Rediscovery of the enigmatic fungus-farming ant "Mycetosoritis" asper Mayr (Hymenoptera: Formicidae): Implications for taxonomy, phylogeny, and the evolution of agriculture in ants

We report the rediscovery of the exceedingly rarely collected and enigmatic fungus-farming ant species Mycetosoritis asper. Since the description of the type specimen in 1887, only four additional specimens are known to have been added to the world's insect collections. Its biology is entirely unknown and its phylogenetic position within the fungus-farming ants has remained puzzling due to its aberrant morphology. In 2014 we excavated and collected twenty-one colonies of M. asper in the Floresta Nacional de Chapecó in Santa Catarina, Brazil. We describe here for the first time the male and larva of the species and complement the previous descriptions of both the queen and the worker. We describe, also for the first time, M. asper biology, nest architecture, and colony demographics, and identify its fungal cultivar. Molecular phylogenetic analyses indicate that both M. asper and M. clorindae are members of the genus Cyphomyrmex, which we show to be paraphyletic as currently defined. More precisely, M. asper is a member of the Cyphomyrmex strigatus group, which we also show to be paraphyletic with respect to the genus Mycetophylax. Based on these results, and in the interest of taxonomic stability, we transfer the species M. asper, M. clorindae, and all members of the C. strigatus group to the genus Mycetophylax, the oldest available name for this clade. Based on ITS sequence data, Mycetophylax asper practices lower agriculture, cultivating a fungal species that belongs to lower-attine fungal Clade 2, subclade F.

Introduction about its biology, we conducted field research in the Floresta Nacional de Chapecó, Santa Catarina, Brazil. Here, for the first time, we: (i) describe the male and larva of Mycetosoritis asper, (ii) document its nest architecture and demographics based on twenty-one colonies collected, (iii) report on the phylogenetic position of M. asper within the subtribe Attina by generating new DNA sequences and conducting multilocus phylogenetic analyses, and (iv) report on the identity of its fungal cultivar and the agricultural system to which it belongs.

Field observations and nest excavations
Field work was conducted [18][19][20][21] October 2014 in the Floresta Nacional de Chapecó (henceforth FLONA Chapecó), located between the municipalities of Guatambú and Chapecó in the west of the state of Santa Catarina, Brazil. The FLONA Chapecó is divided into three zones or Glebas: Glebas I and III are located in the municipality of Guatambú, whereas Gleba II is located in the municipality of Chapecó [27,28]. The FLONA Chapecó contains remnants of Atlantic Forest, including Araucaria angustifolia, as well as plantations of A. angustifolia, pines (Pinus elliottii, P. taeda), and Eucalyptus species. The FLONA Chapecó is surrounded by intensively disturbed habitat used mostly for agriculture, pasture, and silviculture [28]. The most common types of soil in the FLONA Chapecó are Cambisols and Latosols. For more information regarding the FLONA Chapecó see ICMbio [28]. The FLONA Chapecó Gleba I, where we located twenty-one nests of Mycetosoritis asper (at 27.10306˚S 52.77898˚W, elevation 595-601 m above sea level), has an estimated area of 1300 ha and is considered a remnant of Atlantic Forest (containing mixed ombrophilous and seasonal deciduous forests), with a mean annual rainfall of 2007 mm and a mean annual temperature of 22˚C [27][28][29].
Foraging activity by workers of Mycetosoritis asper was observed during the day. Foragers were located by baiting the area with Cream of Rice1 cereal spread generously on the ground. The workers carrying the bait were then followed to their nest entrances. Nest entrances were marked with flagging and dug up when a substantial number of nest entrances had been located. Nests were excavated following Schultz [30], Rabeling et al., [31], and Sosa-Calvo et al., [32]. The twenty-one excavated colonies were transferred, using flame-sterilized forceps and spoons, from their subterranean chambers into plastic nest boxes containing a layer of plaster at the bottom, which was saturated with water [32]. Eleven of the twenty-one colonies collected are still maintained alive in artificial nest boxes in the AntLab at the Smithsonian Institution in Washington, DC (Table 1). Fungus garden fragments were isolated and axenically cultured on PDA (potato dextrose agar) medium with three antibiotics (Penicillin G, Streptomycin sulfate, and Chloramphenicol) in the Smithsonian Institution AntLab. Mycelia were transferred from agarose to PDA liquid broth medium with no antibiotics and cultured at 30˚C under constant agitation in a New Brunswick Scientific Series 25 Incubator Shaker, after which the tissues were filtered, lyophilized, and placed into cryo-storage for later DNA extraction.
Nest architecture was recorded, photographed, and measured following Sosa-Calvo et al., ([32]; page 307) Fungus garden fragments and a subset of workers were preserved in 95% ethanol in the field. Fungal vouchers are deposited in the USNM and ant vouchers are deposited in the insect collections of the following institutions:

Material examined
In addition to the material collected by us in the FLONA Chapecó, we examined the type specimen, a dealate queen from Santa Catarina, Brazil, described by Mayr [23], and a worker from Puerto Piray, Misiones, Argentina, described by Emery [26], but in that publication erroneously recorded as collected in Chubut (see below). Other material examined includes a single worker collected in 1957 by F. Plaumann in Chapecó, Santa Catarina, and two workers collected in 1999 by Rogerio da Silva in a Winkler sample in Seara, Santa Catarina. These specimens are deposited in the following institutions:

Morphological measurements and specimen preparation
All measurements were taken to the nearest 0.001 mm and, unless otherwise noted, are in millimeters. Composite images were generated at the USNM Ant Lab using a JVC KY-F75U digital camera mounted on a Leica Z16 APO stereomicroscope attached to a Dell Optiplex GX620 computer. Composite images were assembled using Auto-Montage Pro1 (Version 5.03.0061 BETA) software (Synoptics Ltd.). Wings of males and queens were removed from the left side of the specimen, placed on microscope slides with Euparal mounting medium, and covered with a circular cover glass. The slides were labeled with the name of the species, sex, country, and locality of collection, and the unique USNMENT number of the specimen to which the wings belong. The larvae were dehydrated sequentially through a series of ethanol concentrations to 100% absolute and then critical-point dried in a Balzers CPD-030 using liquid CO 2 at the Scanning Electron Microscopy (SEM) Lab in the SI-NMNH and mounted on aluminum stubs. The prepared larvae, as well as a worker, a queen, and a male, were sputter-coated with 60:40 wt% gold: palladium alloy on a Cressington Scientific 108 auto/SE sputter coater to a thickness of 20-25 nm. Scanning Electron Micrographs (SEMs) of these specimens were generated using a Philips XL-30 ESEM with Lanthanum Hexaboride (LaB6) source and with a backscatter detector.
Anatomical abbreviations are as follows:

Molecular phylogenetics
Ant and fungal cultivar DNA extraction, amplification, and sequencing were conducted at the Laboratories of Analytical Biology (LAB) at the National Museum of Natural History, Smithsonian Institution, Washington, DC. Genomic DNA was extracted using the Qiagen DNEasy Blood and Tissue kit (Quiagen, Inc.) for the ants and the Plant DNeasy kit (Quiagen, Inc.) for the fungus. For the ants, five nuclear protein-coding genes (EF1α-F1, EF1α-F2, wg, LW Rh, and TOP1) were amplified and sequenced following the methodology outlined in previous studies [12,39,40]. For the fungal cultivar, a ribosomal gene fragment, internal transcribed spacer (ITS), was amplified and sequenced following [41][42][43]. New sequences generated for this study are deposited in GenBank under accession numbers KY809160-KY809180 for the HL Head Length: in full-face view, the maximum vertical distance from the posteriormost margin of the head to the midpoint of the anterior clypeal margin (clypeal apron), excluding the mandibles.

HW
Head Width: in full-face view, the maximum horizontal width of the cephalic capsule excluding the eyes.

ML
Mandible Length: in full-face view, the maximum diagonal-line distance from the base of the external mandibular insertion to the apical tooth. When mandibles were closed, the mandible on top was measured. When mandibles were open, then the left mandible was measured. fungal cultivar and KY828479-KY828592 for the ants. Nexus and tree files can be found in Dryad (http://datadryad.org/resource/doi:10.5061/dryad.64r0j). Ant DNA sequences, consisting of~3.3 kbp, were added to the aligned data set of Schultz and Brady [39] and Sosa-Calvo et al., [7] and aligned first by eye in Mesquite [44] and subsequently by using MAFFT v7.017 [45][46][47] as implemented in Geneious R9 v8.1.8 [48]. Data were partitioned and modeled using the program PartitionFinder v1.1.0 [49] under the Bayesian Information Criterion (BIC) with 15 data blocks consisting of the first, second, and third codon positions of each of the five gene fragments and with a user tree resulting from an unpartitioned maximum-likelihood best-tree analysis conducted in RAxML v.8.2 [50]. The eight partitions and models identified by PartitionFinder were employed in Bayesian analyses using MrBayes 3.2.2 [51] with nucmodel = 4by4, nruns = 2, nchains = 8, samplefreq = 1000, and 20 million generations, with a burn-in of 2 million generations. To address known problems with branchlength estimation in MrBayes [11,[52][53][54][55][56], we set brlenspr = unconstrained:Exp (100). Burn-in, convergence, and stationarity were assessed using Tracer, version 1.5 [57], by examining potential scale reduction factor values and.stat output files in MrBayes, and by using Bayes factor comparisons of harmonic-mean marginal likelihoods of pairs of runs with standard error estimated using 1,000 bootstrap pseudoreplicates in Tracer 1.5 [57], which employs the weighted likelihood bootstrap estimator of Newton and Raftery [58] as modified by Suchard et al., [59].

MSLI
Fungal ITS sequences for M. asper were added to a preexisting data set of agaricaceous (Basidiomycota: Agaricales: Agaricaceae: Leucocoprineae) ant-associated and free-living fungi and aligned in MAFFT, producing a matrix consisting of 506 taxa and 1281 characters, including indels. Data were partitioned and modeled using the program PartitionFinder 2.0 under the kmeans algorithm [60], which does not require initial data blocks, and the corrected Akaike information criterion (AICc). The two partitions and models identified by PartitonFinder were employed in Bayesian analyses using MrBayes 3.2.2 as described above for the ant phylogenetic analyses, except that burn-in was set to 5 million generations.

Systematics
Etymology of "Mycetosoritis". Despite a tradition of treating it as feminine, the genus name "Mycetosoritis" should be regarded instead as masculine and should be combined with masculine-form adjectives such as "asper" and "explicatus." W. M. Wheeler [15] originally described Mycetosoritis as a subgenus of Atta. A rereading of Wheeler's [15] description of Mycetosoritis hartmanni provides no guidance on the gender of Mycetosoritis because (i) "hartmanni" is genitive and (ii) the adjective "aspera," which Wheeler uses when he transfers the former Cyphomyrmex asper into Atta (subgenus Mycetosoritis), matches the (feminine) gender of Atta, the genus name, and not that of Mycetosoritis, the subgenus name. Because there is thus no indication of gender in the original description of Mycetosoritis, Section 30.1.4.2 of the International Code of Zoological Nomenclature [61] applies, requiring the use of the masculine gender for genus-group names with endings of ambiguous gender for which the author did not indicate the gender either explicitly or via a clear adjectival species-group name. The assertion by G. Wheeler [62] that the second part of "Mycetosoritis" is derived from "Sorîtis," an alternate name for the Greek goddess Ceres, is not, even if true, sufficient to supersede the ICZN rule. Wheeler's assertion is, moreover, unsupported based on the opinion (pers. comm.) of E. Adler, Associate Professor of Classics, University of Maryland, College Park, who (i) confirmed that the noun ending "-is" can be masculine or feminine and (ii) could find no reference to a god or goddess "Sorîtis" in The Oxford Latin Dictionary [63]

or in Liddell and Scott's
Greek-English Lexicon [64].
Mycetophylax Emery (1913). The multilocus analyses presented here (Fig 1), as well as phylogenomic analyses of 950 UCE loci [14], indicate with strong support that the species Mycetosoritis asper and Mycetosoritis clorindae are members of the genus Cyphomyrmex as currently defined and, more specifically, of the C. strigatus group. Both data sets further indicate that, as currently defined, the genus Cyphomyrmex is paraphyletic with respect to Mycetophylax, Mycetagroicus, and the higher Attina, and that the C. strigatus group, as currently defined, is paraphyletic with respect to the species Mycetophylax conformis, Mycetophylax morschi, and Mycetophylax simplex, which are derived members of the clade containing the C. strigatus group. Based on these results, and in the interest of taxonomic stability, we here recognize the C. strigatus group as a separate genus. Mycetophylax conformis is the nominal species for the genus name Mycetophylax [19] and "Mycetophylax" is the oldest available species-group name for the clade containing the C. strigatus group. We therefore transfer all members in the C. strigatus group, including Mycetosoritis asper and Mycetosoritis clorindae, to the genus Mycetophylax.
Taxonomic synopsis of the species in Mycetophylax Emery. The species list presented here is modified from Kempf [17] and Klingenberg    Diagnosis. Inner margin of mandibles with 6-7 teeth; preocular carina extending posterad to cephalic corner, forming a well-developed antennal scrobe (shared with members of the former Cyphomyrmex strigatus group, Mycetophylax clorindae, and Mycetophylax morschi, but not with Mycetophylax conformis and Mycetophylax simplex, in which the scrobe is secondarily reduced); anterior margin of antennal scape with a broad carina, gradually expanding apically; body covered with short, erect, simple hairs; single median pronotal tubercle present, well-developed (shared with members of the former Cyphomyrmex strigatus group, but vestigial in species of the former Mycetophylax s.s.); node of petiole with a pair of tubercles or spines; gastral tergite I strongly rugose and with small tubercles from each of which arises a single, short, erect hair.
Description. The worker of Mycetophylax asper was first described by Emery [26] from a single specimen collected in Argentina. Here, we complement Emery's description of the worker based on the specimen studied by him and additional workers from multiple nest series collected in the municipality of Chapecó, Santa Catarina, Brazil.

Fig 1. Phylogeny of fungus-farming ants based on Bayesian analysis of five nuclear protein-coding genes. Mycetophylax asper is indicated in red. Red box indicates our newly expanded definition of the genus Mycetophylax (see text for details).
https://doi.org/10.1371/journal.pone.0176498.g001 Rediscovery of the enigmatic fungus-farming ant Mycetosoritis asper Mayr and 4A); dorsum of antennal scapes covered with erect, simple hairs, leading margin of scapes with decumbent simple hairs, posterior margin of scapes with appressed simple hairs. Palpal formula 4,2 ( Fig 4D).
Mesosoma: In fronto-dorsal and lateral view, pronotum with single median anterior pyramidal tubercle, connected by thin carinae to robust lateral pronotal tubercles, somewhat separating pronotum from mesonotum (Figs 2D and 2F; 3A and 3B; 4G and 4J); dorsum of pronotum rugulose-foveolate (best seen with high magnification, see Fig 4G and 4J); pronotum lacking humeral tubercles; in lateral view, inferior corner of pronotum armed with  anterior mesonotal margin broadly convex; in dorsal view, area circumscribed by lateral and posterior mesonotal tubercles slightly concave and weakly sculptured; in lateral view, posterior mesonotal tubercles triangular ( Fig 4G); in lateral view, mesopleural margin (katepisternum) with thin but conspicuous carina ( Fig 3A); metanotal groove deep. In lateral view, anterior portion of propodeum with pair of small tubercles (Figs 2D, 3A and 4G); posterior propodeal tubercles larger and acute (Figs 2D, 3A and 4G); anterior and posterior propodeal tubercles connected by lateral carinae; declivity of propodeum shorter than base of propodeum and lacking lateral carinae; in dorsal view, lateral face of propodeum with conspicuous carina, arising anterior to propodeal spiracle to form small but conspicuous tubercle; second, smaller, rounded tubercle arising posterior of spiracle, near the margin of metapleural gland bulla; propodeal lobes vestigial to absent. Hind femur and, to a lesser extent, mid femur with pair of conspicuous ventral carinae, produced in the basal one-third into strong ventro-posterior lobe (Fig 5E). Outer margin of mid and hind tibia with short, erect hairs (Fig 5E).
Color: Head and body mostly yellowish to ferrugineous; dorsum of head, mesosoma, and gaster tend to be darker in color. Pilosity as in the worker.
Wings smoky, covered with minute pilosity. Forewing lacking pterostigma and with five closed cells present. Hindwing with reduced venation, a single closed cell present (Fig 6A).
Legs very long, length of hind femur longer than mesosoma (HFL 1.25-1.60, WL 1.18-1.39, HFI 106-116); ventral margin of mid and hind femur with erect simple hairs on basal half, apical half lacking hairs or with very short, appressed hairs.
Fore and hind wings as in the queen (Fig 6B).
Color: Antennal funicular segments, mandibles, and tarsomeres yellowish to light brown, rest of body dark brown (Fig 3E, 3F  Larva: Description based on SEM study of three prepupal worker larvae from a single nest (AJ141020-02).
Body profile "attoid" sensu Wheeler [79] and Wheeler & Wheeler [80], i.e., longitudinally curved, bean-shaped, with ventral profile shorter than dorsal (Fig 7A). Thoracic-abdominal articulation absent, thoracic intersegmental constrictions superficial, deep lateral depressions associated with abdominal spiracles absent, and leg vestiges present and visible as open slits ventrally on thorax. Dorsal and lateral body surfaces without setae ( Fig 7A); setae on head and venter mostly simple, a few with multifurcate tips (Fig 7B). Genal lobes apparently absent. Supra-antennal and supraclypeal setae absent, four setae on each gena and two on clypeus. A few papilliform spinules present on head, restricted to clypeus and genae. Labrum monolobate, narrow, and inflated, with two distinctly setiform anterior setae (Fig 7B-7D). Mandibles fleshy and subconical. Spinules on mandibles densely covering entire surface (Fig 7C). Mandible with distinct, undivided apical tooth and with no subapical teeth. Mandibular gnathobases absent. Basal portion of maxilla fused with head capsule and maxillary palp widely removed laterad from galea. Galea remarkably enlarged and covered with denticles ( Fig 7C). Maxillary palp digitiform, maxillary accessory palpal sensillum apparently absent (Fig 7C). Two setae between galea and palp ( Fig 7C). Labium feebly protruding, lateral sericteral protuberances absent, labial palps reduced to sensilla. Spinules present only on anterior surface of labium dorsal to sericteries. Hypopharyngeal spinules densely distributed and predominantly unidentate (a few two-toothed spinules present (Fig 7D)). Ventral surface of first thoracic segment lacking ventromedian lobe and papilliform spinules, and with only four long, simple hairs on ventral surface. Ventromedian surfaces of first and second thoracic segments bearing multiple multidentate spinules (Fig 7D). Second and third thoracic segments each with six long, simple hairs ventrally. First and second abdominal segments each with a non-lobiform ventromedian protuberance, protuberance on first abdominal segment more pronounced than that on second ( Fig 7D). Four long, simple hairs arise ventrally on each side of protuberance on abdominal segment one, and two to three hairs (varying between specimens) arise on each side of protuberance on abdominal segment two (Fig 7D). Abdominal segment three with one or two pairs of setae; ventral setae absent on abdominal segments four to nine. A single pair of setae anterior to anal opening and another pair lateral to opening, all four setae arising on abdominal segment ten (Fig 7F and 7G). Ventral anal lip absent. by rootlets. Most of the fungus gardens extracted from these colonies were of a light yellow color and honeycomb-like in appearance, with some parts of the garden greenish in color due to the presence of newly added substrate (Fig 8D and 8E). In six of the larger colonies, a wet mud pellet was found at the bottom of the chamber, possibly consisting of dirt or refuse ( Fig  8E), differing from the infrabuccal pellet piles described by Little et al., [83]. Demography. Colonies of Mycetophylax asper are relatively small, containing up to 100 workers. A single, dealate queen was found in every excavated colony, indicating that M. asper is monogynous. As of April 2017, eleven of the twenty-one colonies remain alive in artificial nest boxes at the USNM (see Table 1). Observations during a one-year period in the laboratory suggest that Mycetophylax asper produces reproductive forms from January to April, corresponding to the austral summer when temperatures are warmest.

Ant morphology
Morphologically, Mycetophylax asper shares with most members of the genus Mycetophylax, as defined here, the circumscribed antennal scrobes formed by the joining of the frontal carinae and the posterad-directed preocular carinae at the occipital corners (secondarily reduced in Mycetophylax conformis and Mycetophylax simplex) (Figs 2B and 2D; 3A and 3G; 4A and 4D), six or more mandibular teeth (Fig 5A), and the presence of a single mid-pronotal tubercle [17,75], secondarily reduced in the species M. conformis, M. morschi, and M. simplex. In the males of Mycetophylax asper the antennal scapes are long (SL 0.52-0.63, SI 104-110), longer than length of funicular segments I-III combined, a condition that is shared with other neoattine species and that contrasts with that of the paleoattines, in which the male antennal scape is shorter than the combined length of funicular segments I-III [19,35]. In addition, the males of Mycetophylax asper, which were unknown at the time of Kempf (17) revision of the Cyphomyrmex strigatus group, have 12-segmented antennae, a deviation from the ancestral condition of 13 found in most fungus-farming ants. Although this condition occurs in other members of the former Cyphomyrmex strigatus-group (Mycetophylax faunulus, Mycetophylax auritus) and in Mycetophylax conformis [34], it remains unclear whether this shared reduction is due to homology or homoplasy because (i) Mycetophylax morschi and Mycetophylax simplex males are reported to have 13-segmented antennae [34] and (ii) the 12-segmented condition has arisen independently multiple times in the Attina, including in Mycetagroicus inflatus, species of Sericomyrmex, Trachymyrmex opulentus [23,34,36], and some social parasites [84][85][86].
Based on larval characters, Mycetophylax asper is clearly a member of the Neoattina and is more closely related to members of the former Cyphomyrmex strigatus group (= Mycetophylax s.l.) than to those of the C. rimosus (yeast-farming) group. Character states shared with all other fungus-farming ants include the absence of a thoracic-abdominal articulation, the absence of deep lateral depressions associated with the abdominal spiracles, the superficiality of the thoracic intersegmental constrictions, the leg vestiges visible as open slits ventrally on the thorax, the mandibles fleshy and subconical, and, shared with most other attines, the absence of setae on the dorsal and lateral body surfaces (Fig 7A). Character states shared with other neoattines include the fusion of the basal portion of the maxilla with the head capsule combined with the position of the maxillary palp, widely removed laterad from the galea; the labium feebly protruding; the absence of lateral sericteral protuberances; and the reduction of the labial palps to sensilla.
A close relationship of Mycetophylax asper to Mycetophylax s.l. (i.e., as here redefined) is indicated by the states of no less than six larval characters. First, the presence of non-lobiform ventromedian protuberances on the second and third abdominal segments is similar to the condition observed in the species Mycetophylax auritus and M. faunulus [87], whereas it differs from the condition in the yeast-cultivating C. rimosus-group species in which the first abdominal segment (and, in some species, the second and even third abdominal segments) bears a triangular, lobiform appendage [79,87]. Second, the hypopharyngeal spinules are predominantly unidentate, a condition previously described only in M. auritus and M. faunulus [87] (Fig 7D). Third, a single pair of setae arise anterior to the anal opening, another pair arise lateral to the opening, and all four setae arise on abdominal segment ten (Fig 7F and 7G). This pattern is apparently related to the conditions observed in both the Cyphomyrmex strigatus-(as formerly defined, = Mycetophylax s.l.) and Cyphomyrmex rimosus-group species [87], in which there are four ventral setae, but it may also be related to the condition reported in Mycetophylax conformis, in which there are only two ventral setae [87]. Fourth, the ventral anal lip is absent, differing from the condition in most but not all other known Cyphomyrmex species, including some members of the newly defined Mycetophylax, but shared with M. auritus and M. conformis. Fifth, the presence of a distinct mandibular apical tooth is shared with all known Mycetophylax s.l. species but not with C. rimosus-group species, in which the apical tooth is reduced to a spinule. Sixth, the spinules on the head are papilliform as in other known Mycetophylax s.l. species, rather than serrate as in most known C. rimosus-group species.
The larva of Mycetophylax asper differs from those of other neoattines in having the galea remarkably enlarged and covered with denticles ( Fig 7C). It differs from other Mycetophylax species in having the hypopharyngeal spinules densely distributed, an apparent symplesiomorphy shared with most attine ants, including members of the C. rimosus group. The larva of

Phylogeny
Results of the ant molecular phylogenetic analyses, which, compared to those of Schultz and Brady [39] and Sosa-Calvo et al., [7], include an additional nuclear gene fragment and additional taxa, place Mycetophylax asper in the former strigatus group of the genus Cyphomyrmex (rendered paraphyletic with respect to Mycetophylax) with strong support (Fig 1). Moreover, Mycetophylax asper is closely related to the recently described species Cyphomyrmex andersoni [75] from Central America (Fig 1), here transferred to Mycetophylax. This curious distribution suggests (i) that either or both Mycetophylax asper (in southern Brazil) and M. andersoni (in Mesoamerica) are more widespread than is currently known or (ii) that the distribution of their unknown most recent common ancestor is or was at some time in the past more widespread than either species is today. This pattern of peripherally and disjunctly distributed relict taxa is common across multiple sister clades in the Neoattina and is certainly deserving of further inquiry [14].