De Meillon and Muspratt (1943) were the first to report sporangial germination. T h i s was observed in their type c which has affinity to Coelomomyces stegomyiae. T h e i r account of this process, somewhat ab
breviated, follows: A sporangium about to germinate loses its oil droplets and the interior becomes granular. A slight bulge appears on one side which enlarges until the outer hard shell ruptures and the thin internal membranes appear. T w o thin membranes become visible, and the con
tents of the sporangium flow slowly out and are confined within the in
ner of the thin membranes. As the spore mass enlarges, the two membranes become more distinct and more widely separated. U p to this stage no sep
arate zoospores could be seen, b u t with the pressure released the spores start
174 J. Ν. COUCH AND C. J. UMPHLETT
moving, slowly at first, b u t gradually gaining speed until the interior of the sporangium and the extruded membranes is a seething mass of zoo
spores. After a few minutes the spores push through the thin membranes and swim away.
Couch (1945a) reported sporangial germination b u t gave no details.
H e pointed out, however, that the structure of the sporangial wall, the method of spore development, and above all the detailed structure of the zoospores indicated a relationship to the Blastocladiales. Additional para
sitized Anopheles larvae containing Coelomomyces dodgei, C. punctatus, and C. lativittatus sent by Dodge from Georgia made possible more com
plete studies of resting sporangial germination (Couch and Dodge, 1947).
These studies confirmed in the main the observations of De Meillon and Muspratt (1943), b u t the interpretations of the structure a r o u n d the ex
truded spore mass differed. De Meillon and Muspratt interpreted the covering as two distinct membranes while Couch and Dodge (1947) con
sidered the covering as one m e m b r a n e (the inner sporangial wall) which gelatinized.
T h e first opportunity to repeat these observations came in the sum
mer of 1960 when Dr. Patrick L u m of Vero Beach, Florida, sent us abun
dant resting sporangial material of two new species or varieties related to C. psorophorae, one on Aedes taeniorhynchus, the other on Psorophora howardii. Dr. L u m h a d seen germinating sporangia in both of these be
fore sending us the material for study. W e were fortunate in getting these two species for germination studies since they are closely related to Mus-pratt's type c, the species used by De Meillon and Muspratt (1943). T h e sporangia of both germinated in large numbers. Those on Aedes taeni
orhynchus were germinating when received o n d a m p filter paper whereas those on P. howardii, also shipped on d a m p filter paper, showed n o signs of germination when received b u t began germinating in large numbers after being kept in water on slides in d a m p chambers for 4 days. T h e germination of several sporangia was followed in both species; the stages were recorded by camera lucida drawings and photographs.
After the slit has opened, one can see the slightly bulging sporangial contents b o u n d e d by the inner sporangial wall, which in section view is clearly double-contoured, i.e., it has an inner and an outer face with a hyaline center and is perceptibly thicker where exposed than where covered by the outer wall. T h e slit is formed 6 to 12 hours before the cleavage of the zoospores is completed (Fig. 13). At this stage, soon after the slit is formed, conspicuous and numerous lipoid bodies are fairly evenly dispersed through the sporangium and are larger than the lipoid bodies of the spores. D u r i n g and after cleavage the spores are so closely packed that individual spores cannot be seen, b u t one can tell when
cleavage is completed by the arrangement of the lipoid bodies. I n both the species on Psorophora howardii and the variety on Aedes taeniorhyn
chus, these are arranged in plates which viewed from the edge (40 χ objective with 15 χ oculars) appear as short rods, one plate of lipoid bodies for each zoospore. These "rods" are conspicuous u n d e r high dry magnification, and one can be sure when this stage is reached that cleav
age is completed and the spore mass is ready to emerge. T h i s can be hastened by transferring the sporangia to fresh water. It is likely that during the spore m a t u r a t i o n process the exposed inner sporangial wall has been modified chemically, making it more elastic and subject to hydrolysis.
T h e observations described here were m a d e on the species of Coelo
momyces related to C. psorophorae on Aedes taeniorhynchus. T h e pe
riod from the extrusion of the spore mass b o u n d e d by the inner spo
rangial wall (Fig. 10) to the dispersal of the spores lasts from about 8 to 10 minutes. T h e mass pushes out slowly at first b u t with gradually increasing speed, the outer covering stretching to accommodate the rapidly enlarging mass, until it reaches a roughly oval shape b u t with a distinct depression on top (Fig. 11). U p to this stage the cover, though obviously elastic, has shown no sign of gelatinizing. Suddenly the de
pression straightens out with a snap, the middle region of the covering appears to swell causing a rapid separation of the outer part of the wall from the m e m b r a n e retaining the spores. T h e outer part of the wall or covering fragments into several parts and moves away from the inner mass of spores now retained by a thin m e m b r a n e (Figs. 12, 14, 15).
It may be that the outer wall fragments are pushed away from the expanding spore mass by the gelatinization of the middle part of the old inner wall. T h i s would explain why all the separating fragments remain about equidistant from the spore mass (Figs. 12, 14, 15). At this stage the extruded spore mass is kidney shaped in side view (Figs.
12, 15) and circular in end view (Fig. 14). A feeble rocking motion now starts near the centers of the internal and external masses of spores, increasing in intensity a n d spreading throughout until the spores are seething with motion. T h e retaining m e m b r a n e or vesicle usually rup
tures about this time (Fig. 12) and some of the spores are shot out under obvious pressure, the remaining ones emerging by their own flagellar motion. T h e empty vesicle soon disappears completely. T h i s appears to be the usual course of events, b u t there may be variations in germination within and between the species.
Germination of the thick-walled sporangia has been reported in Coelomomyces (type c) close to C. stegomyiae and C. psorophorae and in Coelomomyces indiana (Muspratt's type a), the former by De Meillon
COUCH AND C. J. UMPHLETT
and Muspratt (1943), the latter by Muspratt (1946a), and in Coelomo
myces dodgei, C. punctatus, and C. lativittatus by Couch and Dodge a n d Coelomomyces psorophorae var. on P. howardii and Coelomomyces n. sp.
on Aedes taeniorhynchus by Couch and U m p h l e t t (first seen by Dr. L u m ) . Failure to get germination in his four types was reported by Walker (1938), by Muspratt (1946a) in his type b, and by U m p h l e t t (1961) in Coelomomyces pentangulatus.
D . Germination of Thin-Walled Sporangia
Muspratt (1946a) reported the germination of thin-walled as well as thick-walled sporangia of type a, C. indiana, and type c, near C. steg
omyiae. T h e thin-walled sporangia, according to Muspratt, germinate if the larval remains are left in water from the breeding place, libera
tion of the zoospores usually taking place within 3 to 6 days after the death of the larvae. T h e water, being turbid, was sometimes diluted with rain water or distilled water. Muspratt (1946a) found, on the other hand, that the thick-walled resting sporangia germinated only after being incubated dry at 28 °C for at least 2 or 3 weeks before being wetted again. I n the Coelomomyces indiana collected in December, 1961, a n d shipped to us fresh by Dr. Iyengar from India there were many colorless, thin-walled sporangia mixed with the brown thick-walled ones. Many of the thin-walled sporangia h a d started to germinate, b u t none h a d gone on to completion. I n shipment the sporangia h a d doubtless lost their viability and we were unable to induce any of them to complete the process. Several stages in germination were evident. I n the m a t u r e
FIGS. 10-15. Germinating thick-walled sporangia of Coelomomyces sp. o n Aedes taeniorhynchus.
FIG. 10. Bulge well formed, outer wall cracked open at the slit and with the hyaline inner wall protruding through the slit. Exposed part is perceptibly thicker than the covered part of the inner wall ( χ 910).
FIG. 11. Same sporangium, 1 m i n u t e later. T h e wall has stretched greatly and is slightly thinner than in Fig. 10 ( χ 910).
FIG. 12. Same sporangium, 8 minutes later. T h e outer part of the covering has n o w become separated from the inner and has fragmented. A rupture in the vesicle immediately surrounding the spore mass occurred at the left end a few seconds after the film for this p h o t o was exposed ( χ 910).
FIG. 13. T o p view of expanded germination slit just prior to extrusion of spore mass ( χ 910).
FIG. 14. End view of germinating resting sporangium showing spore mass extruded and outer layer of covering of extrusion floating freely ( χ 800).
FIG. 15. Side view of resting sporangium at about the same stage as in Fig. 14 ( χ 800). All living material.
178 J. Ν. COUCH AND C. J. UMPHLETT
thin-walled sporangia, a slit was present as in the thick-walled ones.
Beneath the slit in many sporangia was a clear area oval in outline.
T h i s clear area has the appearance of the gelatinous material formed in the papilla of the thick and thin-walled sporangia in many of the Blas-tocladiales, as shown by Couch (1945b) in Catenaria allomycis and C.
anguillulae. Other sporangia were seen in which the process of germi
nation was more advanced, showing a distinct lateral bulge and the slit wide open with the clear gel area between the protoplasm and the inner wall layer. A similar gelatinous (?) area has been seen in the resting sporangia of C. africanus in material preserved in formalin and in living sporangia of Coelomomyces psorophorae var. on P. howardii. It is pos
sible that this gel (?) may play a role in the dehiscence of both the thick-walled and thin-walled sporangia, in which case the account as given above of the germination of the thick-walled sporangia will have to be changed.
VI. LABORATORY AND FIELD INFECTION OF MOSQUITO LARVAE
A. Walker's Experiments
Manalang (1930) and Walker (1938) tried to induce resting sporangial germination in the laboratory, b u t without success. I n spite of this, Walker (1938) was able to carry out some significant experiments in the infection of laboratory-bred larvae. First he p u t mosquito eggs, fresh larvae, and p u p a e in water containing sporangia of Coelomomyces, b u t the results were negative. According to Walker (1938) larvae may ingest the sporangia in large numbers b u t pass them out unchanged.
U n a b l e in laboratory tests to get infection, Walker next tried field experiments. O n e particular pool, which h a d produced infected larvae throughout the year was selected as an experimental area. T h i s pool was flooded by irrigation from a nearby brook after it h a d been almost dry for 2 days. N o larvae of any kind were visible in the pool. A very fine-mesh copper sieve was placed in the pool and m a d e to float about half-submerged by means of corks. Twenty-one laboratory-bred A. gam
biae larvae were p u t in the sieve. D u r i n g the next 3 days larvae were removed and examined microscopically for signs of fungal infection, b u t none was seen. By the fourth day all b u t three of the larvae had disappeared from the sieve, and two of these were heavily infected with Coelomomyces africanus. T h e day after the area was flooded it was noticed that numerous tiny larvae, evidently hatching from eggs laid on the moist earth, were developing in other parts of the pool away from the sieve. Of these, 162 were examined and 66 percent were infected. Apparently the infected larvae harbored C. africanus. A shallow concrete tank was then constructed on the laboratory grounds where
the conditions of shade, sun, etc., were not unlike those of the pool.
Some thirty gallons of water, sediment, and vegetation were transferred from the pool to the tank. All the larvae seen, a total of ten, were removed from the tank. T w o of these were infected with Coelomomyces.
D u r i n g the next 2 days 175 laboratory-bred larvae of A. gambiae were added to the tank. Of these, 12 became infected d u r i n g the next 3 days.
Almost all the uninfected larvae p u p a t e d normally. Twenty-eight days after the concrete tank was filled, 1000 young larvae and some 600 eggs of A. gambiae were added, b u t none of the larvae or of the large num
bers of adults which emerged showed any signs of infection. T h e con
crete tank was then cleaned and filled with 30 gallons of clear water from the brook, well below the point where the irrigation ditch was led off. A b o u t 900 laboratory-reared larvae and eggs of A. gambiae gave an equally high yield of adults, b u t there were no infections. In these observations and experiments Walker found that only A. gambiae and rarely A. funestus were infected, although many other species were examined. H e does report finding on one or two occasions a small n u m b e r of sporangia in the larvae of Culex sp., b u t in each instance these were traced to the gut.
T h e most significant conclusion to be drawn from these experiments is that infection occurred only when the larvae were in contact with water plus soil which contained the inoculum, as in the freshly flooded pool or in the cement tank to which water and sediment from the pool were added. N o infection occurred when fresh larvae were added to dishes containing only water and sporangia, b o t h fresh and dried.
B. M u s p r a t t ' s Experiments
Muspratt (1946a, b) reported experiments indicating that the source of infection may be the zoospores liberated from the sporangia. I n the first experiment (Muspratt, 1946a) small potholes, about 18 inches deep, were dug in the m o p a n e clay soil of an infected area after a heavy rain, when the soil was saturated with water. Larvae of A. gambiae, together with water were then transported from an uninfected habitat and p u t into the potholes; some of the larvae were kept as controls. T h e larvae in the potholes became infected after about 48 hours. T h i s experiment indicates, according to Muspratt, that the zoospores are able to travel through the soil water for some distance. Attempts to infect larvae with zoospores in the laboratory were unsuccessful.
In another experiment performed later the same year, Muspratt (1946b) extended the experimental infection studies carried out by Walker (1938). Muspratt (1946a) m a d e a rather thorough study of the Coelomomyces parasites of mosquitoes a r o u n d Livingstone in N o r t h e r n
180 J. Ν. COUCH AND C. J. UMPHLETT
Rhodesia during the rainy seasons 1941 to 1945. As indicated in Section III he found heavy infections of Anopheles gambiae larvae in the sur
face pools in areas where the soil was a heavy, dark brown, loamy clay, locally called " m o p a n e " clay. Muspratt h a d about 100 pounds of nearly dry clay shipped from Livingstone to Johannesburg. From the same area several h u n d r e d infected A. gambiae larvae packed full of thick-walled sporangia were collected and p u t into jars containing water and soil from the breeding place. After the larvae were dead, the soil was allowed to become nearly dry, and the lids were placed on the jars.
T h e soil and larvae were left in the laboratory over winter. T h e fol
lowing summer, 8 months later, a concrete trough was constructed and placed outside the laboratory in a position where it was exposed to 3 to 4 hours of sunlight daily. T h e dry m o p a n e clay was d u m p e d into the center of the trough and the soil in the bottles containing the larvae filled with resting sporangia of Coelomomyces (type a) was scattered a r o u n d the edge of the m o u n d of soil in the trough. R a i n water was then poured over the m o u n d of soil until the trough was full, and larvae of A. gambiae, hatched from eggs in the laboratory, were p u t in the trough. T h e water was allowed to evaporate to dryness every 2 or 3 weeks and left dry for 3 or 4 days. It was then refilled with fresh rain water and a new batch of laboratory-hatched larvae was p u t in. T h i s schedule was repeated several times. After the water had evaporated once and the trough was refilled, about fifteen out of a h u n d r e d larvae of the second batch became heavily infected and a few in later batches. Un
fortunately the experiment h a d to be discontinued because of the dif
ficulties of getting A. gambiae eggs and because the climate of Johan
nesburg was too cool for the proper growth of the larvae. I n conclusion Muspratt says: "Although the above experiment did not prove that indefinite infection of A. gambiae larvae by the fungus can be obtained in a confined space, I feel confident that, given suitable climatic condi
tions, this would be the case."
Muspratt (1946a) has perhaps carried out more field observations on the infection of the larvae than any other worker. H e states that the larvae may become infected during any stage of their development, that even the very smallest larvae may contain fungal thalli and sporangia, having contracted the infection a day or two after hatching. T h e infec
tion may also appear in the blood fluid just before the larva pupates, and in this case the sporangia develop in the p u p a or adult. H e states that in the majority of cases in which the infection is contracted as late as this, the growth of sporangia has been sufficient to kill the p u p a or adult.
According to Muspratt the weather has an important influence on
the infection of the larvae of Anopheles gambiae. First there must be a dry spell followed by regular intermittent rainfall, enough to fill the pool basins; this must be followed by fine weather. U n d e r such con
ditions the larvae become infected when 3 or 4 days old, in the late first or the second instar. T h e infection can be recognized in such larvae as m i n u t e specks in the body fluid about a week after the pools have been filled by rain. W h e n there is n o rain after the pools have first been filled, all the larvae become infected at about the same time; b u t when rain is continuous this is less evident. T h i s more or less simultaneous infection of the larvae, Muspratt suggests, occurs when the pools have reached a certain concentration (of mineral salts) as a result of evap
oration.
Muspratt found that infected larvae of A. gambiae, if left in the natural habitat became so packed with sporangia that their bodies were discolored by the fungus, b u t if such larvae were brought into the lab
oratory early, such heavy infestations did not develop. Muspratt sug
gests that this may be caused by the u n n a t u r a l conditions of the lab
oratory, particularly the lack of sunlight. As another possible explana
tion, he suggests that continual infection of the larvae may occur in nature, as indicated by the presence of small thalli intermingled with sporangia.
These observations by Muspratt are interesting because of the prob
lems they suggest. W e must admit that at the present time practically nothing is k n o w n about infection by these fungi. From the observations of Walker (1938) and Muspratt (1946a), it would seem that infection may occur at any stage in larval development. However, finding young thalli in the fourth instar does not necessarily m e a n that the infecting agent has just penetrated. It could have entered earlier and developed slowly.
W e have n o idea about the portal or portals of entry of the parasite and we do not know what are the agents of infection. These problems can best be solved by carefully controlled experiments.
C. Infection Experiments of Couch and Dodge and of Umphlett