Fish Physiology and Biochemistry vol. 12 no. 2 pp 101-110 (1993) Kugler Publications, Amsterdam/New York
Cardiac activity, ventilation rate and acid-base regulation in rainbow trout exposed to hypoxia and combined hypoxia and hypercapnia Kristian Borch, Frank B. Jensen and Bent B. Andersen Institute of Biology, Odense University, DK-5230 Odense M, Denmark
Accepted: January 29, 1993 Keywords: heart rate, heart rate variability, ventilation, acid-base status, hypoxia, hypoxia-hypercapnia
Abstract A computerised system for non-invasive monitoring of heart and ventilation rates and the time intervals between heart beats and between breaths was developed and used to investigate cardio-respiratory changes in rainbow trout exposed to hypoxia and to combined hypoxia and hypercapnia. Upon exposure to hypoxia and hypoxia-hypercapnia the arterial 02 tension decreased from about 90 mmHg to about 30 mmHg. Acid-base changes were small in hypoxia whereas exposure to combined hypoxia-hypercapnia caused a large extracellular respiratory acidosis. This acidosis was completely compensated within 24h by accumulation of bicarbonate in plasma to concentrations twice the normoxic values. The ventilation rate was increased to higher values in hypoxic-hypercapnic trout than in hypoxic trout. In contrast to previous reports, the heart rate increased in hypoxia. On top of the tachycardia response to hypoxia, the heart rate was governed by circadian rhythms, with higher heart rates during the day than during the night. The time interval between heart beats varied considerably in normoxic fish. Hypoxia strongly reduced this variability, which may originate in a reduced cholinergic tone to the heart. The width of the frequency distribution of the time intervals between breaths was not affected by hypoxia. The degree of cardio-respiratory synchronization was low in both normoxic and in hypoxic and hypoxic-hypercapnic trout.
Introduction Environmental hypoxia induces physiological adjustments at several steps in the oxygen transport cascade in order to safeguard respiratory function in fishes (Jensen et al. 1993). The ventilatory flow rate increases immediately by elevations in both ventilatory frequency and stroke volume, and the gill diffusion conductance is improved (Randall and Daxboeck 1984). The blood convective conductance can be increased both by an elevation of
cardiac output and by an increased blood 02 capacitance coefficient. However, whereas the 02 capacitance coefficient increases in hypoxia (Jensen et al. 1993), cardiac output is typically considered to be unchanged (Randall and Daxboeck 1984). It appears, that in many fish species heart rate is reduced by hypoxia whereas stroke volume increases, leaving cardiac output unchanged (Randall and Daxboeck 1984; Fritsche and Nilsson 1993). Examples of an unchanged heart rate or an increased heart rate and/or cardiac output are, however, also
Correspondence to: Dr. F.B. Jensen, Institute of Biology, Odense University, DK-5230 Odense M, Denmark.
102
Fig. 1. Scheme of experimental set-up. GMP, gas mixing pump. A, amplifier. F1, F2, electronic filters. C, comparator. V, monostable multi-vibrator. A/D, multi channel A/D converter. PC, IBM compatible PC. See text for details.
available (Cech et al. 1977; Berschick et al. 1987; Glass et al. 1991). Oxygen depleted aquatic habitats often tend to be both hypoxic and hypercapnic, since the processes that consume the water 02 (microbial degradation of organic matter) also produce CO 2. In contrast to the many studies on hypoxia in fishes, only few studies have addresed the physiological consequences of combined hypoxia and hypercapnia (Jensen and Weber 1982, 1985a, b; Thomas 1983). In comparison with fish exposed to hypoxia alone, fish exposed to combined hypoxia and hypercapnia faces the additional problem of an elevated CO 2 tension, which causes a respiratory acidosis. This respiratory acidosis raises the requirement for acidbase regulation and may further compromise blood 02 transport. The present study investigated acid-base changes and ventilatory and cardiac activities in rainbow trout exposed to hypoxia and to combined hypoxiahypercapnia. In order to enable measurement and analysis of heart and ventilation rate, the beat-tobeat variability in these and the degree of cardiorespiratory synchronization, a computerised system for continuous non-invasive monitoring of ventilation and heart activities was developed.
Materials and methods Animals and experimental set-up Rainbow trout (Oncorhynchus mykiss), 350-450 g in weight, were obtained from a local fish farm and kept at a 12h light: 12h dark rhythm in 5001 holding tanks with 14°C normoxic water for a minimum of 14 days before experimentation. The fish were cannulated to the dorsal aorta (Soivio et al. 1975) under 2-phenoxy-ethanol anaesthesia, whereupon the fish were transferred to a flow chamber of a water recirculating system and allowed to recover for 72h in 14°C normoxic water. The water recirculating system consisted of a reservoir aquarium from which water was pumped through the flow chamber at a rate of 901 h-1 (Fig. 1). The flow chamber was submerged in an isolated Faraday box, which minimized electric noise and shielded the animal from visual disturbance. The total water volume of the system was 401 and water was renewed daily. Water composition was: Ca + + = 3.5 mM, Na + = 1.2 mM, C1- = 1.3 mM, total CO2 = 5 mM. Normoxic (N) conditions were obtained by bubbling of air through the equilibration aquarium. Hypoxia was induced by bubbling with an air/N 2 gas mixture, and hypoxia-hypercapnia was induced by bubbling with an air/CO2/N 2 gas mixture. Gas mixtures were delivered by a Wosthoff (Bochum, FRG) gas mixing pump.
103 Collection and storage of ventilatory and cardiac activity data
--
110 -
uu-
Pl
N
90 -
A pair of stainless steel electrodes (2 mm diameter) placed at the top and bottom of the experimental fish chamber were used as recording electrodes. Two similar electrodes on each side of the chamber acted as ground electrodes (Fig. 1). The electrical signal from the fish was monitored by the electrodes and amplified one thousand times in a differential low noise AC amplifier (input resistance > 1 MD, highpassl.5 Hz, 20 dB/dec). The signal was then separated by two filters. One filter (lowpass, 5 Hz, 40 dB/dec) isolated the ventilatory signal, and the other filter (bandpass, 15-90 Hz, 80 dB/dec, 50 Hz notch) isolated the electrocardiogram (ECG) signal. The ventilatory signal was sent directly to an A/D converter. The ECG signal was sent to an A/D converter via a comparator (C), triggering a monostable multi-vibrator (V) that acted as pulse stretching circuit. The multi channel A/D converter was connected to the parallel printer port of an IBM-compatible PC (Fig. 1). A turbo pascal programme (modified from the programme of Depledge and Andersen 1990) controlled the A/D conversion via the parallel port. Counting of the ventilatory and cardiac signals was controlled by a software variable (high/low level) hysteresis discriminator for each channel. Two windows on the monitor acted as an oscilloscope, showing the ECG and ventilation signals live. The number of breaths and heart beats were stored on disk every minute, and shown in two further windows on the monitor as histograms. A separate turbo pascal programme measured and stored the time periods between heart beats and the time periods between breaths.
Experimentalprotocol and measurements After 72h of normoxic-normocapnic exposure (Po2 = 150 mmHg, P 0 2 = 0.7 mmHg) the gas supply to the equilibration aquarium was switched to either the hypoxic (Po2 = 60 mmHg) gas mixture (N = 5) or the hypoxic-hypercapnic (Po2 = 60 mmHg; Po0 2 = 4.8 mmHg) gas mixture (N= 5). This slowly changed water P 02 and Pco 2 towards
E E
70- -
O 50tL
30 -
10 -
-1
3 6
2426
48
time (h) Fig. 2. Arterial oxygen tension (means ± SEM) in rainbow trout in normoxa (N), upon exposure to hypoxia (H) (, N = 5) and to combined hypoxia and hypercapnia (H + H) ( , N = 5), and upon return to normoxia (N). The dashed line depicts the mean values in fish kept in normoxic water (N = 3) throughout the experiment. A significant difference (paired Students t-test) compared to normoxic control values (t = - Ih) is depicted by asterisks (* p < 0.05; ** p < 0.01).
the new equilibrium values in about 5-6h. After 24h, normoxic conditions were reestablished and maintained for additional 24h. Some control fish (N = 3) were kept in normoxic water throughout the experiment. Blood samples (0.6 ml) were taken through the dorsal aortic catheter at - lh (normoxic control sample), at 3, 6 and 24h (hypoxia or hypoxiahypercapnia exposure) and at 26 and 48h (normoxic return). Ventilation and heart rates were monitored continuously and are illustrated as the means of the average individual rates during 10 min of measurement around every full hour. The time intervals between breaths and between heart beats were recorded in the hour before t = 0 and after 23h of hypoxia or hypoxia-hypercapnia exposure. About 1000 intervals (15-30 min of recording) were used for analysis in each case. Arterial Po2 and pH were measured at 14°C with a Radiometer (Copenhagen, Denmark) BMS 3 electrode assembly connected to a PHM 73 pH/ blood gas monitor. Plasma total C02 (CT) was measured by the Cameron (1971) method and arterial Po 0 2 was calculated as
104 8.2
N
N
H/H+H
..
30
I
l iI
··
I I
8
a. 20 7.8
E
E
7.6 r I*
7.4
* a hypoxia
I
7.2
,
,
10
0
hyp.oxi.hyPerCAPnIa
0
i
,
0
I
4
10 E
8
o 16 E E
E
E
2
54 u
0
v
-1
3 6
2426 time (h)
-
48
time (h)
Fig. 3. Changes in arterial blood pH, arterial carbon dioxide tension, plasma bicarbonate concentration and plasma lactate concentration in rainbow trout upon exposure to hypoxia or hypoxia-hypercapnia and upon return to normoxia. Symbols as in Fig. 2.
Pco 2 = CT/CCO2 (1O(pH-pK') + 1)
using the values of pK' and aCO 2 for rainbow trout (Boutilier et al. 1984). Plasma bicarbonate concentration was calculated as CT- acO2 PCO2 . Plasma lactate was assessed by the BoehringerMannheim lactate dehydrogenase method.
Results and discussion Arterial P o 2 and acid-base status Exposure to hypoxia and to hypoxia-hypercapnia gradually decreased the arterial oxygen tension from values around 90 mmHg to values slightly above 30 mmHg within 6h, whereafter Po2 stayed constant up to 24h. Upon return to normoxia, arterial P 0 2 returned to values around 90 mmHg (Fig. 2). Acid-base changes were small during exposure to hypoxia. Arterial Po0 2 decreased slightly but the respiratory alkalosis expected from the Po 2 decrease was countered by a minor metabolic acidosis component from lactic acid release, leaving pH unchanged whereas plasma bicarbonate decreased
(Fig. 3). Exposure to combined hypoxia and hypercapnia produced more profound acid-base changes. Arterial Po0 2 rose from below 3 mmHg to about 6.5 mmHg in 6h, which caused a significant predominantly respiratory acidosis (Fig. 3). This acidosis was effectively compensated by accumulation of bicarbonate in the plasma to values twice as high as in normoxia, restoring pH to control values within 24h (Fig. 3). When the fish were returned to normoxic conditions, arterial PCO2 quickly decreased and plasma [HCO 3 -] was rapidly reduced. Thus, bicarbonate was only transiently elevated above normoxic control values, causing an only small and transient metabolic alkalosis (Fig. 3). It appears, that the rate of acid-base compensation upon exposure to hypoxia-hypercapnia and upon return to normoxia is faster in rainbow trout (Fig. 3; Thomas 1983) than in tench, which is a fish species that is very tolerant to hypoxic-hypercapnic waters (Jensen and Weber 1982, 1985b). Water quality is known to influence the rate of acid-base compensation (e.g., Heisler 1993), but the present study on trout was conducted in water that had the same ionic composition as used in the studies on tench. Thus, the faster pH compensation in rainbow trout than in tench is more likely to result from
105 UU -
N
H/H+H
N
r
E 50 -
100 -
b
__
__I
___
E
E
50 -
> J V_ L) V V__A
-1
6
12
18
24
30
36
42
48
time (h)
1 S Fig. 4. Oscilloscope traces of simultaneous recordings of (a) heart activity by the external electrodes (upper trace) and by dorsal aortic blood pressure changes (lower trace) and (b) ventilatory activity by the external electrodes ower trace) and by pressure changes in a catheter to the operculum chamber (upper trace).
species differences in the rate of bicarbonate gain via branchial ion transfer mechanisms and from the higher arterial control pH in tench, which demands accumulation of larger amounts of bicarbonate in order to restore pH in hypoxia-hypercapnia (Jensen and Weber 1982, 1985b). The respiratory acidosis in hypoxic-hypercapnic trout tends to reduce blood 02 affinity (via the Bohr effect), decreasing the arterial 02 saturation to values lower than in hypoxic trout, which may impair tissue 02 delivery and stimulate lactic acid
Fig. 5. Heart rate (fh) and ventilation rate (fr) in normoxic rainbow trout exposed to hypoxia and to hypoxia-hypercapnia and subsequently returned to normoxia. The dashed line depicts heart rate and ventilation rate in fish kept in normoxic water. Symbols as in Fig. 2.
production. In agreement with this possibility, plasma lactate increased to higher values in hypoxichypercapnic trout than in hypoxic trout (Fig. 3). Absolute values were, however, low. It is possible that -adrenergic red cell responses caused selective protection of red cell pH (Nikinmaa 1990), which limited adverse effects on blood 02 affinity.
Ventilatory and cardiacactivities The measurement of ventilatory and cardiac activities via the external electrode system was validated by comparison with simultaneous recordings of ventilatory and heart activities via pressure transducers connected to an operculum catheter (recording pressure changes caused by breathing) and to
106 the dorsal aortic catheter (recording changes in blood pressure). As shown in Fig. 4, the measurements of heart rate and ventilation rate via the two methods matched each other. The blood pressure pulse associated with heart contraction occurred slightly later than the pulse in the electrocardiogram, reflecting the time required for the pressure pulse to travel through the blood vessels and catheter. Normoxic rainbow trout had a ventilation frequency (fr) of about 70 breaths/min. The frequency increased significantly during exposure to both hypoxia and combined hypoxia-hypercapnia (Fig. 5). Constant levels were reached after approximately 6h when Po2 had reached stable low values (Fig. 2). The mean value of fr,, using all data in the time period 6-24h, was significantly higher (p < 0.01, unpaired Students t-test) in hypoxic-hypercapnic trout (90.4 min-1) than in hypoxic trout (85.0 min-1). Ventilation in fish is primarily controlled by 02 conditions, leading to a rapid hyperventilation in hypoxia (Dejours 1973), but recent studies have suggested that Pco 2 and acid-base status also exerts a regulatory influence (Thomas et al. 1983; Heisler et al. 1988; Perry and Wood 1989). The higher ventilation frequency in hypoxia-hypercapnia than in hypoxia could be caused by a stimulatory effect of high PC0 2 and low pH on ventilation. A firm conclusion regarding the ventilatory minute volume cannot be made, however, since only ventilatory frequency and not stroke volume was measured. Additionally, the lowering of pH may have tended to reduce arterial 02 content more in hypoxic-hypercapnic trout than in hypoxic trout, which could have caused a stronger 02 mediated drive for ventilation. It is of interest, however, that pH recovered towards control values in hypoxic-hypercapnic trout (Fig. 3) without causing a reduction in fr (Fig. 5), which may reflect a stimulatory effect of CO 2. The heart rate (fh) in normoxia was about 50 min -1, which compares well with the resting value reported by Priede (1974) at the same temperature. Upon exposure to both hypoxia and hypoxiahypercapnia the heart rate increased for some 10-12h, whereupon fh decreased (Fig. 5). At the end of the exposure period, however, the heart rate
again increased, and this increase continued into the subsequent period with re-established normoxia beforefh again decreased. A similar cyclic variability in heart rate was observed in fish kept in normoxic water but absolutefh values were lower (Fig. 5). This suggests, that on top of a tachycardia response to hypoxia, heart rate was governed by circadian rhythms, with increasing values offh during the day and decreasing values during the night. A similar observation of higher heart rates during the day than during the night was previously noted in rainbow trout (de Vera and Priede 1991). It is likely, that the circadian rhythm in heart rate reflects an adjustment of cardiac output to a higher activity level and 0 2-consumption during the day than during the night. The elevation of heart rate during hypoxia (Fig. 5) contrasts with earlier studies on rainbow trout, which have shown a decrease fh when Po was lowered below a critical point (Holeton andRandall 1967; Randall and Smith 1967; Smith and Jones 1978). Bradycardia was accompanied by an increased stroke volume, whereby cardiac output was unchanged (Holeton and Randall 1967). A hypoxic bradycardia has been reported in a number of fish species, including lingcod (Farrell 1982), spangled perch (Gehrke and Fielder 1988), tunas (Bushnell et al. 1990), shorthorn sculpin, eel-pout (Fritsche 1990) and Atlantic cod (Fritsche and Nilsson 1993) but other species have been reported to show an unchanged heart rate (e.g., the giant goby and the five-bearded rockling) (Berschick et al. 1987; Fritsche 1990) or an increased fh (e.g., winter flounder and carp) (Cech et al. 1977; Glass et al. 1991). In carp, heart rate increases down to a water Po 2 of 30 mmHg, then it decreases but absolute values remain higher than in normoxia down to 5 mmHg (Glass et al. 1991). In many previous studies, fish were acutely and short-term exposed to severe hypoxia. In the present study, water Po2 was changed slowly (i.e., within 6h) towards a value of 60 mmHg, where it was kept for further 18h. The relatively slow change in P2, together with the relatively moderate degree of hypoxia, may be influential for the tachycardia observed here. Indeed, in a few cases where water Po 2 was allowed to decrease below 60 mmHg (down to 40 mmHg), the
107 normoxia
15 ------
il
hypoxia
II 10 -
L
C 0 ED
5-
.LL II
,.
L'
0 700
900
1100
1300
1500
1700
1200
1400
Int(H) (ms) 25
20
,, 15 U c
a
10
U.
400
600
800
1000
Int(R) (ms) Fig. 6. Frequency distributions of the time interval between heart beats, Int(H), and the time interval between breaths, Int(R), in a rainbow trout in the normoxic control situation and after 23h exposure to hypoxia (water Po2 = 60 mmHg). Approximately 1000 intervals were used for analysis in each case. The data were grouped into size classes of 10 ms duration. The number of observations within each size class is given as a frequency in percent of the total number of observations. Note that the frequency distribution of Int(H) is much more narrow in hypoxia than in normoxia. See text for further details.
tachycardia was succeeded by a bradycardia, which could be reversed by reoxygenation of the water (not illustrated). Additionally, the normoxic control heart rate in the present study (around 50 beats/min) was lower than in the study of Holeton and Randall (1967) on hypoxia in rainbow trout at the same temperature (around 70 beats/min), which may have given the present trout the potential for increasing heart rate in hypoxia. The low control fh in the present study may relate to the
long recovery time used (72h following operation), since we have observed a decline infh for up to 48h after insertion of dorsal aortic chateters under anaesthesia (unpublished). The time interval between heart beats, Int(H), varied considerably in individual normoxic fish but exposure to hypoxia reduced this variability. Thus, the typical distribution of Int(H) was broad with a maximum around 1250 ms in the normoxic control situation at t = - lh, whereas the distribution was narrowed and centered around lower Int(H) values (reflecting the elevated mean heart rate) after 23h of hypoxia (Fig. 6). Such changes in the heart rate variability may provide important information on regulation of cardiac activity. The heart rate in fish is modulated by cholinergic and adrenergic fibers and by circulating catecholamines. Inhibitory cholinergic influences often exceeds stimulatory adrenergic influences, whereby the resting heart rate is lower than the intrinsic pacemaker rate (Farrell 1991). The heart rate variability appears to originate in the opposite influences of cholinergic and adrenergic stimulation (de Vera and Priede 1991), which is supported by the disappearance of the variation by section of the vagus nerves (Priede 1974). Thus, the broad distribution of Int(H) in normoxia may reflect the opposing influences of a simultaneous cholinergic and adrenergic innervation of the heart, whereas the narrow distribution in hypoxia may reflect a reduced (or absent) cholinergic tonus on the heart, which (perhaps aided by an increased adrenergic tonus) increases heart rate and decreases the variation in the beat-to-beat time interval. During normoxia, the time intervals between breaths, Int(R), varied less than time intervals between heart beats. In hypoxia, the Int(R) distribution midpoint moved towards lower values (elevatingf,) but the width of the distribution was not significantly affected (Fig. 6). The possibility of cardio-respiratory synchronization was investigated by calculating the frequency distribution for the ratio between the heart beat-tobeat interval and the breath-to-breath interval, Int(H)/Int(R). If synchronization between heart beats and breaths occurred, this distribution should be expected to show nodes around the integers, with ideal coupling at an Int(H)/Int(R) ratio of 1. Ear-
108
r
LL
0
20 -
hypoxia
0 C e 10
L
0
,, I
hypercapnia+hypoxia
0 Cgo 10 o
0
1
2
3
Int(H)/Int(R) Fig. 7. Frequency distributions for the Int(H)/Int(R) ratio in normoxic, hypoxic (23h) and hypoxic-hypercapnic (23h) rainbow trout. Note that the distribution midpoints are not centered around the integers as would have been the case if cardiac and ventilation rhythms had been tightly coupled. See text for further details.
lier reports have suggested synchronization between heart beats and ventilation movements in hypoxic but not normoxic rainbow trout (Randall and Smith 1967). In the present study the midpoint of the Int(H)/Int(R) frequency distribution was between 1 and 2 in normoxia as well as in hypoxia
(Fig. 7). The same was the case in hypoxia-hypercapnia but an additional peak was present around the integer 2 (Fig. 7). Thus, whereas a minor degree of synchronization may have occurred, a general and tight coupling of respiratory and cardiac rhythms was not present. The pulsatile water and blood flows associated with cyclic ventilatory and cardiac pumping may reduce the efficiency of gas exchange across the gills. The degree of inefficiency depends on the physical dimensions of the gill and the gas capacitance coefficents of water and blood (Malte 1992). Theoretical analysis employing the gill dimensions of rainbow trout suggests that no significant improvement of 02 exchange efficiency would be obtained by a coupling between the heart and ventilatory frequencies (Malte 1992). The demonstrated absence of a tight synchronization in rainbow trout (Fig. 7) is in line with this prediction. In other fish species (including carp) cardio-respiratory synchronization can have a positive effect on the gas exchange efficiency (Malte 1992). Synchronization between heart beats and respiratory movements was recently demonstrated in hypoxic carp (Glass et al. 1991). The present study exemplifies that bioelectric signals associated with breathing and heart contractions can be simultaneously monitored by electrodes in the water outside the fish, which provide a method for non-invasive recordings. In the present case, the monitoring system was used on rainbow trout with indwelling dorsal aorta catheters. Cannulated fish was used in order to validate the electrode measurements (cf. Fig. 4) and in order to enable investigation of acid-base regulatory responses in parallel with the measurements of cardio-respiratory changes. The capability of the monitoring system to measure cardiac and ventilatory activities non-invasively will be particularly useful in studies where minimal disturbance of the fish is crucial and in studies where even minor surgical procedures are undesirable. The system has the drawback that it does not provide direct information on stroke volumes. The noticeable advantages of the system are, however, that it can monitor fr and fh or Int(R) and Int(H) non-invasively and continuously for indefinite periods. The data acquired is stored on disk and can be read into vari-
109 ous soft-ware programmes for further analysis. Computer data-processing of long-term data recordings may prove to be essential in future studies on topics such as circadian rhythms, heart rate variability or cardio-respiratory coupling. Furthermore, given that changes in water quality influences ventilatory and cardiac activities in fish, the system may be used to assess biological effects of pollutants and to judge the physiological state of fish in aquaculture systems.
Acknowledgements We thank Torben Andreasen for help with constructing and fine-tuning the electronics. The work was supported by the Danish Natural Science Research Council (11-9659-1).
References cited Berschick, P., Bridges, C.R. and Grieshaber, M.K. 1987. The influence of hyperoxia, hypoxia and temperature on the respiratory physiology of the intertidal rockpool fish Gobius cobitis Pallas. J. Exp. Biol. 130: 369-387. Boutilier, R.G., Heming, T. and Iwama, G. 1984. Physiochemical parameters for use in fish respiratory physiology. In Fish Physiology. Vol. XA, pp. 403-430. Edited by W.S. Hoar and D.J. Randall. Academic Press, New York. Bushnell, P.G., Brill, R.W. and Bourke, R.E. 1990. Cardiorespiratory responses of skipjack tuna (Katsuwonus pelamis), yellowfin tuna (Thunnus albacares), and bigeye tuna (Thunnus obesus) to acute reductions of ambient oxygen. Can. J. Zool. 68: 1857-1865. Cameron, J.N. 1971. Rapid method for determination of total carbon dioxide in small blood samples. J. Appl. Physiol. 31: 632-634. Cech, J.J. Jr., Rowell, D.M. and Glasgow, J.S. 1977. Cardiovascular responses of the winter flounder Pseudopleuronectes americanus to hypoxia. Comp. Biochem. Physiol. 57A: 123-125. De Vera, L. and Priede, I.G. 1991. The heart rate variability signal in rainbow trout (Oncorhynchus mykiss). J. Exp. Biol. 156: 611-617. Dejours, P. 1973. Problems of control of breathing in fishes. In Comparative Physiology. pp. 117-133. Edited by L. Bolis, K. Schmidt-Nielsen and S.H.P. Maddrel. North-Holland, Amsterdam. Depledge, M.H. and Andersen, B.B. 1990. A computer-aided physiological monitoring system for continuous, long-term
recording of cardiac activity in selected invertebrates. Comp. Biochem. Physiol. 96A: 474-477. Farrell, A.P. 1982. Cardiovascular changes in the unanaesthetized lingcod (Ophiodon elongatus) during short-term, progressive hypoxia and spontaneous activity. Can. J. Zool. 60: 933-941. Farrell, A.P. 1991. From hagfish to tuna: a perspective on cardiac function in fish. Physiol. Zool. 64: 1137-1164. Fritsche, R. 1990. Effects of hypoxia on blood pressure and heart rate in three marine teleosts. Fish Physiol. Biochem. 8: 85-92. Fritsche, R. and Nilsson, S. 1993. Cardiovascular and ventilatory control during hypoxia. In Fish Ecophysiology. pp. 180206. Edited by J.C. Rankin and F.B. Jensen. Chapman and Hall, London. Gehrke, P.C. and Fielder, D.R. 1988. Effects of temperature and dissolved oxygen on heart rate, ventilation rate and oxygen consumption of spangled perch, Leipotherapon unicolor (Giinther 1859), (Percoidei, Teraponidae). J. Comp. Physiol. 157B: 771-782. Glass, M.L., Rantin, F.T., Verzola, R.M.M., Fernandes, M.N. and Kalinin, A.L. 1991. Cardio-respiratory synchronization and myocardial function in hypoxic carp, Cyprinus carpio L. J. Fish Biol. 39: 143-149. Heisler, N. (1993). Acid-base regulation in response to changes of the environment: characteristics and capacity. In Fish Ecophysiology. pp. 207-230. Edited by J.C. Rankin and F.B. Jensen. Chapman and Hall, London. Heisler, N., Toews, D.P. and Holeton, G.F. 1988. Regulation of ventilation and acid-base status in the elasmobranch Scyliorhinus stellaris during hypoxia-induced hypercapnia. Respir. Physiol. 71: 227-246. Holeton, G.F. and Randall, D.J. 1967. The effect of hypoxia upon the partial pressure of gases in the blood and water afferent and efferent to the gills of rainbow trout. J. Exp. Biol. 46: 317-327. Jensen, F.B., Nikinmaa, M. and Weber, R.E. 1993. Environmental perturbations of oxygen transport in teleost fishes: causes, consequences and compensations. In Fish Ecophysiology. pp. 161-179. Edited by J.C. Rankin and F.B. Jensen. Chapman and Hall, London. Jensen, F.B. and Weber, R.E. 1982. Respiratory properties of tench blood and hemoglobin. Adaptation to hypoxichypercapnic water. Mol. Physiol. 2: 235-250. Jensen, F.B. and Weber, R.E. 1985a. Kinetics of the acclimational responses of tench to combined hypoxia and hypercapnia. I. Respiratory responses. J. Comp. Physiol. 156B: 197203. Jensen, F.B. and Weber, R.E. 1985b. Kinetics of the acclimational responses of tench to combined hypoxia and hypercapnia. II. Extra- and intracellular acid-base status in the blood. J. Comp. Physiol. 156B: 205-211. Nikinmaa, M. 1990. Vertebrate Red Blood Cells. Adaptations of Function to Respiratory Requirements. Springer Verlag, Berlin. Malte, H. 1992. Effect of pulsatile flow on gas exchange in the
110 fish gill: theory and experimental data. Respir. Physiol. 88: 51-62. Perry, S.F. and Wood, C.M. 1989. Control and coordination of gas transfer in fishes. Can. J. Zool. 67: 2961-2970. Priede, I.G. 1974. The effect of swimming activity and section of the vagus nerves on heart rate in rainbow trout. J. Exp. Biol. 60: 305-319. Randall, D.J. and Daxboeck, C. 1984. Oxygen and carbon dioxide transfer across fish gills. In Fish Physiology. Vol. Xa, pp. 263-314. Edited by W.S. Hoar and D.J. Randall. Academic Press, New York. Randall, D.J. and Smith, J.C. 1967. The regulation of cardiac activity in fish in a hypoxic environment. Physiol. Zool. 40: 104-113.
Smith, F.M. and Jones, D.R. 1978. Localization of receptors causing hypoxic bradycardia in trout (Salmo gairdneri). Can. J. Zool. 56: 1260-1265. Soivio, A., Nyholm, K. and Westman, K. 1975. A technique for repeated sampling of the blood of individual resting fish. J. Exp. Biol. 63: 207-217. Thomas, S. 1983. Changes in blood acid-base balance in trout (Salmo gairdneri Richardson) following exposure to combined hypoxia and hypercapnia. J. Comp. Physiol. 152: 5357. Thomas, S., Fievet, B., Barthelemy, L. and Peyraud, C. 1983. Comparison of the effects of exogenous and endogenous hypercapnia on ventilation and oxygen uptake in the rainbow trout (Salmo gairdneriR.). J. Comp. Physiol. 151: 185-190.
Cardiac activity, ventilation rate and acid-base regulation in rainbow trout exposed to hypoxia and combined hypoxia and hypercapnia Kristian Borch, Frank B. Jensen and Bent B. Andersen Institute of Biology, Odense University, DK-5230 Odense M, Denmark
Accepted: January 29, 1993 Keywords: heart rate, heart rate variability, ventilation, acid-base status, hypoxia, hypoxia-hypercapnia
Abstract A computerised system for non-invasive monitoring of heart and ventilation rates and the time intervals between heart beats and between breaths was developed and used to investigate cardio-respiratory changes in rainbow trout exposed to hypoxia and to combined hypoxia and hypercapnia. Upon exposure to hypoxia and hypoxia-hypercapnia the arterial 02 tension decreased from about 90 mmHg to about 30 mmHg. Acid-base changes were small in hypoxia whereas exposure to combined hypoxia-hypercapnia caused a large extracellular respiratory acidosis. This acidosis was completely compensated within 24h by accumulation of bicarbonate in plasma to concentrations twice the normoxic values. The ventilation rate was increased to higher values in hypoxic-hypercapnic trout than in hypoxic trout. In contrast to previous reports, the heart rate increased in hypoxia. On top of the tachycardia response to hypoxia, the heart rate was governed by circadian rhythms, with higher heart rates during the day than during the night. The time interval between heart beats varied considerably in normoxic fish. Hypoxia strongly reduced this variability, which may originate in a reduced cholinergic tone to the heart. The width of the frequency distribution of the time intervals between breaths was not affected by hypoxia. The degree of cardio-respiratory synchronization was low in both normoxic and in hypoxic and hypoxic-hypercapnic trout.
Introduction Environmental hypoxia induces physiological adjustments at several steps in the oxygen transport cascade in order to safeguard respiratory function in fishes (Jensen et al. 1993). The ventilatory flow rate increases immediately by elevations in both ventilatory frequency and stroke volume, and the gill diffusion conductance is improved (Randall and Daxboeck 1984). The blood convective conductance can be increased both by an elevation of
cardiac output and by an increased blood 02 capacitance coefficient. However, whereas the 02 capacitance coefficient increases in hypoxia (Jensen et al. 1993), cardiac output is typically considered to be unchanged (Randall and Daxboeck 1984). It appears, that in many fish species heart rate is reduced by hypoxia whereas stroke volume increases, leaving cardiac output unchanged (Randall and Daxboeck 1984; Fritsche and Nilsson 1993). Examples of an unchanged heart rate or an increased heart rate and/or cardiac output are, however, also
Correspondence to: Dr. F.B. Jensen, Institute of Biology, Odense University, DK-5230 Odense M, Denmark.
102
Fig. 1. Scheme of experimental set-up. GMP, gas mixing pump. A, amplifier. F1, F2, electronic filters. C, comparator. V, monostable multi-vibrator. A/D, multi channel A/D converter. PC, IBM compatible PC. See text for details.
available (Cech et al. 1977; Berschick et al. 1987; Glass et al. 1991). Oxygen depleted aquatic habitats often tend to be both hypoxic and hypercapnic, since the processes that consume the water 02 (microbial degradation of organic matter) also produce CO 2. In contrast to the many studies on hypoxia in fishes, only few studies have addresed the physiological consequences of combined hypoxia and hypercapnia (Jensen and Weber 1982, 1985a, b; Thomas 1983). In comparison with fish exposed to hypoxia alone, fish exposed to combined hypoxia and hypercapnia faces the additional problem of an elevated CO 2 tension, which causes a respiratory acidosis. This respiratory acidosis raises the requirement for acidbase regulation and may further compromise blood 02 transport. The present study investigated acid-base changes and ventilatory and cardiac activities in rainbow trout exposed to hypoxia and to combined hypoxiahypercapnia. In order to enable measurement and analysis of heart and ventilation rate, the beat-tobeat variability in these and the degree of cardiorespiratory synchronization, a computerised system for continuous non-invasive monitoring of ventilation and heart activities was developed.
Materials and methods Animals and experimental set-up Rainbow trout (Oncorhynchus mykiss), 350-450 g in weight, were obtained from a local fish farm and kept at a 12h light: 12h dark rhythm in 5001 holding tanks with 14°C normoxic water for a minimum of 14 days before experimentation. The fish were cannulated to the dorsal aorta (Soivio et al. 1975) under 2-phenoxy-ethanol anaesthesia, whereupon the fish were transferred to a flow chamber of a water recirculating system and allowed to recover for 72h in 14°C normoxic water. The water recirculating system consisted of a reservoir aquarium from which water was pumped through the flow chamber at a rate of 901 h-1 (Fig. 1). The flow chamber was submerged in an isolated Faraday box, which minimized electric noise and shielded the animal from visual disturbance. The total water volume of the system was 401 and water was renewed daily. Water composition was: Ca + + = 3.5 mM, Na + = 1.2 mM, C1- = 1.3 mM, total CO2 = 5 mM. Normoxic (N) conditions were obtained by bubbling of air through the equilibration aquarium. Hypoxia was induced by bubbling with an air/N 2 gas mixture, and hypoxia-hypercapnia was induced by bubbling with an air/CO2/N 2 gas mixture. Gas mixtures were delivered by a Wosthoff (Bochum, FRG) gas mixing pump.
103 Collection and storage of ventilatory and cardiac activity data
--
110 -
uu-
Pl
N
90 -
A pair of stainless steel electrodes (2 mm diameter) placed at the top and bottom of the experimental fish chamber were used as recording electrodes. Two similar electrodes on each side of the chamber acted as ground electrodes (Fig. 1). The electrical signal from the fish was monitored by the electrodes and amplified one thousand times in a differential low noise AC amplifier (input resistance > 1 MD, highpassl.5 Hz, 20 dB/dec). The signal was then separated by two filters. One filter (lowpass, 5 Hz, 40 dB/dec) isolated the ventilatory signal, and the other filter (bandpass, 15-90 Hz, 80 dB/dec, 50 Hz notch) isolated the electrocardiogram (ECG) signal. The ventilatory signal was sent directly to an A/D converter. The ECG signal was sent to an A/D converter via a comparator (C), triggering a monostable multi-vibrator (V) that acted as pulse stretching circuit. The multi channel A/D converter was connected to the parallel printer port of an IBM-compatible PC (Fig. 1). A turbo pascal programme (modified from the programme of Depledge and Andersen 1990) controlled the A/D conversion via the parallel port. Counting of the ventilatory and cardiac signals was controlled by a software variable (high/low level) hysteresis discriminator for each channel. Two windows on the monitor acted as an oscilloscope, showing the ECG and ventilation signals live. The number of breaths and heart beats were stored on disk every minute, and shown in two further windows on the monitor as histograms. A separate turbo pascal programme measured and stored the time periods between heart beats and the time periods between breaths.
Experimentalprotocol and measurements After 72h of normoxic-normocapnic exposure (Po2 = 150 mmHg, P 0 2 = 0.7 mmHg) the gas supply to the equilibration aquarium was switched to either the hypoxic (Po2 = 60 mmHg) gas mixture (N = 5) or the hypoxic-hypercapnic (Po2 = 60 mmHg; Po0 2 = 4.8 mmHg) gas mixture (N= 5). This slowly changed water P 02 and Pco 2 towards
E E
70- -
O 50tL
30 -
10 -
-1
3 6
2426
48
time (h) Fig. 2. Arterial oxygen tension (means ± SEM) in rainbow trout in normoxa (N), upon exposure to hypoxia (H) (, N = 5) and to combined hypoxia and hypercapnia (H + H) ( , N = 5), and upon return to normoxia (N). The dashed line depicts the mean values in fish kept in normoxic water (N = 3) throughout the experiment. A significant difference (paired Students t-test) compared to normoxic control values (t = - Ih) is depicted by asterisks (* p < 0.05; ** p < 0.01).
the new equilibrium values in about 5-6h. After 24h, normoxic conditions were reestablished and maintained for additional 24h. Some control fish (N = 3) were kept in normoxic water throughout the experiment. Blood samples (0.6 ml) were taken through the dorsal aortic catheter at - lh (normoxic control sample), at 3, 6 and 24h (hypoxia or hypoxiahypercapnia exposure) and at 26 and 48h (normoxic return). Ventilation and heart rates were monitored continuously and are illustrated as the means of the average individual rates during 10 min of measurement around every full hour. The time intervals between breaths and between heart beats were recorded in the hour before t = 0 and after 23h of hypoxia or hypoxia-hypercapnia exposure. About 1000 intervals (15-30 min of recording) were used for analysis in each case. Arterial Po2 and pH were measured at 14°C with a Radiometer (Copenhagen, Denmark) BMS 3 electrode assembly connected to a PHM 73 pH/ blood gas monitor. Plasma total C02 (CT) was measured by the Cameron (1971) method and arterial Po 0 2 was calculated as
104 8.2
N
N
H/H+H
..
30
I
l iI
··
I I
8
a. 20 7.8
E
E
7.6 r I*
7.4
* a hypoxia
I
7.2
,
,
10
0
hyp.oxi.hyPerCAPnIa
0
i
,
0
I
4
10 E
8
o 16 E E
E
E
2
54 u
0
v
-1
3 6
2426 time (h)
-
48
time (h)
Fig. 3. Changes in arterial blood pH, arterial carbon dioxide tension, plasma bicarbonate concentration and plasma lactate concentration in rainbow trout upon exposure to hypoxia or hypoxia-hypercapnia and upon return to normoxia. Symbols as in Fig. 2.
Pco 2 = CT/CCO2 (1O(pH-pK') + 1)
using the values of pK' and aCO 2 for rainbow trout (Boutilier et al. 1984). Plasma bicarbonate concentration was calculated as CT- acO2 PCO2 . Plasma lactate was assessed by the BoehringerMannheim lactate dehydrogenase method.
Results and discussion Arterial P o 2 and acid-base status Exposure to hypoxia and to hypoxia-hypercapnia gradually decreased the arterial oxygen tension from values around 90 mmHg to values slightly above 30 mmHg within 6h, whereafter Po2 stayed constant up to 24h. Upon return to normoxia, arterial P 0 2 returned to values around 90 mmHg (Fig. 2). Acid-base changes were small during exposure to hypoxia. Arterial Po0 2 decreased slightly but the respiratory alkalosis expected from the Po 2 decrease was countered by a minor metabolic acidosis component from lactic acid release, leaving pH unchanged whereas plasma bicarbonate decreased
(Fig. 3). Exposure to combined hypoxia and hypercapnia produced more profound acid-base changes. Arterial Po0 2 rose from below 3 mmHg to about 6.5 mmHg in 6h, which caused a significant predominantly respiratory acidosis (Fig. 3). This acidosis was effectively compensated by accumulation of bicarbonate in the plasma to values twice as high as in normoxia, restoring pH to control values within 24h (Fig. 3). When the fish were returned to normoxic conditions, arterial PCO2 quickly decreased and plasma [HCO 3 -] was rapidly reduced. Thus, bicarbonate was only transiently elevated above normoxic control values, causing an only small and transient metabolic alkalosis (Fig. 3). It appears, that the rate of acid-base compensation upon exposure to hypoxia-hypercapnia and upon return to normoxia is faster in rainbow trout (Fig. 3; Thomas 1983) than in tench, which is a fish species that is very tolerant to hypoxic-hypercapnic waters (Jensen and Weber 1982, 1985b). Water quality is known to influence the rate of acid-base compensation (e.g., Heisler 1993), but the present study on trout was conducted in water that had the same ionic composition as used in the studies on tench. Thus, the faster pH compensation in rainbow trout than in tench is more likely to result from
105 UU -
N
H/H+H
N
r
E 50 -
100 -
b
__
__I
___
E
E
50 -
> J V_ L) V V__A
-1
6
12
18
24
30
36
42
48
time (h)
1 S Fig. 4. Oscilloscope traces of simultaneous recordings of (a) heart activity by the external electrodes (upper trace) and by dorsal aortic blood pressure changes (lower trace) and (b) ventilatory activity by the external electrodes ower trace) and by pressure changes in a catheter to the operculum chamber (upper trace).
species differences in the rate of bicarbonate gain via branchial ion transfer mechanisms and from the higher arterial control pH in tench, which demands accumulation of larger amounts of bicarbonate in order to restore pH in hypoxia-hypercapnia (Jensen and Weber 1982, 1985b). The respiratory acidosis in hypoxic-hypercapnic trout tends to reduce blood 02 affinity (via the Bohr effect), decreasing the arterial 02 saturation to values lower than in hypoxic trout, which may impair tissue 02 delivery and stimulate lactic acid
Fig. 5. Heart rate (fh) and ventilation rate (fr) in normoxic rainbow trout exposed to hypoxia and to hypoxia-hypercapnia and subsequently returned to normoxia. The dashed line depicts heart rate and ventilation rate in fish kept in normoxic water. Symbols as in Fig. 2.
production. In agreement with this possibility, plasma lactate increased to higher values in hypoxichypercapnic trout than in hypoxic trout (Fig. 3). Absolute values were, however, low. It is possible that -adrenergic red cell responses caused selective protection of red cell pH (Nikinmaa 1990), which limited adverse effects on blood 02 affinity.
Ventilatory and cardiacactivities The measurement of ventilatory and cardiac activities via the external electrode system was validated by comparison with simultaneous recordings of ventilatory and heart activities via pressure transducers connected to an operculum catheter (recording pressure changes caused by breathing) and to
106 the dorsal aortic catheter (recording changes in blood pressure). As shown in Fig. 4, the measurements of heart rate and ventilation rate via the two methods matched each other. The blood pressure pulse associated with heart contraction occurred slightly later than the pulse in the electrocardiogram, reflecting the time required for the pressure pulse to travel through the blood vessels and catheter. Normoxic rainbow trout had a ventilation frequency (fr) of about 70 breaths/min. The frequency increased significantly during exposure to both hypoxia and combined hypoxia-hypercapnia (Fig. 5). Constant levels were reached after approximately 6h when Po2 had reached stable low values (Fig. 2). The mean value of fr,, using all data in the time period 6-24h, was significantly higher (p < 0.01, unpaired Students t-test) in hypoxic-hypercapnic trout (90.4 min-1) than in hypoxic trout (85.0 min-1). Ventilation in fish is primarily controlled by 02 conditions, leading to a rapid hyperventilation in hypoxia (Dejours 1973), but recent studies have suggested that Pco 2 and acid-base status also exerts a regulatory influence (Thomas et al. 1983; Heisler et al. 1988; Perry and Wood 1989). The higher ventilation frequency in hypoxia-hypercapnia than in hypoxia could be caused by a stimulatory effect of high PC0 2 and low pH on ventilation. A firm conclusion regarding the ventilatory minute volume cannot be made, however, since only ventilatory frequency and not stroke volume was measured. Additionally, the lowering of pH may have tended to reduce arterial 02 content more in hypoxic-hypercapnic trout than in hypoxic trout, which could have caused a stronger 02 mediated drive for ventilation. It is of interest, however, that pH recovered towards control values in hypoxic-hypercapnic trout (Fig. 3) without causing a reduction in fr (Fig. 5), which may reflect a stimulatory effect of CO 2. The heart rate (fh) in normoxia was about 50 min -1, which compares well with the resting value reported by Priede (1974) at the same temperature. Upon exposure to both hypoxia and hypoxiahypercapnia the heart rate increased for some 10-12h, whereupon fh decreased (Fig. 5). At the end of the exposure period, however, the heart rate
again increased, and this increase continued into the subsequent period with re-established normoxia beforefh again decreased. A similar cyclic variability in heart rate was observed in fish kept in normoxic water but absolutefh values were lower (Fig. 5). This suggests, that on top of a tachycardia response to hypoxia, heart rate was governed by circadian rhythms, with increasing values offh during the day and decreasing values during the night. A similar observation of higher heart rates during the day than during the night was previously noted in rainbow trout (de Vera and Priede 1991). It is likely, that the circadian rhythm in heart rate reflects an adjustment of cardiac output to a higher activity level and 0 2-consumption during the day than during the night. The elevation of heart rate during hypoxia (Fig. 5) contrasts with earlier studies on rainbow trout, which have shown a decrease fh when Po was lowered below a critical point (Holeton andRandall 1967; Randall and Smith 1967; Smith and Jones 1978). Bradycardia was accompanied by an increased stroke volume, whereby cardiac output was unchanged (Holeton and Randall 1967). A hypoxic bradycardia has been reported in a number of fish species, including lingcod (Farrell 1982), spangled perch (Gehrke and Fielder 1988), tunas (Bushnell et al. 1990), shorthorn sculpin, eel-pout (Fritsche 1990) and Atlantic cod (Fritsche and Nilsson 1993) but other species have been reported to show an unchanged heart rate (e.g., the giant goby and the five-bearded rockling) (Berschick et al. 1987; Fritsche 1990) or an increased fh (e.g., winter flounder and carp) (Cech et al. 1977; Glass et al. 1991). In carp, heart rate increases down to a water Po 2 of 30 mmHg, then it decreases but absolute values remain higher than in normoxia down to 5 mmHg (Glass et al. 1991). In many previous studies, fish were acutely and short-term exposed to severe hypoxia. In the present study, water Po2 was changed slowly (i.e., within 6h) towards a value of 60 mmHg, where it was kept for further 18h. The relatively slow change in P2, together with the relatively moderate degree of hypoxia, may be influential for the tachycardia observed here. Indeed, in a few cases where water Po 2 was allowed to decrease below 60 mmHg (down to 40 mmHg), the
107 normoxia
15 ------
il
hypoxia
II 10 -
L
C 0 ED
5-
.LL II
,.
L'
0 700
900
1100
1300
1500
1700
1200
1400
Int(H) (ms) 25
20
,, 15 U c
a
10
U.
400
600
800
1000
Int(R) (ms) Fig. 6. Frequency distributions of the time interval between heart beats, Int(H), and the time interval between breaths, Int(R), in a rainbow trout in the normoxic control situation and after 23h exposure to hypoxia (water Po2 = 60 mmHg). Approximately 1000 intervals were used for analysis in each case. The data were grouped into size classes of 10 ms duration. The number of observations within each size class is given as a frequency in percent of the total number of observations. Note that the frequency distribution of Int(H) is much more narrow in hypoxia than in normoxia. See text for further details.
tachycardia was succeeded by a bradycardia, which could be reversed by reoxygenation of the water (not illustrated). Additionally, the normoxic control heart rate in the present study (around 50 beats/min) was lower than in the study of Holeton and Randall (1967) on hypoxia in rainbow trout at the same temperature (around 70 beats/min), which may have given the present trout the potential for increasing heart rate in hypoxia. The low control fh in the present study may relate to the
long recovery time used (72h following operation), since we have observed a decline infh for up to 48h after insertion of dorsal aortic chateters under anaesthesia (unpublished). The time interval between heart beats, Int(H), varied considerably in individual normoxic fish but exposure to hypoxia reduced this variability. Thus, the typical distribution of Int(H) was broad with a maximum around 1250 ms in the normoxic control situation at t = - lh, whereas the distribution was narrowed and centered around lower Int(H) values (reflecting the elevated mean heart rate) after 23h of hypoxia (Fig. 6). Such changes in the heart rate variability may provide important information on regulation of cardiac activity. The heart rate in fish is modulated by cholinergic and adrenergic fibers and by circulating catecholamines. Inhibitory cholinergic influences often exceeds stimulatory adrenergic influences, whereby the resting heart rate is lower than the intrinsic pacemaker rate (Farrell 1991). The heart rate variability appears to originate in the opposite influences of cholinergic and adrenergic stimulation (de Vera and Priede 1991), which is supported by the disappearance of the variation by section of the vagus nerves (Priede 1974). Thus, the broad distribution of Int(H) in normoxia may reflect the opposing influences of a simultaneous cholinergic and adrenergic innervation of the heart, whereas the narrow distribution in hypoxia may reflect a reduced (or absent) cholinergic tonus on the heart, which (perhaps aided by an increased adrenergic tonus) increases heart rate and decreases the variation in the beat-to-beat time interval. During normoxia, the time intervals between breaths, Int(R), varied less than time intervals between heart beats. In hypoxia, the Int(R) distribution midpoint moved towards lower values (elevatingf,) but the width of the distribution was not significantly affected (Fig. 6). The possibility of cardio-respiratory synchronization was investigated by calculating the frequency distribution for the ratio between the heart beat-tobeat interval and the breath-to-breath interval, Int(H)/Int(R). If synchronization between heart beats and breaths occurred, this distribution should be expected to show nodes around the integers, with ideal coupling at an Int(H)/Int(R) ratio of 1. Ear-
108
r
LL
0
20 -
hypoxia
0 C e 10
L
0
,, I
hypercapnia+hypoxia
0 Cgo 10 o
0
1
2
3
Int(H)/Int(R) Fig. 7. Frequency distributions for the Int(H)/Int(R) ratio in normoxic, hypoxic (23h) and hypoxic-hypercapnic (23h) rainbow trout. Note that the distribution midpoints are not centered around the integers as would have been the case if cardiac and ventilation rhythms had been tightly coupled. See text for further details.
lier reports have suggested synchronization between heart beats and ventilation movements in hypoxic but not normoxic rainbow trout (Randall and Smith 1967). In the present study the midpoint of the Int(H)/Int(R) frequency distribution was between 1 and 2 in normoxia as well as in hypoxia
(Fig. 7). The same was the case in hypoxia-hypercapnia but an additional peak was present around the integer 2 (Fig. 7). Thus, whereas a minor degree of synchronization may have occurred, a general and tight coupling of respiratory and cardiac rhythms was not present. The pulsatile water and blood flows associated with cyclic ventilatory and cardiac pumping may reduce the efficiency of gas exchange across the gills. The degree of inefficiency depends on the physical dimensions of the gill and the gas capacitance coefficents of water and blood (Malte 1992). Theoretical analysis employing the gill dimensions of rainbow trout suggests that no significant improvement of 02 exchange efficiency would be obtained by a coupling between the heart and ventilatory frequencies (Malte 1992). The demonstrated absence of a tight synchronization in rainbow trout (Fig. 7) is in line with this prediction. In other fish species (including carp) cardio-respiratory synchronization can have a positive effect on the gas exchange efficiency (Malte 1992). Synchronization between heart beats and respiratory movements was recently demonstrated in hypoxic carp (Glass et al. 1991). The present study exemplifies that bioelectric signals associated with breathing and heart contractions can be simultaneously monitored by electrodes in the water outside the fish, which provide a method for non-invasive recordings. In the present case, the monitoring system was used on rainbow trout with indwelling dorsal aorta catheters. Cannulated fish was used in order to validate the electrode measurements (cf. Fig. 4) and in order to enable investigation of acid-base regulatory responses in parallel with the measurements of cardio-respiratory changes. The capability of the monitoring system to measure cardiac and ventilatory activities non-invasively will be particularly useful in studies where minimal disturbance of the fish is crucial and in studies where even minor surgical procedures are undesirable. The system has the drawback that it does not provide direct information on stroke volumes. The noticeable advantages of the system are, however, that it can monitor fr and fh or Int(R) and Int(H) non-invasively and continuously for indefinite periods. The data acquired is stored on disk and can be read into vari-
109 ous soft-ware programmes for further analysis. Computer data-processing of long-term data recordings may prove to be essential in future studies on topics such as circadian rhythms, heart rate variability or cardio-respiratory coupling. Furthermore, given that changes in water quality influences ventilatory and cardiac activities in fish, the system may be used to assess biological effects of pollutants and to judge the physiological state of fish in aquaculture systems.
Acknowledgements We thank Torben Andreasen for help with constructing and fine-tuning the electronics. The work was supported by the Danish Natural Science Research Council (11-9659-1).
References cited Berschick, P., Bridges, C.R. and Grieshaber, M.K. 1987. The influence of hyperoxia, hypoxia and temperature on the respiratory physiology of the intertidal rockpool fish Gobius cobitis Pallas. J. Exp. Biol. 130: 369-387. Boutilier, R.G., Heming, T. and Iwama, G. 1984. Physiochemical parameters for use in fish respiratory physiology. In Fish Physiology. Vol. XA, pp. 403-430. Edited by W.S. Hoar and D.J. Randall. Academic Press, New York. Bushnell, P.G., Brill, R.W. and Bourke, R.E. 1990. Cardiorespiratory responses of skipjack tuna (Katsuwonus pelamis), yellowfin tuna (Thunnus albacares), and bigeye tuna (Thunnus obesus) to acute reductions of ambient oxygen. Can. J. Zool. 68: 1857-1865. Cameron, J.N. 1971. Rapid method for determination of total carbon dioxide in small blood samples. J. Appl. Physiol. 31: 632-634. Cech, J.J. Jr., Rowell, D.M. and Glasgow, J.S. 1977. Cardiovascular responses of the winter flounder Pseudopleuronectes americanus to hypoxia. Comp. Biochem. Physiol. 57A: 123-125. De Vera, L. and Priede, I.G. 1991. The heart rate variability signal in rainbow trout (Oncorhynchus mykiss). J. Exp. Biol. 156: 611-617. Dejours, P. 1973. Problems of control of breathing in fishes. In Comparative Physiology. pp. 117-133. Edited by L. Bolis, K. Schmidt-Nielsen and S.H.P. Maddrel. North-Holland, Amsterdam. Depledge, M.H. and Andersen, B.B. 1990. A computer-aided physiological monitoring system for continuous, long-term
recording of cardiac activity in selected invertebrates. Comp. Biochem. Physiol. 96A: 474-477. Farrell, A.P. 1982. Cardiovascular changes in the unanaesthetized lingcod (Ophiodon elongatus) during short-term, progressive hypoxia and spontaneous activity. Can. J. Zool. 60: 933-941. Farrell, A.P. 1991. From hagfish to tuna: a perspective on cardiac function in fish. Physiol. Zool. 64: 1137-1164. Fritsche, R. 1990. Effects of hypoxia on blood pressure and heart rate in three marine teleosts. Fish Physiol. Biochem. 8: 85-92. Fritsche, R. and Nilsson, S. 1993. Cardiovascular and ventilatory control during hypoxia. In Fish Ecophysiology. pp. 180206. Edited by J.C. Rankin and F.B. Jensen. Chapman and Hall, London. Gehrke, P.C. and Fielder, D.R. 1988. Effects of temperature and dissolved oxygen on heart rate, ventilation rate and oxygen consumption of spangled perch, Leipotherapon unicolor (Giinther 1859), (Percoidei, Teraponidae). J. Comp. Physiol. 157B: 771-782. Glass, M.L., Rantin, F.T., Verzola, R.M.M., Fernandes, M.N. and Kalinin, A.L. 1991. Cardio-respiratory synchronization and myocardial function in hypoxic carp, Cyprinus carpio L. J. Fish Biol. 39: 143-149. Heisler, N. (1993). Acid-base regulation in response to changes of the environment: characteristics and capacity. In Fish Ecophysiology. pp. 207-230. Edited by J.C. Rankin and F.B. Jensen. Chapman and Hall, London. Heisler, N., Toews, D.P. and Holeton, G.F. 1988. Regulation of ventilation and acid-base status in the elasmobranch Scyliorhinus stellaris during hypoxia-induced hypercapnia. Respir. Physiol. 71: 227-246. Holeton, G.F. and Randall, D.J. 1967. The effect of hypoxia upon the partial pressure of gases in the blood and water afferent and efferent to the gills of rainbow trout. J. Exp. Biol. 46: 317-327. Jensen, F.B., Nikinmaa, M. and Weber, R.E. 1993. Environmental perturbations of oxygen transport in teleost fishes: causes, consequences and compensations. In Fish Ecophysiology. pp. 161-179. Edited by J.C. Rankin and F.B. Jensen. Chapman and Hall, London. Jensen, F.B. and Weber, R.E. 1982. Respiratory properties of tench blood and hemoglobin. Adaptation to hypoxichypercapnic water. Mol. Physiol. 2: 235-250. Jensen, F.B. and Weber, R.E. 1985a. Kinetics of the acclimational responses of tench to combined hypoxia and hypercapnia. I. Respiratory responses. J. Comp. Physiol. 156B: 197203. Jensen, F.B. and Weber, R.E. 1985b. Kinetics of the acclimational responses of tench to combined hypoxia and hypercapnia. II. Extra- and intracellular acid-base status in the blood. J. Comp. Physiol. 156B: 205-211. Nikinmaa, M. 1990. Vertebrate Red Blood Cells. Adaptations of Function to Respiratory Requirements. Springer Verlag, Berlin. Malte, H. 1992. Effect of pulsatile flow on gas exchange in the
110 fish gill: theory and experimental data. Respir. Physiol. 88: 51-62. Perry, S.F. and Wood, C.M. 1989. Control and coordination of gas transfer in fishes. Can. J. Zool. 67: 2961-2970. Priede, I.G. 1974. The effect of swimming activity and section of the vagus nerves on heart rate in rainbow trout. J. Exp. Biol. 60: 305-319. Randall, D.J. and Daxboeck, C. 1984. Oxygen and carbon dioxide transfer across fish gills. In Fish Physiology. Vol. Xa, pp. 263-314. Edited by W.S. Hoar and D.J. Randall. Academic Press, New York. Randall, D.J. and Smith, J.C. 1967. The regulation of cardiac activity in fish in a hypoxic environment. Physiol. Zool. 40: 104-113.
Smith, F.M. and Jones, D.R. 1978. Localization of receptors causing hypoxic bradycardia in trout (Salmo gairdneri). Can. J. Zool. 56: 1260-1265. Soivio, A., Nyholm, K. and Westman, K. 1975. A technique for repeated sampling of the blood of individual resting fish. J. Exp. Biol. 63: 207-217. Thomas, S. 1983. Changes in blood acid-base balance in trout (Salmo gairdneri Richardson) following exposure to combined hypoxia and hypercapnia. J. Comp. Physiol. 152: 5357. Thomas, S., Fievet, B., Barthelemy, L. and Peyraud, C. 1983. Comparison of the effects of exogenous and endogenous hypercapnia on ventilation and oxygen uptake in the rainbow trout (Salmo gairdneriR.). J. Comp. Physiol. 151: 185-190.
