Pulmonary compliance and lung volume varies with ecomorphology in anuran amphibians: implications for ventilatory-assisted lymph flux

The lungs of anuran amphibians, in addition to gas exchange, are also crucial for vocalization, buoyancy in the aquatic environment and defensive behaviors (Hillman et al., 2009). We have also recently described a new role for lung ventilation in the vertical movement of lymph from ventral regions of the body to the dorsally located lymph hearts in anurans (Hedrick et al., 2007; Hillman et al., 2010). We hypothesize that lung deflation moves lymph vertically by creating a negative pressure ‘suction’ effect to lift lymph from the limbs to the subvertebral lymph sac overlying the lungs. Several lines of evidence support this hypothesis. First, direct measurement of subvertebral lymph sac pressure during ventilation indicates that the negative pressures that develop in the subvertebral sac during lung exhalation are sufficient to move lymph from ventral regions to the subvertebral sac, which communicates directly with the posterior lymph hearts (Hedrick et al., 2007). Second, lung inflation compresses the lymph sacs surrounding the lungs (e.g. lateral and abdominal sacs) (Carter, 1979) and pushes lymph toward the lymph hearts (Hedrick et al., 2007). Third, insertion of a plastic coil into the subvertebral sac interferes mechanically with the capacity of the lungs to completely inflate or deflate, and significantly reduces lymph flux from the hind limbs (Hillman et al., 2010). Finally, lung ventilation is also coordinated with contraction of a variety of skeletal muscles that insert on the skin in the axillary regions and the urostyle (Drewes et al., 2007), which also contribute to lymph flux (Hillman et al., 2010).

Anurans vary markedly in body shape, from extremely dorso-ventrally flattened aquatic (e.g. Pipa pipa) to globose fossorial species (e.g. Breviceps spp.) (see Hillman et al., 2009), and their lungs occupy a large fraction of the pleuroperitoneal space (Lauridsen et al., 2011). Thus, lung volume relative to body mass should be largest for globose anurans, but nevertheless large for other amphibians, although this has not been systematically examined.

Pulmonary compliance (C) is defined as the change in lung volume per unit change in intrapulmonary pressure (ΔV/ΔP). Compliance is also the product of distensibility (D; percent change in volume/change in pressure) and lung volume (V_lung), thus differences in body mass and relative lung volume (e.g. globose vs dorso-ventrally flattened) are likely to contribute to species differences in lung compliance. Distensibility is a measure of the surface tension of the air–lung interface and the elastic forces of the lung tissue and body wall, and hence might be similar for different anurans. Although the role of lung compliance and volume to the contribution of lymph movement is unclear, it is possible that the ability of anurans to mobilize lymph is related to pulmonary factors such as pulmonary compliance and lung volume, particularly as lung ventilation is important for lymph movement (Hedrick et al., 2007).

The lymph flux of anurans is correlated with ecomorphology; that is, terrestrial species have significantly higher lymph flux rates than semi-terrestrial or aquatic species (Hillman et al., 2011). Because lung ventilation plays a key role in the movement of lymph, we hypothesize that pulmonary compliance and lung volume is also related to anuran ecomorphology, with greater pulmonary compliance and lung volume in more terrestrial species that require greater lymph mobilization. Our hypothesis is that species with higher pulmonary compliance in the physiological range of intrapulmonary pressures should be able to more effectively move lymph and stabilize plasma volume during dehydration. Consequently, compliance should be greater for terrestrial species than for aquatic species.

To our knowledge, the pulmonary compliance of anurans has only been examined previously for two species within the family Ranidae, Rana (=Lithobates) pipiens (West and Jones, 1975b; Hughes and Vergara, 1978) and Rana (=Lithobates) catesbeiana (Dupré et al., 1985). These studies compared the inflation/deflation characteristics of anuran lungs with those of mammalian species. In light of our recent studies demonstrating a role for ventilation in the movement of lymph (Hedrick et al., 2007; Hillman et al., 2010), the present study examined pulmonary compliance and lung volume in selected representatives of three anuran families. The members of these families are found in habitats including terrestrial/semi-terrestrial (Bufonidae), semi-aquatic (Ranidae) and aquatic (Pipidae) (Hillman et al., 2009). We further examined the relationship between pulmonary variables and lymph flux data that are available (Hillman et al., 2011) for three species representing these families: Rhinella marina (semi-terrestrial, xeric), Lithobates catesbeianus (semi-aquatic) and Xenopus laevis (aquatic). These species differ in dehydration tolerance (R. marina>L. catesbeianus=X. laevis) (Hillman, 1978; Hillman, 1987; Hillman et al., 1987) and the degree of development of skeletal muscles associated with lymph flux (R. marina>L. catesbeianus>X. laevis) (Drewes et al., 2007). Our prediction was that pulmonary compliance and lung volume would show a similar pattern given the importance of lung ventilation to lymph flux in these species.

Amphibians were purchased from commercial suppliers. They were maintained at 24°C in conditions appropriate to their natural habitat: water for aquatic species, water and dry areas for the other species. Five species of Bufonidae [green toad, Anaxyrus debilis Girard 1854, 3.9–5.2 g, N=6; Great Plains toad, Anaxyrus cognatus Say 1822, 66.7–109.2 g, N=7; red-spotted toad, Anaxyrus punctatus Baird & Girard 1852, 14.0–25.2 g, N=6; cane toad, Rhinella marina (Linnaeus 1758), 39.0–160.0 g, N=19; common Asiatic toad, Duttaphrynus melanostictus Schneider 1799, 23.7–63.3 g, N=10], two species of Ranidae [North American bullfrog, Lithobates catesbeianus Shaw 1802, 55.0–151.0 g, N=17; southern leopard frog, Lithobates sphenocephalus (Cope 1886), 8.1–48.5 g, N=8] and two species of Pipidae [African clawed frog, Xenopus laevis Daudin 1802, 25.6–69.2 g, N=11; Surinam toad, Pipa pipa (Linnaeus 1758), 93.5–204.5 g, N=9] were studied. These species were chosen to encompass a large range of body masses to investigate family-level allometric relationships for pulmonary variables. All experiments were conducted at 20°C. All animals were salvaged from experiments that measured apparent lymph volumes in the femoral lymph sac of frogs that were anesthetized in buffered MS-222 (see Hillman et al., 2011). Following lymph volume measurement, and while anesthetized, the animals were cranially pithed and then decapitated for measurements of pulmonary compliance, distensibility and lung volume as detailed below. All procedures were conducted at Portland State University with IACUC approval (protocol no. 10.01.11.1).

For each animal, a flared tracheal catheter, fashioned from polyethylene tubing of appropriate size for the species used, was inserted through the glottis. The catheter was anchored and sealed with a combination of sutures (4-0 silk) and cyanoacrylate adhesive. The catheter was connected to a T-junction, which had a syringe and infusion pump (Harvard Apparatus, model 975, Holliston, MA, USA) on one arm and a volumetric pressure transducer (Grass model PT5A, Braintree, MA, USA) on the other. The pressure transducer was calibrated using static water column pressure. Voltage output from the pressure transducer was digitized and recorded with a data acquisition system (MacLab Chart, ADInstruments, Colorado Springs, CO, USA) and computer. An infusion pump delivered air to inflate the lung at a rate of approximately 20% of body volume min^–1. Because inflation was continuous, the measurements represent dynamic compliance. The compliance of the syringe, tubing and transducer was 0.12±0.002 ml cmH₂O^–1, and was subtracted from the total pulmonary compliance measurement.

We chose to determine compliances (ΔV/ΔP) within approximately 1–4 cmH₂O (0.1–0.4 kPa) because anurans normally experience this range of in vivo intrapulmonary pressures over a ventilatory cycle (DeJongh and Gans, 1969; West and Jones, 1975a; Vitalis and Shelton, 1990; Wang, 1994). The mean compliance over the pressure intervals of 0.5–1.5, 1.5–2.5, 2.5–3.5 and 3.5–4.5 cmH₂O were calculated from each trace, corresponding to means of 1.0, 2.0, 3.0 and 4.0 cmH₂O. Lung inflation was repeated five to 10 times for each individual, and the results are the means of those repeated inflation values (see Fig. 1). The mean volume change over each of these intervals was derived from the air infusion rate (i.e. 20% of body volume min^–1). The mean pulmonary compliance () was the mean compliance of these four pressure intervals, and the maximal compliance (C_max) was the highest compliance measured at the four pressure intervals over which compliances were determined. The volume of air necessary to achieve an intrapulmonary pressure of 4 cm H₂O was also calculated from the pressure trace and reported as total lung volume (V_lung). Pulmonary distensibility (D_pul) was calculated from the pressure–volume relationship at 1 cmH₂O intervals by dividing the compliance by C_max measured at each 1 cm H₂O interval and is expressed as % cmH₂O^–1.

Data are reported as means ± s.e.m. unless otherwise noted. Standard linear regression analysis and one-way ANOVA were used as appropriate. All data analyses were conducted using GraphPad Prism v. 5 (San Diego, CA, USA).

There were highly significant (P<0.001) correlations (r²=0.58 to 0.83) for the allometric relationships of C_max, and V_lung for each family (Table 1; Fig. 2). Furthermore, there were clear family-level differences in the relationship between pulmonary variables and body mass. Bufonidae had higher intercepts and lower slopes for C_max, and V_lung, whereas Pipidae had lower intercepts and higher slopes; Ranidae were intermediate for both intercepts and slopes.

Pulmonary compliance and lung volume measured for the species in this study (Table 2) represent the mass-specific values based upon family-level allometric equations (Table 1). Bufonidae species had higher values for C_max, and V_lung compared with Ranidae, which had higher values than Pipidae (Table 2). Overall, C_max for Bufonidae was 105.1±4.6 ml cmH₂O^–1 kg^–1 (N=48), which was significantly greater (P<0.001) than that for Ranidae (55.7±0.3 ml cmH₂O^–1 kg^–1, N=28) and both were significantly greater than that for Pipidae (34.4±1.0 ml cmH₂O^–1 kg^–1, N=20). There was considerable variation in C_max among the Bufonidae species, with values ranging from 83.2±1.6 cmH₂O^–1 kg^–1 in A. cognatus to 177.5±2.2 cmH₂O^–1 kg^–1 in A. debilis. Lung volume followed a similar pattern, with Bufonidae (310.2±13.6 ml kg^–1, N=44) being significantly higher (P<0.001) than Ranidae (144.0±1.3 ml kg^–1, N=20), both of which were significantly higher (P<0.001) than Pipidae (74.6±3.8 ml kg^–1, N=20). Lung volume also exhibited a large variation among the Bufonidae, ranging from 242.1±5.4 ml kg^–1 in R. marina to 503.8±5.8 ml kg^–1 in A. debilis.

There was no significant allometric relationship for D_pul and body mass for Bufonidae, Ranidae or Pipidae at any lung pressure, except in two cases: D_pul for Bufonidae at pulmonary pressures of 3 and 4 cmH₂O were significantly related to body mass (P<0.02 and <0.01, respectively). However, body mass explained very little of the variation in D_pul (r²=0.11 at 3 cmH₂O and r²=0.15 at 4 cmH₂O) (data not shown). Because there was no significant allometric relationship with D_pul, we examined family-level differences in D_pul at each pulmonary pressure. We did find significant differences in D_pul at 2, 3 and 4 cmH₂O, but these differences were not consistent with differences in ecomorphology. For example, at 2 cmH₂O, D_pul for Ranidae was 32.6±1.0% cmH₂O^–1 and was significantly higher (P<0.005) than that for either Bufonidae (29.2±1.1% cmH₂O^–1) or Pipidae (26.1±1.4% cmH₂O^–1) (Fig. 3). D_pul for Pipidae was significantly higher (P<0.01) than that for either Bufonidae or Ranidae at 3 and 4 cmH₂O (Fig. 3). However, the differences in D_pul at these lung pressures were small and were revealed because of the large sample size and low variance of D_pul (Fig. 3).

The relationships between C_max, and V_lung measured in this study and lymph flux measured for R. marina, L. catesbeianus and X. laevis (Hillman et al., 2011) were significant for the femoral (Fig. 4) and subvertebral (Fig. 5) sac lymph for all three species. The consistent pattern was that lymph flux and each pulmonary variable was significantly higher (P<0.001) for the most terrestrial species (R. marina), the lowest for the aquatic species (X. laevis) and intermediate for the semi-aquatic species (L. catesbeianus).

Consistent with our hypothesis, pulmonary compliance and lung volume were correlated with habitat type (ecomorphology) for the nine species of anurans examined in this study. The terrestrial and semi-terrestrial Bufonidae species had the highest values for C_max, and V_lung, the semi-aquatic Ranidae species had intermediate values, and the aquatic Pipidae had the lowest values. These pulmonary values were also strongly correlated with the ability to mobilize lymph for three representative species, with higher lymph fluxes in R. marina, followed by L. catesbeianus and X. laevis (Hillman et al., 2011) (Figs 4, 5). These patterns suggest that variation in ecomorphology and pulmonary factors such as compliance and volume may, in part, account for some of the variation in lymph mobilization capacity.

Previous work has examined pulmonary compliance for only two Ranidae species (Lithobates pipiens and L. catesbeianus). Maximum mass-specific pulmonary compliance of adult L. pipiens during lung inflation ranged from 22 to 49 ml cmH₂O^–1 kg^–1 (West and Jones, 1975b; Hughes and Vergara, 1978), which is slightly lower than our values for L. catesbeianus and L. sphenocephalus (Table 2). Because we did not measure pulmonary compliance for L. pipiens, these differences may be due to variation between species. The values from previous studies are total pulmonary compliance (i.e. lungs and body wall measured in situ) and are therefore directly comparable to the values obtained in this study. Dupré et al. measured volume-specific compliance of isolated lungs during development for bullfrogs, but did not provide values for body mass (Dupré et al., 1985). Their value for lung compliance, measured in air, was approximately 35 ml cmH₂O^–1 for adult L. catesbeianus. Our measurement of absolute C_maxin situ was 6.2±0.8 ml cmH₂O^–1 for L. catesbeianus, approximately five times lower than the value for isolated lungs. Because our measurement included the lungs and body wall, we would expect our values for C_max to be significantly less than the compliance of isolated lungs. Body wall compliance (C_B) can be estimated as C^–1_B=C^–1_T–C^–1_L, where C_T is total pulmonary compliance (=C_max) and C_L is compliance of the isolated lungs. Combining the data for C_L from Dupré et al. (Dupré et al., 1985), we calculate a C_B of 7.5 ml cmH₂O^–1 for L. catesbeianus, indicating that compliance of the body wall is nearly equal to the total respiratory system compliance. A previous study of the Tokay gecko (Gekko gecko) also found that C_T and C_B were approximately equal, owing to the high C_L (Milsom and Vitalis, 1984).

Pulmonary compliance of the anurans in this study and from previous studies (Hughes and Vergara, 1978; Dupré et al., 1985) is approximately an order of magnitude greater than for mammals of comparable body mass (Stahl, 1967). Indeed, our values for Bufonidae species, and for A. debilis in particular, are among the highest mass-specific pulmonary compliance and lung volume values reported for vertebrates. Mass-specific pulmonary compliance and lung volume is also much higher for reptiles than mammals (Tenney and Tenney, 1970; Perry and Duncker, 1978; Perry and Duncker, 1980; Milsom and Vitalis, 1984; Bartlett et al., 1986; Lutcavage et al., 1989; Klein et al., 2003). Reptiles also have a very large variation in total pulmonary system compliance ranging from approximately 8 ml cmH₂O^–1 kg^–1 for the turtle Pseudemys scripta (Vitalis and Milsom, 1986) to 466 ml cmH₂O^–1 kg^–1 for the chameleon Chamaeleo chamaeleon (Perry and Duncker, 1978). Whether this large variation in pulmonary compliance represents family-level differences, as we have shown with anurans, is not clear. It has been suggested that the high pulmonary compliance and lung volume of reptiles also represents a functional compromise between gas exchange and other important functions such as buoyancy and display (Perry and Duncker, 1978).

Although we have shown that pulmonary compliance and lung volume are well correlated with the degree of terrestriality for anurans, gas exchange does not show the same correlation for these families. Maximal oxygen consumption (V_O2,max) from 16 studies of Bufonidae was 0.97±0.06 ml O₂ g^–1 h^–1, compared with 0.50±0.05 ml O₂ g^–1 h^–1 for Ranidae (17 studies) and 0.77±0.08 ml O₂ g^–1 h^–1 for Pipidae (four studies) (Gatten et al., 1992). If gas exchange were the principal driving force for the variation in lung volume, we would expect species with the highest metabolic rates to also have the highest values for lung volume. Because this was not the case, the results of the present study suggest that habitat preference, degree of terrestriality and lymph mobilization capacities may explain more of the variation in pulmonary compliance and lung volume. However, it will be necessary to investigate these relationships for species representing a broader array of families, using a more detailed phylogenetic analysis, before any firm conclusions can be drawn.

Unlike pulmonary compliance and lung volume, pulmonary distensibility was not consistently correlated with body mass. We found that D_pul varied slightly as a function of lung pressure, but the variation was not consistent with the ecomorphology of these species. The lungs of anurans comprise relatively simple sac-like structures in some species or have ‘sac alveoli’ bounded by primary, secondary or tertiary septa that increase the respiratory surface area (Tenney and Tenney, 1970; Goniakowska-Witalinska, 1986; Maina, 1989; Maina, 2002). Pipidae have lower lung septation compared with Ranidae and Bufonidae (Tenney and Tenney, 1970), and this may account for our finding that D_pul was higher for Pipidae at higher pulmonary pressures. The ratio of total lung septal area to total lung surface area is roughly equal for Bufonidae and Pipidae, but lower in Ranidae (Czopek, 1965). It is likely that such variations in lung structure account for the variation in D_pul between the three families in the present study. A previous study found that volume-specific lung compliance did not vary significantly during development for L. catesbeianus (Dupré et al., 1985). Because pulmonary compliance is the product of distensibility and lung volume, and D_pul was essentially constant, our data indicate that variation in pulmonary compliance is primarily a function of differences in lung volume.

It has been noted previously that V_lung and pulmonary compliance are larger for amphibians compared with other vertebrates (Tenney and Tenney, 1970; Hughes and Vergara, 1978). Maximal lung volume scaled to body mass with a slope of 1.05 for a number of different amphibians (Tenney and Tenney, 1970). The slopes for V_lung in our study ranged from 0.76 for Bufonidae to 1.33 for Pipidae, which encompasses the range found previously. It should be noted that the allometric slopes obtained previously were for dried, isolated lungs inflated to a pressure of 20 cmH₂O, well outside the normal physiological range of lung pressures (Tenney and Tenney, 1970). However, despite the very different methods for measuring volume, V_lung scaled similarly as a function of body mass in our study and previous studies.

Lung volume measured in previous studies also exhibits considerable variation. For example, V_lung of L. catesbeianus ranges from approxiately 80 to 100 ml kg^–1 for awake animals (Tenney and Tenney, 1970; Glass et al., 1981), but can reach as high as 300 ml kg^–1 during anoxia (Tenney and Tenney, 1970). Maximal V_lung of L. pipiens ranged from 35 ml kg^–1 for awake animals (West and Jones, 1975b) to approximately 210 ml kg^–1 for anesthetized animals with artificial lung inflation (Hughes and Vergara, 1978). Our values for two species of Ranidae (139–151 ml kg^–1) encompass previous values for anesthetized and awake animals, indicating that our measurements from pithed animals were not substantially different.

We have shown previously that lung ventilation is important in anurans for lymph movement (Hedrick et al., 2007; Hillman et al., 2010). In the present study, we found that the more terrestrial species (Bufonidae) have relatively larger lungs (cf. Lauridsen et al., 2011) than either Ranidae or Pipidae, and are likely to be more effective at transmitting pressure to the surrounding lymph sacs during lung inflation. Inflation of the lungs in anesthetized animals causes increases in pressure of the lymph sacs surrounding the lungs (Hedrick et al., 2007), thus lung inflation would facilitate movement of lymph toward the lymph hearts by ‘squeezing’ lymph sacs against the body wall.

We hypothesize that during lung deflation when lung volume decreases, dorsally located lymph sacs, such as the subvertebral and pubic sacs (Kampmeier, 1969; Carter, 1979), expand, thus creating a negative lymphatic pressure that serves to ‘suction’ lymph from the forelimb and hindlimb lymph sacs. The forelimb and hindlimb lymph sacs are connected to subvertebral and pubic sacs via one-way valves [see fig. 3 in Hillman et al. (Hillman et al., 2011)]. In support of this hypothesis, negative pressures measured in the subvertebral and pubic lymph sac are of sufficient magnitude to move lymph vertically during lung deflation in R. marina and L. catesbeianus (Hedrick et al., 2007). In addition, interfering with the lungs' ability to inflate and deflate normally by placing a plastic ‘coil’ in the subvertebral sac of awake R. marina significantly reduced lymph flux from the hindlimb femoral sac (Hillman et al., 2010). The reduction in lung volume and, therefore, pulmonary compliance, clearly interferes with the mechanism involved in the vertical movement of lymph from ventral to dorsal locations. The more terrestrial species, R. marina, generates greater negative pressures in the pubic lymph sac than does L. catesbeianus (Drewes et al., 2007), and this is associated with higher lymph flux rates in R. marina (Hillman et al., 2011) (Figs 4, 5). Because changes in lung volume result in negative pressures in lymph sacs that assist in lymph movement, a more compliant lung should produce a greater change in volume for a given change in pressure during exhalation. For example, if we compare pulmonary compliance for 100 g animals from each family using the equations in Table 1, C_max is 7.85 ml cmH₂O^–1 for Bufonidae, 5.0 ml cmH₂O^–1 for Ranidae and 3.6 ml cmH₂O^–1 for Pipidae. It is clear that with a larger compliance and lung volume, an equivalent reduction in lung pressure will result in a larger change in lung volume for Bufonidae than for either Ranidae or Pipidae. The larger change in lung volume would translate into a larger change in lymph sac volume and pressure. Thus, we hypothesize that more compliant lungs, associated with more terrestrial species, are more effective for lymph mobilization than smaller, less compliant lungs.

Another key component of this model is likely the large difference between C_L and C_B (see above), which facilitates volume expansion of the subvertebral lymph sac. The subvertebral lymph sac is bounded dorsally by the rigid backbone and is ventrally attached to the lung surface by a thin membrane. During expiration, when lung volume decreases and pressure decreases, the relatively stiffer body wall (lower compliance) compared with the lungs and lymph sacs would allow the expansion of the subvertebral sac overlying the lungs. The expansion of the subvertebral sac creates the negative pressure necessary to move lymph from the limbs to the subvertebral space. However, more work is necessary to explicitly test the role of compliance and volume in the movement of lymph in anurans. Overall, our results suggest that pulmonary compliance and lung volume may play important roles in ventilatory-assisted lymph flux.

We are indebted to Mohammedreza Atashzareh for his outstanding assistance in performing the experiments.