I!OR ,Fl'OLOGY, DEVELOPTF :lT. Ar;C UJLT.RA TPLrCTu' E
Oit m -.:.-F ? T'h:, Th.ter
A DISSF.TATIO''! PPFSEiTE[, TO THE :-,LDUTE COUNtCIL OF
THE .JIVE;l ITr; OF :LC'PIA
IN PARTIAL FULFILL;;MET OF TH- PEI:'-PE;.ENTS FOP THE OEGREE OF
DOCTOR OFr PFHLGEflO'
L:IVEF:.ITr OF FLC.FILIA
1 7 3
I wish to express my gratitude to Dr. James W. Kimbrough, the
chairman of my supervisory cor.-ittee, for his willing, able and
constant guidance and advice throughout the course of this study.
My special thanks are due to Dr. Henry C. Aldrich, who opened a
new field of ultrastructure to me, for his patient assistance and
his criticism on the preparation of the manuscript.
I also wish to extend my thanks to Drs. Leland Shanor, Dana G.
Griffin, 111, and Pobert E. Stall who worked on my supervisory
committee along with Drs. Kimbrough and Aldrich.
Thanks are accorded to Institute of International Education,
New York, and U. S. State Department for the financial assistance
provided, without which this study could not have been possible.
Finally, I wish to thank my wife, Patricia, for her continuous
encouragement and generous understanding, and for typing and proof-
reading the manuscript.
TABLE OF CONTENTS
Acknowledgerents. . . . . . . .
List of illustrations . . . . . .
Abstract. . . . . . . . ...
I The M:orphology and Development.
II Conidiogenesis. . . . .
III Haustorial Mechanis m. . . .
[V Taxonomic Position. . . ...
Bibliography . . . . . . .
Biographical Sketch . . . . . .
. . . . . .
. . . . . .
LIST OF ILLUSTFATIONS
Plate I. :-.i ... ,.. i.-: habit and habitat. . . 13
Plate II. ";:':-,~: .-*.,i.- light micrographs of the
cros.7 sections of ;porocarp. . . . . .. 15
Plate III. 7 :r-..:rt .:.' r:, haustoriial mother cells. . 1
Plate IV. ;:-':..-' r: ',mtr-, electron micro graph of the
hau;troiumi connected to a h5usEori~ l fniother
cell by a neec .and light micrographs of the
young thallus growing on an antennal segment 19
Plate V. .-;r.,lt. i c .:~- l r. ', electron nicro.-raph of th?
;subh ,enial cell ; and light Ticirograph: of
marginal tip of mtture thalluS ....... 21
Plate VI. ;.:-.-:n:', s,.,;-rt, electron nicrograph of
the subh.,rieniun. ... . . . . . . 23
Plate VII. T-: :.! '-a r- :-.: r conidia and the
phial ide . . . . . . . . . 25
Plate VIII. r- -:"- r:a ~'::,ir', light and electron
nicrow,-ph of the e>cipulum. .... ...... 27
Plate IX. :,r-'":,.-, ., ; developmental stage; . . 29
Plate X. rf'.:t.:r-.: c,:./;.ri, sporodochiun and the
conidi . . . . . . . . . .. 45
Plate (I. .'r-:..e ": ,':l. : electron micrographs of
the conidiophores. .. . . . . .... .. 47
Plate XII. T 'rr-,':..:' -r .1 --:, electron micrographs of
the conidiophores. . . . . . . .. 19
Plate XIII. elrn'.:t.:-J .-: Ar, electron micrcgraph: of
the con diophores. . . . . . . .. 51
Plate XIV. :'i..;:. .'.: : I3. .:', different stages in the
formation of dcjble septum . . . . .. 53
Plate XV. Per'it..l a .,*:,'.;; ', diagranmm tic representation
of the conidiogenesis . . . . . . . 55
Plate XVI. I;rnicitr-, ni,, ri, electron micrographs of
haustorial mother cells. . . . . . . 71
Plate XVII. T-r-.,t..ria ,:'dJeri, electron micrographs of
the hiustorial neck. . . . . .. ..... 73
Plate XVIII. Ter" itzrfa n~,:zi' light and electron microgriphs
of haustoria.. . . . . . . . . 75
Plate .IX ;-m:tar.~ :-*' electron micrographs of
the haustoria. . . . . . . . . . 77
Plate y",. 'ertm"ita'ric s',:.'C, electron micrographs of
haustoria showing mini-microtubules. . . .. .79
Plate XXI. crnit.ri.i r, .-i, electron micrographs of
the haustoria showing lysosones . . . .. 81
Plate .XII. ,'!vti CZ-CllZ.a ,'r.e:, lignt microgriphs of
the cross sections of sporccirp . . . ... 93
bL tract of Disser tt on Present-c to t're Graduate Council
of the ULniers itL of Florida in Partial Ful fiT:mert of the FPuiremr.nts
for tr, Degree of oioc.tor of Philocophy
tORPHOLOGY, DEVELOP:I'P:;;T, 'ND LLTF.ATPUCITURE
OF T.:-;:. '._. -~. ": Tha.ter
ChairmT3an: Dr. James U ?.iia.brough
M'jor D[eparrmrtent: Bltany
!r-.- 'i' a '.z--r:. Theater is an entorr phlilous fungus g"o..ing
on the exoskeleton of different species of ter-rites. The fungus
is an ectclparas'to Special :ed thick.-,alled haustorial nm.ther cell;
of the baTal lajyr of the fungal spcrodocnium send raustoria ir-to
the integu,.-int of the r.host. The haLstorial peg enters the noet
through the pore canals of the host cuticle.
The haustoria are lobed, uninucleate, surrounded by' a thick
wall ani Separated from the ho:t protcplast by its plasma membrnr.e.
Alon3 with usual organelles the haustoria contain certain *-acuolei
that are: 1sosorTal in nature, and a new organelle, mini-micrDtubule:
that are only 6 to 100 A in diameter and have ;i>. ub-units. The
mini-TicroLutules are found in tun.jies free in tne general cyto-
plasm. between :he cisternae of smooth endoplasmic reticulun,
between ER cisterna and plazT.a rmenbrane or between ER ci5sterna and
the nuclear envelope.
Tre only known means of reproduction is the formation of
conidia that are cut off endogenou;ly :t the phiilide; at the
conidiogerous loci into long collarettes in basipetal succession.
Phialides are closely clustered into a sporodochium. The conidia
are catenate, cylindrical, uninucleate and hyaline.
It is proposed here that the conidia germinate to form a
crust-like primary thallus that later, by the development of
phialides over it, matures into a sporodochium.
The ta.onomic position of T 7niitaria anj its relationship unth
other imperfect fungi are discussed. Within the fungi imperfecti
a new family and a new order, Termitariaceae and Termitariales
respectively, are proposed to accommodate Teirri .a'ia and ,.'2t;rItl .2'..
A new species of ,al.lttcro2.,a based on the slide of Tenit."::riz made
by Tha:ter is also described.
THE MORFPHOLOGI Ar;N DEvELOPFMEINT
In 1920 Tha.tter described two rnew fungi growing on trhe E .os el.eon
of three species or termites, -- :,t*a :;r-.; ;'L. : .' ; rol ler), ,.
' *; ::, '..5 (E.inls) and .. ,::tr,'T .- ctc-.. (PolaTiren), collected
from Washington, D.C., Sardinia, and Granada rcspectielly. He
proposed a new q tnis w'r~"- .n o ith t.lo spec ies T,J'. :-:,i;ri: and
T'. er-..~- Peichensperger (1921) reported a similar fiun ,us growing
on the e,.oskeleton of ,:; .. ::;t-rr.- -' :r:- (P mt-un), ,.. .. :: ,
(Hagen), C:.- :':-_:.---_ .:, ..- .7 dollarr) ,.--::: ..; ft;:.-T .: i (Fuller)
an v',. z;i'.,; the first three were collected from Brazil and the
others frcrm. the Congo in Africa. He recognized them as I-T:I-.'Ir-
.*tcL r'. Colla (1929) adde3 F. r-, tc .-: 7, '-., : -.' (PR.a tun) and
l'-:, ; s i ;;::;;,: (Silvesteri), both fronr South America, as hosts
of l'-rt c -ic ,Z .-: Tha .ter, and ."z:,:':c r- --r..2 : (MIotschuls' ) ,
A'. ruair,=: (HolrrgrE n) and ,'. :..r.: ::- (Oshi- ) to those of I-:' :.:.'...
co*:'*c.,1 ThaRter. Pickens (1952) mentioned another species ..
p.:c,-:i; :'.3 tut no details were gi'en.
Thaxt:-r considered the furious as an e,.ternil parasite growing
on the surface of the host without any indication of actual penetration
by the parasite through tnr integumrert of the host. He reported that
the fungus did not caus; any irconvieenierc to the insect. However,
Peichensperger (1923) reported defor-mtion of the infected legs and
antennae. Feytaud and Dieu:eide '1927) considered it )n internal
parasite, and showed that the adipose tissue w:as invaded and
altered by the mycelium. The cuticle was penetrated by a root-like
pedicel and in the process the hypodermis .nd cuticle were altered,
cells disappeared near the break, color became abnormal, and large
amoebocytes were found. Heim e. aL. (1951) also thought that this
fungus is an internal parasite which enters the body through
ingestion or licking, invades the fatty tissues, and finally
erupts to the exterior by breaking the exoskeleton of the host.
Thaxter (1920) considered ze -ic. ia a Fungus Imperfectus ano
placed it in family Leptostromaceac, though he noted that its
position there was an isolated one. Heim (1972, personal communication)
thinks this fungus belongs to Tuberculariales of the Fungi Imperfecti.
In and around Gainesville, Florida, the subterranean termites,
Re2tieZeeitez-ie fiavipes and .A. virgainicus have Ter-ritf.-ia sr;yd~r
Thaxter growing on their exasleleton. The rediscovery of Irit-. ri-
has provided an excellent opportunity to study and resolve many
questions about its morphology, development, taxonomy, and mode of
Materials and rMethods
The termites were collected and the studies were done on
naturally growing fungus. Different media here tried but the fungus
could not be cultured. Methods for the studies are described in
detail in the ne.tt two chapters.
The fungus grous on any of the body parts antennae (Fig. 2),
legs (Fig. 5), mandibles, head, thorax (Fig. 1), and abdomen (Fig. 3) -
the most common and most prominent infection be'ng that of the
abdomen. When growing on an antenna the fungus forms a crust that
completely encircles it (Figs. 2, 15). Mature stages were not
encountered on antennae.
Termitaria does not appear to disturb or irritate the host.
Termites with multiple infections have been found to differ little
in their activity from normal ones. HowE'.er. when the legs rnd
antennae are infected certain d formations result. The number of
segments in artenni decrease (Fig. 2) and leg tecones swollen
(Fig. 5). The atdcomen ray al.o swell but it is hard to detect
because of its globose nature.
During the process of collection, to *,ery irinortant ctbser-
vaticn: were made. The fungus co.:uld be collected year rourd, and
only 6-IO of the tme..ber: of a colony i.ere found infected. Earlier
workers h..ve also reported the scarcity of the fungus in infected
colonies (Tha.ter, 1920; eichenrsperger, 1923).
The riature fungus foris a lenticular to hnsterioid sporocarp,
spherical in outline or variously elongated depending upon the
position of its growth. It is closely appressed t. the cuticle of
the termite rad corsists of a ITmainy-l]yered basal pEeudopar'ench.mrtous
suDh.menium from which firmly coherent, simple, parallel conodiophores
arise vertically, forming an even hyTenial surface (Fig. 7). The
whole sporocarp is 70-20 j thick the subhymenium being 11-20 u
and the hyneniuni 50-60 u. The size of the sporocarp ranges from
375 X. 400 1000 u. A thick-walled area divides the h;,enium
into an upper conidium-containing layer and a lower conidium-free
one. The upper zone is 22-25 u thick and the lower 30-25 u.
The peripheral cells are dark and thick walled and form a
well defined, sterile rim or excipulum. The margin in contact with
the substratum is spreading and only one cell thick at the tip
The basal layer is comprised of light-colored cells with dark-
colored, thic-;-walled cells interspersed among them. The structure
and arrangement of these cells can be easily understood by looking
at a younger stage (Figs. 12, 14, 27). They appear pyriform when
viewed from the top, and are often in shiny rows and have a hyeline
spot in the center (Fig. 12). They were considered chlamydospores
by Thaxter (1920), Feytaud and Dieuzeide (1927), and Heim (1952).
This study has shown that the s3-called chlamydospores are actually,
haustorial mother cells (for details please see Chapter III) that
send penetration pegs into the integument through the cuticle.
The hyaline spot in the center in fact marks the position of the
penetration peg. The haustoriun is lobed, uninucleate, the nuclei
lying in the broad base, and has a very thick wall (Figs. 8, 9,
13). The host plasma membrane forms a sheath around the haustorium,
thus at no time is the haustorium in direct contact with the host
cytoplasm. The haustoria could not be traced beyond the basement
membrane, thereby suggesting that the fungus is an ectoparasite.
The haustorial mother cell has been seen both with and without
a nucleus, the forTer situation being rare. It appears as if,after
the haustorirT has reached a certain size, the nucleus of the
haustorial mIc.thrr cell migrates to th7e haustoribm. The haustorial
mother cell is separated from the cell above by a typical ascomycete
septum with Woronin bodies in the vicinity (Fig. 11) and is connected
with the haustorium by a narrow nec! that passes through the cuticle
of the host (Figs. 9, 13).
The subhymenium is comprised of nucleate cells separated from
each other by an ascomycete septum (Figc. 16, 18). These nuclei
are bounded by a double membrane and have a .ell defined nucleolus.
Mitochondria of different shapes ;uith plate-like cristae are
abundant. Endoplasmic reticulum of any ;ind is scarce. Ribosome;
are free in the general cytoplasm. The plasm.aler-ma is elaborated
in plasmalemr.asomes at tar'ious places (Fig. 16). In sore sections
the cells were found full of gltcogcn-l ie bodies (Fig. 13) whilee
in others big globular structure; iere present. inside the sub-
hymenial cells (Fig. 16). The chemical nature of both could not
be ascertained but they appear to be some rind of reserve food
material. The big globular structures are often interconnected and
are full of small light-and dark-c)lored globular bodies. These
also have memb'anous structures inside that run from one to another
globular structure through the connection (Fig. 16).
Closely paced phialides make up the hyrenium. Halfvwa from
the base of the phialide there is a conidiogenous locus where the
wall Is very thick and the conidia are cut off apically into a
collarette (details in Chapter II). The conidiogenous loci of
all the phialides lie at the same distance front the base creating
a distinct thick walled area in the hjneniun (Fig. 7). The
phialides are uninucleate, have riecsores, mitochondria, and rough
endoplasmic reticulum. The nuclei and mitochondria both are very
elongated. Normally four conidia Ean be seen inside a mature
collarette (Fig. 19). The conidia are 3.5-4.5 1.5-2 w, cQ llndrical,
hyaline, thin walled and uninucleale (Figs. 19, 20). The tip of
the phialide is very thick and does not appear to have a pore as
has been reported earlier (Tha.ter, 1920). Discharge of the conidia
was not observed nor was any sporulating sporocarp collected. How-
ever, when a termite with nature famgus was brought into the
laboratory the fungus was found cowved with a white mass of
spores the r.e.t day (Fig. 4). It e.pears as if the whole tip is
blown off the phialide by the force applied by the formation of
new conidia at the conidiogenous lovs.
The cells of the e.cipulum are very thick walled, the outer
surface as thick as 2 L. Though the e.cipular cells in the
hymenium are elongated like conidisquores and are separated from the
underlying cells by typical septa as conidiophores are, they are
sterile (Fig. 22). The cells of the e.,cipulum are uninucleate,
have long mitochondria with plate-like cristae, ribosomres free in
the general cytoplasm, cisternae of endoplasmic reticulum and are
bounded by a tnick wall that is comprised of two layers, the outer
dark and the inner light colored (FiS. 21).
The sporocarp i: traersel by projecting bristles of the host.
The fungus forms a protective filamewous sheath around it (Fig. 10).
The mature sporocarp is a composite structure. The hyphae
appear to emerge at different places and then grow towards each
other (Fig. 61. I think tnat these places dark tne position of
original primsry infe:tion. As far as this study is concerned
these have alr..ys been found lying in a furrow or the inter-
segmental aria. (Figs. 9, 141 where no s leratization occur- and
the cuticle is ver) thin. A11 cells in the basal layer here are
haustorial mother cells sending man) hausroria all around.
Because of its deep lying nature Fe/taud ard Dieuzeide (1927)
called it "pce-icule de l'hi.eriuiT" and Coltl (1929) thought of
it as a foot :o anchor the fungus. Reichensperger (1923) al:o
noted that th: fungus was anchoreJd a..h knob-like intrusions in
the chitinoua matri,.
Because .f the negative response of =-i"'.:1f,_1 .-l..la:. to the
media tried for its culture I had to rely crnpletely on field
collections fcr the study of its mode of infection, growth, and
It is assumed that the conidia, like the spores of .some other
entorogernous fungi, become thick walled and high, pigmented
before germination. The cuticle of the termite proves to be 3
barrier to infection. Only the conidia that have come to lie in
the thin intersegmental area appear to initiate the infection.
Not all of the conidia c n find their way to that spot and this
perhaps is why such a small percentage of termites are found
infected. Peichensperger (192I) has suggested earlier that the
infection could occur at the time of molting when the chitir is
still soft. This is still a possibiiit, an. in certain cases it
may be that the infection does start at that time. But the origin
of primary infection in the intersegmental furrow strongly supports
my assumption. The integument does not seem to be trolen at the
time of infection, suggesting that penetration is accomplished by
some enzyre action. It is already known that certain bacteria and
fungi have an enzyme system that can digest insect cuticle
First, a group of thick walled cells is formed (Figs. 23, 24)
and all of them send haustoria into the host. The hyphae originate
from this mass of cells and diverge in all directions, keepir.ng in
tight lateral contact. The cell at the tip of the hypha is cut off
from the rest by an anticlinal wall, and then divides in only one
plane, periclinal (Fig. 26), thereby resulting in circumferential
as well as radial growth. Thus, the margin of the growing thallus
remains only one cell thick (Fig. 25). During the radial growth
certain cells next to the cuticle of the host modify and become
thick walled haustorial mother cells. It is not clear why one
cell turns into a haustorial mother cell while the next one does
not. There may be certain weak spots in the cuticle and only the
cells opposite them become modified; or perhaps after reaching a
certain size the fungus has to send haustoria; or it may be that
the growth of the fungus is cyclic like that of certain fungi in
agar culture and haustorial mother cells are formed at the beginning
of each new period of growth. Termites, like other arthropods,
have minute ducts extending vertically through the procuticle. These
are called pore canals and one might speculate that haustorial
passage cccurs through them. By observing Fig. 13, v.here a pore
canal and the hiustorial path through the cuticle lie in the same
vicinity, it can be seen that the ha'jstjrial passage appears to be
at the original site of a pore car,n. Thus, it is possible that
only those cells that are opposite pore canals modify, into haustorlal
mother cells. But the dianrater of the pore canals is very small
compared to that of the haustorial path and it array be suggested
that after the initial penetration certain enzymatic activity widers
The young thallus is represented by Figs. 14 and 7. It
forms a circular or variously elongated cruit on the host surface.
It is parenchymatous and a fer cells thick (Fig. 25) depending
upon its age. The margin is spreading, only one cell thick, and
remains that way even after the thallus has matured into a mature
spoiocarp (Figs. 17, 26).
The intermediate stages showing the transformation of this
crust-like primary thallus into a sporocarp have not been obtained,
but it appears likely that after reaching a certain size and
thickness the hyphal branches start to grob, upward, very closely
applied to each other. The tips of these hyphae become thicL
called. These are now the conidiopheres, which ;tart cutting
off phialcconidia apically in collarettes as discussed elsehere
(Chapter II) in detail. The peripheral cells teccire thick walled.
remain sterile and male the eocipulum. Thus, the whole thallus
turns into a Soorocarp.
?:.c-it-Co spp. were considered eo.ternjl parasites by Thaxter
(1920), Reichensperger (1923), Colla (1929) and an internal
parasite by Feytaud an. Dieuzelde (1927) and Heim (1952). Theater
did not see any actual penetration but s;a hypertrophied cells of
the host opposite the dark cells. By looking at his figures it is
apparent that he was actually] looking at the haustoria and he
mistook them for hypertrophied host cells. Reichensperger
actually reported that root-like threads emanate from the basal
layer into the epithelium and either perforate the cells or
encircle them. However, he did not realize the importance of
the dar. cells and did not regard the penetration pegs as haustoria.
It is reported here for the first time that the dark cells are
haustorial mother cells and that the fungus is a parasite and
probably is nourished by the host. The sections of the sporocarps
that were observed by light or electron microscope did not show
the presence of any hyphae in the host, only the haustoria that
were uninucleate and did not have any sept, unlike the sporocarp
above the cuticle. N:o haustoria were seen in underlying adipose
tissue or anywhere beyond the basement membrane. The channels
through which the haustoria go into the integument are ,ery smooth
and do not look as if they have been formed by eruption of the
fungus emerging to sporulate, which should have been the case if
the fungus was an internal p.rasite. RTer i-iria2 nd.ai is an
external fungus that penetrates into the host by haustoria.
The termites with Te;:-i't 'a infection do not look nor behave
abnormally. All the knowledge at hard suggests that the infection
is not lethal. But we also know that the infection of antennae
and legs results in malformation and deformation. PerhapF at a
certain stage of the life cycle the termites are susceptible and
maybe the fungus is lethal. The parasitism of Termitai:..; .-.-1
has been established. The role of this fungus in biological
control for termites awaits more extensive field and laboratory
research. Rearing of termites in the laboratory; ic nece:ar-r to
study the stage it which the termites are moSt suLceptible, howi
much tine it tales for the fungu' to sporulate rrom the tim.- of
the germination of the conidia, whether the fungus lills the
termite; or not, an,3 if it does how it does So.
The geographical distribution is another interesting 3sicct
of the Trnirar:, life cycle. It does not appear to be influenced
by climatic or en ironrental conditions of the area, but corresponds
to the geographical distribution of the hosC anO is such that
almost all the host general ire infected.
The absence of, or our inability to discover, a sejul phase
in the life cycle still leaves the ta.,onomic position of this
unique fungus in dout..
Termites showing the habitat of ? .~'~t.,'L a s:'.der'.
1. Tern7tri'a s:..:lrwi growing on the thorax of the
termite (arrow) .'15. 2. growing on the antenna of
the host (arrow ) .10. 3. growing on the abdomen
of the host (arrow) XIO.
Sporulating X'rrtizria X20. Arrow points to the
white mass of conidia covering the surface of
the whole sporodochium.
Whole mount of terr.ite legs X20. Compare the
infected leg with the uninfected one and notice
the swelling and deformation of the former.
A close-up of the basal layer of the mature
sporocarp '100. Arrows indicate the multiple
points of origin of infection, Raking the thallus
a composite structure.
Figure 7. Cross section of mature sporocarp of ".'r-:,..-i?
ar.'.:dqr showing basal la;er (BL), subhjmenium
(SH), hjTmenium (Hi), e.:ipulum (E), and the
bristle 1() of the termite cuticle passing
through the sporocarp 71,000.
Figure 8. Plastic section of the sporocarp shoiiing the
subhy~.eniui (CH) outside and the haustoria (H)
with distinct nuclei (N) inside the host
Figure 9. Plastic section of the sporocarp at the point of
the origin of infection showing many thick-walled
haustorial mother cells (H1C I and nucleate (N)
haustoria (H) .2,000. One of the haustorial
mother cells is connected with the underlying
Figure 10. Cross section of 3 portion of sporocarp X2,000.
Bristle (B) of the termite cuticle passes through
the subhmenium (SH), hymenium (Hi) and is surrounded
by the sterile filament (arrow points to one such
Figure 11. Haustorial r.:.
peg into the
separated r ,..
a typical as .
(W) in the vi:
Figure 12. Light micro.,r
ell (HMC) sending a penetration
.e cuticle (CLI) \22,500. The
1ll has endoplasmic reticulum
(11), a nucleus (N) and is
cells of the subhymenium by
.e septum with Woronin bodies
- haustorial mother cells with
-.ots in their centers x2,0013,.
Figure 13. A near median section through the neck (n) that
connects the haustorial mother cell (HilC) with
the haustorium .(14,100. The neck appears to
pass through a pore canal, since its passage
through the cuticle (CU) looks similar to the
pore canal (PC). Haustroium has a thick -'all
(W) and contains a nucleus (ri) with a single
nucleolus (NU), mitochondria (i), and endo-
plasnic reticulu (EP.).
Figure 14. Young thallus of Tcr-r-i.:.a growing on an
antennal seryr.it of the termite Y200. It
appears to originate at the base of the
segment ani ro.ws upward in many directions.
Dark haustorial mother cells (HM.C) are very
Figure 15. Cross section of the antenna segment showing
it surrounded by the fungus crust :?50. The
bristles (B) of the termite cuticle are passing
through the young thallus. Basal layer of the
thallus is very dark and prominent.
ITR T' I A
.U4 N L
EtvJ'B'~ q y
( ^ ) -Bi*0
Figure 16. Subhymenial cells of the sporocarp separated from
each other by a typical ascomycete septum having
a central pore (arrow head) and Woronin bodies (W)
.433,000. Plasmalerra is forming plasmalemasomes
(PL). Subhirienial cells have mitochondria (M1)
with plate-like cristae, ribosomes in general
cytoplasm, and globular structures (GS) that have
membranes in them that sometimes run from one
globular structure to another (arrow).
Figure 17. Saggital section through the margin of mature
sporocarp X2,000. Note a single cell at the
Figure 18. Excipular region of tne sporocarp showing
subhymenial layer and the base of the hynenial
excipulumT X15,000. The cells are filled with
glycogen-like bodies (G). Some of the sub-
hmenial cells have vacuoles (V), globular
structures (GS) and one of the cells has a
nucleus (N). One subhvnenial cell is connected
to two sterile hyrenial cells of the excipulum
with ascomycete septa having central pores
(arrows) and Woronin bodies (1).
:' N 18
Figure 19. This section of the hyjr.enum showing conidiophores
(CP) and the excipulum (E) X3,000. The former cut
off conidia (C) with nuclei (N) into the collarettes
(CO) at the conidiogenous loci (CL).
Figure 20. Whole mount of conidia X1,00.
E~-IEI~~ C~I~ ('~~4
Figure 21. Thin section of the e.
sporocarp 037,500. The cell is thick walled,
has a nucleus (N) with a nucleolus (NU) and
Figure 22. Cross section of the e.'cipul3r region of
sporocarp showing thick-walled cells in the
ecipulurn (F) and a portion of the h.renium
' -. ^-'.)--
Figure 23. A section of the very young thallus comprised of
only a few thick-walled cells growing on the host
cuticle (CU) X2,000.
Figure 24. Whole mount of a very young thallus comprised of
thick-walled cells that has just started sending
hyphae around itself X2,000.
Figure 25. Cro'-s section of a young thallus a few cells
thick in height growing on the host cuticle (CU)
and sending haustoria (H) into the integument from
its haustorial mother cells (HIIC) Y2,000.
Spreading margin is only one cell thick (arrow).
Figure 26. Surface view of the growing margin of the young
Figure 27. Whole mount of a young Ter.t .:a thallus viewed
from the top 2~50. Neatly arranged filaments
originate from a group of thick-walled cells
(arrow) in all directions. Dark haustorlal
mother cells (HMC) are arranged in rows.
- fI-, -,
Hughes (1953) emphaSized that "morphologically related imperfect
states can be brought together wrien the precise methods of conidium
origin take first place in the delimitation of the major groupings."
Since then considerable attention has been focused on the structure
of conidiophores and the mode of conidial formation (Tubaki, 1952
and 1963; Sutramanian, 1962). Hughes classified the Hypomycetes
into eight large sections, each based upon different mechanisms of
conidiogenesis. In section IV of his classification he included
those Hypom)retes that have conidiaa developing in rapidly
maturing basipetal series from the apex of a conidiophore which
may or may not possess an evident collarette." Terr-i.ar~-a ,'T.-.ri
belongs in this section. Hughes restricted the term "phialide"
to those unicellular structures which are usually terminal but
sometimes intercalary as well, on simple or branched conidiophores.
They are oval to subcylindrical to flask shaped or subdulate, often
with a well differentiated basal swelling and a narrower distal neck,
with or without a terminal collarette. From the ape* of each
phialide develops a basipetal succession of phialaspores without
an increase in the length of the phialide itself. This section
corresponds to the Phialasporae of Tubaki (1963) and Tuberculariaceae
of Subramanian (1962).
Cole and Fendrick (1l69) studied the phial ides of i'-ialaphora
;,j-r;r,'j:: (f'lel in and IJannf.) Conant, F- i.e: :. .:'d ophilZtm
Dierckx and Th:'..:.::.:p-r.; ,:r,:::. (De Seynes) vro Hohrel by time
lapse photomicrography. In all three they found a fixed endogenous
meristei responsible for conidium formation. The conidiophore
ceases elongiaion once the Teri:tiim become: active and its outer
wall is ruptured at the ape\ by the emerging first conidium or its
initial. In each case a portion of the conijiophore wall rerrains
above the meristeematic one and acts as a collarette through which
the conidiu.i or its initial protrudes. They defined the phialide
as "a sporogenous cell with only one functioning, fixed, endogernous
meristem whose position is marked by the deposition of an inner cr
secondary wall which surrounds each of a basipetal succession of
physiologically independent conidia. Spore production incurs no
concomitant increase or decrease in the length of the iporogenous
cell." At the recent Kananaskis conference on Ta.,onomy of Fungi
Imperfecti (Kendrick, 1971) the phialide was defined as 'a
conidiogenous cell in which at least the first condium initial is
produced within an apical e.,tension of the cell but is liberated
sooner or later by the rupture or dissolution of the upper wall of
the parent cell. Thereafter, from a fixed conidiogenous locus a
basipetal succession of enteroblastic conid a is produced, each
clad in a newly deposited wall :o which the wall of the conidiogenous
cell does not contribute. An, phialide wall distal to the conidiogenous
locus is the collarette. The length of the phialide does not change
during the production of succession of conidia."
Materials and Methods
The subterranean termites Ef:c:icitceez spp. bearing typical
Termitzr.2 lesions were collected from the woods around GainesJille,
Florida. The lesions were removed along with the adjoining
integument, under 2.5% buffered (I:a-Cacodylate pH 7.35) gluteraldehyde
and fixed for four hours at 3VC. The material was washed in buffer
and post-fixed in buffered 1.5. osnium tetro.ide overnight in the
refrigerator. After several rinses in water it was dehydrated in
a graded series of ethanol and in the end washed in reagent grade
acetone. It was infiltrated with graded acetone-plastic mixtures
and finally embedded in 100l plastic. Mollenhauer's (1964) plastic
mixture #2 was used (62 ml Epon 812, 81 ml Araldite 506, 3-4 ml
Dibutyl phthalate; 15 ml of this mixture, 10 ml DDSA and 45 drops
of DIP 30 to make the final plastic mi.ture). For better penetration
the material in the miAture was put on a shaker for eight hours, at
every change during the graded acetone-plastic mixture series.
Bubbes in the plastic were removed in vacuum at 60C. Polymerization
of the plastic was carried out in an oven at 60C for three days.
The material was stained for two hours in 2% UAc in 70; ethanol
during the dehydration and post-stained in 0.5% aqueous UAc for
45 minutes arw with lead citrate for 30 minutes. Some grids
were stained with methanolic UPc as described by Stempak and Ward
The sections were cut on a Porter-Blum MT-2 ultramicrotome
with a diamond knife and examined with a Hitachi-HU 11 E electron
microscope. Some half-micron sections were cut and stained with
NIl (Juniper r ,J:., 1970) for light microscopy. Cryostat sections
of the sporocarp were cut without fixation and stained with aniline
blue in Hoyers. Some whole lesions were mounted in Hoyers and
aniline blue, crushed, and ob-erved under the light microscope.
The conidia are cylindrical 3.5 4.5 u 1 1.5 2 u (Figs.
30, 31, 32, 41). The conidial wall is composed of two distinct
layers, in outer thin electron dense layer and an inner thick electron
transparent layer (Figs. 31, 41). Like other fungal spores the
conidium surface is smooth. Conidia are uninucleate, the nucleus
is bound by a double-imeirtrane, nuclear envelope. Usual or;anclle:
sucn as mitochondria, rough endoplasmic reticulum and ribosoTrie; are
present (Figs. 31, 43, 43). Many lipid droplet; are also present.
In certain conidia the droplets look interconnected (Fig. 43).
Unusual vacuoles have been seen in some conidia (Fig. 31).
The conidiophores are long, cylindrical, vertically parallel
in a sporodochium (Fig. 29). They appear to te fused laterally
with each other, giving a honeycomb appearance when viewed from the
top. The conidiophores are 55 65 p long and 2.0 2.5 u wide,
with a very thick ]nd rounded tip (.ig. 40). Less than half way
down from the tip is the conidiogern s locus (Fig. 28). The part
of the conidiophore is constricted at the locus forming a very
short neck (Figs. 34, 37, 43). The conidiophore wall in the region
of the collarette consists of two layers, an outer more electron
dense, and inner more transparent layer. We could not differentiate
the layers in the rest of the conidiophore wall. At the base of the
collarette there is a pad of wall material between the outer and
inner layers of the wall (Fig. 47). The apparent thickening in the
neck is actually the inner wall la)er of the conidiophore that has
been sectioned obliquely as is evident by the similarity in electron
density. The plasmalemma of the conidiophore is highly convoluted
forming praramral bodies (ilrchant and Robards, 196S) all along the
length of the conidiophore. Sometimes connections can be seen
between the vesicle of paramural body and the plasrralemma (Fig. 35).
The coniiophore is separated from the vegetative cell under-
neath by a septum perforated by a septal pore. There are darkly
staining Woronin bodies near the septum (Fig. 42). The Woronin
bodies are bewnded by a unit-membrane and have a granular matrix.
Plasmalenmasomes h3ve been seen projecting into the protoplast of
the cell underneath the conidiophore (Fig. 39).
Each conidiophore is uninucleate, has mitochondria, endoplasmic
reticulum, ribosomes and vesicles (Figs. 33, 34). The nuclei and
mitochondria are very elongated. Each nucleus is delimited by a
double membrane and has a conspicuous nucleolus at one end. The
nuclei of all the conidiophores are situated halfway from the
conidiogenous locus, and we found all of them in interphase stage
even though in many cases conidium initials were forming at the
loci (Fig. 28).
The conidium initial buds out at the locus into the collarette.
The wall of the conidium initial appears to originate within the
neck of the crnidiophore (Fig. 47). After the conidium initial
has reached a certain size and cell organelles have moved into
it, a delimiting septum starts growing centripetally near the
conidiogenou: locus. The septum has a central electron-transparent
layer between two electron dense ones (Fig. 47). At the rim, of
the growing septum an electron-dence material is present (Figs. 45,
46). The septumT appear: to become functionally complete by
continued cein.ripecal growth (Figs. 44-43). Most of the septum
grows at the conidiogenous locus except for a central pore that
remains open untill the ne.w conidiun directly underneath is fully
grown (Figs. 36, 38). The thickness: of the conidium wall does not
increase noticeably during its forration. However, after it has
become delimited by the septum an inner electron-light wall layer
is deposited Shat grows in thickness as the conidium matures (Fig.
41). The young conidium is connected with the newi initial through
a central pore in the septum (Figs. 36, 33, 41). After the septum
is complete the electron-light layer of the septum appears to
break down, allowing the two halves of the septum to separate and
the conidiuim to secede (Figs. 47, 48). In rare cases Ve have seen
Woronin bodies near the growing septum (Figs. 36, 46, 47). The
shape of the young conidium is determined by the tubular collarette
and the constricted conidiogenous locus. The young conidium is
typically narrow at both ends (Fig. 43) because the septum at
both ends originates at the constricted conidiogenous locus.
From the above observations we reach the following conclusions:
1. The conidia are produced in basipetal succession from a
fixed coniditgenous locus.
2. The conidia are clad in an entirely new wall not derived
from any existing layers of the conidiophore wall.
3. There is no increase in the length of the active cor.diophore
i.e. from locus to the base of the conidiophore.
I am aware of only three reports of the type of Phialoconidio-
genesis where the conidia are produced in a tubular collarette. Tuo
of them are of species of T;:hciaiopfi. Cole and Yendrick (1969)
have studied phialoconidium ontogeny of ". p:,.ariz.?2 (De Seynes)
von Hohnel by time lapse photomicrography, and Del Vecchio .a .Z.
(1969) of r. bsicola by light and electron microscopy. The
conidiogenesis of an Indian isolate of 7. paradc was studied by
Seshadri as reported by Subramanian (1971).
According to Cole and Yendrick (1969) the phialide of T. a'~. :'
is long and cylindrical. One to several conidia differentiate
basipetally within the conidiogenous cell. During early stages of
differentiation growth of the outer wall of the conidiogenous
cell and the conversion of the protoplast into conidia proceeds
simuTtaneously. The center of activity (conidiogenous locus)
continues to move downward until the conidiogenous locus is fixed.
The phialide stops growing and the continued production of the
conidia at che fixed locus e.erts a pressure at the tip of
phialide resulting in the rupture of the vall thereby releasing
the conidia. The cylindrical portion of the phialide beyond the
locus is a collarette. The locus is not marked by any constriction
or other morphological evidence of the phialide wall. The
conidiophore does not elongate after its apex has ruptured.
Seshadri's (S4iAramanian, 1971) work confirmed Cole and Kendrick's
(1969) findings. Moreover, Subramanian (1971i reported the conidia
did not appear to be delimited by a process of septation or double
septation in the Indian isolate of T. paradoxa.
Del Vecchio et at. (1969) described phialides of T. basicola
enclosed inside "hyphal tubes." Tney contained endoconidia and
were composed of cells which were always devoid of cytoplasmic
components. The walls of these cells were never attache. to thE
walls of the endoconidia an, were identical to the cell walls of
the vegetative hyphae. We think that they risinterpreted the
tubular collarette as a hyphal tube. It becomes clear by looking
at their Fig. #1 Plate 1 that the conidiogenous locus is deep seated,
is not marked by any constriction of the phialide -.all, and the
conidia are released by the rupture of the phialide tip.
Subramanian (1971), by looking at tte reports of Cole and
Kendrick (1969), Del Vecchio t a:. (1969), and an Indian isolate
of the r. pmr.~I:& z, suggested that the events of conidiogenesi: are
as follows. The first step in the development of each conidium
appears to be the protoplasmic cleavage at the conidiogenous locus
within the phialide, followed by development of a totally new wall
around the cleaved mass. This process goes on in production of
several conidia in basipetal succession.
Our findings agree with those of the above in the presence of
tubular collarettes, the wall of the conidia being separate irom
that of collarettes, the phialide wall not contributing to the wall
af tne conidia, there being no increase in the length of the phialide
after its apex has ruptured, and that the conidia are produce- in
As opposed to the findings of others, the conidiogenous locus
in the remnnirza ,i:eri phialide is marked by a constriction in its
wall. It appear; as if after reaching a certain stage the conidium
initial buds out through the thick walled tip of the phialide but
the phialide wall is not broken (Fig. 49 a and b). Thus, a
conidiogenous locus is established, marked by a constriction. The
conidia are produced in basipetal succession from this locus (Figs.
41, 49 c-e). The wall of the conidium appeared to originate at
the locus as has also been reported for f,'l.vrpor.i cracc by Lowry
et a1. (1967), for 'ercic: :iin a io-2:rwn by Buckley et .;:. (1969),
for PF-ni:.lin c-jitfci"C by Zachariah and Fitz-Janes (1967) and
for tit.zf'lhiZcn :;deplia. by Harmill (1972a). The growth of the
wall of the phialide and the formation of conidia proceed simultaneously
during early stages. However, we presume that after a while the
phialide stops growing and the pressure thus exerted at its tip
results in the rupture, releasing the conidia. After that, no
increase in the length of the phialide takes place. We were unable
to see the formation of the first conidium initial as well as
rupture of the phialide, because the fungus does not grow in
culture and we could not collect it in those two stages of develo-
ment. However, Thaxter (1920) has reported that the conidia are
released by the rupture of the phialide tip.
The growth of phialides and formation of conidia looks quite
synchronized as is apparent in Fig. 28, where the conidiogenous
loci of all the phialides appear to be formed at more or less the
same level, resulting in a clear cut zone in the sporodochium. All
the phialide: are of the same helgL.r and mo:t of them have four
conidii. This type of synchronous deseloprTent of adjacent phialides
has also been reported by Trinci1 :: c. (19.S) for 4zi :;.- .;z:i...
The rmde of conidial deli.-.itation is also different from that
of ". par'.:= 1a and T. -az:-'.--:;. It is by double septation. The
septum grows centripetally and for quite a while has a central pore
which may or ray not have Wloronin bodies in there vicinity The
septal pore closes by centripetal growth. We have not seen Woronin
plug: in ? ri.:. f lire those reported by Cole and Aldrich (1971)
for 4:;: K.n.ic;r.' te.."',c':f: (Sac:.) Bain. However, the
septun is similar to the septum reported for b. .i;.:..-i.:.'z by
Cole and Aldrich in thit it has an electron transparent layer
sandwiched between two electron dense layers (Fig. 47). Hanmrill
(1972b) reported a similar type of septum in 0 ::.:-'.m-~c-. ,-:. but
as the conidiogenesis continued the electron transparent layer
underwent a transition and became mere electron dense than the other
two. Buckley er *a. (1969, Fig. 12) also show the same construction
of the basal conidiuTm septumT of i'-ri-d:- L:in atc.-.:r., an electron
transparent layer between tio electron dense ones.
The dividing septum undergoes lysis along its aris as has also
been reported for budding of yeast rF--.;::lr;.c: 2:;.':'z by Marchant
-t al. (1967a.Fig. 21).
Thus, phialoconidiogene-is of .'- rn:'r-.2 z:J,-;'. Thaster differs
from that of 7;:'ic.',- .:s p.-tccr~ an.d ",Z;:oZ.U in the following
very important details: (1) the conidiogenous locus is fi>ed in the
beginning; (2) it is marked by a constriction in the phiallde; and
(3) the conidia are delimited by a double septum.
Fletcher (1971) reported that the conidial chains of three
different species of FEr.cilic: F. aiiger:,.. Damelius, P. c c:, w.?
Bainier and P. cn.Fr~crji n Westling, were enclosed within an
electron opaque surface layer that appeared continuous with the
surface layer of the phialide wall after gluteraldehyde and Osj4
fixation. He did not say whether he considered it as a collarette
Growth of Conidiophores
Several schemes have bee:i proposed to explain the phenomenon of
apical growth. According to Eracker (1967), Grove et a_. (1970a and
1970b), fkClure e: a'. (1968), cytoplasmic vesicles are present to
the exclusion of other organelles in the very tip of the hyphae.
Grove et al. (1970b) reported that the single cisternae Golgi play
an important part in the vesicle formation. The vesicles are forced
posteriorly, migrate to the apex, fuse with the plasma membrane, and
liberate their contents as part of the process of growth. However,
Marchant e: al. (1967a) found two different vesicular systems
associated with w3ll synthesis in regions of active growth and
wall synthesis. In the apical region of the hyphae vesicles pro-
duced by the endoplasmic reticulum moved to the plasma membrane,
fused with it, and were responsible for the primary wall formation.
In the older region of the hyphae multivesicular bodies were found
that fused with the plasma membrane and gave rise to lomasomes and
were apparently associated with secondary wall syntnesis. It was
questiornci whether all the loadsoi;e-1l e bodies were honologoI: in
structure and function. Therefore, Marchar.t -;. ,:. (19 ?) proposed
later that all the memibrarous or vesicular structures associated with
the plasma nerbrane be called paraitral bodies, that the uord
loadsore be limited to the structures deri ed from cytoplasnic
vehicles or mulctiesicular bodies, and that all structures derived
from the plasMaleama be called plas..aleiT-rasom.-. They also suggested
that lo,.isomes might be involved in the incorporation of wall pre-
cursors by transport across the plasmale.iia in th-e same ,ay as the
single vesicles originating from ER and Golgi do.
The parai.-ral bodies have also been regarded as artifacts of
fixation and their existence in live iiaterial and their role in
wall synthesis has beer questioned. Recently, evidence that the
paraTural bodies may not be artifacts comes from wor.k of different
people. Griffiths (1970) found theA in frozen etched replicas of
hyphae of rve:cici::s.! ,.:i. Kleb. Heath ,t S:. (1970) demonstrated
their presence by using three types of fiations in the wall of the
growing hyphae of :pr.-::ic fe (ra.-r (Gruithuisen) Thuret and in the
walls of primary spores and their exit papillae in Di.accyic-.'
:-.ri>z Coker. They also used SITS stain which is claimed to bind
to specific sites in plasmalem i After incubation in SITS stain
a strong pale blue flloresence was detected in all hyphae, most
intense at their apices, sometimes occurring in diffuse patches,
the patches rmarking the sites of planmalem-maomes. They also
suggested that the plasrralemrasomes are produced r.hen more plasmalerima
is produced than is needed to line tre cell wall and the plas-,alemmasomes
may become sequestered in tne jevelcsing wall.
The paramural bodies have been reported in iLrrti-;;::;.X a:;.-at;:,.
at the double septum by EucFley 6t a:. (1969), and in 4srEer'":l:,
n:., i ans in the subapical and basal cytoplasn of growing sterigmata
and near their developing septa by Oliver (1972). Multivesicular
bodies have been seen in .'"e:-rr;:=tz:. ,c:'icli.ae by Harinill (1972a)
at both sides of growing delimiting septa whEn wall material was
being deposited, which suggests that they may have played a role in
The cytoplasmic vesicles have also been reported associated with
conidiogenesis. Trinci t aL. (1968) reported the presence of
vesicles in some sections inside phialide primordia. At the time
of bud formation in Rhodc .r.lj gl:,:;'.-, Marchant ae a (196bi)
reported vesicles in the bud at the time of active Y.all synthesis.
Carroll (1972) found fibrillar material and masses of gray and
speckled substances associated with plasmalemiasomes at particular
stages of conidiophore and conidiun development and near growing
septa of te Ip..h2,:L: kL.rlryca '. She supported the conclusions of
Marchant ot :;. (1968) that plasmsilermasomes nay be involved in
secondary transformation of wall materials.
In Territrr-. An.e!" ThaYter phialides the paramural bodies
are present all along the plasmalenma. However, here the tip of
the growing conidium initial is filled with the vesicles (Fig. 10,
16, 18, 19). The vesicles are seen only in the vicinity of tne
growing delimiting septum and only at that stage. After the septum
is compelte in the old conidium initial they are not seer.. Their
location and timing suggest that they may be involved in the synthesis
of the septum.
I agree with Heath and Cre.neood (1970) that piasmalemmasomes
are produced when more plasmalemma is formed than is needed to line
the cell wall. I suggest that at :t.e time of conidium production
the convoluted plasmalemma stretches a.nd covers the initial and
helps in the formation of it: first wall. I have not seen any
fibrillar or gray or specked -ub:tances associated with the
plasma emma some.
Figure 23. Cross section of a portion of the mature sporocarp
stained with AMB (2,000. It shows the haustoria (H)
beneath the cuticle (CU) and the subhymenium and
hyrrenium above it. The basal layer has many thick
walled haustorial mother cells (HMC). The conidio-
phorez (CP) have elongated nuclei (II) and prominent
conidiogenous loci (CL) all of which lie at the same
distance from the base of the hyenium making a very
distinct line across the hymenium.
Figure 29. A mature sporocarp on the dorsal side of the abdomen
of a termite (arrow) (20.
Figure 30. Whole mcunt of conidia under phase contrast Y2,000.
Lipid droplets are distinctly visible.
Figure 31. Thin section of a conidium inside the collarette
showing outer (OL) and inner (IL) layers of
conidial wall and lipid droplets (L), vacuole
(V) and nucleus inside the conidium X28,000.
Figure 32. Whole mount of conidia under bright field Q2,000.
.-. ..,-+ -. I-- L ,"
Figure 33. Thin section through the conidiophores /15,000.
Elongated nuclei (0) with nuclear envelope (NE)
and nucleolus (;U) are present in the conidiophores.
Surface view of the plas.malerma of two conodiophores
shows plasmalemmasises (PL).
Figure 34. Thin section through the apical region of the
conidiophores X15,0GO. Conidiogenous loci (CL),
elongated ait3chondria (M) and rough endoplasmic
reticulum (ER) are evident.
Figure 35. Plasma membrane (PM) of the conidiophore with
plasmalemmasoTes (PL) X88,000. Arrous point to
the connections between plasma membrane and the
plasmalemmasomes. Cisterna of rough endoplasmic
reticulum (ER) is lying close by.
., .: .. 1
-'" ,. t
Jr. ^ *
Figure 36. Thin median section through the conidiogenous loci
showing budding and growing conidia 715,000. Arro,
points to the connection between the budding and
maturing conidiun. lMitochondria (M), lipids (L)
and a Woronin bod. (W) are also there.
Figure 37. Thin section through the tip of the conidiophore
filled with vesicles (VS) X62,000.
Figure 38. Thin section of young and maturing conidia inside
the collarette X40,000. Arrow points to the
connection between the two conidia. ;ote that
the conidial wall is separate from the w~ll of
the collarette. Lipids (L) are present inside
Figure 39. Thin section through the septum between the conidio-
phore and the cell underneath X52,000. A big
plasmalemrasone (PL) connected with the plasma
membrane is projecting into the cell cytoplasm
and a Woronin body is present.
Figure 40. A near median section through the tip of the
Figure 41. A near median section through the conidiogenous
locus (CL) and the collarette (CO) having conidia
in different stages of maturity Y3,500. The
conidla are uninucleate (N).
Figure 42. A thin section of the septum at the conidiophore
base .'.94,000. Arrow points to the unit membrane
around the ol'ronin body (W). Plasma membrane (PM)
is continuous around the edge of the septum towards
Figure 43. Thin section through a youngest conidium of the
chain ,'24,500. The conidium has a nucleus (I0)
bounded by a nuclear envelope (NiE), rough endo-
plasmic reticulum (ER) mitochondria (M) and
lipid droplets (L) that are interconnected. Apex
of the conidiophore is filled with vesicles (VS).
A Woronin body (W) inside the conidium is also
present. Note that the young conidium tapers
toward both ends.
Approximately median sections of the conidio-
phores and the collarettes representing
different stages in the formation of double
septum and secession of conidia.
44. Young septum originating at the conidio-
genous locus \28,000. Notice a mitochon.jrion (1)
between conidium and the conidiophore. Lipid
droplets (L) and rough endoplasmic reticulum are
present in the developing conidium. 45. Later
stage in septal growth X30,000. The tip of the
conidiophore is filled with vesicles (VS). Arrow
points to the electron dense material. 46. Slightly
later stage than the previous one in the centri-
petal growth of the septum X52,S00. Arrows point
to the electron dense material at the rim of the
growing septum. Vesicles (VS), Woronin bodies (l)
and lipid droplets (L) are also there. 47. Fully
developed double septum with a central pore Y40,000.
Black arrow points to the electron transparent layer
of the septum while the white one indicates the
point of conidium wall formation. A granular
material is present between the inner (il) and
outer (ol) layers of collarette wall. Lipid
droplets (L), Woronin bodies (W), and rough endo-
plasmic reticulum (ER) are also there. 48. Conidial
secession X48,000. The arrow points to the place
of secession-the electron transparent layer of the
Figure 49. Diagrammatic interpretation oi seq'uerce of conidio-
genesis. a. conidiophore tip. b. budding of first
conidium and forr.ation of conidiogenous locus tCL)
and the collarette (CO). c. sop:am initiation at
the base of the conidiuri. budJing of second
conidium, first septum is complete e-ccpt for a
central pore. e. first conidiuri is seceded irom
the one underneath, a new.. inrer layer (IL) is
being deposited inside th". outer one (OL) of the
conidium will, a third conidiuja is starting to
bud out at the conidiogencus locus (CL).
L~___ -- 1
Studies about fungal haustoria inside the insect hosts are
scarce. Richards' work about Herpnr-gee infecting cockroaches is
on the light microscope level. However in recent years a number of
workers have published about the haustorial structure inside plant
hosts (Hawker, 1964; Bracker, 1967; Ehrlich and Ehrlich, 1971). A
generalized concept of the host parasite interface, with a few
exceptions, has emerged. A haustorial mother cell, in most cases
thick walled, is present. The haustorium and the haustorial mother
cell are connected by a narrow neck. The fungal wall is continuous
around the neck, an exception being Albu:qc ndti~ haustoria which
lack part of the wall around the distal end of the neck (Berlin and
Bowen, 1964). Haustoria are uninucleate in all cases except for
Albqg orid, FP t.:pk 'r,; ir.festwias, P. pararit:ica, P:F;:ersv.cpra
cuberasi (Ehrlich and Ehrlich, 1971). The haustoria are surrounded
by a sheath (Encapsulation or Zone of Apposition) and are separated
from the hcst protoplast by a membrane that is continuous with the
host plasma membrane.
In this study the haustoria of the entomogenous imperfect
fungus Te-rnt.;ria e :.dri Thaxter infecting the subterranean
termites Re-Kec itr-ea spp. were investigated. The fungus forms
lenticular to hysterioid sporocarps on the exoskeleton of the host
irre:picLvi'e of thi body position and is found closely appressed to
the cuticle. Tne basal laer of the ;porocarp is comprised of thin-
wa;lcd cells intersperred uith thicker-balled, bigger, darK-colored
cells. These cells act as haustorial mother cells.
la trials and 'etrods
Tre subterranean termites ,c:,i;.'t: ,rr ..: spp. bearing typical
ri:..7arI' lesions were collected from t e woods around Gainesville,
Florida. The lesions were removed along with the adjoining integuiTent
under 2.51 btiffred (Na-Cacodylate pH 7.35) glutaraldehyde and fi.ned
for I hours zt 4VC. The material was washed in buffer and post-fi.,ed
in buffered T.5, osmium tetroxide overnight in the refrigerator.
Afrtr sc.vera'. rinses in water it was dehydrated in a graded series
of athnrils and in the end washed in reagent grade acetone. It was
ir.fillrated H itch graded acetone-plastic mir.itures and finally, embedded
in 100. plastic. Mollenhauer's (1964) plastic mixture 42 (62 ml
Epon S12, 81 ml Araldite 506, 3-4 ml Dibutyl phthilate; 15 ml of
this mi.ture, 10 ml DDSA and 45 drops of DMP 30 make the final
plastic ni.'ture) was used. For better penetration the material was
put on a shaker for eight hours at every change during the graded
acetone-plastic mixture series. Bubbles in the plastic were reToved
in vacuum at 50"C and the plastic pol)ynerized at 60'C for three days.
The material was stained for two hours in 21 uranyl acetate in
701 ethanol at room temperature during the dehydration and post-
stained irn C.5 aqueous uranyl acetate for 45 minutes and with lead
nitrate for 30 minutes, on the grids. Some grids were stained with
methanolic ursanyl acetate as described by Stempak and Ward (1964).
The sections ..ere cut on a Porter-Blum MT-2 ultramicrotome with
a diamond knife and examined nith a Hitachi-HU 11E electron microscope.
To determine acid phosphatase activity the rcr- t:p.2 lesions
were fixed in glutarzldehdde as described above. After fixation the
material was rinsed in Nra-Cacodylate buffer (7.35 pH) four times,
15 minutes each. Then it was incubated in a modified Gomori's
Medium (Barka and Anderson, 1962) at room temperature for 30
minutes at pH 5. After incubation it was washed in Na-Cacodylate
buffer at pH 7.35. It was later post-fixed in 2% osmium tetro.ide
buffered at 7.35 overnight at 4C. After that it was dehydrated in
a graded ethanol series and embedded in an Epon-Araldite mixture as
described earlier. The sections wre cut, stained and observed in
similar manner as for general haustorial study. Control experiments
included incubation at pH 7.0, incubation without 6-glycerophosphate
and incubation without lead nitrate.
Periodic acid silver stain (PAS) (Martino and Zamboni, 1967)
was used to test for the presence of polysiccharide. Four different
aqueous solutions were made: 2% periodic acid, 3% examine, 5t
silver nitrate, 5' borax. The staining solution was prepared
shortly before use by mining 23 ml of hetamine, 25 ml of AgNrO3 and
4 ml of borax. The solution was centrifuged at 2,000 rpm for
30 minutes. Grids containing sections were transferred to the sur-
face of periodic acid solution for 15 minutes at room temperature.
Later they were rinsed twice nith water and were floated over the
staining solution at 60C for 30 mites. After staining, the
grids were floated individually upon hypo for 30 seconds and
later rinsed in water. orne grid: containing sections were stained
without prior oxidation by periodic acid.
The haustorial mother cell is very thick walled (Fig. 50, 52,
53). The cell wall is distinctly divided into two zones the outer
dark colored and the inner light colored. The inner light-colo'ed
zone may be comprised of many layers. The haustorial mother cell
is anucleete, has mitochondria, lipid bodies, v3cuoles, Cisternae
of smooth and rough ER, ribosomes and microtodies. Ver) rarely some
of the haustorial mother cells are uninucleate (Fig. 50). The
nucleate haustorial mother cells were not found connected to the
haustoria and it is difficult to say whether the haustoria belong-
ing to then were nucleate or not, because serial sectioning could
not be done, since sections of this material tear very easily.
The mother cell is separated fr.jm the adjacent fungal cell by a
typical ascomycete septum having Woronin bodies in the vicinity.
The mother cell is connected to the haustorium underneath the cuticle
through a neck and its inner wvall ler is continuous around the
neck to the haustroium (Fig. 52, 53). Mitocondria, unique 80 A
microtubules, normal microtubules, and ribosomes were seen in the
neck region. The hole in the cuticle through which the neck runs
is very srooth. Often at the base of the haustroium there 1i a
collar of some granualr material (Fig. 52) in which no membranous
structure could be resolved. The electron density of the collar
is the saze as that of the innermost layer of endocuticle.
The haustorium is a branched structure having its only nucleus
in the base. The nucleus has a nucleolus and is surrounded by a
membrane envelope (Fig. 53, 54, 55). Certain me-bbranous structures
are present in the nucleoplasm.
The wall of the haustoriun is two layered with the outer being
darker than the inner layer and being PAS positive (Fig. 53, 62).
The inner layer of the wall is continuous with the inner la er of
the haustorial mother cell wall.
The haustoria have not been found in direct contact with the
host protoplast but are separated from the host protoplast by a
membrane that can be traced to the host plasmale.1ma.
The haustoria displace the epidermal cells and make room for
themselves (Fig. 54), thereby resulting in the localized swelling
of the body part the fungus is growing on. The fungus evidently
does not penetrate beyond the basement membrane of the host
The older haustoria appear to be hanging in the space they
created for themselves, and are seen in close contact bith the
host protoplast only apically (Fig. 54). However, the younger
haustoria may be surrounded by the host plasma membrane. In older
haustoria the space between the haustorium wall and the membrane
separating it from the host is filled with fibrous material (Fig. 56).
This material is PAS negative (Fig. 62) suggesting that it is not a
polysaccharide. I tried pronase digestion but due to certain
technical problems could not reach a conclusion. However, I
think that it is proteinaceous in nature. In the young haustoria
either there is no such space (Fig. 59) or a very small one. The
host cells around the haustoria haie the usual organelles:
mitochondria, ritocones, eni jla;mic reticulum, nuclei, microtubules.
The most common organelles closest to the haustoria are the micro-
The haustoria have Titcchondria, microtubules, microbodies,
endoplasmic reticulum, lysosomes, vesicles, multivesicular bodies,
and nuclei. The haustorial Flascaler.ma may form plasmalemmasomes
projecting into the cytoplasm of the haustorium (Fig. 56).
Autophagic vacuoles have been seen with ER, mitochondria
(Fig. 72), and lipid (Fig. 71) inside. The autophagic vacuoles or
lysosomes are surrounded by a unit membrane, and are filled *rith
debris and vesicles.
Both smooth and rough endopl]:m'c reticulum are present.
Smooth endoplasmic rEticului is stacked and its citccrnae are
continuous with those of rough Endoplismic reticulumT. Stacks Of
smooth ER have been observed completely failing the center of the
haustorium. Between the cisternae of the sticied smooth erdoplasmic
reticulum very thin microtubules only 5O-100 A in diameter are
present (Fig. 66, 67). From nom on these microcubules will be
referred to as mini-microtubules, because of their smaller
diameter, as distinct from the normal microtubules. Mini-micro-
tubules appear to be thrown out to the periphery of the ER stacks
(Fig. 64). In many cases mini-microtubules have been seen in
bundles with or without ER associated with them (Fig. 65, 68).
They have also teen found tetween ER cisternae and the nuclear
envelope (Fig. 61), and smooth EP and plasmalemrira of the haustorium
(Fig. 60). Single mini-microtubules have not been seen. At higher
magnification these tubules appear to have subunits, most probably
six (Fig. 63, 69). Use of Markham et a!. (1963) image reinforcement
technique gives the strongest reinforcement at n = 6, n being the
number of equal arcs of a complete circle used in making multiple
exposures of the image. The mini-microtubules have been seen
running the entire length of the stacked smooth ER cisternae. They
appear to be a different organelle than the microtubules with 13
subunirs. The stacked smooth ER cuts off vesicles at its periphery
(Fig. 67). Acid phosphatase activity has been noticed over the
smooth ER stacks (Fig. 70).
The haustoria are packed with mitochondria which have plate-
like cristae. Some of the mitochondria have a peculiar shape, very
narro, in the middle a; if they are undergoing division (Fig. 53,
57). Some very long mitochondria have been observed (Fig. 55).
Microbodies have been seen in the vicinity of mitochondria (Fig. 63).
During the study an atypical haustorium was seen. This
haustorial mother cell had a large amount of rough endoplasmic
reticulum. Towards the periphery of the rough ER stacks, closed
spheres of rough E. appeared, with ribosomes inside (Fig. 51).
Between the vacuole and the ER, mitochondria could be seen. The
cytoplasm looked very diffuse and mitochondria stained differently
(Fig. 58). The plasmalemma of the haustorium appeared to be
invaginating at many places. The invaginations contained vesicles.
Free vesicles and multivesicular bodies were also present in the
cytoplasm. Thus, the haustorium had all the organelles that other
normal haustoria had except mini-microtubule;. This haustorium was
separated from the host protoplast by the host plasma membrane.
At tne si.- .
in the in:er-: -
is very tnir.,
The haustor.:-> r .:
substance : te=..:.
Two cntcr:, -
the body of ti.-
and Smith, Ir :- :
f.2,w;;" : :.e.,'" . :
that the per.t, :
case here. Th
is very sr.C.o o .
laminate of th, r.--
not penetrated t.
From theth a
of the ;r-" :._ .
general fungr l F.:,
1967, 1963; Coff-.
1963b, 1966, 1V-i.
Peyton and Ecucrln.
cell that is c.r',-
iginjl infection, which normally lies deep
area of the host exoskeleton where the cuticle
inter of haustoria are ;ent by the mother cells.
rarely visible. There is a dark staining
:ises of adjacent haustoria (Fig. 54).
funi have been sho'in sending haustoria into
.-, .,F-rr c- S pp. on cockroaches (Richard;
.'-:ctr-:.J.2 r C...'t-.IoC:?arZ on lesser housefly
isler, 1968). Pichards and Smith suggested
Sthe tost in the case of &r "-.. ..: 7,_,, ; is
tic activity. The same appears to be the
through h ihich the neck of the haustorium runs
an and there appears to be no distortion of
cuticle. This suggests that the cuticle was
--.rvatioi s it becomes clear that the haustoria
'-. Thaxter are not very different from the
l-a (Berlin and Bowen, 1964; Bracker, 1964,
.i:., 1912; Ehrlich and Ehrlich, 1962, 1963a,
S* ~~outh, 1956; Hardwick .- .22. 1971 ; Hawker,
Eracker, 1970, 1972; McKeen c. .2.., 1966;
. ). There is a thick walled haustorial mother
. ith the Rain body of the hauStorium through
a neck. The hau:.: r. is packed with nitochondria and endoplasmic
reticulumT, sg:: high rate of physiological activity. No
direct contact t-r. .;i the host protcplast and the h.austorium
exists. There is always a membrane that separates the two. This
membrane is continuous with the host plasma membrane.
However, in finer details the haustoria here are Yery different
from all others so far reported. The major difference is due to tile
structure of the host. The cells of termite epidermis, like other
animal cells, do not have rigid cell walls. The haustorial mother
cell is not in contact with the host cell at all. The neck in this
case passes through the integument of the termite and not the host
cell wall. The haustorium does not hang in the cytoplasm of many
epidermal cells. Thus the sheath around the haustorium is hounded
by the plasmalemma of as many cells as come in contact with it. As
reported above, in sone cases the host plasia membrane is very close
to the hausbrium while in others a space exists between the two and
in still others no contact between the two can be seen. The latter
condition is common in older haustoria. It appears as if with age
the haustoria push the epidermal cells aside and make room for
themselves. A similar situation has been reported for Kr.r-5yzs
spp., using the light microscope, where the basement membrane of the
integument has been seen bulging out into the body cavity (Richards
and Smith, 1956).
There fs some controversy and confusion concerning the nomen-
clature for structures associated with haustoria. I do not want
to add to this; therefore, I am not proposing any new terms. I
use the term haustorial sheath as was suggested by Etacker (1967)
and is presently being used by Bracker (1963), Littlefield and
Bracker (197G, 1972), and Coffey ct .:. (1972). However, I would
like to re-emphasize that haustoria of ...: "Lna :;;..:.: are not
sheathed like other haustcria and that unlike others (Littlefield
and Bracker, 1970) the neck is not Dsunded by an extrahaustorial
membrane. The collar is occasionally present at the base of the
haustorium but unlike other hau:toria does not surround the neck.
A dark-staining ring of wall material occurring miidway
between proxinal and dirtal endr of the haustorial neck in sections
stained with uranyl acetate and po'-titained with lead citrate has
been reported for sc.eral rust fungi (Ehrlich and Ehrlich, 171;
Hardwick et al., 1971, Cc.ffe)' .:2., 1972; Littlefield and er.cker,
1972). It has been suggested that the neck ring repre-int an
abrupt transition from the wall of the penetration pe. to the wall
of the haustorium (Littlefield .and Bracker, 1972). I .aT unable
to see any such band in the wall of the hauzcorial neck.
The wall of the hustorium is continuous all around without
any interruptions. Channels e. tendinm through the haastorial wall
providing physical continuity between the haustorial protoplast and
the boundary of the sheath as observed by Ehrlich and Ehrlich in
.Pajccirz-i n dr.r.:i z trI-t:- (1963, 1971) have not been seen.
The haustoria of Zriu--:. a n.; -2~ ': are unique in having mini-
microtubules. Similar structures have not been observed in fungi or
in any other group of animals or plants. Ilewcomb (1969) has recently
reviewed the literature concerning plant microtubules. Ordinary
wicrotubules are 180-300 A in diameter. They have an electron
lucent core about 100 A in diameter b.-unded by an electron opaque
carte.x or 3i11 about 70 A chick. They are separated by a space
of about 200 A or more from Each other suggesting that each micro-
tubule may be surrounded by a specialized zone. They appear to be
rather rigid unbranched structures usually following a straight
path. The wall of plant iTiicrotubules consists of thirteen fila-
mentous subunits (Ledbe'ter and Porter, 1964).
Steer and Newcomb (169) h, e reported some 290 A and 560-660 A
tubules in bean leaf glands. They appear to be different from
typical microtubules cheni;cailly. The smaller tubules have been
observed connected to the endoplasmic reticulum and they first
appeared in the perinucl-ar c3:oplasm between the layers of endo-
plasmic reticulun. llini-nicrotubules bear closest resemblance to
the tubules of P-protein bodies found in the phloem of the higher
plants. Cronshaw and Esau (1967, 1968) observed tubules in the
P-protein bodies of the sieve elements of uc.arbI;'ta nanmna and
Nicotiza, abac:,.-. They called the tubular protein P1-protein.
The tubules were 231 A in diameter in ,. tc:5bice and 242 A in
C. raUxir.. They also found fibrillar protein in the P-protein
bodies that replaced the tubular protein and assumed that the
tubular form becomes reorganized into the fibrillar form. The
fibrillar form of the protein t.as designated as P2-protein. The
Pl-protein tubules appear to have sub-units and have a central
non-staining core. Parthasar.thy and Mijhlethaler (1969) reported
that in N. crba..c:e the protein tubule wall consists of 6 nearly
spherical units of 0O-70 A. The mini-microtubules are much smaller
in diameter; only 80-100 A across. They also have a central
nonstaining core and peripheral subunits. The number of subunits
is w it probably si,. Outside the haustorium in the snea:h of older
haustoria a fibrillar material resembling the P2-protein or tr:-
phloem is present. The mini-microtJtules are most commonly foaind
between the tubular cisternae of stacked smooth endopla:ici: reticulum.
They have, however, been seen in bundles in the general cytoplasmi
also. The cisternae of the stacked :Tmooth ER are continuous .:ith
the cisternae of the rough ER. I propose that the .roat'inr are
synthesized in the rough EF with the aid of ribosomnes in are
transferred to the smooth ER. These proteins are Fpol:i'rizcd and
mini-r.icrotubules are thus formed between, the cistern- of ti.e
stacked smooth ER. The newly formed raini-microtubules arr f.orceJ
out to the periphery and later In:o the general cytoplas-i. Fibrillar
material in the sheath of the haustorium might be the product of
depolyirerization of mini-microtubule protein, suggesting soire kind
of exchange between the host and the sheath.
The one atypical haustorium of r-i:' i. an-i.-: ,huich I
observed is very different from those of other fungal parasites.
Its protoplast is diffuse. Unlike other sister haujtoria it does not
have mini-microtubules. It looks as if it is being e3ten. The
increase in the number of inuaginations of the pladmIalere. iich are
unlike plaimalermasor.es may represent sone kind of increased
activity there. These invaginations may iTmark some kind of pinocytotic
activity. Wheeler ;: :'. (1972) have recently reported pirccytosi;
in root cap cells. Pinocytosis on the host parasite interface has
been proposed by Ehrlich and Ehrlich (19E- ) and Berlin and Coten
(1964). Other workers have also reported aberrant h'ustoria
(Berlin and Boven, 1964; Littlefield and Bracker, 1972). The
necrotic haustoria were completely walled off from the host cyto-
plasm by a material continuous with the host cell wall. In termites
there are no cell walls sn tie aberrant haustorium might be cut
off by a physiological barrier that results from the host activity.
The other departure from normal in Zrmitaria : ,eri haustoria
is the presence of lysosomes that have not been reported so far in
other parasites. Autophagic vacuoles with mitochondria, ER, ribo-
somes and lipid droplets in them are commonly seen. There are other
big vacuoles with vesicles in them that exhibit acid-phosphatase
activity. The haustoria are packed with multivesicular bodies in
the vicinity of the lysosomes. The vesicles inside look similar
to the ones cut off on the periphery of the stacks of smooth ER
cisternae. Though these multihesicular bodies have not been seen
fusing with the lysosomes, they may be carrying lysosomal enzym-es.
It has already been suggested that the vacuolar system of plants
may represent the lysosomes, and that lysosomal enzymes are associated
with EP-derived microsomal vesicles (Matile, 1969). It is possible
that the haustoria of the erm.tar-ia .svneri are getting different
kinds of food from the termite than the haustoria inside a plant
host thereby making lysosomal activity necessary. By using Ei
autoradiography Ehrlich and Ehrlich (1970) have observed transfer
of 14 C from Pu.-nia' y-rmiz. :rstiei uredospores to wheat host
The haustoria of Ter:t- r s :,-ri are the first of its
kind studied. More work is needed to ascertain whether this
type of hustorial structure is unique or is shared with other
external ento-agenou parasi:es. Further study of the haustoria
of this fungus is also required to '.now iore about the nature of
mini-microtubules and of parasitism in general.
Figure 50. A thin section of old haustorial mother cell
X20,000. Notice the many-layered inner wall
(iw), smooth endoplasmic reticulum (er), lipid
droplets (L), mitochondria (M) and nucleus (N)
with nembranou2 structure inside the nucleoplasm
Figure 51. The rough endoplasmic reticulum (ER) of the
atypical haustorial mother cell 774,000. Arrows
point to the spheres of rough endoplasmic reticulum
with ribosores inside.
.... ... m..:: E ... "
B~~.;;i,,.~ ~~........ .: :
SF ^A -s
Figure 52. A median section through the neck of the haustorium
(n) X24,500. Notice the thich outer layer (ow) of
haustorial mother cell wall and the continuity of
the inner layer (iw) of the haustorial mother cell
wall with the inner layer (l) of the wall of the
haustorium. ilitochondrion (M), bundle of mini-
microtubules (DlT) and nicrotubules (MIT) are
present in the neck.
Figure 53. Near median section through a penetration site X20,750.
Haustorial mother cell (H.C) Is connected to the
haustorium (H) by a neck (n). Outer layer of
haustorial mother cell wall (oW) and that of
haustorial wall (ON) are dark while the inner
layers of both haustorial mother cell wall (IW)
and haustorial wall (IW) are electron light and
the two appear to be continuous. The hole in the
cuticle (CU) through which the neck runs appears
to be very smooth. Nucleus (rN) has a nucleolus
(NU) and is bounded by a double membraned nuclear
envelope (NE). Certain miembranous structures are
present in the nucleoalasm (arrows). Open arrows
point to the constricted mitcchordria.
* - '! ~
Figure 54. Thin section of old hausto-ia (H) at the site of
original infection Y3,20'iC. Hautoria have lysorres
(1) and nucleus and are separated from the host
protoplast (h) by a sheath (S).
Figure 55. Thin section of hausteria (H) underneath the cuticle
(CU) of the termite X7,200. HausLoria are separated
from the host protoplast (h) by a sheath (S). Notice
the enormous size of the mitochondria (11).
Figure 56. Thin section of the hauscorium X.35,750. Plasma
membrane is connected (arrow) to a plasrialemmasome
(PL) projecting into the haustorial cytoplasm.
Outside the haustoriu. in the sheath there is
fibrillar material (fm). A microtubule (IIT) is
present inside the haustorium cytoplasm.
Figure 57. Haustorial mitochondrion constricted in the middle
X82,000. Arrow points to the connection between
the crista and the inner mitochondrial rmerbrane.
Notice the nearby mini-microtubules (MI1T) and
rough endoplasiric reticulum (ER).
ira -~i I ., k
Figure 58. A portion of a typical hau.torium with aberrant
mitochondria X35,750. Protoplasm of the haustorium
looks diffuse. Plasma meiTi brare is inv'aginated at
many places (arrows). The invaginations have
vesicles in then. Few vehicles are present in the
cytoplasm. Uall of the haustorium (WI) has two
distinct layers. Host protoplasm (h) is also in
Figure 59. Host-parasite interface 82,000. The host plasma
membrane (hPM) tightly fits against the haustorium
wall (U). Compare the diameter of mini-microtubules
(MIIT) of the haustoriun with that of the micro-
tubules (MT) of the host cytoplasm (hi.
Figure 60. Transversely cut mini-microtubules between the
plasr.a membrane (P;I) of the haustroium and its
endoplasmic reticulum (ER). Wall of the haustorium
(W) is .'ery thick.
Figure 61. flini-microtubule running (arrow) between nuclear
envelope (NiE) and the endoplasmic reticulum-
Figure 62. Periodic acid silver-stained haustoria ,43,750.
Haustorial wall (W) is PAS-positive while fibrillar
material (fm) in the sheath is PAS-negative.
Figure 63. Ilicrobody (ME) of the haustorium X54,500. Notice
the close association with the endoplasmic reticulum
(ER). Mlitochondrion (M) is also in vicinity.
.- .1 -. I I .
I .. .., i8
,<. *- .- i..
L-. .. .. r I "' "
V .-:-. :/ ;. '*-*
Figure 64. lini-microtubules (r-IT) between the cisternae of
endoplasmic reticulum (ER) .109,000. The mini-
microtubules appear to be produced by endoplasmic
reticulum and extruded towards the periphery.
Figure 65. A bundle of longitudinally cut mini-microtubules
Figure 66. A stack of endoplasmic reticulum with longitudinally
cut mini-micratubules between the cisternae (arrows)
Figure 67. A stack of endoplasmic reticulum with transversely
cut mini-microtubules between the cisternae X30,000.
The endoplasmic reticulum is cutting off vehicles
(arrow) at the periphery. Certain vesicles (VS)
are present in the cytoplasm.
Figure 68. A bundle of mini-nicrotubules cut transversely
Figure 69. Mini-microtubules cut transversely X824,000. Six
subunits are clear (arrow) in the viall of the
mini-microtubule. Compare the endoplasmic reticulum
membrane with the subunits of the wall of mini-
Figure 70. Acid phosphatase activity in the hauscorium
X56,000. The activity can be seen in the lysomer
(1) and over the steck of smooth endoplasnic
reticulun (er) cisternae with mini-microtubules
between then. The mitochondria (I) and the
haustorial wall (W) show no such activity.
Figure 71. A lipid droplet being surrounded by endoplasmic
reticulum forming autophagic vacuole .51,500.
Figure 72. Autophjgic vacuole (1) with mitochondrion (H)
inside X51,500. The wall of the haustorium (U)
and the host protoplasm (h) are also seen.
Figure 73. Haustorium with lysosome (1), mitochondrion (1)
and multivesicular bodies (rlVB) '53,000. Vehicles
can also be seen inside the lysosome.
j".4 '',^". ^' -S-r
,"fc "- '.*' 'l *
.C ~1 r
When Tha>ter (1920) described ?er-itri.c as a new genus of
entomophilous fungi on termites, he visualized it as a closely
compacted sporodochium and suggested that "its mature condition,
however, evidently a Fungus Imperfectus, seems to give it a formal
place among the Leptostromaceae." Others (Reichensperger, 1923;
Feytaud and Dieu:eide, 1927; Colla, 1929; Heim et 2:., 1951) also
studied this genus but only Colla (1929) and Heim (personal commu-
nication) had a different opinion about its taxonomic position. Colla
suggested that a new family should be created and Heim thinks it
belongs in Tuberculariales of Fungi Imperfecti. My studies (Chapters
I, II, Ill) have shown a number of features of Te'r-ircar a that were
not reported by these authors. These include the ecto-parasitic
nature of the fungus, the so-called bhlamydospores being haustorial
mother cells in reality, sending haustoria into the host integument,
and the phialoconidial nature of the spores. These discoveries
necessitate an emendation of the genus.
ernita ri: Thaxt. emend. Khan
Entomophilous; primary thallus crust-like, stromatic, few cells thick,
variously shaped with even margin; haustoria penetrate the integument
from specialized thick-walled haustorial mother cells; whole primary
thallus matures into a sporodochium of tightly held phialides making
a hymenium over a basal pseljdoparenchrymtous subhymenium, with a
peripheral excipulum of black thick-walled, sterile cells, margins
lichenoid, spreading but even; phialides cut off conidia apically
in basipetal succession at conidiocenoiu loci into distinct collar-
ettes; collarettes long containing four or more conidia; conidia
catenate, 3.5 4 u X 1.5 2 p, cylindrical, truncate or rounded
at ends, hyaline in mass.
Thaxter originally proposed two species, T. snyderi and T.
corc,:::c, on the basis of the length of the phialides, the length
of the collarettes and the structure of the tip of the phialide.
The phialides and collarettes of .;r--:.:,r ca~ro..a-z are larger in
diameter and length than those of T. ,kI-'r. The phialide tip of
T. n'siri is blunt and that of ;. Joor, r: has minute pointed
prolongations resulting in a minute, echinulate hynenial surface.
Later, Reichensperger (1923) described 2 -r-ni:cia from the Belgian
Congo and Brazil and proposed a new name .. t;ku--rr. for the Brazilian
specimen "should it turn out to be a new species." Having not
examined the type it is impossible for me to know whether it was a
new species. Later. Colla (1929) reported the presence of ;.
,3':.ri and 2. corP.ira, the former from Chile and Brazil and the
latter from British Guayana, Costa Rica and Phillipines, growing
on termites' exoskelecons.
Colla (1929) also described a new genus from British Gua/ana,
parasitizing the integument of the termite ilcre-e r.~'::~a =
(L) and proposed a new genus ;a:ciazt l!iz with a single species
M. sig:s:ri. .at::rol-lZa closely resembles TerSitri'a but its
sporodochium has 8 to 10 special cavities containing fertile hyphae
surrounded by sterile ones. Colla described 'I. silv stri? as having
a star-shaped thallus in young as well as adult stages, the arms
being intensely black, carbonaceous and sterile. The mature thallus
which is a sporodochium 60-70 u in thickness has three distinct
regions, a basal 12-20 p thick subh;-inium, a middle 37-40 u thick
hymenium, and an apical epihjirenium. The basal layer of the
pseudoparenchyTratous subhirieniui is comprised of thick walled
cells. A foot is sent into the integunent which according to
Colla (1929) reacts with maximum hypertrophy of the epidermal cells.
Thaxter (1920) reported a similar response to ?Ter?.t:: a infection
by the epidermal cells of the termite integument. Through my
studies I know that what ThaAter thought were hypertrophied
epidermal cells actually were haustoria of the Teriit.ria. It can
be assumed from Colla's figures that in the case of ,-cctircleaZ
too, those hpertrophied epidermal cells actually here haustoria.
Thus, in the basal layer of the subhymenium there are certain
specialized haustorial mother cells that send haustoria into the
integument. The hymenium is loculate, locules being separated from
each other by the columns of sterile hyphae that run from the sub-
hymenium upward, anstamose at the top of the hymenium and form an
epihymenium. These locules are full of what Colla called fruiting
bodies. She thought they resembled asci. However, shewas not sure
of this nor was shesure of the phialidic nature of the conidiogenous
cells of Territ.ria. She described them as having spores like that
of 7tr-.;r-5 : but smaller in size. She also desc:rited a dark line
across the sporogenous cells at about two-thir.-; of their length.
This dark line resentles the dark line across the hymenium of the
Termitaria that marks the position of the conidiogenous loi of the
phialides. I think that the fruiting bodies were actually phialides
and as Colla herself said it was not possible for her to confirm the
nature of these sporogenous structures because of their small size.
The epihyrenium is sterile and i.akes an arch over the phialides
in the locular areas.
During this study I had the goodJ fortune of looking at some of
the slides of r.;.r:c-'r; made by Tnadter (courtesy Farlow Her'barium).
These slides were in very good condition and I found that slide
nu-ber 3240i of iar -'.-i;a r:r.c.::- s not a : ,- tri but
t'.ittir~clUl The furigus was collected froni Panama and the slide
was made in 1924. The "lide has rany cross sections of the
sporocarp showing subhymenium, hymenlum, epihrmenriumii and the sterile
hyphae dividing the hymenium into r-ny chambers containing phialides
(Fig. 75). There is a basal layer of thick called cells many of
which are haustorial mother cells sending haustoria through the
cuticle (Fig.re 79). Over this there is a pseudoparench.,Tatcous
subhyr'enium 13 u in thickness from which ari;e the phialides
arranged in a palisade-like layer making a broad 52 u thick
hymenium. Tre hymenium is traversed by sterile hyphae that divide
it into chai kers. The apices of all the fertile hyphae (phialides)
and sterile nyphie are fused with each other and it appears that at
maturity the sterile hyphae elongate, as a result of which the thick
tips of the phialides break off (Fig. 78) and become a part of the
epihymenium. The broken ends of the phialides remain free in the
cavity under the eplhymenium. It is assumed that the conidia are
released in those cavities, put a pressure on the epihymenium that
is broken and flaked off. The epihymenium is 5 u thick and has a
reticulate surface (Fig. 80). The space between it and the broken
phialide tips is about 8 u thick. The conidiogenous loci of all the
phialides lie at their bases and each collarette which is approAimately
as long as the thickness of the hyrenium has 10-12 truncate, hyaline
conidia. The conidia are 2.5 3.0 u X 1.5 n, much smaller than that
of the Ti-itaric and broader compared to their size (Fig. 74). The
whole sporocarp is 85-90 u thick.
From the description of the fungus it becomes apparent that this
one is slightly different from :!. sirlestrii in having longer phialides,
thicker sporocarp, conidiogenous loci lying at the bases of the
phialides and that the surface of the epihymenium is reticulate in
this species. Colla's figures also show that the sterile hyphae
partitioning the hymenium into chambers make wider columns in
tU. siZest rii. Thus, it is proposed here that this fungus be
regarded as a new species:
MattirellZZ.: c -lsosa sp. nov. Figs. 74-80.
Entomogenous; thallus maturus sporodochium, densus 85-90 u;
hymenium densum 52 u divisum in receptacula numerosa colu.nis
hypharum sterilium latarum paucilocularibus; phialides strict,
loca conidiogenera sunt in fundamentis phialidum, apices crassi
phialidum et apices hypharum steriliur, sese conglutinanat epihymeniumi
formantes, prolon.jatio hypharuaTi sterilium phialidium apices frangunt,
cavitatem super apices fractos efficientes; conidii enodojeniter
format, 10-12 in coll retta singular, truncati, citenr ti, 2.3 3.0 .
1.5 u, hyalini et cylindrati; superficies epihymerni reticulatus.
Entonogenous; mature thallus a sporodochijum, 35-90 u thick;
hyaienium 52 u thick, divided into rany chambers b frew cells wide
columns of sterile hyphae; phialides tightly held, conidioglenous loci
lie at the bases of the phialides, tnick tips of phialijes and the
tips of sterile hjphae fuse with each other forming an epih,ieniun,
elongation of the sterile hyphae break the phialide tips mal ing
cavity over the broker tips, conidia formed endogenously, 10-12
in each collarette, truncate, catenate, 2.3 3.0 X 1.5 u, h-,aline
and cylindrical; the surface of the epihimenium reticulate.
uct:ircLoea and e ni:ria have iany characteristic: in common.
They both are entomophilous, sending haustoria from specialized
thick walled haustorial mother cells into the inte-gument through
the cuticle of the host, and the haustoria do not go beyond the
epidermis. Both have a foot-like structure nhich according to this
study of 'Te ..ii.a 'i .- .'f- marks the position of original infection
and penetration. Both have a crustose thallus which is stroTmatic in
young stages. In both cases the whole thallus matures into a
sporodochium that has a pseudoparenchynaatous subhymenium, and a
hjnenium of phialides. There is an excipulun around the sporodochium
in both genera. However, they differ in iany respects. All cells
in the basal layer of the thallus are thick walled in Mtti'r:Zejll
while in 7e. r.iar.ia only a select number are thickened. There is
an epihymenium present in L:i:zle21 while in :e- ::-".i there is
none. The 7ymenium of X:ti;rc:>EiZ is divided into many chambers
containing phialides separated from each other by the columns of
On the basis of major difference; between the t.o, Colla (1929)
proposed the genus .larir,!;' .2 but suggested that it be put in a
single family. She did not suggest any name for the family.
The characters of the two genera are so unique that it is very
difficult to put them into any recognized group of Fungi Imperfecti.
That may be the reason why the taxonomic position of the two is
The imperfect fungi are placed under subdivision Deuteromycotina
by Ainsworth (1971). The Deuteromycotina is subdivided as follows:
Blastomycetes -- yeast and yeast-like forms.
Hyphomycetes -- mycelial formTs with or without conidia.
Agonomycetales -- without conidia.
Hyphomycetales -- conidia not on synnemata or sporodochia.
Stilbellales -- synnemateous.
Tuberculariales -- sporodochial.
Coelomycetes -- conidia within acer'uli or pycnidia.
Sphaeropsidales -- pycnidial.
Melanconiales -- acervulial.
Two other orders of Fungi Imperfecti have been proposed,
Peltasterales and Gloeohaustoriales, by Batista and Ciferri (1959)
and Heim (1952) respectively, the former for the imperfect states of
Microtheriales that have pycnidia with inverted hymenium and the
litter for such entoogenou eno nera as 4:,; .'... 3oZ and
C;.;.tra,!c-T: r i Ob.'viousl. ".':: -rc.' j and ta-n'..:v, do not
belong to these to order;. They resemble only rmeiTiers of the
orders Sphaeropsidales or Tuber:ul.ariales because the sporocarp
of :-t.:.:,:-L ri is closer to the sporodochiur of tie latter while
that of .'2:s:ir-J, ;;.. is closer to the stron having many pycnidi
embedded in it like in Sphaeropsidales. The spo'rodochium i a
determinate, pulvinate masi of conidiophores tightly clustered on
a stromra 3'd ojo- not have an e-cipulum of fungal material. The
sporocarp of T?'-f i.'rf. develops on 3 stroratic primary thallus,
is comprised of tightly held phial ide, and has a peripheral e'cipulum
of black thick walled sterile cells. Thus, it 15 a modified form of
sporodochium. The pycnidiujm is defined as a variously shaped cavity
having a pseudoparenchniatous wall that is lined with conidiogenous
cells. It may be inmersed within a stroma or formed superficially
on the surface of the substratum. In ,.' ri'r-.:.::.w the h menial
cavities containing phialides can be superficially compared with the
pycnidia eurtedded in a strom,.
As has been discussed earlier 7ar-L-ia and !!:: .,,Z.rl appear
to be very closely related but differ in their sporocarp structure.
The two on the basis of their general morphology and the structure
of their sporocarps should not be put into any of the existing orders.
Therefore a new order is proposed here.
Termitariales order nov.
Entorogenous; ectoparasiticus, haustoria in integumentum
hospitis penetrans ex cellis matribus crassitunicatis specialibus;
haustoria restrict in texto epidennali hospitis; thallus principals
crustosus, stromaticus sine hyphis 1a.is qui maturat in sporocarpo
perobscuro; sporocarpum coniritens subh)nenio pseudoparenchymate,
hymenio phialidibus haesis strictis et cincto e.cipulo sterili
cellarum crassitunicatarum; phialis cylindratus, phialonconidios
in ordine basipetalo in locis conidiigeneribus in collarettes
apicaliter abscindens, conidii eseptatis, hyalini, cylindrati,
catenati, marginibus truncatis aut rotundis.
Entomogenous; ectoparasitic, sending haustoria into the integument
of the host from specialized thick walled haustorial mother cells,
haustoria limited only to the epidermal tissue of the host; primary
thallus crustose, stromatic without any loose hjphae, matures into a
very black sporocarp; sporocarp consisting of a pseudoparenchymatous
subhymenium, a hyi enium of tightly held phialides and surrounded by
a sterile excipulum of tnick walled cells; phialides cylindrical,
cutting off phialoconodia apically in basipetal succession at conidio-
genous loci into collaret:es; conidia aseptate, hyaline, cylindrical,
catenate with truncate or round ends.
This is a provisional order created for Ter-i. sii and
-Stti rcJell.. Tentatively, a single family Termitariaceae will
be included, but when other entomogenous taxa are investigated or
perhaps sexual features of these genera are discovered, a complete
revision of these taxa will be necessary.
The phialides of 7...'-.':-,:': :d ,'L:'I.rcl .'a closely resemble
those of c--.-:'..7 rC r.::.: Eerk. ar.d Br. (Pirozynski and Mor.;an-Jones,
1968). C::- (Corda) F'penh., ,.';.KZ-:;~Z: S Peyron and Thielaviopsie
Went. The phialide: of all of them cut off chains of conidia in
basipetal succession at conidiogen:.us loci into the long collarettes.
The conidia are cylindrical, a:eptate and hyaline or lightly pigmenteJ.
Bloxamia belong: to Tuberculariales and i: saprophjtc uhile the
other three belong to H..phonyce.tals and can be found as -aprophytes.
Thil- .j;...f: is a serious: pathogen of higher plant:. The conldio-
phores of trm synnematio'us fungus E:*.'r'c:.; s Subram. also sho.a
some reserrbTance to the philides o r" -t-.r;,- ,iand '::: .'.;.:.l? ..
They are grzsjped in a single synnewj cutting off aseptate, hyaline
conidia erl' encusly at conidiog-n'ous loci in small collarettes.
The conidia are e.truded in basipetal succession through the open
ends of collarettes. The sporocarps of .n:r-I,-rt.I and 5:.:: ;'..z are
alike in .iiany respects. They are black, comrprised of tightly held
phialidjes whose conidiogenous J clci lie at the same distance from the
base of the phialides making a distinct line running across the
It can be said chat the members of the proposed order closely
resemble ipembers of the orders of class H)phomiycetes and therefore
it should be placed under the same class ne.t to Tuberculariales.
PLATE X II
Figures 74-80. Mictit cKl, Ua .T:.ttC& sporocarp.
Figure 74. A portion of tne cro;s section of the sporocarp show-
ing subhyrenium (SH), h)yenium (HY), the zone of
conidiogenous loci (CL) and the conidia (C) inside
the collarettes X1,250.
Figure 75. Cross section of the sporocarp X250. The sporocarp
has three distinct regions, epihTnenium (EH). hyrenium
(HY), and subhm-nenium (SH). Columns of sterile hyphae
(ST) divide the hymenium into many chambers. Conidio-
genous loci (CL) at the bases of the phialides rake
a distinct zone. Basal layer (EL) and the excipulum
(E) are also seen.
Figure 76. Cross section of the sporocarp 1,000. Column of the
sterile hyphae (STJ divide the hmenium. Notice the
cavities (cv) over tUe phialides.
Figure 77. Cross section of the sporocarp X320. A column of
sterile hyphae (arrod) is seen all the wa) through
the hir.enium running between the epihymenium (EH)
and the basal layer (BL).
Figure 78. Cross section of the sporocarp K.1,000. The broken
tips of the phialides (arrovs) that fuse to help
make the epihymreniun (EH).
Figure 79. Cross section of the basal lajer of the sporocarp
X1,250. Thick walled haustorial mother cells (hr c)
send haustoria into the integument of the termite
through the channels (arrows) in the host cuticle
Surface of the epihywnium X1,000. Note its
~-1' ~ ( 80i