Plasmodium pteropi Breinl, 1913
(Text-figure 1, Figs. 1–36)

Plasmodium pteropi was originally described from the flying fox (fruit bat) Pteropus gouldi from Queensland, Australia. Although the report containing Breinl's description was published in 1913, it bears the date 1911 in its title. This discrepancy has caused some authors to fall into error and attribute the date of description variously to 1911 (e.g., Manwell, 1946) and even 1912 (e.g., Wenyon, (1926). Bhatia (1938) mistakenly gave West Australia as the type locality, and further stated that Johnston described Plasmodium pteropi as new in 1913, giving as the reference the page number of the description in Breinl's report. The error probably arose through the fact that Johnston collaborated with Breinl in these studies.

The earliest reports of plasmodia from fruit bats are those of Ziemann (according to Breinl, 1913) and Durham (1908), the latter author recording "a small malaria-like parasite" from the red cells of Pteropus natalis on Christmas Island, Straits Settlements. Mackie (1914), without knowledge of Breinl's paper, described as new Plasmodium pteropi from (Pteropus edwardsii) = Pteropus medius in Assam, India. Wenyon (1926) considered it probable that Breinl and Mackie were dealing with the same species, although neither the descriptions nor the figures of either author are full enough for a confident assertion on this point to be made. The parasite recorded as Plasmodium pteropi from the epauletted bat Epomophorus gambianus in Senegal, by Leger and Leger (1914), is stated by M. Leger (in a footnote to Rodhain, 1926) to be conspecific with Plasmodium epomophori Rodhain, 1926.

Donovan (1920) mentioned finding Plasmodium in the blood of the Fijian flying fox, without making it clear whether his material was obtained in Fiji. Wenyon (1926) listed the Javanese Pteropus hyomelanus as a host for Plasmodium pteropi, on the basis of an examination, by himself and Scott, of material from a bat which had died in the Gardens of the Zoological Society of London during 1925. Coloured illustrations of Plasmodium pteropi (?) in Wenyon (1926) were drawn from blood smears collected by Manson-Bahr from "the flying fox in Ceylon" (Pteropus medius?). Scott (1927) reported P. pteropi from Pteropus medius, in the Gardens of the Zoological Society of London. Manwell (1946) recorded Plasmodium from Pteropus gouldi and Dobsonia moluccensis from New Guinea, considering that the page 5 parasites from the former host "probably belong to the species seen by Breinl (1911) in the same host, and named by him Plasmodium pteropi." Pteropus scapulatus and Pteropus conspicillatus were listed by Bearup and Lawrence (1946) as additional Queensland hosts for Plasmodium pteropi. The host list and distribution of this parasite were further extended by McGhee (1949), who recorded it from Pteropus geddiei and Pteropus eotinus at Espiritu Santo, New Hebrides, and from an unidentified member of the family Pteropodidae at San Fernando, La Union, Philippine Islands.

Breinl's description of Plasmodium pteropi consists of an account of the broad morphological features and the staining reaction with Giemsa. It mentions that the host cell is sometimes slightly enlarged, also that "the parasite, during its growth, replaces the cytoplasm of the red blood corpuscles, and finally the remainder of the blood corpuscle surrounds the parasite in the form of a thin pellicle." Breinl stated that gametocytes occur in fairly large numbers, and described and figured what he apparently took to be young schizonts. Mackie (1914) called attention to the fact that gametocytes were very numerous in his material, while young ring forms were not common. The last-named author's figures are not very informative, and, as Bearup and Lawrence (1946) suggested, his "schizonts," together with those referred to by Breinl, might well have been male gametocytes. Manwell (1946) described pigment-free, extracellular segmenters and schizonts from his smears of the peripheral blood of Pteropus gouldi. Neither Bearup and Lawrence (1946) nor McGhee (1949) found segmenters or exoerythrocytic forms in the blood or organs of the bats which they examined, although these authors studied a large number of infected specimens.

The present account is based on material from two heart-blood smears from a heavily parasitized mature example of Pteropus nawaiensis (Gray) shot in the lower reaches of the Wainimbuka River, Viti Levu, Fiji, during July, 1948. My own visit to this locality in May and June of the following year was a little too early in the season to allow of my obtaining more material from P. nawaiensis. The only bat from which I was able to obtain smears was another member of the Pteropodidae, the long-tailed fruit bat Notopteris macdonaldii Gray, 135 examples of which were shot in a limestone cave beneath the native village of Kalabo, a few miles inland from the mouth of the Nasinu River. None of these bats, which were representative of all ages, were parasitized by haematozoa. Fifty of the nycteribiids with which they were infested likewise proved negative for these parasites.

The cytoplasm of the signet ring forms of Plasmodium pteropi (Figs. 3–9) stains a very pale blue with Giemsa, while the nucleus, which is usually somewhat reniform in shape (Figs. 5–7), stains deep red. Appreciable hypertrophy of the host cell (Fig. 8) is very seldom apparent. More usually the size of a parasitized cell remains unchanged (Fig. 3, etc.), but sometimes (Figs. 4, 7, etc.) it is distinctly less than that of the smallest uninfected erythrocytes of the host (Fig. 2). Such a reduction in size of parasitized host cells is characteristically found in Plasmodium malariae infections (Wenyon, 1926). The diameter of the ring forms ranges from 1.9μ to 3.7μ, averaging 2.2μ, while the size of nuclei of the reniform type ranges from 0.8μ by 0.4μ to 2.2μ by 0.5μ. Double infections by small ring forms (Fig. 9) are common. The cytoplasm of two ring forms may fuse in such a way as to simulate early stages of schizogony (Figs. 10, 11). Partial nuclear fusion may also take place, giving rise to an appearance (Figs. 12, 13) strongly suggestive of the band forms characteristic of Plasmodium malariae. It is possible that the page 6 parasites interpreted by Breinl and Mackie as young schizonts may in reality have been such fused ring forms. A thorough search of my material failed to reveal the presence of any schizonts whatever.

Growth beyond the signet ring stage may be initiated by a reduction in size of the vacuole associated with a broadening and slight vacuolation of the cytoplasmic ring (Fig. 14), or by the formation of anastomoses (Figs. 15–18). Some developing gametocytes (Figs. 19, 20) still preserve the large vacuole of the signet ring

Text-figure 1

Figs. 1 and 2: Erythrocytes of Pteropus nawaiensis (Gray). Figs. 3–36: Plasmodium pteropi Breinl, 1913, from P. nawaiensis. Fig. 37: Erythrocyte of Salarias periophthalmus Val. Figs. 38 and 39: Haemogregarina salariasi n.sp., from S. periophthalmus.

page 7 stage, although their cytoplasm has increased very considerably in bulk. The fact that the nucleus of such forms is intimately connected with this vacuole suggests a likely origin of the vacuole associated with the nucleus of many older gametocytes (Figs. 22, 23, 33, 35, etc.). Rodhain (1926) stated that a similar clear area is associated with the gametocyte nucleus of Plasmodium epomophori. The cytoplasm of young macrogametocytes stains a deep greenish-blue and contains numerous small vacuoles (Figs. 22, 23, etc.). A varying number of rather fine pigment granules are present, the earliest stage encountered showing pigment being that illustrated in Fig. 19. These granules are of a dark golden-brown colour. The cytoplasm of erythrocytes containing older gametocytes is often apparent only as an uneven rather dark staining pellicle surrounding the parasite, as Breinl (1913) has already pointed out (Figs. 22, 23, 35, etc.). Rodhain (1926) remarked that gametocytes of Plasmodium epomophori completely replace the cytoplasm of the host erythrocyte, so that no trace of the outline of the cell is left.

Macrogametocytes of P. pteropi (Figs. 21–30) are usually round or more or less ovoid in shape. A few band forms (Fig. 25) occur, and these again call to mind the very similar forms of Plasmodium malariae. The cytoplasm of macrogametocytes of P. pteropi, unlike that of the equivalent stages of Plasmodium epomophori, often contains small vacuoles. The nucleus is compact, and stains densely. This structure is of irregular outline, and is often rather cruciform in shape (Figs. 22, 23, etc.). Both the cytoplasm and the nucleus of the largest macrogametocytes (Fig. 30) are lighter staining than in the smaller examples, the pigment granules are more apparent, and the cytoplasm is differentiated into light and dark staining areas. These facts suggest the possibility that such large forms may have been somewhat distorted during the smearing or drying operations. Macrogametocytes range in size from 4.4μ by 3.1μ to 7.6μ by 6.0μ, and average 5.4μ by 4.5μ.

Microgametocytes (Figs. 31–36) are also round to ovoid in shape. The cytoplasm takes a very light blue stain, finely diffuse pigment often giving large areas of it a light golden colour (central shaded area, Fig. 31). The nucleus is larger than that of the macrogametocyte, and is frequently marginal in position (Figs. 31, 33). Those microgametocytes which appear to be most mature have a triangular nucleus (Figs. 34, 36), which stains very light pink. Microgametocytes of P. pteropi differ markedly from those of Plasmodium epomophori in regard to the shape of the nucleus, that of the latter species being of regular round or oval shape (Rodhain, 1926). The cytoplasm scarcely stains at all, while numerous rather large pigment granules of a dark golden-brown colour are very conspicuous. Microgametocytes are less variable in size than macrogametocytes, ranging from 4.8μ by 3.8μ to 5.8μ by 5.0μ and averaging 5.3μ by 4.6μ.

The search for the vector of Plasmodium pteropi has so far proved fruitless. The uniformly negative results of Bearup and Lawrence (1946) and McGhee (1949) in their examination of ectoparasitic Nycteribiidae from fruit bats infected with P. pteropi do not support Manwell's (1946) suggestion that by analogy with the hippoboscid transmission of Haemoproteus such ectoparasites might prove to be the vectors. At all events the present record serves to eliminate Anopheles from the search, as mosquitoes of this genus do not occur in Fiji. Perhaps further studies aimed at the examination of day-biting jungle mosquitoes which are in a position to be able to bite the bats in their roosting places might be productive of more fruitful results.

Both the smears from Pteropus nawaiensis discussed in this account are in the author's private collection (catalogue numbers HP1 and HP2).