New Haemogregarines from New Zealand Marine Fishes*
[Read before the Wellington Branch, October 25, 1950; received by the Editor, February 19, 1951]
Introduction and Acknowledgments
Prior to this survey no haemogregarines were known from New Zealand fish. Blood examinations of 500 fish of 52 marine and fresh-water species, carried out over the last four years, proved 9 of the marine species to be hosts for Haemogregarina (Protozoa: Coccidia: Adeleidea: Haemogregarinidae). A total of 40 fish belonging to these 9 species were parasitized. Fifteen of these, of 5 of the species, harboured varieties of Haemogregarina bigemina Laveran and Mesnil and of H. platessae Lebailly. These are being dealt with in a separate account, and will not be considered further here. A detailed list of the species of fish investigated with negative results is included in an account of the trypanosomes discovered during the same survey (Laird, 1951).
All the material considered in the present account, with the exception of that from Hoplichthys haswelli McCulloch, was collected during a trawling trip of the S.T. Maimai. I again take the opportunity of acknowledging the kind assistance of the master of this vessel, Captain Cardno. The specimen of H. haswelli, a species only rarely recorded from the New Zealand coast, was forwarded to the Department of Zoology at Victoria College by Mr. F. Abernethy of the S.T. Phyllis.
Thin smears of heart blood taken in the field were air dried on collection, being subsequently fixed in absolute methyl alcohol and stained with Giemsa. The whole area of each preparation was searched microscopically, a × 5 ocular and × 97 oil immersion objective being employed for this task.
Haemogregarina coelorhynchi n.sp. (Plate 128, figs. 1–34)
This species is described from Coelorhynchus australis (Richardson) [javelin fish] and Physiculus bachus (Forster) [red cod]. Eighteen of 28 examples of the former fish and 4 of 8 of the latter were parasitized, there being no significant morphological differences between the haemogregarines from these two hosts. All the fish were trawled at 50 fathoms off Cape Campbell, Cook Strait, in March, 1948. The heaviest infections (up to 50 haemogregarines per 10,000 red cells) were recorded from the smaller examples of both hosts, those measuring up to 24 cm. in length. Some very light infections (less than one haemogregarine per 10,000 red cells) were recorded from medium-sized to large fish measuring up to 46 cm. in length, while no parasites
[Footnote] * This contribution forms part of a protozoological survey undertaken during the tenure of a New Zealand University Research Fund Fellowship.
could be found in smears from 4 examples of C. australis of greater length than this.
One of the infected examples of C. australis and all those of P. bachus were also parasitized by a species of Trypanosoma, which is described in the account of this genus mentioned earlier.
The smallest stages of this parasite so far encountered are free forms (Pl. 128, figs. 3 and 4), which measure 5.0μ to 5.9μ by 2.5μ to 3.1μ in their greatest dimensions. As these occur abundantly in all but chronic infections, it is most unlikely that they are sporozoites, which only make a transitory appearance in the initial stages of an infection. It is probable that they correspond to the smaller of the two types of merozoite described from Haemogregarina stepanowi Danilewsky, a parasite of turtles, by Reichenow (1910), that destined to give rise to gametocytes. The small free forms under consideration have a large nucleus which is rich in chromatin and stains a very deep red with Giemsa. There is a relatively small amount of cytoplasm which stains light blue, and appears as a little tongue at one or both of the extremities. Some of these forms have an amoeboid appearance (Pl. 128, fig. 4), as have the earliest intraerythroblastic stages (Pl. 123. fig. 5). The latter stages have a similar staining reaction to that of the small free forms, but unlike these they have from two to four large granules which stain a dense blackish red situated towards the extremities of the body. These granules persist until the next stage of development (Pl. 128, figs. 5–8), but afterwards disappear. They are probably products of the metabolic processes concerned with the act of invasion of the host cell. The most recently invaded corpuscles have a rounded outline, a round nucleus with conspicuous chromatin blocks, and polychromatophilic cytoplasm (Pl. 128, figs. 5–8). These features distinguish them as late erythroblasts. A normal cell of this type is illustrated (Pl. 128, fig. 1) for comparison with a normal erythrocyte (Pl. 128, fig. 2). The earlier stages of the other piscine haemogregarines studied during the present survey, also the small and intermediate stages of a haemogregarine of the American leopard frog (Bailey, 1948), are likewise found in erythroblasts, as are the ring stages of several species of Plasmodium.
Growth in length leads to the attainment of the vermicular form seen in Pl. 128, figs. 7 and 8. This form has a size range of 7.2μ to 9.5μ by 1.8μ to 2.0μ. It closely resembles the equivalent stage of Haemogregarina aeglifini,* described from the European haddock, (Gadus) = Melanogrammus aeglifinus, by Henry (1913), which measures 8.5μ to 10.5μ by 2.0μ to 3.0μ. The nucleus, which is usually central in position, stains lighter red than does that of the preceding stages. This structure occupies the full width and something under half the total length of the body. The rather alveolar cytoplasm stains light blue. Dense granules like those already described are often present, sometimes at both ends of the body (Pl. 128, fig. 7), but more often at the more pointed extremity only. Henry (1913) described similar granules from the equivalent stage of H. aeglifini (his Pl. xvii, figs. 2–4).
[Footnote] * Amended from aeglefini by Fantham et al. (1942).
Further development leads on to the stage illustrated in Pl. 128, figs. 9–11, which has average measurements of 10.5μ by 2.3μ (range, 8.5μ to 12.3μ by 1.8μ to 2.7μ). The nucleus ranges from 4.8μ to 5.8μ in length and from 1.8μ to 2.2μ in breadth, and is situated towards one extremity. It may appear to bulge out beyond the sides of the body (Pl. 128, figs. 9 and 10), as does that of the parallel stage of Haemogregarina rovignensis, described from the Italian streaked gurnard (Trigla lineata) by Minchin and Woodcock (1910). The half of the body containing the nucleus is narrower than the other, and its extremity is less broadly rounded. The cytoplasm stains light blue to purple and sometimes contains small masses of extranuclear chromatin. A red staining polar cap is developed at the end farthest removed from the nucleus (Pl. 128, fig. 11).
In all stages so far described the host cell nucleus is scarcely if at all displaced, and the cell itself shows no signs of hypertrophy.
It is by no means clear what place in the developmental series is filled by the stout haemogregarines illustrated in Pl. 128, figs. 12 and 13. These have average dimensions of 10.5μ by 3.6μ (range, 8.5μ to 11.0μ by 2.9μ to 4.2μ), and their nuclei measure 4.1μ by 3.1μ (range, 2.9μ to 5.8μ by 2.5μ to 4.0μ). The nucleus is usually situated just within the half of the body furthest removed from the extremity having a polar cap. These stout haemogregarines superficially resemble the stages of H. rovignensis interpreted by Minchin and Woodcock as female gametocytes, which measure 10.6μ by 3.4μ (extreme size about 11μ by 3.5μ), their nuclei averaging 4.8μ by 2.9μ. The available evidence, however, suggests a different interpretation in the case of the haemogregarines under consideration. These stout forms do not retain a regular and well-defined outline, but instead show considerable plasticity, one of the extremities often being markedly amoeboid in appearance (Pl. 128, fig. 13). The earliest stages definitely identifiable as young schizonts being those of irregular outline illustrated in Pl. 128, figs. 14 and 15, there would seem to be a high degree of probability that the stout forms described above are directly antecedent to these undoubted schizonts. By this hypothesis the free forms illustrated in Pl. 128, figs. 33 and 34, are to be identified as analogous with Reichenow's macromerozoites of H. stepanovi, and, like these, are destined to initiate a further schizogonous cycle.
Young schizonts round up and assume a pyriform or ovoid (Pl. 128, figs. 16 and 17) shape. The alveolar cytoplasm, which stains a very light whitish-blue, may contain a number of extranuclear chromatic granules. The chromatin material of the nucleus adopts a loose, trabecular arrangement. Larger schizonts (Pl. 128, fig. 18) have similar staining reactions to the young ones, but the cytoplasm may be somewhat maculated, and a prominent red-staining cap is usually present at one of the extremities. These large forms have a size range of from 8.1μ to 10.8μ by 4.6μ to 5.4μ. Erythrocytes parasitized by haemogregarines at this stage of development become considerably hypertrophied. They may attain 15.0μ by 11.0μ in their greatest dimensions, as compared with the 10.7μ by 9.1μ of normal erythrocytes. Merozoite formation has not been observed. It is, of course, possible that this takes place in the capillaries of internal organs.
Large schizonts are present in only two smears, significantly the most heavily parasitized ones. Gametocytes are present throughout my material, and are the dominant forms in preparations from sparse and presumably chronic infections. These (Pl. 128, figs. 19–32) have a considerable size range. The large stout forms illustrated in Pl. 128, figs. 19 and 20, are from the infections referred to above in which active intraerythrocytic schizogony was proceeding. Very few such forms have been encountered, and although their size as compared with that of the majority of the gametocytes, and the relatively greater bulk of the nucleus, is suggestive of their being macrogametocytes, this explanation of their nature is considered to be an unlikely one. Were they actually female gametocytes, it is to be expected that they would be of constant occurrence throughout the infections, including the sparse ones. The other gametocytes, those present in all the infected smears, are all characterized by the possession of a relatively small, compact nucleus. It thus appears likely that the sexual dimorphism of the haemogregarine under discussion does not become apparent during the sojourn of the gametocytes in the blood of the vertebrate host, and that the large stout gametocytes just described are in reality aberrant forms.
It is considered that further development of the vermicular type of haemogregarine illustrated in Pl. 128, fig. 11, leads directly to the formation of the sexual stages, gametocyte formation thus being initiated by the invasion of a red cell by a micromerozoite as in H. stepanowi Danilewsky (Reichenow, 1910).
Gametocytes having the small, compact type of nucleus (Pl. 128, figs. 21–32) average 12.8μ in length and 3.6μ in breadth (range, 10.0μ to 16.7μ by 2.3μ to 4.7μ). In their morphology they resemble the slender (male?) gametocytes of H. rovignensis, the average length (12μ) of which corresponds with their own, although the breadth (2.1μ) is less. Erythrocytes containing mature gametocytes are markedly distorted, particularly in the long diameter, as are those of Trigla lineata parasitized by the equivalent stage of H. rovignensis. The measurements of such parasitized erythrocytes of Coelorhynchus australis average 13.8μ by 8.7μ (range, 11.9μ to 17.5μ by 6.9μ to 10.4μ), as compared with the 10.7μ by 9.1μ (9.8μ to 12.6μ by 8.4μ to 9.8μ) of normal erythrocytes. As is the case in H. rovignensis infections, the nucleus of a parasitized corpuscle may be indented (Pl. 128, figs. 21 and 25) or flattened (Pl. 128, figs. 19 and 24) by the pressure of the haemogregarine, although it does not seem to be other than mechanically affected.
The gametocyte nucleus stains deep red with Giemsa, and is a compact structure having dense aggregations of chromatin. Measuring some 3.0μ by 2.4μ (range, 2.1μ to 3.5μ by 1.9μ to 3.1μ), it is slightly larger than the nucleus of the equivalent stage of H. rovignensis, the dimensions of which average 2.6μ by 1.7μ. The nucleus is usually situated a third of the body length distant from one of the extremities. The opposite extremity is sometimes rounded (Pl. 128, figs. 22, 25, etc.), but is more usually bluntly pointed or wedge-shaped (Pl. 128, figs. 23, 27, etc.). This end is occupied by a varying amount of finely granular material which stains light red to deep rose red with Giemsa, and
extends back to fill at least half of the body of older gametocytes (Pl. 128, figs. 26, 30, etc.). There is usually a smaller polar cap at. the end of the body nearest the nucleus, in addition to the large one just described. In some cases (Pl. 128, figs. 22 and 26) the only blue-staining cytoplasm remaining apparent occupies the area in the immediate vicinity of the nucleus. H. rovignensis was described as having only one polar cap, that at the broader end of the body, although one example figured by Minchin and Woodcock (their Pl. 8, fig. 19) also has a smaller cap like that of the gametocytes under discussion. Red-staining polar caps occur in other piscine haemogregarines, notably Haemogregarina anarhichadis,* described from the European catfish (Anarchichas lupus) by Henry (1912), and H. hoplichthys n.sp. and H. leptoscopi n.sp. described below. The nature of the polar substance, which stains like chromatin with Giemsa, but which, according to Minchin and Woodcock (1910), is hardly apparent after iron haematoxylin or Twort's stain, is as yet unknown. From the fact that the oldest gametocytes generally have the most extensive polar masses and at the same time the smallest and most compact nuclei, it seems possible that the red staining mass may consist, at least in part, of chromatin liberated from the nucleus.
Mature gametocytes are often slightly curved or bow-shaped, like the equivalent slender forms of H. rovignensis. Sometimes the curve is more strongly accentuated than usual (Pl. 128, fig. 21), and occasionally one end is sharply bent over (Pl. 128, fig. 24). Two instances have been noticed in which the gametocyte has folded over on itself and grown along both top and bottom surfaces of the host cell (Pl. 128, fig. 23). This same phenomenon is seen in H. rovignensis (Minchin and Woodcock, 1910, Pl. 8, figs. 18 and 21). A large lymphocyte of C. australis, containing a mature gametocyte which has caused fission of the host cell nucleus, is illustrated in Pl. 128, fig. 30.
Evidence of necrosis among the older haemogregarines is often apparent in chronic infections, in which most of the parasites present are mature gametocytes (Pl. 128, figs. 27–29). Degenerating gametocytes undergo radical nuclear changes as a result of the liberation of an intracellular enzyme at death, the nucleo-protein then gradually splitting up into simpler compounds with the ultimate liberation of nucleic acid (Muir, 1924). These changes (karyorrhexis) are manifested by the nuclear chromatin breaking up into small, intensely staining granules which become irregularly arranged in the cytoplasm (Pl. 128, figs. 27 and 28). Sometimes the chromatin becomes aggregated into a few discrete, deeply staining masses (Pl. 128, fig. 29), this appearance being known as pyknosis.
In one smear only, from one of the most heavily parasitized examples of C. australis, free gametocytes are abundant in the plasma, each curved in U-form about a mass of chromatic material which represents the greatly hypertrophied remains of the nucleus of the host erythrocyte (Pl. 128, figs. 31 and 32). Fantham (1925) recorded free vermicides “with fragments of red blood corpuscles attached” from his H. bitis, a reptilian haemogregarine. These free gametocytes, which
[Footnote] * Amended from anarrhichadis by Fantham et al. (1942).
one would normally expect to find in the gut of the invertebrate vector, owe their presence in the plasma to the disintegration of the host erythrocyte.
No instances of the double infection of host erythrocytes with H. coelorhynchi n.sp. have been observed. Minchin and Woodcock (1910) found only two such instances in their material of H. rovignensis.
Of the piscine haemogregarines already described, H. rovignensis comes closest to the species under discussion. The gametocyte of the parasite of Coelorhynchus australis and Physiculus bachus bears a close resemblance to the long, slender form of H. rovignensis, from which it differs in characteristically having a small polar cap at the end nearest the nucleus and in having an appreciably larger amount of its cytoplasm infiltrated with the red staining material described earlier. The large broad form of the former haemogregarine differs from that of the latter in its morphology, particularly in the looser structure and more central position of the nucleus. H. anarhichadis Henry resembles the New Zealand species in the morphology of the earlier vermicular stages, and in having a conspicuous polar cap, but in so far as is at present known does not appear to have long slender gametocytes comparable with those of the latter species. In view of the differences between it and previously described piscine haemogregarines, H. coelorhynchi is accordingly described as a new species having the features detailed in the above account.
The type slide of Haemogregarina coelorhynchi n.sp. has been deposited in the collection of the Dominion Museum, Wellington (catalogue number Z13). A paratype slide has been placed in the collection of the Department of Zoology, Victoria College, and four paratypes are in my own collection (catalogue numbers CHg 1–4).
Haemogregarina hoplichthys n.sp. (Plate 129, figs. 1–15)
This haemogregarine, the gametocytes of which closely resemble those of H. coelorhynchi n.sp., is notable for the fact that an intraerythrocytic schizogony cycle leading to the production of two or four merozoites—not gametocytes, as in H. bigemina Laveran and Mesnil and allied species—takes place in the circulating blood of the vertebrate host. It is described from the heart blood of a 29 cm. example of Hoplichthys haswelli McCulloch, trawled at 70 fathoms off Cape Palliser, Cook Strait, in October, 1949. The parasite rate averaged 30 per 10,000 red cells.
Intraerythrocytic schizogony into two merozoites (Pl. 129, fig. 7) takes place either longitudinally (Pl. 129, figs. 2 and 3) or transversely (Pl. 129, figs. 4–6). When four merozoites are produced (Pl. 129, fig. 10) the schizont characteristically adopts a cruciform shape (Pl. 129, figs. 8 and 9). With Giemsa, the cytoplasm, which may be markedly vacuolated (Pl. 129, fig. 5), stains light blue, with darker blue maculation, particularly towards the periphery (Pl. 129, figs. 6 and 9). The nucleus has ground material staining deep red, and chromatin aggregations which stain blackish red. The schizonts never attain a size comparable with that of the larger examples of this stage in H. coelorhynchi n.sp., neither do they cause such marked hyper-
Explanation of the Plates
All figures drawn at a magnification of × 2,400, with the aid of an Abbé camera lucida.
Haemogregarina coelorhynchi n.sp.
Fig. 1—Late erythroblast of Coelorhynchus australis.
Fig. 2—Erythrocyte of C. australis. Figs. 3 and 4—Free micromerozoites. Fig. 5—Intracrythroblastic micromerozoite. Figs. 6–11—Early gametocyte development. Figs. 12–18—Early schizogony. Figs. 19–26—Gametocytes. Figs. 27–29—Degenerating gametocytes. Fig. 30—Intraleucocytic gametocyte. Figs. 31 and 32—Free gametocytes, attached to the hyptertrophied remains of the nucleus of the host erythrocyte. Figs. 33 and 34—Macromerozoites,
Figs. 1–15—Haemogregarina hoplichthys n.sp.
Fig. 1—Erythrocyte of Hoplichthys haswelli. Figs. 2–10—Schizogony. Figs. 11 and 12—Developing gametocytes. Figs. 13 and 14—Mature gametocytes. Fig. 15—Gametocyte freed from host erythrocyte, following the death of the infected fish.
Figs. 16–33—Haemogregarina leptoscopi n.sp. Fig. 16—Erythrocyte of Leptoscopus macropygus. Figs. 17 and 18—Free macromerozoites. Figs. 19–23—Intraerythroblastic (Figs. 19, 21, 22) and intraerythrocytic merozoites. Figs. 24–28—Doubtful forms, interpreted as young schizonts. Figs. 29–33—Maturing schizonts.
trophy of the host cell as do those of the last-named species, although they do cause displacement of its nucleus.
Merozoites range from 4·2μ to 6·3μ in length and from 1·4μ to 2·1μ in breadth, the average dimensions being 5·4μ by 1·6μ. They resemble schizonts in their staining reactions. The nucleus occupies from a third to a little over a half of the total length of the body, and is situated towards the narrower extremity. It often bulges out beyond the visible lateral extent of the cytoplasm. Small extranuclear chromatic granules may be present (Pl. 129, fig. 10).
Intraerythrocytic gametocytes (Pl. 129, figs. 11–14) always occur singly. Their general morphology and staining reaction is closely similar to that of the equivalent stages of H. coelorhynchi n.sp. From the measurement of 20 examples their average dimensions were found to be 12·6μ by 3·8μ (range, 9·8μ to 17·0μ by 3·2μ to 4·7μ), these figures comparing favourably with those obtained from the measurement of a similar number of gametocytes of H. coelorhynchi n.sp. (average, 12·8μ by 3·6μ; range, 10·0μ to 16·7μ by 2·3μ to 4·7μ). The mature gametocyte illustrated in Pl. 129, fig. 15, was drawn from a preparation made several hours after the death of the host, and probably owes its freedom in the plasma to post mortem temperature and other changes in the blood.
The cytoplasm of mature gametocytes (Pl. 129, figs. 13–15) is almost completely infiltrated with red-staining material analogous to that occurring in H. rovignensis Minchin and Woodcock, H. anarhich-adis Henry and H. coelorhynchi n.sp. As in the equivalent stages of the last-named species, blue-staining cytoplasm is restricted to the immediate vicinity of the nucleus, which develops into a relatively small, compact structure situated towards the narrower extremity of the body.
Erythrocytes containing mature gametocytes are, like those of Coelorhynchus australis and Physiculus bachus parasitized by the equivalent stages of H. coelorhynchi n.sp., somewhat narrowed and longitudinally distorted. The narrowing effect is more marked, while the longitudinal distortion is less marked than in these two hosts. The average measurements of such infected corpuscles of Hoplichthys haswelli are 11·3μ by 6·5μ, as compared with the 10·4μ by 9·1μ of normal erythrocytes. When fully developed, the gametocyte occupies virtually the entire free space within the membrane of the host erythro-cyte (Pl. 129, fig. 14), and is curved about the host cell nucleus. The latter structure is displaced, usually in a lateral direction, and becomes longitudinally distorted as a result of its being compressed between the gametocyte and the cell membrane.
It is considered that on purely morphological grounds it is impossible satisfactorily to differentiate the gametocytes of H. hoplichthys n.sp. from those of H. ceolorhynchi n.sp. The differing reactions of parasitized erythrocytes of the respective hosts can hardly be viewed as evidence of specific distinction between the haemogregarines concerned. The two proposed species, in so far as these are at present known, do, however, differ in that intraerythrocytic schizogony terminating in the production of two or four merozoites occurs in the circulating blood in the one under discussion, while H. coelorhynchi n.sp.
has schizonts which attain a much larger size than those of the haemo-gregarine of Hoplichthys haswelli and do not appear to complete their development in the circulating blood. Because of the rarity of Hoplichthys haswelli, it may be a considerable time before it is possible to make a fuller comparison of its haemogregarine with that of Coelorhynchus australis and Physiculus bachus. Thus it is proposed that Haemogregarina hoplichthys n.sp. be considered a distinct species, the occurrence and form of its merozoites in the circulating blood of the host differentiating it from H. coelorhynchi n.sp. and from previously described piscine haemogregarines.
The type slide of H. hoplichthys n.sp. has been deposited in the collection of the Dominion Museum, Wellington (catalogue number Z15). One paratype slide is in the collection of the Department of Zoology, Victoria College, while another is in my own collection (catalogue number CHg 6).
Haemogregarina leptoscopi n.sp. (Pl. 129, figs. 16–33)
Haemogregarina leptoscopi n.sp. is described from heart blood smears from 2 of 4 specimens of Leptoscopus macropygus (Richardson) [stargazer, monk] trawled at 50 fathoms off Cape Campbell, Cook Strait, in March, 1948. The parasite rate averaged from 10 to 15 per 10,000 red cells in both cases, one of the infected fish being barely 15 cm. and the other 60 cm. in length.
Mature gametocytes of this species have not yet been identified. The stages present in my preparations are large free forms (Pl. 129, figs. 17 and 18), which are taken as being analogous with the larger of the two types of merozoites described by Reichenow (1910) from H. stepanowi, intraerythroblastic (Pl. 129, figs. 19, 21, 22) and intra-erythrocytic (Pl. 129, figs. 20 and 23) merozoites, and intraerythro-cytic schizonts (Pl. 129, figs. 24–33).
In both free and intracorpuscular merozoites about half of the body length is occupied by the nucleus, which is situated towards one (usually the narrower) extremity. With Giemsa the nucleus stains deep red, its chromatin aggregations appearing blackish red. The cytoplasm is alveolar, and stains light blue. Towards the anuclear end of the body is an area which stains deeper blue than the rest, and in the free forms this area invests a characteristically keyhole-shaped mass of lighter staining alveolar cytoplasm (Pl. 129, figs. 17 and 18). Numerous small granules staining a faint pink colour may be present in the cytoplasm (Pl. 129, fig. 18). Intracorpuscular merozoites bear a close morphological resemblance to the equivalent stages of H. coelorhynchi n.sp. Their average dimensions are 11·4μ by 2·8μ (range, 9·6μ to 12·2μ by 1·9μ to 3·2μ). Once again the cells initially invaded are late erythroblasts (Pl. 129, fig. 19, etc.). Host cells containing merozoites are not hypertrophied, and their nuclei are seldom displaced.
Schizonts (Pl. 129, figs. 29–33) have alveolar cytoplasm staining a light blue with Giemsa. A polar cap of red-staining material of similar appearance to that already discussed in connection with H. coelorhynchi and H. hoplichthys spp.n. is present at one extremity, although it is often (Pl. 129, fig. 30) of relatively small size and does
not extend across the full width of the haemogregarine. From one to several large vacuoles of up to 1·7μ in diameter are often present in contact with the edge of the polar cap (Pl. 129, figs. 30 and 32) or in the cytoplasm generally (Pl. 129, fig. 33). These resemble the vacuoles described by Henry (1912) from his “Type 3” form of Haemogregarina anarhichadis. Vacuoles are most numerous in schizonts in the more advanced stages of development (Pl. 129, fig. 33), and their probable purpose is to increase the available surface of the cytoplasm before the commencement of merozoite formation. The younger schizonts (Pl. 129, fig. 29) have 24 round to ovoid chromatin masses clearly defined against a diffuse red-staining background of chromatic material. At this stage the general appearance of the nucleus resembles that figured by Henry (1913) for his “Type II” form of Haemogregarina aeglifini, which is, however, interpreted as a macrogametocyte. Towards the end of the body opposite to that having a polar cap, there are frequently dense masses of cytoplasmic material staining deep blue (Pl. 129, figs. 29–31). Residual nuclear material is no longer evident in those schizonts having 24 completely separated chromatin masses illustrated in Pl. 129, figs. 30 and 31. Schizonts of this type are rarely found in the circulating blood of fish, although Fantham (1919) records a schizont having 12 nuclei from a blood film from the South African panga (Pagrus laniarus) and one with 6 nuclei, 3 of which were in process of a second division into two, from a haemogregarine-infected South African harder (Mugil capito). Occasional schizonts of H. leptoscopi n.sp. have 32 nuclei (Pl. 129, fig. 32), 8 of these resulting from a second binary division of 8 of the original chromatin masses. The most mature schizont seen (Pl. 129, fig. 33) has rounded up as a preliminary to merozoite formation. From the presence of such an advanced schizont as this in a heart blood smear, there is a likelihood that merozoite formation may, as in H. hoplichthys n.sp., take place in the circulating blood.
Ten of the vermiform schizonts measured at their greatest dimensions were found to average 13·1μ in length and 5·2μ in breadth (range, 12·5μ to 15·9μ by 4·3μ to 5·6μ). The rounded example figured in Pl. 129, fig. 33, measures 9·2μ in diameter. Erythrocytes containing schizonts become greatly hypertrophied, the average dimensions of the ten cells containing the vermiform schizonts referred to above being 14·6μ by 10·8μ (range, 13·1μ to 16·1μ by 9·6 to 11·1μ), as compared with the 12·0μ by 9·6μ of normal erythrocytes. The host cell nucleus is usually displaced to one side and mechanically distorted.
The nature of the haemogregarines illustrated in Pl. 129, figs. 24–28, has not been definitely ascertained. These forms have a similar staining reaction to the intracorpuscular merozoites, from which they are distinguished by the more central position of the nucleus and the presence of a well-defined red-staining polar cap at one end of the body. The average measurements for twenty examples are 13·9 by 3·5μ (range, 12·2μ to 15·6μ by 2·5μ to 4·8μ). They are thus of very similar dimensions to the gametocytes of H. coelorhynchi n.sp., but can be at once distinguished from these by the very much larger size of the nucleus (average, 6·4μ by 3·2μ), this structure measuring some 3·0μ by 2·4μ in the last-named species. In several respects, notably in the possession
of a large nucleus, they resemble the stages of H. rovignensis interpreted by Minchin and Woodcock (1910) as macrogametocytes, but they differ from these in their greater length (the broad forms of H. rovignensis never attain a greater length than 11·0μ) and in the larger size and more central location of the nucleus, the latter structure being terminally positioned in the European species. While these doubtful forms of H. leptoscopi n.sp. may indeed be gametocytes—indeed, if they are not, it is difficult to account for the absence of sexual stages of this haemogregarine from the circulating blood, for no forms comparable to the undoubted gametocytes of H. coelorhynchi n.sp. and H. hoplichthys n.sp. are present—there is strong presumptive evidence for identifying them as immature schizonts. There is a progressive size increase from the intracorpuscular merozoites illustrated in Pl. 129, figs. 19–23, through the doubtful forms to the schizonts in which reorganization of the nuclear material is taking place. This development is associated with an increase in size and adoption of a more central position by the nucleus; and the extent of the area occupied by the future merozoite nuclei and the residual chromatin in the schizont seen in Pl. 129, fig. 29, is not disproportionate to that taken up by the nucleus of the larger doubtful forms. Schizont formation on this pattern is to be expected, if the free forms illustrated in Pl. 129, figs. 17 and 18, are indeed macromerozoites analogous with those of H. stepanowi Danilewsky. It is thus concluded that the schizogony cycle described above for H. leptoscopi n.sp. is destined to terminate in the production of micromerozoites, which eventually initiate gametocyte formation; but that gametocyte formation has not yet begun in either of the infections studied, the doubtful forms described above as resembling (macro) gametocytes being in reality immature schizonts.
Counterparts to the mature gametocytes and immature schizonts of H. coelorhynchi n.sp. have not been recorded for the haemogregarine of Leptoscopus macropygus. Neither have mature schizonts comparable with those of the latter parasite been recorded for H. coelorhynchi n.sp. There is thus a possibility that future investigations may prove that these two apparent species represent stages in the life history of one and the same organism. The haemogregarine of Leptoscopus macropygus differs radically from H. hoplichthys n.sp. in its intracorpuscular schizogony, and in view of the differences between it and previously described piscine haemogregarines it is proposed to consider it as a distinct species, Haemogregarina leptoscopi n.sp.
The type slide of H. leptoscopi n.sp. has been deposited in the collection of the Dominion Museum, Wellington (catalogue number Z14), and a paratype slide is in my own collection (catalogue number CHg 5).
No complete life cycles of the haemogregarines of fish have yet been clucidated. Furthermore, some species have been described as new purely on the basis of their occurrence in a new host or in a new geographical region. In consequence, our knowledge of these parasites is in a state of confusion, and there is little doubt that future investigations will disclose that considerable synonymy exists. The available
evidence does evince that piscine haemogregarines are broadly separable into two categories. In the first of these a final schizogony terminating in the production of gametocytes take place in the circulating blood, each parasitized red cell thus coming to contain two or more mature gametocytes. Members of this group, distinguished by some authors by the designation of “schizohaemogregarines”, form a distinctive association to which it may ultimately be found necessary to accord separate generic status. In haemogregarines belonging to the second category micromerozoites which develop in the bloodstream or in the capillaries of internal organs invade red cells, and there develop into single gametocytes. This category may be further subdivided on the grounds of the presence or absence of polar masses of extranuclear material of uncertain nature, which stains red with Giemsa, in the older gametocytes and schizonts. It is to the group characterized by the presence of such polar masses—which may conveniently be called the rovignensis group—that the New Zealand haemogregarines described in the present account belong. In view of the confusion already existing in regard to the systematics of the haemogregarines of fish, it is with considerable diffidence that I propose three new specific names for addition to the already considerable list. When more is known concerning host-induced morphological variation among these parasites, and when complete life histories have been elucidated—the only too frequent use made of such phrases as “it thus appears likely” in this contribution bears eloquent testimony to our lack of exact knowledge in this respect—it may well eventuate that the criteria taken as being of specific significance are inadequate in this connection. Nevertheless, I have adopted the course of describing these New Zealand haemogregarines as fully as the nature of the available material allows, assigning as their specific names the generic name of the host (or of one of the hosts, in the case of H. coelorhynchi n.sp.), in the belief that this practice is less reprehensible than that of unduly “lumping” species.
Haemogregarines of the rovignensis group have not been described from Australian fish. Although the published work of several investigators indicates that at least 258 Australian marine fish of 68 species have been examined for haematozoa, only three species of Haemogregarina (all of which, from the descriptions given by Mackerras and Mackerras, 1925, appear distinct from the known New Zealand forms) have been recorded from 15 (6%) of these fish, belonging to 4 (6%) of the species concerned. Of the 500 New Zealand fish examined during the present survey, 115 were from fresh-water habitats, while 6 of the 52 species concerned were fresh-water ones. Forty (10·4%) of the 385 marine fish dealt with were parasitized by haemogregarines, 9 (19·6%) of the 46 marine species acting as hosts. The seemingly greater diversity of the haemogregarine fauna of New Zealand fish indicated by the comparison of these figures with the Australian ones is perhaps more apparent than real, for all four hosts of the species described in this account occur in Australia (McCulloch, 1927), although they have not yet been examined for haematozoa in that country.
Haemogregarines are described as follows from New Zealand marine fish, and their morphology and relationships are discussed: Haemogregarina coelorhynchi n.sp.; hosts, Coelorhynchus australis (Richardson), Physiculus bachus (Forster): Haemogregarina hoplichthys n.sp.; host, Hoplichthys haswelli McCulloch: Haemogregarina leptoscopi n.sp.; host, Leptoscopus macropygus (Richardson).
Bailey, J. K., 1948. Studies on a Haemogregarine from the Leopard Frog. Jour. Parasitol., 34, Section 2, 23–24.
Fantham, H. B., 1919. Some Parasitic Protozoa found in South Africa—II (Abstract). S. Afr. Jour. Sci., 16, 185–191.
—— 1925. Some Parasitic Protozoa found in South Africa—VIII. Ibid., 22, 346–354.
——, Porter, Annie, and Richardson, L. R., 1942. Some Haematozoa Observed in Vertebrates in Eastern Canada. Parasitol., 34, (2), 199–226.
Henry, H., 1912. Haemogregarina anarrhichadis from Anarrhichas lupus, the Catfish. Parasitol., 5, (3), 190–196.
—— 1913. A Haemogregarine and a Leucocytozoon from Gadus aeglefinus. Jour. Path. Bact., 18, 232–239.
Laird, M., 1951. Studies on the Trypanosomes of New Zealand Fish. Proc. Zool. Soc. Lond., 121, (2), 285–309.
McCulloch, A. R., 1927. Check List of the Fish and Fish-like Animals of New South Wales. Austr. Zool. Hndbk. No. 1, Roy. Soc. N.S.W., 2nd Ed., xxvi + 1–104.
Mackerras, I. M., and Mackeeras, M. J., 1925. The Haematozoa of Australian Marine Teleostei. Proc. Linn. Soc. N.S.W., 50, 359–366.
Minchin, E. A., and Woodcock, H. M., 1910. Observations on Certain Blood-parasites of Fishes occurring at Rovigno. Quart. Jour. Micr. Sci., 55, 113–154.
Muir, R., 1924. Text-book of Pathology. Edward Arnold Co., London. vii + 1–774.
Reichenow, E., 1910. Haemogregarina stepanowi. Arch. Protistenk., 20, 251–350.