Proceedings of the Royal Society B: Biological Sciences
You have accessResearch article

The benefit of being still: energy savings during winter dormancy in fish come from inactivity and the cold, not from metabolic rate depression

Ben Speers-Roesch

Ben Speers-Roesch

Department of Ocean Sciences, Memorial University of Newfoundland, St John's, Newfoundland and Labrador, Canada A1C 5S7

[email protected]

Google Scholar

Find this author on PubMed

,
Tommy Norin

Tommy Norin

Institute of Biodiversity, Animal Health and Comparative Medicine, University of Glasgow, Graham Kerr Building, Glasgow G12 8QQ, UK

Google Scholar

Find this author on PubMed

and
William R. Driedzic

William R. Driedzic

Department of Ocean Sciences, Memorial University of Newfoundland, St John's, Newfoundland and Labrador, Canada A1C 5S7

Google Scholar

Find this author on PubMed

Published:https://doi.org/10.1098/rspb.2018.1593

    Abstract

    Winter dormancy is used by many animals to survive the cold and food-poor high-latitude winter. Metabolic rate depression, an active downregulation of resting cellular energy turnover and thus standard (resting) metabolic rate (SMR), is a unifying strategy underlying the persistence of organisms in such energy-limited environments, including hibernating endotherms. However, controversy exists about its involvement in winter-dormant aquatic ectotherms. To address this debate, we conducted simultaneous, multi-day measurements of whole-animal oxygen consumption rate (a proxy of metabolic rate) and spontaneous movement in a model winter-dormant marine fish, the cunner (Tautogolabrus adspersus). Winter dormancy in cunner involved a dampened diel rhythm of metabolic rate, such that a low and stable metabolic rate persisted throughout the 24 h day. Based on the thermal sensitivity (Q10) of SMR as well as correlations of metabolic rate and movement, the reductions in metabolic rate were not attributable to metabolic rate depression, but rather to reduced activity under the cold and darkness typical of the winter refuge among substrate. Previous reports of metabolic rate depression in cunner, and possibly other fish species, during winter dormancy were probably confounded by variation in activity. Unlike hibernating endotherms, and excepting the few fish species that overwinter in anoxic waters, winter dormancy in fishes, as exemplified by cunner, need not involve metabolic rate depression. Rather, energy savings come from inactivity combined with passive physico-chemical effects of the cold on SMR, demonstrating that thermal effects on activity can greatly influence temperature–metabolism relationships, and illustrating the benefit of simply being still in energy-limited environments.

    1. Background

    The cold, food-poor winter of temperate to high latitudes creates a significant bottleneck on the poleward persistence of animals, and has led to the repeated occurrence of winter dormancy, a reversible seasonal phenotype characterized by inactivity, a low body temperature, fasting and a low metabolic rate [13]. A dormant overwintering strategy may facilitate the persistence of species at the cool limit of their range, including marine ectotherms [4], and may be viewed as a tactic to expand geographical ranges to the cold extreme of the thermal niche. However, the mechanisms underlying winter dormancy remain poorly understood, especially in ectotherms [5].

    Metabolic rate depression, a reversible and active downregulation of resting cellular energy turnover to well below the standard or basal (i.e. resting) metabolic rate (SMR or BMR; the baseline cost of living in ectotherms or endotherms, respectively), is a common strategy used by organisms to endure energy-limited environments [6,7]. In hibernating mammals, a profound metabolic rate depression is common and results from active depression of energy metabolism as well as passive Arrhenius physico-chemical effects of cooling due to a resetting of the body temperature set-point [7]. However, excepting when certain species encounter anoxic waters in winter (e.g. some freshwater turtles) [7], there is controversy about the use of metabolic rate depression by winter-dormant ectotherms, which typically overwinter under normoxic conditions [1,8]. In part, this controversy exists because dormancy and metabolic rate depression in ectotherms can be difficult to differentiate from lethargy and low metabolic rates resulting from passive physico-chemical effects of frigid temperatures [1].

    Biologists have used the thermal sensitivity (Q10) of metabolic rate over the transition from an active to dormant state as a tool to identify involvement of metabolic rate depression in winter-dormant ectotherms. A Q10 > 3.5 is thought to indicate an active depression of metabolic rate beyond the passive physico-chemical effects of temperature on metabolism where the typical Q10 is approximately 2–3 [7,9,10]. Such analyses have suggested considerable interspecific variation in the capacity for metabolic rate depression among winter-dormant ectotherms [1,11,12]. For example, among a diverse range of winter-dormant fish species, metabolic rate depression has been either implicated [10,1318] or excluded [9,19,20]. Among the latter species, winter dormancy has been suggested simply to be a period of inactivity [8,9]. Inactivity alone could lead to substantial decreases in measured metabolic rates because voluntary activity, which underlies fundamental behaviours such as foraging and patrolling territories, has been estimated to represent up to 67% of routine metabolic rate in fishes [21]. Indeed, activity is a significant component of daily energy expenditure in animals [22,23]. Thus, while never assessed in earlier studies on winter-dormant fishes, it is possible that high Q10 values for measured metabolic rates, traditionally interpreted as a metabolic rate depression (i.e. active downregulation of SMR), could be caused entirely by inactivity in the cold, which would greatly lower metabolic rate to resting levels (i.e. SMR) compared with warm, active individuals exhibiting routine levels of metabolic rate [24]. However, the roles of reduced activity versus metabolic rate depression in determining variation in metabolic rate in winter-dormant ectotherms have never been elucidated, in part because the relationships between metabolic rate and activity are challenging to measure, especially at frigid temperatures.

    A mechanistic understanding of the relationship between temperature, activity and metabolism in winter-dormant ectotherms is important: winter dormancy is a key phenology representing up to half of the lifespan, and thermal constraints on activity and metabolic budgets can have major ecological and life-history consequences [25,26]. Furthermore, metabolic theories of ecology depend upon the assumption of predictable thermal effects on metabolism, requiring a better understanding of the origins of variability in metabolic rate Q10 [5,27,28], including the potential, but typically overlooked, influence of activity [24].

    To investigate the puzzle of whether metabolic rate depression is involved in winter dormancy in fishes, we studied the cunner (Tautogolabrus adspersus), an abundant western North Atlantic wrasse. Like other temperate wrasses [16,29,30], cunner are winter-dormant: they seek refuge within the substrate and become inactive when the ocean cools below approximately 5°C in autumn, and emerge at approximately 5°C the following early summer [3133]. This winter dormancy in cunner has been associated with a large decrease in metabolic rate that occurs rapidly (within hours) below 5°C and is maintained over the winter [10,18]. The Q10 of metabolic rate over the transition from active to dormant temperatures has been reported to be greater than 10 in cunner, as in other winter-dormant wrasses [16], whereas at warmer active temperatures, the Q10 is between 2 and 3, a typical value for fishes [10,34]. Based on this, and consistent with simultaneous reductions in tissue protein synthesis [35] and suppression of appetite and digestion [33,36,37], metabolic rate depression has been implicated as a central component of winter dormancy in cunner. However, the possible role of behavioural modulation of metabolic rate in dormant fishes has been overlooked. Using cunner as a model, we investigated the hypothesis that the mechanism underlying the energy savings (i.e. low metabolic rate) of winter dormancy in fishes is not metabolic rate depression, but rather a behavioural reduction in activity. We carried out three experiments using automated optical respirometry to allow for multi-day, high-resolution monitoring of whole-animal oxygen consumption rate (Inline Formula; a proxy for metabolic rate) even at frigid temperatures. In experiment 1, we examined the influence of acute exposure to low winter temperature on the diel cycle of metabolic rate. In experiment 2, we examined the effect of acute exposure to darkness and low temperature, which are characteristic of the winter refuge, on the diel cycle of metabolic rate and spontaneous activity (measured simultaneously). In experiment 3, we investigated whether chronic acclimation to low temperature can trigger a metabolic rate depression. If metabolic rate depression is involved in winter dormancy, we predicted that the thermal sensitivity (i.e. Q10) of metabolic rate would remain high at all times when cooled below approximately 5°C, including when fish are at rest (i.e. at their SMR at night, as cunner are active during the day [32]). Alternatively, if reduced activity explains energy savings under winter dormancy, then the thermal sensitivity of metabolic rate during resting periods would indicate physico-chemical effects alone (Q10 ≈ 2–3) regardless of acute or chronic cold exposure and, in experiment 2, variation in activity would largely explain variation in metabolic rate.

    2. Material and methods

    (a) Animals

    Adult cunner of mixed sexes were captured with hoop traps in summer 2013 in Conception Bay (47°37′42″ N, 52°51′31″ W), Newfoundland, Canada. The fish were transferred to holding tanks at the Ocean Sciences Centre (OSC), Memorial University of Newfoundland, supplied with flow-through, temperature-controlled seawater (8–10°C) and exposed to a winter photoperiod (11 L : 13 D). The fish were fed to satiation once a week with chopped herring.

    Juvenile cunner of mixed sexes were the 2013 offspring of wild-caught parents from Placentia Bay (47°42′47″ N, 53°58′06″ W) and Conception Bay, Newfoundland. Spawning, hatching and rearing occurred at the OSC at 15°C and 12 L : 12 D photoperiod. Three months prior to experiments, juveniles were transferred to holding tanks, supplied with flow-through, temperature-controlled seawater (8–10°C) under a winter photoperiod (11 L : 13 D), and fed dry pellets (Gemma; Skretting, St Andrews, NB, Canada).

    An 11 L : 13 D photoperiod was used throughout the study because it occurs in southeastern Newfoundland, when cunner are active but preparing to enter dormancy (October; ocean temperature: approx. 9°C and cooling) or in winter dormancy (February; approx. 0°C) [3133]. Experiments were conducted between December 2014 and April 2015, within the normal Newfoundland dormancy period (November–June) [31].

    The fish were fasted beginning one week (adults) or 2 days (juveniles) before experiments until the end of experiments, to ensure a post-absorptive state and because winter dormancy involves minimal appetite and feeding [36].

    (b) Measurement of metabolic rate

    Metabolic rate of the fish was estimated by measuring their Inline Formula using automated intermittent-closed optical respirometry and best practices [3840]. Briefly, we used 2.6 l polypropylene respirometers for adults and 145 ml glass respirometers for juveniles. Inline Formula was calculated from the continuously measured oxygen decline at 0.5 Hz in the well-mixed respirometer water over 40 min periods irrespective of temperature, interspersed by 5 min flush periods. Further details are in the electronic supplementary material.

    (c) Experiment 1: the influence of temperature on diel variation in metabolic rate

    In experiment 1, we examined how low temperature affects the diel pattern of metabolic rate and compared thermal sensitivity of metabolic rate during day versus night. We carried out separate measurements on adult cunner (body mass: 98.7 ± 11.4 g, n = 9) and juvenile cunner (2.17 ± 0.06 g, n = 8) to reveal whether life stage influences the capacity for metabolic rate depression during winter-dormant conditions. Each respirometry trial first involved moving the fish the night before the experiment began from holding tanks to the respirometers. The water surface above the respirometers was covered with black plastic to prevent visual disturbances. The fish settled in the respirometers overnight at a normal, ‘warm’ active temperature (5.6 ± 0.2°C; all experimental temperatures in the Material and methods section are given as mean ± s.d.; figure 1a), and then were held in the respirometers over a 2-day period at the active temperature and 11 L : 13 D while we continuously monitored metabolic rate. On the third day, we cooled the fish starting at noon at 1°C h−1 until a stable dormant temperature (0.6 ± 0.2°C) was reached. The dormant temperature was maintained overnight and then for a further 2-day period. At noon on the sixth day, a subset of the fish (n = 5) was warmed at 1°C h−1 to an active temperature (5.9 ± 0.5°C) to confirm that the diel rhythm of metabolic rate was re-established the following (seventh) day and night. The experiment was then terminated and the fish were weighed and returned to holding tanks. The trial for juveniles was the same as that for adults, except for slightly different exposure temperatures (active temperature, 6.7 ± 0.08°C; dormant temperature, 1.4 ± 0.06°C), one (not two) diel cycle was measured at each temperature (see supporting data for time course), and fish were not returned to active temperature.

    Figure 1.

    Figure 1. The influence of acute temperature change on diel variation of metabolic rate (Inline Formula) in cunner, a winter-dormant fish. Cunner are active at the warm temperatures used, while the cold temperatures are typical of winter conditions and below the threshold at which cunner enter into a dormant state. (a) Diel cycle in Inline Formula of adult cunner (n = 9) held under a winter photoperiod (11 L : 13 D) and sequentially exposed to a warm, a cold and finally the same warm temperature (n = 5 for return to warm, because only a subset of fish were re-warmed). Data for each temperature treatment are overlapped. White segments indicate daytime periods with lights on and grey segments indicate night-time periods with lights off. The black arrow indicates when fish were placed in respirometers. (b) Average Inline Formula and Q10 values of the adult cunner (i.e. averages of data shown in a). (c) Average Inline Formula and Q10 of juvenile cunner (n = 8). All data are means ± s.e.m. Different lower case letters in (b,c) represent significant differences in Inline Formula between groups, while an asterisk indicates a significant difference in Q10. (Online version in colour.)

    (d) Experiment 2: influence of temperature and darkness on metabolic rate and activity

    In experiment 2, we examined how low temperature combined with darkness influenced the diel pattern of metabolic rate and spontaneous activity measured simultaneously. Metabolic rate of adults (71.1 ± 3.8 g, n = 8) was monitored as in experiment 1, except black plastic did not cover the water in order to video-record fish activity. The activity (i.e. spontaneous movement; see Data analysis) of the cunner inside the respirometers was monitored continuously using an infrared-sensitive digital video recording system (512 × 384 p, 30 fps; Peep-A-Roo Video Probes, Sandpiper Technologies, Manteca, CA, USA). Each infrared camera recorded two respirometers. Infrared illumination was provided by the cameras (940 nm) and by two illuminators (850 nm). Whereas the cunner noticed a human observer appear above its respirometer under visible red LED light, as indicated by eye and postural responses, there was no such indication that they could see under infrared illumination (B.S.-R. 2015, personal observation). After transfer to the respirometers, the fish settled overnight at a ‘warm’ active temperature (approx. 6.2°C). Metabolic rate and activity was then monitored at the warm temperature (6.2 ± 0.1°C) for 2 days under a winter photoperiod (11 L : 13 D), and then for another 2 days under constant darkness (0 L : 24 D). On the fourth day at 18.00, under constant darkness, we started cooling the fish over 2 days by approximately 1.5°C every 14 h until a stable dormant temperature (1.0 ± 0.1°C) was reached (by 18.00 on the sixth day). We used a longer cooling period than in experiment 1 to verify that the cooling period used there was not too short to stimulate potential metabolic rate depression. The dormant temperature was maintained for a 2-day period under constant darkness. On the morning of the ninth day, room light was returned to 11 L : 13 D before terminating the experiment that evening.

    (e) Experiment 3: influence of low-temperature acclimation on metabolic rate

    In order to determine if metabolic rate depression takes longer to manifest in winter-dormant fish, we exposed a subset of fish (n = 6) from experiment 1 to a further four weeks of winter-dormant conditions (0.5–1.5°C, 11 L : 13 D) in individual plastic holding containers containing a shelter and submerged in a larger fibreglass tank (initial fish weight: 94.2 ± 13.7 g; weight after four weeks: 91.0 ± 12.6 g; n = 6; paired t-test, p = 0.06). After four weeks, we repeated measurements of metabolic rate at the same dormant temperature at which the original measurements were made (0.6 ± 0.2°C) and under a 11 L : 13 D photoperiod. Activity of these fish was not measured.

    (f) Data analysis and statistics

    Data for Inline Formula were analysed using an automated Mathematica (Wolfram; Champaign, IL, USA) script that extracted the slopes of the oxygen decline inside the respirometers during each measurement period over the multi-day trials, and then calculated Inline Formula after accounting for the volume of the respirometer, body mass of the fish and background respiration. Certain Inline Formula measurements were excluded from analysis because the fish had temporarily rested upon the optode during the measurement period, causing obvious interference.

    Activity was quantified from recorded videos as spontaneous movement by manually measuring the time spent moving during each time period over which metabolic rate was measured. Activity is presented as seconds moving per minute. Cunner locomotion involves periods of stationary rest (often on the bottom) interspersed by movement periods in which the pectoral (in particular), caudal and sometimes dorsal fins are engaged for thrust. Thus, movement was identified as the start and end of any change of body and head position that involved associated movement of any or all fins, independent of distance travelled. Under this criterion, minor postural adjustments were excluded. To account for the potential time lag between movement and detection of a change in Inline Formula, measurement of movement was offset from the metabolic rate measurement period by −30 s, a time period in which there was circulation of approximately one respirometer volume (2.6 l).

    For adult fish in experiment 1, Inline Formula during day and night at warm and cold temperatures was taken for each fish as the average Inline Formula across the two daytime periods and the second two night-time periods (the first night, a settling period after placement in the respirometer, was excluded). The effect of diel period (night and day) and temperature on metabolic rate was tested with a two-way analysis of variance (ANOVA) with Bonferroni's post hoc test. The same analysis was conducted for the juveniles, except that the Inline Formula values were taken as the average across the single night and day measured at each temperature. Thermal sensitivity of the day-time and night-time Inline Formula values in adults and juveniles was calculated using the temperature (T) coefficient: Inline Formula. The day-time and night-time Q10 values were tested for a significant difference using a Student t-test.

    For experiment 2, the average Inline Formula for each fish was taken within each of the experimental combinations of true diel period (true day or true night), temperature and experimental photoperiod as explained in figure S1 of the electronic supplementary material, and displayed in figure 2b. Pairwise comparisons were made using a Tukey honest significant difference test. Inline Formula Q10 under each experimental condition was calculated. The relationship between spontaneous movement and Inline Formula was examined at each temperature using a linear regression and Q10 of the y-intercept (i.e. extrapolated Inline Formula at zero activity) was calculated.

    Figure 2.

    Figure 2. The influence of acute temperature change and darkness on the spontaneous activity and metabolic rate (Inline Formula) of cunner, a winter-dormant fish. (a) Spontaneous activity and Inline Formula of cunner (n = 8) measured simultaneously over approximately 8 days and sequentially exposed to warm water under a normal winter photoperiod (11 L : 13 D), warm water in total darkness (0 L : 24 D), cold water in total darkness and cold water after return to the normal winter photoperiod. White segments indicate true daytime periods with lights on, grey segments indicate true night-time periods with lights off, and the hatched grey segments represent the circadian (11 L : 13 D) daytime periods with lights off. (b) Average Inline Formula of the same cunner under each experimental photoperiod during true day (white area) or true night (grey area) (see electronic supplementary material, figure S1 for details on calculations). Different lower case letters above bars represent significant differences between groups within either day or night. Asterisks represent significant differences between day and night for the same groups (e.g. 11 L : 13 D during the day versus 11 L : 13 D during the night). (c) Relationships between spontaneous activity and Inline Formula for the same cunner. Shaded areas around regression lines are 95% CIs and regression equations are provided in the panel. All data are means ± s.e.m. (Online version in colour.)

    For experiment 3, the average Inline Formula values for each four-week acclimated fish during daytime and night-time periods were compared with the daytime and night-time Inline Formula values from the same fish during the first 0–2 days exposure to cold. We used a two-way ANOVA with Bonferroni's post hoc test to test the effect of acclimation time and diel period on Inline Formula.

    All data are presented as means ± s.e.m., unless otherwise noted. Statistical analyses were done using GraphPad Prism v.5.0b or R v.3.4.3. In cases where assumptions of normality or equal variance were not met, we square root transformed data before analysis. Statistical significance was accepted at p < 0.05.

    3. Results

    (a) Experiment 1: the influence of temperature on diel variation in metabolic rate

    Adult cunner at the warm temperature (5.6°C) showed a strong diel rhythm in Inline Formula, with elevated, variable rates during the day when the fish were more active and a stable, significantly lower rate at night when the fish were inactive (figure 1a,b). After acute cooling to 0.6°C, the diel rhythm was dampened such that Inline Formula was no longer significantly different between day and night (figure 1a,b). The average daytime Inline Formula was more thermally sensitive than the average night-time Inline Formula, as indicated by a significantly higher Q10 (5.1) for the daytime comparison between the warm and cold compared with the night-time comparison (Q10 = 2.8). A similar result was found for juvenile cunner (figure 1c).

    (b) Experiment 2: influence of temperature and darkness on metabolic rate and activity

    As found in experiment 1, cunner at the warm temperature (6.2°C) under a normal winter photoperiod (11 L : 13 D) showed a strong diel rhythm in Inline Formula (figure 2a), with significantly higher average Inline Formula during the day compared with at night (figure 2b; see electronic supplementary material, figure S1 for detailed explanation). When the lights were then turned off (0 L : 24 D) at the warm temperature (6.2°C), the constant darkness was associated with a dampening of the diel rhythm such that Inline Formula during day and night (circadian clock-time from the previously experienced 11 L : 13 D photoperiod) were not significantly different (figure 2a,b). Following acute cooling to 1.0°C in constant darkness, Inline Formula significantly decreased to a new steady state that was unchanged through several circadian cycles (figure 2a,b). When the lights were turned back on while the fish remained cold (return to 11 L : 13 D), there was a significant increase in daytime Inline Formula that peaked in the morning and returned to baseline levels as the evening approached (figure 2a,b). Table 1 shows the Q10 values for Inline Formula under each photoperiod and circadian clock (day versus night) comparison. The salient finding from calculating Q10 values was that thermal sensitivity was high (Q10 ≈ 7.2) when Inline Formula was compared between the warm and light (6.2°C, 11 L : 13 D) fish versus the cold and dark fish (1.0°C, 0 L : 24 D), but relatively low (Q10 ≈ 2.8 to 3.2) when Inline Formula was compared between the warm and dark (6.2°C, 11 L : 13 D at night, or under 0 L : 24 D) fish versus the cold and dark fish (table 1).

    Table 1.The effect of experimental photoperiods on thermal sensitivity (Q10) of mean metabolic rates during exposure to acute temperature change (6.2 to 1.0°C) in cunner, a winter-dormant fish. The average metabolic rate values used to calculate Q10 are displayed in figure 2b. For example, a Q10 of 7.2 was found when comparing the average metabolic rate for day, 11 L : 13 D at 6.2°C versus the average metabolic rate for day, 0 L : 24 D at 1.0°C.

    6.2°C
    day, 11 L : 13 D day, 0 L : 24 D night, 11 L : 13 D night, 0 L : 24 D
    1.0°C
     day, 0 L : 24 D 7.2 4.0 3.1 2.8
     day, return to 11 L : 13 D 2.9 1.6 1.3 1.1
     night, 0 L : 24 D 7.2 4.0 3.1 2.8
     night, return to 11 L : 13 D 7.3 4.1 3.2 2.8

    The changes in Inline Formula were mirrored by changes in spontaneous activity, with high activity in warm and light conditions, very little activity in warm and dark conditions regardless of circadian clock-time and virtually no activity in cold and dark conditions regardless of clock-time (figure 2a). There was a strong, linear relationship between activity and Inline Formula in both warm and cold conditions across all light treatments (figure 2c). Thus, variation in metabolic rate was almost entirely explained by variation in activity (r2 ≥ 0.90 at both temperatures; figure 2c). The extrapolated Inline Formula at zero activity (i.e. y-intercept) was 20.3 mg h−1 kg−1 in the warm and 11.7 mg h−1 kg−1 in the cold, giving a Q10 of 2.9 for extrapolated SMR (figure 2c).

    (c) Experiment 3: influence of low-temperature acclimation on metabolic rate

    After four weeks of acclimation to a winter cold temperature (0.5–1.5°C), the metabolic rate of cunner measured during day and night at 0.6 ± 0.2°C was not significantly different from that measured during the initial 2 days of exposure at the same temperature (figure 3).

    Figure 3.

    Figure 3. Changes in daytime and night-time metabolic rates (Inline Formula) of cunner (n = 6) at 0.6 ± 0.2°C, after four weeks exposure to winter dormancy conditions (0.5–1.5°C, 11 L : 13 D photoperiod). Data are presented as the percentage of the Inline Formula measured in the same fish during the initial 0–2 days of exposure to 0.6 ± 0.2°C. There were no significant changes in Inline Formula after four weeks of dormancy. Data are means ± s.e.m.

    4. Discussion

    Our salient discovery was that the low metabolic rate that is characteristic of winter dormancy in cunner does not arise from metabolic rate depression, but rather from the striking reduction in activity between active and dormant states combined with the passive physico-chemical effects of cold on SMR. When variation in spontaneous movements between active and dormant states is controlled for by making comparisons of metabolic rate during night when activity was minimal (figures 1b,c and 2b), or when activity was known to be zero in both states (figure 2b,c), the acute thermal sensitivity (Q10) of metabolic rate of 2.8–3.2 showed typical passive thermal effects on metabolism (Q10 ≈ 2–3 [5]). By contrast, Q10 values were elevated (4.8–7.2) only when comparing periods where the fish were active in the warm but inactive in the cold (e.g. during daytime; figures 1 and 2). Thus, it is likely that previous reports of highly thermally sensitive metabolic rates (i.e. elevated Q10 values) with cooling to winter temperatures in cunner, and possibly other winter-dormant fishes, were in fact caused by an arrest of activity in the cold, which led to a decrease in metabolic rate to resting levels that was misinterpreted as metabolic rate depression because activity was never monitored [10,1318].

    Correlating metabolic rate and spontaneous movements is onerous and, consequently, our study is one of few that have directly investigated this relationship in fishes [4144], despite the importance of voluntary movements and associated energy expenditure in day-to-day life [22,45]. Thus, winter-dormant fishes put life on hold by sacrificing continued activity in order to minimize energy expenditure during winter, similar to hibernating mammals [1,46]. The energy savings of being inactive, combined with the still considerable slowing of metabolism resulting from passive effects of cold (Q10 ≈ 2–3), are probably sufficient to match energy demand to lowered energy supply in winter, obviating the requirement to evolve the more complex response of metabolic rate depression. In fact, after four weeks of dormancy and associated fasting in cunner, we found that the mean body weight only decreased (non-significantly) by 3% (see the Material and methods section). Compared with metabolic rate depression, inactivity is a more flexible way to accrue large energy savings. Essentially, the energy savings of winter dormancy come about because activity is minimized at cold temperatures, especially in the darkness typical of the winter refuge, and thus SMR persists continuously (figure 2a,b).

    The possibility of an uncaptured scope for metabolic rate depression beyond our measurements is remote. Metabolic rate of the cunner remained unchanged after chronic cold acclimation, compared with acute cold exposure (figure 3), contradicting the possibility of a slower initiation of metabolic rate depression as seen in certain estivating amphibians [12]. However, the absence of cold compensation of SMR, also observed in the reportedly winter-dormant northern killifish [47], is consistent with a strategy of minimizing energy expenditure during dormancy. Our study fish were held under winter photoperiod, and remained active and feeding as long as they were warm, so photoperiod does not trigger dormancy or a metabolic rate depression. The values for metabolic rate we found are similar to those previously measured in juvenile and adult cunner, even after several months of winter acclimation [10,18], suggesting that we captured the annual nadir of SMR in cunner.

    (a) Inactivity as the central mechanism of energy savings during winter dormancy in fishes

    The notion that reduced activity explains low metabolic rates in winter-dormant fishes originated in early work by Crawshaw and co-workers [9,19,20]. They found typical thermal sensitivities for metabolic rates (Q10 ≈ 2–3) in largemouth bass and brown bullhead catfish when cooled to winter temperatures, concluding that metabolic rate depression is not involved in these species during winter dormancy. Crawshaw and co-workers recognized that cold-induced decreases in activity were characteristic of winter dormancy, but they were cautious about specifying its contribution to reductions of metabolic rate [9]. In part, this reflected the greater logistical challenges at that time of making simultaneous measurements of metabolic rate and activity, as well as the fact that they seem to have serendipitously measured metabolic rate under relatively inactive periods so never saw substantial variation in metabolic rate. Additionally, they calculated Q10 values over wider temperature ranges, which can obscure more sensitive periods that may occur when the fish transitions across the active–dormant threshold temperature. In any case, their results have been widely interpreted as being species-specific responses that contrasted with the numerous other winter-dormant fishes where metabolic rate depression has been implied on the basis of relatively high thermal sensitivity of metabolic rate at low temperatures [10,1318]. However, the thermal sensitivity of SMR in cunner is no different from that in the largemouth bass or catfish studied by Crawshaw and co-workers [9,19,20] and, combined, our studies suggest the absence of metabolic rate depression in winter-dormant fishes.

    (b) The influence of darkness on activity and metabolic rate of winter-dormant fishes

    The persistent inactivity in winter-dormant cunner is induced by low temperatures, but an additional key environmental influence to consider is the darkness that is characteristic of the winter refuge buried within the substrate. Wrasses, including cunner, are diurnal fishes with distinct day–night cycles of activity and sleep-like behaviour at night [48]. Typical of day-active fishes [49,50], cunner under warm conditions showed a diel rhythm of elevated metabolic rate during the active daytime and stable low metabolic rate during night-time rest (figures 1a and 2a). Low temperature greatly dampens these diel cycles (figure 1a) such that day- and night-time metabolic rates are similar under winter-dormant conditions (figure 1b,c). To our knowledge, this is the first demonstration that low temperature can abolish diel cycles in activity and metabolic rate in fish. Perpetual darkness (0 L : 24 D) at warm temperature has the same effect, similar to other wrasse species [51], though more variability in rates is visually apparent (figure 2a,b). The combination of cold and dark resulted in a particularly marked flatline in activity and metabolic rate (figure 2a). Although darkness has been implicated as a metabolic signal in hibernating mammals [52] and possibly reptiles [53], darkness is not a trigger for metabolic rate depression in winter-dormant cunner; metabolic rate in the cold and dark did not show elevated Q10 when compared with warm inactive periods (figure 2b and table 1). Interestingly, even after 2 days of cold and darkness, cunner remain vigilant of light as they responded to reappearance of light by becoming temporarily active and thus increasing metabolic rate (figure 2a). Thus, although low temperature alone can trigger the behavioural and metabolic phenotype of winter-dormant cunner, the darkness of the refuge may be an additional cue that further minimizes fluctuations in activity and thus metabolic rate during winter dormancy. Vigilance to light occurs even in profoundly metabolically depressed hibernating turtles [54], and may be a general attribute of winter-dormant animals to detect changes in seasonal photoperiod in preparation for exit from dormancy. The influence of darkness may be particularly important for winter-dormant fishes in polar regions [17], where the seasonal change in photoperiod is massive while temperature changes marginally.

    The similar effects of cold and darkness on activity and metabolic rate in cunner hint that winter dormancy, even in the absence of darkness, is a sleep-like state. Winter-dormant cunner are sluggish and initially non-responsive to handling, similar to when cunner and other wrasses are ‘sleeping’ at night at warm temperature [48]. Thus, we speculate that winter dormancy in temperate wrasses evolved from the distinctive night-time sleeping behaviour of their tropical relatives. In fact, certain subtropical wrasse species respond to the seasonal cooling that occurs for short periods (up to approx. one month) in their environment by apparently burying within the substrate [55]. The traits that promote poleward expansion of tropical fish lineages are poorly understood [56]; a cold-induced dormancy marked by inactivity and its energy savings may be a common response among wrasses that has facilitated the expansion of this speciose family into temperate regions.

    (c) The influence of activity on estimates of thermal sensitivity of metabolic rate

    The assumption of a predictable, universal thermal sensitivity of metabolic rate among animals is central to several physiological and ecological theories, which are broadly applied, for example, to model the impacts of climate change [5,27,57,58]. Yet Q10 values for metabolic rate (as well as other biological rates) vary considerably, while the origin of this variation remains incompletely understood [28]. We have directly demonstrated how thermal modulation of activity levels can lead to elevated metabolic rate Q10 values that reflect massive energy savings, at the same time that the thermal sensitivity of SMR conforms to typical Q10 values used in metabolic theories of ecology (Q10 ≈ 2–3). This is an important distinction as voluntary movements can be more easily adjusted than SMR; also, the scaling of energy demand differs for locomotion versus SMR [59]. Thus, elevated Q10 values for metabolic rates, which have been reported in many overwintering ectotherms (e.g. [1018,60,61]) as well as more widely (e.g. [6264]), should be interpreted carefully and with an understanding of the origins of the variation. Measured thermal sensitivities of metabolic rate can be confounded if thermal sensitivity of activity changes disproportionately; for example, across the transition from active to dormant states. This is an important consideration when building mechanistic predictive models of animal responses to thermal variability and illustrates the value of exploring the role of behavioural influences in physiological studies [24,25].

    (d) Perspectives

    Inactivity is a major part of daily time budgets of animals, and it may or may not have a functional benefit [65]. Our study supports the energetic benefit of being inactive, especially in resource- or performance capacity-limited environments such as winter. Similarly, reductions in activity probably benefit energy balance for hypoxia survival in fishes [21]. The simple behavioural response of being still for prolonged periods may be a common strategy for survival in energy-limited environments among fishes. Metabolic rate depression, which is absent in cunner and possibly other fishes during winter dormancy, is a more complex response that may be selected for only in environments that present immediate and/or grave challenges to energy supply, such as severe hypoxia or desiccation. Prolonged inactivity can trade-off against fitness-related performance such as growth and reproduction, but presumably represents an optimal strategy in conditions where activity becomes risky due to resource lack and/or poor performance.

    Ethics

    All experimental procedures were approved by the animal care committee of Memorial University of Newfoundland in compliance with Canadian Council of Animal Care guidelines.

    Data accessibility

    Supporting data available as part of the electronic supplementary material.

    Authors' contributions

    B.S.-R. conceived of and designed the study, with input from T.N. and W.R.D. B.S.-R. and T.N. carried out the experimental work and analyses. B.S.-R. drafted the manuscript. T.N. and W.R.D. provided editorial input. All authors gave final approval for publication.

    Competing interests

    We declare we have no competing interests.

    Funding

    Funding was provided by an NSERC Postdoctoral Fellowship and an NSERC Discovery grant to B.S.-R., an NSERC Discovery grant to W.R.D., and a Danish Council for Independent Research Individual Post-doctoral grant (DFF-4181-00297) and a Danish Council for Independent Research Individual Post-doctoral grant (DFF-4181-00297) to T.N. Respirometers were supplied with funding from an NSERC Discovery grant to Kurt Gamperl.

    Acknowledgements

    We thank Dr Kurt Gamperl for lending respirometers, Ms Connie Short and Ms Kathy Clow for fish husbandry, and Ms Michelle Fitzsimmons and Dr Anne Storey for lending infrared cameras.

    Footnotes

    Electronic supplementary material is available online at https://dx.doi.org/10.6084/m9.figshare.c.4203116.

    Published by the Royal Society. All rights reserved.