Changes in haematocrit value are known to occur during hypoxia of rainbow trout and this has sometimes been interpreted as a result of an increase in red cell volume which is observed in vitro following equilibration with gas mixtures of low P02 (Black & Irving, 1938; Soivio, Westman & Nyholm, 1974). The possibility that there is also an increase in red cell number is still under discussion (Thomas & Hughes, 1982). The rise in blood haemoglobin content associated with such an increase would have physiological advantages for oxygen transport. However, an increase in haematocrit may increase resistance to blood flow and could impair the supply of oxygen at the tissue level. It is of interest, therefore, to know what are the effects of hypoxia both environmentally induced and under in vitro conditions on the flow properties of fish blood, for such effects have received little attention. This paper summarizes measurements of red cell deformability following changes from normoxic to hypoxic conditions, both in vivo and in vitro. A technique of filtration through a Nuclepore membrane has been used to provide an index of red cell deformability. During studies on yellowtail blood (Hughes, Kikuchi & Watari, 1982) variability in red cell deformability appeared to depend upon conditions of sampling and differences were observed in the filtration time of blood samples from normoxic and asphyxic fish. This effect has now been studied under more controlled conditions in rainbow trout.

Fish kept at 15 ± 1°C were cannulated through the dorsal aorta. Samples were taken immediately, while the fish was still under MS222 anaesthesia (0·1 g l-1), or some days later after complete recovery from surgery. Blood samples were taken from the same specimens kept in well aerated water or following the bubbling of nitrogen through water in the holding tanks so that fell to 40-60 mmHg. The particular level to which the hypoxia was extended varied according to the response of individual fish. In some individuals a lowering of below 60 mmHg initiates quite strong movements and such effects were avoided during these experiments. The slightest indication of any increase in locomotory activity was taken as a warning when to reduce the N2 bubbling and to maintain that particular . In this way the results are considered to show the effects of lowering of blood without complications due to effects induced by increased activity. The reduction from normoxia to this hypoxic condition occurred over a period of 30 min. No blood samples were taken from a fish until it had been held at the hypoxic level for at least 30 min. In no cases were fish maintained at the lower for longer than 2h. Blood samples obtained under nor-moxic and hypoxic conditions were equilibrated in vitro with N2/air or O2/air mixtures respectively. Syringes containing the blood and gas mixture were shaken as in a tonometer. In this way a comparison was made of the pore passage time of blood at similar values, in one case from fish equilibrated at low and the sample equilibrated in vitro at high and vice versa.

The filtration technique is essentially that summarized earlier (Hughes et al. 1982) and in more detail by Kikuchi, Arai & Koyama (1983). The Nuclepore filters contain pores of 8 μm diameter, the blood flow was due to a pressure difference of 10 cm H2O. Timing was by an electronic clock and in all cases measurements were preceded by control measurements using saline which was left in the pores of the filter so as to reduce variability due to enclosed air bubbles. From the data obtained for blood passage time (BPT), calculations were made of the pore passage time for individual red blood cells (RBC—PPT) using equations already given (Hughes et al. 1982). Duplicate haematocrit measurements and red cell counts were carried out from which mean corpuscular volumes were calculated. Mean values were derived and paired t-tests carried out for the blood obtained from the same fish, or for the same samples when they were equilibrated under normoxic or hypoxic conditions in vitro.

From previous experiments it was known that the results obtained from individual fish varied. In particular there was a range in normoxic haematocrit values which was reflected in the passage times (Fig. 1). Although the mean values summarized in Table 1 are of interest, the wide standard errors tend to obscure changes which were very clear when results using blood from individual fish were analysed separately. This is apparent when data for individual specimens is plotted as in Fig. 1. For nearly all individuals the following changes were observed:

Table 1.

Summary of results of measurements using blood samples from Salmo gairdneri kept under normoxic and hypoxic conditions

Summary of results of measurements using blood samples from Salmo gairdneri kept under normoxic and hypoxic conditions
Summary of results of measurements using blood samples from Salmo gairdneri kept under normoxic and hypoxic conditions
Fig. 1.

Relationships between measurements of haematocrit value (%), blood passage time (s/0·3 ml) and the pore passage time (ms) for single red cells in blood samples taken from trout under normoxic (N, open symbols) and hypoxic (H, closed symbols) conditions. Measurements obtained for individual fish are joined.

Fig. 1.

Relationships between measurements of haematocrit value (%), blood passage time (s/0·3 ml) and the pore passage time (ms) for single red cells in blood samples taken from trout under normoxic (N, open symbols) and hypoxic (H, closed symbols) conditions. Measurements obtained for individual fish are joined.

  1. There was a decrease in passage time for 0·3 ml of blood (BPT) through the S/tm pores.

  2. The haematocrit value (HCt) of samples from fish equilibrated in hypoxic water was higher than that of blood samples from the same specimens under normoxic conditions.

A decrease in the passage time calculated for individual red blood cells (RBC—PPT) was observed, and this was in spite of an increase in mean red cell volumes. As is apparent from the analyses summarized in Table 1 these differences in passage time between samples taken from the same individuals were significant at the 0·1 % level when compared using Student’s paired t-test.

Similar changes to those observed in trout equilibrated with normoxic water and hypoxic water were found when comparisons were made between blood samples taken post-operatively under anaesthesia, when the arterial was low (10-25 mmHg), and following recovery. These results are set out in Fig. 2 and again paired t-tests showed these differences to be statistically significant.

Fig. 2.

Comparison between the RBC pore passage time of blood samples obtained from trout under three conditions: anaesthesia (), normoxia (○), hypoxia (•). Lines join measurements obtained from the same individuals.

Fig. 2.

Comparison between the RBC pore passage time of blood samples obtained from trout under three conditions: anaesthesia (), normoxia (○), hypoxia (•). Lines join measurements obtained from the same individuals.

Blood from normoxic trout equilibrated with high nitrogen-containing mixtures seemed to show an increase in RBC pore passage time (Fig. 3A) but statistical analysis showed that this was not significant at the 5 % level. Blood taken from hypoxic fish when equilibrated at high values showed no change in mean RBC pore passage time (Fig. 3B).

Fig. 3.

Mean values (±s.d.) for RBC pore passage time obtained from blood samples taken from: (A) normoxic fish and equilibrated with low oxygen gas mixtures (PO2 < 5mmHg), (B) blood samples taken from fish in hypoxic water and equilibrated with high oxygen mixtures (PO2 > 160 mmHg). Arrows show the direction of the in vitro change in PO2.

Fig. 3.

Mean values (±s.d.) for RBC pore passage time obtained from blood samples taken from: (A) normoxic fish and equilibrated with low oxygen gas mixtures (PO2 < 5mmHg), (B) blood samples taken from fish in hypoxic water and equilibrated with high oxygen mixtures (PO2 > 160 mmHg). Arrows show the direction of the in vitro change in PO2.

The most striking result of these experiments is the finding of an increase in deformability of red cells in blood from hypoxic fish. Furthermore, this occurred in spite of an increase in cell volume associated with a rise in haematocrit. That no significant increase in deformability was observed in vitro, when blood from normoxic fish was subjected to comparable lowering of , suggests that changes other than corpuscular volume alone are involved in vivo.

From results obtained by previous investigators with rainbow trout under hypoxia there are indications that a rise in haematocrit may be due to at least two other causes : (a) increase in red cell number, by the liberation of fresh red cells from organs such the spleen and (b) a reduction in total plasma volume. Exactly what has happened in the present experiments is not certain, but there is little doubt that haematocrit increased and was accompanied by an increase in mean red cell volume. Both of these features are normally associated with an increase in the time for a given volume of blood to pass through a Nuclepore filter containing 8μm pores. The average dimensions of trout red blood cells are 9·0×11·0μm and it is clear, therefore, that the red cells must undergo deformation. The results from blood samples taken from individual trout during hypoxia clearly involve several other factors which could change haematocrit, whereas for in vitro samples an increase in red cell number is not possible nor can there be any reduction in plasma volume. Thus any change in haematocrit must be associated with a change in corpuscular volume. The absence of any statistically significant changes in RBC—PPT of in vitro samples, therefore, tends to emphasize the possibility that liberation of RBCs and haemoconcentration are involved in the changes which occur in the whole animal.

Additional factors which could be involved in the whole fish are changes in the composition of the RBC population, plasma proteins and inorganic constituents of the blood, or in the concentration of circulating catecholamines. Fish red cells are, in fact, known to be more sensitive to catecholamines than are those of mammals (DeVries & Ellory, 1982). The fact that changes in red cell deformability were observed in samples obtained from fish that had fully recovered from hypoxia, but without marked change in blood pH, indicates that at least some of these factors are not involved.

It would appear that during hypoxia trout blood undergoes several adaptive modifications. An increase in cell size must bring them closer to the water/blood barrier and hence reduce the plasma component of the resistance to oxygen uptake. Any increase in haematocrit resulting from a release of RBCs would raise the O2 carrying capacity. It is now apparent that these changes are accompanied by an increase in RBC deformability which effectively reduces the resistance to blood flow through the gills and other parts of the microcirculation. In view of the known differences in response of the trout ventilatory and cardiovascular systems to different regimes of hypoxia, a more detailed study of rheological aspects of their adaptation would elucidate further details of the modifications indicated by the present study.

Black
,
E. G.
&
Irving
,
L.
(
1938
).
The effect of hemolysis upon the affinity of fish blood for oxygen
.
J. cell, comp. Physiol
.
12
,
255
262
.
Devries
,
A. L.
&
Ellory
,
J. C.
(
1982
).
The effect of stress on ion transport in fish erythrocytes
.
J. Physiol.,Lond
.
324
,
51P
.
Hughes
,
G. M.
,
Kikuchi
,
Y.
&
Watari
,
H.
(
1982
).
A study of the deformability of red blood cells of a teleost fish, the yellowtail (Serióla quinqueradiata), and a comparison with human erythrocytes
.
J. exp. Biol
.
96
,
209
220
.
Kikuchi
,
Y.
,
Akai
,
T.
&
Koyama
,
T.
(
1983
).
An improved filtration method for red cell deformability measurement
.
Med. Biol. Eng. Comput
.
21
,
270
276
.
Sorvio
,
A.
,
Westman
,
K.
&
Nyholm
,
K.
(
1974
).
Changes in haematocrit values in blood samples treated with and without oxygen: A comparative study with four salmonid species
.
J. Fish Biol
.
6
,
763
769
.
Thomas
,
S.
&
Hughes
,
G. M.
(
1982
).
A study of the effects of hypoxia on acid-base status of rainbow trout blood using an extracorporeal blood circulation
.
Respir. Physiol
.
49
,
371
382
.