Lung mechanics and pulmonary function testing in cetaceans

Breath-hold diving mammals live a life of dual constraints. Time underwater needs to be maximized to enhance foraging efficiency; however, animals must ultimately return to the surface to exchange metabolic gases. A central concept in breath-hold diving research has been that the available O₂ and its utilization rate during diving determine the length of a single dive. Studies in both cetaceans (Ridgway, 1986; Ridgway et al., 1969) and pinnipeds (Boutilier et al., 2001; Fahlman et al., 2008) suggest that the available O₂ is seldom the limiting factor for the length of a single dive and that CO₂ may be dictating the surface interval between dives. If O₂ is not the limiting factor, what other physiological limitations affect the depth and duration of a dive? A multitude of pressure-related variables may impose physiological challenges that limit diving (Leith, 1989). One important, yet poorly understood, element of diving capability is the effect of pressure on gas exchange, and thus basic respiratory physiology. The focus of this work is to illuminate the basic elements of respiratory physiology in cetaceans.

Of all the mammals on earth, marine mammals face the widest range of environmental pressures while foraging. It has generally been accepted that marine mammals possess physiological and morphological traits that prevent, or reduce, the occurrence of pressure-related pathologies, such as lung squeeze and decompression sickness (DCS). Scholander (1940) proposed that the conducting airways of marine mammals resist compression, while the alveolar space and chest are flexible, and therefore easily compressed. Theoretical modeling work indicates that the structural properties (e.g. compliance) of the various parts of the respiratory system dictate how pressure affects the distribution of gas between conducting airways and alveolar space (Bostrom et al., 2008; Fahlman et al., 2009,, 2011; Fitz-Clarke, 2007). High pulmonary compliance may be an important component of minimizing transpulmonary pressure gradients (Fahlman et al., 2014; Leith, 1976,, 1989; Ridgway et al., 1969), and has also been suggested to reduce the risk of elevated blood and tissue N₂ levels, thereby minimizing the risk for DCS (Ridgway and Howard, 1979; Scholander, 1940).

Practical and ethical constraints have historically limited experimental options for studying how pressure affects gas exchange in marine mammals (Kooyman and Sinnett, 1982; Ridgway and Howard, 1979). Theoretical studies have indicated that pulmonary gas exchange and cardiac output during diving are the most important variables affecting N₂ levels in blood and tissue, and thereby the risk for developing DCS (Fahlman et al., 2009,, 2001; Hooker et al., 2009; Kvadsheim et al., 2012). Despite the value of these theoretical studies, the results should be interpreted with caution, as the current understanding of how gas exchange is affected by compression of the respiratory system is rudimentary at best. Recent studies suggest that there is considerable structural (anatomical) and functional (compliance) variability in the upper and lower airways of different orders, families and species of marine mammals (Moore et al., 2014; Piscitelli et al., 2010,, 2013).

The respiratory physiology of marine mammals has been insufficiently studied and there is a paucity of information related to the mechanical properties of the lung and chest. More scarce yet is experimentally derived information on the mechanical properties of the respiratory system and how its physiology is affected by pressure (Bostrom et al., 2008; Fahlman et al., 2009,, 2006; Falke et al., 1985; Kooyman and Cornell, 1981; Kooyman and Sinnett, 1979,, 1982; Piscitelli et al., 2010). To investigate the mechanical and physiological properties of the cetacean respiratory system, pulmonary mechanics and gas exchange in the bottlenose dolphin were studied. The results of this study provide baseline data on lung function and respiratory mechanics in healthy dolphins. This work provides a reference to compare these traits and characteristics amongst different marine mammals and their terrestrial counterparts. It is critical knowledge to understand the physiological limits of cetaceans, and how they routinely perform extreme physiological feats as part of their daily life.

Data from a total of 163 spontaneous breaths and 45 maximal effort breaths (trained ‘chuffs’) were collected from 6 male bottlenose dolphins housed at Dolphin Quest, Oahu in April, 2013, using a custom-built pneumotachometer (Fig. 1). For the purpose of evaluating lung mechanics, a subset of these breaths was analyzed (see below).

Fig. 2A shows the respiratory flows, the expiratory O₂, and esophageal pressure during a voluntary breath and a chuff. The same data are presented in Fig. 2B on a larger scale to illustrate the differences in the respiratory flows for relaxed, spontaneous breaths versus maximal respiratory effort. For most inhalations, the flow increased rapidly to a rate that remained fairly constant until nearly the end of the inspiratory phase. Flows for exhalations were more variable, with a very rapid increase in rate, a less consistent plateau than inhalations, and a rapid decrease at the end of the expiratory phase (Fig. 2B). At flows >60 l s⁻¹, the relationship between differential pressure and flow became non-linear, which necessitated the use of different calibration factors for the inspiratory and expiratory phases.

The average respiratory frequency during trials was 3.4±1.1 breaths min⁻¹ (range: 2.7–4.5), which was not significantly different (P>0.1, paired t-test) from values observed during visual follows of animals observed in their enclosure during free time (3.7±0.6 breaths min⁻¹). Peak inspiratory flow of spontaneous breaths ranged from 3.1 to 33.4 l s⁻¹ while peak expiratory flow ranged from 3.6 l s⁻¹ to 137.6 l s⁻¹. The average maximum inspiratory flow of spontaneous breaths was 19.3±2.2 l s⁻¹, which was significantly lower than the maximum expiratory flow of 43.6±7.1 l s⁻¹ (P<0.01, paired-t-test). Both the average maximum inspiratory and expiratory flow-rates were higher during chuffs (inspiration: 24.9±4.8 l s⁻¹; expiration: 83.2±22.0 l s⁻¹) than during spontaneous breaths (inspiration: 18.3±3.6 l s⁻¹, P<0.01, AIC_null=1226, AIC_breath-type=1138, Fig. 3A; expiration: 30.4±10.6 l s⁻¹, P<0.01, AIC_null=1941, AIC_breath-type=1659, Fig. 3B). The duration of the expiratory phase was significantly shorter than that of the inspiratory phase, for both spontaneous breaths (P<0.01, AIC_null=1028, AIC_breath-type=1011, expiratory: 0.31±0.04 s; inspiratory: 0.43±0.05 s) and chuffs (P<0.01, AIC_null=1421, AIC_breath-type=1356, expiratory: 0.26±0.04 s; inspiratory: 0.66±0.11 s). The tidal volume did not differ between animals (P>0.1, Fig. 4), but was significantly higher during chuffs (11.9±3.7 l, AIC_null=1099.2, AIC_breath-type=913.3) compared with spontaneous breaths (5.7±1.7 l, P<0.01). The largest tidal volume recorded during this study was 18.4 l for Kolohe (94 ml kg⁻¹), which was greater than that animal's estimated total lung capacity (TLC_est) of 17.3 l, based on models from excised lungs (TLCest=0.135M_b0.92 see references Fahlman et al., 2011; Kooyman, 1973).

The mean (N=208) end-expiratory O₂ and CO₂ were 12.6±2.2% and 6.9±2.2%, respectively. End-expiratory O₂ was significantly lower during spontaneous breaths (12.3±2.1%) compared with chuffs (13.7±2.3%, P<0.01, AIC_null=916.7, AIC_breath-type=902.8, Fig. 5A), but end-expiratory CO₂ did not differ based on breath-type (spontaneous: 6.9±1.3%; chuff: 6.7±1.1, P>0.1, AIC_null=672.6, AIC_breath-type=675.1, Fig. 5B).

V̇_O2 and V̇_CO2 were estimated for resting durations from 0.6 to 4.5 min. Consequently, the estimated metabolic rates varied substantially within and between animals (Table 1). The mass specific V̇_O2 (sV̇_CO2) and V̇_CO2 (sV̇_CO2) ranged from 3.11 to 5.08 ml min⁻¹ kg⁻¹.

The esophageal pressure during the inspiratory pause (with the blow hole closed) ranged between 25 and 40 cmH₂O (Fig. 2A) with no significant difference between animals (P>0.1, t-test). During the inspiratory phase of spontaneous breaths, there was a drop in the esophageal pressure ranging from 30 to 50 cmH₂O (Fig. 2C, e.g. from 25 to −14 cmH₂O). During the expiratory phase of chuffs there was an initial increase in esophageal pressure, ranging from 18 to 44 cmH₂O (e.g. 25 to 60 cmH₂O), followed by a decrease in esophageal pressure of 21 to 81 cmH₂O (e.g. 60 cmH₂O to −20 cmH₂O). The drop in esophageal pressure is probably attributed to activation of the inspiratory muscles (Fig. 2A).

Flow–volume curves for two voluntary breaths and two chuffs are presented in Fig. 6. By convention, expiratory flow rates are positive. There was no clear dependence of expiratory flow rate on lung volume during chuffs, and expiratory flows are nearly constant over most of the vital capacity. The spontaneous breaths are shown to illustrate the differences in shape between voluntary breaths and chuffs.

Fig. 2C shows selected end-expiratory and end-inspiratory points of zero flow during a spontaneous breath (a and b) which were used to determine the dynamic compliance (C_L, Fig. 7). Of 194 breaths analyzed (Table 1), 74 were considered to provide reliable and repeatable data for these parameters and were included in this analysis. Chuffs were omitted from compliance calculations because the transpulmonary pressure (P_L, the difference between pressure at the airway opening and pleural pressure) was often highly negative, which suggests closed alveoli, airway closure or limitation of expiratory flow (see Discussion for more detail). The mean±s.e.m C_L was 0.37±0.04 l cmH₂O⁻¹ (Fig. 7). As C_L varies with lung size (Stahl, 1967), the specific lung compliance (sC_L, cmH₂O⁻¹, Fig. 7) was computed by dividing C_L by the minimum air volume (MAV), which was estimated to be 7% of total lung capacity based on previous experiments with excised lungs (Fahlman et al., 2011). In animals with very high chest wall compliance (i.e. a chest wall that does not resist compression), MAV can be used as an estimate of the residual volume (V_R) and also of functional residual capacity (FRC) (our unpublished observations). Thus, we assume that the elastic recoil of the chest does not limit emptying of the lung. The estimated MAV ranged from 1.0 to 1.5 l and the average sC_L was 0.31±0.04 cmH₂O⁻¹ (Fig. 7).

The current study confirms the findings of Kooyman and Cornell (1981), which indicate that bottlenose dolphins have the ability to generate extremely high maximal expiratory (>130 l s⁻¹) and inspiratory (>30 l s⁻¹, Fig. 2) flow, and can exchange as much as 95% of their estimated total lung capacity (TLC_est; Fahlman et al., 2011; Kooyman, 1973) in a single breath. In addition, our results provide preliminary data on dynamic lung compliance in bottlenose dolphins that are repeatable and consistent with previous measurements in odontocetes (Olsen et al., 1969).

The esophageal pressures recorded suggest that the inspiratory phase is active (driven by muscle activity) during both spontaneous breathing, and during maximal respiratory efforts. In contrast, the expiratory phase appears to be active only during maximal effort (chuffs). During spontaneous, relaxed exhalations, the absence of a rise in esophageal pressure preceding the expiration (Fig. 2A) indicates a passive process. The decrease in esophageal pressure (P_eso), which begins just before the end of exhalation during spontaneous breaths, probably indicates initiation of activity in the inspiratory muscles (Fig. 2A). Because experimental data suggest that the expiratory phase of spontaneous breaths is a passive process, the animals remain stationary during this testing. Morphological, anatomical and observational data suggest that the expiratory phase of normal respiration while swimming is coupled with locomotion and physical posturing (Cotten et al., 2008). We acknowledge that spontaneous breaths decoupled from locomotion may not represent the exact phenomenon of respiration while swimming. Nonetheless, esophageal pressures during the inspiratory pause ranged from 25 to 40 cmH₂O, which suggests that the chest wall is elastic and under significant recoil between breaths. Hydrostatic pressure may exert force on the chest, and it is likely that the subdermal connective tissue sheath and connective tissue in the blubber play a significant role in this process, effectively storing energy from respiratory and locomotory musculature as tension.

The custom-made pneumotachometer and fast-response gas analyzers allow us to estimate the V̇_O2 and V̇_CO2 on a breath-by-breath basis. The minimum and maximum values of the end-expiratory O₂ and CO₂ provided repeatable values for expirations >200 ms and 300 ms, respectively. The mean±s.e.m. sC_L of dolphins was 0.31±0.04 cmH₂O⁻¹, which is considerably higher than that of humans (West, 2012).

Similar to previously reported results in bottlenose dolphins (Kooyman and Cornell, 1981), most expiratory volumes were 5–40% larger than the inspiratory volumes, especially at high flows. One possible explanation for this disparity is related to a changing relationship between differential pressure and flow through the pneumotachometer (see additional discussion of equipment limitations below) (Finucane et al., 1972). Correction for the changes noted in the relationship over a wide range of flows, was made by performing calibration injections at varying flow rates, before and after each trial. Despite prior training and the ability of animals to ‘opt out’ of participation in the study, it is impossible to determine whether the chuffs where truly maximal breathing efforts. There was no difference in the breathing frequency before, during or after trials. We believe that these data represent valid, normal respiratory parameters for this species and may underestimate the maximal respiratory capacity of this species.

The respiratory flows, flow–volume curves and tidal volumes reported in the current study are slightly lower (140 l s⁻¹) than those reported for a single 285 kg female Pacific bottlenose dolphin (Tursiops truncatus gilli, Kooyman and Cornell, 1981). There are several possible explanations for this disparity, including differences in equipment, experimental protocols and physiological differences between sub-species. In the current study, most of the measured breaths were spontaneous rather than chuffs. Maximum flows were then averaged over a 20 ms period surrounding the peak measured flow rate. In the previous study, the peak flows were reported, and a much higher percentage of the breaths measured were chuffs. Despite the differences in absolute values, the overall shape of the flow–volume curves agrees well with those of the Pacific bottlenose dolphin (Kooyman and Cornell, 1981). The expiratory flows lasting approximately 300 ms show a rapid increase to a relatively constant expiratory flow followed by a rapid decline. The lack of a clear volume dependence of maximal expiratory flow and the variations in flow from effort to effort suggest that expiratory flows were effort dependent over most of the vital capacity, not limited by the lung as in terrestrial animals (Kooyman and Cornell, 1981; Kooyman and Sinnett, 1982). The inspiratory flow–volume curves in both studies were functionally square, with a fairly even flow at all inspiratory volumes and efforts (Fig. 6 in the current study and Fig. 5 in Kooyman and Cornell, 1981).

The maximal tidal volumes reported in this study were considerably lower than those previously reported by Kooyman and Cornell (1981). While this could represent true variation in respiratory capacity between subspecies, it may also indicate differences in the shape, size and configuration of the equipment used. These differences, as well as length, shape and symmetry of the pneumotachometer, significantly affect the flow characteristics and ability to achieve laminar flow, and thereby accurately measure flow over a wide range (Finucane et al., 1972). The pneumotachometer developed for this study was designed to optimize flow characteristics while at the same time minimizing dead space. In its industrial application, the laminar flow matrix used is rated to provide a linear response from 7.5 to 75 l s⁻¹, with minimal increase in resistance. Altering the housing of the flow cell is likely to have altered the flow characteristics (Finucane et al., 1972). The non-linear relationship between flow and differential pressure at flow rates >60 l s⁻¹ was accounted for by systematic calibration of the system at multiple flow rates (see Materials and methods). When corrected for the non-linear response at high flow rates, the very high flow rates measured during the expiratory phase were decreased by up to 40% and most of the disparity between expiratory and inspiratory volumes was resolved. Thus, the discrepancy between our flow rates and tidal volumes and those reported by Kooyman and Cornell (1981) could also be a result of non-linear response in the pneumotachometer at high flows.

The allometric mass exponent for tidal volume is close to 1 for terrestrial mammals (see table 1 in Stahl, 1967). Assuming isometry within cetaceans (but see Piscitelli et al., 2010), the average mass-specific tidal volumes recorded in the current study were 30.8 ml kg⁻¹ (range: 3.8–37.8 ml kg⁻¹) for voluntary breaths and 62.9 ml kg⁻¹ (25.5–70.5 ml kg⁻¹) for chuffs. The maximum mass-specific tidal volume reported for the Pacific bottlenose dolphin was 91.2 ml kg⁻¹ (Kooyman and Cornell, 1981) (Table 2). A range of 20.0 to 88.9 ml kg⁻¹ was reported in the pilot whale (Olsen et al., 1969), 27.5 to 58.8 ml kg⁻¹ in the Atlantic bottlenose dolphin (Irving et al., 1941; Ridgway et al., 1969), 39.3 to 52.6 ml kg⁻¹ in the harbor porpoise (Irving et al., 1941; Reed et al., 2000), 12.0 to 35.3 ml kg⁻¹ in the gray whale (Wahrenbrock et al., 1974) and 42.2 ml kg⁻¹ in the killer whale (Spencer et al., 1967). It is noteworthy that many of the parameters reported in the literature are for spontaneous breaths from animals that were not desensitized to the equipment and were either restrained or out of the water, versus this study, which utilized non-restrained, trained animals voluntarily making maximal respiratory efforts. The data collected in this study agree well with estimated tidal volumes previously reported in cetaceans. Furthermore, these data indicate considerable variation in tidal volume both within and between species, and is a subject that warrants further investigation.

Minimally invasive methods to determine end-tidal gas composition are of considerable importance, both for developing a basic understanding of physiology and because they have clinical relevance for veterinary medicine. Respiratory disease is a significant problem in cetaceans (Baker and Martin, 1992; Sweeney and Ridgway, 1976). Diagnostic methods can be invasive and are usually employed when the disease process is well progressed (Medway and Schryver, 1973). Pulmonary function testing and assessment of changes in gas exchange are established methods to assess respiratory disease in humans and animals. Use of these methods in cetaceans may provide veterinarians access to the tools currently available in the human medical field. For example, variation in end-tidal CO₂ levels is used to evaluate the appropriate matching of ventilation and perfusion in humans (Fletcher, 1980). The methods developed in this study simultaneously assess lung function and end-tidal gas levels. This provides an important, non-invasive tool to assess the clinical health of the respiratory system, in addition to illuminating new information in the study of basic cetacean respiratory physiology.

For breaths that exceed the dead space volume (volume of conducting airways and instrumentation combined), exhaled gas composition approaches alveolar concentrations (West, 2012). If gas exchange is not perfusion or diffusion limited, the end-expiratory P_O2 and P_CO2 should reflect pulmonary venous gas tension, and thereby provide a non-invasive estimate of systemic arterial blood gas P_O2 and P_CO2. Previously reported estimates of end-expiratory P_CO2 from healthy bottlenose dolphins range from 36.5 to at least 57 mmHg (Mortola and Sequin, 2009; Van Elk et al., 2001), which is within the average peak end-expiratory P_CO2 recorded in this study (Fig. 5B, average: 6.6±0.8% or 47.5±5.7 mmHg, assuming that the partial pressure of H₂O in air, at 37°C, is 47 mmHg). In the previous studies, the respiratory flows and respiratory timing (the duration of exhalation and inhalation) were not measured and the end-expiratory P_CO2 was assumed based on the shape of the capnogram. The gas analyzers used in the previous studies have reported response times that are significantly shorter (20–75 ms) than the unit used in the current study and were probably fast enough to record accurate end-expiratory gases, including end-tidal O₂ and CO₂ levels. The response time of gas analyzers in the current study were tested by bolus injections of a gas mixture of 5% CO₂ 95% N₂ into the line that pulled air into the gas analyzer. The bolus injections were made of varying durations (20 to 1000 ms) to determine the response times for the O₂ and CO₂ analyzers. The error of estimated end-tidal gas composition was estimated as the difference between the expected and observed gas content for varying bolus injection durations. The response time for a 95% response was 201 ms for the O₂ and 282 ms for the CO₂ analyzer (3×τ_1/2). Because of the short duration of the expiratory phase (310 ms during spontaneous breaths and 260 ms during chuffs) in the few breaths with a clear end-tidal plateau, the reported CO₂ end-expiratory values, in this study are likely to underestimate the true end-tidal CO₂. The end-expiratory O₂, however, is probably a good reflection of the true end-tidal value. However, the significantly lower end-expiratory O₂ for spontaneous breaths compared with chuffs suggest either hyperventilation or expiratory durations that are too short for the gas analyzer. Concurrent arterial blood samples will be required for confirmation of these values.

There was considerable variability in the estimated V̇_O2 and V̇_CO2. One possible reason is that trials in this study varied greatly in duration, ranging from 40 s to 4.5 min. Thus, for an animal with an average breathing frequency of 3 breaths min⁻¹ short trial duration may have a large influence on the estimate. Another possibility for the variability could be this study's use of breath-by-breath analysis in a species with very high expiratory flows. Flow-through respirometry is a common method used to estimate the metabolic rate (Bartholomew et al., 1981; Fahlman et al., 2004; Fedak et al., 1981; Withers, 1977). Flow-through respirometers are generally large compared with the study species, and the flow-through rate and effective volume are used to determine the duration required to reach a new steady state after changes in the V̇_O2 and/or V̇_CO2 (Bartholomew et al., 1981; Fahlman et al., 2008). However, breath-by-breath measurements provide direct quantification of the O₂ consumed and CO₂ produced for each breath. This type of measurement requires accurate concurrent quantification of the inspiratory and expiratory volumes as well as the gas composition. Several studies using breath-by-breath measurements to determine gas exchange have been conducted in pinnipeds and sirenians (Gallivan, 1980, 1981; Kooyman et al., 1971; Reed et al., 1994). For odontocetes, breath-by-breath analysis becomes more complicated because of the very high respiratory flows and short breath durations. To our knowledge, there has only been one previous study, in the harbor porpoise (Phocoena phocoena), where breath-by-breath analysis was attempted for estimation of metabolic rate (Reed et al., 2000). The measured V̇_O2 for the two harbor porpoises (each weighing 28 kg), ranged from 175 to 365 ml O₂ min⁻¹ (average 264 and 226 ml O₂ min⁻¹). Accurate matching of the instantaneous flow and expired gas composition is critical to estimate the V̇_O2 and V̇_CO2 in breath-by-breath analysis and rapid response gas analyzers are required for these applications. The short expiratory durations and high flows therefore may limit metabolic measurements with the current system, especially for breaths <200 ms. Despite these limitations, metabolic estimates from this study are close to those obtained from the harbor porpoise (Reed et al., 2000) and to metabolic rates reported in bottlenose dolphins using flow-through respirometry chambers (4.0–7.6 ml O₂ kg⁻¹ min⁻¹; see table 3 in Williams et al., 1993; Yazdi et al., 1999).

Few studies have tested the generally accepted hypothesis that repeated exposure to pressure during foraging has resulted in a more compliant respiratory system in marine mammals than their terrestrial counterparts (Fahlman et al., 2011; Kooyman and Sinnett, 1979; Leith, 1976; Olsen et al., 1969; Piscitelli et al., 2010). To our knowledge, the only study that has tested the mechanical properties of the respiratory system in an awake cetacean was that of a single pilot whale (Olsen et al., 1969). The estimated sC_L in that pilot whale (Globicephala scammoni) was 0.13 cmH₂O⁻¹, based on the reported lung compliance of 0.77 l cmH₂O⁻¹ and a measured residual volume at 5.6 l. An sC_L of 0.13 cmH₂O⁻¹ is lower than the average sC_L (0.31 cmH₂O⁻¹) found in the 6 bottlenose dolphins in the current study, but considerably higher than that of humans (approximately 0.083 cmH₂O⁻¹, West, 2012). The slightly lower sC_L in the pilot whale could reflect variation between species, the experimental methods, equipment used, or restraint of the pilot whale in a stretcher maintained approximately 10–20 cm above the whale's free-floating position (self-regulated buoyancy at the surface). The dolphins in this study were measured while free floating at the surface, under their own kinematic control. It is possible that even a small amount of pressure on the thorax and/or abdomen of a cetacean partially suspended in a stretcher may measurably affect the estimated sC_L.

When using esophageal pressures to estimate lung compliance, it is extremely important that the entire airway be open at the point of zero flow through the blowhole. Measurements may yield erroneous compliance estimates if the blow-hole or larynx is closed, if there is flow in the distal airways, or if there is volume acceleration. The timing for accurate estimates is critical, and the sequence of events occurs extremely rapidly during chuffs, making it challenging to determine transpulmonary pressure when flow is precisely zero. Direct measurement of distal tracheal pressure will be critical to verify these assumptions, but it was not possible in this study. During chuffs, P_L was often highly negative, which could suggest closed alveoli, airway closure, or expiratory flow-limitation. Therefore, chuffs were omitted from the calculations used to estimate C_L.

This study reports data on respiratory physiology in healthy bottlenose dolphins using a custom-made pneumotachometer, which support previous reports that this species has an exceptionally large respiratory capacity. Dolphins are able to exchange as much as 95% of their TLC during a single breath of <1 s with respiratory flows >130 l s⁻¹ during exhalation and >30 l s⁻¹ for inhalation. The pneumotachometer, coupled to a fast response gas analyzer, allowed investigators to measure metabolic rate and also provide data for end-expiratory O₂ and CO₂ that are estimated to approximate end-tidal values. This newly developed system for the study of lung function and mechanics in cetaceans may have significant value in respiratory physiology and clinical medicine.

Six adult male bottlenose dolphins (Tursiops truncatus Montagu 1821) of varying age and size (Table 1) were used for all experiments. The name, body mass, straight length and approximate age of the animals are summarized in Table 1. All experiments were performed using operant conditioning and participation by the dolphins was voluntary (the animals were not restrained and could refuse to participate or withdraw at any point during the experimental trial). Prior to initiating the study, animals were desensitized to the equipment and trained for novel research-associated behaviors using operant conditioning. Each experiment (trial) consisted of one animal staying stationary in the water, allowing placement of the appropriate equipment, and either breathing spontaneously or being asked to make a maximal respiratory efforts (chuffs), while continuous measurements were made. This approach allowed for collection of experimental data on lung function and respiratory mechanics in dolphins that were in a relaxed, normal physiological state, under a variety of circumstances.

A custom-made Fleisch type pneumotachometer (VMD Consulting, Miami, FL, USA) utilizing a low-resistance laminar flow matrix (Z9A887-2, Meriam Process Technologies, Cleveland, OH, USA) was placed over the blow-hole of the dolphin (Fig. 1). Differential pressure across the flow matrix was measured using a differential pressure transducer (MPX-2.5 mbar type 339/2, Harvard Apparatus, Holliston, MA, USA), connected to the pneumotachometer with two, 310 cm lengths of 2 mm i.d. firm-walled flexible tubing. The pneumotachometer was calibrated using a 7.0 l calibration syringe (Series 4900, Hans-Rudolph Inc, Shawnee, KS, USA). The signal was integrated and the flow determined assuming a linear response between differential pressure and flow. The linear response of the pneumotachometer was confirmed by calibrating with the 7.0 l syringe immediately before and after each trial, through a series of pump cycles at various flows. The pump cycles allowed the relationship between differential pressure and flows for the expiratory and inspiratory phases to be determined. To avoid spurious peaks, the reported maximal inspiratory and expiratory flows are the average flows over 20 ms; 10 ms on either side of the maximal recorded inspiratory or expiratory flow.

Airway opening pressure (P_ao) was measured via a sample port immediately above the blow-hole connected to a differential pressure transducer (MPX-100 mbar type 339/2, Harvard Apparatus) connected to the pneumotachometer with a 310 cm length of 2 mm I.D, firm walled, flexible tubing. Esophageal pressure (P_eso) was measured using an esophageal balloon catheter (47-9005, Cooper Surgical, Trumbull, CT, USA) connected to a differential pressure transducer (MPX-100 mbar type 339/2, Harvard Apparatus) by a 288 cm length of 2 mm i.d. firm-walled flexible tubing, through a 3-way stopcock. The balloon catheter was manually inserted into the esophagus to the approximate level of the heart and inflated with 1.0 ml of air (Fig. 1). Reference pressure for both P_ao and P_eso was ambient atmospheric pressure (P_amb).

Differential pressure transducers were connected to an amplifier (Tam-A, Harvard Apparatus). The data from the transducers were captured at 400 Hz using a data acquisition system (Powerlab 8/35, ADInstruments, Colorado Springs, CO, USA), and displayed on a laptop computer running LabChart (v7.3.7, ADInstruments). All differential pressure transducers were zeroed immediately before each trial.

The dynamic response of the system configured as described above was evaluated for a step response by balloon deflation test. Popping an inflated balloon with a needle provided an immediate step change in pressure. The time constant (τ) was estimated as the time to 50% pressure reduction, or τ_1/2. The dynamic constant was 7 ms for the pneumotachometer pressure line and 40 ms for the esophageal catheter.

Respiratory gasses were subsampled via a port in the pneumotachometer and passed through a 310 cm length of 2 mm i.d. firm-walled flexible tubing and a 30 cm length of 1.5 mm i.d. Nafion tubing, to fast-response O₂ and CO₂ analyzers (ML206, Harvard Apparatus) at a flow rate of 200 ml min⁻¹. The gas analyzers were connected to the data acquisition system and sampled at 400 Hz. The gas analyzers were calibrated before and after the experiment using a commercial mixture of 5% O₂, 5% CO₂ and 90% N₂. Ambient air was used to check the calibration before and after each experimental trial. Mean daily air temperature and humidity were 24.3±0.6°C (range 23.2–24.8°C) and 76.0±1.9% (72–79%). The average water temperature in the lagoon was 23.6±0.3°C.

The manufacturer-specified response times for a 90% change to equilibrium for O₂ and CO₂ were 130 ms and 90 ms, respectively, at a flow rate of 200 ml min⁻¹. Thus, the analyzers may not respond fast enough to accurately measure end-tidal gas composition in species with short exhalation durations (<300 ms). The dynamic response time of the gas analyzers was tested by bolus injections of a gas mixture containing 5% CO₂, 95% N₂ into a stream of air using a 3-way solenoid valve (Clippard Instrument Laboratory Inc. EVO-3-6) that was controlled by the LabChart program. The duration of bolus injections was varied from 20 ms to 1000 ms, a range that spanned all of the expiratory durations measured during these experiments. The measured τ_1/2 for the O₂ (τ_1/2,O2) and CO₂ (τ_1/2,CO2) analyzers were 67 ms and 94 ms, respectively.

C_L estimated as the tidal volume divided by the tidal change in transpulmonary pressure (P_L=P_ao−P_eso), measured at zero flow (Olsen et al., 1969), which was estimated as the tidal volume divided by the tidal change in transpulmonary pressure (P_L=P_ao−P_eso) measured at zero flow. The esophageal pressure trace was phase corrected by 40 ms (the dynamic response time of the esophageal catheter system) and P_eso was determined at the points of zero flow of the expiratory phase and inspiratory phases (a and b, respectively, in Fig. 2C). End-expiratory transpulmonary data measurements of <−5 cmH₂O were excluded, as these data were most likely flow-limited or suggestive of closed alveoli. Therefore, no chuffs were used to estimate C_L.

The respiratory gas signals were phase corrected by 3.94 s for CO₂ and 3.79 s for O₂, to match the respirations, and the expiratory flow-rate and expired O₂ and CO₂ content were multiplied to calculate the instantaneous V̇_O2 and V̇_CO2. The instantaneous V̇_O2 and V̇_CO2 were integrated over each breath to yield the total volume of O₂ and CO₂ exchanged during each breath. The volumes were summed for each trial period and divided by the duration of the trial to provide an estimate of the oxygen consumption and carbon dioxide production rates for that time period.

All gas volumes were converted to standard temperature pressure dry (STPD; Quanjer et al., 1993). Exhaled air was assumed saturated at 37°C, inhaled air volume was corrected for ambient temperature and relative humidity. Metabolic data are reported as the average O₂ consumption rate or CO₂ production rate for an entire trial. Compliance data are reported as the average for all breaths in a trial (Table 1). The relationship between a dependent variable and experimental covariates was analyzed using linear-mixed effects models (lme, R v3.1.0, R Foundation for Statistical Computing, 2014). The individual animal was treated as a random effect, which accounted for the correlation between repeated measurements on the same individual (Littell et al., 1998). Best models were chosen by the Akaike information criterion (AIC) against the null model (AIC_null) and significant parameters assessed by the t-value between the estimate and its standard error. In this study P-values ≤0.05 were considered as significant and P≤0.1 were considered a trend. Data are presented as the mean±s.d., unless otherwise stated.

A special thanks to all the trainers and staff at Dolphin Quest, Oahu, who made this study possible. Also, thanks to Danielle Kleinhenz and Jamie Kleinhenz for assistance during the trials. We are grateful to Dr Robert Elsner for providing insights into the lung mechanics study on the Pilot whale published in 1969. We would like to thank ADI, Harvard Apparatus, Meriam Process Technologies, Stephanie Pazniokas and Nicole Ruscavage for their continuous support with our numerous calls for technical support.