;495:757–65. this article. All authors have contributed to writing the manuscript. All authors 11. Brijs J, Sandblom E, Axelsson M, Sundell K, Sundh H, Huyben D, Bro- read and approved the final manuscript. ström R, Kiessling A, Berg C, Gräns A. The final countdown: continuous F øre et al. Anim Biotelemetry (2021) 9:3 Page 13 of 13 physiological welfare evaluation of farmed fish during common aquacul- the cardiovascular performance and stress responses of freely swimming ture practices before and during harvest. Aquaculture. 2018;495:903–11. rainbow trout in aquaculture. Sci Rep. 2019;9(1):9090. 12. Johansson D, Ruohonen K, Juell JE, Oppedal F. Swimming depth and 32. Hvas M, Folkedal O, Oppedal F. Heart rate bio-loggers as welfare thermal history of individual Atlantic salmon (Salmo salar L.) in produc- indicators in Atlantic salmon (Salmo salar) aquaculture. Aquaculture. tion cages under different ambient temperature conditions. Aquaculture. 2020;529:735630. 2009;290(3–4):296–303. 33. Coyle SD, Durborow RM, Tidwell JH. Anesthetics in aquaculture, vol. 3900. 13. Wright DW, Stien LH, Dempster T, Oppedal F. Differential effects of Texas: Southern Regional Aquaculture Center; 2004. internal tagging depending on depth treatment in Atlantic salmon: a 34. Leys C, Ley C, Klein O, Bernard P, Licata L. Detecting outliers: do not use cautionary tale for aquatic animal tag use. Curr Zool. 2018;65:665–73. standard deviation around the mean, use absolute deviation around the 14. Campbell HA, Bishop CM, Davies DA, Egginton S. Recording long-term median. J Exp Soc Psychol. 2013;49(4):764–6. heart rate in Paranotothenia angustata using an electronic datalogger. J 35. Miller J. Reaction time analysis with outlier exclusion: bias varies with Fish Biol. 2005;67(4):1150–6. sample size. Q J Exp Psychol. 1991;43(4):907–12. 15. Connors KB, Scruton D, Brown JA, McKinley RS. The effects of surgically 36. Lucas MC. Heart rate as an indicator of metabolic rate and activity in implanted dummy radio transmitters on the behaviour of wild Atlantic adult Atlantic salmon, Salmo salar. J Fish Biol. 1994;44(5):889–903. salmon smolts. In: Aquatic telemetry. Dordrecht: Springer; 2002. p. 231–7. 37. Becker R, Chambers JM, Wilks AR. 2018. The new S language. CRC Press. 16. Jepsen N, Davis LE, Schreck CB, Siddens B. The physiological response of 38. Cleveland RB, Cleveland WS, McRae JE, Terpenning I. STL: a seasonal-trend Chinook salmon smolts to two methods of radio-tagging. Trans Am Fish decomposition procedure based on loess. J Off Stat. 1990;6:3–73. Soc. 2001;130(3):495–500. 39. Bartels-Ferreira R, de Sousa ED, Trevizani GA, Silva LP, Nakamura FY, Forjaz 17. Thoreau X, Baras E. Evaluation of surgery procedures for implanting CLM, Lima JRP, Pecanha T. Can a first-order exponential decay model fit telemetry transmitters into the body cavity of tilapia Oreochromis aureus. heart rate recovery after resistance exercise? Clin Physiol Funct Imaging. Aquat Living Resour. 1997;10(4):207–11. 2015;35(2):98–103. 18. Ellis T, North B, Scott AP, Bromage NR, Porter M, Gadd D. The relationship 40. Cao Y, Tveten AK, Stene A. Establishment of a non-invasive method between stocking density and welfare in farmed rainbow trout. J Fish for stress evaluation in farmed salmon based on direct fecal corticoid Biol. 2002;61:493–531. metabolites measurement. Fish Shellfish Immunol. 2017;66:317–24. 19. Pickering AD. Stress responses of farmed fish. Biology of Farmed Fish. 41. Kalmijn AJ. Bioelectric fields in sea water and the function of the ampul- Sheffield: Sheffield Academic Press; 1998. p. 222–55. lae of Lorenzini in elasmobranch fishes (No. SIO-REF-72-83). La Jolla: 20. Schreck CB. Accumulation and long-term effects of stress in fish. In: Scripps Institution of Oceanography; 1972. Moberg GP, Mench JA, editors. The biology of animal stress. Wallingford: 42. Altimiras J, Larsen E. Non-invasive recording of heart rate and ventilation CAB International; 2000. p. 147–58. rate in rainbow trout during rest and swimming. Fish go wireless! J Fish 21. Wedemeyer GA. Eec ff t of rearing conditions on the health and physi- Biol. 2000;57(1):197–209. ological quality of fish in intensive fish culture. In: Iwama GK, Pickering 43. Gräns A, Sandblom E, Kiessling A, Axelsson M. Post-surgical analgesia in AD, Sumpter JP, Schreck CB, editors. Fish stress and health in aquaculture, rainbow trout: is reduced cardioventilatory activity a sign of improved vol. 62. Society for Experimental Biology Seminar Series. Cambridge: animal welfare or the adverse effects of an opioid drug? PLoS ONE. Cambridge University Press; 1997. p. 35–72. 2014;9(4):e95283. 22. Iwama GK, Afonso LOB, Vijayan MM, Iwama GK, Afonso LOB, Vijayan MM. 44. Cubitt KF, Winberg S, Huntingford FA, Kadri S, Crampton VO, Øverli Ø. Stress in fishes. In: Ewans DE, Claiborne JB, editors. The physiology of Social hierarchies, growth and brain serotonin metabolism in Atlantic fishes. Boca Raton: CRC Press; 2006. p. 319–42. salmon (Salmo salar) kept under commercial rearing conditions. Physiol 23. Mulcahy DM. Surgical implantation of transmitters into fish. ILAR J. Behav. 2008;94(4):529–35. 2003;44(4):295–306. 45. Sloman KA, Metcalfe NB, Taylor AC, Gilmour KM. Plasma cortisol concen- 24. Bridger CJ, Booth RK. The effects of biotelemetry transmitter presence trations before and after social stress in rainbow trout and brown trout. and attachment procedures on fish physiology and behavior. Rev Fish Sci. Physiol Biochem Zool. 2001;74(3):383–9. 2003;11(1):13–34. 46. Oppedal F, Dempster T, Stien LH. Environmental drivers of Atlantic salmon 25. Martinelli TL, Hansel HC, Shively RS. Growth and physiological responses behaviour in sea-cages: a review. Aquaculture 2011;311(1–4):1–18 to surgical and gastric radio transmitter implantation techniques in 47. Hvas M, Folkedal O, Oppedal F. Heart rates of Atlantic salmon Salmo salar subyearling chinook salmon (Oncorhynchus tshawytscha). In: Advances in during a critical swim speed test and subsequent recovery. J Fish Biol. invertebrates and fish telemetry. Dordrecht: Springer; 1998. p. 79–87. 2020. https ://doi.org/10.1111/jfb.14561 . 26. Barton BA, Schreck CB. Metabolic cost of acute physical stress in juvenile 48. Eliason EJ, Higgs DA, Farrell AP. Postprandial gastrointestinal blood flow, steelhead. Trans Am Fish Soc. 1987;116(2):257–63. oxygen consumption and heart rate in rainbow trout (Oncorhynchus 27. Svendsen E, Føre M, Økland F, Gräns A, Hedger RD, Alfredsen JA, Uglem mykiss). Comp Biochem Physiol A Mol Integr Physiol. 2008;149(4):380–8. I, Rosten CM, Frank K, Erikson U, Finstad B. Heart rate and swimming 49. Gräns A, Albertsson F, Axelsson M, Olsson C. Postprandial changes in activity as stress indicators for Atlantic salmon (Salmo salar). Aquaculture. enteric electrical activity and gut blood flow in rainbow trout (Onco - 2020;531:735804. rhynchus mykiss) acclimated to different temperatures. J Exp Biol. 28. Brijs J, Sandblom E, Rosengren M, Sundell K, Berg C, Axelsson M, Gräns A. 2009;212(16):2550–7. Prospects and pit-falls of using heart rate to the welfare of rainbow trout 50. Johnsson JI, Höjesjö J, Fleming IA. Behavioural and heart rate responses (Oncorhynchus mykiss) in aquaculture. Aquaculture. 2019;509:188–97. to predation risk in wild and domesticated Atlantic salmon. Can J Fish 29. Hjelmstedt P, Sundh H, Brijs J, Ekström A, Sundell KS, Berg C, Sandblom Aquat Sci. 2001;58(4):788–94. E, Bowman J, Morgenroth D, Gräns A. Eec ff ts of prophylactic antibiotic- 51. Jensen LB, Wahli T, McGurk C, Eriksen TB, Obach A, Waagbø R, Handler treatment on post-surgical recovery following intraperitoneal bio-logger A, Tafalla C. Eec ff t of temperature and diet on wound healing in Atlantic implantation in rainbow trout. Sci Rep. 2020;10(1):1–9. salmon (Salmo salar L.). Fish Physiol Biochem. 2015;41(6):1527–43. 30. Kolarevic J, Aas-Hansen Ø, Espmark Å, Baeverfjord G, Terjesen BF, Damsgård B. The use of acoustic acceleration transmitter tags for moni- Publisher’s Note toring of Atlantic salmon swimming activity in recirculating aquaculture Springer Nature remains neutral with regard to jurisdictional claims in pub- systems (RAS). Aquacult Eng. 2016;72–73:30–9. lished maps and institutional affiliations. 31. Brijs J, Sandblom E, Axelsson M, Sundell K, Sundh H, Kiessling A, Berg C, Gräns A. Remote physiological monitoring provides unique insights on http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Animal Biotelemetry Springer Journals

Heart rate and swimming activity as indicators of post-surgical recovery time of Atlantic salmon (Salmo salar)

Loading next page...
 
/lp/springer-journals/heart-rate-and-swimming-activity-as-indicators-of-post-surgical-zrXK2IKduh
Publisher
Springer Journals
Copyright
Copyright © The Author(s) 2021
eISSN
2050-3385
DOI
10.1186/s40317-020-00226-8
Publisher site
See Article on Publisher Site

Abstract

Background: Fish telemetry using electronic transmitter or data storage tags has become a common method for studying free-swimming fish both in the wild and in aquaculture. However, fish used in telemetry studies must be handled, anaesthetised and often subjected to surgical procedures to be equipped with tags, processes that will shift the fish from their normal physiological and behavioural states. In many projects, information is needed on when the fish has recovered after handling and tagging so that only the data recorded after the fish has fully recovered are used in analyses. We aimed to establish recovery times of adult Atlantic salmon (Salmo salar) after an intraperitoneal tag- ging procedure featuring handling, anaesthesia and surgery. Results: Based on ECG and accelerometer data collected with telemetry from nine individual Atlantic salmon during the first period after tagging, we found that heart rate was initially elevated in all fish and that it took an average of ≈ 4 days and a maximum of 6 days for heart rate to return to an assumed baseline level. One activity tag showed no consistent decline in activity, and two others did not show strong evidence of complete recovery by the end of the experiment: baseline levels of the remaining tags were on average reached after ≈ 3.3 days. Conclusion: Our findings showed that the Atlantic salmon used in this study required an average of ≈ 4 days, with a maximum interval of 6 days, of recovery after tagging before tag data could be considered valid. Moreover, the differences between recovery times for heart rate and activity imply that recovery time recommendations should be developed based on a combination of indicators and not just on e.g. behavioural observations. Keywords: Fish telemetry/biologging, Atlantic salmon, Post-tagging recovery, Heart rate, Swimming activity Background tags (DSTs) that store data in internal storage mediums Fish telemetry/biologging is a method of monitoring free- accessible only after the fish (and tag) has been recap - swimming fish where individual animals are equipped tured [2]. Irrespective of tag type, most studies using such with electronic tags that often contain sensors for collect- methods aim to assess the status of wild fish in ecologi - ing data on the conditions within or near the fish [1, 2]. cal settings (e.g. [4, 5]), to evaluate how fish communities Such tags may either be transmitter tags transferring data respond to man-made structures (e.g. [6]), or as a tool to wirelessly to the user (see [3] for details on the structure provide knowledge for fisheries management (reviewed of an electronic transmitter tag) or data storage/archival by [7]). The interest in using this approach in aquaculture is also increasing, both because ongoing technological advances are rapidly expanding the possibilities [8], and because new production philosophies such as Precision *Correspondence: martin.fore@ntnu.no Department of Engineering Cybernetics, Norwegian University Fish Farming promote monitoring at an individual level of Science and Technology, 7491 Trondheim, Norway [9]. Example uses of telemetry/biologging in aquaculture Full list of author information is available at the end of the article © The Author(s) 2021. This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creat iveco mmons .org/licen ses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creat iveco mmons .org/publi cdoma in/ zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Føre et al. Anim Biotelemetry (2021) 9:3 Page 2 of 13 include studies to assess fish responses during welfare- common indicators of metabolic effects due to stress is critical operations such as crowding (e.g. [10]) and the increase in plasma glucose concentration [22]. Such transport (e.g. [11]), and responses to environmental var- changes have recently been shown to lead to increased iability such as temperature variations (e.g. [12]). heart rates also in fish [27]. Other studies have aimed In animal monitoring, it is essential to ensure that to evaluate post-surgery recovery by comparing the the observed animals are representative of the targeted behaviour of the tagged fish to their behaviour before population. When using telemetry, the fish selected for surgery or in untagged cohabitant fish. This method tagging must therefore be representative both before has for instance been applied in laboratory experiments and after the tags are deployed. Ideally, this means that with tilapia (Tilapia sp.) who appeared fully recovered the selection of fish should be truly random and repre - 24  h post-surgery after displaying loss of equilibrium sentative and that the tags do not influence physiology or and reduced swimming activity and feeding just after behaviour in such a way that the tagged fish differ signifi - tagging [17]. Swimming activity was then assessed by cantly from untagged fish (e.g. [13]). In addition, tagging measuring the posture of the fish, and presented as the procedures include several steps (e.g. handling, anaes- percentage of the time the fish was resting (assuming thesia and surgical procedures) that may induce stress, an oblique angle with the snout towards the surface) that in turn may lead to physiological and/or behav- or actively swimming (horizontal orientation or snout ioural changes in the fish [2, 14–17]. Acute (short term) pointing toward the bottom). followed by chronic (long term) stress in farmed fish Recovery after tagging may also be studied with sen- may lead to undesirable effects such as reduced disease sor telemetry. The information conveyed by the tag must resistance, reduced growth rates, impaired health, and then reflect the state of the fish, and typical sensor val - increased mortality [18–21]. Stress responses in fish are ues for unstressed fish should be available as a baseline described by primary responses that include the release for comparison. Previous studies using this approach of stress hormones such as catecholamines and corti- include using heart rate tags to compare tagging methods sol into the circulation system, followed by secondary for black cod (Paranotothenia angustata, [14]), and more responses such as changes in glucose levels, electrolyte recently to study post-surgery stress-responses [28] and balance and heart rate and, finally tertiary (whole ani - potential effects of antibiotics on post-surgical recovery mal) responses. If the fish is unable to acclimate to the [29] in rainbow trout (O. mykiss). While Brijs et  al. [28] stressor at this stage, effects such as behavioural changes, implied a recovery from surgical implantation > 72  h, decreased reproductive capacity and growth may occur, Hjelmstedt et  al. [29] demonstrated a decrease in heart sometimes even resulting in that the animal dies (see [22] rate to within baseline levels 72–96  h after anaesthe- and references therein). If such changes are chronic, the sia and surgery. Other sensor measurements that could fish cannot be considered representative of the popula - potentially be used in this way include tri-axial accel- tion and should be excluded from further analyses [1, 23]. erometers, as previous studies have identified links Conversely, if the changes are transient, the fish may be between accelerometer-based activity proxies that are considered fully recovered once the response patterns particularly sensitive to tail beat frequency and amplitude return to those expected from an untagged fish. This and orientation changes, and stress in salmon [10, 30]. means that tagged fish can be used in analyses if the data Although Atlantic salmon (Salmo salar) has been fre- from the period of recovery are excluded. However, this quently studied using telemetry, there is still a lack of also raises the question: how can we define when a fish is detailed quantitative information on the post-surgery properly recovered after a tagging procedure? recovery of this species. We, therefore, sought to identify Jepsen et  al. [16] sought to identify the duration of the recovery time of Atlantic salmon after intraperitoneal post-surgery recovery for Chinook salmon (Oncorhyn- tagging. This was done using heart rate and acceleration chus tshawytscha) by studying changes in commonly data collected using intraperitoneally implanted elec- used blood indicators of the primary (cortisol) and tronic tags, meaning that data could be collected with- secondary (glucose and lactate) stress responses in tel- out introducing the additional handling stress that would eosts. The authors found that all measured parameters accompany other methods such as blood sampling. The decreased from initially elevated levels to within nor- parameters were chosen because they have previously mal ranges within 7  days post-surgery, with glucose been found to be linked with stress (e.g. [10, 28, 31]) and and lactate (substrate and by-product, respectively, of welfare [32] in salmonids and are commercially available elevated anaerobic metabolism) normalising during in archival and telemetry tags. The data were collected the first 24  h, a recovery time resembling that seen in in a controlled experiment in tanks studying how stress several studies (e.g. [24, 25]). Coping with stress is also responses in Atlantic salmon can be measured using an energy-demanding process [26] and one of the most state-of-the-art technology. The stress response part of F øre et al. Anim Biotelemetry (2021) 9:3 Page 3 of 13 this experiment is described in greater detail by Svendsen 600-60A 7  mm pellets). The fish were not subjected to et al. [27]. any fasting during the experiment period. Biotelemetry/logging systems and surgical procedures Materials and methods All 12 tagged fish (Table  1) were equipped with one Experimental site and fish of three different types of heart rate monitoring Data The experiments were conducted at the NINA Ims Storage Tags (DSTs, Star Oddi Ltd.): 4 × DST milli- Research Station near Stavanger, Norway, between HRT (39.5 × 13  mm, 11.8  g in air); 4 × DST centi- January and March 2019, using 60 hatchery-reared HRT (46 × 15  mm, 19  g); 4 × DST centi-HRT ACT adult Atlantic salmon of the Aqua Gen strain (mean (46 × 15 mm, 19 g). Using different DST types rather than 55.5 ± stdev 5.7  cm fork length, mean weight 2100  g). equipping all fish with the same tag types allowed us to The experiment started on January 28th by stocking also investigate whether all three tag varieties would be four square tanks (tank 1–4, 215 cm side, 122 cm depth, suitable for experiments with Atlantic salmon, which is 5600 L) with seven fish each. The fish were then allowed relevant because this is one of the first applications of this to habituate to the tanks for a period of 21  days until technology on this species. Furthermore, since all three February 18th when three fish in each of tanks 1–4 were tag types were from the same provider, contained the selected at random and equipped with tags, resulting in same type of heart rate sensor and comparable sampling 12 tagged fish in total (Table 1). frequencies (80  Hz over 7.5  s per HR sample point for The tanks were set up with flow-through configura - the centi tags and 100 Hz over 15 s per HR sample point tion, with filtered fresh water from the nearby Imsa river for the milli tags), and applied the same post-processing mixed with small amounts (3–6  ppt, average 5  ppt) of methods to the resulting data, they provided heart rate seawater supplied from seawater inlets at 30  m depth data sets that were comparable among tags. The milli- to ensure a stable and homogeneous water quality and HRT type was set with a higher sample storage interval avoid the introduction of parasites and pathogens to the (10  min) than the others (5  min) as they used more of tanks. Consequently, tank water properties followed the their internal storage medium for raw ECG traces. All ambient conditions in the river, temperatures increas- data were timestamped using the tag internal clocks to ing from 3.9 to 5.0 °C and with DO varying between 93.8 facilitate comparison, and eventual clock drift between and 101.2% between the start and end of the experiment individual clocks was negligible compared to the time (March 15th). Oxygen sensors and oxygenation were also scale of the experiment. One tag type (DST centi-HRT used to prevent unfavourable DO levels. The fish were ACT) also measured activity using an embedded tri-axial fed once per day between 08:00 and 10:00 in the morn- accelerometer (1 Hz sampling rate). ing throughout the entire experimental period, with each In addition to the DSTs that were applied, a total of 4 −1 meal consisting of 2  dL  tank (Skretting Røye Vitalis tagged fish (two fish each from tanks 1 and 2, Table  1) Table 1 Information about the tagged individuals used in the study Fish # Body length Est. body Tank DST type Acoustic tag Time Time surgery Sex Mature (cm) weight (kg) anaesthesia (mm:ss) (mm:ss) F1 59.5 2.3 1 centi HRT X 08:00 08:03 F F2 52 1.6 1 centi HRT X 08:20 08:12 F F3 57 2.1 1 milli HRT 08:45 06:31 F F4 63 2.8 2 centi HRT ACT X 08:06 06:30 M x F5 53 1.7 2 centi HRT ACT X 08:30 06:50 F x F6 53 1.7 2 milli HRT 07:30 06:00 M F7 74 4.5 4 centi HRT ACT 07:40 07:00 F F8 67 3.3 4 centi HRT ACT 08:00 06:30 M x F9 55 1.9 4 milli HRT 08:10 08:00 M F10 62.5 2.7 3 centi HRT 05:30 07:00 M x F11 61.5 2.6 3 centi HRT 07:00 06:30 F F12 55 1.9 3 milli HRT 07:00 06:50 F x Individual body weight was not measured and was hence estimated using an allometric model (W = aL ) with a = 0.0142 and b = 2.9401 Føre et al. Anim Biotelemetry (2021) 9:3 Page 4 of 13 experiment produced 12 data sets on heart rate, and 8 were fitted with acoustic tags (A MP-9, 24.4 × 9  mm, data sets on swimming activity. With mean fish weight 3.6 g; Thelma Biotel AS) that contained tri-axial acceler - being 2100 g and a maximum total tag weight carried by ometers (5  Hz sampling rate) and transmitted an activ- an individual at 22.6  g (DST centi-HRT + A MP-9) the ity proxy derived from the accelerometer measurements tag vs. fish weight ratio of all fish were well within the every 40  s. These tags compute the proxy by first high informal rule of thumb of 2% for maximum tag mass rela- pass filtering the accelerations from all three axes using tive to fish mass [2]. a cutoff frequency of 0.2  Hz to remove low-frequency Each tag implantation was started by capturing a ran- acceleration components due to gravity and body orien- dom fish from an experiment tank using a knotless dip tation. The remaining high-frequency components then net and immediately transferring it to an anaesthetic bath mainly contain accelerations caused by features related (Benzoak Vet, 70  mg/L) where the fish was kept until it to bodily movement that are of interest when evaluating lost its equilibrium and stage III anaesthesia [33] was activity levels, such as tail beats (frequency and ampli- reached (average time 7.7  min). The fish was then care - tude) and rapid changes in attitude/orientation. The fully placed with its ventral side up on a specialised sur- Euclidian norm of the three high pass filtered acceler - gical table with a v-shaped mid-section designed such ometer axes is then computed to yield the magnitude of that the head of the fish was immersed in water through - the total high pass filtered 3D acceleration sensed by the out the whole procedure. A hose circulating anaesthetic accelerometer. Although Føre et  al. [10] used the same −2 (Benzoak Vet, 35  mg/L) through the orobranchial cavity activity proxy with a maximum value of 3.465 m s , we −2 of the fish was inserted into its mouth and the head was chose to limit the proxy to 0–2.1  m  s in our study as covered by a moist cloth (Fig. 1). this gave us a higher resolution and hence precision for A 2–3  cm incision was made along the sagittal plane the activity measures. Moreover, Føre et al. [10] observed −2 starting slightly more than one tag length (i.e. the length very few activity values above 2 m s in Atlantic salmon of the tag to be implanted) posterior from the transverse during stressing, implying that using a lower range would pericardial septum. not compromise the ability to capture the dynamics A finger was inserted through the incision to locate associated with salmon swimming activity. To be com- the transverse pericardial septum. While retaining the parable with the data from the acoustic tags, the activ- finger inside the peritoneal cavity for support, a needle ity data from the centi-HRT ACT DSTs were analysed was positioned in the skin just posterior to the trans- similarly by applying filtering and computing the Euclid - verse septum and slightly laterally from the sagittal plane. ian norm as explained for the acoustic tags (see [27] for The finger was withdrawn, and a smooth plastic spoon more details). Adding the acoustic tags thus allowed us inserted through the incision until it was just below the to compare their activity proxies with those based on the needle insertion point. The needle was then pushed acceleration data from the DSTs and resulted in that the Fig. 1 Fish in the surgical table with anaesthetic circulation tube and head cover, indicating approximate locations of DSTs (white tag) and acoustic tags (black tag) after implantation F øre et al. Anim Biotelemetry (2021) 9:3 Page 5 of 13 through the peritoneal wall while simultaneously with- above which samples are rejected. The MAD decision drawing the spoon to extract the needle out through the criterion typically ranges from 2 (poorly conservative) incision while protecting the viscera. One end of a suture to 3 (very conservative). In this study, the choice of 3 is threaded through the end of the tag was inserted into the justified by the measured heart ranges compared to typ - tip of the needle. The needle was then withdrawn to pull ical heart rates published in the literature (15 < HR < 80) the suture out through the needle’s entry point. This pro - for Atlantic salmon and comparable species [28, 36]. cedure was then repeated on the other side of the sagit- Activity data from the DST centi-HRT ACT tags were tal plane. The tag was then inserted through the incision downloaded as raw acceleration values along all three and anchored anteriorly in the peritoneal cavity using the axes and then subjected to similar post processing as suture and an (external) surgical knot. For the four fish that used to compute the activity proxy in the A MP-9 also equipped with separate acoustic tags, the second acoustic transmitter tags to yield a comparable measure tag was inserted into the peritoneal cavity through the of activity between the two tag types. same incision. Finally, the incision was closed using inter- In a non-decomposed time-series, circadian vari- rupted sutures. The fish was then transferred to a recov - ation (that between day and night) and irregular vari- ery tank with circulating seawater where it was kept until ation (that other than circadian of long-term) had the it regained consciousness, upon which it was transferred potential to obscure long-term trends in heart rate and back into the tank it was collected from. See Table  1 for activity. Time-series of heart rate and activity were anaesthesia bath and surgery durations for all tagged fish. therefore first decomposed into circadian, long-term trend, and irregular components. Decomposition, and subsequent removal of the circadian and irregular com- Timeline and experimental design ponents of the time-series, leaving a long-term com- Since the present study focused on investigating the post- ponent (that showed the long-term growth or decline tagging recovery, the analyses only included data from of the time-series values over the temporal extent of the 2  weeks following tagging. To avoid inducing other the series), allowed for the examination of the form of stress effects that could disturb their recovery, the fish the long-term trends towards recovery. To decompose were sheltered from all potential stress factors except each time-series, it was first binned into 15  min inter - those necessary to feed and provide for the fish in this vals (each 15  min interval showing a mean heart rate period. or intensity over that interval) and then converted into None of the fish exhibited signs of adverse health after a time-series object [R function ts {stats}; Becker et  al. tagging or during the trials, and all fish were euthanised [37]]. Time-series objects were then decomposed using after the conclusion of the experiment. Posthumous the Seasonal Decomposition of Time Series by Loess R pathology of all remaining experimental fish at the end function stl {stats} (B. D. Ripley; Fortran code by Cleve- of the experiment (19 female, 23 male) revealed that land et  al. [38] from “netlib”). Long-term trend com- about one-third of these fish (14 in total, 8 F, 6 M) exhib - ponents were then analysed for a systematic change ited signs of sexual maturation through the experimen- in heart rate or activity that could be indicative of a tal period, including 5 of the tagged individuals (Table 1). post-surgery recovery by first modelling the temporal Although this appeared to have a little direct impact on relationship and then compartmentalising this into pre- the fish in three of the tanks, the data from the fish in one and post-recovery phases. of the tanks (tank 3) were excluded from the statistical The relationship between the long-term trend compo - data analyses due to perpetual inter-individual aggression nent of heart rate or activity (y) and time post-tagging between two matured males in that tank throughout the (t) was modelled using an exponential decay model: experimental period. This left nine fish tagged with DSTs measuring heart rate, six of which also measured activ- −αt y(t) = y + y − y e , p 0 p ity. Since two of these fish contained both a DST and an acoustic tag measuring activity, this resulted in a total of where α defines the decay constant from y (at eight time-series of activity. time zero) to y , the model plateau. Models were fit - ted with the nls {stats} R function (D. M. Bates and S. Data processing and statistics DebRoy: D. M. Gay for the Fortran code used by algo- Heart rate data were used as downloaded from the rithm = "port"), using the self-starting asymptotic DSTs. Outliers were removed using the Median Abso- regression function SSasymp {stats} (J. Pinheiro and D. lute Deviation (MAD) approach [34], using a MAD M. Bates). Most trend components followed an expo- decision criterion of 3, which is a conservative value nentially decaying pattern, ensuring model conver- (see [35]). The MAD decision criterion denotes the gence, but some included parts that were inconsistent standard deviation from the dataset’s sample average Føre et al. Anim Biotelemetry (2021) 9:3 Page 6 of 13 with an exponential decay. First, some tags (three asymptotes in the final 3  days, so it was reasonable to heart-rate tags and four activity tags) showed a short assume that values from these days represented post- initial post-surgery increase in registered values at the recovery signature. Thresholds were established on an beginning of the experiment. Secondly, some tags (one individual basis to allow for post-recovery heart rate or heart rate and two activity tags) showed an increase in activity to change according to individuals. registered values after ≈ 5–6 days. This late increase in activity or heart-rate was likely a result of a separate, Results post-recovery change in behaviour of these individuals. Post‑surgery recovery To ensure model convergence, these parts of the long- Daily heart rate significantly declined from a mean of term trend components were removed prior to model 36.0  bpm (range = 24.6–45.6, SD = 5.6, n = 9) on the day fitting. That is, the exponential model was only fitted of surgery to a mean of 22.3  bpm (range = 17.5–26.6, to parts of the long-term trend component that were SD = 2.6, n = 9) 13 days later (one-sided Wilcoxon’s rank consistent with a post-surgery exponential decline. One test, V = 45, p = 0.002) (Fig. 2a; raw data for all tags shown activity tag (fish F4 in tank 8) did not show an expo - in). Daily activity significantly declined from a mean of −2 nential decline with time and was thus not fitted with 0.57  m  s (range = 0.26–0.92, SD = 0.23, n = 8) on the −2 a model. day of surgery to a mean of 0.33  m  s (range = 0.27– Identification of breakpoints between pre- and post- 0.41, SD = 0.06, n = 8) 13  days later (one-sided Wil- recovery phases was done on an individual basis. The coxon’s rank test, V = 45, p = 0.008) (Fig.  2b). However, breakpoint between pre- and post-recovery for each tag individual variation in activity was high (Fig.  2b). Both was set where the heart rate or activity reached a recov- heart rate and activity displayed circadian variation. ery threshold, defined as the heart rate or activity level Heart rate was greater during daytime (mean = 25.8 bpm, delimiting those pre- and post-recovery. A recovery range = 22.2–26.7, SD = 1.9, n = 9) than during night threshold was defined for each tag as the mean + 2SD of (mean = 22.7  bpm, range = 19.6–24.9, SD = 1.9, n = 9) the long-term trend component values calculated from (Fig.  2a; raw data for all tags shown in Additional file  1: the final 3  days of the fitted series. Inspection of the Figures  S1, S2). In contrast, activity was greater during −2 tags showed that trend components were approaching night (mean = 0.47  m  s , range = 0.32–0.60, SD = 0.11, Fig. 2 Average a heart rate and b activity values (blue line) from all fish in tanks 1, 2 and 4 (tank 3 was excluded because of aggressive behaviour between two males) for the first 2 weeks post tagging. The light-blue envelope shows the range in heart rate values of all individuals F øre et al. Anim Biotelemetry (2021) 9:3 Page 7 of 13 −2 circadian components from the time series (Additional n = 8) than during daytime (mean = 0.34  m  s , file  1: Figures  S5, S6). There was generally greater activ - range = 0.25–0.43, SD = 0.06, n = 8) (Fig.  2b; raw data ity during the evening than during morning (Additional for all tags shown in Additional file  1: Figures  S3, S4). file 1: Figure S6). Circadian differences were present throughout the The heart rate trend component showed a decline experiment. However, the circadian difference in activ - that could be modelled with an exponential decay ity declined throughout the experiment. This circadian function (Fig.  3). However, the trend component still variation was shown by all individuals, as revealed by the Fig. 3 Long term trend in heart rate. Each panel represents data from one individual fish. The continuous black line shows the long-term trend component; the continuous blue line shows the fitted exponential decay model. The dashed horizontal line shows the threshold used to define a structural change; the dashed vertical line shows the breakpoint indicating when the structural change occurs between classified pre- and post-recovery phases Føre et al. Anim Biotelemetry (2021) 9:3 Page 8 of 13 showed considerable temporal variation, depending with an exponential decay (Fig.  4), except for one fish on the tagged individual. For example, the trend com- (fish F8) where an exponential decay model could not ponent for fish F4 showed a sharp decline during the be fitted due to the activity trend component peaking first day after tagging, but this then fluctuated for the ≈ 7 days after tagging. Two fish (fish F1 and F2) showed remainder of the 2-week post-tagging period. The activ - an exponential decline in activity but did not reach a ity trend component also showed a pattern consistent Fig. 4 Long term trend in activity. Each panel represents data from one individual fish. The continuous black line shows the long-term trend component; the continuous blue line shows the fitted exponential decay model. The dashed horizontal line shows the threshold used to define a structural change; the dashed vertical line shows the breakpoint indicating when the structural change occurs between classified pre- and post-recovery phases. It was not possible to fit an exponential decay model to the long-term trend component of fish F8. Thresholds and breakpoints were not established for fish F1 and F2 where the exponential model did not plateau. Tag type (DST = Data Storage Tag; Aco = Acoustic tag) is shown in parentheses following the fish ID F øre et al. Anim Biotelemetry (2021) 9:3 Page 9 of 13 plateau during the study period, suggesting that these individual (F8, Fig. 4), there was a similar trend for activ- −2 fish has not fully recovered in terms of activity. ity: day of anaesthesia and surgery, mean = 0.64  m  s , Time to recovery (as defined by the location of the range = 0.39–0.92, SD = 0.20, n = 7; recovery threshold, −2 breakpoint between pre- and post-recovery phases) var- mean = 0.36  m  s , range = 0.28–0.43 SD = 0.07, n = 7. ied between individuals, and the metric used (heart rate For both heart rate and activity, raw values pre-recovery or activity, Figs.  3, 4, Table  2). The mean threshold value were significantly greater than those post-recovery (one- for heart rate in a ‘recovered’ individual was 23.8  bpm sided Wilcoxon’s signed-rank test: heart rate, V = 45, (range = 21.2–26.0, SD = 1.18, n = 9). The mean time to p = 0.002, n = 9; activity, V = 28, p = 0.008, n = 7). reach this threshold (i.e. breakpoint between pre-recov- ery and post-recovery) was 4.1  days (range = 1.3–5.8, Discussion SD = 1.7, n = 9). The threshold for activity recovery was The current study showed plateauing of most time-series, −2 greater for the acoustic tags (mean = 0.44  m  s , n = 2) indicative of recovery, within the 14  days of the experi- −2 than the DSTs (mean = 0.29  m  s , n = 3), reflecting the ment. Two activity tags, F1(Aco) and F2(Aco), however, higher activity values registered by the acoustic tags. For did not show plateauing, suggesting that the tagged fish the activity tags where there was evidence of recovery, had not fully recovered in terms of activity during this the mean time taken to reach the threshold was similar to period. Other time-series showed gentle gradients even that for the heart rate tags (mean 3.3  days, range = 2.1– after the recovery breakpoint (for example, the F6 heart 5.7, SD = 0.09). For the two individuals that were each rate tag) so the definition of the point of recovery of some tagged with two activity tags, the identified breakpoints individuals as having fully recovered is less robust. How- between the parts of the time series classified as pre- and ever, identified breakpoints generally corresponded with post-recovery depended on the tag: in both individuals, systematic changes in the time-series. For example, the the threshold to reach post-recovery occurred later for breakpoint on the F6 heart rate tag occurred in a trough the acoustic tag than the DST. separating the sharp initial decline over the first 5.75 days Although raw values of mean heart rate on the day of with the gentle gradient afterwards, so it is reasonable anaesthesia and surgery (mean = 36.0 bpm, range = 24.6– to infer that the identified breakpoint corresponded to 45.6, SD = 5.6, n = 9) varied more than the recov- the transition to post-recovery. The modelling approach ery threshold (mean = 23.8  bpm, range = 21.2–26.0, used here allowed for a consistent method for establish- SD = 1.8, n = 9, Table 2), there was a clear declining trend ing the time until recovery among a group of time-series. for all tagged individuals. With the exception of one It should be noted, however, that estimated times until recovery is dependent on modelling approach used. For instance, fitting an exponential model to raw—rather than detrended timeseries, or using a different method Table 2 Recovery based on heart rate and activity sensors to establish a breakpoint between pre- and post-recovery Tag ID Heart rate recovery Activity recovery parts of the time-series, would yield different estimates. Threshold Time (days) Threshold Time (days) The exponential model used in this study is a well-vali - −2 (bpm) (m s ) dated method for modelling physiological recovery [39] but alternative approaches may also be considered (e.g. F1 24.90 4.33 No rec No rec [27]). The sample size of fish in this study was small F2 25.83 5.27 No rec No rec (N = 9); a larger sample size would allow a better quan- F3 23.17 4.42 tification of the range of behaviour during recovery and F4 21.43 2.92 0.31 3.59 allow better selection of the modelling approach. 0.45* 5.08* The heart rate data suggest that the tagged Atlantic F5 25.63 1.89 0.28 2.10 salmon in our study could only be considered fully recov- 0.43* 3.30* ered from the anaesthesia and surgical procedure of F6 21.24 5.75 intraperitoneal tag implantation after an average of ≈ 4 F7 25.97 5.41 0.28 2.60 and up to a maximum of 6 days post-surgery. While some F8 23.68 1.27 No fit No fit studies have indicated longer recovery times post-tagging F9 22.71 5.29 [32], our observations concur with several previous stud- Mean 23.84 4.06 0.35 3.3 ies that have reported similar lengths of recovery post- Activity sensors with a * suffix indicate acoustic tags. “No fit” indicates that the tagging as our study [11, 16, 24, 25, 28]. Although some long-term component of the time-series did not follow an exponential decline and that an exponential model could not be fitted; “No rec” indicates that it data series from the tagged fish in our study may visually was possible to fit an exponential model to the time-series but that recovery appear to continue declining after fulfilling the recovery thresholds and times were not assigned because the fitted exponential model threshold criteria, these changes were not found to be did plateau Føre et al. Anim Biotelemetry (2021) 9:3 Page 10 of 13 statistically significant. Recovery results based on activity accelerometers were found to be comparable to those data varied more in the recovery threshold criteria and measured by the acoustic tags (see [27] for details on this time to recovery than heart rate, suggesting it might be comparison). The lower absolute amplitude of the activ - a less consistent indicator of recovery between individu- ity proxies computed from the DST data was probably als. Moreover, both the temporal patterns and absolute caused by them sampling at a lower frequency (1  Hz) values changed less for activity than heart rate between than the acoustic tags (5  Hz), thereby capturing fewer post-tagging and post-recovery periods, implying a lower high-frequency components. The surgical procedure ratio between the baseline pattern (i.e. circadian varia- used to implant the heart rate tags was much simpler tions) and the changes in activity caused by the tagging than the procedure needed for multivariate implants procedure. Together, these factors suggest that activity recently used in rainbow trout by Brijs et al. [31] but was may be a less consistent indicator of post-tagging recov- more comprehensive and invasive than that used for con- ery than heart rate, and that heart rate might be a gener- ventional intraperitoneal tag placement. It is likely that ally more sensitive indicator than activity, especially for less complex surgical procedures would lead to shorter post-tagging recovery. recovery times in Atlantic salmon, as previously found for It is also important to note that there were individual rainbow trout [42, 43]. However, it is probably reasonable variations in the recovery time assessed from heart rate. to be conservative with respect to recovery times, espe- Although inter-individual variation in recovery time cially if the data are to be used e.g. as a management tool might be an inherent effect one should expect when tag - in aquaculture applications or to evaluate stress effects ging A. salmon, we did find that mature fish had a lower on fish in conjunction with ecological studies. Using data heart-rate recovery time than immature fish. However, from fish that are still recovering from post-anaesthesia/ the low sample size did not provide enough statistical surgery effects in such applications could result in sub- power to robustly test influences on recovery time, so we optimal management decisions or erroneous conclusions recommend further studies with larger sample sizes to that could have ramifications beyond the study itself. increase power in analyses of potential influences. The fish included in the analyses exhibited heart rates Based on these results, we urge caution on using that gradually stabilised at daily means between 21 and telemetry data collected after anaesthesia and surgery 26  bpm (daily variations between 15 and 30  bpm, simi- without first ensuring that the fish are fully recovered lar to that observed by for adult A. salmon of mean fork [1, 23]. Biosensors that measure heart rate and/or activ- length of 62.3 cm at 4 °C by Lucas [36]). Due to the simi- ity can be potent tools in such evaluations, as they pro- larities across tanks and individuals, this range in heart vide quantitative, high-resolution data that will be both rate may be typical for Atlantic salmon of this size and more consistent, precise and objective in capturing the with the prevailing temperatures. Moreover, all indi- full post-anaesthesia/surgery effects than e.g. comparing viduals in tanks 1, 2 and 4 had similar circadian rhythms behavioural observations of tagged vs. untagged fish. (higher heart rates during daytime than at night) and Alternative parameters that could be used to assess gradual post-surgery declines in mean daily heart rate post-tagging recovery in individual fish include blood (from more than 30 bpm after surgery to 21–26 bpm after glucose, lactate or pulse oximetry/ppg. These could pro - up to 6 days). This implies a regularity across individuals vide a more direct assessment of stress levels in salmon, that increases the likelihood that heart rate may func- but we are not aware of any commercial electronic tags tion as a consistent stress indicator in Atlantic salmon able to sense such parameters in live fish. Other tech - that may be used to assess fish recovery after tagging. The niques based on measuring cortisol in faecal matter [40] tagged fish in tank 3 were excluded from the study due or bioelectric field monitoring akin to that used by sharks to inter-individual aggression. These individuals demon - [41] could potentially result in future solutions that could strated measured heart rates that differed from the others be used evaluate recovery in a less invasive and independ- both in individual and aggregate values. Although these ent manner, where the fish are monitored before, during fish also showed signs of circadian variation in heart and after the procedure. However, these methods are still rate, the mean value did not appear to decline over the to be developed to a stage where they can be applied to days following tagging, an effect that was attributed to free-swimming fish, at least in large groups under com - inter-individual aggression, all else being held equal. This mercial production conditions, and would only be able to may indicate that the stress induced by the aggression provide information on a group level. between the two males in this tank overrode the stress All three DST types tested in this experiment appeared response due to recovery. A potential interpretation of to be suitable for applications on Atlantic salmon as all this is that the aggressive encounters caused chronically tagged fish provided valid heart rate data. Moreover, elevated stress levels that masked the recovery stress the activity proxy computed from the DSTs containing caused by handling, anaesthesia and surgery. This could F øre et al. Anim Biotelemetry (2021) 9:3 Page 11 of 13 further mean that recovery stress can be difficult to 2% rule”, [2]). Since we worked with adult salmon with a monitor if the fish are simultaneously influenced by inde - mean weight of 2100 g, and the maximum tag weight car- pendent external events, such as individual interactions ried by the fish was 22.6  g (around 1% of the fish body due to dominance hierarchies [44, 45]. mass) this was not a challenge in our study. Based on established knowledge on how salmon swim- ming speeds are affected by variations in light intensity Future research and potential technological improvements [46], as well as previous telemetry studies applying simi- Since this study only focused on Atlantic salmon exposed lar activity proxies on salmon in sea-cages (e.g. [10]), to one set of environmental conditions, it is difficult to we expected to see a circadian rhythm in activity where assess if these concerns are also relevant for other spe- activity was higher during the day than at night in the cies, and/or fish under different conditions. Similar stud - present study. In contrast to these expectations, the cir- ies on rainbow trout using the same tag type found that cadian trends in the activity of our fish were on average they recovered 72–96  h after surgery [28], which was higher during night-time than during the day. A simi- shorter than the Atlantic salmon in the present study. lar “inverse circadian” rhythm was observed in salmon Moreover, wounds in Atlantic salmon are known to heal reared in fish tanks during the period after tagging by faster in warmer temperatures than in cold water [51], Kolarevic et  al. [30] and could imply that a “normal cir- suggesting that the low water temperatures in the pre- cadian” activity rhythm may arise only after the salmon sent study may have contributed to longer recovery peri- have recovered after tagging in tanks. Conversely, the ods. These elements suggest that species-specific effects circadian rhythm in heart rate was more like expected or differences in external environmental conditions are (higher during daytime), meaning that the fish displayed important to consider when studying recovery times. generally higher heart rates when measured activity was Future studies on the relationship between heart rate and low than when activity was high. This may seem counter- post anaesthesia/surgery recovery time should therefore intuitive as one would expect more active fish to display be conducted for other species of interest, across relevant higher heart rates since salmon tend to display increased temperature ranges, to obtain a more complete picture of heart rates with increased swimming activity [47]. How- this relationship. ever, it is possible that the higher heart rates during day- In the present experiment, the fish were kept in groups time were caused by effects such as feeding activity [48, in small tanks. To investigate how recovery time is 49] or perceived increased predation risk due to higher affected by eventual scaling effects and social/inter-indi - light levels [50]. These results are unexpected and very vidual effects arising due to group dynamics, future stud - interesting, but further extrapolations and discussions on ies addressing post-tagging effects should be done with this matter would probably require further experiments a larger number of tagged fish at larger spatial scales. with more data. This would also enable a deeper scrutiny into individual Although this study underlines the importance of criti- variations in recovery, as a higher number of tagged fish cal evaluation with regards to recovery from anaesthesia would provide a good foundation for finding statistical and surgery when using telemetry, the data collected also relationships on the individual level. Although our pre- highlight the importance of telemetry as a method for sent results imply that inter-individual variations are a studying free-swimming fish. The heart rate and activity prominent feature in the recovery time of tagged salmon, values for all tagged fish eventually plateaued, possibly a larger sample number will be necessary to properly indicating that they all recovered from the anaesthe- conclude upon the nature of such variations. To increase sia/surgery, and posthumous pathology revealed no the relevance of a larger follow-up study, it could be done inflammations or other apparent morphological signs in fish cages in the marine environment, perhaps first by of reduced welfare due to the surgical procedures. Even using meso-scale size cages containing fewer fish than a though the low water temperatures during the experi- commercial cage but at similar densities, and then mov- ment may have led to handling and surgery having less ing to full-scale studies to cover all steps in the transition impact on the fish, the tagging procedure used here was from lab to industrial scale. more complex than conventional intraperitoneal tag- ging. It is thus reasonable to conclude that fish carrying Conclusion telemetry tags can be considered representative members The main conclusion from this study is that the Atlan - of the group they were selected from once they are fully tic salmon in these experiments required an average of recovered from anaesthesia and surgery, provided that ≈ 4 and up to a maximum interval of 6 days of recovery they were a representative selection to begin with. How- after anaesthesia and surgery before their heart rates ever, this also requires that the recommendations on ratio returned to assumed baseline routine values. Moreover, between tag size and fish size are not exceeded (e.g. “the although observation of behaviour and/or activity may Føre et al. Anim Biotelemetry (2021) 9:3 Page 12 of 13 Funding alone be insufficient to assess that the fish has physi - This study was funded by the Research Council of Norway (NFR Project ologically recovered, activity measurements indicated Number 280864). similar recovery periods to those based on heart rate, Availability of data and materials although there was a longer maximum period of 10 days. The datasets used and/or analysed during the current study are available from We, therefore, urge caution when using data collected the corresponding author on reasonable request. after surgery and anaesthesia in studies using biologging/ Ethics approval and consent to participate telemetry tags. Assuming that we want all individuals to All fish handling and surgery were made in compliance with the Norwegian be recovered, our study thus implies that only data col- animal welfare act and were approved by the Norwegian Animal Research lected after 6 days recovery time should be used for fur- Authority (Permit No. 18/18431). ther analyses. However, this recommendation would only Consent for publication be applicable to studies featuring Atlantic salmon reared Not applicable. in similar experimental conditions as we used. Since Competing interests recovery time will vary with factors such as fish species, The authors declare that they have no competing interests. water temperature, invasiveness of the surgery, anaesthe- sia time, fish density and physical scale, it is difficult to Author details Department of Engineering Cybernetics, Norwegian University of Science make general recommendations on when one can assume and Technology, 7491 Trondheim, Norway. SINTEF Ocean, 7465 Trondheim, the fish to be recovered from tagging, and the data to Norway. Norwegian Institute for Nature Research, 7485 Trondheim, Norway. be safe for use in biological analyses. However, by con- Department of Animal Environment and Health, Swedish University of Agri- cultural Sciences, 532 31 Skara, Sweden. Department of Biology, Norwegian ducting experiments similar to the present study where University of Science and Technology, 7491 Trondheim, Norway. these parameters are varied, a more complete picture of how we need to account for fish recovery after tagging in Received: 16 April 2020 Accepted: 9 December 2020 telemetry studies may be obtained. Supplementary Information References The online version contains supplementary material available at https ://doi. 1. Cooke SJ, Woodley CM, Eppard MB, Brown RS, Nielsen JL. Advancing org/10.1186/s4031 7-020-00226 -8. the surgical implantation of electronic tags in fish: a gap analysis and research agenda based on a review of trends in intracoelomic tagging Additional file 1: Figure S1. Measured heart rate for Tag IDs F1–F4. The effects studies. Rev Fish Biol Fish. 2011;21(1):127–51. y-axis upper limit has been constrained to 50 bpm. Figure S2. Measured 2. Thorstad EB, Rikardsen AH, Alp A, Økland F. The use of electronic tags in heart rate for Tag IDs F5–F9. The y-axis upper limit has been constrained fish research—an overview of fish telemetry methods. Turk J Fish Aquat to 50 bpm. Figure S3. Measured activity for acoustic tags. Figure S4. Sci. 2013;13(5):881–96. Measured activity for DSTs. The y-axis upper limit has been constrained 3. Føre M, Alfredsen JA, Gronningsater A. Development of two telemetry- −2 to 5 m s . Figure S5. Circadian component of heart rate. Each panel based systems for monitoring the feeding behaviour of Atlantic salmon represents data from one individual fish. Figure S6. Circadian component (Salmo salar L.) in aquaculture sea-cages. Comput Electron Agric. of the activity. Each panel represents data from one individual fish. 2011;76(2):240–51. 4. Taylor MD, Babcock RC, Simpfendorfer CA, Crook DA. Where technology meets ecology: acoustic telemetry in contemporary Australian aquatic Acknowledgements research and management. Mar Freshw Res. 2017;68(8):1397–402. In addition to thanking the Research Council of Norway for financial support 5. Welsh JQ, Bellwood DR. Spatial ecology of the steephead parrotfish (see “Funding”), we would like to thank the crew at NINA’s research station at (Chlorurus microrhinos): an evaluation using acoustic telemetry. Coral Ims, particularly Knut Bergesen, for their effort in the experiments. We would Reefs. 2012;31(1):55–65. also like to thank Asgeir Bjarnason at Star Oddi LTD and Erik Høy at Thelma 6. Cooke SJ, Bunt CM, Schreer JF. Understanding fish behavior, distribution, Biotel AS for collaboration in specifying the loggers and transmitter tags used and survival in thermal effluents using fixed telemetry arrays: a case study in the study. Bengt Finstad acknowledges the financial support obtained from of smallmouth bass in a discharge canal during winter. Environ Manag. DNV GL, while Martin Føre acknowledges the financial support from Salmar 2004;33(1):140–50. ASA/Salmar Ocean. 7. Crossin GT, Heupel MR, Holbrook CM, Hussey NE, Lowerre-Barbieri SK, Nguyen VM, Raby GD, Cooke SJ. Acoustic telemetry and fisheries man- Authors’ contributions agement. Ecol Appl. 2017;27(4):1031–49. MF manages the funding project, participated in the planning and execution 8. Hussey NE, Kessel ST, Aarestrup K, Cooke SJ, Cowley PD, Fisk AT, Harcourt of the experiments, and did the main job in preparing the manuscript. ES RG, Holland KN, Iverson SJ, Kocik JF, Flemming JEM. Aquatic animal participated in the experiments, prepared the data, and was central in analys- telemetry: a panoramic window into the underwater world. Science. ing the data. FØ was central in planning the experiments and conducted the 2015;348(6240):1255642. surgery to equip the fish with the tags (together with AG). AG programmed 9. Føre M, Frank K, Norton T, Svendsen E, Alfredsen JA, Dempster T, Eguiraun the heart rate loggers, conducted the surgery to equip the fish with the tags H, Watson W, Stahl A, Sunde LM, Schellewald C, Skøien KR, Alver MO, (together with FØ) and contributed in the data analyses. JAA participated in Berckmans D. Precision fish farming: a new framework to improve pro - the planning of the experiments and data analyses. BF was the main person duction in aquaculture. Biosyst Eng. 2018;173:176–93. behind the planning and execution of the experiment and participated in the 10. Føre M, Svendsen E, Alfredsen JA, Uglem I, Bloecher N, Sveier H, Sunde experiment and data analyses. RH was responsible for the statistical analyses LM, Frank K. Using acoustic telemetry to monitor the effects of crowding and contributed to the general processing of data. IU had a key role in plan- and delousing procedures on farmed Atlantic salmon (Salmo salar). ning the experiments and had the original idea behind the study resulting in Aquaculture. 2018;495:757–65. this article. All authors have contributed to writing the manuscript. All authors 11. Brijs J, Sandblom E, Axelsson M, Sundell K, Sundh H, Huyben D, Bro- read and approved the final manuscript. ström R, Kiessling A, Berg C, Gräns A. The final countdown: continuous F øre et al. Anim Biotelemetry (2021) 9:3 Page 13 of 13 physiological welfare evaluation of farmed fish during common aquacul- the cardiovascular performance and stress responses of freely swimming ture practices before and during harvest. Aquaculture. 2018;495:903–11. rainbow trout in aquaculture. Sci Rep. 2019;9(1):9090. 12. Johansson D, Ruohonen K, Juell JE, Oppedal F. Swimming depth and 32. Hvas M, Folkedal O, Oppedal F. Heart rate bio-loggers as welfare thermal history of individual Atlantic salmon (Salmo salar L.) in produc- indicators in Atlantic salmon (Salmo salar) aquaculture. Aquaculture. tion cages under different ambient temperature conditions. Aquaculture. 2020;529:735630. 2009;290(3–4):296–303. 33. Coyle SD, Durborow RM, Tidwell JH. Anesthetics in aquaculture, vol. 3900. 13. Wright DW, Stien LH, Dempster T, Oppedal F. Differential effects of Texas: Southern Regional Aquaculture Center; 2004. internal tagging depending on depth treatment in Atlantic salmon: a 34. Leys C, Ley C, Klein O, Bernard P, Licata L. Detecting outliers: do not use cautionary tale for aquatic animal tag use. Curr Zool. 2018;65:665–73. standard deviation around the mean, use absolute deviation around the 14. Campbell HA, Bishop CM, Davies DA, Egginton S. Recording long-term median. J Exp Soc Psychol. 2013;49(4):764–6. heart rate in Paranotothenia angustata using an electronic datalogger. J 35. Miller J. Reaction time analysis with outlier exclusion: bias varies with Fish Biol. 2005;67(4):1150–6. sample size. Q J Exp Psychol. 1991;43(4):907–12. 15. Connors KB, Scruton D, Brown JA, McKinley RS. The effects of surgically 36. Lucas MC. Heart rate as an indicator of metabolic rate and activity in implanted dummy radio transmitters on the behaviour of wild Atlantic adult Atlantic salmon, Salmo salar. J Fish Biol. 1994;44(5):889–903. salmon smolts. In: Aquatic telemetry. Dordrecht: Springer; 2002. p. 231–7. 37. Becker R, Chambers JM, Wilks AR. 2018. The new S language. CRC Press. 16. Jepsen N, Davis LE, Schreck CB, Siddens B. The physiological response of 38. Cleveland RB, Cleveland WS, McRae JE, Terpenning I. STL: a seasonal-trend Chinook salmon smolts to two methods of radio-tagging. Trans Am Fish decomposition procedure based on loess. J Off Stat. 1990;6:3–73. Soc. 2001;130(3):495–500. 39. Bartels-Ferreira R, de Sousa ED, Trevizani GA, Silva LP, Nakamura FY, Forjaz 17. Thoreau X, Baras E. Evaluation of surgery procedures for implanting CLM, Lima JRP, Pecanha T. Can a first-order exponential decay model fit telemetry transmitters into the body cavity of tilapia Oreochromis aureus. heart rate recovery after resistance exercise? Clin Physiol Funct Imaging. Aquat Living Resour. 1997;10(4):207–11. 2015;35(2):98–103. 18. Ellis T, North B, Scott AP, Bromage NR, Porter M, Gadd D. The relationship 40. Cao Y, Tveten AK, Stene A. Establishment of a non-invasive method between stocking density and welfare in farmed rainbow trout. J Fish for stress evaluation in farmed salmon based on direct fecal corticoid Biol. 2002;61:493–531. metabolites measurement. Fish Shellfish Immunol. 2017;66:317–24. 19. Pickering AD. Stress responses of farmed fish. Biology of Farmed Fish. 41. Kalmijn AJ. Bioelectric fields in sea water and the function of the ampul- Sheffield: Sheffield Academic Press; 1998. p. 222–55. lae of Lorenzini in elasmobranch fishes (No. SIO-REF-72-83). La Jolla: 20. Schreck CB. Accumulation and long-term effects of stress in fish. In: Scripps Institution of Oceanography; 1972. Moberg GP, Mench JA, editors. The biology of animal stress. Wallingford: 42. Altimiras J, Larsen E. Non-invasive recording of heart rate and ventilation CAB International; 2000. p. 147–58. rate in rainbow trout during rest and swimming. Fish go wireless! J Fish 21. Wedemeyer GA. Eec ff t of rearing conditions on the health and physi- Biol. 2000;57(1):197–209. ological quality of fish in intensive fish culture. In: Iwama GK, Pickering 43. Gräns A, Sandblom E, Kiessling A, Axelsson M. Post-surgical analgesia in AD, Sumpter JP, Schreck CB, editors. Fish stress and health in aquaculture, rainbow trout: is reduced cardioventilatory activity a sign of improved vol. 62. Society for Experimental Biology Seminar Series. Cambridge: animal welfare or the adverse effects of an opioid drug? PLoS ONE. Cambridge University Press; 1997. p. 35–72. 2014;9(4):e95283. 22. Iwama GK, Afonso LOB, Vijayan MM, Iwama GK, Afonso LOB, Vijayan MM. 44. Cubitt KF, Winberg S, Huntingford FA, Kadri S, Crampton VO, Øverli Ø. Stress in fishes. In: Ewans DE, Claiborne JB, editors. The physiology of Social hierarchies, growth and brain serotonin metabolism in Atlantic fishes. Boca Raton: CRC Press; 2006. p. 319–42. salmon (Salmo salar) kept under commercial rearing conditions. Physiol 23. Mulcahy DM. Surgical implantation of transmitters into fish. ILAR J. Behav. 2008;94(4):529–35. 2003;44(4):295–306. 45. Sloman KA, Metcalfe NB, Taylor AC, Gilmour KM. Plasma cortisol concen- 24. Bridger CJ, Booth RK. The effects of biotelemetry transmitter presence trations before and after social stress in rainbow trout and brown trout. and attachment procedures on fish physiology and behavior. Rev Fish Sci. Physiol Biochem Zool. 2001;74(3):383–9. 2003;11(1):13–34. 46. Oppedal F, Dempster T, Stien LH. Environmental drivers of Atlantic salmon 25. Martinelli TL, Hansel HC, Shively RS. Growth and physiological responses behaviour in sea-cages: a review. Aquaculture 2011;311(1–4):1–18 to surgical and gastric radio transmitter implantation techniques in 47. Hvas M, Folkedal O, Oppedal F. Heart rates of Atlantic salmon Salmo salar subyearling chinook salmon (Oncorhynchus tshawytscha). In: Advances in during a critical swim speed test and subsequent recovery. J Fish Biol. invertebrates and fish telemetry. Dordrecht: Springer; 1998. p. 79–87. 2020. https ://doi.org/10.1111/jfb.14561 . 26. Barton BA, Schreck CB. Metabolic cost of acute physical stress in juvenile 48. Eliason EJ, Higgs DA, Farrell AP. Postprandial gastrointestinal blood flow, steelhead. Trans Am Fish Soc. 1987;116(2):257–63. oxygen consumption and heart rate in rainbow trout (Oncorhynchus 27. Svendsen E, Føre M, Økland F, Gräns A, Hedger RD, Alfredsen JA, Uglem mykiss). Comp Biochem Physiol A Mol Integr Physiol. 2008;149(4):380–8. I, Rosten CM, Frank K, Erikson U, Finstad B. Heart rate and swimming 49. Gräns A, Albertsson F, Axelsson M, Olsson C. Postprandial changes in activity as stress indicators for Atlantic salmon (Salmo salar). Aquaculture. enteric electrical activity and gut blood flow in rainbow trout (Onco - 2020;531:735804. rhynchus mykiss) acclimated to different temperatures. J Exp Biol. 28. Brijs J, Sandblom E, Rosengren M, Sundell K, Berg C, Axelsson M, Gräns A. 2009;212(16):2550–7. Prospects and pit-falls of using heart rate to the welfare of rainbow trout 50. Johnsson JI, Höjesjö J, Fleming IA. Behavioural and heart rate responses (Oncorhynchus mykiss) in aquaculture. Aquaculture. 2019;509:188–97. to predation risk in wild and domesticated Atlantic salmon. Can J Fish 29. Hjelmstedt P, Sundh H, Brijs J, Ekström A, Sundell KS, Berg C, Sandblom Aquat Sci. 2001;58(4):788–94. E, Bowman J, Morgenroth D, Gräns A. Eec ff ts of prophylactic antibiotic- 51. Jensen LB, Wahli T, McGurk C, Eriksen TB, Obach A, Waagbø R, Handler treatment on post-surgical recovery following intraperitoneal bio-logger A, Tafalla C. Eec ff t of temperature and diet on wound healing in Atlantic implantation in rainbow trout. Sci Rep. 2020;10(1):1–9. salmon (Salmo salar L.). Fish Physiol Biochem. 2015;41(6):1527–43. 30. Kolarevic J, Aas-Hansen Ø, Espmark Å, Baeverfjord G, Terjesen BF, Damsgård B. The use of acoustic acceleration transmitter tags for moni- Publisher’s Note toring of Atlantic salmon swimming activity in recirculating aquaculture Springer Nature remains neutral with regard to jurisdictional claims in pub- systems (RAS). Aquacult Eng. 2016;72–73:30–9. lished maps and institutional affiliations. 31. Brijs J, Sandblom E, Axelsson M, Sundell K, Sundh H, Kiessling A, Berg C, Gräns A. Remote physiological monitoring provides unique insights on

Journal

Animal BiotelemetrySpringer Journals

Published: Jan 5, 2021

There are no references for this article.