J. exp. Biol. (1979), 79, 115-126 I I 5With afigura

Printed in Great Britain

BLOOD ACID-BASE BALANCE IN BROOKTROUT (SALVELINUS FONTINALIS)*

BY RANDALL K. PACKER AND ARTHUR L. SUNKIN

Department of Biological Sciences, George Washington University,Washington, B.C. 20052, U.S.A.

(Received 15 June 1978)

SUMMARY

A nomogram is presented which allows the rapid assessment of acid-basestatus in brook trout (SalveKnus fontinalis Mitchill) at 18 °C.

Due to a PCOt of 3-5—4*3 mmHg in the respirometer water, the fish fromwhich data were obtained for determination of zero base excess values weresuffering from slight respiratory acidosis. Trauma of surgery and anaesthe-tization as well as confinement in the respirometer might also have generateda slight metabolic acidosis.

Base excess values for trout with a chronically implanted dorsal aorticcannula ranged from — 3 to +3 m-equiv/1. Blood buffer values of<f[HCO3-]/JpH = 7-5o±4'42 slykes and -dlog10POoJdpH = 1-31 ±0-14were found. Mean plasma protein concentration was 3-6 ±0-8 g/100 mlplasma and the buffering power of plasma was approximately 40% that ofwhole blood. Plasma proteins are more important blood buffers in fish thanin humans.

INTRODUCTION

Various nomograms designed for assessment of acid-base balance in humans havebeen in use for a number of years. These nomograms, in their various forms, allow thederivation of a number of parameters pertaining to acid-base status from a fewrelatively easily measured factors such as blood pH, POo,, and haemoglobin concen-tration ([Hb]) (Singer & Hastings, 1948; Siggaard-Andersen, 1962).

The aim of this research was the construction of a nomogram consisting of baseexcess (BE) and bicarbonate scales on a pH-PC0 co-ordinate system which could beused for the rapid assessment of acid-base status in brook trout (Salvelinus fontinalis).While producing the nomogram, we hoped to establish 'normal' ranges for someindicators of acid-base status such as BE, actual bicarbonate and standard bicarbonatein cannulated fish. Standard bicarbonate is a measure of what the plasma bicarbonateconcentration would be at any given pH if Pco, w a s 'normal', and is used to assessthe non-respiratory (or metabolic) component of blood acid-base status, as is BE.A further goal was to investigate blood buffering.

• A portion of this work was submitted as a thesis by A. L. Sunkin to George Washington Universityin partial fulfilment of the requirements for the M.Sc. degree.

n 6 R. K. PACKER AND A. L. SUNKIN

MATERIALS AND METHODS

A. Source and maintenance offish

The Pennsylvania Fish Commission Hatchery at Huntsdale, Pa. provided 44 troutof both sexes ranging in weight from 185 to 380 g. These trout are considered to bethe Huntsdale strain (strain H). We also used seven strain O (wt 285-465 g) and threestrain 15 (wt 235-485 g) trout of both sexes obtained from the Benner Spring FishResearch Station, Pennsylvania Fish Commission, Bellefonte, Pa. Strains 15 and Owere derived from nine and seven generations of inbreeding, respectively, from theBellefonte Open strain (Robinson et al. 1976).

Trout were maintained in the laboratory at 18 ± 1-5 °C in a 680 1 epoxy coated steeltank for at least 2 weeks prior to use in experiments. Approximately 8 lh~x of charcoal-filtered tap water flowed through the tank. Tank water was also recirculated through aglass wool and charcoal filter at a rate of approximately 7 lh"1. An ultraviolet light waspositioned over the filter. The fish were fed Purina Trout Chow three times per weekand were exposed to the natural photoperiod. Trout were not fed for 48 h prior to anexperiment. All experiments were conducted in a controlled temperature room at awater temperature of 18 ±0*5 °C.

B. Instrumentation and techniques

All measurements of blood pH and Pco were made using a Radiometer BloodMicro System model BMS 2 Mk II in conjunction with a Radiometer PHM 71Mk II acid-base analyser at I 8 ± O - I °C. This instrument was calibrated with Radio-meter buffers S 1500 and S 1510 at pH 6-888 and 7-434. Buffer pH values werecorrected to 18 °C using data provided by Radiometer. Blood samples were collectedanaerobically and the pH of a 60 fi\ aliquot was measured. The original Poo, °f theblood sample was then determined using the Astrup technique. Two 80 fil aliquots ofthe blood sample were equilibrated with water-saturated gases of 4% (RadiometerCCS) and 8 % (Radiometer CCS) or 4% (Radiometer CCS) and 1 -45 % (Air Products)CO2, 21% O2, balance N2. Following equilibration, the pH of each aliquot wasmeasured and a pH-POOi titration line w,as thus determined. When the Astruptechnique is used the PCOj of the original sample, prior to equilibration, is determinedby the position of the original pH on the pH-PC0> titration line which is plotted on apH-PCOj co-ordinate system.

Blood [Hb] in gram percentage was determined by measuring cyanmethemoglobinusing Hycel reagent and standards (Hycel, Inc.) and a Coleman Jr Spectrophotometermodel 6A, following the methods of Blaxhall & Daisley (1973).

Dorsal aortic cannulations were done following the procedures of Smith & Bell(1964; 1967). Cannulae consisted of a 23-gauge Huber point syringe needle tipinserted in a length of PE-50 Intramedic polyethylene tubing. The cannulae werefilled with heparinized (1000 i.u./ml) Courtland saline (Wolf, 1963). Following cannu-lation the fish were transferred to a closed respirometer system having a 35 1 totalcapacity. The respirometer as well as the Radiometer blood gas system and theoperating table used in cannulation were all situated in the controlled-temperatureroom at 18 ±0-5 °C. The aortic cannula was connected to a 23-gauge syringe needl^tip which passed through a rubber stopper in the lid of the respirometer. A 6-10 cm

Blood acid-base balance in brook trout

length of saline-filled PE-50 tubing was attached to the other end of this needle onthe outside of the respirometer lid. Blood was sampled by releasing a clamp on thistubing. After a blood sample was taken, the blood remaining in the cannula wasflushed back into the fish with heparinized saline. The exterior of the respirometerwas painted so that the fish could not see the investigator. Cannulated fish wereallowed to recover for at least 24 h prior to any data collection in all experiments.Houston, Czerwinski & Woods (1973) and Soivio, Nyholm & Huhti (1977) report thatmost physiological parameters altered by anaesthestization and cannulation stabilizeafter 24 h.

Haematocrit and plasma protein concentrations were measured in blood samplesfrom cannulated trout. Haematocrit was determined by centrifugation of blood incapillary tubes at 11500 rev/min for 5 min in an IEC model MB haematocrit centri-fuge. Total protein concentration of 20 fil plasma samples was measured using themethods of Palma (1971).

C. Determination of standard values of blood pH and PO O i

Prior to the construction of the nomogram, it was necessary to establish values ofpH and JPQO,

m dorsal aortic blood of cannulated trout which could be used to definethe condition of zero BE. Aortic blood pH and PCOj were determined in anaerobicallydrawn blood samples from 19 trout with the dorsal aorta cannulated. Of this group,9 were strain H, 7 were strain O, and 3 were strain 15.

D. Construction of the nomogram

The nomogram constructed was based on the curve nomogram of Siggaard-Andersen (1962). This form was chosen, rather than an alignment nomogram such asthose of Singer & Hastings (1948) or Siggaard-Andersen (1963), because the curvenomogram is well suited for use in conjunction with the Astrup technique for Poo%

determination. The pH-PCOt titration lines generated by the Astrup technique maybe plotted directly on the pH-Poo, co-ordinate system of the curve nomogram. Theoriginal Po0t of the blood sample and the blood acid-base parameters given on thevarious nomogram scales may then be determined directly from the p H - P ^titration line. In addition, the slope and position of the pH-PCOt titration line may beused to calculate blood buffer values.

The nomogram, consisting of BE and standard bicarbonate scales on a pH-POOi

axis system, was constructed following the methods of Siggaard-Andersen & Engel(i960) and Siggaard-Andersen (1962). Modifications of their methods are given in thefollowing brief description of construction of the nomogram.

In order to construct the BE scale, blood was collected into a heparinized glasssyringe from the dorsal aorta of 35 strain H trout which were anaesthetized in asolution of 50 mg/1 of MS 222 (Crescent Research Chemicals). The blood of two orthree fish was pooled to obtain sufficient quantities (6-7 ml) for titration. Blood wasstored on ice for no more than 1 h until titrations and measurements were completed.

Each pool of blood was used in the titrations as whole blood, as plasma obtained bycentrifugation, or as whole blood to which red blood cells had been added to produce aHigh [Hb], Each pool was divided into a series of 490 fil aliquots, to which were added12-5 fil of acid (HC1) or base (Na2CO8) solution. The amounts of acid added were

n 8 R. K. PACKER AND A. L. SUNKINx> 3> 5> 7> 9, or i i m-equiv/1 of blood and the amounts of base were i, 3, 5or 7 m-equiv/1. Although the proportion of blood to acid or base added was the sameas that used by Siggaard-Andersen (1962), we used smaller total volumes to conserveblood. A complete titration series consisted of three pools of blood of differing [Hb]to which the quantities of acid or base noted above were added. NaCl was added to theHC1 and Na^COg solutions to give a total [Na+] of 150 HIM. Volumes of blood andacid-base solutions were measured using 1000 and 25 fi\ Hamilton syringes fitted withChaney adaptors.

Following the addition of acid or base to a given blood or plasma sample, the pHwas measured. Then, two 80 fi\ aliquots of blood were placed in the tonometers of theRadiometer apparatus and were equilibrated with water-saturated gases of 4% CO2,21 % O2, balance N2 and 8% CO2, 21 % Oj, balance Na. Following equilibration, thepH of each aliquot was measured and a pH-PCOt titration line was thus determined.The slope of this titration line is directly related to the [Hb] of the blood sample.Because of differences in slope resulting from differences in [Hb], the pH-POOf linesintersect at a single point for aliquots of plasma, whole blood, and whole blood plusred blood cells to which, for example, 3 m-equiv/1 of acid had been added. Any bloodsample that had the same initial acid-base characteristics and to which 3 m-equiv/1 ofacid had been added should yield a pH-PCOi titration line which intersects the samepoint. Such intersection points were determined for each quantity of acid or baseadded. Four series of titrations were done, and yielded a minimum of three intersectionpoints for each quantity of acid or base added.

All intersection points were then plotted and after standard values of arterial bloodpH and Poot were determined in cannulated trout, the BE curve was drawn on thenomogram. The BE value of a blood sample is read at the point the pH-Pco,titration line intersects the BE scale.

In addition to the BE curve, a standard bicarbonate scale was placed on the nomo-gram (see Results section).

From pH-Poo, titration lines plotted on the completed nomogram, we calculatedtwo different blood buffer values. An assessment of non-bicarbonate buffering powerwas made using the relationship <f[HCO3~]/dpH, which corresponds closely to-/?a( -fla = dB/dpU + d[HCO3-]/d pH + 2d[COa~]/d pH) in titrations with carbondioxide rather than with strong acids and bases. Due to this similarity we designatedd[HCO3-]/dpH as - / ? ; . We also calculated -dlog^P^JdpH which has beensymbolized as — yffOOt (Burton, 1973). — AGOI*3 a measure of the combined bufferingpower of bicarbonate and non-bicarbonate buffers.

RESULTS

Construction of the BE curve necessitated the establishment of ' normal' values forpH and PQQ^ of trout blood from the dorsal aorta at 18 °C in cannulated trout. Datafrom cannulated trout of all three strains yielded the following values:pH = 7797±0-065 (-X'is.D., n = 19); PCOl = 5-2± I-I mmHg (X±S.T>., n = 18).A one-way analysis of variance (Steel & Torrie, i960) showed no significantdifferences in pH (F = 1-50; P > 01) and P(X)t (F = 1818; P > o-i) among th«|three strains, so mean values were calculated by lumping data from fish of all th re '

Blood acid-base balance in brook trout 119

Table 1. Co-ordinates of points comprising the base excess scale

Base excess(m-equiv/1)

— 1 1

- 9- 7- 5- 3— 1

o+ 1+ 3+ 5+ 7

Co-ordinates

pH

76627-6947-7177-742776677897-79778127-8427-8767-917

i'oo,

i-942 7 03283-8S

4-434-98S-2O5-465-986-43670

strains. These values were then used to define the condition of zero BE. It must bepointed out that these trout were suffering from slight respiratory acidosis and perhapsmetabolic acidosis as well. The POOi of water in the respirometer ranged from 3-5 to4-3 mmHg. This led to the relatively high blood PQQ,. A slight metabolic acidosismay have been generated by the stresses of anaesthetization and surgery as well as bythe stress of confinement in the respirometer. Determinations of true zero BE valuescould be obtained only from completely unstressed fish. After adjustment of all datafrom the acid or base titrations to these normal values, the intersection points ofpH-P00| titration lines were plotted for blood of different [Hb] to which a givenamount of acid or base was added. A smooth curve, representing the BE curve, wasfitted by eye through the point clusters (Fig. 1) and was placed on the nomogram(Fig. 2). The co-ordinates of points comprising the BE curve are given in Table 1.BE of any blood sample is then determined by the point of intersection of a pH-POOi

titration curve of the sample with the BE scale.A standard bicarbonate scale was also constructed and placed on the nomogram. On

this nomogram, standard bicarbonate is the plasma [HCO8~] in blood having aPoot of 5-2 mmHg with the haemoglobin fully Oa saturated. Using the Henderson-Hasselbalch equation, it is possible to calculate plasma [HCOS~] for any pH as follows:

From a plot of the data given by Severinghaus (1965), we determined that the solubilityof CO2 in human plasma at 18 °C is 0-0492 mmol/1 per mmHg Pcot- Since troutplasma is similar in ionic strength to human, we felt safe using this value. Values forpK, which change with pH, were read from the nomogram of Severinghaus, Stupfel& Bradley (1956). Bicarbonate concentrations were calculated for pH ranging from6-8 to 8*2 at a PCOs of 5-2 mmHg, and the scale was placed on the nomogram (Fig. 1).Standard bicarbonate values are read from the point of intersection of a p H - P ^titration line of a blood sample with the standard bicarbonate scale.

The 'actual bicarbonate' value, i.e. the bicarbonate concentration of anaerobicallydrawn plasma, is found by connecting the original blood pH-PCOi point to thestandard bicarbonate scale with a line having a slope equal to the slope of a pure

ticarbonate buffer. The slope of an equilibrium line for a bicarbonate solution can beetermined from calculation pH or PClOt values for a given constant bicarbonate

1 2 0 R. K. PACKER AND A. L. SUNKIN

10987

6 -

_ 5 -

e 4

o 3 -

2 -

6-8

-7

+7

7-4 7-5 7-6l

7-7

PH

I7-8

I7-9 8 0

Fig. i. Base excess curve. Individual points (open circles) represent intersection points ofpH-Poo. titration lines of blood samples having varying [Hb], to which a known amount ofacid or base had been added. All intersection points resulting from experiments in which aparticular amount of acid or base was added are encircled and the amount of acid or baseadded (m-equiv/1) is shown.

X

E

O

60-i

40

30

2 0 -

1Q -

3 -

2 -

1 1-2 1-41-6 2 3 4 5 6 7

Standard bicarbonate m-equiv/1

24 28

Base excessm-equiv/1

6-8 6-9 7 0 7-1 7-2 7-3 7-4

PH

7-5 7-6 7-7 7-8 7-9 8 0 81

Fig. 2. The acid-base nomogram, with scales for standard bicarbonate and base excess; onwhich is shown a pH-Poo, titration line of a blood sample from a cannulated trout. Point Adesignates the original pH of the sample. Open circles indicate blood pH values after equili-bration with 4 and 1-45 % COf

Blood acid-base balance in brook trout 121

concentration according to the modified Henderson-Hasselbalch equation (Siggaard-Andersen, 1964; Davenport, 1975).

where a is the solubility of CO2. Rearranging,

Plotting values of pH and -Pco, obtained from this equation on the nomogramyielded a line having a slope of — 280. Therefore, to determine the actual bicarbonatevalue of any blood sample, a line with a slope of — 280 is drawn from the actual bloodpH value to the standard bicarbonate scale. The intersection point is the actualbicarbonate value.

To illustrate the use of the nomogram, an example of a pH-POOt titration line for ablood sample taken from a cannulated trout is shown on Fig. 2. Point A on the titrationline designates the original pH of the sample, 7-881. From the nomogram, one candetermine that original Poot = 3-8 mmHg; standard bicarbonate = 11 -4 m-equiv/1;base excess = o; and actual bicarbonate = 10-7 m-equiv/1.

Table 2 shows mean values for various blood parameters for cannulated trout.These data were derived from not only the 19 cannulated trout which were used todetermine the 'normal' pH and PCOt values needed to define zero BE, but alsoinclude data from an additional 21 cannulated trout used in experiments done afterconstruction of the nomogram (see Packer, 1978).

The data were analysed for differences between strains using a one-way analysis ofvariance as described above. Values of F indicating no significant differences betweenstrains were found for haematocrit, total plasma protein, base excess, and the twomeasures of blood buffering, — fi* and — /?oo,- F values indicating significantdifferences (P < 0-05) among strains were found for haemoglobin concentration,actual [HCO3-] and standard [HCO3-].

The application of Duncan's New Multiple Range Test (Steel & Torrie, i960)indicated that [Hb] in strain H was lower (P < 0-05) than that of strain 15, but notstrain O. There was no significant difference in [Hb] between strains O and 15. Thestandard and actual bicarbonate values for strain H were lower (P < 0-05) than bothstrains O and 15, but there were no differences between strains O and 15.

DISCUSSION

The nomogram we have constructed is valid only for brook trout at 18 °C havingPCOl within the range of the fish used in this study. A further limitation is that, asmay be seen in Fig. 1, the accuracy of the points on the BE curve decreasesat numerically high BE values. Additionally we were unable to extend the BE curvebeyond —11 m-equiv/1 because the pH of acidified and COa equilibrated blood wastoo low to be read on the expanded scale of the pH meter. Thus the BE scale of thenomogram is not useful for the assessment of severe acid-base disturbances. TheIstandard bicarbonate scale may be used to assess the degree of metabolic disturbancesin extreme conditions.

Tab

le 2

. V

alw

of va

riou

s b

hd

pmam

eltr

s for

cann

- tr

out

Str

nin

H X S.D

. (n

) R

anee

0

3

S.D

.

Ran

ee

15

X

B.D

.

Ran

se

Poo

led xf

S.D.

To

tal

prot

ein

(g/1

=

ml

PI-=

)

3'7

f 0.

7 (7

) 2.

6-5'

0

3 '7

f 1

'0 (

10)

2.0-

5.1

3'5

f 0.

8 (7

) 2'

1-4'

3

3.6 f 0

.8

Sta

nd

ard

IH

CO

a-I

(m-e

qu

iv/l

)

10.4

f 1

.6 (

9)

7'6

13

.0

12'1

f 1

.8 (

6)

9'8-

13-8

12.3

f 0

.7 (

3)

I 1.

5-12

.7

-

Tab

le 3

. A

um

apm

isa of

erah

rcs of

arte

rial

blood Pco, fd

in t

his

std

y m

Mth

th

ose

repo

rted

by

othe

r in

vest

igat

ors

Blw

d P,,

Blw

d p

H

(mm

He)

Tem

p.

(OC)

sp

ecies

Salm

o ga

irdn

mn

m

S. g

crir

dnm

' S

. gai

rdnm

' S

. gai

rdne

ri

S. g

nir

drm

i S. ga

b&

(gro

up I

, Aug

.)

S. g

aird

nm'

(gro

up 1

1, N

ov.)

S. g

crir

drm

i (g

roup

111

, N

ov.)

S-

fontinolir

S. f

onho

nh&

Ref

eren

ce

Eddy,

1976

J

m

& Randall,

I975

H

olet

on &

Randall,

1967

C

amer

on &

Randall,

1972

R

anda

ll &

Cam

eron

, 19

73

Soi

vio

et al. I

977

S

oivi

o et

al. 1

977

Soi

vio

et al. 1

977

Mil

ne &

Randall,

1976

This s

tud

y

Blood acid-base balance in brook trout 123

The average pH of dorsal aortic blood from trout in neutral environments was7797 ±0-065. Taking temperature differences into consideration, it can be seen inTable 3 that this value is within the expected range. Aortic blood PCOt (5-2 mmHg),on the other hand, is higher than values for rainbow trout reported by other workers(Table 3). The respirometer system used in this study was closed, and despitevigorous air bubbling in the reservoir, the PcOi °f water entering the respirometerranged from approximately 3-5 to 43 mmHg. Therefore the high POOi of inspiredwater resulted in a blood Pc 0, that was high, but within a range easily tolerated bytrout (Eddy & Morgan, 1969). Due to the elevated blood PCo,> actual plasma[HCO3~] (11 -3 m-equiv/1) was higher than that reported by other workers for trout(Table 3).

BE values ranged from — 6 to +3 m-equiv/1 in fish in neutral environments, withmost values falling between — 3 and + 3 m-equiv/1. The normal range of BE values inhumans is —3 to +3 m-equiv/1 (Tietz, 1976), indicating a similarity in control of therange of non-volatile acid concentrations in the blood of the two species.

We found blood buffer values of - # J = 7-5014-42 slykes and -/?oo, =1-31 ±0-14. In order to make comparisons with the data of other workers we areassuming that — dC^^Jd pH (CCOi = total CO2 content) would be only slightlygreater than — fl* for the same blood sample (Burton, 1973). As may be seen inTable 4, our buffer values are somewhat lower than those reported by other workersfor rainbow trout. No obvious reasons for the differences, such as large differences in[Hb], are evident.

Average plasma protein concentration was 3-6 + 0-8 g/100 ml plasma, and apparentlyplasma protein contributes significantly to blood buffering. — J3* and —fiOOt werecalculated from CO2 titrations of seven separated plasma samples, —fit —3-25 ±0-95 slykes/1 of plasma (X±s.D.), showing that the average non-bicarbonatebuffering capacity of plasma was about 40% that of whole blood. In humans, thenon-bicarbonate buffering power of plasma is about 24% that of whole blood (Tietz,1976). The average [Hb] of whole blood samples for which — /?* was determined was73 g/1 °f blood, and the —/?J was 7-5 ± 4-42 slykes/1 (X± S.D.). Assuming an averageHct of 35%, and a plasma protein concentration of 36 g/1 of plasma, there were23 g plasma protein/1 of blood. If the factor — /?* for whole blood is divided by theg/1 of plasma protein plus [Hb] in whole blood, a value of 0-08 slykes/g protein I"1 isobtained. The non-bicarbonate buffering power ( — /?J) of plasma, expressed asslykes/g protein I"1 = 0-09. A comparison of these two values indicates that plasmaproteins and Hb have similar buffering power per unit weight.

~Ax>i> which is a measure of both bicarbonate and non-bicarbonate bufferingpower, was 1-18 + 0-04 in plasma, whereas in whole blood —/?co, w a s i"3i±o-i4.Therefore, even though blood [Hb] is quite variable and leads to great variability in—/?J in trout blood, changes in Hb content have a smaller effect on total bloodbuffering power (as measured by — ̂ co,) than might be expected.

No physiological significance could be ascribed to the differences in [Hb] andstandard and actual bicarbonate values which were found between strain H andstrains O and 15. A number of workers (Robinson et al. 1976; Falk & Dunson, 1977;iwarts, 1978) found strain 15 trout to be much more acid tolerant than strain O, andwe had originally hoped to find a physiological basis for the difference in tolerance. We

Tab

le 4

.

spec

ies

A c

ompa

rison

of

bM

buffe

r va

lues

foun

d by

us

and

thos

e re

port

ed b

y ot

her

inve

stig

ator

s R

efer

ence

Blood acid-base balance in brook trout 125

were not successful. The fact that [Hb] and bicarbonate values were lower in strain Hcould not be interpreted in terms of resistance to acid environments since this strainhad not been tested for acid tolerance as strains O and 15 had been.

This work was supported by N.S.F. grant BMS 7406257 and the GeorgeWashington University. We would like to thank D. Graff, T. Dingle, and W. Kennedyof the Pennsylvania Fish Commission and Dr J. E. Wright, Professor of Genetics,Pennsylvania State University for providing trout for this study. We would also like tothank Dr George M. Hughes, Professor of Zoology, University of Bristol and DrWilliam A. Dunson, Professor of Biology, Pennsylvania State University, for theirhelp at various stages in this project, as well as Ms Adrienne Fleury for her aid inmanuscript preparation.

REFERENCES

BLAXHALL, P. C. & DAIBLEY, K. W. (1973). Routine haematological methods for use with fish blood.J. FishBiol. 5, 711-781.

BURTON, R. F. (1973). The roles of buffers in body fluids. Retpir. Pkysiol. 18, 34-42.CAMERON, J. N. & RANDALL, D. J. (1972). The effect of increased ambient CO, content and pH in

rainbow trout. J. exp. Biol. 57, 673—678.DAVENPORT, H. W. (1975). The ABC of Acid-Base Chemistry, 6th ed. Chicago: The University Press.EDDY, F. B. (1974). In vitro blood carbon dioxide of the rainbow trout, Salmo gairdneri. Comp. Biochem.

Physiol. 47 A, 129-140.EDDY, F. B. (1976). Acid-base balance in rainbow trout (Salmo gairdneri) subjected to acid stress.

J. exp. Biol. 64, 159-171.EDDY, F. B. & MORQAN, R. I. G. (1969). Some effects of carbon dioxide on the blood of rainbow trout

Salmo gairdneri, Richardson. J. Fish Biol. 1, 361—372.FALK, D. L. & DUNSON, W. A. (1977). The effects of season and acute sub-lethal exposure on survival

times of brook trout at low pH. Water Res. 11, 13-15.HOLETON, G. F. & 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.HOUSTON, A. H., CZBRWINSKI, C. L. & WOODS, R. J. (1973). Cardiovascular-respiratory activity during

recovery from anesthesia and surgery in brook trout (Salvelinus fontinalis) and carp (Cyprinuscarpio).J. Fish Res. Bd Can. 30, 1705-1712.

JANSSEN, R. G. & RANDALL, D. J. (1975). The effects of changes in pH and Poo, in blood and water onbreathing in rainbow trout, Salmo gairdneri. Resp. Physiol. 35, 235-245.

MILNE, R. S. & RANDALL, D. J. (1976). Regulation of arterial pH during fresh water to sea watertransfer in the rainbow trout (Salmo gairdneri). Comp. Biochem. Physiol. 53A, 157-160.

PACKER, R. K. (1979). Acid-base balance and gas exchange in brook trout (Salvelinus fontinalis) exposedto acidic environments. J. exp. Biol. 79, 127-134.

PALMA, A. V. (1971). Procedures in Clinical Chemistry, pp. 85-87. Philadelphia: Lippincott.RANDALL, D. J. & CAMERON, J. N. (1973). Respiratory control of arterial pH as temperature changes in

rainbow trout Salmo gairdneri. Am. J. Physiol. 23$, 997-1002.ROBINSON, G. D.t DUNSON, W. A., WRIGHT, J. E. & MAMOLITO, G. E. (1976). Differences in low pH

tolerance among strains of brook trout (Salvelinus fontinalis). J. Fish Biol. 8, 5—17.SEVERINGHAUS, J. W. (1965). Blood gas concentrations. In Handbook of Physiology, vol. 2 (ed. W. D.

Fenn and H. Rahn), pp. 1475-1487. Washington: American Physiological Society.SBVKRINGHAUS, J. W.( STUPFEL, M. & BRADLEY, A. F. (1956). Variations of serum carbonic acid pK'

with pH and temperature. J. appl. Physiol. 9, 197-200.SIOGAAHD-ANDERSBN, O. (1962). The pH-log Poo, blood acid-base nomogram revised. Scand. J. din.

Lab. Invest. 14, 598-604.SIGGAARD-ANDERSEN, O. (1963). Blood acid-base alignment nomogram. Scand. J. din. Lab. Invest. 15,

211-217.SIGOAARD-ANDERSEN, O. (1964). Symposium: pH and Blood Gases. Presented at the 1964 Annual Meeting

of the Canadian Society of Clinical Chemists, pp. 2-15. London and Ontario: Bach-Simpson.SIGOAARD-ANDERSEN, O. & ENGEL, K. (i960). A new acid-base nomogram. An improved method for the^ calculation of the relevant acid-base balance data. Scand. J. dm. Lab. Invest. 13, 177-195.BINGER, R. B. & HASTINGS, A. B. (1948). An improved clinical method for the estimation of disturbances

of the acid-base balance of human blood. Medicine 34, 51-95.5 EXB79

126 R. K. PACKER AND A. L. SUNKIN

SOIVIO, A., NYHOLM, K. & HUHTI, M. (1977). Effects of anesthesia with MS 222, neutralized MS 222Iand benzocaine on the blood constituents of rainbow trout, Salmo gairdneri. J. Fish Biol. io, 91—101.

SMITH, L. S. & BELL, G. R. (1964). A technique for prolonged blood sampling in free swimmingsalmon. J. Fish Res. Bd Can. a i , 711-717.

SMITH, L. S. & BELL, G. R. (1967). Anesthetic and surgical techniques for the pacific salmon. J. FishRes. Bd Can. 24, 1579-1587.

STEEL, R. D. G. & TORRIE, J. H. (i960). Principles and Procedures of Statistics, pp. 112-114. New York,and London: McGraw-Hill Book Company, Inc.

SWARTS, F. A. (1978). Factors involved in the resistance of brook trout (Salvelinus fontinalis) to sulfuricacid solutions and mine acid polluted waters. M.Sc. thesis, Pennsylvania State University.

TIETZ, N. W. (1976). Fundamentals of Clinical Chemistry, pp. 934-972. Philadelphia, London andToronto: W. B. Saunders Company.

WOLF, K. (1963). Physiological salines for fresh water teleosts. Progve Fish Cult. 2$, 135-140.