Camp.
Biochem.
Vol. 1024 No. 2, pp. 31 l-321, 1992
Physiol.
0300-%29/92 SS.00 + 0.00 @ 1992 Pcrgamon Press Ltd
Printed in Great Britain
CO~PA~T~VE
HEMATOLOGY
IN MARINE FISH
DANILO WILHELM FILHO,+ GUNTHERJmsm EBLB, GILSON KASWER, FRANCISCOXAVIER CAPRARIO,ALCIR LUIZ D&f and MASANAO Gmw#
~pa~ento de Biologia, and ~~pa~mento de Ci&ncias Estatisticas e da Computatio, Univ~i~de Federal de Santa Catarina, Florianbpolis, 88049, SC, Brazil. Telephone: (0482) 319226. (Received 17 September 1991) Abstrsct-1 . A comparative study involving 80 species (14 ray, 14 shark and 52 teleost species) of marine fish found at the southeastern Brazilian coast is presented. 2. Active species displayed higher values for all hematological parameters studied when compared to the less active forms. 3. Mean values of hematocrit, hemoglobin concentration and red blood cell counts increased according to the sequence: rays, sharks, teleosts. 4. As a group, cartilaginous fish blood displayed larger and fewer erythrocytes containing more hemoglobin than teleosts; mean cell hemoglobin concentration was significantly higher in rays and sharks than in teleosts. 5. For all but the hemoglobin concentration, the hematological values studied revealed a marked contrast between bony and ~tila~nous fishes which suggests distinct ways to accomplish their oxygen demands.
INTRODUCITON
Hematological analysis in fish is always difficult to interpret due to intra and interspecific variations. These discrepancies are attributable to many causes that involve mainly the following features: blood sampling and laboratory procedures (Hesser, 1960, Schaefer, 1961; Dacie and Lewis, 1968; Ferrando et al., 1981; Soivio and Nikinmaa, 1981; Hlrdig and Hiiglund, 1983; Duthie and Tort, 1985; Rail0 et al., 1985); seasonal oscillations (Denton and Yousef, 1975; Bridges et al., 1976; Cech and Wohlschlag, 1982; H&dig and Hoglund, 1984; Andersen et al., 1985; L~hmiller et al., 1989); size and ontogeny (Potter et al., 1974; Potter and Beamish, 1978; Clark et al., 1979; Macey and Potter, 1981; Aldrin et al., 1984); genetic determinants (Benfey and Sutterlin, 1984; Murray, 1984); sex (Atkinson and Judd, 1978; Murray, 1984); population density (Burton and Murray, 1979; Murray and Burton, 1979; Murray, 1980); geographical distribution (Wintrobe, 1933; Haws and Goodnight, 1962; Bridges et al., 1984); habitat (Haws and Goodnight, 1962; Val et al., 1985) and exercise and stress (Schaefer, 1961; Yamamoto et al., 1980; Swift, 1982; Wood et al., 1983; Ling and Wells, 1985; Torres et al., 1986). Despite the massive information available in the literature concerning fish hematology, only few works refer to a comparative approach comprising broad taxonomic fish groups, or attempt to find novel insights in this field (cf. Wells et al., 1980; Graham et al., 1985). In the present study more than 500 specimens belonging to 80 different marine fish species found in the southwest Atlantic (southeastern coast of Brazil) *To whom all correspondence should be addressed. $Present afBliation: Department0 de Ciencias FisiolcIgicas, UFSC, FlorizuGpolis, 88049, SC, Brazil.
analysed in a comparative way concermng hematological parameters and indices. Possible relationships between cardiovascular and hematological features of each major fish group are suggested.
were
MATERIALS AND METHODS
Blood samples were obtained from adult forms through dorsal aorta puncture in teleost and shark individuals, and through cardiac puncture in rays, using minute amounts of heparine as anticoagulant, and were immediately stored on broken ice (co 2’C). Most Esh were already dead when bled, and were caught by artesanal fishery in the vicinities of Santa Catarina Island (27”40’ south latitude; 48”30’ west longitude), south Brazil. Hematological procedures were carried out essentially according to Hesser (1960) and Dacie and Lewis (1968). The hematocrit was determined by the ~cromethod using cap illary tubes (70 x 1 mm) and centrifuged at 5OOOg for 10min (Fernando et al., 1981; H&dig and Hoglund, 1983). The cyanomethemoglobin form was employed for blood hemoglobin concentration evaluations. Red blood cell counts were performed with improved Neubauer chambers, comprising 400 small square counts, to minimize errors. AH the three parameters were evaluated in duplicates or triplicates. Hematological indices were calculated as usual. Exploratory data analysis (EDA) (Tukey, 1977) and regression analysis were employed together with a Statgraphics 2.6 computer program, using a 95% confidence interval. Despite the obvious distortions associated with sample unevenness concerning each species analysed, and considering that many species here studied occur with a very low frequency, ~metirn~ represented by only one or two individuals in a sampling period of 5 years (e.g. Alopias lapis, Isurus oxyrhinchus, Squalus c&en&, Rhinobatos percellens, Astroscopus ygraecum and Echeneis mucrates), mainly due to the fishing methodology employed by the &shermen, the principal sample goal in the present work was to increase the species diversity for comparative purposes.
311
312 RESULTS
IHSCUS!BON
Table 1 contains the hematological parameters and corresponding indices concerning 80 species studied. All the values found in the present study are within the range already described in the literature. More active fish possess higher hematocrits, hemoglobin concentrations and red cell counts when compared with less active forms. Examples for the first category are the endotherm sharks Isurus uxyrhynchus and Alopias yulpimcs and the scombrids, besides the non-endothe~s SFhyrna zygaena, S. lewini, Carcharhinus porosus and C. macuiipinnis, and also the teleosts Pomatomus saltator, Trichiurus lepturus, ~ligoplites satiens and Seriala ~al~di. Conversely, lower hematolo~cal values are related to sluggish benthonic or epibenthonic species like Astroscopus sexspinasus and A. ygraecum, Paralichthys brasiliensis and Prionotus
The early and classic comparative study of Wintrobe (1933) suggested that MCHC was the only relatively uniform hematogical variable in a variety of vertebrates studied, the single exception being the class of fishes which exhibited higher inter and even intraspecific MCHC variations. Nevertheless, our study showed very small intraspecific oscillations, indicating that, at least for matured specimens of similar age and length, MCHC is fairly constant in ahnost all species examined (see Table 1). On the other hand, we found a higher mean MCHC value for cartilaginous (27.2 g%) than for bony (21.2 g%) fishes. This intergroup distinction was not revealed in that work, probably due to sample limitations concerning elasmobranch species. Furthermore, the values of this variable revealed in each fish group were fairly constant (Table 2). The MCHC values of rays and sharks examined, and to a lesser extent the teleost values, were slightly lower and closer to those relating to humans and mammals. Unlike the enucleated mammalian red cells, the nuclei of fish erythrocytes are very large and can comprise one fourth or even more of the cell volume (Wintrobe, l933; Coburn and Fisher, 1973). This could explain, in purely physical terms, the lower fish MCHC values compared to other vertebrates. The relative constancy of MCHC verified within and among species belonging to each major group could perhaps better be explained as a reflection of developmental constraints (i.e. the intrinsic high correlation between Hb and Ht; see Fig. 7) than as adaptive adjustments to en~ronment. It is tempting to attribute at least part of the existence of the high hemoglobin heterogeneity that typify fish in general (Reischl and Tondo, 1974; Fyhn et al., 1979; Wilhelm Filho and Reischl, 1981; Perez and Rylander, 1985) to counteract the solubility limitations that high hemogiobin con~ntrations impose to blood oxygen capacitance (Riggs, 1976). An interesting inverse correlation between MCHC and latitude was described by Wells et al. (1980). In that hypothesis, four ranges of MCHC values were confronted with latitude. Mean cell hemoglobin concentration values for the species here examined (390-143 g/l) fall within the tropical/temperate/cold ~tabIished in that work temperate ranges (366-150 g/l). These discrepancies could probably be explained by the following features: the small number of species (11) used to characterize a geographical distribution of more than 60” of latitude, compared to nearly 40 species almost restricted to polar and subpolar areas (Larsson et al., 1976; Wells et al+, 1980), the large geographical distribution that characterizes many fish species and/or specific hematological determinants related to habitat and habit concerning fish species inhabiting different latituoes. The use of MCHC as an indicator of blood oxygen transport capacitance shouid be used with caution, keeping inferences concerning this hematological character more restricted to the i~~~llular microenvironment adjustments of oxygen transport (Weber and Wells, 1989). Blood hemoglobin concentration and the related intrinsic 0, affinity, besides the rheological and diffusional implications of the hematocrit
punctatus.
Comparing cartila~nous with boned fishes (Table 2), it is apparent that they behave as two distinct groups for all the parameters and indices examined (Figs l-6), with the exception of the blood hemoglobin concentration (Fig. 21, where the sharks exhibited an intermediary position between rays and teleosts. This intergroup contrast is more evident when shown through the means and the corresponding confidence intervals (Figs l-6B). Cartilaginous fish have larger (ca 3-4-fold) and fewer erythrocytes than the teleosts, thus confirming previous data in the literature (Wintrobe, 1933; Saunders, 1966; Pica er al., 1983). As a consequence, they have a cl-S-fold higher mean ceil hemoglobin content (MCH) than the teleosts, meaning that, in spite of the lower blood hemoglobin concentration, their large erythrocytes contain more hemoglobin than the teleost ones. The intrae~thr~ytic hemoglobin concentration values revealed for sharks, rays and teleosts were 4.4, 4.2 and 3.3 mM, or, as expressed in mean cell hemoglobin concentration (MCHC!) values, 28.4, 27.0 and 21.2 g%, respectively (Table 2). Regression analysis revealed high correlations between hemoglobin concentration (Hb) vs hematocrit (Ht) parameters (Fig. 7; correlation coefficient = 0.84), a feature frequently reported for fish (e.g. Wintrobe, 1933; Schaefer, 1961; Swift, 1982). On the other hand, the high correlation found for erythrocyte counts vs hematocrit (r = 0.81), and also between MCH vs mean cell volume (MCV) (r = 0.93) is rarely found in the literature con~ming fish (Figs 8 and 9, respectively). Less active teleosts generally occupy a clearly separate place in relation to more active ones (most plots below the regression line in Fig. 9, for instance). The variables Hb and red cell counts revealed different regression fitness for cartilaginous fish (r = 0.59) and for teleosts (r = 0.72) (Fig. 10). Nonlinear regression fits better for both red cell counts vs erythrocytic volume and for cell counts vs hemo~obin content (Figs 11 and 12, respectively. Intraspecific variations of MCHC were, in general, relatively low, some species exhibiting remarkably low oscillations, especially in the teleosts.
Iimbotas
Carcharhh mactd&~is Carcharhitu miltvrti Catvharhitw ob.wtww Ciuchnrhirau poro~us Family Lmnidae Ioxyrinchw Family Odontaspididae odtmtaspis tFamily Sphymidae SpAynra Inulnd WY@= tyEIreM Family Squalidae $qtI&Kc!l&i%T& Family Squatinidae sq#ari?la atgentkka Family Triakidae Mtstefw fosciaus Mwtelus schmitti
Cm&rhimss
R&a qc~ophora Rq&zpkltmre Sympterygio acuta Sympterigia banapartei Family Rhinobatidae ~~~~ kke#t RkWtoJ percelteru zqpleryx bre&astrik Family Rhinopteridae Rhitwtera botwsw Ofdez EkpLliformes Family Alopiidae Alopiar vul~inuc Family Carcbarhinidae
Rgja cmteimnd
Rqjiog&zf
Class choadrichthyes Subetass ~~ob~~h~~ onfcl RajifDrmcs Family Dasyatidae Dasyalli gtittatu Dtuyaiis say Family Myliobatidac MyUotvtis go&i Family Narcinidae
Taxonomicai position and scientific names
0.62 i: 0. I4 (6) 0.70 k 0.21(8) 0.41 + 0.18 (4) 0.91 (I) 0.64 + 0.18 (2) 0.73 + 0.18 (5) 0.50 & 0.15 (7) 0.95 f 0.15 (5) 0.76 + 0.18 (4) I.04 k 0.31 (2)
1.84(l) I.31 +0.37(2) f.i2+0.5O(S) 0.58+0.1i (2) 0.75 (I) 0.30 + 0.08 (2) 2.17(l)
17.1 rt &l(8) 17.7 & 6.1 {W) 15.4 + 5.3(M) 17.9 + 2.7(2) 16.3 & 4.445) 21.6+ 3,6(8) 13.8 + 3.0 (9) 18.9 + 7.0 (7) 13.341) 19.0 k 5.7 (5) 25.1 + 5.743)
33.0(l) 22.3 + 8.6 (3) 30.1 t 12.0(5) 16.1 +?.0(2) 15.0(l) 29.9 rt 6.3 (7) 28.7 + 10.6 (2)
0.173 +0.081 (IO) 0.298 (I) O.Ul +0.127(4)
0.66 + 0,17(10) 0.76(l) 0.65 rf;5.32 (2)
23. I + 8.7 (14) 23.5 t 2.1 (2) 20.4 & 8.2 (4)
0.250(l)
1.13(l)
31,0(I)
738.3 (I) 618.8 f Xti.0 (4)
1496.6 f 737.6 (10)
1192.3(l)
475Sf 126.Q(I8) 516.8 It 215.2 (18)
0.592fO.l75(18) 0.559rt0.2H1(18)
22.2(l) 22. I f 4.0 (2)
24.6& 13.O(lO)
23.4(i)
34.3 f 6.9 (14) 28AI f 6.6 (20)
28.3 (1)
780.2 * 294-7 (3) 0.297 f 0.072 (3)
0.91 (I) I.30 i 0.22 (14) 1xl2 f: 0.27 (20)
35.9(I)
563.5 k I 19.4 (2) 0.492 + 0. IS6 (2)
21.9 + 3.5 (3)
27.8 f 6.3 (2) 26.5 & 5.2 (5) 31-l f &f(2) 32.4 (1) 24.8 f 3.3 (2)
717,4(l)
36.0 (I)
28.6 k: 0.1(2)
1I.3 (3)
475.7 k
26.0 + S-9 (5) 32.9(k)_ 24.7 f 8.6 (4)
23.1 f 5.5 (8) 2&s f 2.8 (8) 36.6(l) 23.4 f 5.2 (2) 22.6 f 5.1(S) 24. I & 7.0 (7)
22.7 k 6-O(6)
29.8 & 9.3 (3)
38.9 f SS (3) 23.5 f 2.1 (3)
MCHC (g%)
618.1 f fZS.4(3) 461.8 (I) 592.5 f 75.4 (4)
649.8f211.8(3) 809.7 + 97.3 (4) 617.1 k216.015)
597.3 rt 152.0(S) 784.6 & 230.2 (8)
1212.8 f 336.2 (4)
581.8 + 298.4 (4)
775.5 4 517.7 (2) 579.3 f 64.6 (3)
MCV (a’)
563.5 f 124.9 (3) 596.8 g: 115.8 (3) 444.0 Ifr Il8.3 (2) 388.611) 523.7 I: 135.0 (4)
0.389 * 0.270 (3) 0.522 & 0.043 (3) 0.353 2 0.064 (2) 0.386(l) 0.508 f 0,038 (4)
0.460(l)
0.527 +- 0.108 (3)
0.303 4 0.066 (3) 0.288 (1) 0.343 f 0.058 (4)
0.242 $020 (3) 0.259 f 0.038 (4) 0.220 + 0.063 (5)
0.257 + O.OS4(5) 0.212 ri 0.105 (8)
0.165 f 0.064 (4)
0.368 & O.lSl (4)
0.284 + 0.082 (2) 0.278 f 0.012 (3)
RBC (lO’/mm’)
27.3 + 7.4 (18) 25.4 f 8.8 (23)
0.68(I)
0.76 k 0.25 (3)
18.0 + 6.3 (7)
-
1.27 F 0.25 (3) 0.56 +O.I0(3)
Hb (mM)
standard deviations {when applicable) concerning 80 specks studied
21.7&6.6(3) 14.3 ): 3.6 (4)
Ht 1%)
Table I. Hematological values and corresponding
iLO(2)
conritaued
164.2 (I) 149.4 f 47.6 (2)
391.5 f 192.4(10)
278.9 (I)
ISo.9&4444.s(14) 127.6 f X3(18)
2oP.3(1)
232.6 (1)
113.7;
117.4 f 30.4 (2) 143.1 f 74.1 (3) 127.3~74:;; (2)
257.8 (I)
135.4 f 4.7 (2)
218.9 It: 78.4 (3) 152.1(l) 142.2 f 30.6 (4)
185.7 f 41.8 (2) I%.3 f 18.6 (4) 146.4 f 36.0 (3)
132.3 & 36.0 (S) 167.3 f 14,7 (4)
289.3 f 136.5 (4)
252.9 + 46.1 (3)
132.0+ 17.1 (3)
298.5 2 160.0(2)
MCH (P&
z W
Taxonomid position antlsdentiknamcs
17.3119
16.3(i)
414S(l) -
n.c.
@.538(i)
n.e. 2.356 & 0.746 (4) 2.816f0.655(17)
0.44(i) 0.69 (I)
1.07(t) 1.14 & 0.07 (4) 1.61 +0.34/17) 1.63 ($9
i.oai~o.30~2j 2.10(i) 274(i) 201(i) 1.63 & 0.43 (4) 1.06~1) i.~iO.23(6) I.%(1) 1.92(i) 0.97(l) i.90f0.46{7) I .a f 0.35(7) I&?(1) O.%f 0.10 (2)
12.2(i)
20.7 + 2.3 (2)
42.6(l) 42.9 + 5.9 (7) 52.4 f I I.3 (24) 6Li (if 35.7 + 5.5 (3) 33.9 + 8.3 (2) 33.0 f Il.8 (2) 0.5 F 10.6(Z) 63.5(i) 61.0(i) 49.2* i3.3{4) 30.0(l) 56.9f 11.7(6) 53.8(i) 426 (I) 31.1(l) 50.1 f 12447) 44.4 f le.0 (7) 33.0(1) 34.0 f 2.8 (2)
156.8 (I) 195&l) 105.8 (1) ni *of 86.2 (3) 225.3 (I) 111.2(l) 165.8 (I) 108.7 f 8.4 (3) 171.4 f 3&O(4)
4.336(i) n.t. i.G-&) 2.835 (1) 2.652 rt 1.W (3)
1.&0((l)
z::; i:llsfi(l) 3.966 f 0.170(3) 2y;o;i;; (4)
177.4(l)
119.7 (1)
237.9 f 49.2 (2) 179.7(l) IM.3 (I)
1A88+ 0.023(2) i.aco(l) 1.574(i)
217.4(l)
ZllM(i)
185.1 f 38,9(17)
189.5 f 45.1(4)
18.2 f 0.3 (29
Et g; 25.2i4.i (7) ~~i~~~~(~
19.9(l) 28.1 (I) 2L2{1) la.?* 3.314) 22.7(l) 20.5& 1.1(6) 23.8 (I)
17.4ji: 1,2(3) 21.1 &4.9{2] Zl.!if i.8(2)
16.9(r)
2i.2& 4.6(17)
17.3*2.8(49
23.4(l)
14.6 + 2.8 (2)
-
n.e.
14.3 4 2,3 (35)
193.8 f 57.6 (35)
0.68 k 0.04 (2)
1.892 f 0.521 (35)
32.4 + 6.0 (2)
-
26.7 + 2. I (2)
36.7 (I)
30&) 24.0 (I) 63.5 It 4.3 13) 53.4(i) 32.3 (I) 33.7 (1) 29.1 f 8.3 (39 38.9 f lLV(4) 25.1 (I)
31.3(l) -
44.1 f 1.3(2) 420(l) 35.5 (I)
36.7 (1)
38.0 f 7.3 (17)
37.8 f 19.7 (4)
-
71.9(i)
-
-
28.9 f 9.3 (359
74.6(l)
MCI-4 (Pg9
38.1(l)
0.83 + 0.16 (35)
19.5.8(l)
MCWC (g%)
35.5 f 7.8 (41)
47.7 + WA
Q.#i(l)
MCV (a’)
2.26 + 0.25 (21
0.77 (1)
13.0(l)
RBC (IO’/mm”)
I I.8(3)
Hb (mM)
HI (%)
C!ompatiltive hematoiogy in mariM: fish
DAMLOWILHELMFILHOet al.
316
Table 2. Mean values and standard deviations related to rays, sharks and teleosts studied Sharks
bYS
Ht (%) Hb (mM) RBC (I06/mm’) MCV @‘) MCHC (g%) MCH (pg)
17.9 f 0.75 f 0.288 f 673.5 + 27.0 f 188.4*
3.3 (14) 0.23 (14) 0.090(13) 194.2(13) 5.7 (14) 59.3(13)
24.8 f 1.09 f 0.403 * 686.8 * 28.4 f 185.0 k
5.5 (14) 0.45 (14) 0.123 (14) 305.5 (14) 4.9 (14) 80.4 (14)
TCkosts 41.0 * 1.34 f 2.289 f 200.0 f 21.2 f 39.1 f
14.3 (52) 0.58 (52) 0.904 (41) 89.8 (41) 4.6 (52) 13.0 (41)
Numbers in parenthesis denote species’ diversity. Ht = Hematocrit; Hb = hemoglobin concentration; RBC = red blood cell count; MCV = mean cell volume; MCHC = mean cell hemoglobin concentration; MCH = mean cell hemoglobin content.
and red cell size and number, together with other morphological and physiological characters related to oxygen transport system, such as the cardiovascular and ventilatory performances, are apparently more reliable in ascertaining respiratory needs of an entire organism. As expected, more active pelagic forms of teleosts and sharks showed higher hematological parameters than sedentary or less active species, a condition that enables the former species to cope with increased oxygen demands (Schaefer, 1961; Haws and Goodnight, 1962; Eisler, 1965; Cameron, 1970; Larsson et al., 1976; Emery, 1986). In this regard, regression analysis revealed that sluggish species (e.g. Prionotus punctatus, Percophis brasiliensis, Balistes capriscus and Stephanolepis hispidus, and the stargazers A. sexspinosus and A. ygraecum) are localized between
active teleosts and the elasmobranch group (open circles in the left lower comer in Figs 10 and 11). Similarly, the intermediary position showed by sharks concerning blood hemoglobin remained as the only variable that overlapped the three fish groups studied (Fig. 2A, B). This feature could be attributable to the common pelagic habit of sharks, contrasting with the sedentary and benthonic rays. An enhanced blood oxygen capacity of active fish fits well with the ca 3-fold higher oxygen extraction of endotherm sharks compared to the ectothenn forms (Emery, 1986). In this regard, elasmobranchs generally seem to extract more oxygen than teleosts (Layton, 1987). Endotherm fish, like the lamnid sharks and some scombrids, bear a countercurrent system with a rete mirabile located in their red muscles (Carey, 1982; Emery, 1986), that allows them to keep these tissues at higher temperatures than the surrounding water, and to exhibit extraordinary swimming performances. Some scombrids here studied revealed hematocrits as high as 80%. Although stress, in some cases, can be responsible for overestimations in hemoglobin content and also concerning hematocrit and red cell numbers, relatively small differences were eventually found (Swift, 1982). Doubling the hematocrit, as in the case of teleosts confronted with rays and sharks, the viscosity (n) and flow resistance (R) of blood is also doubled, according to Pouseille’s equation, thereby increasing the cardiac work in teleosts. Vascular resistance and cardiac work are further increased by the high capillary density of teleosts (Hartmann and Lessler, 1964; Schmidt-Nielsen, 1983). Nevertheless, the high prebranchial compliance of the ventral aorta of the teleosts, when compared to the elasmobranchs (Taylor, 1964; Langille et al., 1983), may constitute physiological and anatomical adjustments to com-
pensate the high blood viscosity of the former. On the other hand, elasmobranchs have broader and less dense capillaries, and possess larger and fewer red cells, besides larger volemia (ca 3 times greater than the teleosts) (Thorson, 1962; Prosser, 1973; SchmidtNielsen, 1983), that may reduce blood viscosity and peripherical resistance (Graham and Fletcher, 1983). Furthermore, elasmobranchs are also characterized by higher cardiac stroke volumes (Prosser, 1973) and lower cardiac work (Side11 and Driedzic, 1985), related to the high pericardic rigidity (Johansen and Gesser, 1986), and also by a low pre-branchial compliance of the ventral aorta (Langille et al., 1983). As a consequence, they show an increased blood flow when compared to the teleosts. The higher blood convection requirement of cartilaginous fish in this manner may counteract the modest hematological parameters of rays and most sharks (Johansen, 1971). It is interesting to verify that a similar condition is also found in the abissal (Graham et al., 1985), and also in the Antarctic teleosts (Wells et al., 1980). The blood oxygen transport capacity (CHbG& is directly proportional to the arterial-venous difference in oxygen concentration (Cao, - Cv,), and, considering that the latter is inversely related to. the cardiac output (Vb), the oxygen consumption (M%) (Weber, 1982; Dejours, 1988) is: nio, = (Cue, - Cv,)
rib i.e. h&,/vb
= Cao, - Cuop.
For the same oxygen consumption rate, the higher the oxygen capacity, the lower the blood flow (and cardiac output) rate needed. This assertion seems to fit well both the phylogenetic positions (e.g. cartilaginous vs bony fishes), and the ecophysiological features concerning each species (e.g. endotherms vs ectotherms, or active vs sluggish forms). For instance, the angelshark Squutina urgentina showed hematological relationships closer to rays than to the squaliforms, but compatible with the habit and habitat that this fish shares with the rajiforms. In short, it seems that elasmobranch and teleost fish have, as two broad groups, different gas transport systems. Instead of high cardiac work and blood pressures accompanied by numerous small red cells held in more viscous blood of higher hematocrits ascribed for teleost fishes, the elasmobranchs, and especially the rajiforms, show, in general, relatively lower energy costs related to modest cardiac work, but exhibit higher gill oxygen extraction than teleosts, higher cardiac outputs, and higher blood volemia and flow rates. These counteract the reduced oxygen transfer conductance that results from low hematocrits, despite the large red cell volumes they normally display (Yamaguchi et al., 1988). These are examples
317
Comparative hematology in marine fish %
40
30
20
10
R
mM
’
f
/
1
HB
r
T
s
mM-’ . HB
-I
lo4mm3
lOe/mm?
23
3.9
13 1.0
0.2
03
R
S
T
R
S
1
Figs l-3. Multiple box plots (A) and means and the correspobding standard errors (B) (confidence intervals of 95%) of ali hematological parameters (XT, HB and RBC counts) analysed. R = Rays; S = sharks; T = teleosts. Values according to the units employed in Tables I and 2.
DANEOWILKBLM FUIO er al.
318
P% 200
300 t20
200
100
40
R
S
t
R
S
T
Figs 4-6. Multiple box ploti (A) and meana aud the cocrcqondi~ stendard ~rrora (B) (con&iencc intervals of 95%) of the three indice (MCV, MCH and MCHC). R = Rays; S = sharks; T = tclcosts.
Comparative hematology in marine fish
, I
VII
, I,,
,,.)
.,
c,
,
3
2
1
‘
I
HB
MCH
300
200
100
300
800
900
‘200 MCV
1
2
3
4
1
2
3
4
RBC
RBC
Figs 7-12. Regression analysis related to some hematological parameters and indices. Values according to the units employed in Table 1 and 2. Rays (A); sharks (A); sluggish teleosts (0) and moderate/active teleosts (a). Outer and inner dotted lines represent prediction and confidence intervals, respectively.
of distinct mo~holo~~a1 and physiolo~cal devices of composing appropriate respiratory needs. It could be argued that the low energy costs of e~~mobr~chs in this respect imply an energy profit
that may be transferred to economic activities, as welf
as ~pr~u~on,
benefbing the existence of wide-
spread viviparity in these fishes, with all its associated advantages. Therefore, a side consequence of physi-
DANILOWmsmul FILHO et al.
320
ology could in a way be thought of as meaning an increase in general exaptation (Gould and Vrba, 1982). Irrespective of whether this is true or not, respiratory physiology tells us that the elasmobranchs are at least as successful as teleosts in what concerns physiological efficiency. Contrary to earlier speculations (see Wintrobe, 1933, and other modem literature), more akin to cultural preferences than to corroborated facts, el~mobranch fish are not necessarily “unsuccessful evolutional experiments”, or members of a group “low in the scale of evolution” (see similar ideas relating to ecto vs endotherms in Pough, 1980). Acknowledgements-We are grateful to the fishermen of Praia Plntano do Sul, Arma@o, Campeche, and Barra da Lagoa, who kindly provided the fish for blood sampling. Also to Dr Eduardo Htmreres and Carlos Eduardo Pinheiro (Universidade Federal de Santa Catarina), for the use of laboratory equipment and computer, respectively, and Dr Evaldo Reischl (Universidade Federal do Rio Grande do Sul) for comments on the manuscript. This study was supported by grants of the CNPq (Con&ho National de D~envolvimento Cientifico e T~ol~~~/Brazil, proc. 402~3/88-6 and ~33~/89-7) to DWF”. REFERENCES
Aldrin J. F., Messager J. L. and Saleun S. (1984) Analyses sanguines de turbots d’elevages immature8 (Scophtalmus maximus L.). Aquaculture 40, 17-25. Andersen N. A., Laursen J. S. and Lykkeboe G. (1985) Seasonal variations in hematocrit, red cell hemoglobin and nucleoside triphosphate concentrations, in the European eel Anguilla anguilla. Comp. Biochem. Physioi. 8lA,
Physioi. 81A, 879-883.
Eisler R. (1965)Erythrocyte counts and hemoglobin content in nine species of marine teleosts. Chesapeak. Sci. 6, 119-120. Emery S. H. (1986) Hematological comparisons of endothermic vs ectothermic elasmobranch fishes. Copeia 3, 700-705. Ferrando A., Bobadilla I. G., Bobadilla M. S., Palou A. and Alemany M. (1981) Comparative estimation of hematocrit and trapped plasma in the packed cell volume in man, rabbit and chicken blood. Comp. Eiochem. Physiol. 7OA, 611-613.
Fyhn U. E. H., Fyhn H. J., Davis B. J., Powers D. A., Fink W. L. and Garlick R. L. (1979) Hemoglobin heterogeneity in Am~onian fishes. Comp. Biochem. Physiol. 62A, 39-66.
Gould S. J. and Vrba E. S. (1982) Exaptation-a missing term in the science of form. Pu~eo&iofogy 8, 415. Graham M. S. and Fletcher G. L. (t983) Blood and plasma viscosity of winter flounder: influence of temperature, red cell concentration, and shear rate. Can. J. Zool. 61, 2344-2350. Graham M. S., Haedrich R. L. and Fletcher G. L. (1985) Hematology of three deep-sea fishes: a reflection of low metabolic rates. Comp. Biochem. Physiol. 8OA, 79-84. HIrdig J. and Hiiglund L. B. (1983) Chr accuracy in estimating fish blood variables. Comp. Biachem. Physiol. 75A, 35-40.
87-92.
Atkinson E. and Judd F. W. (1978) Comparative hematology of Lepomis mi~ro~oph~ and Ciehlasoma eyanoguttatum. Copeia 2, 230-237.
Benfey T. J. and Sutterlin A. M. (1984) The haematology of triploid landlocked Atlantic salmon, Salmo salar L. J. Fish Biol. 24, 333-338. Bridges D. W., Cech J. J. Jr and Pedro D. N. (1976) Seasonal hematological change in winter flounder, Pseudooteronectes
Dacie J. V. and Lewis S. M. (1968) Practical Haematology, 4th edition. J and A Churchill. Dejours P. (1988) Respiration in Water and Air: Adaptations-Regulation-Evolution. Elsevier, Amsterdam. Denton J. E. and Yousef M. K. (1975) Seasonal changes in hematology of rainbow trout, Salmo gairdneri. Comp. Eiochem. Physiol. SlA, 151-153. Duthie G. G. and Tort L. (1985) Effects of dorsal aorta cammlation on the respiration and hematology of Mediterranean living S~y~~orhin~ cat&u/a L. Comp. Biochem.
americanus.
Trans. Am. Fish. Sot.
18%
596~600. Bridges C. R., Taylor A. C., Morris S. J. and Grieshaber M. K. (1984) Econhvsioloaical adaptations in Blennius pholis (i.) blood io-inter~dal rockpool environments. J. exp. Mar. Biol. Ecol. 77, 151-167. Burton C. B. and Murray S. A. (1979) Effects of density on goldfish blood-I. Hematology. Comp. Biochem. Physiol. 62A, 555-558.
Cameron 3. N. (1970) The influence of en~ronmental variables on the hematolo~ of pi&h (Lagodon r~o~oides) and striped mullet (~ugiI cephalus). Comp. Biochem. Physiol. 32, 175-l 92.
Carey F. G. (1982) Warm fish. In A Companion to Animal Physiology (Edited by Taylor C. R., Johansen K. and Bolis L.), op. 217-231. Cambridge University Press, ,- __ Cambridge. Cech J. J. Jr and Wohlschlan D. E. (1982) Seasonal patterns of respiration, gill ventila%on and he&atological‘characteristics in the striped mullet, Mugil cephalus L. Bulf. Mar. Sci. 32, 130-138.
Clark S., Wbitmore D. H. Jr and McMahon R. F. (1979) Considerations of blood narameters of largemouth bass, Micropterus sabrwides. J.* Fish Biai. 14, t
Biochem.
Vol. 1024 No. 2, pp. 31 l-321, 1992
Physiol.
0300-%29/92 SS.00 + 0.00 @ 1992 Pcrgamon Press Ltd
Printed in Great Britain
CO~PA~T~VE
HEMATOLOGY
IN MARINE FISH
DANILO WILHELM FILHO,+ GUNTHERJmsm EBLB, GILSON KASWER, FRANCISCOXAVIER CAPRARIO,ALCIR LUIZ D&f and MASANAO Gmw#
~pa~ento de Biologia, and ~~pa~mento de Ci&ncias Estatisticas e da Computatio, Univ~i~de Federal de Santa Catarina, Florianbpolis, 88049, SC, Brazil. Telephone: (0482) 319226. (Received 17 September 1991) Abstrsct-1 . A comparative study involving 80 species (14 ray, 14 shark and 52 teleost species) of marine fish found at the southeastern Brazilian coast is presented. 2. Active species displayed higher values for all hematological parameters studied when compared to the less active forms. 3. Mean values of hematocrit, hemoglobin concentration and red blood cell counts increased according to the sequence: rays, sharks, teleosts. 4. As a group, cartilaginous fish blood displayed larger and fewer erythrocytes containing more hemoglobin than teleosts; mean cell hemoglobin concentration was significantly higher in rays and sharks than in teleosts. 5. For all but the hemoglobin concentration, the hematological values studied revealed a marked contrast between bony and ~tila~nous fishes which suggests distinct ways to accomplish their oxygen demands.
INTRODUCITON
Hematological analysis in fish is always difficult to interpret due to intra and interspecific variations. These discrepancies are attributable to many causes that involve mainly the following features: blood sampling and laboratory procedures (Hesser, 1960, Schaefer, 1961; Dacie and Lewis, 1968; Ferrando et al., 1981; Soivio and Nikinmaa, 1981; Hlrdig and Hiiglund, 1983; Duthie and Tort, 1985; Rail0 et al., 1985); seasonal oscillations (Denton and Yousef, 1975; Bridges et al., 1976; Cech and Wohlschlag, 1982; H&dig and Hoglund, 1984; Andersen et al., 1985; L~hmiller et al., 1989); size and ontogeny (Potter et al., 1974; Potter and Beamish, 1978; Clark et al., 1979; Macey and Potter, 1981; Aldrin et al., 1984); genetic determinants (Benfey and Sutterlin, 1984; Murray, 1984); sex (Atkinson and Judd, 1978; Murray, 1984); population density (Burton and Murray, 1979; Murray and Burton, 1979; Murray, 1980); geographical distribution (Wintrobe, 1933; Haws and Goodnight, 1962; Bridges et al., 1984); habitat (Haws and Goodnight, 1962; Val et al., 1985) and exercise and stress (Schaefer, 1961; Yamamoto et al., 1980; Swift, 1982; Wood et al., 1983; Ling and Wells, 1985; Torres et al., 1986). Despite the massive information available in the literature concerning fish hematology, only few works refer to a comparative approach comprising broad taxonomic fish groups, or attempt to find novel insights in this field (cf. Wells et al., 1980; Graham et al., 1985). In the present study more than 500 specimens belonging to 80 different marine fish species found in the southwest Atlantic (southeastern coast of Brazil) *To whom all correspondence should be addressed. $Present afBliation: Department0 de Ciencias FisiolcIgicas, UFSC, FlorizuGpolis, 88049, SC, Brazil.
analysed in a comparative way concermng hematological parameters and indices. Possible relationships between cardiovascular and hematological features of each major fish group are suggested.
were
MATERIALS AND METHODS
Blood samples were obtained from adult forms through dorsal aorta puncture in teleost and shark individuals, and through cardiac puncture in rays, using minute amounts of heparine as anticoagulant, and were immediately stored on broken ice (co 2’C). Most Esh were already dead when bled, and were caught by artesanal fishery in the vicinities of Santa Catarina Island (27”40’ south latitude; 48”30’ west longitude), south Brazil. Hematological procedures were carried out essentially according to Hesser (1960) and Dacie and Lewis (1968). The hematocrit was determined by the ~cromethod using cap illary tubes (70 x 1 mm) and centrifuged at 5OOOg for 10min (Fernando et al., 1981; H&dig and Hoglund, 1983). The cyanomethemoglobin form was employed for blood hemoglobin concentration evaluations. Red blood cell counts were performed with improved Neubauer chambers, comprising 400 small square counts, to minimize errors. AH the three parameters were evaluated in duplicates or triplicates. Hematological indices were calculated as usual. Exploratory data analysis (EDA) (Tukey, 1977) and regression analysis were employed together with a Statgraphics 2.6 computer program, using a 95% confidence interval. Despite the obvious distortions associated with sample unevenness concerning each species analysed, and considering that many species here studied occur with a very low frequency, ~metirn~ represented by only one or two individuals in a sampling period of 5 years (e.g. Alopias lapis, Isurus oxyrhinchus, Squalus c&en&, Rhinobatos percellens, Astroscopus ygraecum and Echeneis mucrates), mainly due to the fishing methodology employed by the &shermen, the principal sample goal in the present work was to increase the species diversity for comparative purposes.
311
312 RESULTS
IHSCUS!BON
Table 1 contains the hematological parameters and corresponding indices concerning 80 species studied. All the values found in the present study are within the range already described in the literature. More active fish possess higher hematocrits, hemoglobin concentrations and red cell counts when compared with less active forms. Examples for the first category are the endotherm sharks Isurus uxyrhynchus and Alopias yulpimcs and the scombrids, besides the non-endothe~s SFhyrna zygaena, S. lewini, Carcharhinus porosus and C. macuiipinnis, and also the teleosts Pomatomus saltator, Trichiurus lepturus, ~ligoplites satiens and Seriala ~al~di. Conversely, lower hematolo~cal values are related to sluggish benthonic or epibenthonic species like Astroscopus sexspinasus and A. ygraecum, Paralichthys brasiliensis and Prionotus
The early and classic comparative study of Wintrobe (1933) suggested that MCHC was the only relatively uniform hematogical variable in a variety of vertebrates studied, the single exception being the class of fishes which exhibited higher inter and even intraspecific MCHC variations. Nevertheless, our study showed very small intraspecific oscillations, indicating that, at least for matured specimens of similar age and length, MCHC is fairly constant in ahnost all species examined (see Table 1). On the other hand, we found a higher mean MCHC value for cartilaginous (27.2 g%) than for bony (21.2 g%) fishes. This intergroup distinction was not revealed in that work, probably due to sample limitations concerning elasmobranch species. Furthermore, the values of this variable revealed in each fish group were fairly constant (Table 2). The MCHC values of rays and sharks examined, and to a lesser extent the teleost values, were slightly lower and closer to those relating to humans and mammals. Unlike the enucleated mammalian red cells, the nuclei of fish erythrocytes are very large and can comprise one fourth or even more of the cell volume (Wintrobe, l933; Coburn and Fisher, 1973). This could explain, in purely physical terms, the lower fish MCHC values compared to other vertebrates. The relative constancy of MCHC verified within and among species belonging to each major group could perhaps better be explained as a reflection of developmental constraints (i.e. the intrinsic high correlation between Hb and Ht; see Fig. 7) than as adaptive adjustments to en~ronment. It is tempting to attribute at least part of the existence of the high hemoglobin heterogeneity that typify fish in general (Reischl and Tondo, 1974; Fyhn et al., 1979; Wilhelm Filho and Reischl, 1981; Perez and Rylander, 1985) to counteract the solubility limitations that high hemogiobin con~ntrations impose to blood oxygen capacitance (Riggs, 1976). An interesting inverse correlation between MCHC and latitude was described by Wells et al. (1980). In that hypothesis, four ranges of MCHC values were confronted with latitude. Mean cell hemoglobin concentration values for the species here examined (390-143 g/l) fall within the tropical/temperate/cold ~tabIished in that work temperate ranges (366-150 g/l). These discrepancies could probably be explained by the following features: the small number of species (11) used to characterize a geographical distribution of more than 60” of latitude, compared to nearly 40 species almost restricted to polar and subpolar areas (Larsson et al., 1976; Wells et al+, 1980), the large geographical distribution that characterizes many fish species and/or specific hematological determinants related to habitat and habit concerning fish species inhabiting different latituoes. The use of MCHC as an indicator of blood oxygen transport capacitance shouid be used with caution, keeping inferences concerning this hematological character more restricted to the i~~~llular microenvironment adjustments of oxygen transport (Weber and Wells, 1989). Blood hemoglobin concentration and the related intrinsic 0, affinity, besides the rheological and diffusional implications of the hematocrit
punctatus.
Comparing cartila~nous with boned fishes (Table 2), it is apparent that they behave as two distinct groups for all the parameters and indices examined (Figs l-6), with the exception of the blood hemoglobin concentration (Fig. 21, where the sharks exhibited an intermediary position between rays and teleosts. This intergroup contrast is more evident when shown through the means and the corresponding confidence intervals (Figs l-6B). Cartilaginous fish have larger (ca 3-4-fold) and fewer erythrocytes than the teleosts, thus confirming previous data in the literature (Wintrobe, 1933; Saunders, 1966; Pica er al., 1983). As a consequence, they have a cl-S-fold higher mean ceil hemoglobin content (MCH) than the teleosts, meaning that, in spite of the lower blood hemoglobin concentration, their large erythrocytes contain more hemoglobin than the teleost ones. The intrae~thr~ytic hemoglobin concentration values revealed for sharks, rays and teleosts were 4.4, 4.2 and 3.3 mM, or, as expressed in mean cell hemoglobin concentration (MCHC!) values, 28.4, 27.0 and 21.2 g%, respectively (Table 2). Regression analysis revealed high correlations between hemoglobin concentration (Hb) vs hematocrit (Ht) parameters (Fig. 7; correlation coefficient = 0.84), a feature frequently reported for fish (e.g. Wintrobe, 1933; Schaefer, 1961; Swift, 1982). On the other hand, the high correlation found for erythrocyte counts vs hematocrit (r = 0.81), and also between MCH vs mean cell volume (MCV) (r = 0.93) is rarely found in the literature con~ming fish (Figs 8 and 9, respectively). Less active teleosts generally occupy a clearly separate place in relation to more active ones (most plots below the regression line in Fig. 9, for instance). The variables Hb and red cell counts revealed different regression fitness for cartilaginous fish (r = 0.59) and for teleosts (r = 0.72) (Fig. 10). Nonlinear regression fits better for both red cell counts vs erythrocytic volume and for cell counts vs hemo~obin content (Figs 11 and 12, respectively. Intraspecific variations of MCHC were, in general, relatively low, some species exhibiting remarkably low oscillations, especially in the teleosts.
Iimbotas
Carcharhh mactd&~is Carcharhitu miltvrti Catvharhitw ob.wtww Ciuchnrhirau poro~us Family Lmnidae Ioxyrinchw Family Odontaspididae odtmtaspis tFamily Sphymidae SpAynra Inulnd WY@= tyEIreM Family Squalidae $qtI&Kc!l&i%T& Family Squatinidae sq#ari?la atgentkka Family Triakidae Mtstefw fosciaus Mwtelus schmitti
Cm&rhimss
R&a qc~ophora Rq&zpkltmre Sympterygio acuta Sympterigia banapartei Family Rhinobatidae ~~~~ kke#t RkWtoJ percelteru zqpleryx bre&astrik Family Rhinopteridae Rhitwtera botwsw Ofdez EkpLliformes Family Alopiidae Alopiar vul~inuc Family Carcbarhinidae
Rgja cmteimnd
Rqjiog&zf
Class choadrichthyes Subetass ~~ob~~h~~ onfcl RajifDrmcs Family Dasyatidae Dasyalli gtittatu Dtuyaiis say Family Myliobatidac MyUotvtis go&i Family Narcinidae
Taxonomicai position and scientific names
0.62 i: 0. I4 (6) 0.70 k 0.21(8) 0.41 + 0.18 (4) 0.91 (I) 0.64 + 0.18 (2) 0.73 + 0.18 (5) 0.50 & 0.15 (7) 0.95 f 0.15 (5) 0.76 + 0.18 (4) I.04 k 0.31 (2)
1.84(l) I.31 +0.37(2) f.i2+0.5O(S) 0.58+0.1i (2) 0.75 (I) 0.30 + 0.08 (2) 2.17(l)
17.1 rt &l(8) 17.7 & 6.1 {W) 15.4 + 5.3(M) 17.9 + 2.7(2) 16.3 & 4.445) 21.6+ 3,6(8) 13.8 + 3.0 (9) 18.9 + 7.0 (7) 13.341) 19.0 k 5.7 (5) 25.1 + 5.743)
33.0(l) 22.3 + 8.6 (3) 30.1 t 12.0(5) 16.1 +?.0(2) 15.0(l) 29.9 rt 6.3 (7) 28.7 + 10.6 (2)
0.173 +0.081 (IO) 0.298 (I) O.Ul +0.127(4)
0.66 + 0,17(10) 0.76(l) 0.65 rf;5.32 (2)
23. I + 8.7 (14) 23.5 t 2.1 (2) 20.4 & 8.2 (4)
0.250(l)
1.13(l)
31,0(I)
738.3 (I) 618.8 f Xti.0 (4)
1496.6 f 737.6 (10)
1192.3(l)
475Sf 126.Q(I8) 516.8 It 215.2 (18)
0.592fO.l75(18) 0.559rt0.2H1(18)
22.2(l) 22. I f 4.0 (2)
24.6& 13.O(lO)
23.4(i)
34.3 f 6.9 (14) 28AI f 6.6 (20)
28.3 (1)
780.2 * 294-7 (3) 0.297 f 0.072 (3)
0.91 (I) I.30 i 0.22 (14) 1xl2 f: 0.27 (20)
35.9(I)
563.5 k I 19.4 (2) 0.492 + 0. IS6 (2)
21.9 + 3.5 (3)
27.8 f 6.3 (2) 26.5 & 5.2 (5) 31-l f &f(2) 32.4 (1) 24.8 f 3.3 (2)
717,4(l)
36.0 (I)
28.6 k: 0.1(2)
1I.3 (3)
475.7 k
26.0 + S-9 (5) 32.9(k)_ 24.7 f 8.6 (4)
23.1 f 5.5 (8) 2&s f 2.8 (8) 36.6(l) 23.4 f 5.2 (2) 22.6 f 5.1(S) 24. I & 7.0 (7)
22.7 k 6-O(6)
29.8 & 9.3 (3)
38.9 f SS (3) 23.5 f 2.1 (3)
MCHC (g%)
618.1 f fZS.4(3) 461.8 (I) 592.5 f 75.4 (4)
649.8f211.8(3) 809.7 + 97.3 (4) 617.1 k216.015)
597.3 rt 152.0(S) 784.6 & 230.2 (8)
1212.8 f 336.2 (4)
581.8 + 298.4 (4)
775.5 4 517.7 (2) 579.3 f 64.6 (3)
MCV (a’)
563.5 f 124.9 (3) 596.8 g: 115.8 (3) 444.0 Ifr Il8.3 (2) 388.611) 523.7 I: 135.0 (4)
0.389 * 0.270 (3) 0.522 & 0.043 (3) 0.353 2 0.064 (2) 0.386(l) 0.508 f 0,038 (4)
0.460(l)
0.527 +- 0.108 (3)
0.303 4 0.066 (3) 0.288 (1) 0.343 f 0.058 (4)
0.242 $020 (3) 0.259 f 0.038 (4) 0.220 + 0.063 (5)
0.257 + O.OS4(5) 0.212 ri 0.105 (8)
0.165 f 0.064 (4)
0.368 & O.lSl (4)
0.284 + 0.082 (2) 0.278 f 0.012 (3)
RBC (lO’/mm’)
27.3 + 7.4 (18) 25.4 f 8.8 (23)
0.68(I)
0.76 k 0.25 (3)
18.0 + 6.3 (7)
-
1.27 F 0.25 (3) 0.56 +O.I0(3)
Hb (mM)
standard deviations {when applicable) concerning 80 specks studied
21.7&6.6(3) 14.3 ): 3.6 (4)
Ht 1%)
Table I. Hematological values and corresponding
iLO(2)
conritaued
164.2 (I) 149.4 f 47.6 (2)
391.5 f 192.4(10)
278.9 (I)
ISo.9&4444.s(14) 127.6 f X3(18)
2oP.3(1)
232.6 (1)
113.7;
117.4 f 30.4 (2) 143.1 f 74.1 (3) 127.3~74:;; (2)
257.8 (I)
135.4 f 4.7 (2)
218.9 It: 78.4 (3) 152.1(l) 142.2 f 30.6 (4)
185.7 f 41.8 (2) I%.3 f 18.6 (4) 146.4 f 36.0 (3)
132.3 & 36.0 (S) 167.3 f 14,7 (4)
289.3 f 136.5 (4)
252.9 + 46.1 (3)
132.0+ 17.1 (3)
298.5 2 160.0(2)
MCH (P&
z W
Taxonomid position antlsdentiknamcs
17.3119
16.3(i)
414S(l) -
n.c.
@.538(i)
n.e. 2.356 & 0.746 (4) 2.816f0.655(17)
0.44(i) 0.69 (I)
1.07(t) 1.14 & 0.07 (4) 1.61 +0.34/17) 1.63 ($9
i.oai~o.30~2j 2.10(i) 274(i) 201(i) 1.63 & 0.43 (4) 1.06~1) i.~iO.23(6) I.%(1) 1.92(i) 0.97(l) i.90f0.46{7) I .a f 0.35(7) I&?(1) O.%f 0.10 (2)
12.2(i)
20.7 + 2.3 (2)
42.6(l) 42.9 + 5.9 (7) 52.4 f I I.3 (24) 6Li (if 35.7 + 5.5 (3) 33.9 + 8.3 (2) 33.0 f Il.8 (2) 0.5 F 10.6(Z) 63.5(i) 61.0(i) 49.2* i3.3{4) 30.0(l) 56.9f 11.7(6) 53.8(i) 426 (I) 31.1(l) 50.1 f 12447) 44.4 f le.0 (7) 33.0(1) 34.0 f 2.8 (2)
156.8 (I) 195&l) 105.8 (1) ni *of 86.2 (3) 225.3 (I) 111.2(l) 165.8 (I) 108.7 f 8.4 (3) 171.4 f 3&O(4)
4.336(i) n.t. i.G-&) 2.835 (1) 2.652 rt 1.W (3)
1.&0((l)
z::; i:llsfi(l) 3.966 f 0.170(3) 2y;o;i;; (4)
177.4(l)
119.7 (1)
237.9 f 49.2 (2) 179.7(l) IM.3 (I)
1A88+ 0.023(2) i.aco(l) 1.574(i)
217.4(l)
ZllM(i)
185.1 f 38,9(17)
189.5 f 45.1(4)
18.2 f 0.3 (29
Et g; 25.2i4.i (7) ~~i~~~~(~
19.9(l) 28.1 (I) 2L2{1) la.?* 3.314) 22.7(l) 20.5& 1.1(6) 23.8 (I)
17.4ji: 1,2(3) 21.1 &4.9{2] Zl.!if i.8(2)
16.9(r)
2i.2& 4.6(17)
17.3*2.8(49
23.4(l)
14.6 + 2.8 (2)
-
n.e.
14.3 4 2,3 (35)
193.8 f 57.6 (35)
0.68 k 0.04 (2)
1.892 f 0.521 (35)
32.4 + 6.0 (2)
-
26.7 + 2. I (2)
36.7 (I)
30&) 24.0 (I) 63.5 It 4.3 13) 53.4(i) 32.3 (I) 33.7 (1) 29.1 f 8.3 (39 38.9 f lLV(4) 25.1 (I)
31.3(l) -
44.1 f 1.3(2) 420(l) 35.5 (I)
36.7 (1)
38.0 f 7.3 (17)
37.8 f 19.7 (4)
-
71.9(i)
-
-
28.9 f 9.3 (359
74.6(l)
MCI-4 (Pg9
38.1(l)
0.83 + 0.16 (35)
19.5.8(l)
MCWC (g%)
35.5 f 7.8 (41)
47.7 + WA
Q.#i(l)
MCV (a’)
2.26 + 0.25 (21
0.77 (1)
13.0(l)
RBC (IO’/mm”)
I I.8(3)
Hb (mM)
HI (%)
C!ompatiltive hematoiogy in mariM: fish
DAMLOWILHELMFILHOet al.
316
Table 2. Mean values and standard deviations related to rays, sharks and teleosts studied Sharks
bYS
Ht (%) Hb (mM) RBC (I06/mm’) MCV @‘) MCHC (g%) MCH (pg)
17.9 f 0.75 f 0.288 f 673.5 + 27.0 f 188.4*
3.3 (14) 0.23 (14) 0.090(13) 194.2(13) 5.7 (14) 59.3(13)
24.8 f 1.09 f 0.403 * 686.8 * 28.4 f 185.0 k
5.5 (14) 0.45 (14) 0.123 (14) 305.5 (14) 4.9 (14) 80.4 (14)
TCkosts 41.0 * 1.34 f 2.289 f 200.0 f 21.2 f 39.1 f
14.3 (52) 0.58 (52) 0.904 (41) 89.8 (41) 4.6 (52) 13.0 (41)
Numbers in parenthesis denote species’ diversity. Ht = Hematocrit; Hb = hemoglobin concentration; RBC = red blood cell count; MCV = mean cell volume; MCHC = mean cell hemoglobin concentration; MCH = mean cell hemoglobin content.
and red cell size and number, together with other morphological and physiological characters related to oxygen transport system, such as the cardiovascular and ventilatory performances, are apparently more reliable in ascertaining respiratory needs of an entire organism. As expected, more active pelagic forms of teleosts and sharks showed higher hematological parameters than sedentary or less active species, a condition that enables the former species to cope with increased oxygen demands (Schaefer, 1961; Haws and Goodnight, 1962; Eisler, 1965; Cameron, 1970; Larsson et al., 1976; Emery, 1986). In this regard, regression analysis revealed that sluggish species (e.g. Prionotus punctatus, Percophis brasiliensis, Balistes capriscus and Stephanolepis hispidus, and the stargazers A. sexspinosus and A. ygraecum) are localized between
active teleosts and the elasmobranch group (open circles in the left lower comer in Figs 10 and 11). Similarly, the intermediary position showed by sharks concerning blood hemoglobin remained as the only variable that overlapped the three fish groups studied (Fig. 2A, B). This feature could be attributable to the common pelagic habit of sharks, contrasting with the sedentary and benthonic rays. An enhanced blood oxygen capacity of active fish fits well with the ca 3-fold higher oxygen extraction of endotherm sharks compared to the ectothenn forms (Emery, 1986). In this regard, elasmobranchs generally seem to extract more oxygen than teleosts (Layton, 1987). Endotherm fish, like the lamnid sharks and some scombrids, bear a countercurrent system with a rete mirabile located in their red muscles (Carey, 1982; Emery, 1986), that allows them to keep these tissues at higher temperatures than the surrounding water, and to exhibit extraordinary swimming performances. Some scombrids here studied revealed hematocrits as high as 80%. Although stress, in some cases, can be responsible for overestimations in hemoglobin content and also concerning hematocrit and red cell numbers, relatively small differences were eventually found (Swift, 1982). Doubling the hematocrit, as in the case of teleosts confronted with rays and sharks, the viscosity (n) and flow resistance (R) of blood is also doubled, according to Pouseille’s equation, thereby increasing the cardiac work in teleosts. Vascular resistance and cardiac work are further increased by the high capillary density of teleosts (Hartmann and Lessler, 1964; Schmidt-Nielsen, 1983). Nevertheless, the high prebranchial compliance of the ventral aorta of the teleosts, when compared to the elasmobranchs (Taylor, 1964; Langille et al., 1983), may constitute physiological and anatomical adjustments to com-
pensate the high blood viscosity of the former. On the other hand, elasmobranchs have broader and less dense capillaries, and possess larger and fewer red cells, besides larger volemia (ca 3 times greater than the teleosts) (Thorson, 1962; Prosser, 1973; SchmidtNielsen, 1983), that may reduce blood viscosity and peripherical resistance (Graham and Fletcher, 1983). Furthermore, elasmobranchs are also characterized by higher cardiac stroke volumes (Prosser, 1973) and lower cardiac work (Side11 and Driedzic, 1985), related to the high pericardic rigidity (Johansen and Gesser, 1986), and also by a low pre-branchial compliance of the ventral aorta (Langille et al., 1983). As a consequence, they show an increased blood flow when compared to the teleosts. The higher blood convection requirement of cartilaginous fish in this manner may counteract the modest hematological parameters of rays and most sharks (Johansen, 1971). It is interesting to verify that a similar condition is also found in the abissal (Graham et al., 1985), and also in the Antarctic teleosts (Wells et al., 1980). The blood oxygen transport capacity (CHbG& is directly proportional to the arterial-venous difference in oxygen concentration (Cao, - Cv,), and, considering that the latter is inversely related to. the cardiac output (Vb), the oxygen consumption (M%) (Weber, 1982; Dejours, 1988) is: nio, = (Cue, - Cv,)
rib i.e. h&,/vb
= Cao, - Cuop.
For the same oxygen consumption rate, the higher the oxygen capacity, the lower the blood flow (and cardiac output) rate needed. This assertion seems to fit well both the phylogenetic positions (e.g. cartilaginous vs bony fishes), and the ecophysiological features concerning each species (e.g. endotherms vs ectotherms, or active vs sluggish forms). For instance, the angelshark Squutina urgentina showed hematological relationships closer to rays than to the squaliforms, but compatible with the habit and habitat that this fish shares with the rajiforms. In short, it seems that elasmobranch and teleost fish have, as two broad groups, different gas transport systems. Instead of high cardiac work and blood pressures accompanied by numerous small red cells held in more viscous blood of higher hematocrits ascribed for teleost fishes, the elasmobranchs, and especially the rajiforms, show, in general, relatively lower energy costs related to modest cardiac work, but exhibit higher gill oxygen extraction than teleosts, higher cardiac outputs, and higher blood volemia and flow rates. These counteract the reduced oxygen transfer conductance that results from low hematocrits, despite the large red cell volumes they normally display (Yamaguchi et al., 1988). These are examples
317
Comparative hematology in marine fish %
40
30
20
10
R
mM
’
f
/
1
HB
r
T
s
mM-’ . HB
-I
lo4mm3
lOe/mm?
23
3.9
13 1.0
0.2
03
R
S
T
R
S
1
Figs l-3. Multiple box plots (A) and means and the correspobding standard errors (B) (confidence intervals of 95%) of ali hematological parameters (XT, HB and RBC counts) analysed. R = Rays; S = sharks; T = teleosts. Values according to the units employed in Tables I and 2.
DANEOWILKBLM FUIO er al.
318
P% 200
300 t20
200
100
40
R
S
t
R
S
T
Figs 4-6. Multiple box ploti (A) and meana aud the cocrcqondi~ stendard ~rrora (B) (con&iencc intervals of 95%) of the three indice (MCV, MCH and MCHC). R = Rays; S = sharks; T = tclcosts.
Comparative hematology in marine fish
, I
VII
, I,,
,,.)
.,
c,
,
3
2
1
‘
I
HB
MCH
300
200
100
300
800
900
‘200 MCV
1
2
3
4
1
2
3
4
RBC
RBC
Figs 7-12. Regression analysis related to some hematological parameters and indices. Values according to the units employed in Table 1 and 2. Rays (A); sharks (A); sluggish teleosts (0) and moderate/active teleosts (a). Outer and inner dotted lines represent prediction and confidence intervals, respectively.
of distinct mo~holo~~a1 and physiolo~cal devices of composing appropriate respiratory needs. It could be argued that the low energy costs of e~~mobr~chs in this respect imply an energy profit
that may be transferred to economic activities, as welf
as ~pr~u~on,
benefbing the existence of wide-
spread viviparity in these fishes, with all its associated advantages. Therefore, a side consequence of physi-
DANILOWmsmul FILHO et al.
320
ology could in a way be thought of as meaning an increase in general exaptation (Gould and Vrba, 1982). Irrespective of whether this is true or not, respiratory physiology tells us that the elasmobranchs are at least as successful as teleosts in what concerns physiological efficiency. Contrary to earlier speculations (see Wintrobe, 1933, and other modem literature), more akin to cultural preferences than to corroborated facts, el~mobranch fish are not necessarily “unsuccessful evolutional experiments”, or members of a group “low in the scale of evolution” (see similar ideas relating to ecto vs endotherms in Pough, 1980). Acknowledgements-We are grateful to the fishermen of Praia Plntano do Sul, Arma@o, Campeche, and Barra da Lagoa, who kindly provided the fish for blood sampling. Also to Dr Eduardo Htmreres and Carlos Eduardo Pinheiro (Universidade Federal de Santa Catarina), for the use of laboratory equipment and computer, respectively, and Dr Evaldo Reischl (Universidade Federal do Rio Grande do Sul) for comments on the manuscript. This study was supported by grants of the CNPq (Con&ho National de D~envolvimento Cientifico e T~ol~~~/Brazil, proc. 402~3/88-6 and ~33~/89-7) to DWF”. REFERENCES
Aldrin J. F., Messager J. L. and Saleun S. (1984) Analyses sanguines de turbots d’elevages immature8 (Scophtalmus maximus L.). Aquaculture 40, 17-25. Andersen N. A., Laursen J. S. and Lykkeboe G. (1985) Seasonal variations in hematocrit, red cell hemoglobin and nucleoside triphosphate concentrations, in the European eel Anguilla anguilla. Comp. Biochem. Physioi. 8lA,
Physioi. 81A, 879-883.
Eisler R. (1965)Erythrocyte counts and hemoglobin content in nine species of marine teleosts. Chesapeak. Sci. 6, 119-120. Emery S. H. (1986) Hematological comparisons of endothermic vs ectothermic elasmobranch fishes. Copeia 3, 700-705. Ferrando A., Bobadilla I. G., Bobadilla M. S., Palou A. and Alemany M. (1981) Comparative estimation of hematocrit and trapped plasma in the packed cell volume in man, rabbit and chicken blood. Comp. Eiochem. Physiol. 7OA, 611-613.
Fyhn U. E. H., Fyhn H. J., Davis B. J., Powers D. A., Fink W. L. and Garlick R. L. (1979) Hemoglobin heterogeneity in Am~onian fishes. Comp. Biochem. Physiol. 62A, 39-66.
Gould S. J. and Vrba E. S. (1982) Exaptation-a missing term in the science of form. Pu~eo&iofogy 8, 415. Graham M. S. and Fletcher G. L. (t983) Blood and plasma viscosity of winter flounder: influence of temperature, red cell concentration, and shear rate. Can. J. Zool. 61, 2344-2350. Graham M. S., Haedrich R. L. and Fletcher G. L. (1985) Hematology of three deep-sea fishes: a reflection of low metabolic rates. Comp. Biochem. Physiol. 8OA, 79-84. HIrdig J. and Hiiglund L. B. (1983) Chr accuracy in estimating fish blood variables. Comp. Biachem. Physiol. 75A, 35-40.
87-92.
Atkinson E. and Judd F. W. (1978) Comparative hematology of Lepomis mi~ro~oph~ and Ciehlasoma eyanoguttatum. Copeia 2, 230-237.
Benfey T. J. and Sutterlin A. M. (1984) The haematology of triploid landlocked Atlantic salmon, Salmo salar L. J. Fish Biol. 24, 333-338. Bridges D. W., Cech J. J. Jr and Pedro D. N. (1976) Seasonal hematological change in winter flounder, Pseudooteronectes
Dacie J. V. and Lewis S. M. (1968) Practical Haematology, 4th edition. J and A Churchill. Dejours P. (1988) Respiration in Water and Air: Adaptations-Regulation-Evolution. Elsevier, Amsterdam. Denton J. E. and Yousef M. K. (1975) Seasonal changes in hematology of rainbow trout, Salmo gairdneri. Comp. Eiochem. Physiol. SlA, 151-153. Duthie G. G. and Tort L. (1985) Effects of dorsal aorta cammlation on the respiration and hematology of Mediterranean living S~y~~orhin~ cat&u/a L. Comp. Biochem.
americanus.
Trans. Am. Fish. Sot.
18%
596~600. Bridges C. R., Taylor A. C., Morris S. J. and Grieshaber M. K. (1984) Econhvsioloaical adaptations in Blennius pholis (i.) blood io-inter~dal rockpool environments. J. exp. Mar. Biol. Ecol. 77, 151-167. Burton C. B. and Murray S. A. (1979) Effects of density on goldfish blood-I. Hematology. Comp. Biochem. Physiol. 62A, 555-558.
Cameron 3. N. (1970) The influence of en~ronmental variables on the hematolo~ of pi&h (Lagodon r~o~oides) and striped mullet (~ugiI cephalus). Comp. Biochem. Physiol. 32, 175-l 92.
Carey F. G. (1982) Warm fish. In A Companion to Animal Physiology (Edited by Taylor C. R., Johansen K. and Bolis L.), op. 217-231. Cambridge University Press, ,- __ Cambridge. Cech J. J. Jr and Wohlschlan D. E. (1982) Seasonal patterns of respiration, gill ventila%on and he&atological‘characteristics in the striped mullet, Mugil cephalus L. Bulf. Mar. Sci. 32, 130-138.
Clark S., Wbitmore D. H. Jr and McMahon R. F. (1979) Considerations of blood narameters of largemouth bass, Micropterus sabrwides. J.* Fish Biai. 14, t
