Coelomomyces opifexi (Pillai & Smith). Coelomomycetaceae: Blastocladiales II. Experiments in sporangial germination

Hydrobiologia vol . 38, 3-4, pag . 425-436 . 1971 .

Coelomomyces opifexi (Pillai & Smith) . Coelomomycetaceae : Blastocladiales . 1 . Its distribution and the ecology of infection pools in New Zealand . by J . S . PILLAI

Department of Microbiology, Medical School, University of Otago, New Zealand .

INTRODUCTION

In recent years there has been considerable interest in the fungi belonging to the genus Coelomomyces as biological control agents of mosquito populations . Although a number of species have been described from different mosquito larvae hosts, in various habitats, very little has been published particularly on aspects of the ecology of their infections in nature . This information is necessary before a species can be mass produced in the laboratory for testing its usefulness as a biological agent . In one such investigation, it was shown that the infection of Anopheles quadrimaculatus larvae by Coelomomyces punctatus in a part of a lake in North Carolina, U .S .A., was related to a rise in water temperature and a lowering of oxygen tension (UMPHLETT, 1968) . Other species of the fungus are known to attack mosquito larvae utilizing temporary or transient breeding habitats (MUSPRATT, 1963) or brackish pools (LUM, 1963), yet nothing is known about the ecology of the infection pools of these Coelomomyces . After the discovery of Coelomomyces opifexi from two coastal mosquitoes in New Zealand, the physical ecology of the infection pools was investigated . During these studies the range of distribution of the Coelomomyces within New Zealand became known and the significance of this is discussed . Coelomomyces opifexi is the only species to be recorded in New Zealand so far . It was isolated in the first instance from the larvae of the endemic mosquito Op fex fuscus (HUTTON) (PILLAI & SMITH, 1968) Received Dec 3, 1970 . 42 5

and later from the larvae of the introduced Aedes australis (ERICHsoN) (PILLAI, 1969) . Both species are known to breed all the year around in supralittoral pools along the coast line in a zone above the meanhigh level water neap, but continually influenced by sea spray or splashing . The salinity factors of the pools fluctuate irregularly and sometimes rapidly (McGREGOR, 1964) . However, the mosquitoes are well adapted to this habitat as they are capable of breeding in pools of wide ranging salinities from fresh water to hypersaline conditions (MCGREGOR, 1965) . Opifex fuscus is the shoreline mosquito of New Zealand . It occurs over most of the coastline except for a stretch south of Dunedin in the South Island . Here it appears to have been replaced by the introduced Aedes australis . The distribution of the two species overlaps near Dunedin on the Otago Peninsula (Fig . 1, inset A) . sr

4

V eifington

ull hristchurth

Fig . 1 . Map of New Zealand . North Island only part shown . Stippled area on S .E . coast of S . Island is present known distribution of the introduced mosquito Aedes australis in New Zealand . Insets A & B show main foci of C . opifexi activity in New Zealand . MATERIALS AND METHODS

Hydrogen ion concentration, temperature and salinity readings were taken from surface samples of infected and non-infected pools . Hydrogen ion concentration was read on a Beckman Meter and salinity was determined by titrating standard silver nitrate (27 .25 grams per litre) against aliquots of pool water using potassium chromate as indicator . Repeat samples from the same pool, unless otherwise stated were collected at intervals of a month or more . 426

FIELD OBSERVATIONS

1. Distribution of C . opifexi Since June 1967 when the presence of C . opifexi was first discovered, larvae from supralittoral pools have been systematically examined from various parts of New Zealand . These studies have indicated that the activity of the C . opifexi was confined to the distribution range of Aedes australis . Within this region there appeared to be 2 main foci of infection . One of these was on the Otago Peninsula where it attacked both host larvae and the second locality was approximately 50 kilometers down the coast at Catlins, where Aedes australis was the only mosquito present (insets fig. 1) . 2 . Infection pools of rocky outcrops In the two localities the infection pools were situated mainly on rocky outcrops inside bays sheltered from the prevailing south or south-westerlies . These outcrops, because of their resistant nature, constitute minor irregularities along precipitous sections of the coastline . They are terrace-like ledges rising from 5 (at sea level) to about 20 meters above sea level at highest point . The surface weathering of soft strata and their excavation by erosion, has provided numerous clefts and trenches which accumulate water and

Fig . 2 . Supralittoral pool on Puddingstone Rock, Otago Peninsula (site A) with permanent colony of mosquito Opifex fuscus infected with C . opfexi .

427

form suitable habitat for host mosquitoes (figs 2 & 3). Although breeding activity of the host was randomly distributed in pools all over the surface of these outcrops, the infection pools were situated mainly on higher elevations or some distance away from the sea. Periodically, however, infected larvae were encountered in pools close to the splash zone. Such pools were either protected from the sea by a rock wall barrier or the infection appeared after a prolonged period of calm sea conditions. I n the two localities where Coelomomyces was active, only 21 pools out of a total of more than 400 contained the infection. A few of these pools were permanent infection sites while in others host larvae were attacked only intermittently. In an effort to describe the conditions surrounding infection of larvae, measurements of temperature, hydrogen ion concentration, pool size, bottom sedimentation, salinity factors and biotic composition of infection and non-infection pools were made.

Fig. 3. Supralittoral rock pool with Aedes australir larvae infected with C. opifexi at Long Point outcrop.

With the exception of salinity readings (see below) these studies did not reveal any significant differences between infected and noninfected pools. The main features of the observations can be summarised as follows:

1 . Temperature

For infected pools the mean temperature was 13°C for 49 readings, with a range of 5°C (winter) and 28°C in summer . For uninfected pools : mean 12 .5 °C for 81 readings, range 5 °C (winter) and 29°C (summer) . (Average mean air temperature for this latitude is 11 °C, winter mean 7 °C and summer mean 15 °C.) 2. Hydrogen ion concentrations

Hydrogen ion concentration of 54 samples taken over a 3 year period at intervals of one month or more from 20 infected pools gave a range of 6 to 8 .53 (7 .29 mean) . The corresponding figures for uninfected pools ranged from 6 .3 to 8 .75 and the mean for 133 readings was 7 .24 . 3. Pool size

Both infected and uninfected pools ranged from relatively small structures, less than a litre in volume to very large bodies containing 10,000 litres or more . The larger pools were mainly permanent breeding habitats and the small ones were utilized during summer months when the mosquito activity reached its peak . 4 . Bottom sedimentation

Pools close to the splash zone contained more of sand and rock fragments or even boulders . Pools further back on the outcrop had a bottom sedimentation of silt or soil . 5. Biota of pools

Biotic composition depended mainly on stability of the pool . Pools less exposed to splashings contained the following organisms, large numbers of the orange copepod Tigriopsus fulvus (FISHER), Cladocerans, amphiopds, ostracods, chironomid larvae, larvae of the syrphid Ephydrella novazealandiae (TONNOIR & MALLOCH) and quite commonly strands of green alga Enter morpha spp . Pools situated in the splash zone, because of the transient nature, often lacked the main elements of the stable pools, though after a period of calm, some of the organisms did appear temporarily . 6. Salinity levels of infected and non-infected pools

429

A total of 166 samples were determined for salinity from 21 infected and 64 non-infected pools over the three year period . They ranged from 0 .60 to 20 .8°/00 NaCl from 54 samples for infected and 8.0 to 42 .5 from 112 samples from non-infected pools . Varying amounts of fluctuations have been recorded in both infected and non-infected pools . For example an increase from 1 .5 to 20.8° /00 NaCl (fourteen fold) within 8 days and a decrease of 17 .0 to 1 .00 /00 (seventeenfold) within 3 days following heavy rainfall was recorded for infected pools . The corresponding figures for uninfected pools were from 6 .0 to 29 .9 (fivefold) within 2 days and a decrease of 23 .0 to 1 .20 (nineteenfold) in 8 days . However, despite these fluctuations recorded from some of the pools, the overall trend showed that infection pools were generally of a lower salinity than non-infected pools . This is demonstrated by the grouping of the salinities recorded in units of 5 .0°1 00 NaCl . It will be seen that 30 out of 54 (or 55%) readings of infected pools were in the category of 5.0°J oo , as against 18 out of 112 (or 16 .0%) for non-infected pools . With an increase in salinity there were fewer infected pools whereas the non-infected pools were more evenly distributed in salinities up to 20.0°1 00 NaCl . (Table I) . The figures confirm field observations that infection was present mainly in pools situated in sheltered aspects of the outcrop or removed some distance away from the splash zone . It seems probable that a few of these pools constitute a permanent source of Coelomomyces . Long term observations of selected infection pools appear to confirm this . For example pool A situated at the Long Point outcrop in Catlins . This is a largish structure with an estimated volume of 10,000 litres which serves as permanent habitat for Aedes australis . Although situated within the splash zone it is usually sheltered from splash and/or spray by a rock barrier . In March 1970, 2nd to 4th instar larvae were found to be infected in a population estimated at approximately 100 larvae per litre volume . New infections continued to appear through to November (a period of eight months), with the host population diminishing to approximately 0 .3 per litre volume . Salinity readings of the pool taken at fortnightly intervals showed that with the exception of two samples for April all the results were at 6.0°/00 NaCl or below (Table II) . The figures 19 .5 and 11 .0°/00 recorded during April were due to very rough coastal weather when the sea entered the pool raising the salinity level . With the return to normal conditions, the salinity dropped and the equilibrium was established .

430



TABLE I

1969 1970

ND 0 .3

Jan .

30(55%) 18(16%)

infected pools non-infected pools

17(31) 13(11)

5 .0-10 .0 4 (8) 30(27)

15 .0-20.0

TABLE III

1 (2) 27(24)

10 .0-15 .0 2 (4) 12(11)

20 .0-25 .0

0 (0) 11(10)

25 .0-30 .0

ND ND

Febr .

April 20 .8* 19 .0

March 1 .5* 6 .0*

23 .0 24.0

May 16 .0 27 .0

June

8 .0* 32 .0

July

6 .0 18 .0

August

19 .0 9 .0

Sept .

28 .0 3 .0*

Oct .

32 .0 8.0*

Nov .

0)0 0(0)

30 .0-35.0

Monthly 0/00 salinity readings for Pool B Purakaunui, 1969-1970, *infection recorded, ND = No data . .

0 .0-5 .00

0/00 NaCl

25 .0 ND

Dec .

0(0) 1(1)

35+

Comparisons of Salinity readings (1967-1970) from 21 infected and 64 non-infected pools grouped in 5.00/00 units of salinities . (Percentages in brackets) .

TABLE II

Fortnightly salinity readings for pool A Long Point . March-November 1970 . March

April

May

June

3 .0 0/ 00 4.8 0/00

19 .5 11 .0

6 .0 4.30

5 .0 4 .8

July August Sept . 3 .5 3 .0

2 .8 2 .0

2 .8 3 .0

Oct .

Nov.

3 .5 3 .0

3 .5 2 .3

A correlation between lower salinity and infected larvae is perhaps even more strikingly demonstrated in pool B which was characterised by fluctuating salinities . This is a smaller pool with an estimated volume of approximately 500 litres and located on the Purakaunui outcrop about a kilometer north of Long Point . Although it is situated some distance away from the splash zone it has a natural communicating channel which is often breached during high tides (fig . 4) . Since 1965 this pool has been known as a perma-

Fig . 4 . Pool B Purakaunui outcrop with permanent colony of Aedes australis, showing only intermittent C. opifexi infection . Arrow points to natural channel often breached by sea at high tides.

nent habitat for Aedes australis . Salinity readings were commenced in March 1969 when a search revealed the presence of infected larvae . Monthly figures for almost 2 years showed that salinity fluctuated from a low 1 .5°/0o to a maximum of 28 .00/°°, but infected larvae appeared during or following periods of low salinities (Table III) . 432

DISCUSSION & CONCLUSIONS

From field observations it appears that C. opifexi occurs mainly within the distribution range of Aedes australis, which is itself restricted to the south east coast of the South Island of New Zealand . Since the presence of A . australis in New Zealand was only recently reported (NYE, 1962) and is presumed to have been derived from Australian sources (BELKIN, 1968) the phytogeographic relationship of C. opifexi is open to question . On the basis of its present distribution pattern in New Zealand NYE (loc. cit .) suggested that australis may have gained entry through shipping at the southern port of Bulff near Invercargill . There appears to be some merit in this theory, since it has been observed that the mosquito is extending its range northwards along the coast of the South Island . It has been suggested (LAIRD & COUCH, pers . comm . 1969) that C. opifexi is conspecific with Coelomomyces tasmaniensis also described from the larva of Aedes australis (ERICHSON) in Tasmania (LAIRD, 1956) . Should this be the case, the possibility of a Coelomomyces identical (?) with tasmaniensis reaching New Zealand at the same time as australis cannot be ruled out . Having arrived here it would have had no difficulty coping with the endemic Opifex fuscus which occupies the same ecological niche in New Zealand . Although the biology of infection in the host larva has been studied previously (see review, COUCH & UMPHLETT, 1963) some aspects of the life cycle of the Coelomomyces remain unknown . Larvae in advanced state of infection are packed with large numbers of sporangia and later become moribund and drop to the bottom of the pool to die . When the sporangia are freed from the dead larva, they germinate releasing numerous motile zoospores which are thought to be the infective agents (COUCH, 1968) . In the supralittoral pool of Coelomomyces opifexi, the risk of the sporangial filled moribund larvae and/or the motile zoospores being swept away by sea action might appear to be very considerable . From the present observations it appears that its survival is probably due to its activity in some of the more sheltered pools of the rocky outcrops, from which it spreads to other pools when conditions are favourable . Except for salinity concentrations there was no evidence to indicate that any distinctions could be drawn between infected and noninfected supralittoral pools on the basis of temperature, hydrogen ion concentration, size, bottom and biotic compositions . It seems more than likely that once the Coelomomyces is introduced into the pool, infection of larvae is influenced by a lower range of salinity as 433

shown both by collective (Tab . I) and individual pool samples over extended period . In pool A at Long Point, from March to November a salinity level of 6 .00/00 NaCl or below was recorded for a 7 month period, while infected larvae appeared continuously . The correlation between low salinity and presence of infected larvae was even more striking in the case of the pool B on the Purakaunui outcrop, where infection was present only during or following periods of low salt concentration . This probably emphasises the requirement of low levels of salinities for infection to take place . Laboratory evidence supports these field observations . For experimental infection of Aedes australis by C. opifexi, a lower range of salinity (10 .0°/°0) appears necessary . Once the infection has occurred a rise in salinity does not have any adverse effect on the larva or fungus, though further infections are inhibited (PILLAT, unpublished MSS.) In the fluctuating supralittoral environment optimum conditions appear to be available in a few sheltered pools for prolonged periods, for Coelomomyces opifexi activity . These probably serve as the main source of infection which is passed on to other pools when salinity level is suitably low . The Coelomomyces is apparently unaffected by a temporary rise in salinity . However in a contaminated pool subjected to prolonged periods of high salinity, diseased larvae seem to appear only when the salt concentration drops to a suitable level for infection to take place .

SUMMARY

Coelomomyces opifexi is a pathogen of two species of brackish mosquitoes in New Zealand, the endemic Opifex fuscus and the introduced Australian species Aedes australis . Since the activity of the fungus is confined to the south east coast of the South Island, which also corresponds to the present distribution pattern of Aedes australis in New Zealand, it is suggested that the Coelomomyces may have also been derived from Australian sources . The host species breed in supralittoral pools of rapidly fluctuating salinities ranging from fresh water to hypersaline conditions . Temperature, hydrogen ion concentration, salinity, pool size, biotic composition factors were compared between infection and non-infection pools . With the exception of salinity readings there was no significant difference in the two . Field studies showed that there was a correlation between a lower range of salinities and the appearance of infected larvae in the pool . A few supralittoral pools appear to be located in sheltered situations

434

where the effect of sea action is minimal . These pools serve as a source of infection which spread to other pools when conditions are favourable . ZUSAMMENFASSUNG

In Neuseeland ist Coelomomyces opifexi pathogen fur zwei Arten brackiger Stechmucken, den endemischen Opifex .fuscus and die eingefiihrte Australische Art Aedes australis . Da die Tatigkeit des Pilzes, wie die heutige Verbreitung von Aedes australis nur auf die siidostliche Kiiste der Siidinsel beschrankt ist, wird vermutet, daB Coelomomyces auch australischer Herkunft ist . Die Wirtspecies erzeugen in Kiistenlachen schnellveranderliche Salzhaltigkeiten wechselnd von frisch bis zu ubersalzig . Faktore wie Temperatur, Ph, Salzhaltigkeit, LachegroBe and biotischer Aufbau in infizierten and nichtinfizierten Lachen Bind verglichen worden . Dabei gab es keine bedeutenden Unterschiede auBer dem Salzigkeitsgrad . Aus den dortigen Forschungen hat sich herausgestellt, daB es eine Wechselbeziehung zwischen einem niedrigen Salzhaltigkeitsbereich and der Erscheinung infizierter Larven gibt . Man hat verschiedene Lachen in beschiitzenden Lagen mit einem minimalen MeereseinfluB gefunden . Unter giinstige Verhaltnisse stellen sie eine Ansteckungsquelle dar . ACKNOWLEDGEMENTS

I wish to thank PROFESSOR J.A.R . MILES of the Department of Microbiology, University of Otago for his continuing interest in Coelomomyces investigations . This research was supported by a grant from the New Zealand Medical Research Council . REFERENCES BELKIN, J . N . - 1968 - Mosquito studies (Diptera : Culicidae) VII . The Culicidae-

of New Zealand . Amer . Entomol. Inst., contrib . 3(l) : 182 . COUCH, J . N . - 1968 - Sporangial germination of Coelomomyces punctatus and the

conditions favouring the infection of Anopheles quadrimaculatus under laboratory conditions . Proc . Joint U.S .-Japan Seminar on microbial control of insect pests . Panel 8. Research on Pesticides : 93-105 . LAIRD, M . - 1956 - A New species of Coelomomyces (fungi) from Tasmanian mosquito larvae - J . Parasitol. 42 : 53-55 . LuM, P . T. M . - 1963 - Infection of Aedes taeniorhynchus WIEDEMANN and Psorophora howardii COQUILLETT by the fungus Coelomomyces J. Insect Pathol . 5 : 157166 .

435

McGREGOR, D . D. - 1964 - Physical ecology of some New Zealand Supralittoral pools. Hydrobiologia. 25 : 277-284 . MCGREGOR, D . D . - 1965 - Aspects of ecology of Opifex fuscus HUTTON (Diptera : Culicidae) . Proc . R. ent . Soc. Lond. (A) 40 : 9-14. MUSPRATT, J . - 1963 - Destruction of the larvae of Anopheles gambiae GILES by a Coelomomyces fungus . Bull. Wld Hlth Org. 29 : 81-86 . NYE, E . R . - 1962 - Aedes (Pseudoskusea) australis ERICHSON (Diptera : Culicidae) in New Zealand . Trans. Roy. Soc . New Zealand, Zoology. 3 : 33-34 . PILLAI, J . S . & SMITH, J . M . B . - 1968 - Fungal Pathogens of mosquitoes in New Zealand . I . Coelomomyces opifexi sp . n . on the mosquito Opifex fuscus HUTTON . J . Invert . Pathol . 11 : 316-320. PILLAI, J . S . - 1969 - A Coelomomyces infection of Aedes australis in New Zealand . J . Invert . Pathol. 14 : 93-95 . UMPHLETT, C . J . - 1968 - Ecology of Coelomomyces infection of mosquito larvae . J . Elisha Mitchell Sci. Soc. 84 : 108-114.

436