The relationship between heart rate and rate of oxygen consumption in Galapagos marine iguanas (Amblyrhynchus cristatus) at two different temperatures

Heart rate has not previously been used to predict FMR of reptiles. In these animals, the relationship between V̇_O2 and fH may be complicated by variations in metabolic rate associated with changes in ambient temperature and the possible variation in shunting between the left and right sides of the heart. Thus, the aim of the present study was to determine the relationship between V̇_O2 and fH in a reptile at two different temperatures that represent the extremes of its average daily range during summer(Fig. 1). Using implantable heart rate data loggers (HRDL; Woakes et al., 1995), this relationship will be used in ongoing field studies employing HRDLs to determine FMR and the energetic costs of specific behaviours (see Fig. 4).

The reptile chosen was the Galapagos marine iguana, Amblyrhynchus cristatus. This is the only lizard that dives beneath the sea to feed and whose food is primarily marine algae. There appears to be a size limit to these animals that is related to the availability of food, with smaller animals out-competing larger ones when food availability declines(Wikelski et al., 1997). During particularly lean (El Niño) years, some animals `shrink'(reduction in body length as well as in body mass), and those that `shrink'the most survive the longest (Wikelski and Thom, 2000). On top of this is the cost of reproduction. Females actively choose mates (Wikelski et al.,2001) and suffer a cost of reproduction in terms of a lower probability of survival during the following season(Laurie, 1989). Thus, there are many potential applications for the fH method in this species,which occupies an unusual niche for a lizard, in order to determine the energy costs of specific behaviours.

This study was performed on seven marine iguanas Amblyrhynchus cristatus Bell at the Darwin Research Station, Santa Cruz Island (SC iguanas), and on four iguanas on board the vessel Quest anchored off the Island of Santa Fé, Galapagos Islands, Ecuador (SF iguanas). The iguanas were captured by noosing or by hand and held in captivity for no longer than 48 h before being used. The body mass (mean ± S.D.) of the SC animals was 1.39±0.31 kg (see Table 2 for individual values),while that of the SF animals was 1.35±0.55 kg. Heart rate data loggers were implanted into the abdominal cavity of the iguanas while they were anaesthetised with a mixture of Isoflurane (Abbott Laboratories, USA) and air. Upon exposure to the mixture, the iguanas usually stopped breathing for several minutes. However, once they started to inhale the Isoflurane, they immediately became deeply anaesthetised. They were then positioned upside down between two foam pads. The skin and underlying muscles were opened along a 3.5-4 cm long incision, which was approximately 2 mm off the ventral midline,in order to avoid a small vein.

The sterilised loggers were inserted with one electrode close to the heart and the other laying along the side of the HRDL. The body of the HRDL was fixed in place with two sutures of surgical silk (#2, Dexon, Germany) through the body wall. The logger incorporated a low-power radio frequency transmitter which emitted a short pulse on each QRS wave of the electrocardiogram (ECG). Detection of this signal by a radio receiver indicated when the electrodes were in the correct position. The muscle layer and skin were individually closed together with dissolvable surgical suture (#1, Dexon, Germany) and an antibiotic/antifungal spray (Chloromycetin, Parke Davis, USA) was used on the wound after surgery. The animal was then taken off Isoflurane and woke within 5 min. The iguanas were left for at least 24 h before they were used in an experiment.

To determine whether iguanas suffered from post-surgical stress after the implantation of data loggers, we took a blood sample from the tail vein of four implanted females approximately 24 h after surgery and of six control females that had not undergone surgery. Plasma corticosterone levels were determined using a standard radioactive immuno-assay procedure (Romero and Wikelski, 2000). There was no difference in the levels of corticosterone between the two groups (implanted, 6.3±1.8 ng ml^-1; control,5.4±1.4 ng ml^-1; means ± S.D., t-test, P=0.39). Thus, we conclude that iguanas do not suffer post-surgical trauma that would be indicated by an increased corticosterone level compared to that of controls. None of the animals showed signs of infection or discomfort and all data loggers were removed 3 days after implantation. All the animals survived the surgical procedures and four of the SC animals were seen 9 months later at the location where they had been caught and released.

The animals were studied either in the early morning or early afternoon when their body temperatures were at approximately 27 °C or 36 °C,respectively. When required, the use of a refrigerator or an infrared lamp enabled us to maintain the animals at these temperatures. Body temperature was determined by a thermocouple placed 3-4 cm into the cloaca. Once body temperature had been maintained close to the required value for at least an hour, the animal was fitted with a loop aerial on the top of its body. This enabled the transmitted heart beat signal from the implanted data logger to be detected by a radio receiver and the output from the receiver to be sent to a pre-amplifier (Isleworth, Electronics, England, model A101) and the signal appropriately filtered.

A transparent mask constructed from a plastic water bottle was placed over the head of the animal and held in place with a rubber collar around the neck. An airtight seal between the collar and the skin of the iguana was achieved with a layer of quick-setting, non-toxic polyether material (Impregum, ESPE Dental AG, Germany). The mask was fitted with inlet and outlet tubes through which air was drawn at a rate of approximately 2.61 min^-1 STPD by a pump (Reciprotor, Denmark, model 506R) on the outlet side. The air flow rate was set and monitored by a mass flow meter and controller (Sierra, models 840L and 902C). A subsample of the air leaving the pump was passed through a drying column (Drierite, Hammond) and analysed for the fractional content of O₂ and CO₂ by a gas analyser (ADInstruments, model ML205). The gas analyser was calibrated with room air and was accurate to 0.01% for both gases. Outputs from the ECG pre-amplifier and gas analyser were collected at 1 kHz (Powerlab 800, ADInstruments) and displayed on a computer using Chart software (ADInstruments) as heart rate and rate of oxygen consumption. Rate of oxygen consumption was determined from the airflow through the mask and the difference between incurrent and excurrent fractional concentrations of dry air following consideration of respiratory quotient(RQ)-related errors (see Appendix in Frappell et al., 1992).

After instrumentation, the iguanas were placed on a variable-speed treadmill (1.2 m long and 0.5 m wide) and allowed to settle for at least 30 min, when fH and V̇_O2 had reached steady (pre-exercise) values. They were then run at the maximum speed that they could comfortably maintain for a few minutes (maximum exercise). Although this included bursts of locomotion, the animals were not run to exhaustion. When an iguana no longer wanted to run, the treadmill was stopped and recordings continued for approximately 60 min during the recovery phase (see Gleeson, 1980, for recovery times of V̇_O2after exhaustive exercise in marine iguanas). Rate of oxygen consumption and fH data were obtained from each animal during the pre-exercise period, at maximum exercise (one datum point at each) and at four approximately equally spaced points during recovery. Data were averaged over 30-60 s. At the Darwin Station, each iguana was run, in random order, at body temperatures of 27 °C and 36 °C. On board the Quest, the animals were run only at a body temperature of 36 °C and values of fH and V̇_O2 were recorded only during the pre-exercise period and at maximum exercise. All values of V̇_O2 are at standard temperature and pressure, dry (STPD).

Least-squares regressions were used to determine the relationships between fH and V̇_O2 for individuals and for the group data at the two different temperatures for the SC animals. Regression equations were compared using an analysis of variance(ANOVA) general linear model (GLM; Zar,1984) and, after testing for normality (Kolomogorov-Smirnov test),a Student's t-test was used to compare the significance of any difference between the means of two populations. When more than two means were compared, a repeated-measures ANOVA was used with two grouping factors(location and level of exercise). Post-hoc modified t-tests with Bonferroni corrections were used to test for differences between the various factors. Two means were considered to be significantly different when P<0.05 and are quoted at the level at which they were found to be significant. All mean values are given ± S.D.

The mean values of fH, mass-specific V̇_O2(sV̇_O2) and mass-specific oxygen pulse (sO₂ pulse, see equation 1) for animals during the pre-exercise period and at maximum levels of exercise at the two different body temperatures (T_b) are given in Table 1. In all animals, both fH and sV̇_O2 exhibit significant increases over their pre-exercise values at maximum levels of exercise. For the SC animals at 27 °C,sV̇_O2 increased 5.2-fold, whereas there was only a 1.9-fold increase in fH. This means that there was a 2.8-fold increase in sO₂ pulse. The comparable factorial increases at 36 °C were:sV̇_O2, 4.9-fold; fH, 1.7-fold and sO₂ pulse, 2.9-fold. Similar factorial increases were seen in the SF animals at 36 °C.

In SC animals, pre-exercise fH was 94 % higher at a T_b of 36 °C compared with that at 27 °C, which is equivalent to a Q₁₀ of 2.1. The comparable pre-exercise values for sV̇_O2 were 55 %higher, equivalent to a Q₁₀ of 1.6. The maximum fH value recorded during exercise was 75 % higher at a T_b of 36°C compared with that at 27 °C, which yields a Q₁₀ of 1.9. The comparable value for maximum sV̇_O2 during exercise was 45 % higher, equivalent to a Q₁₀ of 1.5. Mass-specific oxygen pulses during the pre-exercise period were not significantly different at 27 °C and 36 °C (P=0.11) whereas sO₂ pulse during maximum exercise at 27 °C was significantly (21 %) greater than that at 36 °C (P=0.02). There was no significant difference between the values obtained from the SC and SF animals at 36 °C during the pre-exercise period and at maximum exercise for fH,sV̇_O2 and sO₂ pulse.

In order to be able to use fH for the estimation of V̇_O2 in the field,it is important that the calibrations are performed under conditions that are representative of those in the natural environment. Fig. 1 shows mean daily variation in T_b from animals in the field during two consecutive summers and the two values of temperature that we used in the present experiments were chosen on the basis of these data. Iguanas, like many other species of lizards, often use relatively short bursts of locomotion. For example, most (>95 %) marine iguanas forage in intertidal areas and make repeated short forays during which they run to particular foraging sites, take a few bites of algae, and run back to safe places to escape huge waves(Wikelski and Trillmich,1994). Similarly, iguanas walk in short bouts from coastal resting areas towards foraging sites. Each walking bout consists of approximately 20-40 fast steps, with interspersed resting pauses (seconds to minutes; Wikelski and Hau, 1995). Likewise, iguana males engage in relatively fast head-bob walking bouts during territorial activities, again with intermittent resting phases. During the mating season, females are constantly harassed by satellite males and have to run away and struggle away from males several times every hour (Wikelski et al., 1996, 2001; Wikelski and Bäurle,1996). Nesting females engage in bouts of intense digging activities lasting several minutes at a time. Even during diving bouts, marine iguanas are only active for a short time while grazing under water for an average of 2-5 min. Animals then resurface and remain floating largely motionless before diving again or swimming back to shore(Wikelski and Trillmich, 1994; Drent et al., 1999). Thus, the inclusion of data during the recovery period after activity is important if the energy cost of a particular behaviour is to be determined in the field(Scholnick and Gleeson,2000).

Perhaps because of their propensity for short bursts of locomotion, we did not find it easy to persuade all the animals to walk/run at speeds below the maximum speed they could maintain for a few minutes. Nonetheless, we were successful in achieving this with three animals at both temperatures and there was no significant difference between the slopes of the regression lines of data obtained from animals during the pre-exercise period and when walking/running at different speeds, and data obtained during the recovery period (P=0.67 at 27 °C and 0.07 at 36 °C, Fig. 4). Thus our experimental procedures simulated as closely as possible what is known for marine iguanas exercising in the wild and we are confident that the regression lines given in Fig. 2 represent both exercise at different levels and recovery from maximum exercise.

The sV̇_O2 data that we obtained are similar to those obtained by other workers studying the Galapagos marine iguana (Bennett et al.,1975, who electrically stimulated the animals in order to obtain activity; Bartholomew and Vleck,1979; Gleeson, 1979, 1980), with the exception of pre-exercise sV̇_O2at 27 °C, where our mean value is approximately twofold greater than the`resting' values obtained by the above workers. On the other hand, our mean value for pre-exercise fH at 27 °C is within the range given by Bartholomew and Lasiewski(1965), but our mean value for pre-exercise fH at 36 °C is approximately 75 % of that reported by the latter authors while the iguanas were being heated and cooled. As far as we can determine, there are no values in the literature for fH of marine iguanas during exercise.

It would appear from the present study that it should be possible to use fH as an indicator of sV̇_O2 for iguanas in the field, as between 86 and 91 % of the variation in the latter could be explained by the fitted regressions from the calibration experiments. The utility of the fH method for the estimation of sV̇_O2 in the marine iguana was further supported by the fact that the mean values of sV̇_O2 from four animals from a population different from that involved in producing the calibration equations were within the 95 % prediction intervals of the regression. However, the effect of temperature is to vary the intercept of the relationship between the two variables, rather than to extend a single regression line (see Fig. 2).

An interesting aspect of the present data is the fact that the increase in sV̇_O2 in response to an increase in T_b is the result of an increase in fH, with no significant change in sO₂ pulse. However,during exercise at both temperatures, the increases in fH are insufficient to provide all of the additional O₂ required and there are significant increases in the sO₂ pulses. Consequently, the situation arises (as illustrated in Fig. 2) whereby an fH value of around 60 beats min^-1 is related to an sV̇_O2 value of approximately 0.4 ml g^-1 h^-1 at 27°C, when the animal is exercising maximally, and to an sV̇_O2 value of approximately 0.1 ml g^-1 h^-1 at 36°C during the pre-exercise period. This means, of course, that the sO₂ pulse is fourfold greater during the former than during the latter. It is apparent from equations 1-4 and 6-9 and Fig. 9 of Bennett(1972) that a similar phenomenon occurs in S. hispidus and in Varanus gouldii,when activity is the result of electrical stimulation, although in the latter species, the values of fH during `exercise' at 27°C and while at`rest' at 36°C do not actually overlap. On the basis of a study on Iguana iguana and Varanus exanthematicus at 35°C, it would seem that the major contribution to the increase in sO₂ pulse during exercise is a twofold increase in Ca_O2—Cv̇_O2(Gleeson et al., 1980).

The authors wish to thank Dr Todd Gleeson for his extremely helpful comments on the manuscript, Dr Jon Green for his assistance with the data analysis and to Qantas Airlines for their logistical assistance. We thank L. B. Marty Martin, Tove Petterson, Lynn Micheletti and the Quest team for their assistance in the field. TAME, the Charles Darwin Research Station and the Parque Nacional Galapagos supported this study. Funding was provided by Princeton University. This is contribution number 492 to the Charles-Darwin Foundation.