Vol. 276, Issue 3, H844-H857, March 1999
Heart rate and behavior of fur seals: implications for
measurement of field energetics
I. L.
Boyd1,
R. M.
Bevan2,
A. J.
Woakes2, and
P. J.
Butler2
1 British Antarctic Survey,
Cambridge CB3 OET; and 2 School of
Biological Science, University of Birmingham, Edgbaston, Birmingham
B15 2TT, United Kingdom
 |
ABSTRACT |
Archival data
loggers were used to collect information about depth, swimming speed,
and heart rate in 23 free-ranging antarctic fur seals. Deployments
averaged 9.6 ± 5.6 days (SD) and totaled 191 days of recording.
Heart rate averaged 108.7 ± 17.7 beats/min (SD) but varied from 83 to 145 beats/min among animals. Morphometrics explained most variations
in heart rate among animals. These interacted with diving activity and
swimming speed to produce a complex relationship between heart rate and
activity patterns. Heart rate was also correlated with behavior over
time lags of several hours. There was significant
(P < 0.05) variation among animals
in the degree of diving bradycardia. On average, heart rate declined
from 100-130 beats/min before the dive to 70-100 beats/min
during submersion. On the basis of the relationship between heart rate
and rate of oxygen consumption, the overall metabolic rate was 5.46 ± 1.61 W/kg (SD). Energy expenditure appears to be allocated to
different activities within the metabolic scope of individual animals.
This highlights the possibility that some activities can be mutually exclusive of one another.
diving; Antarctica; metabolic rate; respirometry; morphometrics
| |
INTRODUCTION |
THE RATE OF ENERGY EXPENDITURE of
free-ranging animals is central to our understanding of many aspects of
their behavior. This is because the profitability of certain patterns
of behavior, especially during foraging, is determined by a balance
between the costs, in terms of the energy expended, and the benefits, in terms of the energy gained. Although the exact pattern of behavior can probably also be modified by other factors, such as predation risk
(38), energy expenditure is still one of the central variables involved
in understanding the relative importance of different behavioral patterns.
Despite this, it has proved to be very difficult to measure energy
expenditures associated with different behaviors. The double-labeled water (DLW) method has been used widely to measure average energy expenditure across periods of several days (see, e.g., Refs. 39 and
47), but this is usually too long to permit estimates of the energetic
costs of individual behaviors to be measured. Only by examining
behavioral time budgets, together with overall metabolic rates from
DLW, is it occasionally possible to estimate energy costs associated
with a specific behavior (22, 51, 52).
A method with a finer time resolution is required, and this may be
achieved by using variations in heart rate
(fH) as an indicator of
metabolic rate (16, 21, 37, 41, 42). This method is based on the Fick
equation (16) and works on the principle that the volume of oxygen
being delivered to and extracted by the tissues is proportional to
fH. It assumes that heart stroke volume and the oxygen extracted by the tissues vary systematically with
fH. Several studies have
successfully tested this approach and have shown monotonic
relationships between fH and
simultaneous measurements of metabolic rate during exercise (6, 8, 9, 20, 35, 40, 54). However, there have been no studies to examine the
relationship resulting from increased metabolic rate caused by, for
example, specific dynamic action or under conditions of thermal stress,
although there is no a priori reason to assume that the relationship
should not also hold in these circumstances.
Measurement of fH in free-ranging
animals has been made possible through the development of digital
electronic recorders (see, e.g., Ref. 57). These types of instruments
have led to an unprecedented increase in our ability to gather
multichannel data concerning physiology and behavior in free-ranging
animals, but there is now a requirement to develop equally robust
approaches to the analysis of these multidimensional data.
Among pinnipeds and other divers
fH varies in relation to diving
behavior (18), but few studies have examined the
fH response of completely
unrestrained pinnipeds foraging in their chosen natural habitat (1, 34,
49). Bradycardia during diving (43) may not be caused by changes in
overall metabolic rate, although it is likely to signal changes in the
instantaneous rate at which oxygen is delivered to the majority of
tissues. Therefore, to use fH to
examine metabolic rate during diving bouts, it is essential to average
over periods that are greater than one diving cycle (18). This is
because bradycardia during dives is evened out through time by a
compensatory tachycardia when the animal is at the surface between
dives (28). Moreover, fH will vary among animals, depending on their characteristics (48, 55), and it will
also be influenced by their level of activity. For example, both rate
of oxygen consumption (
O2)
and fH in sea lions vary with
swimming speed (15, 20, 55).
Measuring metabolic rate associated with particular activities in
free-ranging animals is complicated by the problem that the total
metabolic rate of an animal may be accounted for by the sum of many
different activities. Not all of these activities can be measured
simultaneously, and each will account for different amounts of
variability in metabolic rate. The effects of some activities on
metabolic rate, such as feeding, may also have a time lag associated
with them that may itself be variable. As with most other higher
vertebrates examined to date (31), animals like the antarctic fur seal
will have an upper limit to their sustained metabolic rate. This may
result in different activities, such as sustained swimming at high
speed or digestion, being mutually exclusive in some circumstances.
Overall, we would not expect a simple bivariate relationship between
fH and observed behavioral variables.
In the present study, we examined the behavior and
fH of lactating antarctic fur
seals. We used multivariate statistics to partition the variance in
fH between the different
behavioral variables while acknowledging that some other variables that
may have affected metabolism and
fH could not be measured. The
foraging behavior of female antarctic fur seals during lactation is
relatively well understood (11, 12), and several studies have also
examined the energetics of foraging using stable isotopes (3, 23, 24).
The importance of this species as a component within the Southern Ocean
ecosystem, especially at South Georgia, also means that refined
measurements of variability in energy expenditure could lead to better
estimates of the spatial and temporal impact of fur seals on their prey
and of the potential interaction between fur seals and commercial
fisheries. The objective was to examine 1) the importance of the time scale
of measurement on the relationship between
fH variation and behavior,
2) the variation in
fH between animals and in relation
to behavior, and 3)
fH as an indicator of field
metabolic rate.
| |
MATERIALS AND METHODS |
Measurement of fH and behavior.
This study was conducted at Bird Island, South Georgia, during the
austral summers of 1991-92 and 1992-93. Lactating females were captured from beaches, together with their pups, using the methods
of Gentry and Holt (30). Most females were anesthetized with tiletamine
hydrochloride mixed with zolazepam (Zoletil; see Ref. 13) and
intubated. An fH data logger (57),
housed within a single machined Perspex casing filled with silicone
fluid to resist pressure during diving and coated in wax and silicone
rubber, was implanted subcutaneously. It was placed dorsally and
between the scapulae through an incision 3 cm in length after
sterilization of the implantation location and administration of local
anesthesia (1% lignocaine). Electrodes, which were composed of flat
stainless steel plates (diameter 10 mm), were placed against the dorsal musculature. Each logger contained a small short-range (1-5 m) radio transmitter that pulsed whenever a heartbeat was detected. Detection of a heartbeat pulse by the instrument was checked using a
radio receiver simultaneously with independent examination of the
pulse, which can be observed visually on the thoracic wall. The
incision was closed by suturing with braided nylon and catgut, and the
animal was treated with 1 mg/kg long-acting oxytetracycline antibiotic
(Terramycin, Pfizer). Animals were kept isolated and monitored during
recovery from anesthesia, after which they were reunited with their
pups and returned to the same location on the beach from which they had
been taken. At the end of each deployment, during which the animal had
made at least one foraging trip, the fH logger was removed using the
same surgical methods as at implantation.
Monitoring of the heartbeat using the radio signal was carried out
during the period of deployment and recovery to verify feasible
fH variation in relation to bursts
of activity. Output from this design of data logger has been verified
previously against electrocardiograms (ECG) in both sea lions and
penguins (4, 20). Although concurrent ECG and logger monitoring has
only been carried out for penguins during exercise (4), Butler et al.
(20) used the radio transmitter within these loggers to verify the
plausibility of fH being recorded
in swimming sea lions. Moreover, all the
fH records were inspected in 6-h
blocks to search for regions in which there was an unusual distribution
of fH, indicated by implausibly
low or high records. In the cases where such records were found, the
record was rejected and was not used in the current analysis.
During 1992, some of the females were held within a restraining board
(30) and were not anesthetized. The radio transmitter-logger unit was
mounted externally after being embedded in epoxy resin (RS Components,
Corby, UK). The electrodes, in this case, were composed of two sterile
stainless steel needles. They were placed though a pinch of the skin,
after administration of local anesthesia, along the dorsal midline at
least 17-26 cm apart. The electrodes were placed approximately at
the level of the heart and held in position with a single suture. Epoxy
glue covered the electrodes and was also used to attach the leads and
the body of the logger to the fur of the animal. The detection of
heartbeat was checked in the same way as for animals with implanted
loggers. These females were also given long-acting oxytetracycline
antibiotic and released immediately back at the location from which
they had been taken.
A total of 9 animals were equipped with
fH loggers in 1991 and 14 in 1992 (Table 1). The total number of heartbeats
was counted and logged at intervals of 15 (in 1992) or 30 (in 1991) s,
and this allowed measurements to be recorded for up to 18-30 days. In addition, 11 animals also carried time-depth recorders (TDR; Mark 4, 5, and 6, Wildlife Computers). These were deployed as described by Boyd
et al. (11), and they measured depth, using a calibrated pressure
transducer, at 5- or 10-s intervals throughout the period of
deployment. The TDR had a saltwater switch that allowed time spent in
and out of the water to be recorded. Six of these instruments also had
the capability of measuring swimming speed using a turbine (stall speed
0.1-0.3 m/s). Speed was calculated on the basis of the number of
revolutions of the turbine measured over consecutive 15-s intervals,
and we used the calibration curve given for these instruments by Boyd
et al. (14). This was checked against the speed of ascent and descent
during dives when seals were diving vertically. In such circumstances,
vertical speed divided by measured speed should equal unity.
Examination of the frequency distributions of this index should show
large numbers of records with an index less than unity (caused by
animals diving at angles of <90° from the horizontal), and there
should be a rapid decline in the number of records at the point of
equality between the vertical speed and the measured speed. If this
transition point departs from unity, then the calibration is incorrect.
In the animals examined in this study, this transition showed that measurements of speed were within ±10% of the measurements based on the calibrated pressure transducer.
Both the fH loggers and the TDR
had accurate clocks (precision better than 1 s in 1 wk), and
instruments deployed simultaneously on individual animals had their
clocks synchronized to the nearest second. After recovery of the
instruments, data from each were placed on a combined time base and
analyzed jointly. Records from all animals were combined within a
single data set using the SAS statistical package (version 6.11, SAS
Institute) on a UNIX work station. Submergence was defined as any
excursion >1 m below the surface, and the predive
fH was the
fH sampled immediately before an
animal was recorded as submerged.
Morphometrics.
Mass was measured to an accuracy of 0.5 kg using a calibrated spring
balance at the beginning and the end of each experiment. However, these
mass measurements provided only a general indication of the mass of
individual animals because, in all cases, these seals were feeding
pups, and the absolute mass measured at the beginning and end of each
experiment will depend on when animals were captured within the
lactation cycle. In addition, all animals that were included in the
group with swimming speed recorders were measured for their
straight-line nose-to-tail length, pectoral girth, maximum foreflipper
span (measured from the tip of one flipper to the tip of the opposite
flipper with the flippers spread perpendicular to the midline), and the
length of the leading edge of the left foreflipper in a straight line
from the proximal insertion of the flipper to the tip. All these
measurements were made to an accuracy of 0.005 m.
Statistics.
Multivariate statistical methods available within SAS were used to
examine the possible factors affecting
fH. An important point to note
about the use of the multivariate statistical procedures in this study,
particularly multiple regression, is that they are used simply as
empirical descriptions of the data. In general, the data do not conform
to the assumptions of normality or independence (lack of serial
correlation) that are required for rigorous hypothesis testing using
multiple linear regression. In the few cases where probability levels
are quoted from multivariate procedures, they are intended to be used
as a guide to the relative importance of different variables rather
than an absolute measure of their importance. Where appropriate, values
are means ± SE unless otherwise stated.
Depth, speed, and fH records were
transformed from a time series to a frequency series with a
fast-Fourier transformation without filtering, using functions supplied
by Vetterling et al. (50). This allowed the principal frequencies
within the time series of diving and swimming behavior, together with
fH, to be examined in relation to
one another and, most importantly, allowed us to determine the dominant
time scales over which variability in
fH occurred.
Calibration of fH with metabolic rate.
In a previous study (20), the relationship between
fH and
O2 was determined in
California sea lions (Zalophus
californianus) of similar body size to the female
antarctic fur seals examined in this study. The relationship was found
to be approximately linear in sea lions. Because it was not possible to
carry out a complete calibration of
fH with
O2 in antarctic fur seals swimming under controlled conditions, we assumed that a similar relationship was present in the antarctic fur seals. We confirmed this
by examining the relationship during three types of activity (resting
on land, resting in water, and active in water) that represented the
full range of potential metabolic rates.
The O2 of resting fur seals
when on land was measured within a respirometer (2 m × 1 m × 1 m), whereas when they were in the water
O2 was measured in a static
water channel (1.22 m × 1.22 m × 14.63 m) that was covered
over its whole length except at each end, where the animals could
surface to breathe within respirometers. The respirometry system and
calculation of O2 are
described in detail in Bevan et al. (9). A gate could be lowered to
confine the experimental animal to one of the respirometers to allow
measurements to be made while animals were resting in water. Air
temperature during experiments was 7.7 ± 1.1°C, and the water
temperature was 6.6 ± 0.6°C. Data were only used after the seal
had acclimatized to the surroundings and when
O2 was steady (±5%)
for at least 5 min. Measurements were made during active swimming when
animals swam directly between the respirometers at each end of the
channel. If a seal attempted to surface between the respirometers these
data were not used.
O2 and
fH were measured simultaneously at
each of the activity levels in seven fur seals. Error associated with
the estimates of oxygen uptake from
fH was estimated as described by
Bevan et al. (8).
Calculation of energy expenditures.
O2, based on respirometry
or estimated from fH, can be
translated into a rate of energy expenditure after an assumption is
made about the metabolic substrate being used. Thus, when animals are
relying mainly on metabolizing fats, which is likely when they are
fasting ashore, they will produce 19.67 J/ml
O2 (46). However, the diet of
antarctic fur seals is mainly composed of antarctic krill
(Euphausia superba) (44, 45) and it
is likely that, when at sea, substrates used for metabolism will
reflect a high-protein diet, because most fats in the diet are likely to be stored for subsequent use in the formation of milk (2). Thus,
while the animals were at sea, we assumed that 18.41 J/ml O2 were produced from the
metabolism of protein substrate (46).
| |
RESULTS |
Short-term variation in fH during diving.
A representative section of the record of simultaneous measurements of
swimming speed, dive depth, and fH
from one animal is illustrated in Fig. 1.
This shows the extreme short-term variability in
fH, in this case from a minimum of
<80 beats/min at the bottom of a dive to a maximum of >160
beats/min at the surface when swim speed dropped to zero. Although this
might have resulted in a bimodal distribution of
fH during diving (one mode for
fH during diving and the other for
fH at the surface), such a pattern
was not generally apparent from frequency distributions of
fH among animals, even when only
the fH from seals when they were
at sea was considered (Fig. 2). In only two
cases (nos. w1571 and
w3835, Fig. 2) was there obvious
bimodality in fH. For other
animals, mean fH varied
sufficiently during the period of measurement that biomodalities caused
by bradycardia during diving and prediving tachycardia were obscured by
variability in fH over a larger
temporal scale than that associated with individual dives or groups of dives.
View larger version (34K):
[in this window]
[in a new window]
|
Fig. 1.
Representative section of heart rate, depth, and swimming speed record
from a female antarctic fur seal
(w1765) during a bout of diving
lasting ~60 min. This illustrates heart rate and swim speed variation
in relation to diving.
|
|
View larger version (25K):
[in this window]
[in a new window]
View larger version (24K):
[in this window]
[in a new window]
|
Fig. 2.
Frequency distributions of heart rate for all female antarctic fur
seals examined in the study during the 1991-92
(A) and 1992-93
(B) austral summers. Filled bars,
frequency based on all data points measured for these animals; open
bars, frequency for heart rate when seals were at sea (only shown for
seals in which diving records were available). Total sample size was
1.2 × 106. Sample sizes for
each animal are given in Table 1.
|
|
Bradycardia during diving was clearly apparent (Fig.
3). The relatively long duration of the
interval at which fH was sampled meant that it was not possible to examine changes in
fH within dives for those lasting
<60 s. fH declined significantly
(P < 0.05) to 70-85% of
predive levels after the dive began. During shorter dives (60-105
s), fH then began to increase
during the second half of the dive and returned to close to predive
levels when the fur seals returned to the surface (Fig. 3).
For dives of most durations, fH on
surfacing returned to less than the predive levels, implying that
fH increased before the next dive.
During longer dives (>105 s), after the initial decline,
fH remained at 70-90% of the
predive fH for an extended period
that increased in duration in proportion to dive duration (Fig. 3).
This appears to represent the progressively longer duration of the
lowest values in the longer-duration dives. From 105 s, the increase in
fH before surfacing became
progressively less until, at 180 s, there was no increase. However, the
pattern of increase for dives of 165 s was anomalous in this respect
(Fig. 3). The lack of increase in
fH before surfacing was not
indicative of shallower diving, because the longest dives were also the
deepest dives.
View larger version (29K):
[in this window]
[in a new window]
|
Fig. 3.
Heart rate in relation to stage of dive for dives of different
durations. Values are means for each animal expressed as an overall
mean. Sample sizes (n) are shown for
total number of dives with number of fur seals in parentheses. These
relationships were for interval between submergence, measured from
time-depth recorder, and return to surface. bpm, Beats per minute.
|
|
The degree to which fH declined,
compared with the predive fH,
during diving varied among animals (Table
2). For example, dives lasting 60 s, in
which there was an average reduction to ~0.7 of the predive
fH (Fig. 3), showed that
the maximum reduction in individual animals varied from 0.45 to 0.89 of
the predive fH. This level
of variation was highly significant (ANOVA,
F10,2795 = 141;
P < 0.001).
Long-term variation in fH.
In addition to short-term variation in
fH during diving,
fH is likely to vary between
animals and with swimming speed and feeding activity and even with time
of day. The overall variance structure of the
fH, depth, and swimming speed is
illustrated for the same animal as in Fig. 1 using spectral analysis
(Fig. 4). This shows a peak in spectral
power for all three variables in the 0.003- to 0.008-Hz band (125- to
333-s intervals) and also at <0.002 Hz (>500-s intervals). Because
the peak at 0.003-0.008 Hz also corresponds to the peaks in
spectral power for depth and speed, the peak in spectral power for
fH across this bandwidth is likely
to correspond to variation induced by behavior associated with diving.
Because of the averaging of fH
over 15- to 30-s recording periods, it was not possible to examine
higher-frequency components of fH
variation potentially caused by, for example, sinus arrhythmia.
View larger version (17K):
[in this window]
[in a new window]
|
Fig. 4.
Spectral density expressed as standardized power for depth
(A), speed
(B), and heart rate
(C) in a single antarctic fur seal
(W1765) across the whole of a
foraging trip at sea lasting 5.5 days.
x-Axis has been scaled to illustrate
frequencies associated with diving.
|
|
The variance structure of the whole data set from animals with a
complete set of measurements involving
fH, swimming speed, dive depth,
and morphometrics (n = 6, Table 1) was investigated by averaging the time-dependent components of
the data set (fH, swimming speed,
and depth) over successively longer intervals. These were then examined
in a multiple-regression model using a wide range of independent
variables (Table 3) to predict
fH. The proportion of variation
explained by the multiple-regression model increased in relation to the
time interval over which measurements were averaged (Fig.
5). However, the explained variation
increased rapidly for intervals up to 300-600 s, and thereafter it
increased at a substantially slower rate. This suggested, together with the spectra illustrated in Fig. 4, that two different processes producing variability in fH are
effective over different temporal scales. The small-scale process,
which can probably be attributed to variations in
fH associated with diving (Figs.
1, 3, and 4) and which would contain most diving cycles in this
species, was included when averaging took place over a duration of 300 s (Fig. 5). When fH was averaged
over 300-s periods the multiple-regression model explained 52% of the
variation. The larger-scale process, involving time intervals >300 s,
was represented by a more gradual increase in the proportion of
variation explained by the model with increasing time interval (Fig. 5)
as well as the lower frequencies illustrated in Fig. 4. In this section
we examine variation in fH at
temporal scales >300 s, which will exclude the short-term variation
in fH caused by diving activity.
View larger version (16K):
[in this window]
[in a new window]
|
Fig. 5.
Proportion of total variation in heart rate explained by a
multiple-regression model of heart rate in relation to
morphometric and behavioral variables (Table 3) and in relation to
interval over which averaging of behavioral variables (depth,
swimming speed) took place.
|
|
In the multiple-regression model involving averaging over 300 s, most
variance was explained by mass, length, girth, and time of day (Table
3). When the individual identity of each fur seal in the analysis was
introduced as a factor on its own, it made a significant contribution
to the fit of the model. However, when individual identity was
introduced together with the morphometric variables (mass, length,
girth, flipper span, left foreflipper length) it was no longer
significant, showing that the morphometric variables were an adequate
description of difference in fH
between animals. fH increased
significantly with swimming speed (expressed as a nonlinear variable),
mass, length, and flipper length and declined significantly with girth
(Table 3). Overall, swimming speed had an important influence on
fH, but this was mainly related to
its interaction with diving behavior represented in the model by mean
depth. Although depth alone had no significant effect on
fH, the effect of depth became
significant when it was combined with speed. When speed was entered as
a simple linear term it did not explain any significant amount of
variation in fH, but when speed
was entered as quadratic and cubic terms there were significant effects
of speed in both cases. This suggests that there were important
nonlinearities in the relationship between swimming speed and
fH (Tables 3 and
4).
There was also an effect of time of day on
fH (Tables 3 and 4), suggesting
that there was underlying diel variation in
fH, even after accounting for diel
variations caused by changes in diving behavior. We also examined the
effect of the proportion of time spent submerged and the dive frequency
during each 300-s interval, but these variables were not important
determinants of fH as judged by
the nominal probability level in the multiple-regression model.
Similarly, the use of mass0.75 and
mass2 in the model were not an
improvement on the use of mass1.
The complexity of the relationship between swimming speed and
fH is illustrated in Fig.
6, which shows
fH in relation to swimming speed
when at the surface for each animal in which speed, depth, and
fH were measured simultaneously.
There was no consistent pattern between animals in
fH variation with speed, which may
indicate the effects of other variables, in addition to swimming speed, on fH. The
fH associated with different
activities can be estimated from the linear regression equation defined
by the parameters in Table 3. An example of the predicted relationship
based on this regression, and the morphometrics of an average animal,
is given in Table 4. Like the empirical relationships in Fig. 6, this
shows that, on average, there was little change in
fH with swimming speed, but it
also suggests that fH was likely
to increase at the very highest swim speeds.
View larger version (35K):
[in this window]
[in a new window]
|
Fig. 6.
Variation in measured mean heart rate (± SE), averaged over 300-s
intervals, in relation to measured swimming speed in female antarctic
fur seals while on surface (depth <1 m) at sea. Also shown is
frequency distribution of swimming speeds measured. Sample sizes are
given in Table 4.
|
|
fH at sea and ashore.
The time spent in and out of the water was recorded for animals that
carried a TDR. There were significant differences in the
fH within animals between the two
environments, but there was no consistent pattern among animals (Table
5). In some cases (n = 3) individual animals had greater
mean fH when they were ashore than
when they were at sea, whereas in others
(n = 8) the opposite was the case.
Prediction of fH.
Morphometric measurements alone accounted for 46.7% of the variation
in fH within and among animals
when averaged over 300-s time intervals. For the nine animals from
which a full set of morphometric measurements were available (Table 1)
the prediction equation
(r2 = 0.47, P < 0.001) for
fH during any 300-s period
was
|
(1)
|
When depth was included as a quadratic term, the equation
was
|
(2)
|
with
r2 = 0.49 and
P < 0.001.
To test the ability of the morphometrics to predict
fH in the nine animals for which
full morphometric measurements were available, nine separate linear
regression equations were derived with one animal excluded from each
equation. The morphometrics of the excluded animal were then used to
predict, on the basis of the relationships for the other eight animals,
the fH of the excluded animal. The results of this are shown in Table 6.
When averaged across all nine animals in the sample, the mean predicted
fH was within 2% (mean observed = 106.1 beats/min; mean predicted = 107.7 beats/min) of the observed
fH.
The duration of sampling required to obtain a mean and standard
deviation of fH that was within
±5% of the values measured during the whole period of deployment
varied considerably among individual animals (Table 6). In some cases,
variability was apparently so large that the sampling duration used in
the experiment was probably insufficient to characterize fully the mean
fH and variability in the
fH. However, for the majority, a
deployment of an instrument for up to 7 days was sufficient to
characterize fH variation in
individual animals.
Time scales of variation in fH and
behavior.
One of the reasons why some of the variation in
fH remained unexplained in the
multiple-regression model is that specific behavior may affect future
fH. The gradually increasing
spectral power of fH and speed at
lower frequencies illustrated in Fig. 4 was probably caused by
autocorrelation, an effect that appeared to be greater in
fH and swimming speed than in
depth. This means that the behavior occurring during one 300-s time
period could influence fH in the
next time period and so on. Partial correlation was used to examine the
relationships between fH lagged
through time and both mean depth and speed during 300-s time periods. Partial correlation allowed the examination of direct effects of speed
or depth in one time period on fH
in a subsequent time period without the added effects of serial
correlation from intervening time periods. In general, there was no
consistent pattern of correlation between
fH and either current or future
mean depth or swimming speed during each 300-s interval, although there
were significant negative and positive partial correlations with both
variables (Fig. 7). Highest correlations
between fH and depth occurred over short time lags, but this was not as obvious for swimming speed (Fig.
7). These results suggest that, in general, the
fH occurring in any particular
time period may be influenced by behavior occurring many hours earlier.
View larger version (30K):
[in this window]
[in a new window]
|
Fig. 7.
Pearson partial correlation coefficients between mean heart rate with
mean depth and mean speed during each 300-s sampling interval with
effects of depth and speed on heart rate lagged through time. Each
animal is shown by a different line. Dotted lines parallel to
x-axis show ~95% confidence
intervals.
|
|
Relationship between fH and
O2.
The mean fH (±SE) of seals
that were resting in air, resting in water, and active in the water
were 77.7 ± 5.4, 110.1 ± 8.6, and 146.7 ± 7.8 beats/min,
respectively. The measured rates of O2 over these periods were
7.27 ± 0.37, 19.70 ± 1.88, 28.03 ± 1.88 ml
O2 · min
1 · kg1,
respectively, which were equal to 2.38 ± 0.12, 6.46 ± 0.62, and
9.19 ± 0.62 W/kg. The relationship between
fH and
O2 was examined for each
individual animal in the calibration experiment using curvilinear
models fitted by least-squares regression (Table 7). There was no significant difference
between the lines for each individual animal (ANCOVA,
F12,7 = 1.557, P > 0.05). The common line, based on
the combined sums of squares of the regression lines for each
individual animal, was therefore used
|
(3)
|
Estimated energy expenditure.
On the basis of the calibrations and the assumptions about the
metabolic substrate being used, the equations for the relationship of
fH to metabolic rate
(M), expressed as watts per kilogram, for periods spent
ashore and at sea are
|
(4)
|
and
|
(5)
|
respectively.
The similarity between the O2
during fat (19.67 J/ml O2) and
protein (18.41 J/ml O2)
metabolism means that the result is relatively insensitive to the
assumptions about which substrate was being used. In cases where there
was no discrimination between periods spent at sea or ashore a mixed
metabolic strategy was assumed, involving roughly equal proportions of
protein and fat. This gave the equation
|
(6)
|
With
Eqs. 4 and 5,
fH produced estimates of energy
expenditure that were not significantly different when the seals were ashore or at sea (Table 5; paired
t-test,
t = 0.671, P = 0.51). The estimated metabolic
rate in individual animals varied from 2.9 to 8.5 W/kg, and the mean
metabolic rate was 5.58 ± 1.64 W/kg (Table 6). The overall
metabolic rates of animals with and without TDR (Table 1) were 5.84 ± 0.50 and 5.11 ± 0.45 W/kg, respectively. These values were
not significantly different from one another (t = 1.11, df = 21, P >0.2).
Mean fH and metabolic rate were
compared during diving and swimming by classifying the 300-s intervals
when the fur seals were at sea into four different categories of
behavior: 1) swimming at the
surface, 2) stationary at the
surface, 3) submerged swimming, and
4) stationary while submerged.
Surface was defined as any 300-s interval in which the mean depth was
<2 m (i.e., the resolution of the instrument). Swimming was defined
as any 300-s interval in which the mean speed was >0.3 m/s (i.e.,
stall speed). This showed that there was no significant difference
between metabolic rate in relation to behavior among the six animals
for which fH, speed, and depth
measurements were available, even after accounting for differences
among individual animals (Table 8; ANOVA,
F3,15 = 1.13, P = 0.370). There was also no
significant difference between the metabolic rate at the surface or
when submerged in the 11 animals for which depth data were available
(ANOVA, F1,10 = 0.04, P = 0.843).
View this table:
[in this window]
[in a new window]
|
Table 8.
Heart rate and metabolic rate among 6 female antarctic fur seals in
relation to behavior recorded within each 5-min time period
|
|
| |
DISCUSSION |
Causes of variation in fH.
This study has shown that behavior and morphometrics have important
effects on fH in fur seals.
Despite caveats associated with hypothesis testing using the
multiple-regression approach (see MATERIALS AND
METHODS), this method of analysis allows empirical comparisons to be made between the magnitudes of the effects of different variables on fH without
the need to resort to experimental conditions in which each variable is
manipulated independently.
The regression model using morphometrics, depth, speed, and time of day
explained 52% of the total variance in
fH (Table 3). If
fH is to be used as a
deterministic measure of metabolic rate, then it is important to
understand the source of the remaining unexplained variation. Potential
causes include the possibility that relationships between variables and
fH were not linear, variables involving the immediate surroundings of individual animals could not be
measured, and the effects of feeding and digestion on
fH as well as psychological state
(48) also could not be included as variables in the multiple regression
because they could not be measured. The effects of nonlinearities in
the relationship between fH and
other variables was considered, and only in the case of swimming speed
was a nonlinear relationship significant. This is supported by
empirical evidence from experimental studies in other marine mammals
(25, 54). However, as illustrated by Fig. 7, one of the effects causing
variability in fH that may be most
difficult to track down involves the lagged effects of previous
activities. For example, diving could lead to feeding, which in turn
would lead to a heat increment associated with digestion, and this
apparent effect of diving on fH
could, therefore, take several hours to dissipate.
Additional reduction in unexplained variation in metabolic rate
measured using fH could be
achieved by including morphometrics within calibrations of
fH with
O2. In this study
fH explained 72% of variation in
O2, but this relationship is
likely to be influenced by other factors such as morphometrics, and
these should be included in future calibration studies to account for
as much of the variation in the fH
calibration relationship as possible.
Previous studies have examined variation in
fH in relation to specific
activities such as incubation in seabirds (7), flight (4, 7, 19),
swimming (25, 36, 40, 54, 55), and diving (4, 28, 43, 53, 56). In
free-ranging conditions, it is rare for a single activity to occur to
the exclusion of another. For example, diving is often combined with
swimming, and resting may be combined with digestion. In addition,
psychological state has been suggested as a factor affecting
fH (48). Therefore, rather than
distinguishing between different discrete activities and assuming that
there was sufficient information to define each activity, as has been
done in many other past studies of time-energy budgets or where
fH has been related to behavior
(4, 8, 11, 23, 40, 51, 52), in the present study we examined fH in a multivariate context that
made no assumptions about the accuracy with which behavior was being
classified. We considered this to be necessary in the present study
because time-depth-speed recorders provided only a partial view of the
true activities of the animals. However, to retain some consistency
with past studies, we classified behavior at sea into four groups in
relation to diving and swimming activity and found no significant
difference in the metabolic rate at the surface compared with diving
(Table 8). The only other classification made was between the time
spent ashore and the time spent at sea, when there was a clear
distinction between the circumstances in which the animals were
operating, and these periods also showed no significant difference in
fH. This approach to analysis of
these data can produce significant differences from conclusions based
on more traditional methods of behavioral classification, as suggested
by a preliminary analysis of these data (17).
By averaging over an appropriate time period (300 s in this case) we
also partitioned out variation in
fH caused by diving-surfacing activity. Fedak et al. (28) found a linear increase in oxygen consumption with fH in gray seals
(Halichoerus grypus) when averaged over whole dive cycles. The effect of averaging over periods of 300-s
duration in this study was to reduce the effects of the extreme
short-term variation in the rate of oxygen delivery to the tissues
during the dive-surface cycle (normally ~60-240 s; Ref. 12), and
this was illustrated by the rapid change in the slope of the plot
illustrated in Fig. 5 and the spectral power distribution in Fig.
4.
Because the sampling interval of the
fH recorder was 15 s for most of
those animals in which diving behavior was recorded, as pointed out in
previous studies (4, 34), it was not possible to be certain about the
patterns of variation in fH for
short dives (<60 s) or during the first and last 15 s of a dive.
However, fH profiles were similar
to those recorded in other species (1, 26, 27, 32, 49), especially in
terms of the rapid decline in fH
at the beginning of the dive. Despite this response to submergence, the
decline in fH during diving never
reached the levels observed in gray or elephant seals (1, 49) in which
diving fH declined to <10% of
surface fH during some of the
longest dives. This compares with
fH of ~70% (60-80
beats/min) of surface fH during
dives of >90-s duration in the present study. Therefore, the lowest
fH recorded in this study are
higher than those recorded for true seals, and this is probably not an
effect of the averaging interval used for
fH in this study.
California sea lions that had been trained to dive to a target also
showed a greater decline in fH
during diving (~25% of predive
fH) than observed in the present
study (43). However, Ponganis et al. (43) also observed that the diving
fH was higher in sea lions when
they were involved in voluntary submerged swimming rather than diving
to a target. Diving to a target involves these animals staying
submerged for an unpredictable duration, and the fH response can be interpreted as
the most conservative that it is possible for the animal to make and
the one that is most likely to extend potential dive duration. The
responses elicited during voluntary submerged swimming in California
sea lions were closer to those observed in wild antarctic fur seals. As
shown in the present study (Fig. 2), female antarctic fur seals were
capable of sustaining fH as low as
25% of predive fH (20-30
beats/min) over the 15-s measurement interval, but such a low
fH was rarely recorded.
There was a complex pattern of fH
variability over periods that exceeded the normal dive duration (Figs.
4 and 7). This included autocorrelation of
fH at time scales >10 min
(<0.001 Hz, Fig. 4), but there was also evidence that
fH was influenced by events occurring up to several hours beforehand. This could have been induced
partly by cyclical variation in behavior associated, for example, with
bouts of diving caused by periodic encounters with prey patches (10) or
caused by diel patterns of foraging (12). Conversely, the effects on
metabolic rate of specific activities, such as feeding, could take
several hours to dissipate and result in the observed lags between
swimming and diving activity with fH. Similarly, if metabolic
changes occur during diving (18), it may take several hours for an
individual animal to recover fully from a sustained bout of diving.
Excluding short-term variation in
fH caused by diving, the
morphometrics of individual animals had the greatest influence on
fH of all the variables examined.
Mass, length, and pectoral girth were particularly influential. On the
basis of morphometrics alone, it was possible to predict the mean
fH of a group of animals to <2%
of the true fH based on a multiple
regression of morphometric variables. This suggests that, for lactating
female fur seals at least, much information about
fH and, by implication, about energy consumption can be obtained from a simple set of morphological measurements. With the addition of information about diving depth, it
is possible to obtain a small amount of additional explanatory power
(Eqs. 1 and 2). Such a result is perhaps not
surprising, because metabolic rates are likely to vary in relation to
features, such as surface area for heat exchange and hydrodynamics, all of which are determined to a large degree by morphometrics (33). It may
be interesting to investigate this relationship with morphometrics during times when food availability is reduced and when, in theory at
least, animals may have to work harder to find food. However, it is
also possible that animals normally work close to their maximum (or
optimum) sustained metabolic scope under most circumstances and that,
as a result, fH may not vary under
different foraging conditions. This could occur because fur seals may
have to trade off metabolic expenditure on different activities to
retain a long-term energy balance and, therefore, remain within their
long-term sustained metabolic scope.
Estimating free-living energetics from
fH.
As with estimates of metabolic rate using DLW, large errors may exist
in the estimates of metabolic rate from
fH (8, 9, 15). These can
be overcome to a large extent, especially for fH (40), by pooling mean metabolic
rates among individual animals. The equation for the relationship
between oxygen consumption and fH
in data pooled by Butler et al. (17) across each measurement and
individual fur seal was weighted toward the measurement for individual
animals resting in air. Conversely, the present analysis takes into
account the differences between individual animals and the number of
points available for individual fur seals at each activity level. When
applied to the fH measurements of
free-ranging fur seals, an overall mean metabolic rate of 5.46 ± 0.34 W/kg was obtained for the complete set of fur seals (Table 6). The metabolic rates ashore and at sea for those animals carrying TDR were
5.65 ± 0.52 and 5.71 ± 0.55 W/kg, respectively, which
were not significantly different from one another (paired
t-test,
P > 0.9). They demonstrate that, at
sea, the metabolic rate was not significantly different from the
metabolic rates measured by direct respirometry for animals resting in
water (6.46 ± 0.62 W/kg; t = 1.41, df = 28, P > 0.1) and that the
metabolic rates of females when they were ashore with their pups was
double the resting metabolic rate measured using direct respirometry
(2.38 ± 0.12 W/kg; t = 4.91, df = 28, P > 0.001). The latter contrasts with results from DLW experiments in which the metabolic rate was
estimated to be 4.96 W/kg when the animals were ashore and 9.52 W/kg when they were at sea (24). Although the results from the present study are not significantly
(t = 0.21, df = 17, P > 0.5) different from the
estimates of metabolic rate based on DLW when the animals were ashore,
they are significantly (t = 3.92, df = 31, P < 0.001) less than
those for lactating females when at sea. Although Arnould et al. (3)
also used DLW, they showed metabolic rates that were similar to those
found in the present study using
fH but only for animals that spent
a large proportion of time diving. Otherwise, the estimates produced by Arnould et al. (3) were also greater than those observed in the present
study. Moreover, when examining metabolic rates in actively swimming
California sea lions, Boyd et al. (15) showed that DLW may overestimate
the metabolic rate by ~34%. Although the reasons for this
overestimate are not completely understood and could be associated with
the methodology of the experiment (47), in such circumstances the
estimate of Costa et al. (23) would be closer to 7.10 W/kg, which is
not significantly different from the estimate produced in the present
study (t = 1.1, df = 31, P > 0.1).
The similarity between the current
fH and past DLW estimates of the
metabolic rate when animals were ashore and the dissimilarity of these
estimates when animals were at sea suggests that one or both of the
methods has inherent inaccuracies. Experimental studies in pinnipeds
have shown that fH has a linear
relationship with oxygen consumption (20, 54) and that oxygen
consumption increases exponentially with swimming speed (25, 54, 55). If, during extreme conditions of exercise,
fH tends to approach a maximum,
then it is possible that fH could
underestimate the metabolism of free-ranging antarctic fur seals.
Alternatively, there are several potential inherent errors associated
with the DLW method (39, 47). The possibility that errors exist within the DLW method is illustrated by the observation that the sustained metabolic rate (measured using DLW) of fur seals foraging at sea over
several days was similar to the maximum rate of energy expenditure that
it was possible to elicit from sea lions swimming in a water channel
(15, 23). Either free-ranging fur seals are capable of increasing their
metabolic rate substantially higher than sea lions in the laboratory,
or the DLW method is overestimating their metabolic rate while at sea.
There was a significant relationship between
fH and time of day indicated by
the differences between fH
estimated from the multiple-regression model (Table 4). This may
reflect higher metabolic rates early in the day because female
antarctic fur seals feed at night (12), and they may therefore be
digesting food in the period represented by the group at 0600 h in
Table 4. Similar results, in terms of rapidly increasing energy
expenditure at high speeds, to those suggested by the
fH data in Table 4 have been
obtained under experimental conditions for both swimming birds (56) and
pinnipeds (25, 29, 49). On the basis of this evidence, it appears that
fur seals may maintain a reasonably constant rate of energy expenditure
but that the energy expenditure may be used for different purposes. For
example, heat production may be required for thermoregulation, but the
heat produced for this purpose may be derived from a number of
different sources, including locomotion (5, 33) when the animals are
swimming and/or diving. Only when swimming speed exceeds ~ 3 m/s (which is rare) does it appear that additional work has to be done
to maintain locomotion.
In conclusion, fH in free-ranging,
lactating antarctic fur seals exhibits a high degree of variability
both within and between animals. Nonetheless, estimates of field
metabolic rate from fH seem to be
reasonable when compared with those obtained from studies using the DLW
method. Morphometrics are particularly important for explaining
variability among animals. These interact with behavioral features to
produce a complex relationship between fH and activity patterns.
fH declined during submergence,
but there was large variation in the average magnitude of this decline in different animals. Although female antarctic fur seals are capable
of extremely low fH during diving,
as observed in other pinnipeds, these rarely occurred for longer than
15-30 s during free-ranging diving behavior. Estimates of
metabolic rate based on fH
suggested there were no differences in the costs of being ashore and
the costs of being at sea. They also suggest that the metabolic costs
of being at sea are lower than previously estimated. Although metabolic
rate changed little across the range of swimming speeds commonly used
by female antarctic fur seals, its relationship with speed was affected
by diving so that, in general, metabolic rate was reduced with
increasing average depth. We conclude that, on the basis of
fH as an indicator of metabolic
rate, antarctic fur seals are likely to partition total energy
expenditure between different activities and, because of limited
long-term metabolic scope, it may not be possible to support some
activities simultaneously.
| |
ACKNOWLEDGEMENTS |
The authors thank the staff at the British Antarctic Survey
Research Station on Bird Island, South Georgia, for assistance throughout this project. Prof. J. P. Croxall and Dr. D. McCafferty kindly provided comments on the manuscript.
| |
FOOTNOTES |
This work was supported by the Natural Environment Research Council
(Grant GR/3/7508).
The costs of publication of this
article were defrayed in part by the
payment of page charges. The article
must therefore be hereby marked
"advertisement"
in accordance with 18 U.S.C. §1734 solely to indicate this fact.
Address for reprint requests: I. L. Boyd, British Antarctic Survey,
Madingley Rd., Cambridge CB3 OET, UK.
Received 2 September 1998; accepted in final form 4 November 1998.
| |
REFERENCES |
1.
Andrews, R. D.,
D. R. Jones,
J. D. Williams,
P. H. Thorson,
G. W. Oliver,
D. P. Costa,
and
B. J. Le Boeuf.
Heart rates of northern elephant seals diving at sea and resting on the beach.
J. Exp. Biol.
200:
2083-2095,
1997[Abstract].
2.
Arnould, J. P. Y.,
and
I. L. Boyd.
Temporal patterns of milk production in Antarctic fur seals (Arctocephalus gazella).
J. Zool. Lond.
237:
1-12,
1995.
3.
Arnould, J. P. Y.,
I. L. Boyd,
and
J. R. Speakman.
The relationship between foraging behavior and energy expenditure in Antarctic fur seals.
J. Zool. Lond.
239:
769-782,
1996.
4.
Bevan, R. M.,
I. L. Boyd,
P. J. Butler,
K. Reid,
A. J. Woakes,
and
J. P. Croxall.
Heart rates and abdominal temperatures of free-ranging South Georgian shags, Phalacrocorax georgianus.
J. Exp. Biol.
200:
661-675,
1997[Abstract].
5.
Bevan, R. M.,
and
P. J. Butler.
The effects of temperature on the oxygen-consumption, heart-rate and deep body-temperature during diving in the tufted duck Aythya fuligula.
J. Exp. Biol.
163:
139-151,
1992[Abstract/Free Full Text].
6.
Bevan, R. M.,
P. J. Butler,
and
J. R. Speakman.
Daily energy-expenditure of tufted ducks
a comparison between indirect calorimetry, doubly labeled water and heart-rate.
Funct. Ecol.
9:
40-47,
1995.
7.
Bevan, R. M.,
P. J. Butler,
A. J. Woakes,
and
P. A. Prince.
The energy expenditure of free-ranging black-browed albatrosses.
Philos. Trans. R. Soc. Lond. B Biol. Sci.
350:
119-131,
1995.
8.
Bevan, R. M.,
A. J. Woakes,
P. J. Butler,
and
I. L. Boyd.
Use of heart rate to estimate energy expenditure of black-browed albatrosses, Diomedea melanophrys.
J. Exp. Biol.
193:
119-137,
1994[Abstract].
9.
Bevan, R. M.,
A. J. Woakes,
P. J. Butler,
and
J. P. Croxall.
Heart rate and oxygen consumption of exercising gentoo penguins.
Physiol. Zool.
68:
855-877,
1995.
10.
Boyd, I. L.
Temporal scales of foraging in a marine mammal.
Ecology
77:
426-434,
1996.
11.
Boyd, I. L.,
J. P. Y. Arnould,
T. Barton,
and
J. P. Croxall.
Foraging behavior of the Antarctic fur seal during periods of contrasting prey abundance.
J. Anim. Ecol.
63:
703-713,
1994.
12.
Boyd, I. L.,
and
J. P. Croxall.
Diving behavior of lactating Antarctic fur seals.
Can. J. Zool.
70:
919-928,
1992.
13.
Boyd, I. L.,
N. J. Lunn,
C. D. Duck,
and
T. Barton.
Response of Antarctic fur seals to immobilization with ketamine, a ketamine-diazepam or ketamine-xylazine mixture, and Zoletil.
Mar. Mamm. Sci.
6:
135-145,
1990.
14.
Boyd, I. L.,
K. Reid,
and
R. M. Bevan.
Swimming speed and allocation of time during the dive cycle in Antarctic fur seals.
Anim. Behav.
50:
769-784,
1995.
15.
Boyd, I. L.,
A. J. Woakes,
P. J. Butler,
R. W. Davis,
and
T. M. Williams.
Validation of heart rate and doubly labelled water as measures of metabolic rate during swimming in California sea lions.
Funct. Ecol.
9:
151-160,
1995.
16.
Butler, P. J.
To what extent can heart rate be used as an indicator of metabolic rate in free-living marine mammals?
Symp. Zool. Soc. Lond.
66:
317-332,
1993.
17.
Butler, P. J.,
R. M. Bevan,
A. J. Woakes,
J. P. Croxall,
and
I. L. Boyd.
The use of data loggers to determine the energetics and physiology of aquatic birds and mammals.
Braz. J. Med. Biol. Res.
28:
1307-1317,
1995[Medline].
18.
Butler, P. J.,
and
D. R. Jones.
Physiology of diving of birds and mammals.
Physiol. Rev.
77:
837-899,
1997[Abstract/Free Full Text].
19.
Butler, P. J., A. J., Woakes, and C. M. Bishop. Behaviour and physiology of Svalbard barnacle
geese, Branta leucopsis, during their
autumn migration. J. Avian Biol. In
press.
20.
Butler, P. J.,
A. J. Woakes,
I. L. Boyd,
and
S. Kanatous.
Relationship between heart rate and oxygen consumption during steady-state swimming in California sea lions.
J. Exp. Biol.
170:
35-42,
1992[Abstract/Free Full Text].
21.
Ceesay, S. M.,
A. M. Prentice,
K. C. Day,
P. R. Murgatroyd,
G. R. Goldberg,
W. Scott,
and
G. B. Spurr.
The use of heart rate monitoring in the estimation of energy expenditure: a validation study using indirect whole-body calorimetry.
Br. J. Nutr.
61:
175-186,
1989[Medline].
22.
Costa, D. P.
Method for studying the energetics of freely diving animals.
Can. J. Zool.
66:
45-52,
1988.
23.
Costa, D. P.,
J. P. Croxall,
and
C. D. Duck.
Foraging energetics of Antarctic fur seals in relation to changes in prey availability.
Ecology
70:
596-606,
1989.
24.
Costa, D. P.,
and
F. Trillmich.
Mass changes and metabolism during the perinatal fast: a comparison between Antarctic (Arctocephalus gazella) and Galapagos fur seals (Arctocephalus galapagoensis).
Physiol. Zool.
61:
160-169,
1988.
25.
Davis, R. W.,
T. M. Williams,
and
G. L. Kooyman.
Swimming metabolism of yearling and adult harbor seals, Phoca vitulina.
Physiol. Zool.
58:
590-596,
1985.
26.
Elsner, R.,
D. Wartzok,
N. B. Sonfrank,
and
B. P. Kelly.
Behavioral and physiological reactions of arctic seals during under-ice pilotage.
Can. J. Zool.
67:
2506-2513,
1989.
27.
Fedak, M. A.
Diving and exercise in seals: a benthic prespective.
In: Diving in Animals and Man, edited by A. Brubakk,
J. W. Kanwisher,
and G. Sundnes. Trondheim, Norway: Tapir, 1986, p. 11-32.
28.
Fedak, M. A.,
M. R. Pullen,
and
J. Kanwisher.
Circulatory responses of seals to periodic breathing: heart rate and breathing during exercise and diving in the laboratory and open sea.
Can. J. Zool.
66:
53-60,
1988.
29.
Feldkamp, S. D.
Swimming in the California sea lion: morphometrics, drag and energetics.
J. Exp. Biol.
131:
117-135,
1987[Abstract/Free Full Text].
30.
Gentry, R. L.,
and
J. R. Holt.
Equipment and Techniques for Handling Northern Fur Seals. Washington, DC: NOAA, 1982 (National Marine Fisheries Service SSRF-758).
31.
Hammond, K. A.,
and
J. Diamond.
Maximal sustained energy budgets in humans and animals.
Nature
386:
457-462,
1997[Medline].
32.
Hill, R. D.,
G. C. Liggins,
A. H. Schuette,
R. L. Elliott,
M. Guppy,
P. W. Hochachka,
J. Qvist,
and
K. J. Falke.
Heart rate and body temperature during free diving of Weddell seals.
Am. J. Physiol.
253 (Regulatory Integrative Comp. Physiol. 22):
R344-R351,
1987[Abstract/Free Full Text].
33.
Hind, A. T.,
and
W. S. C. Gurney.
The metabolic cost of swimming in marine homeotherms.
J. Exp. Biol.
200:
531-542,
1997[Abstract].
34.
Hindell, M. A.,
and
M.-A. Lea.
Heart rate, swimming speed, and estimated oxygen consumption of a free-ranging southern elephant seal.
Physiol. Zool.
71:
74-84,
1998[Medline].
35.
Kooyman, G. L.,
and
P. J. Ponganis.
Emperor penguin oxygen consumption, heart rate and plasma lactate levels during graded swimming exercise.
J. Exp. Biol.
195:
199-209,
1994[Abstract].
36.
Kooyman, G. L., and P. J. Ponganis. The
physiological basis of diving to depth: birds and mammals.
Annu. Rev. Physiol. 60: 19-32.
37.
Lund, G. F.,
and
G. E. Folk.
Simultaneous measurements of heart rate and oxygen consumption in black-tailed prairie dogs (Cynomys ludovicianus).
Comp. Biochem. Physiol. A Physiol.
55:
201-206,
1976.
38.
Mangel, M.,
and
C. W. Clark.
Dynamic Modeling in Behavioral Ecology. Princeton, NJ: Princeton Univ. Press, 1988.
39.
Nagy, K. A.
CO2 production in animals: analysis of potential errors in the doubly labeled water method.
Am. J. Physiol.
238 (Regulatory Integrative Comp. Physiol. 7):
R466-R473,
1980[Free Full Text].
40.
Nolet, B. A.,
P. J. Butler,
D. Masman,
and
A. J. Woakes.
Estimation of daily energy expenditure from heart rate and doubly labeled water in exercising geese.
Physiol. Zool.
65:
1188-1216,
1992.
41.
Owen, R. B.
Heart rate, a measure of metabolism in blue-winged teal.
Comp. Biochem. Physiol. A Physiol.
31:
431-436,
1969.
42.
Pauls, R. W.
Heart rate as an index of energy expenditure in red squirrels (Tamiasciurus hudsonicus).
Comp. Biochem. Physiol. A Physiol.
67:
409-418,
1980.
43.
Ponganis, P. J.,
G. L. Kooyman,
L. M. Winter,
and
L. N. Starke.
Heart rate and plasma lactate responses during submerged swimming and trained diving in California sea lions, Zalophus californianus.
J. Comp. Physiol. [B]
167:
9-16,
1997[Medline].
44.
Reid, K.
The diet of Antarctic fur seals (Arctocephalus gazella Peters 1875) during winter at South Georgia.
Antarct. Sci.
7:
241-249,
1995.
45.
Reid, K.,
and
J. P. Y. Arnould.
The diet of Antarctic fur seals Arctocephalus gazella during the breeding season at South Georgia.
Polar Biol.
16:
105-114,
1996.
46.
Schmidt-Nielsen, K.
Animal Physiology (5th ed.). Cambridge, UK: Cambridge Univ. Press, 1997.
47.
Speakman, J. R.
Doubly Labelled Water: Theory and Practice. London, UK: Chapman and Hall, 1997.
48.
Stephenson, R.,
P. J. Butler,
N. Dunstone,
and
A. J. Woakes.
Heart-rate and gas-exchange in freely diving American mink (Mustela vison).
J. Exp. Biol.
134:
435-442,
1988[Free Full Text].
49.
Thompson, D.,
and
M. A. Fedak.
Cardiac responses of grey seals during diving at sea.
J. Exp. Biol.
174:
139-164,
1993[Abstract].
50.
Vetterling, W. T.,
S. A. Teukolsky,
W. H. Press,
and
B. P. Flannery.
Numerical Recipes in C. Cambridge, UK: Cambridge Univ. Press, 1996.
51.
Weathers, W. W.,
W. A. Buttemer,
A. M. Hayworth,
and
K. A. Nagy.
An evaluation of time-budget estimates of daily energy expenditure in birds.
Auk
101:
459-472,
1984.
52.
Weathers, W. W.,
and
K. A. Nagy.
Simultaneous doubly labeled water (3HH18O) and time budget estimates of daily energy expenditure in Phainopepla nitens.
Auk
97:
861-867,
1980.
53.
Webb, P. M.,
R. D. Andrews,
D. P. Costa,
and
B. J. LeBoeuf.
Heart rate and oxygen consumption of northern elephant seals during diving in the laboratory.
Physiol. Zool.
71:
116-125,
1998[Medline].
54.
Williams, T. M.,
W. A. Friedl,
and
J. E. Haun.
The physiology of bottlenose dolphins (Tursiops truncatus): heart rate, metabolic rate and plasma lactate concentrations during exercise.
J. Exp. Biol.
179:
31-46,
1993[Abstract].
55.
Williams, T. M.,
G. L. Kooyman,
and
D. A. Croll.
The effect of submergence on heart rate and oxygen consumption of swimming seals and sea lions.
J. Comp. Physiol. [B]
160:
637-644,
1991[Medline].
56.
Woakes, A. J.,
and
P. J. Butler.
Swimming and diving in tufted ducks, Aythya fuligula, with particular reference to heart rate and gas exchange.
J. Exp. Biol.
107:
311-329,
1983[Abstract/Free Full Text].
57.
Woakes, A. J.,
P. J. Butler,
and
R. M. Bevan.
An implantable data logger system for heart rate and body temperature: its application to the estimations of field metabolic rates in antarctic predators.
Med. Biol. Eng. Comput.
33:
145-152,
1995[Medline].
Am J Physiol Heart Circ Physiol 276(3):H844-H857
0002-9513/99 $5.00
Copyright © 1999 the American Physiological Society
Top
Abstract
Introduction
