PATHOBIOLOGY OF BURENELLA DIMORPHA JOUVENAZ
AND HAZARD (MICROSPORA: MICROSPORIDA)
DONALD P. JOUVENAZ
A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
00 Many people have assisted, advised, and encouraged me
Sin my graduate studies. I am deeply grateful to the members
of my graduate committee, Drs. Donald W. Hall, Clifford S.
Lofgren, Edward M. Hoffman, and Robert C. Wilkinson, for
their patience and guidance, and for the time and effort
they have spent on my behalf. I also wish to thank Dr.
George E. Allen, who served as chairman of the graduate
committee before his departure from the University for
In addition to the members of my graduate committee,
several people have given me invaluable assistance. Fore-
most among these is Ms. E. Ann Ellis, Electron Microscopist
at the Insects Affecting Man and Animals Research Laboratory,
USDA, Gainesville. Mr. John D. Atwood, Mrs. Anita Lemire,
Mrs. Susan W. Avery, and Dr. Albert Undeen have also helped
me in various ways. Dr. William F. Buren of the University
of Florida encouraged me over a long period of time.
I cannot adequately express my gratitude to my wife,
Judy, for her unfailing encouragement and patience throughout
my graduate studies.
TABLE OF CONTENTS
LIST OF TABLES.................................... V
LIST OF FIGURES.................................... i
ONE INTRODUCTION............................ 1
TWO GENERAL LITERATURE REVIEW............. 4
Synopsis of the Microsporidia......... 4
Microsporidia Parasitic in Ants....... 6
Solenopsis geminata................... 9
THREE GENERAL MATERIALS AND METHODS......... 11
Collection and Maintenance of Colonies 11
Temperature and Humidity.............. 15
Laboratory Propagation of B. dimorpha 16
Harvest of Spores...................... 16
FOUR TRANSMISSION AND INFECTIVITY OF SPORES
OF BURENELLA DIMORPHA................. 18
Materials and Methods ................ 19
FIVE HOST SPECIFICITY OF BURENELLA DIMORPHA 27
Materials and Methods ................. 29
SIX ABUNDANCE OF BURENELLA DIMORPHA IN
Materials and Methods.................. 37
Results and Discussion................ 37
SEVEN PATHOLOGY OF BURENELLA DIMORPHA
Materials and Methods ................
EIGHT TEMPERATURE-DEPENDENT SPORE DIMORPHISM
IN BURENELLA DIMORPHA................
Materials and Methods.................
NINE AMENDMENTS TO THE DESCRIPTION OF
Materials and Methods.................
TEN SUMMARY AND CONCLUSIONS...............
LIST OF TABLES
1. B. dimorpha infection rates in colonies
of S. invicta and S. richteri. 31
2. Susceptibility of selected species of ants
and a moth to infection by B. dimorpha. 33
3. Relative abundance of B. dimorpha MB spores
in pupae of S. geminata reared from larvae
at various temperatures. 69
4. B. dimorpha spore yield of pupae of
S. geminata reared from larvae at 20,
28, or 32 C. 69
5. Relative abundance of B. dimorpha MB spores
in pupae of S. geminata placed at 20 or 32 C
at different ages. 71
LIST OF FIGURES
1. A six-cell colony of S. geminata. 14
2. A miniature nest cell. 14
3. Infrabuccal cavity. 24
4. Enlarged view of infrabuccal cavity and
5. Diseased S. richteri pupa compared to
healthy and diseased S. geminata pupae. 36
6. Diseased (left) and healthy male pupae
of S. geminata. 44
7. Diseased (left) and healthy female pupa
of S. geminata. 44
8. Healthy (left) and diseased worker pupae
of S. geminata. 44
9. The development of pathognomonic signs
in pupae of S. geminata infected by
B. dimorpha .47
10. Frontal section through the head of a
young (ca 10 days postpupation) pupa
infected by B. dimorpha. 51
11. Eye, optic nerve, and optic stalk of a
young (ca 10 days postpupation) pupa
infected by B. dimorpha. 51
12. Frontal section through the head of an
older (ca 16 days postpupation) pupa
infected by B. dimorpha. 51
13. Cuticle of a healthy pupa. 54
14. Body surface of a pupa infected by
B. dimorpha. 54
15. Detail of the cuticle in a healthy
16. Detail of microvilli extending into
the fluid space of a pupa infected by
B. dimorpha. 54
17. Sagittal section through the developing
eye of a healthy pupa. 56
18. Sagittal section through the eye of a
pupa infected by B. dimorpha. 56
19. Sagittal section through the eye of a
healthy pupa. 56
20. Sagittal section through the eye of a
diseased pupa. 56
21. Sagittal section through the developing
eye of a healthy pupa showing detail of
optic nerve and base of the developing
22. Sagittal section through the eye of a
pupa infected by B. dimorpha. 58
23. Fat body of a healthy pupa. 58
24. Fat body of a pupa infected by B. dimorpha. 58
25. Tissue specificities of MB and NMB spores
of B. dimorpha. 60
26. Tissue specificities of MB and NMB spores
of B. dimorpha. 60
27. Tissue specificities of MB and NMB spores
of B. dimorpha seen at higher magnification. 60
28. Transmission electronmicrograph of a
mature NMB spore of B. dimorpha. 80
29. Transmission electronmicrograph of a
mature MB spore of B. dimorpha. 80
30. Scanning electronmicrograph of a mature
NMB spore of B. dimorpha. 80
31. Scanning electronmicrograph of a
mature MB spore of B. dimorpha. 80
32. Transmission electronmicrograph of an
immature NMB spore. 83
33. Transmission electronmicrograph of two
immature MB spores within a delicate
pansporoblast membrane. 83
Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
PATHOBIOLOGY OF BURENELLA DIMORPHA JOUVENAZ
AND HAZARD (MICROSPORA: MICROSPORIDA)
Donald P. Jouvenaz
Chairman: Donald W. Hall
Major Department: Entomology and Nematology
Burenella dimorpha Jouvenaz and Hazard is a host-specific,
dimorphic microsporidium that parasitizes the tropical fire
ant, Solenopsis geminata (F.). The red and black imported fire
ants, Solenopsis invicta Buren and Solenopsis richteri Forel,
can be infected perorally by B. dimorpha, but the parasite
survives only a few generations in colonies of these factitious
hosts. Nine species of ants other than fire ants (representing
eight genera) were refractory to infection.
Infected pupae develop pathognomonic signs (eye teratology
and blister-like clearings in the occiput and petiole) that are
due to destruction of the cuticle. In an advanced stage of
disease, the pupa ruptures and is cannibalized by worker ants.
The spores and other particulate matter are not ingested, but
are diverted to the infrabuccal cavity, formed into a pellet,
expelled, and fed to fourth-instar larvae only. The intra-
colonial cycle of infection is thus from ruptured, diseased
pupae to fourth-instar larvae via the adults, who are mechan-
ical vectors. The binucleate, nonpansporoblast membrane-
bounded (NMB) spores are infective perorally for larvae; the
uninucleate, pansporoblast membrane-bounded (MB) spores are
not infective, and their function is unknown. These spore
types are tissue specific: NMB spores develop from disporous
sporonts in the hypodermis; MB spores develop in octets from
multinucleate sporonts in the fat body.
The development of MB spores is temperature-dependent.
The lower thermal threshold for MB spore development is
between 20 and 22.5 C; the upper thermal threshold is below
32 C. A hypothesis is advanced that these limits are set by
the stability of an enzyme(s), and that the physiological
function blocked in MB sporulation is meiosis.
Burenella dimorpha has two cycles of merogony (vegetative
multiplication). The meronts of the first cycle were originally
described as uninucleate cells that become binucleate and divide.
However, these nuclei are actually pairs of nuclei in a diplo-
Surface structure has been seen for the first time in MB
spores. The surface of NMB spores is smooth.
The red and black imported fire ants, Solenopsis
invicta Buren and Solenopsis richteri Forel, are medical
and agricultural pests which infest ca 9.3 x 107 hectares
(2.3 x 10 acres) in the Southeastern United States. Both
species were apparently introduced into the United States
with products shipped from South America to Mobile, Alabama,
about 1918 and 1940, respectively. The tropical fire ant,
Solenopsis geminata (Fabricius), may also be an introduced
species; however, it has been a resident of this country
for so long it is generally regarded as native. S. geminata
is not important as a pest, except in Hawaii where it has
Efforts to control the imported fire ants by chemical
means have been the subject of serious controversy since
before 1960. Consequently, the discovery of microsporidian
infections in these ants in their native lands by Drs. G. E.
Allen, W. F. Buren, and A. Silviera-Guido greatly stimulated
interest in research on their possible use for biological
control. Earlier, surveys by several investigators had
failed to detect specific pathogens of fire ants in the
In response to the reports of Allen, Buren, and Silviera-
Guido, the Agricultural Research Service of the United States
Department of Agriculture funded several trips to South
America by federal and university scientists to search for
potential biological control agents for fire ants. I was
priviledged to participate in the first trip to Mato Grosso
and Mato Grosso do Sul, the homeland of S. invicta, as a
member of a team of five investigators. For five weeks
we collected fire ants and laboriously examined them indi-
vidually for evidence of infection. Despite the handicap
of not knowing how to efficiently screen ants for disease,
we collected spores of two species of microsporidia. Unfor-
tunately, all of our attempts to infect healthy laboratory
colonies of S. invicta with these spores failed.
Since we were unable to propagate and study the exotic
microsporidia, a survey of pathogens of S. geminata in Florida
was initiated to find a locally available host-pathogen
model for studies of basic pathobiology, and to develop
techniques for the detection and study of disease in ants.
The pathogen selected for study in detail was Burenella
dimorpha Jouvenaz and Hazard, a microsporidium that is trans-
missible per os and which has the advantage of producing
pathognomonic signs of infection in pupae of S. geminata.
I have attempted to trace the history of this microsporidiosis
form the transmission of spores to a healthy host, through
the life cycle of the parasite and pathogenesis in the host,
to the production of spores of the succeeding generation.
An understanding of the biology of the natural enemies
of insects is fundamental to the development of strategies
for their employment in pest management. The information
gained in this study will contribute to our knowledge of
the microsporidia in general, and, hopefully, in some small
way to the development of effective biological control of
the imported fire ants.
GENERAL LITERATURE REVIEW
As a general reference to the microsporidia, the reader
is referred to the two volume monograph by Bulla and Cheng
(1976, 1977). These two volumes constitute the only recent
monograph on these protozoa. Earlier monographs on the
microsporidia were published by Kudo (1924) and Weiser (1961);
however, the development and recent extensive application
of electron microscopy to the study of microsporidia has
rendered these works largely obsolete. Since microsporidia
are little known to most entomologists and even protozoolo-
gists, a brief synopsis of their systematics and biology is
presented. A glossary of terms specifically used in the
study of microsporidia and abbreviations used in this disser-
tation has been included.
Synopsis of the Microsporidia
Microsporidia are extremely small protozoa which are
obligate (lacking mitochondria), intracellular parasites of
invertebrate (primarily arthropod) and, less commonly, verte-
brate animals. Their life cycles include the production of
spores which contain a single sporoplasm and a long, coiled,
tubular organelle, the polar filament (or a rudiment thereof).
Upon ingestion by a suitable host, the polar filament uncoils
and extrudes with extreme rapidity from the spore, remaining
attached anteriorly. The sporoplasm is expelled through
the polar filament, which apparently evaginates as it ex-
trudes from the spore. If the spore is near the gut wall
and is properly oriented, the wall is penetrated by the
violently extruding filament and the sporoplasm is injected
into a host cell as though by a hypodermic syringe. There
it multiplies and develops in direct contact with the host
cytoplasm, there being no parasitophorous vesicle. After a
large number of vegetative parasites have been produced by
cycles of multiplicative development (merogony), they trans-
form into sporonts, or cells which give rise to spores after
a set number of divisions characteristic of the species
Sprague (1977), considering the protozoa a polyphyletic
group, assigned the microsporidia to a new phylum, Microspora,
consisting of two classes, Rudimicrosporea and Microsporea.
The typical microsporidia, including the subject of this
dissertation, are assigned to the order Microsporida, one
of the two orders of Microsporea (Levine et al., 1980).
Microsporidia Parasitic in Ants
Only two species of microsporidia, Thelohania solenopsae
Knell et al. (Thelohaniidae), and the subject of this disserta-
tion, Burenella dimorpha Jouvenaz and Hazard (Burenellidae),
have been described from ants.
T. solenopsae, the first specific pathogen known from
fire ants and the first protozoan known from Formicidae,
was discovered by Dr. W. F. Buren during a taxonomic study
of S. invicta (Allen and Buren, 1974). While examining alcohol-
preserved specimens from the state of Mato Grosso, Brazil,
Buren observed subspherical, cyst-like bodies in the partially
cleared gasters of workers. These cysts contained spores
of the microsporidium, which was subsequently described from
fresh material by Knell et al. (1977). Very soon after Buren's
observation, Allen and Silviera-Guido (1974) reported similar
microsporidia from S. richteri in Uruguay and Argentina,
and from an unidentified Solenopsis species in Uruguay.
T. solenopsae (or T. solenopsae and sibling species which
cannot be differentiated) has since been detected in ca 22
described and undescribed species of fire ants in South
America (Jouvenaz et al., 1977).
T. solenopsae infects fat body cells of workers and
sexual, and the ovaries of females. Infected cells hyper-
trophy, forming the cysts observed by Buren. Within the
cysts, the spores occur in octets bounded by a membrane
(pansporoblast membrane). The infection is not rapidly
fatal, but destruction of the fat body occurs which results
in premature death of adult ants. Consequently, colonies
are debilitated (Knell et al., 1977). Attempts to transmit
T. sdlenopsae to healthy colonies in the laboratory have
failed, and the mode of transmission of this parasite is
unknown (Jouvenaz et al., 1981).
Burenella dimorpha was described by Jouvenaz and Hazard
(1978) as the type species of a new genus which represents
a new family, Burenellidae. This family includes those
species of microsporidia having two sporogonic sequences,
one producing non-pansporoblast membrane-bounded (NMB)
spores from disporous sporonts, and the other producing
octets of spores bounded by a pansporoblast membrane (MB).
Hazard et al. (1981) listed three additional genera as
members of Burenellidae: Culicosporella Weiser and Hazardia
Weiser (monotypic genera parasitic in mosquitoes) and
Vairimorpha Pilley (parasitic in a variety of Lepidoptera).
Certain genera of the family Thelohaniidae Hazard and Oldacre
also have dimorphic (producing two morphologically distinct
types of spores) species; however, their NMB spores arise
from plasmodia rather than disporous sporonts.
Little is known of the pathobiology of B. dimorpha beyond
those minimal aspects of spore morphology and life cycle
necessary for classification. This information will be
reviewed in conjunction with the specific studies reported
in this dissertation.
In addition to B. dimorpha, at least three undescribed
species of microsporidia infect S. geminata (Jouvenaz et
al., 1977), and at least one undescribed species infects
Solenopsis spp. in South America (Jouvenaz et al., 1980).
Other Pathogens of Ants
Because of the paucity of information on the diseases
of ants, it is feasible to summarize briefly the literature
on pathogens of ants other than microsporidia. The only
remaining pathogen of ants which has been described is
Mattesia geminata Jouvenaz and Anthony (Neogregarinida:
Ophrocystidae), which also infects S. geminata in Florida.
This protozoan develops in the oenocytes of the hypodermis,
causing destruction of the hypodermis, melanization, and
eye malformation in pupae. The infection appears to be
invariably fatal in the pupal stage of development. Attempts
to transmit the infection per os (using fresh, aged, and
variously treated spores) and by placing infected pupae
in healthy colonies (conspecific pupae are adopted) have
failed (Jouvenaz and Anthony, 1979). A similar or identical
neogregarine occurs in fire ants in Brazil (Jouvenaz et
Virus-like particles have been detected in an undescribed
Solenopsis species from Brazil and in S. geminata from
Florida (Avery et al., 1977). These particles are morpho-
logically similar, being non-occluded, rod-shaped, and
bound by double membranes. They occur in hypertrophied
nuclei of fat body cells, but their pathogenicity is as
yet undetermined. Virus-like particles have also been
reported from apparently healthy wood ants, Formica lugubris
Zett. (Steiger et al., 1969). These particles occurred
in the cytoplasm of nerve cells, and were morphologically
quite different from those found in fire ants.
Only one species of fungus, which remains unidentified,
is known to be specifically associated with fire ants (Jouvenaz
et al., 1977). Other fungi which have been reported as
causing infections in various species of ants include
Metarrhizium anisopliae (Metschnikoff) Sorokin, Beauveria
bassiana (Balsamo) Vuillemin, and several species of Cordyceps.
Allen and Buren (1974) summarized and discussed these reports.
Broome et al. (1976) discussed the mechanism by which B.
bassiana infection is initiated in S. richteri.
The literature on the tropical fire ant is very sparse;
most studies of fire ants have been concerned with the
imported species, which are pests. For information on
the biology and behavior of S.geminata and fire ants in general,
the reader is referred to the excellent review by Lofgren et al.,
An outline of the life cycle of Solenopsis species
is as follows: mating takes place during nuptial flights,
after which the females (queens) return to earth, dealate,
and secrete themselves in a closed, subterrainian chamber.
During this claustrall" period, the queens, subsisting
on their histolyzing flight muscles and food stored in
their fat body and crop, lay eggs and rear their first
offspring. As the number of workers increases, the queen
stops caring for the immatures, but continues to lay eggs
throughout her lifespan. The workers forage for food,
extend and maintain the nest structure, defend the nest
and territory, and care for the queen and brood. A mature
colony may contain as many as 200,000 or more individuals.
The developmental stages of fire ants are egg, four
larval instars, pupa, and adult. The development period
is temperature dependent, but averages about three to four
GENERAL MATERIALS AND METHODS
Collection and Maintenance of Colonies
Laboratory colonies of fire ants were obtained by
collecting queens and contingents of their workers from es-
tablished field colonies, or (during the spring of the year)
by capturing newly mated queens from under debris or as
they wandered over the surface of the ground. Queens of
other species of ants (used for host range studies) were
collected exclusively by the latter method.
Ants were collected from field colonies by excavating
the mounds with shovels and transporting the soil containing
ants to the laboratory in plastic buckets, the inner walls
of which were coated with Fluon GP-1 (ICI United States,
Wilmington, Delaware 19897) to prevent escape of the ants.
The soil in the buckets was slightly moistened (if necessary)
and left undisturbed overnight to allow the ants to establish
tunnels and to collect buried immatures. Water was then
slowly dripped from medical intravenous fluids tubes into
the buckets, forcing the ants to the surface of the soil.
When the soil was completely submerged, masses of adult
and immature ants floated or clung to the sides of the bucket,
and were easily transferred with a ladle to Fluon-coated
trays (ca 52 x 40 x 7 cm) containing several layers of paper
towels on their bottoms. After the wet masses of ants had
dried and dispersed, the collections of ants were searched
for queens. Those containing queens were transferred to
laboratory nests in Fluon-coated trays held in metal racks
Newly mated queens were placed in miniature nest cells
or in glass culture tubes (150 x 17 mm) containing a mass
of wet cotton and held in a Fluon-coated tray.
Nest cells for large colonies were constructed by pouring
liquid Castone (Ransom and Randolph Co., Toledo, Ohio) to
a depth of 10-12 mm in 25 x 150 mm disposable plastic petri
dishes, allowing the Castone to solidify, and melting four
exit ports in the sides of the bottom dish above the surface
of the Castone (ports were unnecessary in the loose-fitting
tops). Solidified Castone is too hard for the ants to tunnel
through, yet remains slightly moist, providing adequate
(but not excessive) humidity and substrate moisture within
the cells for two to three months.
When the nest cells became dry and soiled, their tops
were removed, and fresh cells were placed in the tray. The
ants quickly moved into the new cells, and the abandoned
old cells were discarded. Up to four cells were used simul-
taneously in each tray, and were replaced on a rotating
Miniature nests for newly mated queens, very small
colonies, or groups of experimental subjects held in isolation
Fig 1. A six-cell colony of S. geminata. Two glass culture
tubes plugged with cotton and containing water or
dilute honey, a dish of insect pupae, and part of a
boiled egg are on the left. Alate sexual are con-
centrated in the reat top cell.
Fig 2. A miniature nest cell. The nest cell has a layer
of Castone (dental casting compound) in the bottom
and an entry port in the side. A wad of wet cotton
helps maintain high humidity. The wall of the large
petri dish is coated with Fluon to prevent escape
of the ants. The large petri dish top is normally
taped securely to the bottom and labeled.
" 9 '
**-* ''*, ., S
from their parent colonies were constructed from 15 x 60 mm
and 25 x 150 mm disposable plastic petri dishes. Liquid
Castone was poured to a depth of ca 5 mm in the smaller
dish, allowed to solidify and two or three exit ports were
melted into the wall. This small nest, with lid in place,
was placed inside the larger dish, the inner walls of which
were coated with Fluon. A wad of wet cotton was also placed
in the larger dish to help maintain high humidity (the much
smaller mass of Castone in these cells dried within a few
days) and the lid was taped to the petri dish bottom to
provide additional security. Thus, a nest and a secure
foraging area were combined in a single, compact unit (Fig 2).
Nest cells of both sizes were prepared in advance and
stored in a refrigerator until needed. Prior to use, they
were soaked in distilled water to saturation, then blotted
with paper towels to remove excess water.
Temperature and Humidity
The colonies were maintained at a ca 280 C in a labora-
tory room having no humidity control; however, the Castone
nest substrate and wet cotton balls (described above) main-
tained humidity within tolerable limits.
Laboratory colonies were fed live or dead (frozen)
insectary-reared insects (cockroaches, mealworms, cabbage
looper pupae), supplemented with boiled eggs, fruit, and 50%
aqueous honey solution. A diet composed of housefly or stable-
fly pupae (1 liter) pureed with cooked ground beef (ca 250 g)
in a blender and thoroughly mixed with whole eggs (1 dozen)
and vitamins (5.0 ml Poly Vi-Sol, Mead Johnson & Co., Evansville,
Indiana 47721) in hot liquid agar solution (50 g agar dissolved
in 1.5 L distilled water) was also used as a supplement
when insects were in short supply. The warm liquid diet
was poured to a depth of ca 1.5 cm in shallow enamel pans
and allowed to gel. It was then cut into cubes for dispens-
ing to the ants. The excess diet was covered with plastic
film and frozen or stored in the refrigerator.
Laboratory Propagation of Burenella dimorpha
Colonies of ants were infected with B. dimorpha by
feeding them boiled egg yolk wetted with a suspension of
spores to a consistency of paste. These colonies were then
maintained and diseased pupae were harvested periodically
as spores were needed. Prior to infection, colonies were
carefully examined for other diseases using the methods
described by Jouvenaz et al. (1977).
Harvest of Spores
Suspensions of spores free of cellular debris were
obtained by density gradient centrifugation using Percoll
(Pharmacia Fine Chemicals, 800 Centennial Ave., Piscataway,
N.J. 08854). Pupae exhibiting signs of advanced infection
were homogenized in an equal volume of distilled water using
a glass tissue grinder, and the homogenate was centrifuged
at 10,000 g for ca 20 minutes through a discontinuous gradient
consisting of layers of 25, 50, 75, and 100% Percoll. Over
97% of B. dimorpha NMB spores purified in this manner ex-
truded their polar filaments when ingested by ant larvae
(determined by phase microscopy of meconia), indicating
that most immature spores had been removed. Slightly less
highly purified suspensions suitable for use in routine
propagation of the parasite were obtained by centrifuging
the crude homogenate at 750 g for 20 minutes in a 50% solution
of Percoll. After all centrifugations with Percoll, the
spores were washed three times with distilled water, and
the volume was adjusted to ca 107 spores/ml. Spore suspen-
sions were stored at 40 C.
TRANSMISSION AND INFECTIVITY OF SPORES OF
Burenella dimorpha produces two morphologically distinct
types of spores. Binucleate NMB spores develop from disporous
sporonts in the hypodermis; uninucleate, MB spores develop
in octets from multinucleate sporonts in the fat body. The
infection is characterized by the development of clear,
blister-like areas in the occiput and petiole which are
due to destruction of the cuticle. As the infection pro-
gresses, the clear areas increase in size, the cuticle becomes
extremely fragile, and eventually it ruptures. The pupa
is then cannibalized by workers (I have observed this in
Suspensions of spores of B. dimorpha containing both
types of spores are infective perorally for S. geminata
(Jouvenaz and Hazard, 1978). However, attempts to separate
the spore types by density gradient centrifugation were
unsuccessful (the pansporoblast membrane is very delicate,
rupturing on dissection of the host, and the free MB spores
are nearly identical in bouyant density to NMB spores).
Therefore, Jouvenaz and Hazard were unable to determine
which or if both types are infective. They described this
parasite as a dimorphic species on the basis of light micro-
scopy studies of the life cycle, and the statistical improb-
ability of a dual infection occurring at high frequency,
but they were not able to experimentally demonstrate by
feeding tests that both spore types were produced by the
Adult fire ant workers feed liquid food to larvae in
all instars, but solid foods to fourth-instar larvae only
(Petralia and Vinson, 1978). A pellet of solid food is
expelled from the infrabuccal cavity and placed on the prae-
saepium of a larva by a worker, and the pellet is consumed
by the pupa. This method of feeding larvae suggested a
cycle of infection from ruptured pupae to fourth-instar
larvae vectored mechanically by adults via the infrabuccal
pellet. This study was conducted to 1) determine the mode
of intracolonial transmission of infection, 2) to determine
whether one or both spore types are infective, and 3) to
verify that B. dimorpha is a dimorphic species.
Materials and Methods
Boiled egg yolk wetted to a paste consistency with
a suspension of spores was offered to a small, healthy colony
of S. geminata, which was allowed to feed for 24 hr. A
sample of workers trapped immediately after feeding was
held one hour, killed by freezing, sectioned into the three
body parts, and examined by phase-contrast microscopy to
determine whether spores were ingested or diverted to the
infrabuccal cavity microsporidiann spores are visible in
slightly compressed, intact fire ant body segments at a
magnification of 300X). Heads of these workers were fixed,
stained with heavy metals, embedded in Spurr-Quetol resin,
sectioned, and examined by phase-contrast microscopy. The
details of these procedures are given in the section on
host pathology. Eight fresh infrabuccal pellets were removed
from the praesaepia of larvae and examined by phase micro-
scopy for spores.
After 24 hours, 109 fourth-instar larvae were removed
from the nest and held in a miniature nest cell with conspe-
cific workers that had not been exposed to B. dimorpha.
These workers functioned as nurses (immatures held in isola-
tion from adults are quickly attacked by fungi). After
21 days these immatures (now pupae) were examined for infec-
tion. A group of 79 prefourth (primarily third) instar
larvae were also removed from the same nest and held in
a similar manner until they closed as adults.
A suspension of mature spores of both types was prepared
by homogenizing diseased pupae in distilled water with a
glass tissue grinder and centrifuging the extract in a dis-
continuous Percoll gradient (100%, 75%, 50%, and 25%) for
20 minutes at ca 10,000 g (Jouvenaz, 1981). This procedure
produced a clean suspension of spores of both types, almost
all of which appeared to be mature. These spores were fed
to fourth-instar larvae in a paste of finely powdered (mortar
and pestle) dry baby cereal and spore suspension. Infrabuccal
pellet-sized quantities of this preparation were placed
on the praesaepia of fourth-instar larvae with a flattened
insect pin, and the larvae were held in a miniature nest
cell until pupation. Adult workers were provided to care
for these larvae ca 4 hours after feeding. When the larvae
pupated, 20 meconia were recovered and examined by phase-
contrast microscopy to determine whether spores had extruded
their polar filaments.
A suspension of NMB spores only was obtained by selecting
diseased pupae which, on the basis of pathognomonic signs,
were estimated to harbor some mature NMB spores but no mature
MB spores (NMB spore development precedes MB spore development).
Wet squashes of these pupae were examined individually by
phase-contrast microscopy, and those that were free of mature
MB spores were washed from the slide, pooled, and cleaned
and concentrated by centrifugation. An examination of 10,000
spores individually in a diluted aliquot (0-5 spores/field)
and careful scanning of the concentrated suspension confirmed
the absence of MB spores.
The suspension of NMB spores was mixed with boiled
egg yolk and fed to a small healthy colony of S. geminata.
After 20 days, 25 pupae in advanced infection were individually
homogenized in ca 0.5 ml water and examined by phase-contrast
microscopy (0-6 spores/field) to determine the ratio of the
spore types in the first 200 spores seen. Twenty-five infected
pupae from a laboratory colony that had been infected with a
crude suspension of mixed spores for routine propagation
of the parasite were also examined.
Solenopsis geminata adults did not ingest spores into
the crop; instead, they diverted them to the infrabuccal
cavity (Figs 3 and 4). One hour after exposure to food
containing spores, 100 of the 160 workers that were examined
contained spores in their infrabuccal cavities. Not one
contained even a single spore in the digestive tract. All
eight infrabuccal pellets removed from fourth-instar larvae
shortly after deposition by workers contained numerous spores.
Only fourth-instar larvae became infected with B. dimorpha.
Seventy-one of the 109 that were exposed as fourth-instar
larvae became infected; none of the 79 immatures that were
exposed to spores as prefourth-instar larvae became infected.
Nonmembrane bound spores were infective perorally for
S. geminata and produced infections with pathognomonic
signs and both spore types. Furthermore, the spore types
were produced in normal ratios. Pupae that had been fed
NMB spores only contained 28.6% 10.2 (range 16-48%) MB
spores; those that had been fed both spore types contained
30.8% 7.8 (range 16-43%) MB spores.
In the meconia of larvae fed the suspension of mature
spores of both types, 1,723 (97.1%) of the 1,774 spores found
Fig 3. Infrabuccal cavity. This medial section of an adult
fire ant worker head shows an infrabuccal pellet in
the infrabuccal cavity. P = pharynx; IBC = infra-
buccal cavity; IBP = infrabuccal pellet.
Fig 4. Enlarged view of infrabuccal cavity and pellet.
Numerous MB and NMB spores may be seen in the pellet.
Arrows point to valves of the orifice of the infra-
F, A 4 r f
r j .y ,
had extruded their polar filaments. The 51 (2.9%) NMB spores
that had not extruded their polar filaments were probably not
quite mature, as they were very slightly less refractile
or less deep amber internally than mature spores.
Membrane bound spores did not appear to be infective
per os for S. geminata. Examination of meconia showed that
few if any MB spores had extruded their polar filaments
while in the larval gut. A total of 274 mature, nonextrud-
ed MB spores were seen in the 20 meconia, but only one body
was seen which strongly appeared to be an extruded MB spore.
Five additional bodies were seen that resembled extruded
MB spores but almost certainly were not. The meconia con-
tained occasional empty walls of ingested or gut unicellular
fungi, some of which superficially resembeled extruded MB
spores; extruded NMB spores were distinct.
The cycle of B. dimorpha infection within an ant
colony may be summarized as follows: NMB spores develop
in the hypodermis, producing clear areas in the heads,
petioles, and gasters of pupae. As the infection
progresses, the cuticle becomes very fragile and eventually
ruptures. The adult ants cannibalize these ruptured pupae
but do not ingest the spores. Instead, the spores, together
with other particulate matter, are diverted to the infra-
buccal cavity and formed into an infrabuccal pellet. This
pellet is expelled and placed on a specialized anteroventral
area, the praesaepium, of fourth-instar larvae. The prae-
saepium, which bears spines specialized for holding solid
food while the larva feeds, is absent from earlier instars,
which are fed only liquid. Because of this method of feeding,
the fourth-instar larva is the only stage which is vulnerable
to infection. Both spore types are ingested, but only the
NMB spore is infective. The MB spores are expelled unextrud-
ed in the meconium upon pupation.
It is evident that B. dimorpha is indeed a dimorphic
microsporidium, since ingestion of NMB spores only resulted
in typical infection in which both spore types were produced
in normal ratios.
The function of the MB spore remains unknown. A most
attractive hypothesis is that it either infects an alternate
host or is primed in the gut of a mechanical vector for
extrusion upon subsequent ingestion by ant larvae. Either
would explain the mode of intercolonial transmission of
the infection (fire ants are territorial and aggressive
towards conspecific ants). Many candidate species exist
for the role of vector; a large and varied arthropod fauna
is associated with fire ants. Collins and Markin (1971)
listed 52 species of insects that have been collected from
fire ant nests; other invertebrates also occur. At least
some of these organisms have symbiotic relationships with fire
ants and are known to travel between fire ant nests (Wojcik,
HOST SPECIFICITY OF BURENELLA DIMORPHA
Until recently, microsporidia were generally assumed
to be highly host specific, and the discovery of an infection
in a new host frequently led to the description of a new
species. Now microsporidia are known to vary greatly in
host specificity. Some species, particularly the parasites
of lepidoptera, are able to infect numerous hosts in different
genera, families, and even orders. Other species appear
to be highly host specific; however, in some cases the diffi-
culty of transmitting infection in the laboratory makes
it impossible to determine host range. The host ranges
of only a few species have been extensively investigated
The successful per os transmission of B. dimorpha infec-
tion to healthy colonies of S. invicta, S. richteri, and
the southern fire ant, Solenopsis xyloni McCook, as well
as the natural host, S. geminata, was reported by Jouvenaz
and Hazard (1978). High rates of infection were obtained
in colonies of these species by feeding them spores mixed
with boiled egg yolk. The infected pupae exhibited typical
pathognomonic signs, and Jouvenaz and Hazard concluded that
"the course of the disease is identical in all four species."
After these initial transmission tests had been published,
I attempted to propagate the parasite S. invicta. This
species is not only the pest we ultimately hope to control,
but I have found it easier to obtain and culture than S.
geminata. Several colonies were fed spores and, as expected,
successfully infected; however, after two or three months,
these colonies were free of the parasite. It appeared then,
that B. dimorpha might not be able to maintain itself in
factitious fire ant hosts, even though these hosts were
readily infected. The studies reported here were conducted
to determine the degree of host specificity of B. dimorpha.
In addition to fire ants, several species of ants of other
genera were tested for susceptibility to infection by B.
The corn earworm, Heliothis zea (Boddie) (Lepidoptera:
Noctuiidae) is susceptible to infection by Nosema algerae Vavra
and Undeen, a microsporidium which is highly pathogenic
to anophiline mosquitoes. Heliothis zea was being used
for mass production of N. alterae spores at the Insects
Affecting Man and Animals Research Laboratory, USDA, Gaines-
ville, during part of the period in which I conducted my
research. I therefore took advantage of the availability
of this moth and tested it for susceptibility to infection
by B. dimorpha.
Materials and Methods
Three small (ca 500 workers) colonies of S. invicta and
one larger colony (ca 10,000 workers) of S. richteri (the
only colony of this species that was available) were infected
by allowing them to consume diseased pupae of S. geminata.
The colonies of S. invicta were fed 15 selected pupae in
the advanced stages of infection on each of the following
days: Colony #1, days 1 and 3; colony #2, days 1, 2, and
3; colony #3, days 1, 3, and 5. The colony of S. richteri was
fed 50 selected pupae on one occasion.
The colonies were examined 3 weeks after the final
feeding to determine the initial rate of infection, and
at monthly intervals thereafter. Pupae old enough (judged
by eye development) to exhibit pathognomonic signs if infec-
ted were examined individually against a black background
with a stereomicroscope illuminated by an annular fiber
optic. All, or a maximum of 1,000, of these pupae were
examined from each colony.
Ants of other species were gifts from colleagues or
were collected after mating flights and reared as described
in General Materials and Methods. Identifications of species
were made or confirmed by Dr. Daniel P. Wojcik, Agricultural
Research Service, USDA, Gainesville, or by Mr. James Trager,
Department of Entomology and and Nematology, University
of Florida. These ants were also fed selected diseased
pupae of S. geminata (ingestion of the pupae was confirmed
by observation), and spores mixed with their regular diet.
In addition to examining pupae for pathognomonic signs,
aqueous extracts of pupae were examined by phase-contrast
microscopy for spores in case external signs were not evident.
Larvae of H. zea and their rearing media were obtained
from Mr. D. W. Anthony, Agricultural Research Service, USDA,
Gainesville. The protocols for conducting infectivity tests
with B. dimorpha included those used by Anthony et al. (1978)
to infect H.zea with H. algerae for mass production of spores.
Twenty larvae 4-5 days old were held in isolation without
food or water for 24 hr, fed 0.1 ml of a suspension containing
ca 106 spores per ml, and then placed on fresh media. Twenty
larvae 8-10 days old were allowed to consume one diseased
S. geminata pupa each, and an additional 20 larvae 8-10
days old were injected with 2-10 ul of spore suspension.
The adult moths were examined by phase-contrast microscopy
ca one month after they had ingested or been injected with
The B. dimorpha did not persist in S. invicta or S.
richteri for more than a few generations (Table 1). Three
weeks after the final feeding of B. dimorpha spores, the
S. invicta colonies were infected at the rates of 6% (colony
#1), 44% (colony #2), and 4% (colony #3). The colony of
B. dimorpha infection rates in colonies of S. invicta
and S. richteri.
Number of days after
ingestion of spores
aOne pupa in the nest population of 786 was infected.
S. richteri had an infection rate of 5%. One month later
(day 49), however, infection could be detected only in S.
invicta colony #2, and the infection rate in this colony
had declined from 44% to 8.5%. One month later (day 80),
only one infected pupa (out of 786 pupae in the age group
that exhibits pathognomonic signs) was found in the colony.
By day 110, all of the 1,000 pupae that were examined were
free of disease. All four colonies remained free of B.
dimorpha on all subsequent examinations. In contrast, the
colony of S. geminata from which the diseased pupae (those
fed to the test colonies) were taken maintained an infection
rate of ca 50% for 14 months.
None of the other species of ants tested were susceptible
to infection by B. dimorpha (Table 2). The H. zea was
not susceptible to infection either per os or by injection.
The B. dimorpha is host specific for S. geminata. Although
S. invicta and S. richteri may be readily infected, the
parasite does not persist in these factitious hosts beyond
a few generations. The 9 species (representing 8 genera)
of ants other than Solenopsis spp. proved refractory to
infection. While this is a very small sample of the living
species of ants, it is highly improbable that a parasite
that cannot maintain itself in factitious species of the
genus of its host would be able to infect species of other
Table 2. Susceptibility of selected species of ants and a
moth to infection by B. dimorpha.
Solenopsis geminata (Fabricius) +
S. xyloni McCook ?
S. invicta Buren
S. richteri Forel
Pheidole morrisi Forel
Monomorium minimum (Buckley)
M. floricola (Jerdon)
Crematogaster clara Mayr
Aphenogaster ashmeadi (Emery)
Ochetomyrmex auropunctatus (Roger)
(= Wasmannia auropunctata)
Pseudomyrmex elongata Mayr
Paratrechina longicornis (Latreille)
Camponotus floridana (Buckley)
Heliothis zea (Boddie)
a+ = susceptible to infection
= susceptible, but infection does not persist
- = not susceptible to infection
Following infection with spores of B. dimorpha from
S. geminata, signs typical of infection in S. geminata
develop in pupae of S. invicta and S. richteri. However,
pupae that are infected subsequently in these colonies (by
spores produced in siblings) develop atypical signs. The
blister-like clearings in the head and gaster are not apparent,
although faint clearing in the petiole may occur. Abnormal
eye development typical of the disease was noted, however
(Fig 5). Jouvenaz and Hazard were premature in their state-
ment that the course of the disease is identical in all
four species; they had examined only pupae infected with
spores produced in S. geminata.
The reasons) for the failure of B. dimorpha to cycle
indefinitely in S. invicta and S. richteri is unknown. Ob-
viously, they are "not well adapted" to these species, but
this merely restates our observations. Perhaps the cuticle
is less likely to rupture and cannibalization is less likely
in these species due to dehydration or putrification in
the intact state. The cadavers may then simply be discarded.
Fig 5. Diseased S. richteri pupa compared to healthy and
diseased S. geminata pupae. The pupa in 5A is
healthy. The signs exhibited by the S. richteri
pupa (5B) are as pronounced as they will become.
Note the faint clearing in the petiole and the
slight blurriness and irregularity of outline of
the eye. The diseased S. geminata pupa (5C)
exhibits typical pathognomonic signs at maximal
ABUNDANCE OF BURENELLA DIMORPHA IN NATURE
As part of a survey for pathogens of native and imported
fire ants in the United States, Jouvenaz et al. (1977) examined
307 colonies of S. geminata from74 collection sites in Florida
and Georgia. Twelve (3.9%) of these colonies from seven
collection sites in Florida were infected with B. dimorpha.
The infection rate at the type locality of B. dimorpha (which
was not included in the survey) was obviously much higher
than 3.9%. Therefore, I decided to investigate the infec-
tion rate at this specific site.
Materials and Methods
Sixty-nine colonies of S. geminata from the type locality
of B. dimorpha (State Road #26, ca 1-3 km east of Interstate
Highway 75, Alachua County, Florida) were screened for disease
as described in General Materials and Methods. Forty of
these colonies were examined in August, 1975, and the remain-
ing 29 were examined in September, 1977.
Results and Discussion
The infection rate at the type locality of B. dimorpha was
much higher than that of the 307 colonies examined by Jouvenaz
et al. Seventeen (42.5%) of the 40 colonies in the first
group, and 4 (13.8%) of the 29 colonies in the second group
(collected two years later) were infected. These limited
data suggest that B. dimorpha may be quite common in specific
localities, but not homogeneously distributed in nature.
Unfortunately, surveys for disease are very expensive, time-
consuming, and laborious, and the prospects of amassing
data sufficient to draw conclusions about the distribution
and abundance of B. dimorpha in nature appear remote. The
S. geminata has now been almost completely displaced from
the type locality of B. dimorpha by S. invicta.
PATHOLOGY OF BURENELLA DIMORPHA INFECTION
Very few protozoan diseases of insects may be diagnosed
by simple observation of pathognomonic signs. Pebrine (infec-
tion of the silkworm, Bombyx mori L., by the microsporidium
Nosema bombycis Naegeli) provides a classic example of patho-
gnomonic signs in insects, being characterized by the appear-
ance of dark, pepper-like spots on the integument. Infection
of the tropical fire ant, Solenopsis geminata (Fabricius),
by the neogregarine Mattesia geminata Jouvenaz and Anthony
(1979), may be readily diagnosed by eye teratology and a
pattern of abnormal melanization in pupae beginning with
a "sooty" black discoloration of the legs. In the same
host, infection by Burenella dimorpha Jouvenaz and Hazard
is manifested by signs of a type that appears to be unique
in insects. In general, however, specific manifestations
of protozoan infections, if present at all, are related
to particular tissues and may be detected only by histolog-
ical examination (Brooks, 1974).
Insects infected by protozoa may exhibit nonspecific
signs such as loss of appetite, diarrhea, sluggishness,
irregular growth, stunted or malformed adults, reduced fecun-
dity, and premature death of immatures or adults. Often,
however, there are no indications of disease other than
premature death (Brooks, 1974). Individual fire ant workers
infected with Thelohania solenopsae (Knell et al., 1977)
or any of four undescribed species of microsporidia, for
example, cannot be differentiated from healthy specimens
by either appearance or behavior. Diagnosis of these infec-
tions must be made by microscopic examination or, in the
case of T. solenopsae, by observing cysts in dissected gas-
ters. Similarly, bees, Apis spp., infected by Nosema apis
Zander are completely devoid of outward signs of disease,
and even histological diagnosis is difficult (Bailey, 1981).
Pupae of S. geminata infected by B. dimorpha exhibit
pathognomonic signs that are obvious, and, to the best of
my knowledge, unique. Jouvenaz and Hazard (1978) attributed
these manifestations to destruction of the hypodermis and
described them as follows:
A clear area in the occipital region of the head,
which appears about the time the developing eyes
become prominent, is usually the first noticeable
change. Later, similar clear areas appear in
the petiole and gaster, and the eyes become irreg-
ular in outline and appear sunken. Pupae having
such changes do not mature or even melanize. In-
stead, the clear areas increase in size, and the
cuticle eventually ruptures.(p. 27)
Jouvenaz and Hazard also noted that NMB spores develop in the
hypodermis, and MB spores develop in the fat body.
This brief description of the appearance of diseased
worker pupae and notation of tissue specificity of the spore
types is all that has been published on the pathology of
B. dimorpha infection. The signs of infection in sexual
pupae have not been described. The development and histo-
pathological basis of the unique manifestations of this
disease are described here.
Materials and Methods
The progressive development of pathognomonic signs
was recorded by periodically photographing healthy and dis-
eased pupae. Diseased pupae exhibiting the minimal degree
of eye development necessary for diagnosis and healthy pupae
of corresponding age and size were paired and held in minia-
ture nest cells. The specimens were positioned on a plate
of nonglare glass above a black parabolic background, illum-
inated by an annular fiber optic illuminator, and photographed
through a stereomicroscope fitted with a polarizing filter.
Photographs were made ca every 48 hours until the healthy
specimen had closed and the diseased specimens had reached
maximum expression of pathognomonic signs.
Specimens for both light and transmission electron
microscopy were fixed and embedded in epoxy resin. Tissue
specimens (heads, gasters, whole larvae, etc.) were prefixed
in buffered 1% osmium tetroxide (0.1 M sodium cacodylate
buffer, pH 7.5) for 30-60 minutes at room temperature, rinsed
in the same buffer, and partially hardened in buffered 2.5%
gluteraldehyde-l% acrolein (same buffer). Specimens were
washed in buffer and usually stored in Histocon (polyvinyl-
pyrrolidone, Tris-HCl, 2% chlorhexidine, distilled water;
Polysciences, Inc.) in the refrigerator overnight or for
several days prior to postfixation. Specimens were washed
in buffer, postfixed in buffered 1% osmium tetroxide for
2 hours at room temperature, washed in deionized water,
and en bloc stained in 0.5% aqueous uranyl acetate overnight.
Specimens were dehydrated with acidified 2,2-dimethoxypropane
(Lin et al., 1977) and infiltrated and embedded in a Spurr-
Quetol 651 resin (Ringo et al., 1979).
Blocks were sectioned with a LKB Huxley ultramicrotome.
For light microscopy, 2-4 ur sections were cut on dry glass
knives, spread in a drop of 10% acetone, mounted in immersion
oil, and studied by phase-contrast microscopy. Thinner
sections, 0.5-1.0 um, were stained with 1% aqueous toluidine
blue in 1% aqueous borax, rinsed in deionized water, mounted
in immersion oil, and examined by either bright-field or
phase-contrast microscopy. For transmission electron micros-
copy, gold sections were poststained with 2% aqueous uranyl
acetate followed by lead citrate (Reynolds, 1963). Grids
were examined and photographed at an accelerating voltage
of 75 kV in a Hitachi H-600 electron microscope.
Portraits of healthy and diseased male, female, and
worker pupae are presented in Figs 6-8. The clearing in
the dorsal thorax of the sexual pupae does not occur in
worker pupae, and occasionally is reduced or absent in
Fig 6. Diseased (left) and healthy male pupae of S. geminata.
Note the clearing of the occiput and dorsal thorax.
The light color of the eye is due to destruction of
the lens and a fluid space over the eye. X 10.
Fig 7. Diseased (left) and healthy female pupa of S. geminata.
The apparent faint clearing in the head of the healthy
specimen is a photographic artifact. X10.
Fig 8. Healthy (left) and diseased worker pupae of S. geminata.
Note eye pathology and clearing in occiput and petiole
of the diseased pupa. X 40.
sexual pupae. The eyes of the sexual pupae appear to be
less affected than the eye of the worker pupa; this may
be an artifact due to the larger (ca 16-20X in area) eyes
of sexual pupae. The diseased worker pupa is in an advanced
stage of infection, and exhibits typical pathognomonic signs.
Note the irregular outline of the eye, the derangement of
the facets, and its fainter color.
The development of pathognomonic signs is recorded
in Fig 9. In the first pair of photographs, the healthy
and diseased pupae (1H and lD, respectively) are almost
indistinguishable. The diseased pupa exhibits a faint clearing
of the petiole, and in life--but not in the photograph--
the earliest faint signs of eye teratology could be detected.
In the photograph taken 48 hours later (3D), the clearing
in the petiole is not evident due to a slight change in
lighting. Subsequent photographs show the development of
clearing in the head and petiole. The eye of the diseased
pupa is rather blurry in these photographs. This is not
due to focus, but to a fluid space over the eye.
The healthy pupa closed on day 18. The last photo-
graph of the diseased pupa (3D) shows an indentation in
the gaster, indicating dehydration. In a colony, rupture
would probably have occurred by this time as a result of
manipulation (being moved, groomed, etc.) by the workers.
Frontal sections of whole heads of younger and older
diseased pupae are presented, respectively, in Figs 10 and 12.
Fig 9. The development of pathognomonic signs in pupae of
S. geminata infected by B. dimorpha. The numerals
indicate the number of days after the series began
that the photographs were taken. The letters H and
D indicate, respectively, the healthy and diseased
specimens. X 30.
I I I i 3
Fig 9 (cont.) The development of pathognomonic signs in pupae
of S. geminata infected by B. dimorpha. The
healthy pupa closed on day 18 (18H). X 30.
Fig 10. Frontal section through the head of a young (ca 10
days postpupation) pupa infected by B. dimorpha.
Note the fluid space between the head and pupal
sheath. B = brain; IBC = infrabuccal cavity; arrows
point to the eyes. X 90.
Fig 11. Eye, optic nerve, and optic stalk of a young (ca 10
days postpupation) pupa infected by B. dimorpha.
OS = optic stalk; ON = optic nerve; arrow points to
eye. X 600.
Fig 12. Frontal section through the head of an older (ca 16
days postpupation) pupa infected by B. dimorpha.
Note the shrunken, fiberous appearance of the brain
and sunken, amorphous eyes. B = brain; IBC =
infrabuccal cavity; arrows point to eyes. X 90.
Note the fluid spaces in both specimens, and the shrunken,
fiberous appearance of the brain and the amorphous, sunken
eyes of the older specimen. The optic stalk, optic nerve, and
flattened, unstructured mass of developing eye tissue of a
younger pupa (corresponding in age to the specimen in Fig
10) are shown in Fig 11.
The destruction or inhibition of development of the
cuticle in infected pupae is shown in Figs 13-16. The healthy,
developing cuticle with its trophic microvilli and two
pores (Fig 13) is completely absent in the diseased specimen
(Fig 14). The microvilli of the latter extend randomly
into the fluid space. Details of the microvilli are pre-
sented at higher magnification in Figs 15 and 16. The tis-
sues shown in these electronmicrographs are from the head
where pathognomonic clearing develops. In tissues from
the gaster, where little or no clearing occurs, some cuticle
is present (Fig 26).
The structures of a healthy, developing pupa eye and
a diseased eye of a similar age are compared at increasing
magnifications in Figs 17-22. The lenses, composed of cu-
ticle, are absent in the diseased eye, and the rhabdoms
and associated cells are twisted, tangled masses. Also,
the basement membrane is absent, and only remnants of the
optic nerve may be seen. Pigment granules are present in
the pigment cells of the diseased eye.
Fig 13. Cuticle of a healthy pupa. Note the trophic
microvilli and two pores. PS = pupal sheath;
MV = microvilli; Cu = cuticle. X 1,500.
Fig 14. Body surface of a pupa infected by B. dimorpha.
There is no cuticle, and the microvilli that
normally penetrate the cuticle extend into the
fluid space. PS = pupal sheath; MV = microvilli.
Fig 15. Detail of the cuticle in a healthy pupa. Note
the microvilli penetrating the cuticle. MV =
microvilli. X 12,500.
Fig 16. Detail of microvilli extending into the fluid space
of a pupa infected by B. dimorpha. X 12,500.
Fig 17. Sagittal section through the developing eye of
a healthy pupa. L = lens; Rh = rhabdom and
developing associated cells; ON = optic nerve.
Fig 18. Sagittal section through the eye of a pupa
infected by B. dimorpha. There is no lens,
and the rhabdoms and the associated cells are
twisted and tangled in an amorphous mass.
Rh = rhabdom and associated cells; ON = optic
nerve. X 450.
Fig 19. Sagittal section through the eye of a healthy
pupa. X 1,500.
Fig 20. Sagittal section through the eye of a diseased
pupa. Rh = rhabdom and associated cells.
". ,,, *
.. .- */ **
S.. .. .
: a ,i
". S,:il "" *"" -" """ e
*- v "' '""
.-., .- .l. '
C"e :. .S ,
Fig 21. Sagittal section through the developing eye of a
healthy pupa showing detail of optic nerve and
base of the developing ommatidium. Rh = rhabdom
and associated cells; BM = basement membrane;
PG = pigment granules; ON = optic nerve; FB =
fat body. X 3,000.
Fig 22. Sagittal section through the eye of a pupa infected
by B. dimorpha. Only remnants of the optic nerve
may be seen. The basement membrane area is destroy-
ed. ON = optic nerve; FB = fat body; PG = pigment
granules; Rh = rhabdom and associated cells.
Fig 23. Fat body of a healthy pupa.
Fig 24. Fat body of a pupa infected by B. dimorpha. Note
the depleted appearance of fat body and the presence
of several MB spores. X 7,500.
..1 ,. -
a ** *
# t.' .
[ '1 "
Fig 25. Tissue specificities of MB and NMB spores of
B. dimorpha. The MB spores in the darkly stained
fat body are seen in groups of less than eight
because of the thinness of this section (o.5 mu).
Tissue specificities of MB and NMB spores of
B. dimorpha. Nonmembrane bound spores are con-
fined to the hypodermis; MB spores are confined
to the fat body. The MB spores are seen in
octets in this section because of its thickness
(6-8 mu). Some cuticle is seen in this section
from the gaster. X 240.
Fig 27. Tissue specificities of MB and NMB spores of
B. dimorpha seen at higher magnification. X 375.
k, ,~~ S.
^Ti~- r *I
The fat body is diminished in infected pupae, as is
evident in Figs 23 and 24. Several MB spores may be seen
in the latter electronmicrograph.
The tissue specificities of the two spore types are
shown at increasing magnifications in Figs 25-27. The MB
spores are confined to the fat body; NMB spores are confined
to hypodermal and connective tissues, including that in
the deeper parts of the body such as the tissue encasing
the brain. The tissue section in Fig 25 is only 0.5 mu
thick; therefore, MB spores are seen in groups of 2-4 rather
than octets. The tissue sections in Figs 26 and 27 are
ca 6-8 mu thick, and thus show the octosporous pansporoblasts
The pathognomonic signs of B. dimorpha infection in
S. geminata reflect damage or destruction of the developing
adult cuticle. In the pupal stage of development, the integu-
ment of the adult ant forms under the integument of the
fourth-instar larva. During this period, the larval integument
is transformed into a protective sheath that adheres closely
to the developing adult integument and is moulted when the
development of the latter is complete (eclosion). Over
most of the body, the adult cuticle of pupae infected by
B. dimorpha appears to be damaged or inhibited only to the
extent that tanning is inhibited (Fig 26; also, the adult
morphology is essentially developed). In parts of the head
and in the petiole, however, the cuticle is completely destroy-
ed (Figs 13-16). In these areas, tissue fluids seep into
the space between the denuded hypodermis and the pupal sheath.
As the infection progresses, the fat body is diminished and
the brain shrinks (due to loss of lipids?). The mass of
these tissues, covered by hypodermis, decreases in size and
recedes from the pupal sheath, which is extended by fluid.
Jouvenaz and Hazard (1978) attributed the development of
clear areas in the head and petiole to destruction of the
hypodermis. More correctly, infection of the hypodermis
results in destruction or inhibition of formation of the
The malformation of the eyes is also due to destruction
of the cuticle. The lens are part of the cuticle, and the
remaining components of the ommatidium (rhabdom and pigment,
retinular, and semper cells) extend between the lens and
a basement membrane. The lens are nonexistent in eyes in-
fected by B. dimorpha, and the rhabdoms and associated cells
become twisted masses. Perhaps the surrounding connective
tissue matrix is also destroyed. The eyes of pupae infected
by B. dimorpha develop after cuticle destruction has begun,
and are abnormal from the time they first become visible.
Thus, the derangement of the eyes is teratologic in nature.
TEMPERATURE-DEPENDENT SPORE DIMORPHISM
IN BURENELLA DIMORPHA
The effect of temperature on the relative abundance
of MB and NMB spores was first studied by Maddox (1966)
in the dimorphic microsporidium Vairimorpha necatrix (Kramer).
This parasite was originally thought to be a mixed infection
involving two species of microsporidia, Nosema necatrix Kramer
and Thelohania diazoma Kramer, that produced NMB and MB
spores, respectively, in lepidopteran hosts (Kramer, 1965).
Maddox demonstrated that only NMB spores are produced if
infected armyworm larvae, Pseudaletia unipuncta (Haworth),
are reared at temperatures of 320 C or above. At lower
temperatures (160 C) 40% or more of the spores are the MB
type. Maddox infected armyworm larvae with an apparently
pure suspension of NMB spores obtained from larvae reared
at elevated temperature, and found that both types of spores
were produced in larvae held at temperatures below 320 C.
This led Maddox to suggest (as the more radical of two pos-
sibilities) that N. nectarix and T. diazoma were not
two species at all, but rather one species with two distinct
developmental cycles and resulting spore forms, the ratio
of which is influenced by temperature" (p. 112).
The dimorphic nature of this microsporidium was experi-
mentally demonstrated by Fowler and Reeves (1974) through
the use of mechanical as well as thermal and temporal methods
to separate the two spore types for transmission studies.
They retained the name N. necatrix, suppressing the name
T. diazoma. Pilley (1976), recognizing that N. necatrix does
not conform to the definitive characteristics of the genus
Nosema, reassigned it to a new genus, Vairimorpha. Jouvenaz
and Hazard (1978) observed that V. necatrix and B. dimorpha
are related and suggested that the genus Vairimorpha properly
belongs in the family Burenellidae.
The temperature dependence of spore dimorphism in V.
necatrix was also demonstrated by Maddox and Sprenkel (1975),
who serially passed V. necatrix spores through eight genera-
tions of P. unipuncta held at 21.10 and 320 C. Maddox and
Sprenkel also demonstrated temperature-dependence of spore
dimorphism of "Nosema plodiae -Thelohania nana" Kellen and
Lindegren, and in an undescribed microsporidium. Maddox
(1966) described the V. necatrix infections of P. unipuncta
reared at elevated temperature as "light."
The primary purpose of the experiments reported here
was to determine whether spore dimorphism in B. dimorpha
is temperature dependent. The effects of temperature on
spore yield ("light infections", Maddox) and size were also
measured. In addition, since the MB spores of B. dimorpha
develop in the fat body while NMB spores have been found
only in the hypodermis (Jouvenaz and Hazard, 1978), I attempted
to answer the following questions: 1) If MB spores do
not develop at all temperatures at which NMB spores develop,
is it because sporulation is inhibited in the fat body?
Or, 2) do NMB spores develop in lieu of MB spores in the
fat body? And 3) are the effects of heat and cold the same
in this respect? Resolution of these questions was attempted
by histological examination of diseased pupae reared at
high or low temperatures.
Materials and Methods
Ratios of Spore Types
Fourth-instar larvae were removed from a heavily infected
laboratory colony and held in miniature nest cells at the
desired temperature (20, 22.5, 28, 32, and 35 C) overnight.
The temperatures in the incubators were monitored by hygro-
thermographs and mercury thermometers. Humidity was main-
tained close to 100% by the design of the cells (see General
Materials and Methods). Larvae that pupated during the
first 16 hr were discarded, as were those that had not pupated
during the next 24 hr. A contingent of nurses (young workers
primarily care for the immatures) captured from the brood
piles was introduced to care for the immatures. These were
removed shortly before the end of pupal life (at that time
they were no longer needed to groom the pupae) to facilitate
the examination and harvest of specimens in advanced disease.
The uninfected pupae served as controls to monitor the rate
of maturation at the different temperatures. Infected pupae
were harvested only after eclosion of the controls, when
the disease had progressed to near the point of cuticle
rupture and spore production was essentially complete. The
completeness of spore production was judged by the percentage
of immature spores seen in phase-contrast microscopic examina-
tion. Spore ratios were determined only for those pupae
in which > 98% of the NMB spores appeared to be mature.
The initial rearing study indicated that the develop-
ment of MB and NMB spores of B. dimorpha is inhibited by
both high and low temperatures. Therefore, an additional
study was conducted to determine the maximum age of the
host at which MB spore inhibition can be effected by raising
or lowering the temperature. Immature workers were divided
into five age groups and half of each group were incubated
at 20 C and the remaining half at 32 C. The age groups
were 1) fourth-instar larvae; 2) larvaform pupae; 3) early
pupaform pupae (eye development not yet evident); 4) pupae
whose eyes had developed sufficiently to permit diagnosis
of infection; 5) pupae slightly older, having first evidences
of clearing in the occiput. These groups were reared and
examined as described above.
Spore ratios were determined by homogenizing individual
pupae in ca 1 ml distilled water in a 3 ml glass tissue
grinder, and examining this extract by phase-contrast
microscopy at a magnification of 600X. The concentration
of spores in these extracts (0-6 spores/field) was low enough
to facilitate counting and individual study of spores. The
percentage of mature and immature MB spores was based on
the first 200 spores observed.
Spore measurements were made using an A. E. I. Cook
Image-Splitting micrometer calibrated for the microscope
used. The spores were immobilized for measurement by trap-
ping them between a layer of agar and the coverslip. This
was accomplished by pipetting a small pool of warm, liquid
Noble Agar (Difco Laboratories, Detroit, Mich.) (1.5%) on
the slide with a dropping pipette and bulb, and withdrawing
a drop of the agar just before gelling occurred. This pro-
duced an agar bed which was rather flattened on top. A
very small drop of spore suspension was then placed on a
coverslip, inverted and placed on the agar.
Spore yield was determined by weighing groups of 25
pupae in advanced disease on an electric balance, homogeniz-
ing these in distilled water, adjusting the volume appropri-
ately, and counting spores with a hemacytometer (phase-
contrast microscopy, 300X).
Tissue specimens from gasters of pupae held at 20 or 32 C
until infection was very advanced were fixed, stained with
heavy metals embedded in Spurr-Quetol resin, sectioned, and
examined by phase-contrast microscopy. The details of these
procedures are given in the section on host pathology.
The development of MB spores was inhibited in pupae
held at high (32 or 35 C) or low (20 C) temperature. At
28 C, the temperature at which S. geminata colonies seem
to live best in my insectary, an average of 35.9% 2.6
of the spores in 25 pupae were of the MB type (Table 3).
Jouvenaz and Hazard (1978) stated that MB spores typically
constitute ca 25-40% of the spores in diseased pupae collected
in the field.
Only 10 of the pupae held at 35 C survived to an advanc-
ed stage of infection. These pupae were devoid of MB spores,
and a small number (ca 2%) of the NMB spores from these
pupae were morphologically aberrant (pairs of spores fused
laterally or in tandem, Y-shaped or triangular spores, giant
spores, etc.). Thirty-five degrees appears to be very near
the upper thermal limit of survival for host and parasite.
Twenty-two of the 25 pupae held at 32 C were negative
for MB spores. Only four immature MB spores were among
the 600 spores from the remaining three pupae (200 spores
from each specimen were examined).
Table 3. Relative abundance of B. dimorpha MB spores in
pupae of S. geminata reared from larvae at various
No. (%) positive
for MB spores
% MB spores in
35.9 2.6 (31-39)
aMean and (range)
bMean and standard deviation (range)
Table 4. B. dimorpha spore yield of pupae of S. geminata
reared from larvae at 20, 28, or 32 C.
Temperature NMB Spores MB Spores
(degrees C) Spores/pupa Spores/mg Spores/pupa Spores/mg
20 6.7 x 105 3.9 x 105 0 0
28a 6.3 x 105 4.4 x 105 2.7 x 105 2.0 x 105
32 4.8 x 105 2.7 x 104 0 0
aTotal spore production at 28 C = 8.9 x 105/pupa or 3.9 x 105/mg.
The development of MB spores was completely inhibited
in 25 pupae held at 20 C. A few immature MB spores were
found in pupae held at 22.5 C, indicating that the lower
thermal threshold of MB sporulation is between 20 and 22.5 C.
Pupae reared from larvae at 20 and 28 C produced similar
numbers of NMB spores (Table 4). Pupae reared at 32 C produced
ca 75% as many spores per pupa and ca 60% as many spores per mg
bodyweight as did those reared at 28 C.
The inhibitory effect of temperature on MB sporulation
in pupae placed at high or low temperature at comparable
ages was essentially the same (Table 5). Specimens placed
at 20 or 32 C as larvae or larvaform pupae (a total of 40
specimens) were negative for MB spores. Of the 10 specimens
placed at 20 C as early pupaform (lacking visible eye develop-
ment) pupae, five had no MB spores, two had one MB spore
each, two had two MB spores each, and the remaining pupa
had seven MB spores among the 200 spores that were examined.
Of the 10 specimens placed at 32 C as early pupaform pupae,
nine had no MB spores and the remaining pupa had only one
MB spore among the 200 spores. Thus, mean spore production
by this age group was less than one percent at either tempera-
The production of MB spores in pupae that had developed
eyes to the extent that diagnosis of disease was possible
at the time they were placed at 20 or 32 C was lower than
in those reared at 28 C for their entire life. Those held
Table 5. Relative abundance of B. dimorpha MB spores in pupae
of S. geminata placed at 20 or 32 C at different ages.
% MB Spores
Age Group 20 C 32 C
lacking eye development
minimum eye development
slight clearing, occiput
aMean and standard deviation
at 20 C produced an average of 10.3% 7.6 (range 2-25%)
MB spores; those held at 32 C produced an average of 16.8%
6.8 (range 6-23%). Pupae exhibiting minimal clearing
of the occiput produced averages of 18.7% 14.0 (range
1.5-40%) and 24.4% 3.7 (range 18.5-33.5%) at 20 and 32 C
respectively. Approximately 90% of the spores in these
latter two age groups appeared to be mature.
The measurements of NMB spores from pupae reared from
larvae at 20, 28 or 32 C were, respectively, 3.0 0.2 X
6.8 0.3 mu, 3.0 0.1 X 6.8 0.3, and 3.0 0.0 X 7.0
0.4 mu. Obviously, there were no differences in NMB spore
sizes (MB spores were not produced at 20 or 32 C).
Neither MB or NMB spores were seen in tissue sections
of fat body of pupae reared at 20 or 32 C. Very few MB
spores were seen in sections of fat body of pupae reared
at 22.5 C.
The development of NMB spores precedes the development
of MB spores, and NMB spores predominate in number. In
field colonies, MB spores typically constitute ca 25-40%
(occasionally fewer) of the spores from pupae in advanced
stages of infection (Jouvenaz and Hazard, 1978). The present
study has demonstrated that MB sporulation also occurs in
a more restricted range of temperature than NMB sporulation.
The lower thermal limit of MB spore development appears
to be between 20 and 22.5 C. At 28 C production ofMB spores
was consistently near maximum. The upper thermal limit
of MB spore development is below 32 C. Thus, optimal MB
spore production occurred near the upper thermal limit of
spore production. This is reminiscent of the activity curve
of heat-liabile enzymes and the growth curves of poikilo-
therms over a range of temperatures. Very possibly the
upper thermal limit of MB spore development is the tempera-
ture of inactivation of an enzyme(s).
The B. dimorpha MB spore production is also inhibited
by low temperature (20-22.5 C). In contrast, V. necatrix MB
spores are produced at temperatures at least as low as 16 C.
This difference in cold sensitivity may reflect different
requirements for enzyme stability due to host behavior.
Loss of quaternary enzyme structure at low temperature is
relatively common in homeotherms and their microbial symbiotes.
The homologous enzymes of poikilotherms do not, at least
in some cases, lose their quaternary structure at low temper-
ature (Hochachka and Somero, 1973). The hosts of both V.
necatrix and B. dimorpha are, of course, poikilotherms;
however, the host of the latter microsporidium is a subtrop-
ical, social insect that actively tends its brood, moving
them in a subterranian environment to regulate temperature
and humidity. The tumulus of the nest is a solar heating
device, and there is evidence that metabolic heat production
is significant (Seeley and Heinrich, 1981). The various
lepidopteran hosts of V. necatrix, however, are exposed
to ambient temperatures as are most poikilotherms. Thus,
the stability of quaternary enzyme structure may be more
critical in this microsporidium.
We may only speculate as to the physiological functions)
that may be inhibited by extremes of temperature. However,
the critical event in MB sporulation may well be meiosis.
If the sister nuclei of the diplocaryon found in NMB spores
and the vegetative stages of B. dimorpha are each diploid,
meiosis would produce eight uninucleate haploidd) spores
(MB spores occur in octets bound by a membrane). Evidence
that meiosis does indeed occur in microsporidia has been
published by Loubes et al. (1976) and Hazard et al. (1979).
The hypothesis that meiosis may be the critical event
in MB sporulation inhibition by extremes of temperature
is at least circumstantially supported by the essentially
complete absence of MB spores in specimens placed at 20
or 320 C at an age before eye development becomes visible,
and the partial and highly variable degree of inhibition
in specimens exhibiting early eye development. The appear-
ance of MB sporoblasts approximately coincides with the
earliest visible eye development. At this stage, second
cycle meronts enlarge to the full size of mature pansporo-
blasts, and the diplocaryotic nuclei become very diffuse.
Successive nuclear divisions then produce plasmodia containing
two, four, and finally eight nuclei. Endogenous cytoplasmic
budding around these nuclei produces an octet of uninucleate
spores bound by a membrane. Thus, meiosis occurs at the
age limit for thermal inhibition of MB spore development.
The high variability of inhibition among pupae in this age
group simply reflects the slight variability in age of indi-
viduals and the degree to which MB sporulation had advanced
at the time of temperature change. Once sporogony is under
way (meiosis has occurred?), it apparently cannot be stopped
by temperatures at least as low as 20 C or as high as 32 C.
AMENDMENTS TO THE DESCRIPTION OF BURENELLA DIMORPHA
The vegetative stages of B. dimorpha were described
by Jouvenaz and Hazard (1978) from light microscope studies
Burenella dimorpha appears to have two sequences
of merogony. The first involves uninucleate
cells with deeply staining cytoplasm (Giemsa)
and compact nuclei that become binucleate and
divide. The second sequence involves binucleate
cells with moderately staining cytoplasm and
less dense nuclei that become tetranucleate and
divide to produce two binucleate cells.(p. 27)
The accuracy of this description is challenged by the follow-
Studies on the transmission of B. dimorpha have demon-
strated that only the NMB spores are infective for fire
ant larvae (see Transmission and Infectivity of Spores).
The NMB spores are diplocaryotic (Fig 32); therefore, the
infecting sporoplasm is (presumably) also diplocaryotic.
The formation of uninucleate meronts would require either
fusion of the sister nuclei or separation of the nuclei
followed by cell division. Studies of Nosema spp. (Sprague
et al., 1968; Vavra and Undeen, 1970; Cali, 1971) and an
Amblyospora sp. (Andreadis and Hall, 1979) indicate that in
these species, the diplocaryotic condition persists throughout
the life cycle.
Recently I collected an undescribed microsporidium
(a parasite of S. invicta in Brazil) that appears to be
closely related to B. dimorpha. This dimorphic species
also has two sequences of merogony; however, its meronts
are diplocaryotic. These observations and reports prompted
me to reexamine the life cycle of B. dimorpha, with particular
attention to the nuclear condition of early meronts. Addi-
tional studies of ultrastructure, including scanning electron
microscopy, were also conducted.
The appearance of sexual castes in colonies of S. geminata
is seasonal (unpublished), and Jouvenaz and Hazard were
unable to examine diseased sexual pupae. Spores from sexual
pupae have now been examined, and found to differ in one
respect from spores from worker pupae.
Materials and Methods
Specimens were smeared on acid-alcohol cleaned, galss
microscope slides, air-dried, and fixed in acetone-free
methanol for 5 minutes. Slides were stained with 10% Giemsa
stain in phosphate buffer, pH 7.41 (Fisher Gram Pac Buffer)
for 12 minutes, rinsed in tap water or acidified deionized
water deionizedd water adjusted to pH 6.8 with acetic acid),
blotted with filter paper, and examined under oil.
Scanning Electron Microscopy
Spores from a suspension in deionized water were attached
to carbon-stabilized, formvar coated grids with polylysine
(Mazia et al., 1975), fixed overnight in buffered 1% osmium
tetroxide, and dehydrated in 2,2-dimethoxypropane. Specimens
were critical point dried, sputter coated with palladium-
gold, and examined in a Hitachi H-6010A high resolution
scanning electron microscope at an accelerating voltage
of 50 kV.
Jouvenaz and Hazard (1978) were correct in their state-
ment that B. dimorpha has two sequences of merogony; however,
I am now convinced that these cells are diplocaryotic and
bidiplocaryotic, rather than uninucleate and binucleate.
The cells of the first merogonic sequence are very small,
and their compact nuclei are very closely appressed; the
cells of the second merogonic sequence are larger, and their
nuclei are rather diffuse. The diplocaryotic condition
of these nuclei is not readily apparent, and the published
diagrammatic life cycle adequately reflects their appearance
in stained smears.
Jouvenaz and Hazard also stated that the surface of
both spores is smooth. This is correct only with respect
to the NMB spore (Figs 28 and 30). The surface of the MB
spore has a reticular pattern of ridges that are evident
Fig 28. Transmission electronmicrograph of a mature NMB
spore of B. dimorpha. Note smooth spore surface.
N=nucleus; Pf=polar filament. X 17,100. Inset:
NMB spores as they appear in phase-contrast micros-
copy. X 2,000.
Fig 29. Transmission electronmicrograph of a mature MB
spore of B. dimorpha. Note the surface sculpture
(arrow) and the polar cap (upper end of spore).
N=nucleus; Pf=polar filament. X 18,000. Inset:
MB spore as it appears in phase-contrast micros-
copy. X 2,000.
Fig 30. Scanning electronmicrograph of a mature NMB spore
of B. dimorpha. Note the smooth surface and lack
of a morphologically differentiated polar cap area.
Fig 31. Scanning electronmicrograph of a mature MB spore
of B. dimorpha. Note the surface sculpture and
morphologically differentiated polar cap area
(bottom of spore). X 16,500.
in both the transmission (Fig 29) and scanning (Fig 31)
electronmicrographs. The ridges may also be seen in the
immature spores in Fig 33.
The pansporoblast membrane of B. dimorpha in worker
pupae was described by Jouvenaz and Hazard (1978) as sub-
persistent and delicate, and ruptured during dissection
of the host. Mature spores are never, and immature spores
are rarely, seen in octets in smears. This is also true of
B. dimorpha from male pupae; however, the pansporoblast mem-
brane is quite persistent in specimens from female sexual
pupae. In the latter, mature spores are commonly seen in
octets in smears or aqueous extracts.
The diplocaryotic condition of meronts of B. dimorpha is
most difficult to observe. I have been unable to obtain
convincing photomicrographs of diplocaryia in these cells,
despite the use of high contrast technical pan film and
manipulation of staining protocols. After studying a very
large number of these cells with a microscope of excellent
quality, focusing up and down to distinguish depth of stain-
ing and faint curvatures, I have become convinced that these
cells are indeed diplocaryotic. Interestingly, I have con-
sistently obtained much superior definition in stains of
an undescribed, closely related microsporidium from S. invicta.
Fig 32. Transmission electronmicrograph of an immature NMB
spore. The spore wall has not yet developed.
N=nuclei; Pf=polar filament. X 12,000.
Fig 33. Transmission electronmicrograph of two immature
MB spores within a delicate pansporoblast membrane.
The surface sculpture is already prominent, although
the endospore is not yet fully developed. PM=pan-
sporoblast membrane. X 9,000.
SUMMARY AND CONCLUSIONS
Burenella dimorpha was described by Jouvenaz and Hazard
(1978) as the type species of a new genus which represents
a new family, Burenellidae. This family includes those
species of microsporidia having two sporogonic sequences,
one producing nonpansporoblast membrane-bounded (NMB) spores
and the other producing octets of spores bounded by a pansporo-
blast membrane (MB). Jouvenaz and Hazard described this
microsporidium as a dimorphic species on the basis of light
microscope studies of the life cycle. They were unable
to separate the two spore types, and therefore could not
confirm by feeding tests that B. dimorpha is a single species.
Neither, of course, could they determine which or if both
spore types are infective. Also, the degree of host-speci-
ficity of the parasite was unknown. Three species of Solenopsis
other than S. geminata were shown to be susceptible to infection,
but later observations cast doubt on the ability of the
parasite to persist in populations of these species.
Fire ant pupae infected by B. dimorpha develop pathogno-
monic signs--malformation of the eyes and blister-like clear
areas in the occiput and petiole. These signs increase
in severity as the infection becomes more advanced, and
eventually the pupa ruptures and is cannibalized by workers.
The histopathological basis of these signs was not understood,
nor was the exact mode of transmission of the parasite.
An effort was made to extend our limited knowledge
of B. dimorpha in the areas outlined above. In addition,
the effect of temperature on spore dimorphism was investi-
gated, and certain aspects of the life cycle and morphology
of the parasite were reexamined. The conclusions reached
in this study are summarized below.
1. It was experimentally confirmed that B. dimorpha is
a dimorphic species, producing two morphologically distinct
types of spores, and not a dual infection of two species
2. Only NMB spores are infective; unextruded MB spores
are expelled in the meconium upon pupation.
3. The intracolonial cycle of transmission of B. dimorpha
infection is from ruptured, diseased pupae to fourth-instar
larvae via the adult workers, who act as mechanical vectors.
4. Adult workers do not ingest spores into the crop,
but divert them to the infrabuccal cavity where they are
formed into infrabuccal pellets with particulate food material.
5. Only fourth-instar larvae are vulnerable to infection.
6. Burenella dimorpha is host-specific for S. geminata.
Infections in S. invicta and S. richteri do not persist
in the colony.
7. Burenella dimorpha may be locally abundant, even
though it is not common in the S. geminata population as
8. Spore dimorphism is temperature dependent. Mem-
brane bound sporulation is inhibited by both low (20-22.5 C)
and high (32 C) temperatures; NMB sporulation is affected
little, if at all.
9. Nonmembrane bound spores do not develop in lieu
of MB spores in the fat body at low or high temperatures.
10. The pathognomonic signs of B. dimorpha infection
are due to destruction or inhibition of formation of the
11. Meronts are diplocaryotic and bidiplocaryotic,
not uninucleate and binucleate as originally described.
12. The NMB spore surface is smooth, but the MB spore
surface has a complex pattern of ridges and a well defined
polar cap area.
13. The pansporoblast membrane is extremely delicate
and subpersistent in workers (as originally described),
and in male pupae; in female pupae it is persistent.
Diplocaryon. Two nuclei in intimate contact, their membranes
adhering to each other over a large area. These nuclei
divide in a plane perpendicular to the plane of their
physical contact, in synchrony.
Disporous. Producing two sporoblasts.
Endogenous sporogony. Sporogony within the limiting membrane
of the sporogonial plasmodium.
Endospore. The chitinous inner spore wall.
Exospore. The proteinaceous outer spore covering or envelope.
MB. As used in this dissertation--bounded by a pansporoblast
Merogony. Vegetative multiplication. Synonym: schizont.
Meront. A cell that undergoes binary or multiple fission in
the vegetative phase of the life cycle. Synonym: schizont.
NMB. As used in this dissertation--not bounded by a pansporo-
Pansporoblastic membrane. A somewhat modified membrane that
encloses a group of spores or sporoblasts.
Polar cap. An internal, chromophilic area at the anterior end
of the spore.
Sporogenesis. The transformation of the sporoblast into a
Sporogenesis. The transformation of the sporoblast into a
Sporogony. The production of sporoblasts.
Sporont. A cell that gives rise to a sporoblast.
Sporoplasm. The nucleus (nuclei) and cytoplasm contained
within the spore that is injected into a host cell.
Sporulation. Spore production; sporogony plus sporulation.
Vegetative stage. That phase of the life cycle when the
parasite is actively feeding and multiplying prior to
Allen, G. E., and W. F. Buren. 1974. Microsporidan and
fungal diseases of Solenopsis invicta Buren in Brazil.
J. N. Y. Entomol. Soc. 82: 125-130.
Allen, G. E., and A. Silveira-Guido. 1974. Occurrence of
Microsporida in Solenopsis richteri and Solenopsis sp.
in Uruguay and Argentina. Fl. Entomol. 57: 327-329.
Andreadis, T. G., and D. W. Hall. 1979. Development, ultra-
structure, and mode of transmission of AMblyospora sp.
(Microspora) in the mosquito. J. Protozool. 26: 444-
Anthony, D. W., K. E. Savage, E. I. Hazard, S. W. Avery, M.
D. Boston, and S. W. Oldacre. 1978. Field tests with
Nosema algerae Vavra and Undeen against Anopheles
albimanus Wiedeman in Panama. Misc. Pub. Entomol. Soc.
Amer. 11: 17-28.
Avery, S. W., D. P. Jouvenaz, W. A. Banks, and D. W. Anthony.
1977. Virus-like particles in a fire ant, Solenopsis
sp. (Hymenoptera: Formicidae) from Brazil. Fl. Entomol.
Bailey, L. 1981. Honey bee pathology. Academic Press, New
York. 124 pp.
Brooks, W. L. 1974. Protozoan infections, Vol. II. p. 327-
330. In Ge. E. Cantwell [ed.] Insect diseases. Marcel
Dekker, Inc., New York.
Broome, J. R., P. P. Sikorowski, and B. R. Norment. 1976.
A mechanism of pathogenicity of Bearuveria bassiana
on larvae of the imported fire ant, Solenopsis richteri.
J. Invertebr. Pathol. 28: 87-91.
Bucher, G. E. 1963. Nonsporulating bacterial pathogens,
Vol. II, p. 117-147. In E. A. Steinhaus [ed.] Insect
pathology--An advanced treatise. Academic Press, Inc.,
Bulla, L. A., and T. C. Cheng [eds.]. 1976. Comparative patho-
biology, Vol. I, Biology of the microsporidia. Plenum
Press, New York and London. 371 pp.
Bulla, L. A., and T. C. Cheng [ed.]. 1977. Comparative patho-
biology, Vol. II, Systematics of the microsporidia.
Plenum Press, New York and London.
Cali, A. 1971. Morphogenesis in the genus Nosema. Proc.
IV Int. Colloq. Insect Pathol., 431-438. University of
Maryland Press, College Park, Md.
Collins, H. L., and G. P. Markin. 1971. Inquilines and other
arthropods collected from nests of the imported fire ant,
Solenopsis saevissima richteri. Ann. Entomol. Soc. Amer.
Fowler, J. L., and E. L. Reeves. 1974. Spore dimorphism
in a microsporidian isolate. J. Protozool. 21: 538-
Hazard, E. I., T. G. Andreadis, D. J. Joslyn, and E. A. Ellis.
1979. Meiosis and its implications in the life cycles
of Amblyospora and Parathelohania (Microspora). J.
Parasitol. 65: 117-122.
Hazard, E. I., E. A. Ellis, and D. J. Joslyn. 1981. p. 163-
182. Identification of microsporidia. In H. D. Burgess
[ed.] Microbial control of pests and plant diseases 1970-
1980. Academic Press, New York.
Hazard, E. I., and S. W. Oldacre. 1975. Revision of Micro-
sporidia (Protozoa) close to Thelohania with descriptions
of one new family, eight new genera, and thirteen new
species. U. S. Dept. Agric. Tech. Bull. No. 1530.
Hochachka, P. W., and G. N. Somero. 1973. Strategies of bio-
chemical adaptation. W. B. Saunders Co., Philadelphia,
Jouvenaz, D. P. 1981. Percoll an effective medium for
cleaning microsporidian spores. J. Invertebr. Pathol.
Jouvenaz, D. P. Some protozoa infecting fire ants, Solenopsis
spp. In L. A. Bulla and T. C. Cheng [ed.] Comparative
pathobiology, Vol. V. Plenum Press, New York and London
Jouvenaz, D. P., G. E. Allen, W. A. Banks, and D. P. Wojcik.
1977. A survey for pathogens of fire ants, Solenopsis
spp., in the Southeastern United States. Fl. Entomol.