Content uploaded by Ngai Chin Lai
Author content
All content in this area was uploaded by Ngai Chin Lai on Aug 15, 2015
Content may be subject to copyright.
Previous papers in this series have examined diverse aspects
of the physiology of facultative and continuously air-breathing
fishes. These works quantified the influence of aquatic O2
tension (PwO∑) on the thresholds for air breathing (Graham and
Baird, 1982), established that hypoxia acclimation favourably
affects aerial respiratory efficiency by modulating air-
breathing frequency, haemoglobin O2-affinity and air-
breathing organ (ABO) volume (Graham, 1983), and
demonstrated that both PwO∑and body size influence
respiratory partitioning and the efficacy of branchial,
cutaneous and ABO gas-exchanging surfaces (Graham and
Baird, 1984; Graham et al. 1987). The objectives of the present
study are to define the relationships between air-breathing and
heart activity in two species of the family Synbranchidae,
Synbranchus marmoratus Bloch and Monopterus albus
(Zuiew), and to compare cardiorespiratory interactions among
the air-breathing fishes.
Synbranchids occur circumtropically in freshwaters.
Commonly referred to as swamp, or rice, eels (however, they
are not real eels which are in the family Anguillidae), these
fishes make extensive mud burrows, are protogynous
hermaphrodites and use their modified buccopharyngeal
chamber as an ABO (Taylor, 1831; Johansen, 1966; Liem,
1963; Rosen and Greenwood, 1976; Bicudo and Johansen,
1979). The ABO surface is covered by a vascular epithelium,
and in Monopterus (an Asian genus) the gills are reduced to
such an extent that this fish is an obligatory air breather
(Lomholt and Johansen, 1974, 1976). Some species (e.g. M.
cuchia) even have auxiliary respiratory sacs within the
buccopharynx (Taylor, 1831; Munshi et al. 1989).
1455
The Journal of Experimental Biology 198, 1455–1467 (1995)
Printed in Great Britain © The Company of Biologists Limited 1995
Extreme heart-rate lability accompanies the air-
breathing cycles of Synbranchus marmoratus and
Monopterus albus. When air is taken into the
buccopharyngeal air-breathing organ of these fishes, heart
rate increases sharply above pre-inspiration rates of
3–25beatsmin21to as high as 40–45beatsmin21. With
time, and as O2is depleted from the air-breathing organ,
heart rate gradually declines and drops to its lowest level
with, or following, exhalation. Relationships between air
breathing and sinus arrhythmia in M. albus were
investigated by injecting variable gas volumes and O2
mixtures into the cannulated air-breathing organ. Tests
were also carried out on undisturbed fish breathing
volitionally in atmospheres containing different O2levels.
Both the volume and O2content of the inspired gas affect
the level and duration of inspiration tachycardia.
Additional factors affecting tachycardia are the heart rate
prior to inspiration and the time since air was last held.
S. marmoratus is a non-obligatory air breather and uses
rhythmic branchial aquatic respiration to a greater extent
than M. albus, an obligate air breather. While the heart
rates of both species are increased during aquatic
ventilation, the higher heart rate to ventilation ratio in S.
marmoratus (2–3 versus approximately 1 in M. albus) seems
attributable to its more proficient aquatic respiratory
system. The available cardiorespiratory data for air-
breathing fishes indicate that the scope of air-inspiration
tachycardia is smaller in lungfishes and other primitive
species than in most teleosts. This difference is mainly
attributable to the greater chronotropic effect of
sympathetic cardiac innervation in teleosts.
Key words: Synbranchus marmoratus, Monopterus albus,
Synbranchidae, air-breathing fish, heart rate, sinus arrhythmia,
tachycardia, bradycardia.
Summary
THE TRANSITION TO AIR BREATHING IN FISHES
V. COMPARATIVE ASPECTS OF CARDIORESPIRATORY REGULATION IN
SYNBRANCHUS MARMORATUS AND MONOPTERUS ALBUS (SYNBRANCHIDAE)
JEFFREY B. GRAHAM, N. C. LAI, DAVID CHILLER AND JOHN L. ROBERTS*
Center for Marine Biotechnology and Biomedicine and the Marine Biology Research Division,
Scripps Institution of Oceanography, University of California San Diego, La Jolla, CA 92093-0204, USA
Accepted 7 March 1995
*Present address: Department of Zoology, University of Massachusetts, Amherst, MA 01003-0027, USA.
Introduction
1456
Synbranchus (a neotropical genus), in contrast, has a complete
set of gills and is not an obligatory air breather (Johansen,
1966; Graham et al. 1987). Other synbranchid specialisations
for air breathing include a reduced, ventrally positioned, single
opercular slit and several longitudinal buccal–branchial tissue
folds that permit marked ABO expansion during air breathing
(see Fig. 1A).
In most vertebrates, regulation of cardiac activity occurs
through balanced inputs from the cholinergic
(parasympathetic) and adrenergic (sympathetic) autonomic
nervous system. Increased vagal (cholinergic) tone usually
reduces both heart rate and contractility, whereas reduced
vagal tone or an increase in adrenergic activity often increases
both variables. Pronounced cardiac arrhythmias during air
breathing have been documented in a number of fish species
(Johansen et al. 1968a; Singh and Hughes, 1973; Lomholt and
Johansen, 1976; Farrell, 1978; Smatresk, 1988, 1990). The
most common pattern of arrhythmia is an inspiration-induced
tachycardia followed by the gradual onset of bradycardia as the
O2content of the breath falls, leading to full bradycardia with
exhalation. This pattern appears to optimize ventilation and
perfusion relationships by elevating blood flow to the ABO just
after a breath is taken and then gradually attenuating flow as
O2is depleted (Johansen, 1966, 1970).
Repetitive cycles of air-breathing tachycardia and ABO-
deflation bradycardia have been demonstrated in both
Synbranchus and Monopterus (Johansen, 1966; Lomholt and
Johansen, 1974, 1976; Roberts and Graham, 1985). Several
classes of receptor input are thought to contribute to the
integration of aerial respiration and sinus arrhythmias in fishes,
and recent studies have implicated mechanoreceptors within
the ABO or along the airway (Milsom, 1990; Smatresk, 1988,
1990). The response levels of the heart-regulating
mechanoreceptors have not, however, been defined for most
air-breathing fishes. Also, few studies have considered whether
the quality of the inspired gas affects the time course of sinus
arrhythmia during the breath-hold. In this paper, we
demonstrate ABO mechanoreceptor effects on the
cardiorespiratory regulation of S. marmoratus and M. albus
and present evidence that an ABO O2receptor influences
instantaneous air-breath tachycardia and other
cardiorespiratory variables.
Materials and methods
Specimens of S. marmoratus (150–960g) were collected
from streams in the Republic of Panama. Specimens of M.
albus (40–700g) were purchased in a Singapore fish market
and from commercial dealers in San Diego and San Francisco,
California. All fish were maintained in aerated laboratory
aquaria (at 24–27˚C) and fed about every 4 days with pieces
of cut mackerel and squid. Experiments with S. marmoratus
were conducted on six fish between 1984 and 1987. With
access to relatively few specimens, the work with S.
marmoratus was more qualitative than for M. albus for which
18 specimens were studied between 1988 and 1993.
Specimen preparation
Fig. 1A illustrates the placement sites of electrocardiogram
(ECG) and branchial electrodes and the branchial cannulae. To
make these attachments, fish that had been starved for at least
48h were anaesthetised in cool (8–10˚C) water for 15–30min
and then placed on a wet surgical table and covered with cool,
wet towels. Pairs of stainless-steel or copper (30 and 36 AWG)
electrode wires were implanted via small (22–25 gauge)
hypodermic needles onto the ventral wall of the branchial
chamber and adjacent to the heart (Fig. 1A), secured by self-
tying skin loops and stitches, and directed dorso-anteriorly
from the body. Branchial tubes (PE 50 or PE 90), with heat-
flared and smoothed tips and side ports, were passed through
the ventral branchial chamber wall via a hypodermic puncture,
secured with two stitches, and directed anteriorly. Two
cannulae were used so that simultaneous withdrawal and
injection tests could be carried out and also to reduce the
possibility that the thick tissue layers in the branchial folds
might occlude the cannula tips and thus prevent experimental
ABO-volume changes.
About 30min was required to instrument a fish. After
surgery, fish were placed in one of two different experimental
chambers (Fig. 1B,C) and allowed 48h for recovery and
chamber acclimation. Experimental data were obtained over
3–6 days following recovery, during which time the fish was
not fed. ECG and ventilation signals were amplified on either
d.c. or impedance circuits and sent to a Gould 2400 chart
recorder. In later studies, long-term records were stored on
reel-to-reel magnetic tape. After the experiments had been
completed, each fish was anaesthetised by chilling, all wires,
catheters and sutures removed, and wet body mass determined.
Effects of changes in ABO volume and the O2content of the
ABO gas
Controlled ventilation chamber
Fish fitted with both electrode sets and the branchial
chamber tubes were slipped gently and tail first into a slightly
inclined horizontal tube (Fig. 1B). The front of the tube was
then sealed with a rubber stopper into which were mounted two
stainless-steel hypodermic needles that were connected to the
branchial PE tubes on the inside of the chamber and to syringes
on the outside. The electrode wires were fed out between the
chamber wall and the stopper. The tube was filled with water
that was recirculated by an oscillating pump (Fisher 13-874-
32) operated at low voltage. An O2electrode (Yellow Springs
Instrument Co., no. 5750) was placed in the loop to monitor
PwO∑, which was kept at, or below, 4kPa to ensure that fish
would remain heavily reliant upon air-breathing. For most
tests, this system was maintained between 25 and 27˚C by
thermostattically controlled water heaters.
Inclining the tube slightly (Fig. 1B) enabled positioning of
an air pocket in the anterior top section of the chamber, just
over the head of the fish. This was convenient in preventing
gas bubble entrainment in the circulating water loop during
experiments and it additionally provided air access for the fish,
J. B. GRAHAM AND OTHERS
1457Synbranchid sinus arrhythmia
thus reducing stress levels as well as the chance of accidental
suffocation (M. albus is an obligatory air breather). Inclination
also aided breath exhalation, which is carried out most
effectively when the fish can raise its snout to about 45˚ with
respect to the horizontal plane (Graham and Baird, 1984).
Inclination also lessened the tendency of the fish to roll up on
and kink its wires and tubes, although this occasionally
happened.
Experiments
The objective of these tests was to determine how heart rate
was affected by ABO inflation and deflation and whether the
O2content of the ABO gas affected heart rate. For S.
marmoratus, volume effects were tested by incremental
injections of small (1–3ml) volumes of air or gas of known O2
content at regular intervals. These began at ‘zero’ ABO volume
(i.e. the fish was ‘between’ air breaths, had no gas in its ABO
and had a low heart rate) and continued until the organ was
full, as indicated by a distended branchial chamber and
expanded longitudinal tissue folds, and by the ‘overflow’ of
gas from the opercular slit during injection. This series was
followed by a stepwise volume withdrawal and the sequence
was then repeated.
Several techniques were employed to ensure that the
injections began at zero ABO volume. First, previous
experiments showed that, provided there is sufficient room for
S. marmoratus to raise its head to an angle of 45˚ or more, this
fish will void all gas from its ABO during a normal expiration
(Graham and Baird, 1984). Also, no gas is present in the ABO
of S. marmoratus during long periods of aquatic ventilation
(Graham et al. 1987). Preliminary tests with M. albus also
indicated complete ABO emptying and the absence of residual
ABO gas. Thus, by beginning experiments after expirations,
the assumption of a ‘zero volume’ ABO was reasonable. Also,
with careful placement of the catheter in the ventral wall of the
branchial chamber, we could extract most of the injected gas,
and this could be verified with a volume measurement.
Nevertheless, it is likely that a fraction of the injected gas was
not retrieved. In cases where the unrecovered volume was
suspected of being excessive (20%) or where we were not
certain that the ABO was gas-free, the zero volume state could
be established by purging the ABO with a large bolus of water
and postponing the experiments for 1 h, during which breathing
behaviour was continually monitored.
For M. albus, two types of experiments were carried out to
ECG
b
c
c
c
fv
ECG
b
oe P
In Out
t
ECG, b
A
B
C
Fig. 1. (A) Lateral and ventral views of the head of Synbranchus
marmoratus showing the positions of the branchial chamber cannulae
(c) and the branchial (b) and ECG electrodes. Also shown are the
small, ventral opercular valve (v) and longitudinal tissue folds (f) in
the floor of the branchial chamber. (B) The horizontal tube system
showing the connection between the branchial cannulae and syringes
used for the ‘push–pull’ and ‘volume overflow’ tests. The branchial
(b) and ECG electrodes were passed between the chamber wall and
stopper. The arrow shows the direction of the water flow through the
pump (p) and the oxygen electrode (oe). (C) L-shaped tube design.
Gases of known PO∑were circulated through the open space at the top
of the chamber via the ‘in’ and ‘out’ ports in the rubber stopper. The
pressure transducer port (t), a few centimetres below the water level,
was connected to a Statham gauge which monitored water-head
pressure to indicate when air breathing occurred (see text and Fig. 2).
The electrodes attached to the fish were passed through the stopper.
Water circulation indicated by arrows.
1458
determine the effects of gas quality on air-breath tachycardia.
First, ‘volume-overflow’ studies tested the effects of O2levels
in a fully inflated organ. This procedure incorporated a series
of large volume injections (see below) of gas of known O2
content, over a relatively short period (2min). An earlier study
on S. marmoratus (Graham and Baird, 1984) had established
the quantitative relationship between ABO volume (vin ml)
and body mass (min g) as:
logv = 20.825 + 0.737logm, (1)
and values determined from this equation are similar to ABO
volume data for M. albus (Lomholt and Johansen, 1974). From
this regression, it was possible to calculate the injection
volume required to overflow the ABO and thus to keep it at,
or near, maximal inflation (as indicated by equation 1, this
volume varied with fish body size). The injections began when
the fish was post-expiration (ABO empty, low heart rate), four
injections were given at 2min intervals, and heart rate was
measured 1min after injection.
A second set of experiments, the ‘push–pull’ tests, was
designed to measure the effects of different O2levels in a
relatively constant ABO volume. This was done by causing an
abrupt change in ABO O2content in the absence of a change
in ABO volume. First, the ABO was partially filled by injecting
enough gas (about 75% of maximum v, see equation 1) to
accelerate the heart while not fully distending the ABO. An
additional volume (20% of maximum v) of experimental gas
was then injected through one branchial cannula and the same
volume was withdrawn (either simultaneously or immediately
following injection) from the second cannula and heart rate
was determined. After the test, the initial gas volume was
withdrawn (i.e. leaving the ABO completely, or nearly, empty)
and, after waiting for 1min, another gas injection and
push–pull test were carried out.
Cardiac responses following volitional inspiration of mixed
gases
Test chamber
Experiments were also carried out using an L-shaped tube
(Fig. 1C) that was placed in a waterbath. The water in this tube
was slowly circulated and oxygen levels monitored. As above,
PwO∑was kept below 4kPa to ensure dependence on the aerial
oxygen supply. Tests were carried out at water temperatures
between 23 and 27˚C. Electrode leads, with sufficient slack to
permit free vertical movement by the fish, were passed up
through the vertical arm. From about 2cm below the water
level, the tube was completely wrapped in black plastic and the
fish often retreated to this area. Tests with S. marmoratus
carried out in this tube included the placement of a peritoneal
catheter for determination of the effect of atropine (1mgkg21)
on cardiorespiratory interactions.
Experiments
For tests of cardiac responses to different gas mixtures, the
vertical tube was sealed with a rubber stopper containing in-
and out-flow ports so that a controlled, mixed-gas respiratory
atmosphere could be maintained above the water surface
(Fig. 1C). These tests were carried out on M. albus fitted with
ECG and ventilation electrodes but without branchial cannulae.
Connection of a Statham pressure transducer to the pressure
port (Fig. 1C) enabled continuous monitoring of the pressure
change caused by air breathing (i.e. inspiration inflated the
ABO and thus displaced water and raised hydrostatic pressure;
Fig. 2). Data obtained with this system allowed investigation
J. B. GRAHAM AND OTHERS
Fig. 2. Sections of a 33min record showing six air breaths by a 560g Monopterus albus in the L-shaped tube (Fig. 1C). Numbers below each
panel indicate breath duration (s). The relative pressure trace (top) shows the step increase in water depth associated with inspiration followed
by a series of five ‘V-shaped’ exhalation–inhalation breathing cycles and then a final decrease with air-release. The ECG trace (middle) shows
the arrhythmia coincident with inflation and deflation. The impedance record (lower trace) registers displacement of the branchial chamber wall
with air-breathing events (T=25˚C).
Pressure
Branchial chamber impedance
640 400 160 220 180 360 Release
ECG
1min
1 2 3 4 5 6 Release
1459Synbranchid sinus arrhythmia
of both instantaneous and long-term features of mixed-gas
inspiration, without the disturbance associated with the close
proximity of an investigator, manual gas injections or a
variable injection rate or volume. A Wösthoff pump (Bochum,
Germany) was used to mix air with either N2or O2.
Experimental atmospheres in these tests ranged from 100 to
1.5% O2, and even 100% N2was used for short periods.
Results
Synbranchus marmoratus
Fig. 3 shows simultaneous branchial and ECG signals
recorded for S. marmoratus while air breathing in normoxic
(air = 21% O2), hypoxic (10.5% O2/89.5% N2), and
hypercapnic (10% CO2/90% air) atmospheres. Sequential
spikes in the branchial records indicate exhalation of a breath
followed by inhalation. For these data, the inter-breath interval
(IBI), the time between the release of old and the taking of new
air, is relatively short (5–40s). Depending upon conditions, the
IBI for S.marmoratus can vary considerably (Johansen, 1966).
Graham and Baird (1984) measured a mean IBI of 15min for
this fish; however, the observational range extended from 1 to
42min. Air exhalation can also lead directly to the onset of
aquatic ventilation (see below).
The ECG records for S. marmoratus (Fig. 3) show a slowing
of the heart just prior to breath release, a brief interval of heart
stoppage, and tachycardia following inspiration; the pattern
first described by Johansen (1966). Fig. 3 also shows that
inspired gas quality affects both the extent and duration of the
inspiration tachycardia. Under normoxic conditions (Fig. 3A),
breaths were held for longer times and supported a greater
range of cardiac activity than breaths taken in hypoxia
(Fig. 3B). Exposure to hypercapnia for 2h or longer (Fig. 3C)
resulted in a more variable, but usually shorter-duration, air-
breathing pattern and a greater degree of arrhythmia.
S. marmoratus frequently utilises aquatic ventilation
(Graham et al. 1987) and Fig. 4 shows the transition from air
breathing to aquatic ventilation. At the onset of aquatic
ventilation, the heart rate was relatively fast and the ratio of
heart rate to ventilation was 3.6. As ventilation rate increased,
this ratio fell to 2.4 (Fig. 4). A 1mgkg21dose of atropine, a
cholinergic antagonist, accelerated the heart rate of S.
marmoratus and obliterated the arrhythmic pattern associated
with normal air breathing (Fig. 4B). Administered via the
peritoneum, atropine began to have an effect within 1h and its
influence on the heart remained strong for nearly 20h. Air-
injection experiments (Fig. 4C) revealed that heart rate is
dependent upon ABO volume. Withdrawal of 3ml of gas from
the ABO caused bradycardia; however, before additional
withdrawals could be made, this fish spontaneously ejected all
the ABO gas, which slowed the heart to its pre-inflation
bradycardic level.
Monopterus albus
M. albus utilises aquatic ventilation to a lesser extent than
S. marmoratus and has a lower heart rate to ventilation ratio.
Fig. 5 compares heart rate during aquatic ventilation and air
breathing. These data were obtained using the system
described in Fig. 1C and, during the aquatic respiration phase,
the fish positioned its head just below the water surface and
thus ventilated in more oxygenated water. The ventilation rate
of this fish was similar to that of S. marmoratus (Fig. 4A);
however, the ratio of heart rate to ventilation was smaller
(above 2.0 in S. marmoratus and approximately 1.0 in M.
albus).
Features of the air-breathing sinus arrhythmia in M. albus
are described in Table 1 and Figs 2 and 6. Table 1 shows mean
data for 10 sequential and spontaneous air breaths recorded for
five fish under an air atmosphere (25–27˚C). This table
expresses the relative increase in post-air-breath heart rate in
terms of the sinus arrhythmia index (SAI) which ranged from
48 to 117% for these five fish. Table 1 also shows that, even
though there was not much variability among the five fish in
terms of average breath duration (5–9min), the individual
variation was quite large, as in S. marmoratus.
Fig. 2 shows the ECG record for a 560g M. albus
immediately before, and after, a series of six volitional air
breaths (in normoxia) spanning 33min. Also shown in the
record are step changes in water pressure and the branchial
activity associated with ABO ventilation. This figure
demonstrates the cumulative effect of sequential air breaths on
heart rate. The first breath was taken after a 35min IBI, during
Table 1. Sinus arrhythmia data for five Monopterus albus measured over 10 sequential air breaths in normoxic water
Pre-air-breath Inspiration Exhalation Mean breath
Mass heart rate heart rate SAI heart rate duration
Individual (g) (beatsmin−1) (beatsmin−1) (%) (beatsmin−1) (s)
1 450 18 (6) 39 (6) 117 22 (9) 287 (115)
2 690 20 (7) 36 (5) 80 33 (4) 423 (51)
3 150 22 (9) 41 (6) 86 31 (9) 498 (149)
4 190 25 (8) 37 (6) 48 27 (6) 368 (74)
5 220 21 (9) 42 (8) 100 26 (7) 551 (138)
SAI is the sinus arrhythmia index and is expressed as a percentage increase in heart rate with inspiration. For fish 1, the calculation is
100(39−18)/18=117%.
Values are means (S.E.M.) for five fish over 10 sequential breaths; T=25–27°C.
1460 J. B. GRAHAM AND OTHERS
A
Normoxic
B
Hypoxic
B
Hypercapnic
ECG
Impedance
1min
Fig. 3. ECG and branchial impedance activity records for a 260 g Synbranchus marmoratus air breathing in (A) normoxic (21% O2), (B) hypoxic
(10.5% O2) and (C) hypercapnic (10% CO2, 90% air) atmospheres in the L-shaped tube shown in Fig. 1C. At the time records were taken, the
fish had been air breathing in the specified atmosphere for several hours. The time scale (1min) shown between A and B applies to all three
records (T=27˚C).
1461Synbranchid sinus arrhythmia
Fig. 4. ECG and branchial chamber impedance records for a 390 g Synbranchus marmoratus. The 1min time scale shown under trace B applies
to both A and B. (A) The effect on heart rate of a spontaneous transition from air breathing to gill ventilation. The impedance trace shows a
typical air-breath signal and the associated cardiac acceleration (ab). Cyclic aquatic gill ventilation (v) begins about 80s after the air breath and
results in a lower level of tachycardia than followed the air breath. After 5min of aquatic ventilation, the maximum ventilation rate is about
18ventilationsmin21and the ratio of heart rate to gill ventilation is 2.4. (B) ECG and branchial impedance records showing air breathing (ab)
following atropinisation (1mgkg21). (C) Effect on heart rate of four manual air injections (3ml each) into the air-breathing organ (ABO),
followed by the withdrawal of 3ml, and then the spontaneous release of all ABO gas by the fish. Numbers below the trace indicate injected
and cumulative volumes at each step (T=27˚C). On the basis of equation 1, the estimated ABO volume of this fish was 12ml.
A
Transition from air to water breathing
B
Atropine
C
ABO inflation
ABO
empty +3ml
3ml +3 ml
6ml +3 ml
9ml +3 ml
12ml −3ml
9ml Release
ABO empty
ECG
ECG
ECG
Impedance
Impedance
ab
1min
ab ab ab
v
1462
which there was no aquatic ventilation. This breath was held
for 640s and was followed immediately by breath 2 (400s).
Heart rate prior to the first breath was about 7beatsmin21,
abruptly increasing to 14beatsmin21(SAI=100%) when
breath 1 was taken. Tachycardia following breath 2 peaked at
about 18beatsmin21and by breath 6 tachycardia just after
inspiration was 21beatsmin21. Thus, there was a steady rise
in the mean heart rate over the six-breath sequence. The
pressure record indicates that breaths 2, 3 and 4 were of a larger
volume than the others. Although our techniques did not allow
quantification of this difference, the possibility that this larger
volume affected heart rate is unlikely as the relatively smaller-
volume breaths (5 and 6) were accompanied by higher heart
rates. Upon the expiration of breath 6, heart rate fell abruptly
to 6beatsmin21.
Additional aspects of the sinus arrhythmia of M. albus are
shown in Fig. 6, which covers almost 3h of air breathing and
three complete air-breath and IBI cycles. Although both air-
breath and IBI durations vary, the effects of inspiration, time
and exhalation on heart rate are clearly evident.
Overflow injections of gases with different O2levels were
given over a short period (8min) in order to assess the effect
of the oxygen content of ABO gas on heart rate. Fig. 7 gives
the results for a 690g M. albus (individual 2, Table 1) given
four 20ml injections at 2min intervals. The injection series
began with the ABO empty and with initial heart rates that
varied between 5 and 12beatsmin21. In all cases, heart rate
increased with the first injection; however, except for 10%
CO2, the relative increase was directly proportional to the
amount of O2contained in the injected gas. Similarly, with the
exception of the 10% CO2injection, and allowing for
differences in the pre-injection rates, heart rate following the
final overflow injection in each series was directly proportional
to the O2content in the injected gas. That O2content was, in
fact, determining maximum heart rate in these tests is
demonstrated by the pronounced tachycardia resulting from the
overflow injection of 20ml of pure O2into the previously N2-
inflated ABO (Fig. 7). Similar mixed-gas effects on heart rate
were verified in the eight other fish tested.
Table 2 shows results of a push–pull test carried out on a M.
J. B. GRAHAM AND OTHERS
A
30s
ab
b
ECG
P1
P2
ab ab
B
Fig. 5. (A) Records of aquatic branchial ventilation (b), the ECG and relative water-head pressure (P1) in the L-shaped tube (Fig. 1C) for a 40 g
Monopterus albus. Also shown is a simultaneous amplified pressure signal (P2). Note that the pressure traces are stable because the fish was
not air breathing. At maximal ventilation rate (23ventilationsmin21), the heart rate to ventilation ratio is 1.0. (B) Record during air breathing
showing episodic branchial movements and corresponding pressure oscillations with the initial, second and third air breaths (ab) in a series (see
Fig. 2) (T=27˚C). Scales for time, relative pressure and ECG are the same in both A and B.
1463Synbranchid sinus arrhythmia
albus (tests of this type were completed on four fish). In this
series, 3ml of experimental gas was injected and the same
volume was then immediately withdrawn (i.e. organ volume
expanded for 10s) and the ensuing heart responses were
monitored over six 10s intervals. The ABO was then deflated
and, after 1min, another gas was injected. Table 2 indicates
that the small, brief inflation pulse initiated tachycardia, but
that the extent to which tachycardia developed depended on
both the pre-inflation heart rate and the relative O2content of
the injected gas. In five of the seven injections shown in
Table 2, tachycardia developed within 10s of gas injection. In
these cases, pre-injection rate was 18–24beatsmin21, whereas
in the two cases where tachycardia did not develop, pre-
inflation heart rate was 30–36beatsmin21. The effect of gas
quality can be seen from the heart rates at 20 and 30s. In three
of the four cases where the injection contained less than 21%
O2(i.e. injection of N2or re-injection of ‘used’ air), heart rate
began to decrease by 20s after injection. This contrasts with
the tachycardia induced by injection of air and pure O2, which
required about 30s to develop maximally.
The rapidity of the onset of a gas-quality component of air-
breath tachycardia in M. albus was investigated with the
apparatus shown in Fig. 1C. Fig. 8, based on experiments
carried out over a period of several days with three specimens,
shows that inspired-breath O2content affects both breath
duration and heart rate. A reduction in inspired O2content
Table 2. Cardiac responses of Monopterus albus following
the nearly simultaneous injection/withdrawal (push–pull
method) of ABO gases containing different amounts of O2
Instantaneous heart rate (beatsmin−1)
Gas Pre- Time post-injection (s)
sequence breath 10 20 30 40 50 60
Air 18 36 36 42 42 42 36
‘Used’ air 36 36 30 30 24 24 24
Air 18 30 30 36 30 30 24
‘Used’ air 24 36 36 30 24 30 30
N230 30 24 24 18 24 18
N218 24 18 18 18 18 18
O218 36 36 42 36 36 42
Gases were delivered in the order shown.
Instantaneous heart rates are indicated for the pre-injection state
and over each 10s interval of the first minute following gas
manipulation.
Data are for fish 3 in Table 1; T=25–27°C.
30
20
10
Heart rate (beatsmin−1)
6 24 42 60 78 96 114 132 150 168
Time (min)
R
R
R
R
II
I
Fig. 6. A 3h record of heart rate and air breathing in a 170g Monopterus albus. Vertical lines indicate times of air-breath inspiration (I) and
release (R) (T=23˚C).
Fig. 7. Effects on heart rate of a series of four 20ml overflow
injections of gases containing different proportions of O2into the
ABO of Monopterus albus (690g, estimated ABO volume 19ml).
Experimental gases are: air (filled squares) and pure N2(filled
triangles). Mixtures are 18.9% O2(open triangles), 15.7% O2(open
circles), 10.5% O2(filled circles) and 90% air/10% CO2(open
squares). Injection 5 shows the effect of replacing N2with O2
(T=25˚C).
30
20
10
40
Heart rate (beatsmin−1)
23 4501
Injection number
1464
causes a general decline in mean breath duration, but post-
inspiration tachycardia (defined here as the time required for
the first five heart beats) was unaffected by oxygen levels as
low as 5.2%. Inspiration of both 3.1 and 1.5% O2, however,
caused the immediate (1–2s) behavioural response of air
ejection and affected air-breath tachycardia (Fig. 8). Many of
the breaths taken in 1.5% O2were released before 10 heart
beats had occurred (11–27s), and exposure to pure N2greatly
slowed the heart and, in most cases, the gas was not held for
long enough to record more than two or three heart beats.
Discussion
Cardiorespiratory integration in the Synbranchidae
In fishes, as in most vertebrates, heart rate depends upon the
balance between inhibitory vagal (cholinergic) inputs and
stimulatory (adrenergic) effects of either sympathetic nerve
fibres or circulating catecholamines (Laurent et al. 1983;
Taylor, 1987; Farrell and Jones, 1992). Vagal inhibition can
affect both chronotropic and inotropic cardiac action and may
be elicited by stress stimuli, by hypoxia or by input from both
mechano- and chemoreceptors. Excitatory heart stimulation
can occur as a result of reduced vagal tone, an increase in the
central-vessel blood volume (i.e. by Starling’s law), the action
of blood-borne catecholamines or the direct effect of
adrenergic nerve (sympathetic innervation) activity.
From experimental manipulation of ABO-gas volume and
content, and by allowing fish to breathe air as required without
disturbance in experimentally determined gas mixtures, this
study has demonstrated the presence of volume- and O2-
mediated ventilatory sinus arrhythmia in both S. marmoratus
and M. albus.
Volume effects
The presence of a volume-mediated cardiac arrhythmia in
these two synbranchids is in agreement with previous studies
on this group (Johansen, 1966; Lomholt and Johansen, 1974).
In the air-breathing teleosts that have been studied to date
(summarised in Table 3), ABO deflation elicits bradycardia
and reduces ABO perfusion while spontaneous or experimental
ABO inflation initiates tachycardia, increases ABO perfusion
and reduces aquatic ventilation. In both S. marmoratus
(Johansen, 1966) and M. cuchia (Lomholt and Johansen,
1974), heart rate directly affects ABO perfusion because blood
ejected from the heart flows into the ventral aorta and to the
branchial arches and buccopharyngeal epithelium (Rosen and
Greenwood, 1976). In most species, the transduction of ABO-
volume change occurs via mechanoreceptors sensitive to either
wall stretch or displacement. Cardiorespiratory integration is
mediated by the interconnection of the ABO and the cardiac
and respiratory control systems via vagal loops consisting of
sensory and motor tracks (Taylor, 1987; Smatresk, 1988,
1990).
Oxygen effects
Evidence for an O2-dependent component of cardiac
regulation in M. albus has been obtained in this study. The
overflow experiments examined the effect of the O2content of
ABO gas on heart rate at maximum ABO volume and
demonstrated (Fig. 7) that a series of gas deliveries, given over
a short period, have an additive effect on heart rate. However,
both the relative increase in heart rate following the first
injection and heart rate after the final delivery were strongly
influenced by gas O2content. The effect of 18.9 % O2(i.e. 90%
air) on heart rate was diminished by the addition of 10% CO2.
Also, continued exposure to N2over the four-injection
sequence resulted in an absence of tachycardia, an effect that
was rapidly reversed by O2injection (Fig. 7).
The ‘push–pull’ experiments had the objective of changing
ABO-O2level while minimally affecting ABO volume. It can
be seen in Table 2 that slight volume changes taking place
during this technique would trigger a tachycardia, but that the
J. B. GRAHAM AND OTHERS
1200
240
180
60
0(39)
(36)
(33)
(30) (18)
(24) (261) (9)
8
7
6
5
4
1300
Mean air-breath duration (s)
Time for five heart beats (s)
0 25 50 75 100 O2
% Air
Fig. 8. Effects of different air-phase O2levels on the mean air-breath
duration and instantaneous heart rates of three Monopterus albus.
Sample sizes are given in parentheses. Values are means ± S.E.M.
(T=24–26˚C).
1465Synbranchid sinus arrhythmia
intensity and duration of this response were ultimately
determined by gas quality and heart rate prior to gas injection.
Observations of freely air-breathing M. albus also indicated an
effect of inspired O2on breath duration and showed that fish
would void severely hypoxic and anoxic breaths within a few
seconds of inspiration (Fig. 8). The presence of an ABO
chemosensor is suggested by the rapidity of this gas-voiding
reflex, which was about 2–4 times faster than would be
expected if stimulation had occurred via remote vascular
receptors located downstream from the branchial chamber
(Eclancher, 1975; Eclancher and Dejours, 1975).
The effect of branchial-volume displacement on heart rate
could also be demonstrated by injecting a moderately large
bolus of water into the branchial chamber of M. albus between
air breaths.In tests on seven fish, water injection elicited a
tachycardia and this tachycardia was greater and was sustained
for longer when hyperoxic as opposed to hypoxic water was
injected. These observations suggest that an aquatic-O2sensing
system, analogous to the one operating in air, has the capacity
to affect heart rate. It is entirely possible that the same O2
chemosensors could function in both water and air; however,
this cannot be verified without further information regarding
sensor morphology, location and development.
Fishes that respire aquatically have the capacity to monitor
inspired PwO∑with externally facing sensors located in the
anterior branchial arches (Taylor, 1987; Smatresk, 1988;
Milsom, 1990). Thus, it is not surprising that a functional air-
breath O2sensor would occur in synbranchids because of their
use of the buccopharyngeal cavity as an ABO. Our results for
S. marmoratus and M. albus suggest that the ABO-O2receptor
transduces information about inspired gas quality (O2content),
which influences the intensity of the tachycardic response to
inflation as well as the rate of bradycardia during the breath-
hold.
Both S. marmoratus and M. albus normally reside in mud
burrows that can be hypoxic (J. B. Graham, personal
observation). These fish may also gulp air from enclosed gas
pockets under floating vegetation. In such circumstances, the
ABO-O2receptor could monitor the O2content of each breath
and elicit a breath-voiding response if required. Pronounced
effects on the heart rate and breath-voiding were found at O2
levels between 1.5 and 3.1%, but it is unlikely that S.
marmoratus and M. albus would consistently encounter air
oxygen levels this low. Thus, the ABO-O2sensor may be
important in the modulation of mechanoreceptor and other
stimuli affecting air-breath tachycardia, in attenuating
tachycardia as breath PO∑declines and, ultimately, in
terminating the breath when the PO∑drops to a level
unfavourable for O2transfer to blood (Graham and Baird,
1984).
Additional aspects of heart control
In addition to the effects of inspired (or injected) gas volume
and quality, this study has shown that inspiration tachycardia
can be influenced by the length of time that a breath is held in
the ABO and the heart rate prior to gas introduction. Even
Table 3. Heart-rate responses to air intake in aquatic air-
breathing fishes
Heart rate
(beatsmin−1)
Temperature Pre-air- Post-air-
Species (°C) breath breath Reference
Neoceratodus 18 14 181Johansen et al.
forsteri (1968a)
Protopterus 25 34 (2.1) 34 (2.0)2Johansen and
aethiopicus Lenfant (1968)
25 33 361,3 Johansen et al.
(1968a)
25 36 36 Szidon et al.
(1969)
Lepidosiren 27 29432 Axelsson et al.
paradoxa (1989)
Lepisosteus 18–21 35 375Smatresk and
oculatus Cameron (1982)
−60 70 Smatresk (1988)
Amia calva 25–30 33 346Johansen et al.
(1970a)
Arapaima gigas 26–30 34 (11) 34 (10) Farrell (1978)
Electrophorus 28 28 66 Johansen et al.
electricus (1968b)
−14 30 Johansen et al.
(1970b)
Hoplerythrinus 26–30 61 (12) 82 (23) Farrell (1978)
unitaeniatus
Ancistrus 25 110 150 Graham (1983)
chagresi
Clarias 25 30 39 Jordan (1976)
batrachus
Monopterus 20 14–16 34–381Lomholt and
cuchia 28–30 12 70 Johansen (1976)
Monopterus 25 9 (2.1) 31 (2.5) This study
albus 30 17 (1.6) 55 (1.2)
Synbranchus 20–22 6 27 Johansen (1966)
marmoratus −21 35 Johansen et al.
(1970b)
25 5–10 40 Roberts and
Graham (1985)
Anabas 25 30 577Singh and
testudineus Hughes (1973)
Values in parentheses, where given, are S.E.M.
The species are arranged in phylogenetic order and, beginning with
Arapaima, all listed species are teleosts.
1Data obtained by manual air inflation/deflation.
2Data compiled from reference given, Figs 2, 10 and 12.
3Lung deflation reduced heart rate from 18 to 12beatsmin−1(see
Fig. 15 of reference).
4Resting heart rate. Rates following vasoactive drug administration
were: atropine 32beatsmin−1, propranolol 25beatsmin−1.
5Pre- and post-air-breath differences not significant.
6Data compiled from reference, Fig. 11.
7Rates and amount of increase varied with activity and water O2
content.
1466
when not air breathing, S. marmoratus and M. albus engage in
long periods of branchial apnoea (see also Lomholt and
Johansen, 1974; Graham et al. 1987). Bradycardia
accompanies branchial apnoea (Figs 4A, 5A), and aquatic
ventilation triggers tachycardia. In the latter case, stimulation
of the heart would appear to result from the cyclic deformation
of the branchial apparatus during ventilation and this contrasts
with air breathing, where a single ABO inflation (wall
stretching) event triggers tachycardia (Fig. 4C). Our data
indicate that these two different volume-change transduction
events have a similar effect on heart rate although, in both
cases, the normoxic stimulation of peripheral chemoreceptors
would also be an important factor in sustaining tachycardia
during air breathing and aquatic ventilation. Similarly,
differences in the heart rate to ventilation ratio found here for
M. albus and S. marmoratus may reflect differences in the
exchange capacity of their aquatic respiratory surfaces as well
as variables such as PwO∑and metabolic demand.
Comparative aspects of cardiorespiratory interaction
In contrast to our findings for synbranchids, in the electric
eel Electrophorus electricus, ABO gas quality does not affect
heart activity. Johansen et al. (1968b) reported no difference
in the extent of tachycardia elicited by the injection of either
pure O2or N2into the ABO of E. electricus. As with the
synbranchids, E. electricus uses its buccal chamber as an ABO
and retains a reduced, although functional, branchial
epithelium. While branchial tissue atrophy might explain the
lack of a specific response to either N2or O2in E. electricus,
Johansen et al. (1968b) based their conclusions on only a few
records made on specimens recovering from anaesthesia,
which may have desensitised possible peripheral
chemosensory responses. The records reported were too brief
to determine whether centrally mediated receptors in either the
blood or myocardium were subsequently affected.
A marked difference in the cardiac response to air breathing
can be seen for lungfishes (Neoceratodus,Lepidosiren,
Protopterus) and to some extent for the garfish (Lepisosteus)
relative to the teleosts (Table 3, note that Arapaima and all of
the species listed below it in this table are teleosts). Although
air-breath-initiated changes in pulmonary and systemic blood
flow are well documented for lungfish (Johansen et al. 1968a;
Szidon et al. 1969), the role played by air-breath tachycardia
in these fishes appears to be minor. The difference between
lungfishes and teleosts could be attributable to differences in
the extent of cholinergic inhibition. However, it seems more
likely that this difference is a phylogenetic consequence of the
divergent patterns of cardiac regulation that exist among air-
breathing fishes. Lungfishes, for example, lack sympathetic
cardiac innervation entirely, and there is only limited
sympathetic innervation to the heart of Lepisosteus (Laurent et
al. 1983).
The need exists for additional comparative studies
controlling for the many variables likely to influence pre- and
post-air-breath heart rate. Table 3 does, nevertheless, suggest
a phyletic difference, based on innervation pattern, in the
cardiac responses to air breathing. It further suggests that these
differences have resulted in radically different vasomotor
responses to ABO inflation. At one extreme is the highly
integrated vasomotor control system of Lepidosiren and
Protopterus, which regulates cardiac output to the nearly
separate and parallel pulmonary and systemic circulations, in
phase with the air-breathing cycle (Johansen et al. 1968a;
Szidon et al. 1969). In contrast to the teleosts, most of which
have an in-series ABO to systemic blood flow, these lungfish
modulate the balance between pulmonary and systemic
circulation by changing peripheral resistance.
References
AXELSSON, M., ABE, A. S., BICUDO, J. E. P. W. AND NILSSON, S.
(1989). On the cardiac control in the South American lungfish,
Lepidosiren paradoxa. Comp. Biochem. Physiol. 93A, 561–565.
BICUDO, J. E. P. W. AND JOHANSEN, K. (1979). Respiratory gas
exchange in the air-breathing fish, Synbranchus marmoratus. Env.
Biol. Fish. 4, 55–64.
ECLANCHER, B. (1975). Contrôle de la respiration chez les poissons
téléostéens: réactions respiratoires a des changements
rectangulaires de l’oxygènation du milieu. C. R. hebd. Séanc. Acad.
Sci. Paris 280, 307–310.
ECLANCHER, B. AND DEJOURS, P. (1975). Contrôle de la respiration
chez les poissons téléostéens: existence de chemorécepteurs
physiologiquement analogues aux chemorécepteurs des vertébres
superiéurs. C. R. hebd. Séanc. Acad. Sci. Paris 280, 451–453.
FARRELL, A. P. (1978). Cardiovascular events associated with air-
breathing in two teleosts, Hoplerythrinus unitaeniatus and
Arapaima gigas. Can. J. Zool. 56, 953–958.
FARRELL, A. P. AND JONES, D. R. (1992). The heart. In Fish
Physiology, vol. XII (ed. W. S. Hoar, D. J. Randall and A. P.
Farrell), pp. 1–88. New York: Academic Press.
GRAHAM, J. B. (1983). The transition to air breathing in fishes. II.
Effects of hypoxia acclimation on the bimodal gas exchange of
Ancistrus chagresi (Loricariidae). J. exp. Biol. 102, 157–173.
GRAHAM, J. B. AND BAIRD, T. A. (1982). The transition to air breathing
in fishes. I. Environmental effects on the facultative air breathing
of Ancistrus chagresi and Hypostomus plecostomus (Loricariidae).
J. exp. Biol. 96, 53–67.
GRAHAM, J. B. AND BAIRD, T. A. (1984). The transition to air breathing
in fishes. III. Effects of body size and aquatic hypoxia on the aerial
gas exchange of the swamp eel Synbranchus marmoratus. J. exp.
Biol. 108, 357–375.
GRAHAM, J. B., BAIRD, T. A. AND STOCKMANN, W. (1987). The
transition to air breathing in fishes. IV. Impact of branchial
specializations for air breathing on the aquatic respiratory
mechanisms and ventilatory costs of the swamp eel, Synbranchus
marmoratus. J. exp. Biol. 129, 83–106.
JOHANSEN, K. (1966). Air breathing in the teleost Symbranchus
marmoratus. Comp. Biochem. Physiol. 18, 383–395.
JOHANSEN, K. (1970). Air breathing in fishes. In Fish Physiology, vol.
IV (ed. W. S. Hoar and D. J. Randall), pp. 361–411. New York:
Academic Press.
JOHANSEN, K., HANSON, D. AND LENFANT, C. (1970a). Respiration in
a primitive air breather, Amia calva. Respir. Physiol. 9, 162–174.
JOHANSEN, K. AND LENFANT, C. (1968). Respiration in the African
lungfish, Protopterus aethiopicus. II. Control of breathing. J. exp.
Biol. 49, 453–468.
J. B. GRAHAM AND OTHERS
1467Synbranchid sinus arrhythmia
JOHANSEN, K., LENFANT, C. AND HANSON, D. (1968a). Cardiovascular
dynamics in lungfishes. Z. vergl. Physiol. 59, 157–186.
JOHANSEN, K., LENFANT, C. AND HANSON, D. (1970b). Phylogenetic
development of pulmonary circulation. Fedn Proc. Fedn Am. Socs
exp. Biol. 29, 1135–1140.
JOHANSEN, K., LENFANT, C., SCHMIDT-NIELSEN, K. AND PETERSEN, J.
A. (1968b). Gas exchange and control of breathing in the electric
eel, Electrophorus electricus. Z. vergl. Physiol. 61, 137–163.
JORDAN, J. (1976). The influence of body weight on gas exchange in
the air-breathing fish, Clarias batrachus. Comp. Biochem. Physiol.
53A, 305–310.
LAURENT, P., HOLMGREN, S. AND NILSSON, S. (1983). Nervous and
humoral control of the fish heart: structure and function. Comp.
Biochem. Physiol. 76A, 525–542.
LIEM, K. F. (1963). Sex reversal as a natural process in the
synbranchiform fish Monopterus albus. Copeia 1993, 303–312.
LOMHOLT, J. P. AND JOHANSEN, K. (1974). Control of breathing in
Amphipnous cuchia, an amphibious fish. Respir. Physiol. 21,
325–340.
LOMHOLT, J. P. AND JOHANSEN, K. (1976). Gas exchange in the
amphibious fish, Amphipnous cuchia. J. comp. Physiol. 107,
141–157.
MILSOM, W. K. (1990). Mechanoreceptor modulation of endogenous
respiratory rhythms in vertebrates. Am. J. Physiol. 259,
R898–R910.
MUNSHI, J. S. D., HUGHES, G. M., GEHR, P. AND WEIBEL, E. R. (1989).
Structure of the air-breathing organs of a swamp mud eel,
Monopterus cuchia. Jap. J. Ichthyol. 35, 453–465.
ROBERTS, J. L. AND GRAHAM, J. B. (1985). Adjustments of cardiac rate
to changes in respiratory gases by a bimodal breather, the
Panamanian swamp eel, Synbranchus marmoratus. Am. Zool. 25,
51A (Abstr. 266).
ROSEN, D. E. AND GREENWOOD, P. H. (1976). A fourth neotropical
species of synbranchid eel and the phylogeny and systematics of
synbranchiforme fishes. Bull. Am. Mus. nat. Hist. 157, 1–69.
SINGH, B. N. AND HUGHES, G. M. (1973). Cardiac and respiratory
responses in the climbing perch Anabas testudineus. J. comp.
Physiol. 84, 205–226.
SMATRESK, N. J. (1988). Control of the respiratory mode in air-
breathing fishes. Can. J. Zool. 66, 144–151.
SMATRESK, N. J. (1990). Chemoreceptor modulation of endogenous
respiratory rhythms in vertebrates. Am. J. Physiol. 259,
R887–R997.
SMATRESK, N. J. AND CAMERON, J. N. (1982). Respiration and
acid–base physiology of the spotted gar, a bimodal breather. I.
Normal values and the response to severe hypoxia. J. exp. Biol. 96,
263–280.
SZIDON, J., LAHIRI, S., LEV, M. AND FISHMAN, A. P. (1969). Heart
and circulation of the African lungfish. Circulation Res. 25,
23–38.
TAYLOR, J. (1831). On the respiratory organs and air-bladder of
certain fishes of the Ganges. Edinb. J. Sci. 5, 33–51.
TAYLOR, E. W. (1987). Cardiovascular–respiratory interactions in fish
and crustaceans. In The Neurobiology of the Cardiorespiratory
System (ed. E. W. Taylor), pp. 277–303. Manchester: Manchester
University Press.
