Acclimation to tidal conditions alters the physiological responses of the green shore crab, Carcinus maenas, to subsequent emersion

The intertidal zone is defined as the area between the high and low water marks during the maximal spring tides (Levinton, 2009). Organisms that inhabit the intertidal zone almost exclusively originate from the marine environment and, consequently, one of the major challenges that they face is the shift from an aquatic to an aerial respiratory medium. Despite the fact that intertidal zones are highly dynamic environments, they support a rich flora and fauna. The exposure of these organisms to large scale and rapid environmental changes, coupled with their ease of access, has resulted in intertidal organisms being an important group for studying physiological responses to environmental change (Somero, 2002).

The intricate nature and complexity of the equipment required to perform physiological experiments makes it difficult to conduct studies in the field, and thus the majority of experiments that have investigated the effects of air exposure on the physiology of intertidal organisms have been carried out in a laboratory setting. For convenience, almost all of these studies maintained intertidal organisms in aquaria in permanently submerged conditions prior to experimentation, rather than a routine cycle of immersion and emersion as they would typically experience in nature (e.g. Burnett, 1988; Burnett et al., 2002; DeFur and McMahon, 1984; Lagos et al., 2014; McGaw et al., 2009; Rastrick et al., 2014). Although a period of acclimation to constant conditions in the laboratory is sound scientific practice, evidence suggests that intertidal animals acclimated to a regular cycle of immersion–emersion exhibit different physiological responses when exposed to emersion from those maintained in permanently submerged conditions. For example, the blue mussel, Mytilus edulis, is a temperate, predominantly intertidal species with a habitat range that includes the shallow subtidal zone. In an in situ study, blue mussels were reciprocally transplanted between the intertidal and subtidal zones over a 7 week period, following which tolerance (survival) of emersion was determined in a laboratory experiment. Intertidal individuals were found to have higher survival rates during air exposure than those from the subtidal zone (Altieri, 2006). Moreover, tolerance to emersion was lost in M. edulis from the intertidal zone when acclimated to subtidal conditions, and vice versa. This suggests that the responses of intertidal animals acclimated to permanently submerged conditions in the laboratory may not reflect the physiological responses that occur in situ. Recognition of this phenomenon has attracted interest, and subsequently there are some studies that have acclimated individuals to simulated tidal conditions in the laboratory (e.g. Altieri, 2006; Dong and Williams, 2011; Drake et al., 2017; Han et al., 2013; Jimenez et al., 2016; Marshall and McQuaid, 1992; McMahon et al., 1991; Paganini et al., 2014; Widdows and Shick, 1985; Yin et al., 2017). Most of these studies investigated how tolerance to regular air exposure affects the subsequent ability of organisms to cope with either hypoxia or elevated temperature stress. There are only a few studies that have investigated how acclimation to a simulated tidal cycle affects an intertidal animal's response to subsequent air exposure (Drake et al., 2017; Weinrauch and Blewett, 2019; Widdows and Shick, 1985; Yin et al., 2017), and these indicate that the response to emersion is plastic. For example, routinely exposing bivalves to air results in an increase in scope for growth during air exposure (Widdows and Shick, 1985) and an upregulation of antioxidant enzymes (Yin et al., 2017). Such studies are enhancing our understanding of the influence of acclimation history on the physiological capacity of intertidal organisms, as they combine the benefits of controlled laboratory experiments with ecologically realistic field observations. However, our knowledge of the effect of routine air exposure on the physiology of mobile intertidal invertebrates, such as sea stars and crabs, is still lacking.

Small decapod crustaceans are common inhabitants of the intertidal zone. The green shore crab, Carcinus maenas, is native to Europe, but has become a global invasive species, where it is found in sheltered bays and estuaries primarily from the mid-intertidal down to the shallow subtidal zone (5–6 m) (Klassen and Locke, 2007). Carcinus maenas is considered an effective air breather (DeFur, 1988; Dejours and Truchot, 1988; Depledge, 1984; Newell et al., 1972; Taylor et al., 1973; Taylor and Butler, 1973, 1978; Wallace, 1972). However, reports on the extent to which aerial respiration can be maintained vary significantly, ranging from 50% (Simonik and Henry, 2014) to 120% (Taylor and Butler, 1978) of aquatic oxygen consumption values. In addition, some articles also report a significant increase in lactate production in C. maenas during short-term air exposure, suggesting an anaerobic component (Santos and Keller, 1993; Simonik and Henry, 2014), while others report that respiration is fully aerobic and no lactate build-up occurs during emersion (Johnson and Uglow, 1985; Taylor and Butler, 1978). At the onset of emersion, C. maenas typically experiences a 50–75% decline in the partial pressure of arterial oxygen (Pa_O2) (Dejours and Truchot, 1988; Depledge, 1984; Taylor and Butler, 1978), despite an increased availability of oxygen in air. Moreover, because of the difficulty in excreting CO₂ in air, C. maenas experiences an increase in P_CO2 levels which results in a decrease in haemolymph pH (Truchot, 1975; Taylor and Butler, 1978), which is sustained throughout the duration of emersion (Burnett, 1988).

Despite being a prominent model species in understanding the effects of emersion on crustacean physiology, previous researchers have maintained C. maenas (prior to experimentation) in conditions more representative of subtidal rather than intertidal conditions (e.g. Dejours and Truchot, 1988; Depledge, 1984; Newell et al., 1972; Taylor et al., 1973; Taylor and Butler, 1973, 1978; Wallace, 1972). The influence of acclimation to tidal conditions on the physiological responses reported in other intertidal species (Altieri, 2006; Jimenez et al., 2016; Drake et al., 2017; Widdows and Shick, 1985; Yin et al., 2017), coupled with the disparity between physiological responses reported for this species, suggests that our understanding of C. maenas’ physiological response to emersion may not be accurate. Therefore, the focus of this study was to determine how acclimation to constant submergence (referred to as ‘non-tidal’) versus a simulated semi-diurnal tidal cycle (referred to as ‘tidal’) influences the physiological responses of C. maenas to emersion.

Large adult male Carcinus maenas (Linnaeus 1758) (carapace width 50–80 mm, 56–105 g) were collected at Fox Harbour (47.3209°N, 53.9082°W), Long Harbour (47.4324°N, 53.8162°W) and Fairhaven (47.5343°N, 53.8998°W), NL, Canada, between June and November using dome net crab traps. Females were not used because of permit regulations in Newfoundland. Only individuals in intermoult stage with no carapace damage or missing chelae were used in experiments. The crabs were transported to the Department of Ocean Sciences, Memorial University, and placed in an aerated flow-through seawater holding tank (salinity 31±0.3‰; oxygen >95% saturation) at ambient temperatures (approximately 3–12°C). The crabs were fed mackerel once a week and any uneaten food was removed after 2 days. Perforated plastic pipes (diameter 10 cm, length 20 cm) were placed in the holding tanks to provide shelter and discourage aggressive behaviour amongst conspecifics.

Prior to experiments, the crabs were transferred to two separate flow-through seawater tanks (155 cm×95 cm×50 cm deep, 1000 l, with the same light source and water source via a splitter) where they were held in perforated containers (with lids) (37 cm×40 cm×17.5 cm deep, 30 l) at 15±0.5°C and a salinity of 31±0.3‰. An air stone was placed in each container to ensure that the oxygen content of the water was close to air saturation (above 95%). Throughout the acclimation period and experiments, temperature, salinity and oxygen content were monitored at least twice daily and consistently remained within the ranges stated here. The crabs were maintained in these tanks in either non-tidal (control) or tidal (experimental) conditions for at least 4 weeks prior to experimentation. The non-tidal tank received unfiltered seawater at 2.2 l min⁻¹, and the perforated containers which held the crabs in this tank were permanently submerged. The tidal tank had a similar seawater supply, but the water level in this tank was controlled so that it simulated a semi-diurnal 6 h tidal regime (i.e. alternating periods of 6 h of immersion and 6 h of emersion). A semi-diurnal tide was deemed the most appropriate as it is the most common tidal pattern in their natural environment (Little and Kitching, 1996). This was achieved by manipulating the water level via a timer-controlled solenoid valve connected to the outflow. When turned off, the solenoid valve closed the outflow, allowing the tank to slowly fill up with seawater and submerge the crabs. When turned on, the valve opened the outflow, allowing the water to gradually drain, exposing the crabs to air. The water temperature of both tanks (15±0.5°C) was maintained via an in-tank heater and salinity remained constant at 31±0.3‰. The air temperature was maintained at 15±0.5°C using an air conditioning unit so that it was similar to that in the water in the tanks. A temperature of 15°C was used as it is the optimum temperature for growth and physiological processes in C. maenas (Robertson et al., 2002). Water oxygen content was maintained above 95% saturation using air stones and the humidity during emersion was monitored using a hygrometer (11-661-16, Fisher Scientific), and varied between 70% and 80%. Animals were held in constant dim red light to reduce any endogenous cycles associated with light change. Prior to experimentation, the crabs were fasted for 3–4 days to ensure that digestive processes did not affect any of the measured physiological parameters (McGaw, 2006; Robertson et al., 2002). Individual crabs from the non-tidal and tidal group were matched as closely as possible with respect to size, mass and colour morph (Styrishave et al., 2004).

To establish whether acclimation to a cycle of emersion entrained any endogenous rhythms (De la Iglesias and Hsu, 2010; Palmer, 2000), oxygen consumption rate was measured for both the non-tidal and tidal crabs (n=12 per group) during 72 h of immersion (100% air saturation) at a salinity of 32‰ and a temperature of 15°C. Individuals were transferred from their holding tanks to Plexiglas respirometry chambers submerged in a seawater table and recording began immediately. Handling of crustaceans can result in an increase in oxygen consumption (Jouve-Duhamel and Truchot, 1985), and in previous work, we found oxygen consumption was elevated for several hours after transfer (Nancollas, 2020; Wilson et al., 2021). Because of timing constraints on oxygen consumption measurements in the simulated tidal cycle experiment (see below), experiments had to be started without a settling period. Therefore, preliminary experiments were carried out to devise the least stressful method for transferring crabs from the acclimation tanks to the respiratory chambers. Briefly, this involved transferring crabs to the respiratory chambers and immediately recording oxygen consumption for a total period of 24 h. The resting rate was determined when no statistically significant difference between each hourly oxygen consumption rate was measured. A ‘stress index’ was then calculated by dividing the oxygen consumption at each hour by the resting oxygen consumption rate. This stress index was used to adjust the oxygen consumption values after transfer; the exact methods are covered in detail in Wilson et al. (2021).

Oxygen consumption was measured using an L-DAQ intermittent flow respirometry system (Loligo Systems, Copenhagen, Denmark), with the oxygen probes calibrated to 0% and 100% saturation using a sodium sulphite solution and air-bubbled seawater, respectively. The system consisted of 4 identical cylindrical chambers (20 cm diameter×12 cm deep), which were submerged in a tank (155 cm×95 cm×50 cm deep, 1000 l) containing aerated seawater (32‰) at 15°C. Each chamber was equipped with two pumps. The first pump continually flushed seawater through the chamber between oxygen consumption measurements. This pump was automatically turned off when the chamber was closed (sealed) for oxygen measurements, and a second pump recirculated water through the chamber at a rate of 5 l min⁻¹ to ensure that there were no oxygen gradients in the chamber. Two crabs from each treatment (one crab in each of the 4 chambers), were used during each measurement period. The experiment was also run with empty chambers and background microbial respiration was found to be negligible. Experiments were carried out in constant dim red light, and black plastic sheeting surrounded the experimental tank to avoid visual disturbance to the animal. Oxygen consumption was measured every hour, during which the chamber was sealed for 30 min, and then flushed with fresh seawater for the remaining 30 min. Oxygen levels in the chamber were continuously recorded using the AutoResp 4 (v.1.7) data acquisition system (Loligo Systems), which calculated oxygen consumption rate as mg O₂ kg⁻¹ h⁻¹.

In a separate series of experiments, crabs (n=12) were subjected to a simulated tidal cycle: 6 h immersed, 6 h emersed, followed by re-immersion for a further 6 h. This was carried out in synchrony with times of anticipated immersion and emersion in the tidal holding tank. Oxygen consumption was measured at hourly intervals during the first 6 h immersion period using the aquatic respirometry system previously described. After 6 h, individual crabs were carefully transferred from the aquatic respirometry system to an aerial oxygen uptake system. Preliminary tests were carried out to assess any effects of handling stress on oxygen consumption of C. maenas when transferred between the aquatic and aerial oxygen apparatus. The results of these tests showed a slight effect of handling on oxygen consumption during the first hour of emersion/re-immersion. To account for this effect, the oxygen consumption values from the first hour after transfer (both non-tidal and tidal crabs) were divided by a stress index, which was calculated as described above to remove this handling effect (details in Nancollas, 2020; Wilson et al., 2021).

The system for measurement of aerial oxygen consumption consisted of separate blacked-out airtight respiration chambers (20 cm×27 cm×12 cm deep) housed in an incubator (MIR-254-PE, Panasonic Biomedical) maintained at 15°C and a relative humidity of 70%. During measurement of oxygen consumption, the chambers were sealed for 30 min, which allowed a measurable drop in chamber oxygen level, without exposing the crabs to hypoxia. Following these measurements, the chambers were opened for 30 min before being sealed for the next reading. Samples of air in the chamber were taken every hour (while they were sealed) during the 6 h emersion period using a 60 ml syringe by inserting the syringe's 18-gauge needle through a small hole in the lid, which was sealed with dental wax. The syringe was pumped in and out 3 times to circulate the air in the chamber before withdrawing an air sample. The sample was injected through a Drierite^® column (to remove any moisture) and into a Q-S102 O₂ analyser (Qubit Systems). The analyser was pre-calibrated to 100% air saturation (20.95% oxygen saturation) with air from the incubator, while nitrogen gas was used to achieve 0% oxygen saturation. Following the 6 h period of emersion, the crabs were carefully transferred back into the aquatic respiration chambers, and oxygen consumption was monitored at hourly intervals for a further 6 h.

Aerial oxygen consumption was calculated by taking into consideration the incubator temperature, the volume of air in the chamber minus the volume of air displaced by the crab, the length of time the chamber remained closed and the mass of the crab. After STP correction, oxygen consumption was converted from millilitres per hour to milligrams per hour by multiplying by 1.43 (32 g mol⁻¹ divided by 22.4 l mol⁻¹ at STP). The experiment was also run with empty chambers to check for microbial respiration and this was found to be negligible.

In a third series of experiments, separate animals were subjected to the same 18 h simulated tidal cycle as used above, and haemolymph samples were collected at regular intervals and analysed for Pa_O2, venous pH (pH_v), and haemocyanin and lactate concentrations. To avoid adversely affecting animals by the repeated collection of large samples from the same crab, separate animals (n=7 per group) were used at each time point.

For Pa_O2 measurements, at least 3 days before sampling, a small hole was drilled directly over the heart, which pierced through the carapace but left the pericardial membrane intact. A section of dental dam was placed over the hole and secured with cyanoacrylate glue. During the experiment, an arterial blood sample (400 µl) was taken by inserting a 21-gauge needle attached to a 1 ml airtight Hamilton syringe through the dental dam and into the pericardial cavity. Samples were taken within 30 s of removing the animal from its chamber, and approximately 200 µl of arterial haemolymph was injected below a layer of mineral oil in an Eppendorf^® tube and immediately transferred to a water bath at 15°C. Pa_O2 was measured using a Fibox-3 O₂ analyser (PreSens, Regensburg, Germany). This meter was calibrated using fully aerated seawater as 100% air saturation and seawater with sodium sulphite solution (0.06 g ml⁻¹) as 0%. The dipping probe was inserted into the sample, and readings were taken once Pa_O2 had stabilized (approximately 3 min) using OxyView software (PreSens).

Once the arterial haemolymph sample had been collected (for Pa_O2 analysis), approximately 400 µl of venous haemolymph was withdrawn from the same individual from the arthrodial membrane at the base of a walking leg. A 200 µl sample was injected below a layer of mineral oil in an Eppendorf^® tube and transferred immediately to the water bath at 15°C. pH_v was measured using a pH mini-V2 analyser (PreSens). This was calibrated using colourless pH reference buffers (Ricca Chemical Company, Arlington, TX, USA). The dipping probe was inserted into the sample and readings were taken once levels had stabilized (approximately 3 min) using pH 1-view software (PreSens). The remaining 200 µl aliquots of arterial and venous haemolymph were immediately transferred to Eppendorf^® tubes, and placed on ice before being transferred to a −80°C freezer for later haemocyanin and lactate analysis.

Haemocyanin concentration was determined spectrophotometrically using a Spectramax M5 multimode microplate reader (Molecular Devices) and an assay adapted from Paschke et al. (2010). Arterial haemolymph was thawed at on ice, then vortexed for 5 s to evenly distribute the protein. A 1:20 dilution was made with deionized water and vortexed for a further 5 s (n=7 per time point). Haemocyanin concentrations were estimated using the Beer–Lambert law from the peak absorbance measured at 335 nm, and using an extinction coefficient of 17.5 mmol l⁻¹ cm⁻¹ based on the specific absorbance (A_1%,1cm) value of 2.33 reported for C. maenas (Nickerson and Van Holde, 1971), and a molecular mass of 75 kDa.

Lactate concentration was determined using an assay that was adapted from Clow et al. (2016) using thawed venous haemolymph samples. Samples were deproteinized using 6% perchloric acid with a dilution ratio of 1:10 (n=7 per time period). The samples were vortexed and then centrifuged at 10,000 g for 10 min. The subsequent supernatant was then extracted, and 25 μl of this extract was added to 200 μl of assay medium containing glycine buffer (Sigma, G5418) and 2.5 mmol l⁻¹ NAD⁺, pH 9.0. Absorbance was determined at 340 nm using a DTX 880 microplate reader (Beckman Coulter) before the addition of 10 IU ml⁻¹ of lactate dehydrogenase (Sigma, L2500). Absorbance was read when stable, after approximately 30 min. Lactate concentration (mmol l⁻¹) was then calculated from a standard curve.

All data passed the assumptions of normality, independence and homogeneity, except for the lactate data, which were square root transformed before statistical analysis. Periodicity in oxygen consumption in the submersion experiment was determined by a Lomb–Scargle periodogram using the software PAST (Hammer et al., 2001). Periodicity was determined using the equation periodicity=1/x, where 1 is total frequency and x is the frequency of peak power. Differences in oxygen consumption rate between groups were identified using a two-way repeated measures ANOVA, where acclimation (non-tidal or tidal) and time (hour) were used as factors. Pa_O2, pH_v and haemolymph lactate and haemocyanin concentrations were analysed using two-way ANOVA, where acclimation (non-tidal or tidal) and time (hour) were used as the two factors. One-way ANOVA, or Tukey's HSD post hoc tests were used when applicable to detect significant differences between levels of individual factors. One-way ANOVA was performed to compare immersion with re-immersion oxygen consumption values. A two-sample t-test was used to compare the drop in oxygen consumption from immersion to emersion in both acclimation groups. The statistical analyses were carried out using GraphPad Prism (version 5.3 for Windows; www.graphpad.com). In all cases, a P-value of <0.05 was utilized as the criteria for statistical significance.

There was considerable variation in oxygen consumption rates during the 72 h period of submersion, both within individual animals and between the non-tidal and tidal crabs (Fig. 1). However, this variation was more pronounced in non-tidal crabs. The mean (±s.e.m.) oxygen consumption rate of non-tidal crabs was 53.4±7.8 mg O₂ kg⁻¹ h⁻¹ during the 72 h experiment, whereas that of individuals from the tidal treatment was 45.8±4.2 mg O₂ kg⁻¹ h⁻¹. There was a significant interaction between acclimation and time (two-way RM ANOVA, d.f.=71, F=1.433, P=0.0118; Fig. 1). When tested separately for main effects, this was due to significant variation in the oxygen consumption of tidal crabs (one-way RM, ANOVA, d.f.=71, F=3.129, P=0.0108), but not in non-tidal crabs (one-way RM ANOVA, d.f.=71, F=0.9391, P=0.4762) over time. Tidal crabs appeared to have an endogenous rhythm, and exhibited a small but significant decrease in oxygen consumption during times of anticipated air exposure in comparison to anticipated immersion (two-way RM ANOVA, d.f.=1, F=6.329, P=0.0306). This periodicity in tidal crabs was confirmed using a Lomb–Scargle periodogram (Fig. 2), which revealed a significant peak (P<0.01) in the spectrum at 0.0809 cycles per hour (i.e. 12.36 h per cycle).

Despite the tidal crabs exhibiting a trend towards a lower oxygen consumption rate during the initial immersion phase, there was no statistically significant difference between the two groups, with levels varying between 40.02±4.62 and 51.41±5.79 mg O₂ kg⁻¹ h⁻¹ (Fig. 3). The oxygen consumption rate of both non-tidal and tidal crabs decreased considerably when emersed, with non-tidal crabs experiencing a larger drop (54%) than tidal crabs (45%) (two-sample t-test, d.f.=22, T=2.171, P=0.0410), and the two groups maintained comparable oxygen consumption rates throughout the duration of emersion (two-way RM ANOVA, d.f.=1, F=0.2655, P=0.6115). When the crabs were re-immersed, there was a significant increase in oxygen consumption rate in both the non-tidal and tidal crabs (two-way ANOVA, d.f.=1, F=86.9, P≤0.0001) with rates returning to initial immersion values in both groups (one-way ANOVA, non-tidal: d.f.=11, F=0.3393, P=0.9753; tidal: d.f.=11, F=0.805, P=0.6325). However, the oxygen consumption rate of non-tidal individuals was significantly elevated over that of the tidal group for the majority of the re-immersion period (two-way RM ANOVA, d.f.=1, F=7.808, P=0.0106).

The haemolymph Pa_O2 of the crabs (Fig. 4) did not change significantly during the simulated tidal cycle in either group (two-way ANOVA, d.f.=8, F=1.868, P=0.0725), and values were not significantly different between the two groups (two-way ANOVA, d.f.=1, F=1.294, P=0.2578). Haemolymph Pa_O2 varied between mean values of 12.37±0.71 and 10.07±0.79 kPa for non-tidal crabs and 13.03±0.76 and 10.21±1.8 kPa for tidal crabs, during the 18 h experimental period.

The pH_v of tidal and non-tidal crabs was stable during the initial period of immersion, with mean values of 7.78±0.04 and 7.80±0.04, respectively (Fig. 5). The pH_v of tidal crabs did not change during the experiment. However, the pH_v of non-tidal crabs decreased significantly (by 0.17 pH units) during emersion, and then returned to values similar to those recorded during the initial immersion period when they were returned to water. This difference in the pattern of changes in pH_v resulted in the pH_v of non-tidal crabs being significantly lower than that measured for tidal crabs (two-way ANOVA, d.f.=1, F=4.666, P=0.0330). This was due to a significant decline in pH during the third (Tukey HSD, P=0.0151) and sixth (Tukey HSD, P=0.0111) hour of emersion.

The haemocyanin concentration of both tidal and non-tidal crabs remained relatively stable over the course of the simulated tidal cycle (Fig. 6; two-way ANOVA, d.f.=8, F=0.7726, P=0.6277). However, the haemocyanin concentration of tidal individuals was significantly higher (0.7–0.85 mmol l⁻¹) compared with that of non-tidal individuals, which varied between 0.54 and 0.75 mmol l⁻¹ (two-way ANOVA, d.f.=1, F=19.32, P≤0.0001). This was largely driven by significant differences between non-tidal and tidal crabs at hours 1, 7, 9 and 15 (Tukey's HSD test, P=0.0466, P=0.0105, P=0.0106 and P=0.0g1442, respectively).

There was considerable inter-individual variation in haemolymph lactate concentration in both non-tidal and tidal crabs (Fig. 7), and a significant interaction effect between acclimation group and time (two-way ANOVA, d.f.=7, F=2.619, P=0.0164). When the main effect of time was analysed separately, it showed that the haemolymph lactate concentration of non-tidal crabs changed significantly over the course of the experiment (one-way ANOVA, d.f.=7, F=2.226, P=0.0495), whereas lactate concentration remained stable in tidal crabs (one-way ANOVA, d.f.=7, F=1.274, P=0.2835). Haemolymph lactate concentration increased from 1.34 mmol l⁻¹ at the beginning of emersion (hour 7) to 2.22 mmol l⁻¹ at the end of this period (hour 12) in non-tidal crabs (Tukey HSD, P=0.0461), which was also significantly higher than that measured in tidal individuals at this time point (Tukey, HSD, P=0.0018). During re-immersion, the haemolymph lactate concentration of non-tidal crabs returned to pre-treatment levels (Tukey HSD, P=0.9993), and was significantly lower than that measured in tidal crabs after 3 h of re-immersion (i.e. at hour 15) (two-way ANOVA, d.f.=1, F=5.255, P=0.0282).

The tidal crabs exhibited a cyclic pattern of oxygen consumption with a periodicity of 12.36 h when monitored for 72 h in seawater. In contrast, there was no periodicity in oxygen consumption in non-tidal crabs, suggesting that acclimation to permanent submersion eliminates important stimuli that initiate circatidal rhythms. Endogenous circatidal rhythms allow intertidal organisms to anticipate and respond to predictable daily changes in tidal height (Tessmar-Raible et al., 2011; Wilcockson and Zhang, 2008), and play a key role in preparing physiological mechanisms for anticipated periods of stress (Schnytzer et al., 2018). Circatidal rhythms can be entrained by a number of environmental variables such as periodic inundation, changes in salinity, hydrostatic pressure, water turbulence, temperature and food availability (De la Iglesia and Hsu, 2010; Palmer, 2000; Reid and Naylor, 1990). As many of these entraining variables (e.g. salinity, food availability and air–water temperatures) were maintained at constant levels, the pressure increases associated with re-immersion were the most likely zeitgebers entraining this rhythm in the present study (Harris and Morgan, 1984; Williams and Naylor, 1969). The cyclic fluctuations in oxygen consumption of tidal crabs, with depressed rates of oxygen consumption during times of expected emersion, were most likely due to changes in locomotor activity (Naylor, 1958). Tidal and non-tidal crabs showed similar levels of activity during times when tidal crabs were expected to be immersed, but tidal crabs became noticeably quiescent during periods of expected emersion, even though they were continuously immersed. Locomotor activity is considered a primary mechanism underlying changes in cardiorespiratory responses (McGaw and McMahon, 1998), and this circatidal pattern of locomotor activity, and concomitant changes in oxygen consumption, are well established in C. maenas (Arudpragasam and Naylor, 1964; Naylor, 1958, 1996; Warman et al., 1993). This suggests that a reduction of locomotor activity is a key behavioural response to air exposure in this species.

When the crabs were exposed to a simulated tidal cycle (6 h immersion–6 h emersion–6 h immersion), the tidal crabs generally had lower aquatic (immersed) metabolic rates compared with the non-tidal crabs. This depressed metabolic rate may be due to acclimation to cyclic conditions, which has been previously shown to lower the resting metabolic rate of a number of organisms. For example, fluctuating temperature regimens depress the oxygen consumption of mud crab Panopeus herbstii and fiddler crab Uca pugilator (Dame and Vernberg, 1978). A similar pattern occurs in juvenile sea cucumber, Apostichopus japonicus, when acclimated to fluctuating temperatures (Dong et al., 2006). In rainbow trout, Oncorhynchus mykiss, exposure to cycles of hypoxia also results in a lower oxygen consumption rate (Williams et al., 2019). The underlying mechanisms were not elucidated in these studies, but Dong et al. (2006) proposed it as a mechanism to make more energy available for growth.

When exposed to air, this difference in oxygen consumption rate between the two groups was abolished, and both non-tidal and tidal crabs maintained oxygen consumption at approximately 50% of the rate measured during the initial period of immersion. This is comparable to previous work on C. maenas under simulated field conditions (50%; Simonik and Henry, 2014), but lower than other studies with C. maenas (75–120% of immersion values; Newell et al., 1972; Taylor and Butler, 1978). It is likely that the decline in oxygen consumption rate was partially due to a decline in locomotor activity, which is a typical response in C. maenas in response to emersion (Depledge, 1984; Simonik and Henry, 2014), and was anecdotally observed with both non-tidal and tidal crabs during this study.

Although the two groups of crabs exhibited similar rates of aerial oxygen consumption, it is worth noting that because the oxygen consumption rate of the non-tidal crabs was higher during periods of immersion they actually experienced a greater relative drop in oxygen consumption rate when transferred to air. The non-tidal crabs also experienced a concomitant decline in pH_v and an increase in lactate production, which was not observed in the tidal crabs. This suggests that the non-tidal crabs were unable to fully support metabolic costs with aerobic mechanisms. The Pa_O2, however, was not different between the two groups, which suggests that the inability of non-tidal crabs to support their metabolic costs using aerobic metabolism may have been partially related to oxygen delivery to the tissues, rather than the amount they were able to uptake. In support of this assumption, haemocyanin concentrations were higher in the tidal crabs. Carcinus maenas sampled in situ also have a higher haemocyanin concentration than those acclimated to permanently submerged conditions in the laboratory (Massabuau and Forgue, 1996), which may reflect the need for increased oxygen delivery during environmental fluctuations. Further, exposure to air for 3 h or more has been shown to result in a higher percentage of oxygen delivered bound to haemocyanin as compared with that dissolved in the haemolymph (Hsia et al., 2013; Lorenzon et al., 2007, 2008; Mangum and Weiland, 1975; Morris et al., 1996; Taylor and Whiteley, 1989). For example, oxygen delivered by haemocyanin in the intertidal crab Hemigrapsus nudus increases from <50% during submersion to over 85% when exposed to air (Mangum and Weiland, 1975; Morris et al., 1996). Even in subtidal crustaceans, such as Homarus gammarus, the participation of haemocyanin in oxygen delivery increases during exposure to air, with 94% of oxygen delivered under this condition (Taylor and Whiteley, 1989). Because haemocyanin levels were lower in the non-tidal crabs, this could have resulted in a lower level of O₂ delivery overall to the tissues to meet metabolic demands. Additionally, bradycardia commonly occurs during emersion in decapod crustaceans (DeFur, 1988), and has been previously reported for C. maenas (Newell et al., 1972; Styrishave et al., 2003; Wallace, 1972; but not always: see Simonik and Henry, 2014). Therefore, while not investigated in this study, it is possible that a more severe bradycardia in non-tidal crabs (i.e. a greater reduction in circulatory capacity) in conjunction with lower haemocyanin levels could have imposed a limitation on aerobic metabolism during emersion in non-tidal crabs.

During emersion, tidal crabs were able to maintain pH_v at pre-emersion levels, whereas non-tidal crabs exhibited a progressive decline, so either tidal crabs do not experience an acidosis during emersion or they can efficiently buffer the pH changes. The decline in the pH_v displayed by non-tidal crabs could, in part, be caused by the progressive accumulation of lactate in the haemolymph. However, significant levels of lactate did not accumulate in non-tidal crabs until the end of the emersion period (6 h), and the decline in pH exhibited prior to this could be due to other factors such as an increase in P_CO2. Although not directly measured in this study, it is important to consider the influence of CO₂ on acid–base balance. Decapod crustaceans experience an increase in P_CO2 during emersion as a result of the difficulty of excreting CO₂ across the gills in air (reviewed in DeFur, 1988; Truchot, 1990). Low intertidal/subtidal species are more limited in their ability to compensate for increased P_CO2 during emersion compared with species that experience regular exposure to air. For example, the velvet crab, Necora puber, a low intertidal/subtidal species, has a limited capacity to mobilize HCO₃⁻ (Rastrick et al., 2014), and C. maenas acclimated to submerged conditions take 100 h to compensate for P_CO2 increases during emersion (Truchot, 1975). Conversely, intertidal species such as the purple rock crab, Leptograpsus variegates, and the granulated crab, Neohelice granulata, can better regulate P_CO2 levels and restore pH within 1–2 h of air exposure (Butler and Morris, 1996; Luquet et al., 1998). This is likely facilitated by elevated carbonic anhydrase activity in the gills, which has been shown to be up to 10-fold higher in intertidal crustacean species in comparison to fully aquatic species (Böttcher et al., 1990; Henry et al., 1994, 2003). In addition, as haemocyanin makes up the majority of the protein component of crustacean plasma (Pascual et al., 2003), it is an important buffer of pH changes (Whiteley, 2011; Rastrick et al., 2014). Therefore, it is possible that while both tidal and non-tidal crabs experienced changes in acid–base balance, the elevated levels of haemocyanin in the tidal crabs may have been more effective at buffering H⁺ ions during emersion.

Although oxygen consumption rate and lactate levels returned to pre-treatment values on re-immersion, haemolymph lactate levels were slightly (but not significantly) higher in tidal crabs and this could suggest that non-tidal and tidal crabs use different mechanisms, or have different capacities, to regulate oxygen consumption and lactate production during emersion. The crayfish Austropotamobius pallipes exhibits elevated lactate levels during re-immersion after a 24 h emersion period (Jackson et al., 2001; Taylor and Wheatly, 1981). This is due to the crayfish sequestering lactate and protons in the calcified skeleton (Jackson et al., 2001), which results in a reduction of lactate in the haemolymph during emersion. Upon re-immersion, there is a lactate wash-out into the haemolymph, which results in elevated haemolymph lactate levels. A similar phenomenon may have occurred in the present study, with lactate produced by tidal crabs being sequestered during emersion but released into the haemolymph during re-immersion. Sequestration by the exoskeleton could also aide in acid–base buffering capacity by the formation of calcium and bicarbonate from calcium carbonate and hydrogen (Jackson et al., 2001).

Although lactate is typically regarded as an end-product of anaerobic metabolism, it may also be an important metabolic fuel (Gladden, 2004; Jayasundara and Somero, 2013). Crustaceans living in more dynamic environments display higher constitutive levels of lactate (Jost et al., 2012; Maciel et al., 2008), which may reflect the need to have fuel readily available to tolerate unpredictable conditions. Furthermore, in many crustaceans, elevated levels of lactate can increase the oxygen affinity of haemocyanin, ensuring optimal oxygen delivery during periods of stress (Truchot, 1980). In the crab Neohelice granulata, elevated haemolymph lactate levels during normoxia are common and utilized as fuel either via oxidation through the tricarboxylic acid cycle (TCA) or by conversion to glucose via gluconeogenesis (Maciel et al., 2008). Therefore, the elevated lactate observed in tidal C. maenas during submerged conditions could serve as a metabolic fuel during changes in respiratory medium, or to enhance the oxygen affinity of haemocyanin (Truchot, 1980).

The results of this study on C. maenas support those of previous studies showing that caution must be taken when acclimating intertidal organisms to laboratory conditions. Although acclimation to constant conditions allows us to associate physiological responses with a particular stressor, it also risks abolishing important physiological responses and adaptations that may play a critical role in physiological performance of organisms in situ. If there is a drawback of our study, it is that we only had a single tank for each treatment (non-tidal and tidal). The conditions in the tidal and non-tidal tanks were replicated as closely as possible (same size, and situated under the same light source, same water supply fed into each tank via a splitter hose and water temperature maintained within precise levels with digital heaters) and therefore it can be assumed that any differences between the two groups of crabs were due to the change in water level only. For logistical reasons (available tanks and time), we were unable to replicate the actual treatment tanks. In future studies, replication of treatment tanks would avoid the problem of pseudoreplication and allow one to perform more robust statistical analysis.

With respect to studies on intertidal organisms, the cyclic exposure to air is a predictable occurrence, and it is clearly an important modulator of physiology and behaviour. The underlying mechanisms and plasticity of traits associated with emersion tolerance warrants further exploration, which could be achieved through reciprocal studies in the field and in the laboratory with subtidal and intertidal crabs. Moving forward, it is paramount that organisms are exposed to realistic ecological scenarios that incorporate a multifaceted design within the laboratory environment if we are to gain an accurate understanding of how these species respond in situ.

We wish to thank Gemma Rayner and Roy Murphy for help with crab collection and Connie Short for assistance with lactate assays. Comments by Drs Bill Driedzic, Kurt Gamperl and John Spicer, and two anonymous reviewers helped improve earlier versions of the manuscript.