Flipper stroke rate and venous oxygen levels in free-ranging California sea lions

Optimal management of oxygen stores underlies the breath-hold capacity of air-breathing divers. The utilization of oxygen stores during dives is dependent on the regulation of heart rate, the magnitude and/or distribution of peripheral blood flow, tissue oxygen uptake and muscle workload. The severe bradycardia and peripheral vasoconstriction observed in the classic dive response conserves oxygen in blood by directing it towards hypoxia-sensitive tissues, such as the brain and heart (Irving et al., 1941; Scholander et al., 1942). However, blood oxygen supplementation to working muscle can occur during moderate bradycardias that are common during routine dives and breath-holds of diving animals (Guyton et al., 1995; Jobsis et al., 2001; Ponganis et al., 2008; Williams et al., 2011). In addition, the positive relationship between heart rate and flipper stroke rate in some diving mammals has led to the hypothesis that exercise and muscle workload modulate the dive response and increase heart rate during short duration dives (Davis and Williams, 2012). In the presence of muscle blood flow, blood oxygen extraction by working muscles during a dive should result in a decline in venous hemoglobin saturation (S_O2). Such declines in venous oxygen content are proportional to muscle workload in terrestrial mammals, despite increases in muscle blood flow and oxygen delivery during exercise (Taylor et al., 1987).

The California sea lion [Zalophus californianus (Lesson 1828)] is an ideal model with which to investigate the effect of muscle workload (flipper stroke rate) on the depletion of blood oxygen stores during dives. These animals are known for being active swimmers and can dive as deep as 540 m and as long as 10 min, during which they demonstrate a wide range of heart rate responses and blood oxygen depletion patterns (Kuhn and Costa, 2014; McDonald and Ponganis, 2012, 2013, 2014). For this study, we equipped free-ranging adult female California sea lions with bio-logging devices that simultaneously recorded depth, tri-axial acceleration and venous partial pressure of oxygen (P_O2). From these data, we documented stroke rate profiles, evaluated the effect of muscle workload on changes in S_O2 and determined whether blood oxygen extraction by muscle contributed significantly to S_O2 depletion in dives of different depths and durations.

This study was conducted on San Nicolas Island, California, in August 2013. Four animals were captured with hoop nets, transferred approximately 1000 m from the capture location site and anesthetized using a portable vaporizer-breathing circuit set up with an initial mixture of 5% isoflurane and 100% oxygen. Once anesthetized, a custom-built P_O2 datalogger (UUB-2PT; UFI, Morro Bay, CA, USA) with custom housing (Meer Instruments, Palomar Mountain, CA, USA), time-depth recorder with three-axis accelerometer (TDR10±2 g; Wildlife Computers, Redmond, WA, USA) and VHF radio transmitter (mm160B; Advanced Telemetry Systems, Isanti, MN, USA) were affixed to the dorsal, midline pelage of the animal using epoxy (Loctite, Henkel Corp., Westlake, OH, USA). Using an ultrasound machine (SonoSite Inc., Bothell, WA, USA), the caudal gluteal vein was identified and percutaneously catheterized with a peel-away catheter (5 Fr, Cook Medical, Bloomington, IN, USA). A P_O2 electrode (Licox C1.1, Integra Life Sciences, Plainsboro, NJ, USA) was then inserted in the catheter and threaded up into the posterior vena cava. To prevent water damage to the P_O2 electrode while out at sea, the P_O2 electrode was connected to a waterproof cable and connector (Impulse Enterprise, San Diego, CA, USA). To protect and secure the electrode, the insertion site was covered with a small patch of neoprene that was affixed to the animal's fur with fast-setting glue (Loctite 401, Henkel). These procedures and calibrations have been explained in more detail in previous publications (McDonald and Ponganis, 2013; Meir et al., 2009; Ponganis et al., 1991, Ponganis et al., 1997, 2007; Stockard et al., 2007). Depth and P_O2 were recorded at 1 Hz, and three axes of acceleration were recorded at 16 Hz.

After instrumentation, the animals were placed in a large canine kennel to allow them to safely recover from anesthesia (∼30–60 min) and weighed (±0.2 kg, MSI-7200 Dyna-link; Measurement Systems International, Seattle, WA, USA). Once the animals were fully alert, they were released back onto the same beach from which they were originally captured. We recaptured the sea lions after one to three trips to sea, removed the instruments under manual restraint, and released them (∼5–10 min procedure). All procedures were approved by the University of California, San Diego, Animal Subjects Committee (no. S11303) and National Marine Fisheries Services (no. 14676).

Dive data were analyzed using a custom-written MATLAB program (MathWorks, IKNOS; Y. Tremblay). Briefly, this dive analysis program calculates a zero-offset correction at the surface and identifies dives using a specified minimum depth (10 m) and duration (20 s). This depth threshold was selected as dives under 10 m were too short to evaluate changes in physiological parameters. The following dive phases were identified in the data: descent (surface to 80% maximum depth), bottom (depths within 80% maximum depth) and ascent (80% maximum depth to surface). Most California sea lion dives are shallower than 100 m, yet many adult female California sea lions from San Nicolas Island are known to be deep divers (Kuhn and Costa, 2014; McHuron et al., 2016). Therefore, to compare shallow and deep dives, dives to maximum depths >100 m were classified as deep, and dives to maximum depths ≤100 m were classified as shallow.

Foreflipper stroke rate was calculated using a custom-written algorithm in MATLAB. The low-frequency static acceleration data were filtered out using a 0.2 Hz high-pass Butterworth filter. The resulting dynamic acceleration was then analyzed using power spectral density analysis to identify the dominant frequency of a stroke for each individual animal (approximately 0.8–1.2 strokes s⁻¹). A peak detection algorithm, similar to those in other studies (Jeanniard-du-Dot et al., 2016; Sato et al., 2011), was used to identify flipper strokes. A single flipper stroke was identified when there was a prominent acceleration peak (≥0.4–0.5 m s⁻²) in the x-axis (‘forward surge’) or z-axis (‘heave surge’) (Fig. 1). Stroke rate was calculated as an average stroke rate along a moving window of 10 s throughout the diving record.

The P_O2 electrodes were calibrated in the laboratory before deployment (Ponganis et al., 2007; Stockard et al., 2007). Values of S_O2 were calculated using the hemoglobin–dissociation curve from McDonald and Ponganis (2013). Because the minimum calculated aerobic dive limit for this species is 3.4 min (Weise and Costa, 2007) and increased P_CO2 contributes to a decrease in pH as dive duration increases (Kooyman et al., 1980), P_O2 values collected prior to 3 min into the dive were converted to S_O2 using the equation for a pH of 7.4 (log[S_O2/(100–S_O2)]=2.473×log(P_O2)–3.632) and all P_O2 values collected beyond 3 min into the dive were converted to S_O2 using the equation for a pH of 7.3 (log[S_O2/(100–S_O2)]=2.363×log(P_O2)–3.576). Previous data suggest that California sea lions, and other deep-diving pinnipeds, typically keep core body temperature relatively constant at 37°C while diving (McDonald and Ponganis, 2013; Meir and Ponganis, 2010); therefore, P_O2 electrodes were only calibrated at this temperature and temperature effects on the S_O2 values were not considered. The time on the TDR/accelerometer instrument and P_O2 logger were synchronized to the same internet-synced computer clock. Mean change in S_O2 (ΔS_O2, Δ% s⁻¹) was calculated for the same 10-s moving window as flipper stroke rates to facilitate comparison. The mean ΔS_O2 was also calculated for the three dive phases (descent, bottom and ascent) and the total dive duration.

To investigate whether flipper stroke rate activity affects ΔS_O2 for dives of different durations, we used linear mixed-effect models with flipper stroke rate, dive duration and an interaction term (flipper stroke rate×dive duration) as fixed effects and individual as a random effect. This model was evaluated independently for shallow and deep dives.

Gliding (i.e. no flipper strokes recorded) was consistently noticed during the descent and ascent phases of dives deeper than 100 m. While dives shallower than 100 m did periodically include periods of gliding, the glides were often too short in duration to examine any impacts on diving physiology. The starting and ending depths of the glide and glide durations were independently compared against mass, maximum depth and ΔS_O2 for those phases of the dives using a linear mixed-effect model, with individual as the random effect. All linear mixed-effect models were evaluated in JMP (V.12, SAS Institute, Cary, NC, USA) using a built-in LME function and random slope or random intercept model. Corrected Akaike's information criterion (AIC_c) values between these two models were then evaluated; however, the random intercept model always gave the lowest AIC_c value and therefore this model was consistently used. Residuals were visually assessed for normal distribution to determine the quality of the model predictions.

Venous P_O2, depth and acceleration data were simultaneously collected from four adult female California sea lions. This resulted in data collection from a total of 3287 dives. Flipper stroke rate analysis revealed a stroke–glide pattern with periods of prolonged gliding during the descent of deep dives (Fig. 1A). Gliding occasionally occurred in dives shallower than 100 m; however, the glides were brief and inconsistent between these shallower dives. Similar to previous reports on California sea lions from San Nicolas Island (McDonald and Ponganis, 2013, 2014; McHuron et al., 2016), all animals reached depths over 200 m, while the majority (52%) of dives were shallower than 50 m. The distribution of maximum dive depths was bimodal, with very few dives (11%) occurring at depths of 100–150 m. This bimodal pattern influenced how we defined shallow (≤100 m) and deep (>100 m) dives. Dive statistics are reported in Table 1.

Posterior vena caval S_O2 profiles were variable during shallow dives, with a wide range of values throughout dives and inconsistent patterns in ΔS_O2 (Fig. 2A). On deeper dives, there was a steady drop in S_O2 throughout the dive until later in ascent, when values often increased prior to surfacing (Figs 2B, 3C, 4C). There was no significant relationship between flipper stroke rate and ΔS_O2 during deep dives (Table 2). However, there was a significant relationship between the interaction term of dive duration and flipper stroke rate with ΔS_O2 during shallow dives (Table 2).

On deeper dives, animals would regularly glide down to depth (Figs 1A, 2B, 3C). This resulted in low stroke rates during the descent of deep dives (Fig. 5). The depth at which gliding started (mean=60.5±1.6 m; Table 3) on the descent phase of deeper dives did not vary with maximum depth of the dives (P=0.8; Table 3). During deep dives, sea lions glided an average of 55% of both the depth and duration of the descent phase, with a maximum of 88% of the descent depth and 92% of the descent duration. As expected, the duration of the descent glide increased significantly with maximum depth of the dive (F_1,159=59.3, P<0.001). The depth at which the descent glide started did not significantly predict patterns of ΔS_O2 during the descent period (P=0.8).

The average flipper stroke rate was 0.42±0.002 Hz during the ascent phase of deeper dives (Fig. 5). However, ascent flipper stroke rates usually were higher at the beginning of the ascent (∼1 Hz) and tapered off to a glide by the time the animals reached 15–25 m (Figs 1A, 3C). The depth at which the glide started varied between individuals, and this was not influenced by sea lion mass or dive depth or duration. At the end of the ascent from deeper dives, animals would also utilize a gliding strategy to reach the surface (Figs 1A, 3C). The depth at which gliding started on the ascent phase did not vary with maximum depth of the dive (P=0.6). The duration of the ascent glide did not vary significantly with maximum depth (P=0.5). The depth at which gliding started on the ascent phase and the duration of the ascent glide phase were not related to ΔS_O2 during the ascent phase (P=0.3, P=0.2, respectively) (Table 3).

In contrast to the consistent stroke rate patterns seen in deep dives (Figs 2B, 5), our data showed that there was a large variation in flipper stroke rates on dives with maximum depths shallower than 100 m (Figs 2A, 5). Similar to deep dives, the highest stroke rates during shallow dives occurred at the beginning of the dive, while the lowest stroke rates were often at the end of the dive (Figs 3A, 4B). This pattern is consistent with expected buoyancy changes with depth. Overall, the mean flipper stroke rate for the entire dive duration was similar between both shallow dives and deep dives (Fig. 5). However, the range and variability of flipper stroke rates for the entire dive duration was larger during shallow dives (Fig. 5).

We documented flipper stroke rate patterns in free-ranging adult female California sea lions for the first time. Similar to techniques using one axis of acceleration to identify flipper strokes in free-diving animals (Maresh et al., 2014; Sato et al., 2003), we used both the surge (x) and heave (z) acceleration axes to reliably identify flipper strokes in all dive phases (Fig. 1). Our results show that the average flipper stroke rates during different dive phases from this study were consistent with data from other free-ranging otariids (∼0.4–0.6 Hz; Fig. 5) (Insley et al., 2008; Jeanniard-du-Dot et al., 2016). Similar to some free-ranging phocids and cetaceans (Davis et al., 2001; Williams et al., 2000), California sea lions exhibited periods of gliding on the descent and ascent phases of deep dives, with maximum glides reaching over 160 s and covering 340 m vertical distance on the descent portions (Figs 1A, 2B, 3C). California sea lions are streamlined, with high fineness ratios resulting in low drag, making them efficient at gliding through the water (Feldkamp, 1987). Gliding during descent will conserve blood and muscle oxygen stores for subsequent activities such as prey capture and stroke effort during ascent.

Prolonged gliding on the descent phase was consistently associated with dives that reached the range of the estimated depth of lung collapse in this species (∼200 m; Fig. 5) (McDonald and Ponganis, 2012). The average depth at which gliding started on the descent was relatively shallow at around 60 m, yet this is not surprising as buoyancy will decrease as lung volume decreases (approximately 50% of surface lung volume at 10 m depth, and 14% at 60 m according to Boyle's law). Interestingly, elephant seals also began prolonged glides during deep dives at approximately 60 m depth (Davis et al., 2001). The gliding pattern seen on the ascent phase of sea lions has also been documented in other marine mammals and diving birds, and was suggested to be influenced by respiratory air volume (Sato et al., 2011; Watanuki et al., 2006; Williams et al., 2000). Ascent glides in sea lions are likely secondary to increased buoyancy near the surface owing to lung re-expansion during ascent.

Profiles of S_O2 were highly variable during shallow dives, with S_O2 even increasing throughout the duration of the dive (Figs 2A, 4D). In addition, some shallow dives began with low S_O2; the sea lions did not completely re-saturate venous blood prior to starting these dives (Fig. 2A). In some cases, ΔS_O2 appeared to be affected by stroke rate patterns as would be expected during exercise and consistent with recent suggestions that heart rate (and presumably some muscle blood flow) is modulated by exercise (Davis and Williams, 2012; Williams et al., 2015). During shallow dives, California sea lions exhibit higher heart rates than those seen during deep dives (McDonald and Ponganis, 2014). These higher heart rates are likely associated with increased blood flow to muscle, and could account for the relationship between flipper stroke rates and venous blood oxygen depletion during shallow dives. However, the weak correlation between the two parameters, and in some instances, increases in S_O2 throughout the dives, despite high stroke rates, argues that this does not always occur.

Full interpretation of S_O2 during shallow dives is also limited by the lack of arterial hemoglobin saturation data during shallow dives of California sea lions. The relationship between S_O2 profiles and the interaction of dive duration and stroke rate of shallow dives may also be partially secondary to increased heart rates, muscle perfusion and greater arterial hemoglobin desaturation during longer shallow dives. In other words, a decrease in S_O2 may reflect a decrease in arterial oxygen content delivery rather than an increase in blood oxygen extraction by tissue. Ideally, simultaneous measurements of arterial/venous blood oxygen content, heart rate, flipper stroke patterns and/or muscle myoglobin saturation would resolve this question.

From our current data, it appears that for shallow, short duration dives, regulation of posterior vena caval S_O2 is not critical or highly controlled, and that the peripheral vascular response is quite variable. Such plasticity in blood flow distribution has also been observed in diving emperor penguins (Williams et al., 2011).

Assessment of locomotory effort on S_O2 profiles during deep dives is aided by our prior study that demonstrated that arterial hemoglobin saturation is maintained during dives over 300 m deep and 6 min in duration (McDonald and Ponganis, 2012). Consequently, changes in S_O2 during deep dives are more likely because of changes in perfusion and tissue oxygen extraction rather than a decrease in arterial oxygen content delivery.

Posterior vena caval S_O2 profiles in the sea lions were similar to those of a previous study from our laboratory (McDonald and Ponganis, 2013). In the present study, we demonstrate that flipper stroke rates had little impact on ΔS_O2 during a majority of deep dives of sea lions (Table 2). We also saw that S_O2 often increased during the ascent, despite the animal's active stroking to reach the surface (Figs 2B, 3C, 4C). Furthermore, the greatest decreases in S_O2 occurred during the low stroke rates and prolonged glides of the descents of deep dives (Figs 3, 5). In contrast, some of the largest increases in S_O2 occurred while the animals were actively stroking during ascent from deep dives (Figs 3C, 5). As in our prior studies, we postulate that the re-oxygenation of venous blood prior to surfacing is indicative of maintenance of pulmonary gas exchange at shallow depths, increased peripheral perfusion and possible use of arterio-venous shunts during the ascent (McDonald and Ponganis, 2012, 2013, 2014).

Based on these findings, we conclude that ΔS_O2 in the posterior vena cava is not significantly affected by flipper stroke rate during deep dives. Rather, as we have previously hypothesized (McDonald and Ponganis, 2013), it is more likely that S_O2 is determined by the degree of bradycardia, the magnitude of reduction in peripheral blood flow and the associated changes in tissue transit time during dives. Prolonged tissue transit times during severe bradycardia should result in greater blood oxygen extraction and lower hemoglobin saturations in any blood slowly draining into the posterior vena cava. In addition, given that arterial hemoglobin saturation is well maintained throughout deep dives (McDonald and Ponganis, 2012), desaturation of venous blood is unlikely secondary to a decrease in arterial blood oxygen content.

The maintenance of high arterial hemoglobin saturation (McDonald and Ponganis, 2012), despite the posterior vena caval S_O2 decreasing to extremely low values during deep dives, suggests the presence of a central venous oxygen pool that is slowly depleted during the severe bradycardias seen in deep dive (McDonald and Ponganis, 2014). Otherwise, in the presence of lung collapse and lack of gas exchange, arterial hemoglobin saturation should reflect the low venous values observed in the posterior vena cava. Such a central blood oxygen store would be analogous to the hepatic sinus oxygen store described in elephant seals (Elsner et al., 1964). The existence of such a central venous oxygen store remains to be demonstrated in otariids with either anterior vena cava or pulmonary artery hemoglobin saturation measurements.

California sea lions use a gliding strategy during the descent phase of most dives beyond 100 m. On these deeper dives, animals also glide to the surface from depths of 15–25 m, probably because of the increased buoyancy associated with lung re-expansion during ascent. The lack of a strong relationship between posterior vena caval S_O2 and flipper stroke rate during deep and some shallow dives demonstrates that posterior vena caval hemoglobin desaturation is not dependent on muscle work load. This independence of blood oxygen depletion from locomotory effort implies that (1) muscle blood flow is restricted during deep dives, and not consistently regulated during shallow dives, and (2) the posterior vena caval oxygen profile is more related to heart rate and the magnitude of tissue perfusion and/or oxygen extraction.

The logistics and field work for this project were made possible by the intense efforts from many volunteers, and we are very thankful for their help. Specifically, we would like to thank J. Ugoretz, G. Smith, G. Kooyman, R. Walsh, E. McHuron, C. Verlinden, C. Stehman, D. Costa and many other members of the Costa lab at UCSC.