CO 2 transport and excretion in rainbow trout (Oncorhynchus mykiss) during graded sustained exercise

Transcription

1 Respiration Physiology 119 (2000) CO 2 transport and excretion in rainbow trout (Oncorhynchus mykiss) during graded sustained exercise C.J. Brauner a, *,1, H. Thorarensen b,2, P. Gallaugher b,3, A.P. Farrell b, D.J. Randall a a Department of Zoology, Uni ersity of British Columbia, Vancou er BC, Canada, V6T 1Z4 b Department of Biological Sciences, Simon Fraser Uni ersity, Burnaby BC, Canada, V5A 1S6 Accepted 18 October 1999 Abstract A quantitative analysis of CO 2 transport and excretion was conducted in seawater acclimated rainbow trout (Oncorhynchus mykiss) swimming at different sustained swimming velocities. CO 2 excretion increased linearly with cardiac output during exercise but arterial P CO2 (Pa CO2 ) and total CO 2 levels also increased indicating a diffusion limitation to CO 2 excretion. The elevated Pa CO2 was not accompanied by a decrease in ph, indicating that the acid base compensation was rapid. Mixed-venous P CO2 increased to a greater extent than Pa CO2 resulting in a large increase in the venous arterial difference in P CO2 (Pv CO2 Pa CO2 ). The Pv CO2 Pa CO2 difference was used to calculate the proportion of total CO 2 excreted comprised of dissolved CO 2 which accounted for less than 1% of total CO 2 excreted in fish swimming at 11 cm sec 1 but increased to about 9% at the greatest swimming velocity indicating that the pattern of CO 2 excretion changes during exercise. There was no effect of exercise on the proportion of CO 2 excreted which was dependent upon HCO 3 /Cl exchange (54%) or that which was dependent upon the dehydration of HCO 3 that resided within the red cell prior to gill blood entry (42%). The large proportion of total CO 2 excreted that was dependent upon HCO 3 /Cl exchange is significant because this is thought to be the rate limiting step in CO 2 excretion Elsevier Science B.V. All rights reserved. Keywords: Acid base balance, CO 2, fish; Carbon dioxide, transport, excretion, fish; Exchange, HCO 3 /Cl ; Exercise, CO 2 excretion, fish; Fish, rainbow trout (Oncorrychus mykiss) * Corresponding author. Tel.: ext ; fax: address: braunerc@mcmaster.ca (C.J. Brauner) 1 Present address: Department of Biology, McMaster University, Hamilton, Ontario, Canada, L8S 4L8. 2 Present address: Holar Agricultural College, 551 Saudarkrokur, Iceland. 3 Present address: Continuing studies in Science, Simon Fraser University, Burnaby, BC, Canada, V5A 1S6. 1. Introduction In fish swimming at maximal levels of sustained exercise, oxygen consumption rate can increase by 5 20-fold over resting levels (Brett, 1964; Kiceniuk and Jones, 1977; Randall and Daxboeck, 1984). There are a suite of physiological adjustments during exercise which ensure oxygen deliv /00/$ - see front matter 2000 Elsevier Science B.V. All rights reserved. PII: S (99)

2 70 C.J. Brauner et al. / Respiration Physiology 119 (2000) ery to the active tissues, including an increase in arterial venous oxygen content difference (Kiceniuk and Jones, 1977), gill ventilation and cardiac output, primarily due to an increase in stroke volume (Kiceniuk and Jones, 1977; Randall, 1982). Haematocrit and the oxygen carrying capacity of the blood have been shown to increase at fatigue (Jones and Randall, 1978; Thomas et al., 1987) and there is evidence of a graded release of red cells from the spleen with increasing swimming velocity (Gallaugher et al., 1992). The elevated capacity for gas transport in the blood is matched with adjustments at the gills. Kiceniuk and Jones (1977) measured an increase in the gill ventilation: perfusion ratio from 12 at rest, to 32 during maximal exercise, indicating that ventilation volume is elevated disproportionately, relative to changes in cardiac output. The increase in ventral aortic pressure during swimming (Kiceniuk and Jones, 1977) elevates the proportion of secondary lamellae which are perfused and reduces epithelial thickness (Randall and Daxboeck 1984), increasing the diffusing capacity of the gills. Despite the increased blood volume of the gills, the elevation in cardiac output reduces blood transit time through the gills from 3 to 1 sec (Randall, 1982), limiting the time available for the processes involved in gas exchange. All of these modifications to the cardio-respiratory system are crucial for the maintenance of elevated gas flux across the gills. Most of the studies to date examining sustained exercise in fish, have concentrated specifically on the oxygen transport system. Respiratory exchange ratios, (e.g. CO 2 excretion rate/o 2 consumption rate) have been measured during exercise in fish and they confirm that CO 2 excretion rate is elevated in accordance with oxygen consumption rate (Randall and Daxboeck, 1984; Kieffer et al., 1998). In general, however, there have been few reports of CO 2 transport and excretion during sustained exercise in fish. The diffusivity of CO 2 is times that for O 2 which has been interpreted to mean that a diffusion limitation will be seen for O 2, long before CO 2.CO 2 removal, however, is dependent on the rate of Cl /HCO 3 exchange, which has been suggested to be the rate limiting step in CO 2 excretion (Perry, 1986). The objective of this experiment was to conduct a quantitative analysis of CO 2 transport and excretion in rainbow trout, based upon arterial and venous blood samples, obtained during different levels of sustained exercise. Of particular interest was the partitioning of CO 2 transport between plasma and red cells, the pattern of CO 2 excretion and the relative role of HCO 3 /Cl exchange during different levels of sustained exercise. 2. Materials and methods 2.1. Experimental animals Rainbow trout (Oncorhynchus mykiss, weight= g; length= cm) were purchased from a local supplier (West Creek Trout Ponds, Aldergrove, BC) and acclimated to sea water [29 parts per thousand (ppt), 9 C] for at least 1 month prior to experiments. Fish were housed and experiments were conducted at the Department of Fisheries and Oceans, West Vancouver. Fish were fed to satiation bi-weekly and starved for 2 days prior to surgery. Throughout the experiments, fish were maintained in 29 ppt sea water Surgery and handling Rainbow trout were anaesthetized in a 1: solution of tricaine methanesulphonate (MS-222) in sea water, adjusted to ph 7.5 with NaHCO 3 and bubbled with air. Each fish was placed on a surgery table and the gills were continually irrigated with a more dilute anaesthetic solution (1: MS-222 in sea water). The dorsal aorta was cannulated with polyethylene tubing (Clay- Adams PE-50; internal diameter, mm; outer diameter, mm) and in some fish, the prebranchial artery of the first gill arch was cannulated (PE-50) to sample mixed-venous blood, in which case the entire cannulated gill arch was tied off. Fish that had both the arterial and venous systems cannulated will be referred to as Series I and those with only the dorsal aorta cannulated will be referred to as Series II. The ventral aorta of Series I fish were also fitted with a Transonic flow probe (Transonic, Ithaca, NY) for direct

3 C.J. Brauner et al. / Respiration Physiology 119 (2000) measurement of cardiac output as described by Thorarensen et al. (1996). Following surgery, weight and fork length were recorded and the fish was revived by irrigating the gills with aerated water. Fish were then transferred to an opaque acrylic box, and left to recover for h. Cannulae were flushed daily with heparinized (10 i.u. ml 1 ammonium heparin; Sigma), teleost saline solution. Fish were placed in a Brett-type swim tube respirometer the night before swimming experiments. During acclimation to the swim tunnel, water flow velocity was maintained at 11 cm sec 1 and a continual overflow of water prevented build up of metabolic waste products Experimental procedure Blood parameters were measured in rainbow trout at four swimming velocities: the velocity to which fish were acclimated over night, two intermediate swimming speeds, and finally, the maximal critical swimming velocity (Ucrit). At the start of each swimming trial, 0.6 ml of blood was withdrawn from the dorsal aorta and the prebranchial artery for measurement of whole blood and plasma CO 2 content (C CO2 ) and plasma and red cell ph (phe and phi, respectively). Plasma P CO2 and [HCO 3 ] were calculated from plasma C CO2 and phe by re-arrangement of the Henderson Hasselbalch equation as described below. Samples were also taken for measurement of whole blood O 2 content (C O2 ) and partial pressure (P O2 ), Hct, [Hb], methaemoglobin concentration and plasma [Cl ] but are reported in the following paper (Brauner et al. 2000). There was a maximum 5 min interval between sampling of arterial (from the dorsal aorta) and mixed-venous blood (from the pre-branchial artery) and the first source of blood sampled was chosen at random. Following blood removal, 1.2 ml of blood from a resting cannulated donor fish was injected into the swimming fish to restore blood volume to presample levels. In Series II fish, sampling was conducted only on arterial blood but the subsequent procedures were the same for both groups of fish. Upon completion of sampling, water velocity was gradually elevated by 0.66 body length (Bl) sec 1 over a 10 min period. The fish were left at the new velocity for 30 min before the sampling procedure described above was repeated. Thirty minutes is sufficient time for blood gas and acid base parameters to stabilize following a change in water velocity (Kiceniuk and Jones, 1977; Thomas et al., 1987). The change in water velocity was increased by 0.66 Bl sec 1 increments followed by the sampling regime until the fish could no longer maintain the swimming velocity. This water velocity and time to fatigue were noted for the calculation of Ucrit as described by Brett (1964) taking into consideration the solid blocking effects of the fish. The water velocity was then slightly reduced to a level at which the fish could sustain swimming, and a final sampling procedure was conducted following 15 min at this velocity Analytical techniques phe and phi were measured using a radiometer micro-capillary ph electrode (G299A) in conjunction with a radiometer BMS3 Mk2 blood microsystem maintained at the temperature to which the fish was exposed. phi was measured according to the freeze thaw method of Zeidler and Kim (1977). Plasma and whole blood C CO2 were measured on 50 l samples using a gas chromatograph (Carle Instruments, USA, Model III), coupled to a chart recorder as described by Boutilier et al. (1985) Calculations Plasma bicarbonate (HCO 3 ) and P CO2 levels were calculated by re-arrangement of the Henderson Hasselbalch equation: Plasma C Plasma P CO2 = CO2 CO2 [antilog(phe pk )+1] (1) Plasma [HCO 3 ]=Plasma C CO2 ( CO2 P CO2 ) (2) where pk is the apparent pk of plasma and CO 2 is the solubility of CO 2 in plasma at 9 C

4 72 C.J. Brauner et al. / Respiration Physiology 119 (2000) taken from Boutilier et al. (1984). The total CO 2 contained within the erythrocyte (red cell C CO2 ) was calculated as: 2.6. Statistics Statistically significant differences between Red cell C CO2 = Whole blood C CO 2 (Plasma C CO2 (1 Hct/100)) Hct/100 where Hct refers to haematocrit. (3) In calculating the relative contribution of various sources to venous arterial CO 2 differences in rainbow trout during exercise (reported in Table 3) the following equations were used: Percent of venous arterial CO 2 content difference comprised of dissolved CO 2 (%Cv CO2 Ca CO2 as dissolved CO 2 ): mean values measured at different swimming velocities were detected using a repeated measures ANOVA, or Friedman repeated measures ANOVA on ranks, followed by a Dunnett s test. Comparisons between arterial and venous parameters were conducted using a paired t-test. In all cases a probability level of 5% was chosen %Cv CO2 Ca CO2 as dissolved CO 2 = ((Pv CO 2 Pa CO2 ) ((1 Hct/100)+(Hct/ )) CO (4) (Cv CO2 Ca CO2 ) where Pv CO2 and Pa CO2 are mixed-venous blood (sampled from the pre-branchial artery) and arterial blood (sampled from the dorsal aorta) partial pressure of CO 2 (P CO2 ), 0.86 is the solubility of CO 2 in red cells relative to that in plasma (Van Slyke et al., 1928), CO 2 is the CO 2 solubility in plasma from Boutilier et al. (1984), and Cv CO2 and Ca CO2 are the total CO 2 content of prebranchial and arterial blood, respectively. Percentage of venous arterial C CO2 difference comprised of plasma HCO 3 (%Cv CO2 Ca CO2 as plasma HCO 3 ) is calculated as: as the limit of statistical significance. Regression coefficients were calculated using least squares regression. 3. Results 3.1. CO 2 transport and excretion during exercise CO 2 excretion rate is elevated during sustained exercise (Table 1) in accordance with an elevation in oxygen consumption rate. The respiratory ex- %Cv CO2 Ca CO2 as plasma HCO 3 = (HCO 3 vp (1 Hct v /100)) (HCO 3 ap (1 Hct a /100)) (Cv CO2 Ca CO2 ) where HCO 3 vp and HCO 3 ap are [HCO 3 ] in venous and arterial plasma, respectively. (5) Percentage of venous arterial C CO2 difference comprised of red cell HCO 3 (%Cv CO2 Ca CO2 as rbc HCO 3 ) is calculated as: %Cv CO2 Ca CO2 as rbc HCO 3 =100 [(Eq. 4) +(Eq. 5)] (6) This equation assumes that there is no oxylabile carbamate which contributes to the venous arterial C CO2 difference (see Section 4). change ratio was the same at all levels of exercise, being (n=28 determinations) for all data combined. The increase in CO 2 excretion rate was associated with an increase in the venous arterial difference in CO 2 content of the blood and an elevation in cardiac output. There is a strong, linear relationship (r 2 =0.87) between CO 2 excretion rate and cardiac output measured during exercise (Fig. 1). Despite the greatly elevated CO 2 excretion rate, arterial P CO2 and C CO2 also in-

5 Table 1 Blood Hct, ph and CO 2 transport characteristics in rainbow trout at different levels of sustained exercise (Series I) a % Ucrit Hcta Hctv PHae phve PHai PHvi Pa CO2 (mmhg) Pv CO2 (mmhg) Plasma [HCO 3 ]a Plasma[HCO 3 ]v M CO (1.7) (0.32) (1.1) (1.3) (0.01) (0.02) (0.01) (0.01) (0.16) (0.19) (0.44) (0.53) (0.14) (1.5) (0.56) (1.1) (1.3) (0.01) (0.02) (0.01) (0.03) (0.15) (0.28) (0.52) (0.64) (0.16) * 27.7* 28.0* *+ 7.61* *+ 5.37* * 13.86*+ 2.58* (1.2) (1.15) (0.9) (0.9) (0.02) (0.04) (0.03) (0.02) (0.22) (0.55) (0.49) (0.66) (0.15) * *+ 7.63* *+ 6.74* * * 2.47* (0.9) (1.40) (0.8) (0.7) (0.02) (0.03) (0.01) (0.01) (0.21) (0.43) (0.47) (0.57) (0.26) Cv CO2 Ca CO2 (mm) a Ca CO2 Ca CO2 refers to the venous arterial difference in CO 2 content of the blood. M CO2 refers to whole animal CO 2 excretion rate (mmol kg 1 h 1 ), calculated from Cv CO2 Ca CO2 and cardiac output. Values are means with S.E.M. in brackets and n beneath. * signifies statistically different from the lowest swimming velocity+signifies statistically different from respective arterial value. C.J. Brauner et al. / Respiration Physiology 119 (2000)

6 74 C.J. Brauner et al. / Respiration Physiology 119 (2000) ing exercise, venous phe was significantly reduced at the two highest swimming velocities, resulting in a significant arterial venous difference in phe (Table 1). Red cell ph increased significantly in arterial blood during exercise but there were no significant differences between arterial and venous phi (Table 1) Partitioning of CO 2 between plasma and red cells Fig. 1. The relationship between CO 2 excretion rate (M CO2 ) and cardiac output (measured using a transonic flow probe) in rainbow trout forced to swim at different levels of sustained exercise (Series I). Points represent individual data points. (Y= X, r 2 =0.865). creased during exercise. Arterial P CO2 and plasma [HCO 3 ] increased from 2.5 Torr and 9.1 mm at the lowest swimming velocity to 3.9 Torr and 13.6 mm at Ucrit, respectively (Table 1). Interestingly, no significant changes were observed in arterial ph at any swimming velocity (Table 1). A ph/ HCO 3 plot of these data reveal that the elevation in arterial blood CO 2 levels during exercise was associated with a net HCO 3 increase (Fig. 2a). For the group of fish used in Series I, the afferent branchial artery of the first gill arch was cannulated and the arch was tied off, reducing the gill surface area available for gas exchange. However, the same trends in dorsal aortic blood CO 2 levels were observed during exercise in fish which did not have the afferent branchial artery cannulated (Series II, Table 2, Fig. 2b). While arterial blood ph remained constant dur- CO 2 content of plasma, whole blood and red blood cells increased significantly during exercise (Fig. 3). The majority of CO 2 (78 90%) was transported in the plasma with a larger proportion of C CO2 in plasma of arterial relative to venous blood (Fig. 4). During exercise there was a significant decrease in the proportion of C CO2 transported in plasma versus red cells in both arterial and venous blood. The red cells accounted for about 11 16% of the C CO2 in arterial and venous blood, respectively, at rest and this amount increased to a maximum of 16 22%, respectively during exercise (Fig. 4) Partitioning of enous arterial differences in CO 2 In mixed-venous blood, the increase in total blood CO 2 and P CO2 during exercise was more pronounced than that in arterial blood (Table 1). Venous P CO2 and plasma [HCO 3 ] increased from 2.6 Torr and 9.9 mm at the lowest swimming velocity to 6.7 Torr and 13.6 mm at Ucrit, respectively. Thus, with an increase in exercise intensity, there was an increased venous arterial difference in P CO2 and total blood CO 2 (Table 1). The increase in Pv O2 Pa CO2 during exercise resulted in an increase in the proportion of venous arterial C CO2 difference comprised of physically dissolved CO 2 (%Cv CO2 Ca CO2 as dissolved CO 2, see Eq. Fig. 2. A ph/hco 3 plot of changes in blood acid base status of rainbow trout during different levels of sustained exercise with (a; Series I fish) and without (b; Series II fish) one gill arch tied off. The data points represent mean values (error bars represent S.E.M.) for arterial phe and plasma [HCO 3 ] from fish swimming at different velocities (16 (A), 55 (B), 91 (C) and 99 (D)% of Ucrit, Tables 1 and 2). The buffer line (dotted line) was calculated from the regression equation for as a function of [Hb] derived by Wood et al. (1982) and a Hb concentration of 8.5 g dl 1 (the value measured at the lowest swimming velocity [16% Ucrit]; accompanying paper, Brauner et al., 2000).

7 C.J. Brauner et al. / Respiration Physiology 119 (2000) Fig. 2.

8 76 C.J. Brauner et al. / Respiration Physiology 119 (2000) Fig. 3. Total CO 2 content (mm) in: (a) plasma; (b) whole blood; and (c) red blood cells in rainbow trout forced to swim at different levels of sustained exercise (Series I). All venous levels (solid bars) are significantly different from arterial values (open bars). * indicates significant differences from lowest swimming velocity of respective arterial or venous blood.

9 C.J. Brauner et al. / Respiration Physiology 119 (2000) Table 2 Arterial blood Hct, ph and CO 2 levels during sustained exercise in rainbow trout without the afferent branchial artery cannulated (Series II) % Ucrit Velocity (cm s 1 ) Hct phe phi Pa CO2 (mmhg) Plasma [HCO 3 ] 18.4 (1.0) (0.4) (1.1) (0.03) (0.03) (0.20) (0.9) 5 (1.0) (0.4) (1.1) (0.03) (0.03) (0.20) (0.9) (6.1) (2.4) (0.9) (0.02) (0.03) (0.17) (1.0) * 10.8 (4.5) (2.6) (1.3) (0.01) (0.02) (0.15) (0.6) * 11.2 (1.4) (1.0) (1.8) (0.02) (0.02) (0.05) (0.6) (4)). In resting animals, physically dissolved CO 2 accounted for 1% of the Cv CO2 Ca CO2 difference but increased to as high as 9% at Ucrit (Fig. 4). The %Cv CO2 Ca CO2 accounted for by plasma and red cell [HCO 3 ] (Eqs. (4) and (5), respectively) did not change significantly during exercise (data not shown) yielding a pooled value of 54% for plasma HCO 3 and 42% for red cell HCO 3 (assuming no oxylabile carbamate exists; Table 3). 4. Discussion Although oxygen transport during sustained exercise in fish has been investigated in detail (Kiceniuk and Jones, 1977; Randall and Daxboeck, 1984), relatively little is known about CO 2 transport. Currie and Tufts (1993) have reported on changes in blood CO 2 transport following a brief bout of exhaustive exercise in rainbow trout, but this is the first detailed study of changes in CO 2 transport in the blood during aerobic exercise in fish. Table 3 Relative contribution of various sources to venous arterial C CO2 difference (Cv CO2 Ca CO2 ) in rainbow trout at all levels of exercise (Series I), see text and Eqs. (4) (6) for further details a Dissolved CO 2 Plasma HCO 3 Red cell HCO 3 a n=29 in each CO 2 transport during exercise % % % The absence of an acidosis in arterial blood (Table 1) and the relatively low levels of lactate in the plasma near Ucrit (Table 1 of Brauner et al., (2000)) indicate that metabolism was predominantly aerobic during exercise in this study. The respiratory exchange ratio was unaffected by exercise level and, because changes in body oxygen and carbon dioxide stores were minor in comparison with excretion rates, this indicates that the respiratory quotient was also unchanged by exercise. Thus both oxygen utilization and carbon Fig. 4. Proportion of CO 2 transported in: (a) plasma; (b) red cells; and (c) as physically dissolved CO 2 in the plasma, in rainbow trout at different sustained swimming velocities (Series I). In (a) and (b) open bars indicate arterial values while closed bars indicate venous values. %Cv CO2 Ca CO2 as dissolved CO 2 represents the % of venous arterial C CO2 difference comprised of dissolved CO 2 (see Eq. (4)). * indicates significant differences from lowest swimming velocity of respective arterial or venous blood, while + indicates significant differences from respective arterial value.

10 78 C.J. Brauner et al. / Respiration Physiology 119 (2000) Fig. 4.

11 C.J. Brauner et al. / Respiration Physiology 119 (2000) dioxide production increased proportionately with exercise and there was a linear increase in CO 2 excretion relative to cardiac output (Fig. 1). There was, however, an increase in both blood P CO2 and C CO2 levels in arterial blood (Tables 1 and 2), with increasing levels of exercise indicating that a diffusion limitation could exist for CO 2 excretion during exercise. This increase in carbon dioxide levels was larger than that observed in Series II fish, where the afferent branchial artery was not cannulated (Table 2), indicating that there was a detectable effect of eliminating blood flow through one of the gill arches on CO 2 transport consistent with a diffusion limitation of some sort during exercise. A similar relationship was also observed in an isolated spontaneously ventilating trout head preparation (Perry et al., 1982), however, to our knowledge this is the first time this relationship has been demonstrated in unrestrained fish. Thomas et al. (1987) also observed an increase in arterial P CO2 which was correlated with swimming velocity, while Wilson and Egginton (1994) did not observe a significant change in arterial P CO2 during sustained exercise. The magnitude of the increase in Pa CO2 in vivo is likely less than that reported in this study due to disequilibrium states in the blood, however, this has not been determined experimentally. An increase in Pa CO2 may be important in maintaining ventilation following exercise, speeding up the recovery process, as has been demonstrated following exhaustive exercise (Wood and Munger, 1994). Arterial plasma HCO 3 levels increased during exercise by as much as 50% at the maximum swimming velocity (Table 1), which was larger in fish with one afferent branchial artery tied off (Series I) compared with intact fish (Series II). Interestingly, the elevation in blood P CO2 and total CO 2 in either group of fish did not result in an acidosis in arterial blood at any swimming velocity (Fig. 2) indicating an efficient and rapid ability to elevate plasma [HCO 3 ] either through acid excretion (coupled with HCO 3 retention) at the level of the gills or kidney, or via direct HCO 3 uptake across the gills CO 2 partitioning between plasma and red cells during exercise The majority of CO 2 transported in the blood was carried within the plasma compartment with only 11 and 16% transported within the red cells in arterial and venous blood, respectively (Fig. 4) at the lowest swimming velocity. These values are consistent with those measured by Heming (1984) in resting rainbow trout. In resting fish, Currie and Tufts (1993) measured a much lower proportion of C CO2 transported in the erythrocytes, about 2% in arterial blood and 9% in venous blood, associated with a larger phe phi difference. Following exhaustive burst exercise, the proportion of C CO2 transported in the red cells increased significantly to 13.5 and 20% in arterial and venous blood, respectively (Currie and Tufts, 1993) similar to the values observed in this study. They attributed the increased proportion of CO 2 within the red cells to the increase in Hct and the effects of catecholamines on red cell ph and subsequent distribution of CO 2 between the red cells and the plasma. The changes in both red cell and plasma ph during sustained exercise in this study were small in comparison with those following exhaustive exercise (Currie and Tufts, 1993) resulting in only modest changes in the proportion of CO 2 transported within the red cell Partitioning of enous arterial differences in CO 2 during exercise In general, CO 2 transport and excretion in fish is similar to that of most vertebrates (Perry, 1986). That is, the majority of CO 2 is transported in the plasma as HCO 3 but almost all CO 2 is excreted across the gills as molecular CO 2. When blood first enters the gills, any physically dissolved CO 2 which exists in pre-branchial blood will rapidly diffuse across the gill epithelium into the ventilatory water, creating conditions for HCO 3 dehydration. HCO 3 which resides within the red cell will then be dehydrated to CO 2 in the presence of carbonic anhydrase. There is probably very little HCO 3 dehydrated within the plasma compartment during gill blood transit because there is no CA activity available to the plasma in teleost fish

12 80 C.J. Brauner et al. / Respiration Physiology 119 (2000) gills (Perry and Laurent, 1990) and the buffer capacity of plasma is low. Thus, the reduction in red cell HCO 3 concentrations will create the condition for HCO 3 entry into the red cell via Cl / HCO 3 exchange permitting continued CO 2 excretion during gill blood transit. In fish swimming at 16% Ucrit, the venous arterial difference in P CO2 was very small and thus the contribution of dissolved CO 2 to total CO 2 excretion was negligible (Fig. 4). However, as exercise intensity increased, this route for CO 2 excretion reached a maximum of 9%, comparable to that in resting humans (Comroe, 1974). Our calculations are based upon P CO2 values calculated in blood at equilibrium following sampling. In actuality, equilibrium conditions between dissolved CO 2 and bicarbonate may never be achieved in vivo (Gilmour et al., 1994). The absence of plasma accessible CA in the gills of fresh water teleost fishes gives rise to a post branchial blood disequilibrium. As blood flows away from the gills, arterial blood ph slowly increases as the plasma HCO 3 is titrated to CO 2 at the uncatalyzed rate, resulting in an elevation in plasma P CO2 (Gilmour et al., 1994). P CO2 and ph also increase during stop-flow in mixed-venous blood (Perry et al., 1997). In an analysis of CO 2 excretion in resting fish subjected to normoxia or hypoxia, there was no significant influence of the P CO2 disequilibrium on the calculation of the contribution of dissolved CO 2 to venous arterial C CO2 difference (Brauner, 1995). Blood gas parameters have been measured during exercise in arterial blood using an extracorporeal circuit (Thomas et al., 1987) but the magnitude of the disequilibrium during exercise has not been assessed. The degree to which disequilibria will alter blood P CO2 at the high swimming velocities used in this study is not known but is assumed to be minor. After accounting for the role of dissolved CO 2 to the Cv CO2 Ca CO2 difference, the remaining CO 2 excreted must have arisen either from red cell or plasma HCO 3 (ultimately dehydrated to CO 2 within the red cell) or from red cell carbamate. CO 2 reacts with the N-terminal amino groups of Hb to form carbamate but is not thought to play a role in CO 2 excretion in most fish because the terminal amines of the Hb subunits are acetylated while the available terminal amines of the -subunits are in direct competition with organic phosphates which are preferentially bound (Gillen and Riggs, 1973; Farmer, 1979). Heming et al. (1986) concluded that appreciable quantities of carbamate exist in rainbow trout blood, however, they only investigated oxygenated blood and therefore could not determine what proportion of the carbamate was oxylabile. Thus, there is no evidence for the existence of oxylabile carbamate in the blood of rainbow trout but this remains to be measured experimentally. It is likely, however, that after accounting for dissolved CO 2, the difference between venous arterial red cell C CO2 reflects the contribution of HCO 3 within the red cell prior to gill entry which is dehydrated to CO 2 during gas exchange. This value did not change significantly with exercise level and accounted for 42% of the Cv CO2 Ca CO2 difference (Table 3). Assuming that all HCO 3 dehydration occurs within the red cell, the venous arterial changes in plasma HCO 3 may be indicative of HCO 3 /Cl exchange during CO 2 excretion. Venous arterial changes in plasma HCO 3 accounted for 54% of the Cv CO2 Ca CO2 difference which did not change significantly with exercise (Table 3). Extrapolating venous arterial changes in plasma HCO 3 to HCO 3 /Cl exchange also assumes that HCO 3 /Cl is complete during blood passage through the gills as has been suggested for resting fish (Gilmour et al., 1994). During exercise, when the residence time of red cells in the gill lamellae is reduced 3-fold, HCO 3 /Cl exchange may not be complete in post-branchial blood (Jensen and Brahm, 1995). This could result in an overestimation of the contribution of HCO 3 /Cl exchange to CO 2 excretion during blood transit through the gills but it is probably very small. Another assumption of this model is that there is no direct uptake of HCO 3 from the water across the gills. The data in Fig. 2 indicate that there may be a direct uptake of HCO 3 across the gills during exercise. Relative to the rate of CO 2 excretion, this pathway is quite small and would tend to underestimate the contribution of HCO 3 /Cl exchange during CO 2 excretion, offsetting the effects of incomplete HCO 3 /Cl exchange in postbranchial blood.

13 C.J. Brauner et al. / Respiration Physiology 119 (2000) That 54% of the total CO 2 excreted was dependent upon HCO 3 /Cl exchange during exercise is in close agreement with the value obtained in resting fish exposed to normoxia and two levels of hypoxia (Brauner, 1995). Such a large dependence of HCO 3 /Cl exchange to CO 2 excretion is significant because this exchange is thought to be the rate limiting step in CO 2 excretion (Perry et al., 1982; Perry, 1986; Jensen and Brahm, 1995). The rate limitation of HCO 3 /Cl exchange across the red cell membrane may be offset during exercise by an increase in hematocrit and venous bicarbonate levels, however, HCO 3 /Cl exchange is probably the site at which CO 2 was diffusion limited during exercise ultimately leading to the observed elevation in arterial P CO2 (and total CO 2 content) required to restore CO 2 excretion rate. Acknowledgements This work was funded by an NSERC operating grant to A.P.F. and D.J.R. P.G. and H.T. were recipients of Graduate Fellowships from Simon Fraser University and C.B. was supported by an NSERC post graduate scholarship. We thank the two anonymous referees for valuable suggestions. References Boutilier, R.G., Heming, T.A., Iwama, G.K., Appendix: physicochemical parameters for use in fish respiratory physiology. In: Hoar, W.S., Randall, D.J. (Eds.), Fish Physiology, vol. 10A. Academic Press, New York, pp Boutilier, R.G., Iwama, G.K., Heming, T.A., Randall, D.J., The apparent pk of carbonic acid in rainbow trout blood plasma between 5 and 15 C. Respir. Physiol. 61, Brauner, C.J., An analysis of the transport and interaction of oxygen and carbon dioxide in fish. Ph.D. Thesis. University of British Columbia, Vancouver, BC, p. 148 p. Brauner, C.J., Thorarenson, H., Gallaugher, P., Farrell, A.P., Randall, D.J The interaction between O 2 and CO 2 in the blood of rainbow trout (Oncorhynchus mykiss) during graded sustained exercise. Respir. Physiol. 119, Brett, J.R., The respiratory metabolism and swimming performance of young sockeye salmon. J. Fish. Res. Board. Canada 21, Comroe, J.H.J., Physiology of Respiration. Year Book Medical Publishers, Chicago, p Currie, S., Tufts, B.L., An analysis of carbon dioxide transport in arterial and venous blood of the rainbow trout, Oncorhynchus mykiss, following exhaustive exercise. Fish Physiol. Biochem. 12, Farmer, M., The transition from water to air breathing: effects of CO 2 on hemoglobin function. Comp. Biochem. Physiol. 62A, Gallaugher, P., Axelsson, M., Farrell, A.P., Swimming performance and haematological variables in splenectomized rainbow trout, Oncorhynchus mykiss. J. Exp. Biol. 171, Gillen, R.G., Riggs, A., Structure and function of the isolated hemoglobins of the American eel (Anguilla rostrata). J. Biol. Chem. 248, Gilmour, K.M., Randall, D.J., Perry, S.F., Acid base disequilibrium in the arterial blood of rainbow trout. Respir. Physiol. 96, Heming, T.A., The role of fish erythrocytes in transport and excretion of carbon dioxide, Ph.D. Thesis. University of British Columbia, Vancouver, BC, p. 177 p. Heming, T.A., Randall, D.J., Boutilier, R.G., Iwama, G.K., Primmett, D., Ionic equilibria in red blood cells of rainbow trout (Salmo gairdneri ): Cl, HCO 3 and H +. Respir. Physiol. 65, Jensen, F.B., Brahm, J., Kinetics of chloride transport across fish red blood cell membranes. J. Exp. Biol. 198, Jones, D.R., Randall, D.J., The respiratory and circulatory systems during exercise. In: Hoar, W.S., Randall, D.J. (Eds.), Fish Physiology, vol. 7. Academic Press, New York, pp Kiceniuk, J.W., Jones, D.R., The oxygen transport system in trout (Salmo gairdneri ) during sustained exercise. J. Exp. Biol. 69, Kieffer, J.D., Alsop, D., Wood, C.M., A respirometric analysis of fuel during aerobic swimming at different temperatures in rainbow trout (Oncorhynchus mykiss). J. Exp. Biol. 210, Perry, S.F., Davie, P.S., Daxboeck, C., Randall, D.J., A comparison of CO 2 excretion in a spontaneously ventilating blood-perfused trout preparation and saline-perfused gill preparations: contribution of the branchial epithelium and red blood cell. J. Exp. Biol. 101, Perry, S.F., Carbon dioxide excretion in fishes. Can. J. Zool. 64, Perry, S.F., Laurent, P., The role of carbonic anhydrase in carbon dioxide excretion, acid base balance and ionic regulation in aquatic gill breathers. In: Truchot, J.P., Lahlou, B. (Eds.), Animal Nutrition and Transport Processes 2. Transport, Respiration and Excretion: Comparative and Environmental Aspects, vol. 6. Karger, Basel, pp Perry, S.F., Brauner, C.J., Tufts, B., Gilmour, K.M., Acid base disequilibrium in the venous blood of rainbow trout (Oncorhynchus mykiss). Exp. Biol. Online 2, 1. Randall, D., Blood flow through gills. In: Houlihan, D.F., Rankin, J.C., Shuttleworth, T.J. (Eds.), Gills. Society for

14 82 C.J. Brauner et al. / Respiration Physiology 119 (2000) Experimental Biology Seminar Series, vol. 16. Cambridge University Press, USA, pp Randall, D.J., Daxboeck, C., Oxygen and carbon dioxide transfer across fish gills. In: Hoar, W.S., Randall, D.J. (Eds.), Fish Physiology, vol. 10A. Academic Press, New York, pp Thomas, S., Poupin, J., Lykkeboe, G., Johansen, K., Effects of graded exercise on blood gas tensions and acid base characteristics of rainbow trout. Respir. Physiol. 68, Thorarensen, H., Gallaugher, P., Farrell, A.P., Cardiac output in swimming rainbow trout, Oncorhynchus mykiss, acclimated to seawater. Physiol. Zool. 69, Van Slyke, D.D., Sendroy, J., Hastings, A.B., Neill, J.M., Studies of gas and electrolyte equilibria in blood. X: the solubility of carbon dioxide at 38 C in water, salt solution, serum and blood cells. J. Biol. Chem. 78, Wilson, R.W., Egginton, S., Assessment of maximum sustainable swimming performance in rainbow trout (Oncorhynchus mykiss). J. Exp. Biol. 192, Wood, C.M., McDonald, D.G., McMahon, B.R., The influence of experimental anaemia on blood acid base regulation in vivo and in vitro in the starry flounder (Platichthys stellatus) and the rainbow trout (Salmo gairdneri ). J. Exp. Biol. 96, Wood, C.M., Munger, R.S., Carbonic anhydrase injection provides evidence for the role of blood acid base status in stimulating ventilation after exhaustive exercise in rainbow trout. J. Exp. Biol. 194, Zeidler, R., Kim, D.H., Preferential hemolysis of postnatal calf red cells induced by internal alkalinization. J. Gen. Physiol. 70,