New species of Troglomyces and Diplopodomyces (Laboulbeniales, Ascomycota) from millipedes (Diplopoda)

We describe five new species of fungi of the order Laboulbeniales Lindau growing on millipedes and belonging to the genera Diplopodomyces W.Rossi & Balazuc and Troglomyces S.Colla. Three new species of Diplopodomyces , viz. Diplopodomyces coronatus Santam., Enghoff & Reboleira sp. nov. living on Serboiulus spelaeophilus Gulicka, 1967 from Bulgarian caves, Diplopodomyces liguliphorus Santam., Enghoff & Reboleira sp. nov. on an unidentified species of Spirobolida from Sri Lanka, and Diplopodomyces ramosus Santam., Enghoff & Reboleira sp. nov. on Pachyiulus spp. from Turkey, Macedonia and Serbia; and two new species of Troglomyces , viz. Troglomyces dioicus Santam., Enghoff & Reboleira sp. nov. on Nepalmatoiulus sp. from Myanmar, and Troglomyces tetralabiatus Santam., Enghoff & Reboleira sp. nov. on Caucaseuma Strasser, 1970 and Heterocaucaseuma Antic & Makarov, 2016 from caves in Western Caucasus. Troglomyces dioicus sp. nov. is the first dioecious species described in the genus Troglomyces . Keys for all hitherto known species of Diplopodomyces and Troglomyces are presented, as is a discussion of the status of both genera. Additional records for Diplopodomyces lusitanipodos Santam., Enghoff & Reboleira and Troglomyces manfrediae S.Colla are also included.


Introduction
Fungi of the order Laboulbeniales Lindau represent more than 2000 species placed in the Phylum Ascomycota (Weir & Hammond 1997;Santamaría et al. 2017).They are well characterized by having very minute compact thalli, mostly 50-500 µm in length.Laboulbeniales are obligate biotrophs that live attached to the tegument of arthropods, mainly insects, but also arachnids and millipedes (Santamaría et al. 2017).It is a common view that these fungi have little or no effect on their hosts (Benjamin 1971;Tavares 1985).Fitness effects have been estimated by a few controlled studies (Whisler 1968;Strandberg & Tucker 1974), as well as some insights into how these fungi may influence the behaviour of their hosts (Konrad et al. 2014;Báthori et al. 2017;Csata et al. 2017).Their source of nutrients is debatable, but they are believed to draw them from their hosts through absorption or via contact with living tissues (Scheloske 1969;Benjamin 1971;Tavares 1985).Traditionally, Laboulbeniales have been considered ectoparasites, because some of them show a more or less conspicuous haustorium that penetrates into the host body (Thaxter 1896).However, recent studies (Tragust et al. 2016) demonstrate that a few species do not show these structures and therefore, at least in these cases, Laboulbeniales should be defined as commensalistic.This point of view is not new, having been thoroughly used by some previous authors (Tamada & Kaya 1993;Lawrence & Milner 1996;Jolivet 1998).It seems almost unquestionable that both life styles, parasitism and commensalism, exist in Laboulbeniales, and the use of the term "ectobionts" for Laboulbeniales, instead of the most restrictive "ectoparasites", is suggested.
Laboulbeniales were assumed to be very uncommon on millipedes (Diplopoda) but recent collaboration of myriapodologists with laboulbeniologists has revealed that this is not beyond doubt.The number of these fungi living on millipedes has been substantially increased during recent years from eight species (Santamaría et al. 2014), to 24 (Enghoff & Santamaría 2015;Santamaría et al. 2016), and 29 if the new taxa described here are included.In the present paper five new species of the genera Troglomyces S.Colla and Diplopodomyces W.Rossi & Balazuc are described, and new information about these genera which are found only on millipedes is also given.

Material and methods
Millipede hosts are preserved in 70% ethanol and were obtained from the zoological collections of the Natural History Museum of Denmark (ZMUC), except for specimens of Pachyiulus cattarensis Latzel, 1884 which are from the Naturhistorisches Museum of Vienna (NHMW).Fungi were detected by the observation of millipedes under 40-80× magnification of a dissecting microscope.Permanent slides were prepared following previously described methods (Rossi & Santamaría 2015), and are kept in the BCB Mycotheca of the Universitat Autònoma de Barcelona (BCB slides), and in the Mycological Collection Herbarium of the Natural History Museum of Denmark, University of Copenhagen (C-F slides).Photomicrographs were made with a Jenoptik ProgRes 10 Plus digital camera on a Leica DMR microscope equipped with differential interference contrast optics (DIC).Images were processed with Photoshop CS5 software.
We mostly follow the terminology according to Tavares (1985).

Brief description
Receptacle three-celled, with superposed cells I and II.Cell III lateral to cell II.Septum II-III vertical to oblique.Perithecium with four or five strongly unequal outer wall cells in each vertical row.The original description was extended by Santamaría et al. (2014).Six species.

Etymology
The species epithet means "crown-like", in relation to the crown-like perithecial tip, resulting from the peculiar disposition of its four apical wall cells.

Description
Thallus hyaline, except for the blackened foot.
Basal cell (I) slightly longer than broad, enlarged distally, with the base abruptly constricted at the point of junction with the blackened foot (it breaks easily with manipulation).Suprabasal cell (II) trapezoidal, slightly longer than broad.Cell III nearly twice as long as broad.Septum I-II horizontal, septa II-III and II-VI oblique.Septum II-VI two times longer than septum II-III.Primary septum (a) slightly constricted, oblique.
Primary appendage strongly divergent from the perithecial axis, forming an angle of 45-70º, unbranched, consisting of up to 4-5 similarly elongated cells, gradually tapering distally towards a blunt apical cell, if unbroken.Spiny remains of spore apex (sx) conspicuous, stout and blunt, located on the distal margin of the 4th cell.A single bottle-shaped, narrow and elongate antheridium, borne on a corner cell at the adaxial side of appendage basal cell.
Perithecial stalk cell (VI) flattened, trapezoidal, 2.5 times as broad as long.Secondary stalk cell (VII) and basal cells of perithecium (n, m, n') well defined.Perithecium almost conical, broadest near the base, gradually tapering into a truncate apex.The latter bears four rather conspicuous and blunt lips, formed by upper tier of perithecial wall cells (w 4 ).Margins of perithecium showing a slight indentation at the junction between what seem the w 3 and w 4 tiers, above which the margin is slightly convex (Figs 1C-E).Length from foot to apex of perithecium 104-139 µm.Perithecium (including basal cells) 70-89 × 18-23 µm.Primary appendage (from primary septum to apex, undamaged) 48-84 µm.

Remarks
Thalli are sparsely distributed on the host body, mostly on the legs, a few on the body but the latter in poor condition.
Exceptionally one immature thallus bears an additional antheridium on the second cell of appendage, identical in shape, size, position and orientation to the usually single antheridium.
As mentioned in the diagnosis, this species seems to be closely related to Diplopodomyces lusitanipodos, but differs from the latter for the position of the single antheridium on the appendage basal cell, instead than on the second appendage cell.Moreover, in Diplopodomyces coronatus Santam., Enghoff & Reboleira sp.nov. the antheridium is born on a corner cell (sc-an, Fig. 1F), which is absent in Diplopodomyces lusitanipodos.The perithecial apex appears to be ornate with four protuberant lips, i.e., the upper outer wall cells (w 4 , Fig. 1C-E), whereas it is blunt although showing four minute preapical rounded protuberances in Diplopodomyces lusitanipodos (Santamaría et al. 2014(Santamaría et al. : 1031)).The sx (Fig. 1F) is hardly distinguishable, short, blunt and stout on the fourth appendage cell, whereas it is spiny, easily recognizable, and usually on the third appendage cell in Diplopodomyces lusitanipodos.

Diagnosis
The presence of a ligula-like protuberance on the perithecial apex is the most striking characteristic of this species which does not seem to be closely related to any other species in this genus.

Etymology
The species epithet means "bearing a tongue", in relation to the perithecial protuberance.

Description
Cells II, III, VI and basal cell of primary appendage bright reddish coloured; the rest of the thallus is hyaline, except for the blackened foot.
Basal cell (I) two times as long as broad, typically obconical, narrow at the base and gradually broadening distally.Suprabasal cell (II) more or less trapezoidal and isodiametric.Cell III smaller than II, slightly longer than broad, laterally adnate and exceeding in length cell VI.Septum I-II horizontal.Septum II-III oblique, almost half in length than I-II.Septum II-VI very oblique, as long as septum I-II.
Primary appendage unbranched, variably bent, consisting of up to 8 cells, gradually tapering distally.Basal cell of primary appendage three times as long as broad, distinctly larger and darker than other cells of appendage.Typically a single flask-shaped antheridium without supporting cell is borne on the second cell of appendage.Spiny remains of spore apex (sx) conspicuous, stout, on the third cell, rarely on the fourth cell of appendage.
Perithecial stalk cell (VI) triangular to rhomboid, obliquely arranged above cell II and under the perithecial body.Secondary stalk cell (VII) and basal cells of perithecium (n, n', m) hardly distinguishable.Perithecium fusiform, with a blunt apex which bears a small but evident, erect, rounded, ligulate protuberance (li), formed by a w 4 perithecial wall cell.

Remarks
Abundant thalli of the new species were scattered on legs and body rings (most on ventral side) of the frontal part of the millipede.Rarely a second, additional antheridium was seen on the third appendage cell (an, Fig. 2D).Interestingly, the nuclei appear to be conspicuous in all cells of the appendage and also inside the basal cell (I) of the receptacle when seen with DIC optics.
This species is well distinguished by the protuberant lip on the perithecium, the reddish colour in some receptacle cells, and poorly defined walls of the basal cells of the perithecium, a characteristic unknown in the tribe Laboulbenieae but found in other genera of other tribes.

Diagnosis
Characterized by a very long and ramified primary appendage, by the scattered antheridia along the main axis of the appendage and by the stocky perithecium bearing four conspicuous acute lips on the apex.

Etymology
The species epithet means "branched", in relation to the large and ramified primary appendage.

Description
Thallus yellowish-brown, often darker on cells II, III, VI, VII and perithecial basal cells, forming a defined stalk below the perithecium.Blackened foot, rather elongate and pointed.
Basal cell (I) 3-4 times as long as broad, with slightly divergent margins towards the apex and nearly parallel margins in the central area; straight except for the constricted, curved and geniculate base above the foot.Suprabasal cell (II) trapezoidal, almost isodiametric or slightly broader than long.Septum I-II horizontal.Cell III 1.5 times as long as broad, cylindrical (rectangular in optical section), adnate to cell VI and to the base of the perithecium.Septum II-III slightly oblique.Septum II-VI strongly oblique.
Primary appendage very long, consisting of a main axis of 6-9 cells which are 3-4 times longer than broad, each bearing 1-3 flask-shaped antheridia on distal and relatively large corner cells.The appendage ramifies approximately above the 4th cell, where corner cells grow into thin, long, and flexuous branches, which often ramify again forming a confused mass of appendages glued together with dirt.Primary septum (a) well defined by a constriction and contrasting colour between the darker cell III and the pale basal cell of primary appendage, below the base of the perithecium.
Perithecial stalk cell (VI) rhomboidal, arranged obliquely between cells II and VII; slightly smaller than cell II.Secondary stalk cell (VII) and basal cells of perithecium (n, m, n') well defined, forming a more or less defined stalk or pedicel under the ascogenic mass of the perithecium.Perithecium broadly elliptical, with a constricted lower third, an inflated venter, and a narrow tip; the apex bears 4 small, acute, teeth-like ligulae, resembling teeth, formed by the w 4 tier of perithecial wall cells.

Remarks
This species grows on male hosts around the gonopods and on females close to the gonopores at the base of leg-pairs 1 and 2. In this feature it is analogous with Diplopodomyces veneris Santam., Enghoff & Reboleira (Santamaría et al. 2014) but despite this coincidence, the morphological characteristics of both species are incomparable.Moreover, Diplopodomyces ramosus Santam., Enghoff & Reboleira sp.nov.does not show dimorphism in relation to the host sex.
The spiny remains of the spore apex (sx) is not visible on mature thalli.Only one very young thallus has been observed showing this characteristic (sx, Fig. 3E).The four small ligulae of the perithecial apex are well distinguished only in over-mature thalli, not when ascospores start being visible inside the perithecium (Fig. 3F) (which is the usual condition for most Laboulbeniales when we referring to a mature perithecium).
The four ligulae protrude when most of the external wall of the perithecium is broken and the ascospores have been released.
Both host species belong to the problematic genus Pachyiulus Berlese, 1883 (Frederiksen et al. 2012).Pachyiulus cattarensis occurs in forests as well as open areas from sea level up to 2400 m (Kime & Enghoff 2017).

Brief description
Receptacle three-celled, with superposed cells I, II and III.Cells II and III narrow and adnate to the perithecium.Perithecium with five, rarely six, outer wall cells in each vertical row.Perithecial apex bearing four conspicuous lobes.The original description was emended by Enghoff & Santamaría (2015).Eight species.

Diagnosis
First dioecious species in the genus, with female thalli characterized by a very short cell III and a septum II-III placed under the level of the perithecial base, whereas in the other species cell III is taller and the septum II-III is clearly higher up.

Etymology
Self-defined by its species epithet, meaning "two households".

Description
Dioecious.Both male and female thalli are hyaline except for the darkened feet.
Male thalli.Consisting of six superposed cells.Basal cell (I) about 3 times as long as broad.Suprabasal cell (II) small and squared.Cell III 1.5 times as long as broad.Primary septum well distinguished.Fourth cell 1.5 times as long as broad.Fifth cell squared, very small, bearing the single, apical, bottleshaped antheridium.No spiny remains of spore apex because it is coincident with the apex of efferent tube, i.e., where the discharging pore is found.
Female thalli.Basal cell (I) twice or more as long as broad, enlarged distally, slightly constricted at the base.Suprabasal cell (II) as long as broad, laterally adnate to cell VI, broadened at base where in contact with cell I, narrowed toward distal end.Cell III very narrow, 3 times as long as broad, laterally adnate to perithecium.Septum I-II horizontal.Septum II-III horizontal or slightly oblique, below the level of septum VI-VII.
Primary appendage unbranched, forming an angle of almost 90º with main axis of thallus, consisting of up to 4 cells, of which the lower is about 4 times as long as broad, the others gradually shorter and tapering distally.Spiny remains of spore apex (sx) difficult to see in mature thalli (Fig. 4A), but well visible in very young thalli (Fig. 4D).

Remarks
This species is unique among the species of Troglomyces for its dioecism and because cell III, although adnate to perithecium (a characteristic of the genus) is located much lower than in other forms.
Male and female thalli are paired, as it is usual in dioecious Laboulbeniales.However as can be seen in Fig. 4A we have found a pair of females without any trace of antheridia or paired males (often seen as remains of the darkened feet, e.g., Fig. 4B).
The species seems very rare and the thalli were scarce on legs 2, 3 and several anterior post-gonopodal legs.Undamaged mature male thalli with spermatia present in the efferent neck have been seen only twice (Fig. 4C); immature thalli are more common and help us to understand that the efferent neck and pore (through which spermatia discharge) coincide with the original acute ascospore apex (Fig. 4D, arrow), an unusual feature among Laboulbeniales.
Dioecism is not unusual among genera of Laboulbeniales although most species are monoecious.A good example is the genus Laboulbenia Mont.& C.P.Robin, with more than 600 species, only three were demonstrated to be dioecious (Santamaría 1998).
The host belongs to the large Southeast Asian genus Nepalmatoiulus Mauriès, 1983, which are mainly inhabitants of forest litter (Enghoff 1987).

Etymology
The species epithet is in relation to the four elongated lips on the perithecial apex.

Description
Thallus hyaline, except for the blackened foot.
Basal cell (I) 1.5-3 times as long as broad, enlarged distally, strongly constricted at the base.Suprabasal cell (II) twice as broad as long, laterally adnate to cell VI, broadened at base where in contact with cell I, narrowed toward distal end.Cell III very narrow, up to 7 times as long as broad; laterally adnate to the perithecium, extending up to a third of the total length of the perithecium (lips included).Septum I-II horizontal.Septum II-III very short and oblique.
Primary appendage branched above the second and successive cells, rarely above first cell (Fig. 5D); short, not extending beyond the perithecial apex.Basal cell of primary appendage longer than other cells of appendage, about 2 times as long as broad.Primary septum constricted, strongly oblique.Antheridia flask-shaped, only clearly seen on immature thalli; usually a single one borne directly on second cell of appendage, sometimes supported by a short branch; sometimes a second and even a third antheridium may be found on the branches above.Antheridia deteriorate after maturation of thallus.Spiny remains of spore apex (sx) not seen.
Perithecial stalk cell (VI) inconspicuous, strongly flattened.Perithecium conical, the broadest part at the base, gradually tapering upwards.The apex bears 4 conspicuous identical, often distinctly divergent and curved lips, about 7 times as long as broad, arising through a scar.

Remarks
Among the hosts examined there are some with heavy infection on legs (Fig. 5A).Thalli found near the gonopods (slide SS E613f) are slightly different from others, therefore may be considered a growth form.Interesting are the digitations penetrating the tegument observed in some feet (Fig. 5I-K).
As mentioned in the diagnosis, this species is similar to Troglomyces bilabiatus, which lives on Acipes spp.(Blaniulidae) from Madeira and Canary Islands (Enghoff & Santamaría 2015).Troglomyces tetralabiatus Santam., Enghoff & Reboleira sp.nov.differs from Troglomyces bilabiatus by having four apical lobes instead of two, its more slender habitus and branched appendage.However, despite the morphological similarity it is hard to imagine a close relationship between two fungi living on such geographically and taxonomically isolated hosts.
The scar near the perithecial apex (Fig. 5C, arrows) was already described and discussed for five other species of Troglomyces (Enghoff & Santamaría 2015).
The hosts are cave-dwelling species of Western Caucasian anthroleucosomatid millipedes, belonging to genus Heterocaucaseuma and Caucaseuma.This species was found living in a cave at the remarkable depth of 1980 meters below the surface in the Krubera-Voronja Cave (Sendra & Reboleira 2012;Antić et al. 2018).This makes it the "deepest" Laboulbeniales species on Earth.No significant difference regarding fungal position or thallus density was found when comparing hosts collected at different depths inside the cave.

Remarks
The host is Guadarramasoma ramosae, a species collected in large numbers in traps set in the Mesovoid Shallow Substratum, MSS (Gilgado et al. 2017).This is the first record of Diplopodomyces lusitanipodos from a species of the order Chordeumatida; previously known hosts belong to the genera Lusitanipus and Cyphocallipus (Callipodida), reported from Portugal and Spain (Santamaría et al. 2014;Reboleira & Enghoff 2015, 2017;Antić et al. 2018).

Remarks
Previously known only from Typhloiulus Latzel, 1884 and Trogloiulus Manfredi, 1931 from Italy (Santamaría et al. 2014).The new records extend both the host range and the geographical distribution of this species.Among the new hosts, Cylindroiulus truncorum, Amblyiulus kovali and Serboiulus kresnik belong to the family Julidae, like previously known hosts, whereas Proteroiulus hispanus belongs to the related family Blaniulidae.Cylindroiulus truncorum is a widespread, largely synanthropic species with a preference for high humidity; Amblyiulus kovali and Serboiulus kresnik are cavernicolous (Haacker 1968;Kime & Enghoff 2017).Infected specimens of Proteroiulus hispanus were collected at high altitude in Sierra Nevada, mainly in the Mesovoid Shallow Stratum, MSS (Gilgado et al. 2015).

Discussion
The new species add significantly to the known geographical distribution of the genera Diplopodomyces and Troglomyces which were until now known only from Europe.Diplopodomyces is recorded from Sri Lanka and Turkey, Troglomyces from Myanmar and Georgia.With millipede-inhabiting species of the genus Rickia now known from Europe, Iran, tropical Africa, Sri Lanka, southeastern Asia, Australia and New Zealand (Santamaría et al. 2016) and the monotypic genus Triainomyces known from New Zealand (Rossi & Weir 1998), millipede-inhabiting Laboulbeniales are now known from all biogeographical regions with the notable exceptions of the Nearctic and Neotropical (and Antarctic) ones.
Until Diplopodomyces and Troglomyces were monotypic, they appeared very distinct.Now that quite a few new species of these genera have been described, it is evident that they are very closely related.The discovery of new species has broadened the limits of the generic descriptions.Tavares (1985), in the key of her monograph, separated Troglomyces because of cells II and III are narrow and adnate to the base and side of the perithecium.At the time of Tavares' work, the only known species in the genus was Troglomyces manfrediae S.Colla; moreover Tavares indicated that she never saw this taxon.Enghoff & Santamaría (2015) emended the description of Troglomyces to broaden and refine some characteristics relating to antheridial structures.The inclusion here of a new dioecious species (Troglomyces dioicus Santam., Enghoff & Reboleira sp.nov.) further modifies the generic limits.
Diplopodomyces was compared with Autophagomyces Thaxter (Rossi & Balazuc 1977), and a common ancestor was even suggested (Tavares 1985).In the key published by Tavares (1985) it is mentioned that Diplopodomyces shows very small and not readily visible basal cells and secondary stalk cell of perithecium.This characteristic is not present at least in Diplopodomyces coronatus sp.nov.(Fig. 1A) or Diplopodomyces ramosus sp.nov.(Fig. 3A).Also, a feature that is broadly accepted as very important to separate genera is the number of perithecial wall cells which, however, for each vertical row varies from the five mentioned in Tavares' key to the four seen in, for example, in Diplopodomyces coronatus sp.nov.In steps 59 and 59' of the key, Diplopodomyces and Troglomyces are separated by the limited growth of the appendage in Diplopodomyces in contrast to its indefinite growth, with branching, in Troglomyces.With our present knowledge this distinction is questionable because only Diplopodomyces callipodos meets this criterion.
Two main types of distribution of thalli on the host body can be discerned among Laboulbeniales growing on millipedes.Some species are exclusively, or at least predominantly, found on parts of the host's body which come in contact during mating.Others are more widely distributed on the body and probably are transferred via the substrate or via non-sexual contacts (Enghoff & Santamaría 2015).Among the new species described here, Diplopodomyces ramosus sp.nov., Troglomyces dioicus sp.nov., and possibly Diplopodomyces liguliphorus Santam., Enghoff & Reboleira sp.nov.and Troglomyces tetralabiatus sp.nov.seem to belong to the first category, Diplopodomyces coronatus sp.nov. to the second.The scattered distribution of Diplopodomyces lusitanipodos on the body of its new host is in line with the one observed on previously recorded hosts (Santamaría et al. 2014).