GOJRATI, Hassan Ali Navvab, 1935 EPIZOOTIOLOGICAL SURVEY Of AVIAN I:iALARIA IN THE HAWAIIAN ISLANDS. ' University of Hawaii, Ph.D.

71.,..12,222 GOJRATI, Hassan Ali Navvab, 1935 EPIZOOTIOLOGICAL SURVEY Of AVIAN I:iALARIA IN THE HAWAIIAN ISLANDS. ' University of Hawaii, Ph.D., 1970 Entomol,ogy University Microfilms, A XEROX Company, Ann Arbor, Michigan THIS DISSERTATION HAS BEEN MICROFILMED EXACTLY AS RECEIVED

EPIZOOTIOLOGICAL SURVEY OF AVIAN MALARIA IN THE HAWAIIAN ISLANDS A DISSERTATION SUBMITTED TO THE GRADUATE DIVISION OF THE UNIVERSITY OF HAWAII IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY IN ENTOMOLOGY SEPTEMBER 1970 Hassan Ali Navvab Gojrati Dissertation Committee D. Elmo Hardy, Chairman Andrew J. Berger Mercedes D. Delfinado Wallace C. Mitchell Minoru Tamashiro

ACh~OWLEDGMENT8 I would like to express my sincere appreciation to many individuals and organizations for their assistance and cooperation in various aspects of this study. Mr Jack Throp, Director of the Honolulu Zoo for allowing complete freedom in trapping and examination of birds at the zoo. Dr. Joseph E. Alicata of the University of Hawaii for examination of some of the smears; Dr. Allen Y. Miyahara of the University of Hawaii for very essential technical assistance; Miss Elaine M. L. Chang, technician of the Board of Agriculture; Mr. James K. Ikeda of the State Department of Health for providing information on mosquito rearing; Mr. Winston Banko of Hawaii National Park for great assistance in mist-netting the birds at higher elevations; Mr. John L. Sincock of the U. S. Fish and Wildlife Service, Koloa, Kauai, for providing large numbers of slides from birds trapped at high-elevations on Kauai in addition to slides from Nihoa Finch, Nihoa Millerbird, Laysan Finch, Laysan Duck, etc. The Pathology section, Queen1s Medical Center for generously permitting me to use some of their laboratory facilities; Dr. Robert S. Gesowitz of the University of Hawaii for examination of slides and technical assistance; and Dr. Marshall Laird of the University of New Foundland, Canada, for indispensable assistance in diagnosis and identification of species of Plasmodium.

TABLE OF CONTENTS Page ACKNOWLEDGMENTS....... ABSTRACT.......... LIST OF TABLES................. INTRODUCTION................ HISTORICAL BACKGROUND........ TAXONOMIC POSITION OF THE AVIAN MALARIA PARASITES..................... Class Sporozoa Leuckart 1879.......... Order Haemosporida Danilewsky 1886...... Genus Plasmodium Marchiafava and Celli 1885. DEVELOPMENTAL CYCLE OF AVIAN PLASMODIA. MATERIAL AND METHODS................. Trapping and Handling the Birds..... Collecting and Handling the Blood..... Collecting and Handling the Mosquitoes... Description of the Collecting Area. Rearing and Feeding the Mosquitoes. Transmission Experiment......... Technical Obstacles............. HISTOLOGICAL METHODS............... Preparation of Slides for Making Blood Smears. Making Blood Smears............ Staining Procedure.......... Buffered \\later............. Technique for Dissecting the Salivary Glands of Mosquitoes.......... Technique for Preparation of Smear of the Salivary Glands for Sporozoites Study............ Technique for Dissecting the Mid-gut of the Mosquitoes for Oocysts Study. Technique for Preparation of Smear of the Mid-gut of Mosquitoes for Ookinetes Study............ ii v vii 1 5 9 9 10 10 12 16 16 18 20 20 21 22 24 25 26 26 27 27 29 30 31 33

iv Technique for Sectioning of Liver for Exoerythrocytic Study................. RESULTS AND DISCUSSION......... Information on Mosquitoes........ Haemoproteus Infections in Birds. Leucocytozoon Infections in Birds. CONCLUSIONS................. SUMtv1A.RY..... LITERATURE CITED.............. 34 38 48 51 53 55 58 60

ABSTRACT Avian malaria is believed to be one of the main factors in the extinction of the native Hawaiian avifauna. An epizoological investigation of the disease was conducted in the Hawaiian Islands primarily to determine whether avian malaria was present in Hawaii. The birds were collected from the islands of Oahu, Maui, Kauai, and Hawaii, at different elevations for a period of three years. Blood samples were taken either from the main wing vein, or from the toe nail. Blood smears were prepared and stained with Giemsa1s stain. Histological preparations for the study of the exoerythrocytic stage of the parasites were also made. A total of 4,988 blood smears from 2,604 birds representing 38 species of native and introduced birds was examined. Plasmodium circumflexum Kikuth, f. gallinaceum Brumpt, f. cathemerium Hartman, and probably f. matutinum (Huff), were found in six species of native and introduced birds. Significantly, these records are the first authenticated reports of these parasites from the Pacific Islands, and this is the first time that f. circumflexum has been reported from this part of the world. The presence of f. gallinaceum indicates that this parasite has accompanied its host to many parts of the tropics. Blood smears taken from pigeons and doves indicated that some of these birds were infected with Haemoproteus and Leucocytozoon.

It was found that 65 percent of the pigeons were infected with vi Haemoproteus. this pathogen. The doves, however, did not seem to be infected by Although both pigeons and doves were infected by the Leucocytozoon, the incidence was relatively low. percent of the two species of birds were infected. Only four This is the first record for Leucocytozoon from the Hawaiian Islands. The presence of the larvae of the potential vector of avian malaria, Culex pipiens quinguefasciatus Say, in a water container at 6511 feet elevation indicates that probably one of the limiting factors in the vertical distribution of Culex is the availability of suitable habitats.

LIST OF TABLES TABLE PAGE I BUFFERED WATER FOR USE WITH GIEMSA r S STAIN... 27 II OCCURRENCE OF PLASMODIUM IN NATIVE AND INTRODUCED BIRDS IN HAWAII. 39 III PREVALENCE OF PLASMODIUM IN NATIVE AND INTRODUCED BIRDS IN HAWAII. 42 IV DISTIRBUTION OF INFECTION RATES BY ISLAND AND ELEVATION. 44

INTRODUCTION Research on avian malaria has offered to microbiologists, physiologists, biochemists, and many other scientists tools for solving many of their problems and this K~ll continue to be true regardless of whether human malaria is eradicated. Moreover, the growing feeling that eradication of malaria is just around the corner has given some people the idea that malarial research is no longer necessary. Because malaria has not as yet been eradicated, our tools for controlling it are rusting away leaving us unprepared for combatting an unexpected return of this great scourge. In this respect research on avian malaria provides us with a means for solving problems of basic importance to many fields of biology. The discovery that birds are subject to malaria by Danielewsky in 1885 followed closely after the finding of malarial parasites in the blood of man by Laveran (1880). Since then, the course of experimental observation with both organisms has proceeded in a parallel course. Unfortunately, the study of bird malaria, and its possible important role in the extinction of the rich Hawaiian avifauna, has been almost completely neglected. It is believed that, the parasites of aaimals and their vectors have gained entrance with the importation of infected animals from various parts of the world. Because of the mild climate and other favorable factors, these parasites have become established.

2 It is characteristic of the parasitic diseases that they do not cause immediate high mortality, and more often their pathological symptoms develop gradually. This has been the case in the Hawaiian Islands. It seems that parasites did not attract the attention of the Hawaiians until about 1800 when the American and European ships began to call frequently. Since that time, parasites have been found in many species of birds and mammals (Alicata, 1969). According to Hardy (1960), the night-biting mosquito (Culex pipiens quinquefasciatus Say) apparently was imported into the state of Hawaii in water casks on a ship (or ships) from Mexico between 1826 and 1830. There are a number of accounts as to how they first arrived here. Van Dine (1904) says, ITprevious to the year 1826 mosquitoes were unknown in Hawaii. During that year they were brought to the port of Lahaina, on the Island of Maui, in the ship TWp.llington T from San BIas, Mexico. 1T Furthermore, up to the year 1826 there was no word in the Hawaiian language for mosquito. The native term is TMakika T, a corruption of the word mosquito (Hardy 1960). On the basis of available evidence, Hardy (1960) also states that the two species of day-biting mosquitoes, Aedes aegypti (Linnaeus) and ~. albopictus (Skuse), apparently did not reach the islands until a much later date. Avian diseases aid greatly in controlling numbers of birds, but we know very little about these in Hawaii. With the exceptions of reports by Alicata (1939) on Haemoproteus columbae Kruse and of Fisher and Baldwin (1947) on Plasmodium vaughani Navy and Mac Neal,

3 practically no information is available. Warner (1968) reported on the susceptibility of drepaniid species to avian malaria and birdpox. He concluded that avian malaria was probably one of the main factors in the extinction of native Hawaiian avifauna in the lowlands. The principle objective of this paper was to acquire more information concerning avian diseases. Therefore, an epizootiological investigation of avian malaria was undertaken in the Hawaiian Islands. This threeyear study confirmed the presence of Plasmodium, determined the species involved, the kind and proportion of the birds infected, the mosquito that probably serves as vector, and the relationships between Plasmodium, mosquito, and bird. It was demonstrated in the laboratory that Culex pipiens quinquefasciatus Say is capable of transmitting Plasmodium to canaries. It also has provided information concerning the prevalence of Haemoproteus and Leucocytozoon infection among Hawaiian birds. In a broader sense, it was possible to study the ecology of avian malaria in the Hawaiian Islands in the hope of contributing to the epizoology of Plasmodium infection in native and wild birds.

HISTORICAL BACKGROUND The first mention of intracorpuscular parasites in the blood cells of birds was made by Danilewsky in 1885. Without knowledge of the description of malarial organisms in man by Laveran (1880), Danilewsky briefly but accurately described similar parasites from birds, although at the time he did not know what they were. Danilewsky described three types of parasites from the blood of birds. The first type occurred free in the plasma and appeared to him to be a type of gregarine or "little blood worm." The second blood parasite was a trypanosome, also noted free in the plasma. A third was found to be both intracorpuscular and, at certain times, free-swimming in the plasma; we now know this to be a species of bird malaria parasite. During the ten years which followed his discovery of malaria parasites in birds, Danilewsky published several papers describing further studies, chiefly in French and German journals. In 1889 Danilewsky published the monograph "La Parasitologie Comparee du Sang," in which he described his observations ffi<::.de on the malarial parasites of wild birds obtained in Southern Russia. Danilewsky continued his investigations with the malaria parasites of birds, and in 1890 published three papers in the Annals of the Pasteur Institute. The first of these deals chiefly with a description of Leucocytozoon. In a second paper his observations on the phagocytosis of haemogregarines and malaria parasites by white blood cells were presented. A

5 distinction is made in the third paper between acute and chronic malaria infections in bi~ds, and a general comparison is given of malaria in birds and man. Following Danilewsky, Grassi and Feletti were next in importance in the early history of bird malaria. In 1890 these authors reported two distinct types of intracorpuscular parasites in sparrows (Passer hispaniolensis) and in pigeons Columba livia. The genus Laverania was established to include the crescent-shaped parasites already described by Danilewsky in birds, as well as similar parasites in man described by Laveran. The parasite in birds was called Laverania danilewsky and in man, Laverania malariae. I believe the latter name is still used by some modern investigators, although others have placed it in synonomy as Plasmodium falciparum. The second parasite described by Grassi and Feletti was closely related to the amoeboid forms of the human malaria parasite. In fact, it was their belief that this parasite was identical with the parasite of human malaria. This has given rise to an interesting situation regarding the correct scientific name for the parasite with which they were working. They suggest the genus Haemamoeba and proposed the name~. praecox for both the bird malaria parasite and the human parasite. A second species was described as ~' immaculata, because of the supposed absence of pigment. Later they changed the name ~. praecox to H. relicta retaining the name H. praecox for the parasite in human malaria. During the Grassi and Feletti period, Kruse in 1890 proposed the

6 generic name Haemoproteus for the crescent-shaped bodies found by Danilewsky in the blood cells of birds. It is of interest that the discoverer of malaria parasites in man should play a part in the early researches on bird malaria. Laveran began to publish a series of papers on this subject, being particularly interested in the use of birds as experimental laboratory hosts and in the relationships between human and bird malaria parasites. In 1891 he published two papers on the general characteristics of malaria parasites in birds, and in the same year presented photomicrographs of both bird malaria and human malaria to the Academy of Sciences in Paris. Following Grassi and Feletti, the two most important Italian workers in the early studies on bird malaria were Celli and San Felice (1891). They reclassified bird malaria parasites, but later work indicated that they were dealing with mixed infections of Haemoproteus and Plasmodium. A more concrete contribution by Celli and San Felice was the first successful transfer of malaria parasites from bird to bird by blood inoculation. In 1894 Labbe's monograph appeared, in which blood-cell parasites were studied from a purely zoological point of view. He then outlined his own classification and suggested the generic name ProteosOffia for the parasites which Danilewsky described as Cytosporon and Grassi and Feletti as Haemamoeba. The endocorpuscular parasites of birds were not seriously considered in America until thirteen years after their discovery in Europe. In 1898 Eugene Opie of the John Hopkins University in

7 Baltimore published a paper on the haemocytozoa of birds. Opie examined a number of species of native American birds and found malaria parasites in several of them. One of the most important in the history of malariology was the discovery of the process of fertilization in Plasmodium falciparum (Welch) by MacCallum in 1898. Thus a sexual cycle in malaria was demonstrated. MacCallum also described the pathology of bird malaria infections. In the same year (1898), the role of mosquitoes as vectors of malaria was described by Ross, a discovery based in part on MacCallum1s description of the fertilization process and Sir Patrick Manson1s encouragement. Koch (1899) verified the work of Ross by successfully transmitting avian plasmodia from bird to bird by mosquitoes. He was also the first worker to transmit bird malaria to canaries. The volume of literature on avian malaria in the 20th Century is immense, many of the contributions having dealt with drug screening and other studies related to malaricidal properties of various drugs. More than 30 species of Plasmodium have been described from birds of various kinds (see Hewitt, 1940), but fewer -than half are presently regarded as valid (see Levine, 1961; Bray, 1957; Laird and Lari, 1958). There are also a great many contributions on other aspects of experimental research, such as studies on immunity, genetics, in vitro cultivation, interrelationship with other infectio~s, and basic physiology and biochemistry. A brief but excellent review of representative contributions to these and other fields of experimental research on avian malaria has been presented by Huff (1963). Besides,

8 several other workers such as Coatney and Roudabush (1937), Coatney and West, (1938), Herman (1938, 1944), Herman et al. (1954), Kikuth (1931), Levine and Hanson (1953), Manwell (l935, 1938), and Wolfson (1941) have made contributions to our knowledge of malariology. In the period between the appearance of Hewitt1s (1940) monograph and this study, a very great increase in the amount of experimental work on avian malaria can be seen. The increase was largely due to the increased importance of human malaria during World War II. Avian malaria parasites were already known to be excellent screening agents for antimalarial compounds, and in fact had contributed significantly to the chemotherapy of malaria before World War II. Also, significant research was carried out on Leucocytozoon and Haemoproteus infections. The history of malariology will record this period as one in which research on avian malaria played a most important part in gaining an understanding of the mode of transmission of malaria. From this great amount of interest and effort many new problems were uncovered and new techniques developed. In the State of Hawaii the research which has been conducted on avian malaria is very limited. According to Alicata (1939), pigeons in Hawaii are commonly infected with Haemoproteus columbae Kruse. Blood smears from 101 adult pigeons in Honolulu showed 83 to be infected; and of a total of 25 vectors, pigeon fly Pseudolynchia canariensis (Macquart) dissected, nine or 36 percent were infected with ~. columbae (Kartman, 1949). The true avian malarial organism, Plasmodium vaughani Navy and Mac Neal, in the Pekin Nightingale,

9 Leiothrix lutea, has been reported by Fisher and Baldwin in 1947. This parasite was noted in one of 11 birds examined from Hawaii National Park, island of Hawaii. TAXONOMIC POSITION OF THE AVIAN MALARIA PARASITES Kudo (1966) proposed the new classification of the Protozoa in which the phylum is divided to two subphyla: Plasmodroma and Ciliophora. The subphylum Plasmodroma is characterized by possession of one to many nuclei of one kind and flagellae or pseudopodia or no organelles of locomotion. This group is subdivided into four classes: 1) Mastigophora, 2) Sarcodina, 3) Sporozoa, and 4) Cnidosporidia. In the subphylum Ciliophora are placed those protozoa which possess two kinds of nuclei (macronucleus and micronucleus) and have cilia or similar locomotor organelles in at least one stage of development. It is subdivided into Ciliata and Suctoria. Each class is composed of several orders. However, we are primarily concerned here with the class Sporoaoz and specifically with the order Haemosporida. Therefore, a very brief discussion of this group is presented. Class Sporozoa Leuckart 1879 Members of the Sporozoa are parasitic and produce spores. They possess no organs of locomotion except in the gamete stage. Reproduction is asexual by binary or multiple fission (schizogony) or sexual (gametogony). Gametogony leads to the formation of a zygote which in turn initiates the process of sporogony or spore formation.

10 The classification of the class Sporozoa has occasioned much discussion, but, for the purpose of this paper, that proposed by Kudo (1966) is being followed. Kudo divided the Sporozoa into four orders: 1) Gregarinida, 2) Coccidida, 3) Haplosporida, and 4) Haemosporida. Order Haemosporida Danilewsky 1886 The Haemosporida require both vertebrate and invertebrate hosts to complete its life cycle. Schizogony occurs in vertebrates, and gametogony and sporogony occurs in blood-sucking invertebrates. The order is divided by Kudo into three families, namely: Plasmodiidae, schizogony in the peripheral blood of vertebrates. Pigment present. This family contains one genus of importance- Plasmodium, the malarial parasite of man and other animals. Haemoproteidae, schizogony in the endothelial cells of inner organs; only gametocytes appear in the peripheral blood. Pigment present. Two genera of interest, Haemoproteus and Leucocytozoon, occur in this family. Babesiidae, small, non-pigmented parasites of erythrocytes. Genus Plasmodium Marchiafava and Celli 1885 The true malarial organisms belong to the genus Plasmodium, which in turn is closely related to Haemoproteus and Leucocytozoon. The principal significant difference between Plasmodium and the other two genera is that the asexual stages (schizonts) of the

11 former occur in erythrocytes of the circulating blood, while those of the two latter genera occur in the internal organs (lung, liver, spleen, kidney, etc.). As a result, Plasmodium can be transmitted regularly from one susceptible host to another by injection of infected blood from the vessels or heart, whereas in the case of the other two this procedure will result in infection only at certain times when merozoites are in the blood by chance, because the only stages ordinarily in the blood are gametocytes which can o~~y ~o~t~~ue development inthe proper invertebrate host. Like Haemoproteus, Plasmodium contains pigment. It is of special interest that mammalian malarias are carried by Anopheles mosquitoes, while those of birds are generally carried by culicine (Culex, Aedes) mosquitoes, although some of the latter also have anopheline vectors. Most of the avian species of Plasmodium are far less host specific than mammalian forms. Some occur naturally in a considerable number of species of wild birds, and some have been adapted by experimental passage to the development in birds in which they are not known to occur naturally. Only a few species have been reported as naturally infecting domestic birds, and it is not known whether all of these are of veterinary importance. In the Haemoproteus, a part of the developmental cycle occurs in Hippoboscid flies, commonly called Louse flies, and the only proven vector is Psuedolynchia canariensis (Macq.). However, it has been shown by Fallis and Wood (1957) that a biting midge, Culicoides (possibly Eiliferus) Root and Hoffman, is an intermediate host and

12 transmitting agent of ~. nettionis Coatney in ducks. The vectors of Leucocytozoon are members of the genus Simulium, and Simulium venustum Say has beer recognized as an important vector species. DEVELOPMENTAL CYCLE OF AVIAN PLASMODIA A major advance in the understanding of the life cycle of the malarial organism was made by the discovery that infective sporozoites did not enter erthrocytes directly, but rather developed as exoerythrocytic forms in cells of the reticuloendothelial system prior to invasion of the erythrocytes. Following the introduction of the sporozoites from infected culicine mosquitoes, numerous pre-erythrocytic schizonts are found in the macrophages and fibroblasts of the skin near the point of entry. These are referred to as cryptozoites. Merozoites from this first generation of pre-erythrocytic schizonts form a second generation of pre-erythrocytic schizonts, the metacryptozoites. Merozoites from the metacryptozoites enter erythrocytes and other cells of the body and in the latter form exoerythrocytic schizonts. In the case of P. gallinaceum, ~. relictum and ~. cathemerium, these other cells are endothelial cells, but in the case of P. elongatum and R. vaughani they are cells of the haemopoietic system. In some species of avian plasmodia, e.g. ~. gallinaceum, and P. elongatum, the exoerythrocytic developmental stages may be added to by forms which are derived from the erythrocytic cycle. These are

13 known as phanerozoites, being derived from the merozoites of the schizonts in the erythrocytic cycle. The erythrocytic cycle is initiated seven to 10 days after infection by merozoites from metacryptozoites and at other times by merozoites ~rom exoerythrocytic schizonts located, according to species, in the endothelial or haemopoietic cells. On entering the red blood cell, the merozoite rounds up to form a trophozoite. This is a small rounded form containing a large vacuole which displaces the cytoplasm of the parasite to the periphe~ of the cell. The nucleus is situated at one of the poles, giving the young form a!signet ring! appearance when stained with Giemsa. The early trophozoites undergo schizogony to produce merozoites, the number produced depending on the species of parasite. During the process of schizogony, the parasite takes in host cell cytoplasm by invagination, haemoglobin is digested and the residual haematin pigment is deposited in granules within the food vacuoles. Apparently, schizogony may continue indefinitely, the length of each cycle of schizogony depending on the species of parasite. The release of merozoites from the schizonts occurs synchronously in the host, and in human malaria this is associated with a paroxysm of fever. Fever does not appear to be a significant part of the syndrome in avian hosts (Russell et al., 1963). -- After a number of asexual generations has occurred, some merozoites undergo sexual development with the formation of microgametocytes and macrogametocytes. Levine (1961) claims that the

'14 female forms should be referred to as macrogametes since they possess a haploid number of chromosomes. The haploid nature is maintained throughout the whole of the life cycle of the malarial parasite, except that a diploid state is found following fertilization and zygote formation. The female forms are generally more numerous than the male forms, and they stain more intensely blue with Giemsa than do the male forms. In addition, of course, the nucleus of the microgametocyte is more diffuse than in the female cell. Further development of the gametocytic stages can take place only when the blood is ingested by a suitable mosquito. Development in the mosquito is rapid. Within 10 to 15 minute~ the nucleus of the microgametocyte divides, and through a process of exflagellation, six to eight long, thin, flagella-like microgametes are extruded from the parent cell. These remain attached to the parent cell for a few minutes, lashing actively; they then become detached and swim away to find, and fertilize, the macrogamete. The zygote resulting from fertilization is motile and is called an ookinete. This ookinete penetrates the mid-gut mucosa and comes to lie on the outer surface of the stomach, forming an early oocyst about 50 1)-60 u in diameter. The nucleus of the oocyst divides repeatedly to produce a very large number of sporozoites. These are about 15 u in leng~h with a central nucleus. Maturation of the oocyst takes a variable period of time depending on the species of parasite, temperature, and the species of mosquito; but in general, it is 10-20 days. When mature, the oocyst ruptures, liberating the sporozoites into the body cavity of the mosquito, and these then

15 migrate allover the body of the mosquito but eventually reach the salivary glands. Here they may lie intracellularly, extracellularly, or in the ducts of the salivary glands. They are now infective to a new host, infection occurring when the mosquito takes a blood meal. A mosquito remains infected for its life span, transmitting malarial parasites every time it takes a blood meal.

16 MATERIAL AND METHODS Trapping and Handling the Birds Birds were collected in a variety of ways. The majority of birds used in this work were captured by trapping. Numerous birds were sampled by mist netting, and collections were occasionally made by shooting. Two types of traps were utilized. One of these traps which was successfully used for capturing most of the common low-land birds was a square-mesh wire netting funnel trap 42 inches wide, 60 inches long, and 42 inches high. The trap had a funnel- or cone-shape entrance opening from both ends, ailu Cl small door on one side for releasing the birds. The trap was baited with natural bird food and placed in the Honolulu Zoo. Sometimes, for capturing a particular kind of bird, a few individuals were intentionally placed in the cage to attract other members of the same species. Another kind of trap utilized for capturing House Sparrows was a door trap. This trap is frequently called a lfpull-string lf trap, and is a device which is merely an adaptation of the old and wellknown lfsieve lf trap. It is easily made at little expense; and although not usually automatic in its operation, it is probably the best trap for a new operator to use until he has acquired proficiency in handling birds. The door trap used in this experiment was a squaremesh wire netting 18 inches wide, 24 inches long, and 10 inches high. The trap was baited with natural bird foods and placed in the backyard of a residential area.

Several mist nets of different lengths and widths were also used 17 for capturing the birds. The desired height of the net was adjusted with the help of telescopic poles. The mist nets were placed among the trees where the birds were actively moving from one place to another. The birds were collected at low elevations on the islands of Oahu and Maui; at different elevations on Kauai; and at 4,000-6,500 feet elevations in the Volcano National Park on Hawaii. All the birds collected in the National Park and some of those collected for me from the island of Kauai were banded and rele~sed after they were sampled. Some of the birds were captured and blood smears made by John L. Sincock from the islands of Kauai, Nihoa, Necker, Laysan, and Midway. These were: NewellTs Shearwater, Wedge-tailed Shearwater, Bulwer's Petrel, Red-tailed Tropicbird, Laysan Duck, American Coot, Golden Plover, Black-necked Stilt, Gray-backed Tern, Barn Owl, Nihoa Millerbird, Elepaio, Iiwi, Anianiau, Nihoa Finch, Laysan Finch. If the bird was sacrificed (as in the case of Sparrows, Canaries, and few Mynahs), the liver was preserved either in 70 percent alcohol or ZenkerTs Fluid.* The former was used only when the latter was not available. The liver was fixed in Zenker's Fluid for approximately 24 harrs. It was then washed in several changes of 70 percent alcohol and finally preserved in 80 percent alcohol. Good results were obtained by using Zenker's Fluid. * K2Cr207-2.5 grn.; HgC1 2-5 to 8 gm.; distilled water - 100 mi.; glacial acetic acid (added at time of use) 5% by volume.

18 Collecting and Handling the Blood Blood was collected from the wing vein or by clipping the center toe nail of the birds. Some of the trapped birds were banded and released after being sampled, and some of them were released without banding. An attempt was made to minimize the handling period of the small birds prior to venipuncture. After removing caged birds from cages or wild birds from mist nets or traps, the birds were handled very gently for short periods before processing, because, even without venipuncture, small caged birds and small wild birds occasionally die if they are handled or manipulated excessively. Such handling or manipulation reduces the chance for survival following venipuncture. It had special importance in the case of this study because I was not allowed to kill the birds. All preparations for venipuncture procedures were made before the blood was taken. These procedures included the banding and the parting of the feather in the blood-collecting site. Venipuncture was the last procedure before the bird was released. If blood collection from the wing vein was desired, one of the wings was held open so that the undersurface of the wing was facing the operator. The wing vein could be seen through the skin by parting the feathers and exposing the surface of the skin of the wing near the birdts body. In most birds, especially the large ones, the wing vein could be seen without much difficulty. In very small birds, application of slight pressure with the thumb against the wing bone facilitated the detection of the wing vein.

19 With the bird in the hand of the partner and the wing vein exposed, a 20-gauge needle attached to a 10 ml. syringe was inserted in the superficial blood vessel of the wing. About 5 mi. of venous blood was taken and poured into a clean blood-collection tube containing one drop of Sequester-Sol* as an anticoagulant agent. Care was taken to use the proper quantity of anticoagulant to prevent any distortion or change in the quality of the blood collected. The blood-collection tube containing the mixture of the blood and anticoagulant was inverted four to five times to insure the complete mixture of the blood with the anticoagulant. This was the procedure when blood was carried to the laboratory for the preparation of the blood smear. When the blood arrived at the laboratory, it was placed in the refrigerator for future use. If the smear was made at the same time that the blood was taken, slightly different procedures were applied. A fingernail clipper was used to clip a toe nail about midway between the base and the end of the vessel visible in the nail. The nail was clipped in an anteroposterior direction, because this tends to dilate the vessel. The-blood was collected either from the wing vein or toe nail directly to a heparinized capillary tube 75 mm. long and 1.4 to 1.6 mm. in diameter. The blood collected in this way was put directly on a clean slide and the smear was immediately made. In general, bird blood coagulates rapidly. Because no anticoagulant was used in * Active ingredient: Dipotassium Ethylenediamine Tetraacetone

20 this method, the smear had to be made very rapidly. In small birds, the blood flows very slowly, so that the blood had to be sucked out by attaching a rubber tube to the capillary tube. Blood flow was stopped almost immediately by application of firm pressure to the toe nail or wing vein with the thumb and forefinger. Collecting and Handling the Mosquitoes Description of the Collecting Area. The mosquitoes, Culex pipiens quinquefasciatus Say, were collected in two different ways. First, by collecting the eggs from the field, and second by collecting larvae in a small can containing water outside of the laboratory. The former were collected for me in the field by the State Department of Health, Mosquito Control Division. The mosquito egg rafts were collected at the Hawaiian Maid farm in Ewa, Hawaii. These were found in a poultry effluent ditch which formed from the overflow of the chicken watering trough. The old feed and chicken feathers in the water provided a slightly acid media (ph=6.5) in the ditches. The fermenting feed attracted many rotifers and other pollution-hardy organisms. The highest concentration of mosquito eggs was amongst the weeds along the ditch and between patches of duckweed, Lemna spp. Heavy breeding is common along the edge of the ditch but the larvae in the deeper areas are eliminated by the guppies, Lebistes reticulatus

21 (Peters). Food sources available to the adult mosquitoes include the chickens and birds that frequent the poultry houses. Rearing and Feeding the Mosquitoes. The rearing procedure discussed below is commonly used by the Mosquito Control Division. Egg rafts were collected and placed on damp Kimwipes in a baby-food jar. Initial field collections were made by transporting late instar larvae in plastic bags. This method proved to be detrimental to the larvae since more than 50 percent were dead upon arrival at the laboratory Five to six egg rafts (approximately 1000 eggs) were placed into a 6 x 8 x 12 inch plastic tub which was filled to a depth of two to three inches with tap water aged for one to two days. If the water was not aged, a pinch of Brewers! yeast was added to the fresh tap water to deoxygenate the media. Hatching occured after 24 hours and a small portion of a one-to-one mixture of Purina Dog Chow and Brewers! yeast (relatively inactive) were provided for larval food. Subsequently, daily portions of the same larval food was given, but care was taken to prevent overfeeding because this caused a scummy layer to appear on the water surface which made it necessary to transfer the larvae to a fresh container. By careful addition of the food, no scum developed. After seven to eight days in the larval stage, the insec~ pupated. The pupae were removed from the rearing tubs by siphoning them through a modified pipitte. This pipitte was essentially a syringe bulb fitted with a wide mouth glass tube (J.D. = 3/16 11 ).

22 The pupae were placed into small dixie cups and placed in 12 x 12 x 12 inch screened cages for emergence. A sleeve was attached to one side to facilitate the admission of birds and withdrawal of mosquitoes. Newly emerged adults were fed a five percent honey or sugar solution which was placed in a baby-food jar with a dental wick protruding th~ough a hole in the lid. The adults were able to feed on the material on the wick. They were also given raisins as additional food. Four days after emergence, the females were offered blood meals from an immobilized bird, as will be discussed later. Deposition of eggs by the engorged mated females took place three to five days after the blood meal. The oviposition media offered was aged water in a 1000 mi. beaker. Transmission Experiment. In connection with the experiment upon infectivity and transmission, two conditions had to be fulfilled-- first, the birds had to have large numbers of gametocytes in their blood, and second, there had to be a supply of adult mosquitoes in readiness for biting. This necessitated the taking and examination of smears frequently to determine whether or not the bird was in condition satisfactory for the feeding experiment. The birds were exposed to the mosquitoes four days after emergence of the latter. At this age; the mosquitoes theoretically should have shown a great tendency for a blood meal, but practically it was not always so, as will be discussed later. The birds used in the present work as a source of the pathogen were House Sparrows, (Passer domesticus), which were previously trapped and found to be infected with malaria by blood examination.

23 In order to allow the mosquitoes to feed on the diseased bird, the bird was immobilized by tying the feet, the wings, and the beak snugly with a piece of masking tape, and the bird was placed on the top of the breeding chamber in such a way that its exposed breast would be accessible to the mosquitoes. The feathers of the bird were then parted in the pectoral region and wetted down. The bird could be left in this position for at least an hour without apparent discomfort. When the female mosquitoes had engorged upon the blood of the infected bird, the bird was removed. Then, all engorged females in the cage were picked out individually and transferred to another breeding chamber which also was provided with a long sleeve. In so doing, it was necessary to wear a rubber glove to prevent the mosquitoes from engorging from my hand, for they might then be mistaken for mosquitoes which had had the infective meal. The engorged mosquitoes were kept from eight to ten days and then dissected. It was soon found, under the prevailing laboratory conditions (76 0 F. - 82 0 F temperature and 65-75% relative humidity), that the oocysts of the parasite reached their maximum size on the tenth day following the feeding. The mosquitoes were kept routinely thereafter for ten days. After a few had been anesthetized and dissected, and the presence of oocysts were observed, the rest of the mosquitoes in the chamber \'~ere kept long enough (i. e. usually about two weeks) to insure the finding of sporozoites in their salivary glands. I found that it was advisable to dissect a few mosquitoes daily in order to observe the progress of infection and the day of

24 invasion of the salivary glands. No attempt at carefully counting the number of oocysts in the stomachs was made, because this count could in most cases have been only an estimate, due to the great number present and the difficulty of viewing all sides of the stomach without duplicating the count of certain oocysts. When the presence of sporozoites in the salivary glands of the mosquitoes was observed, they were allowed to feed on healthy birds. Five uninfected canaries, Serinus canaria were then exposed to the mosquitoes which had been fed on the infected birds fifteen days before, and they were allowed to be bitten by mosquitoes. In this way, the pathogen was transferred from the diseased House Sparrows to the healthy canaries. Canaries are usually used as an experimental animal for the study of malaria parasites because of their high susceptibility to the pathogen. A period of six to eight days elapsed until the parasites first appeared in the blood of the canaries bitten by infected mosquitoes. An attempt was made to allow the mosquitoes to repeatedly feed on the infected birds because it is never possible to be quite certain that the gametocytes are matured. But, only partial success was obtained. Technical Obstacles~ A great deal of difficulty was encountered at first in getting the adult females to feed upon birds. Three factors were essential for success in this attempt. It was found necessary to keep the mosquitoes away from water for at least 24 hours before attempting the feeding experiment. During extremely

25 hot weather, it was necessary to keep the breeding cages over moistened cotton so as to keep the air moist, but in such a way that the mosquitoes could not imbibe the liquid. The other important factor was darkness. Nearly all individuals of the species of mosquitoes dealt with in this experiment would bite much better in darkness than they would in the light. Hence all of the feedings ~ere made at night. Another problem in infecting the mosquitoes was that a certain percentage of the mosquitoes which were fed on parasitized birds failed to become infected even though other mosquitoes fed on the same bird exhibited full development of the parasite. DifficUlty was frequently encountered in getting the mosquitoes to take a second blood meal. Some died immediately after depositing eggs; others drowned in the water tray; and some died shortly af~er the first feeding before the blood meal was digested. Many refused to feed a second time although they were given an opportunity to do so. It was necessary, therefore, in order to insure enough mosquitoes for successful transfers, that considerable allowance be made for the difficulties involved in carrying the transmission to completion. Histological Methods In searching for blood protozoa in general, thick or thin smears of the blood are prepared and stained with one or another of the Romanowsky (methylene blue-eosin combination) stains. Thick smears are preferred to thin ones for mammalian blood because their use

26 permits one to examine a relatively large amount of blood in a relatively short time. However, they cannot be used for avian blood because of its nucleated erythrocytes. The protozoa may be distorted in thick smears enough so that much care and practice is needed to differentiate and interpret the species, especially of the malaria parasites. Therefore, I used only thin smears. Among many different kinds of stains available, Giemsa has been the most frequently used by different investigators of avian Plasmodia. Rapid stains, such as Wright1s and Field1s stain, are used only if speed is necessary, because they stain unevenly; and they are not as precise as the slow stains. Preparation of Slides for Making Blood Smears. The first step in making a good thin film and easily diagnosable smear is to have very clean slides. Therefore, the slides were rinsed in 95 percent alcohol and wiped with a cj~an cloth Making Blood Smears. To prepare a thin smear, a small drop of fresh blood was placed at one end of a slide, so that it was about 1 1/4 inch from the end. Quickly, the underside of a second spreader slide, the corner of which has been cut away, was touched to the drop of blood which would spread along the edge of the spreader slide in contact with the lower slide. Then, with the spreader slide held at a 30 _40 angle, rapidly drew (without pushing) the spread drop over the horizontal lower slide, and a uniform thin smear resulted. well-spread smear appeared very pale with a feather-shaped end. The The smears were allowed to dry in the air (a matter of a few seconds)

27 until they changed color. The host number or species was scratched on the slide with a diamond-point stylus. The smears were fixed and stained within four or five hours of smearing. Staining Procedure. After the thin blood smears had dried, they were immersed in absolute methyl alcohol (acetone-free) for three to five minutes for fixation. The slides were removed and allowed to dry in the air. Then, they were placed in GiemsaTs stain. The Giemsa's stain was prepared fresh and diluted before staining. Buffered Water. A phosphate buffered water having a ph of 6.5 to 6.8 was used for the dilution of the stock Giemsa and for rinsing the stained slides. The two salts were mixed thoroughly, as shows in Table I, in a mortar and 1 gm. to 2000 c.c. of distilled water was used. It was necessary to vary the amounts of phosphates to obtain the desired results. Red is increased by lowering, and blue is increased by raising the ph. TABLE I. BUFFERED WATER FOR USE WITH GIEMSATS STAIN ~ 6.5 6.8 gm 2.723 4.539 gm 8.316 5.940 Each drop of GiemsaTs stain was diluted with 1 c.c. of such a buffer. After dilution of Giemsa's stain, the smears were placed in the

28 stain for twenty to thirty minutes; then they were removed and washed with just enough buffered water to remove the excess stain. The slides were then placed on end on a piece of blotting paper, and allowed to dry. The nuclear chromatin stains garnet or TIlby red and the cytoplasm a delicate sky blue, thus contrasting the reddish-purple nuclei of the leucocytes and thrombocytes (platelets). Many investigators do not cover blood smears because they think it is unnecessary to cover stained blood preparations unless they are to be observed with dry objectives. I covered the blood smears, and it seemed to me that this makes the color slightly more brilliant and the small details a little sharper. Moreover, if a given slide had to be checked several times, there was no danger of destroying the erythrocytes. If a smear was covered, a neutral medium such as diaphane was used, because fading may otherwise be very rapid. Diaphane-mounted Romanowsky-stained blood films will keep for years without much deterioration. All smears were examined under oil immersion objectives of a compound microscope for a sufficiently long period, usually at least ten minutes before the result was considered negative or positive. Great care was exercised to avoid mistaking the blood platelets accidentally superimposed upon red cells for malarial parasites. These platelets are frequently surrounded by an unstained halo. Precipitated stain, dirt, or bacteria may constitute other sources of error. It should be mentioned that a thorough working knowledge of the

29 thin blood film, i.e. the appearance of the normal constituents of blood, of the more common pathological changes in the blood cells, as well as of the different species of Plasmodia in their various stages, is necessary before attempting to learn to identify malaria parasite in a thin film. The thin film has the great disadvantage of failing to reveal a great number of positive cases, particularly of light infections. One also should be aware of the fact that the identification of species of naturally occurring malaria is not easily made from single blood smears even by specialists. Due to the lack of a protozoologist and expert on avian malaria in the Islands, I have spent a great deal of time familiarizing myself with the most commonly found species by consulting P. C. C. Garnham1s (1966) book. Finally, an attempt was made to assign each organism to the proper species. After preliminary identification of species, the slides were sent to Dr. M. Laird for final diagnosis. Technique for Dissecting the Salivary Glands of Mosquitoes. Several methods have been suggested by different investigators for extracting salivary glands from the mosquito. I have adopted the Shute and Maryon (1966) method. These authors clearly describe and illustrate four stages of the process. This is one of the most usual methods for the removal of the salivary glands, and it is the modification and refinement of the method which is given by Barber and Rice (1936) and Giovannola (1934). Chloroform was used to anesthetize the mosquitoes. The wings and legs were then removed. The fly was laid on one side of a slide where a small drop of normal

30 saline was available. The head was severed by a sharp cut. Then, pressure was applied with a sharp dissecting needle to the side of the thorax near the base of the fore-legs whereupon the salivary glands would emerge, usually accompanied with a drop of a haemocoelic fluid. The salivary glands were separated quickly from adjacent structures and covered with a drop of saline. The glands were dissected in saline solution and were examined under the low power objective of the microscope without any coverslip in order to confirm the presence of sporozoites. They were then placed in a one percent aqueous solution of osmium tetroxide, kept at a ph of 7.3 by MichaelisTs buffer. Fixation proceeded at 32 _38 F. for half an hour, after which the glands were dehydrated rapidly through a series of graded alcohols. After a third change of ethyl alcohol, the salivary glands were stained for half an hour in a one percent solution of phosphotungstic acid in alcohol. Finally, a rinse in absolute alcohol was given, and they were embedded in Araldite. For more detailed information concerning this procedure, the reader can refer to Shute and Maryon (1966) and Barber and Rice (1936). Technique for Preparation of Smear of the Salivary Glands of Mosquitoes for Sporozoites Study. To prepare the smear of the salivary glands, the slide containing the salivary glands was transferred to the microscope and the specimen was brought into focus with low power, using a reduced light. The glands were placed in the center of the slide, and a large square coverslip was dropped onto