Phylogeny and biogeography of Asthenopodinae with a revision of Asthenopus, reinstatement of Asthenopodes, and the description of the new genera Hubbardipes and Priasthenopus (Ephemeroptera, Polymitarcyidae)

Abstract The Neotropical species of Asthenopodinae are revised in a formal phylogenetic context. The five known species of Asthenopus Eaton, 1871, together with other five new species were included in a cladistic analysis using morphological characters (continuous and discretes). Representatives of the Afro-Oriental group of the subfamily (Povilla Navás, 1912 and Languidipes Hubbard, 1984) were also included to test the monophyletic hypothesis traditionally accepted for the group. Additional taxa representing the other subfamilies of Polymitarcyidae were incorparated: Ephoron Williamson, 1802 (Polymitarcyinae) and Campsurus Eaton, 1868, Tortopus Needham & Murphy, 1924 and Tortopsis Molineri, 2010 (Campsurinae). A matrix of 17 taxa and 72 characters was analyzed under parsimony resulting in a single tree supporting the monophyly of the subfamily Asthenopodinae. Other results include the monophyly of the Afro-Oriental taxa (Povilla and Languidipes), the paraphyletic nature of Neotropical Asthenopodinae, and the recognition of four South American genera: Asthenopus (including Asthenopus curtus (Hagen), 1861, Asthenopus angelae de Souza & Molineri, 2012, Asthenopus magnus sp. n., Asthenopus hubbardi sp. n., Asthenopus guarani sp. n.), Asthenopodes Ulmer, 1924, stat. n. (including Asthenopus picteti Hubbard, 1975, stat. n., Asthenopodes traverae sp. n., Asthenopodes chumuco sp. n.), Priasthenopus gen. n. (including Priasthenopus gilliesi (Domínguez), 1988, comb. n.), and Hubbardipes gen. n. (including Hubbardipes crenulatus (Molineri et al.), 2011, comb. n.). Descriptions, diagnoses, illustrations and keys are presented for all Neotropical taxa of Asthenopodinae (adults of both sexes, eggs and nymphs). Additionally a key to the subfamilies and genera of Polymitarcyidae is included. A quantitative biogeographic analysis of vicariance is presented and discussed through the study of the “taxon history” of the group.


Introduction
Polymitarcyidae Banks (Ephemeroptera) have long attracted the attention of freshwater biologist because their nymphs burrow tunnels in submersed wood, live in aquatic plants and sponges, and some inorganic sediment as clay, mud or sand (Arndt 1938, Hartland-Rowe 1958, Sattler 1967. Vejhabongse (1937) reported the damage caused by larvae of Povilla Navás, 1912 (Asthenopodinae) in woody structures. Leal et al. (2003) studied the capacity of Campsurus Eaton, 1868, larvae as bioturbators in soft mud bottom of Amazonian lakes; and Tortopsis Molineri, 2010, nymphs play an important role in the erosion of river clay banks (Molineri, unpubl.).
The family Polymitarcyidae is composed of three subfamilies (Polymitarcyinae, Asthenopodinae, and Campsurinae, Traver 1954, Bae andMcCafferty 1995), and shows a broad distribution in the Holartic, Paleartic, Oriental and Neotropical regions (McCafferty 2004). Ogden et al. (2009) included one representative of each of these subfamilies in a combined molecular and morphological study of the entire order Ephemeroptera -in their work the traditionally known relationships were recovered with Polymitarcyinae spliting first and then Campsurinae as sister to Asthe-

Morphological characters
Characters are scored from external morphological features of adults (male imago unless otherwise indicated), nymphs and eggs. Dissected parts of the nymphs and adults were mounted on microscope slides using Canada Balsam, except wings that were mounted dried. All the material is preserved in ethyl alcohol 96%. Photographs were taken using a NIKON SMZ-10 stereomicroscope or a microscope, with a Nikon D5000 digital camera; some pictures were modified with Combine ZP (Hadley 2010). Line drawings were done using a camera lucida attached to a microscope. Structures used for SEM study were dehydrated in a graded ethanol series, dried by the critical point method, sputter coated with gold and observed with a JEOL 35 CF scanning electron microscope.
The measures and ratios used in some characters are explained and illustrated to permit repeatability. To score variation in shape of genitalic structures (chars. 12-17) some measures were defined (see Appendix 1). Ratios between some of them were used instead of the original data to avoid the spurious differences occasioned by the great variation in size common in the group. Characters 0 to 11 also are ratios or counts to represent characteristics commonly used in the taxonomy of the group, for example the length of forelegs of male and its relation to other structures. Fore and hind wings are abbreviated FW and HW respectively throughout the text. A complete list of the characters, their definitions and character states is given in Appendix 1.

Cladistic analysis
Searches were conducted in TNT (Goloboff et al. 2008) under implied weights (k = 3), and using the "implicit enumeration" command (under "Analyse"). Implied weighting was suggested to ameliorate the problems of scaling in continuous characters (Goloboff et al. 2006). By the implicit enumeration, the complete universe of possible trees for the matrix is calculated, then picking the shortest ones. All characters were treated as non-additive except for continuous characters (chars. 0 to 26). Absolute and relative Bremer supports were calculated with 2000 suboptimal trees (up to 8 steps longer than shortest tree), with the commands "suboptimal" (under "Analyse") and "Bremer Supports" (under "Trees"). Frequency difference (GC, Goloboff et al. 2003), using 400 replications of symmetric jacknifing, was also calculated as a measure of group support. The optimizations of individual continuous characters given in Figs 21 and 22 were obtained in TNT using the commands "Optimize: Characters: Character Mapping".

Biogeographical methods
All the available geographic records (448 records for 17 taxa) of the species or genera included in the phylogeny were compiled. Concerning Asthenopus species, only records from specimens (or photographs) revised by us were included, since much confusion exists in the literature in relation to A. curtus and similar species. All records are exact points of occurrence except some for Ephoron species from North America and Europe, which were only roughly approximated from the maps in McCafferty (1975) and de Jong (2012), respectively. Some of the records (ca. 100) for Tortopsis, Ephoron and Povilla were downloaded from Global Biodiversity Information Facilities (http://www. gbif.org/, last accesed February 10th 2014).
The biogeographical analysis was performed through spatial analysis of vicariance ), a method that reconstructs taxon biogeographic history by looking at disjoint sister pairs in a given phylogeny. Besides a cladogram, it uses as input the distributional records of the terminals; and is implemented in the software VIP (Vicariance Inference Program) available at http://www.zmuc.dk/public/phylogeny/vip (Arias 2010). A grid of 1° × 1° was used (maximum fill = 0) in a world vegetation map (obtained in http://neo.sci.gsfc.nasa.gov) to represent distributions as absence/presence data in each cell. The "OR reconstruction" was used under the default settings except that cost of distribution removal was set to 1.5, and cost of partial removal to 0.75. Reconstructions using different grid size (5° × 5°) and costs were also conducted for comparative purposes, but results were similar.

Phylogeny and taxonomic status of genera and species
Parsimony under implied weights (IW) resulted in a single tree (Figs 1-2), with an adjusted homoplasy of 14.64 and a total fit of 54.36. Neotropical Asthenopodinae was not recovered as a monophyletic group: Asthenopus s.s. is sister to the Afro-Oriental clade (Povilla-Languidipes). Asthenopodes is revalidated as a distinct genus since its type species (A. picteti Hubbard) was recovered in a group with other two new species. Asthenopodes stat. n. presents defining characters in all stages (see below). The species described from adults by Molineri et al. (2011) as Asthenopus crenulatus is now known in all the stages (see below) and a new genus Hubbardipes is proposed for it. Hubbardipes gen. n. is supported as sister to the remaining Asthenopodinae (i.e., excluding Asthenopodes). Another new monotypic genus, Priasthenopus gen. n. is proposed for A. gilliesi, only known from adults and eggs. Priasthenopus is sister to the clade Povilla + Asthenopus.
The type species of Asthenopus (A. curtus) forms a clade with A. angelae and three new species. Synapomorphies of each clade are numerous; changes on continuous and discrete characters are presented separately in Figs 1 and 2, respectively. For details on changes see generic diagnosis and Appendix 2. Group support is strong for most clades (Fig. 3).
Summarizing, the taxonomic changes needed to adjust formal classification with the phylogeny include: 1) revalidation of Asthenopodes as a distinct genus and the corresponding new combination for its type species (Asthenopodes picteti), 2) the erection of Hubbardipes gen. n. and the new combination Hubbardipes crenulatus (Molineri et al.) for its single species, and 3) the erection of Priasthenopus gen. n. and the new combination P. gilliesi (Domínguez). This phylogeny also was the framework used to describe the five new species in the corresponding genus (see "Descriptions"): Asthenopodes traverae, Asthenopodes chumuco, Asthenopus hubbardi, Asthenopus magnus, and Asthenopus guarani.
An additional taxonomic change is proposed for Campsurus paraquarius Navás (1920) that undoubtedly pertains to Asthenopodinae as recognized by Lestage (1923) and Kluge (2004), most likely it is related with P. gilliesi (Domínguez). C. paraquarius is here considered a nomen nudum (see discussion under Priasthenopus below).
Asthenopodinae is defined by numerous derived character states (see Appendix 2 for details) including: shorter male foretibia, wider forceps, shorter female cerci, female sternum VIII with anteromedian keel; eggs with large and small disk-like chorionic structures; nymphs with wider mandibular tusks, tusk with two or three apical denticles (right and left tusks respectively) and with a small basal tubercle on outer dorsal surface, occipital region strongly expanded and convex, and foretarsal claws with marginal row of denticles. It is interesting to note some characters that simultaneously change in opposite "directions", showing a certain tendency in Asthenopodes but the contrary in it sister group (remaining Asthenopodinae). For example the following features define Asthenopodes and its sister group: the male foretibia is thinner distally (vs wider in the remaining Asthenopodinae, Figs 20, 22A), median remnant of styliger plate shorter (vs longer), larger HW (vs smaller, Fig. 22B). Other synapomorphic changes associated to Asthenopodinae excluding Asthenopodes are given in Appendix 2.
The sister relation between Priasthenopus and the clade Povilla + Languidipes + Asthenopus is supported by (see Appendix 2, node 23): additional shortening of foreleg,  Fig. 1) showing changes on discrete characters (number above and below circles are character and state number, respectively, see Appendix 1). Black circles indicate unique apomorphies.
shorter and less anastomosed cross veins in the margin of FW, additional increase in forceps width, even wider and shorter female prothorax.
The clade formed by Asthenopus together with the Afro-Oriental genera share (Appendix 2): even wider forceps, shorter and less anastomosed intercalary veins in both wings, slightly shorter female pronotum, female sternum VIII with long and slender anteromedian keel, sockets small, and nymphal mandibular tusks with large basal tubercle in inner margin (this last state may be present also in Priasthenopus gilliesi, but nymphs remain unknown).
Povilla and Languidipes form a monophyletic group because both show shorter forelegs in male, blade-like penes relatively straight not curved inward, MP 2 shorter than IMP in FW (non-unique, also in Campsurinae), IMP connected to MP 1 , forceps two-segmented, pedestals much narrower basally, and nymphal left mandibular tusk with an additional apical denticle.

Campsurinae
In spite of Campsurinae being only marginally represented in the matrix, we consider useful to note the derived character states defining the subfamily and genera (see Appendix 2 for details): much shorter forelegs in male, slender forceps, ICu 1 joins hind margin of FW close to tornus while ICu 2 joins it on basitornal margin, intercalary marginal veins absent or if present not anastomosed, median remnant of styliger plate absent, double penial arm articulation (sternum IX and base of pedestal), nymph with dorsum of head with dense patches of short setae on frons. Campsurus is very sparsely represented in the matrix (only 2 spp from more than 40 known) so only the character states defining Tortopus + Tortopsis may have some interest: slender penes, male foreleg slightly shorter, foretarsal segment 5 apically trilobed, intercalary marginal veins rarely present in male FW, HW anal area with many crossveins forming a network, forceps two-segmented.

Evolution of selected features
Variation in the relative length of male foreleg segments has been repeatedly used by different authors in the taxonomy of the family. One of our characters concerning this variation (char. 1, ratio length tarsal segment 2/tibia) presented many changes: most important are those defining the genus Asthenopus and the pair Asthenopodes traverae -A. picteti, indicating opposite changes in these nodes (encircled arrows in Fig. 21A).
In the nymphs, the robustness of the mandibular tusks is an important feature distinguishing different subfamilies. The optimization of character 27 (ratio length/ width of tusk, Fig. 21B) shows an apomorphic change in the base of Asthenopodinae, indicating that the slender tusks of Campsurinae are plesiomorphic.
Trends in some charecteristics are markedly manifested, for example the width of male foretibia (char. 0 ratio width tibia/tarsal segment 2, Fig. 22A) decreases at the base of Asthenopodes and become thinner again later in Asthenopodes picteti. On the contrary, it becomes more robust in the base of the sister clade to Asthenopodes (the remainning Asthenopodinae) and wider again in Hubbardipes crenulatus (Fig. 22A).
The length of male foreleg is another character commonly used to define groups (here represented as the ratio length of FW/foreleg, Fig. 22C): a paulatine shortening of male foreleg in two lineages (Fig. 22C) and female pronotum (in just one of them, Fig. 22D) were obtained in the phylogeny. Both trends are only coupled (i.e., show simultaneous variation in the same node) in the clade including Priasthenopus gilliesi, Povilla and Asthenopus s.s. It is interesting to note that a character commonly used to define the subfamily Asthenopodinae (ring-like prothorax in the male) did not present changes in our phylogeny because of large intraspecific variation. Another character used as defining the subfamily (parallel ICu veins in forewings) is a plesiomorphy (the apomorphic state is present in Campsurinae).
General shape of the forceps in male adults shows opposing tendencies, becoming slender in Campsurinae but stouter in Asthenopodinae (Fig. 21C). The reduction in the number of segments from three in Ephoron, to one in the base of Campsurinae + Asthenopodinae is independently reversed twice (in the ancesters of Tortopus + Tortopsis, and Povilla + Languidipes). Independent losses involve also to Languidipes and Campsurinae in relation to median remnant of styliger plate. A reduction in the length and anastomosis of marginal intercalary veins (reduction of a marginal archedictyon) is present independently in both subfamilies (Campsurinae and Asthenopodinae).

Geographical comments
Polymitarcyidae is widely distributed (Fig. 23), including tropical and temperate areas, with the exception of Australia and New Zealand (Kluge 2004, McCafferty 2004. Polymitarcyinae (Ephoron and Eopolymitarcys) is widely distributed in the Holarctic, Ethiopian and Oriental regions, sharing southern portions of its range with selected species of the other subfamilies (Asthenopodinae in Africa, and Campsurinae in North America, Fig. 23). Neotropical Asthenopodinae is not a monophyletic group, because Asthenopus s.s. is sister to the Ethiopian-Oriental clade (Povilla and Languidipes). Campsurinae is mainly a Neotropical group, with only 6% of its species reaching the Nearctic (squares in Fig. 23).
The spatial analysis of vicariance found 6 disjoint sister pairs (

Key to the subfamilies and genera of Polymitarcyidae
A key to the subfamilies of Polymitarcyidae is presented because new species described here show some of the characteristics (e.g., many crossveins, numerous anastomosed marginal intercalaries, etc.) previously used to diagnose other subfamilies. A key to the genera is proposed to include new or recently described genera (Tortopsis, Hubbardipes and Priasthenopus) or those raised to generic status (Languidipes, in Baumgardner et al. 2012, and Asthenopodes, here). The triangullar cell in Cu sector mentioned by Baumgardner et al. (2012) as diagnostic for Languidipes is not used in the key because it is also present in some Asthenopus females.
Key to the species of Asthenopodes and Asthenopus are given after each generic section (see below). Please note that in male genitalia, the pedestals are considered part of the styliger plate (Kluge 2004), thus number of forceps segments does not include them.
Key to adults (imagos and subimagos, except measures and ratios aplicabble only to imagos) and eggs of Polymitarcyidae

1
FW with vein Sc ending before the tip of the wing, its apical portion not curved posteriorly; cubital area of FW broadly expanded, usually with 3-5 intercalaries and many cross veins and marginal intercalaries; FW with MA fork 1/3 or more distance from wing base (i.e., a long MA stem is present); HW with convex intercalary between R 1 and Rs field (x.i. in Fig. 16  Penes blade-like (Fig. 19D, G); female sternum 8 with anteromedian keel basally swollen (Fig. 19C, E-F); egg without polar caps, chorion mainly smooth but few disk like structures may be present (Fig. 19K)  Penes with many spines on outer subapical margin (Fig. 7A); male foretarsal segment 1 partially fused to tarsal segment 2 (Fig. 20D); female sternum VIII with a protruding subcircular anteromedian structure (figures 12-13 in Molineri et al. 2011); eggs with the space between large plates completely covered by smaller plates (Fig. 7D-E  Penes apically rounded or slightly pointed but never with a spine (Fig. 8A-C); female sternum VIII with a short blunt keel with larger sockets (Fig. 8J); eggs with polar caps formed by 14-16 tightly twisted threads (Fig. 8K)

Discussion
Taxonomy and phylogeny Lestage (1922) reported a concise summary of the meandering story of the early taxonomic stages of Asthenopus: Asthenopus was erected by Eaton (1871) for the species Asthenopus curtus (Hagen) and also for A. dorsalis (Burmeister). Later Eaton (1883), without mentioning Asthenopus, treated these species at Campsurus. Asthenopus is not mentioned in the systematic literature until Ulmer (1921) reinstated it as a valid name for Asthenopus albicans Pictet (nec Percheron), A. curtus (Hagen) and A. amazonicus (Hagen 1888), but treating A. dorsalis as a Campsurus. Shortly after, Ulmer (1924) erected Asthenopodes for A. albicans. Ulmer's consideration of A. albicans as a species authored by Pictet was an error (Hubbard 1975), and it was renamed by this last author as Asthenopodes picteti Hubbard. In 1978 Berner synonymyzed A. curtus with A. amazonicus. Thus, Neotropical Asthenopodinae was at that moment known by two species: Asthenopus curtus (Hagen) and Asthenopodes picteti Hubbard. Domínguez (1988) described Asthenopus gilliesi, a species showing some characters shared by Asthenopus and Asthenopodes and based on this Hubbard and Domínguez (1988) combined both genera. More recently, two species were added to Asthenopus: A. crenulatus Molineri, Cruz & Emmerich (2011) and A. angelae de Souza & Molineri (2012). Ulmer (1942) discussed many characters to differentiate Asthenopus from Campsurus, most of them measures and ratios of different structures (prothorax, forewing veins, legs, forceps). In contrast, Ulmer (1942) found only two characters to distinguish Asthenopus from Povilla, an African Asthenopodinae described by Navás (1912): 1) ICu 2 free at base or joined by crossveins to nearby veins in Asthenopus, but ICu 2 springing from CuP in Povilla; and 2) penis lobes curved and cylindrical in Asthenopus but straight and rod-shaped in Povilla. The first character is present in P. adusta but not in the other species of Povilla (including Languidipes corporaali), so it is not useful at the generic level. The second, may still be used to distinguish both genera, but it is not useful to separate Hubbardipes gen. n. from Povilla (both showing blade-like penes). Domínguez (1988) presented a more detailed description concerning these genera, and reported that Povilla differs from Asthenopus because the former shows: 1) forceps 2-segmented (1-segmented in Asthenopus, note that here the genitalia was reinterpreted following Kluge 2004 and the basal segment or pedestal is considered part of the styliger plate), 2) penes bladelike (cylindrical in Asthenopus), 3) bifurcation Rs from base to margin: 1/10 (2-2.5/10 in Asthenopus), and 4) MP 2 shorter than IMP (subequal in Asthenopus). Kluge (2004) did not differentiate both genera, apparently treating them as synonyms. But he summarized the features traditionally associated with the subfamily Asthenopodinae (at that time = Asthenopus + Povilla), distinguishing two "synapomorphies" in the nymph: mandibular tusks specialized in biting (not long, very thick and stout, with serrate inner margin), and foretarsal claw with a row of denticles. Kluge (2004) also listed the following "plesiomorphies" shared by this group: 1) FW with CuA not so strongly curved as in Campsurinae, thus both ICu veins go nearly parallel to basitornal margin and terminate near or anteriad to tornus; 2) genitals retain small median remnant of styliger which is articulated to posterior margin of sternite IX and bears immobile pedestals (narrow basally, and with muscles that move forceps); 3) penial arms retain lateral articulations with postero-lateral angles of tergite IX; 4) imaginal moult is present in both sexes (not in female Polymitarcyinae); 5) egg ellipsoid as usual, two polar caps or none; 6) nymphal gill I bilamellate. Note that a formal quantitative phylogeny was not presented by Kluge (2004), so his use of "synapomorphy" and "plesiomorphy" are ad hoc hypotheses.
The inclusion of many of these characters in a formal cladistic analysis permitted us to recognize apomorphic from plesiomorphic states, showing that some of the hypothesized synapomorphies for Asthenopodinae (or particular genera) were not homogeneous in the corresponding clade (e.g., some members of Asthenopodinae may show Campsurinae features, etc.). For this reason, a list of the synapomorphies for each genus is given either in the descriptions and taxon discussion (for Hubbardipes, Priasthenopus, Asthenopus and Asthenopodes) or in Appendix 2 (for Asthenopodinae, and the other genera).
The revalidation of Asthenopodes is in general coincident with the observations of previous authors (i.e., the character changes defining the group are those previously reported by Ulmer 1924, Traver 1956and Hubbard 1975, and new characters are added to its diagnosis due to the knowledge of the egg and nymph. The generic status of Hubbardipes and Priasthenopus are supported by many autapomorphic character states and its position in the phylogeny is intermediate between Asthenopodes and Asthenopus. That the Neotropical species do not form a monophyletic group is a novel hypothesis obtained in our study.

Biogeography
Polymitarcyidae probably originated in Pangea and with the break up of the supercontinent in Laurasia and Gondwana, the first division of the family occurred. The subfamily Polymitarcyinae diferentiated as a Laurasian group (but this would require the ad hoc hypothesis of a later expansion to Africa) (triangles in Fig. 23), indicating a vicariant pattern with the ancestor of Asthenopodinae + Campsurinae, a Gondwanic group. This Gondwanic group probably was always restricted to temperate or warm climates because it is absent from Southern South America, Australia and New Zealand. Asthenopodinae still shows this pattern but Campsurinae extended its range (originally Neotropical) to the Nearctic region but marginally (only 4 of the 62 species, Fig. 23). The ancestor of Campsurinae and Asthenopodinae was most probably distributed in tropical-subtropical Gondwana, where it originated at least five groups: 1) Campsurinae, 2) Asthenopodes, 3) Hubbardipes, 4) Priasthenopus, and 5) the ancestor of Asthenopus + Povilla + Languidipes. Subsequently, the formation of the Atlantic Ocean separated the South American Asthenopus from the Ethiopean-Oriental group (ancestor of Povilla + Languidipes).
Two fossils may falsify these hypotheses. Mesopalingea Whalley & Jarzembowski (a Campsurinae sensu McCafferty 2004) is known in the nymphal stage from the late Jurasic-early Cretaceus of Spain. Kluge (2004) classifies it as Fossoria insertae sedis (a group including almost all the burrowing families), thus not recognizing this genus as a Polymitarcyidae. The tusks of Mesopalingea strongly resemble those of Campsurus, but we suspect that this similarity is only superficial because at least two features would place the fossil outside Campsurinae: a somewhat prominent frontal projection and a very short ring-like prothorax.
The second fossil with contradictory information, Asthenopodichnium  is known from 2 trace species (tubular marks in wood and bones) from the Miocene of Vienna. These tubular marks were atributed to Polymitarcyinae by Thenius ( , 1988 or Asthenopodinae by McCafferty (2004). There is no evidence that they were produced by a mayfly; actually other animal taxa produce similar traces, so we agree with Kluge's (2004) treatment of Asthenopodichnium as Animalia insertae sedis.
Five groups (Campsurinae, Asthenopodes, Hubbardipes, Priasthenopus and Asthenopus) constitute independent sources to study biogeographic patterns inside the Neotropical region. Vicariant patterns in Campsurinae are only known for Campsurus (Molineri & Salles, 2013), which can be compared to those here indicated by Asthenopodes and Asthenopus (Hubbardipes and Priasthenopus are monotipic). Asthenopodes shows a single disjoint sister pair (A. chumuco in the Amazonas subregion vs A. traverae + A. picteti in the Parana subregion, barrier 6 in Fig. 25), coincident with the disjunction between Asthenopus guarani vs the remaining species of Asthenopus (barrier 3 in Fig. 24). This pattern resembles that of Campsurus amapaensis vs C. argentinus + C.major (Molineri & Salles, 2013) and was probably initiated by the appearance of drier areas (Chacoan biogeographic subregion) separating Amazonas subregion to the North from the Paranaense subregion in the SE (Morrone 2001). The two more recent vicariant events involve species in Asthenopus: barrier 4 separating A. hubbardi from A. curtus + A. angelae + A. magnus (detail in Fig. 24) and barrier 5 with A. magnus in the Napo Region vs A. angelae more to the East, Fig. 24). Without a molecular dating of these clades we can only speculate possible explanations related with pleistocene refugia (Hooghiemstra and van der Hammen 1998).
But perhaps the most interesting biogeographic pattern found in the present work is the paralogy in tropical South American areas caused by the paraphyletic nature of the Neotropical Asthenopodinae. The accepted hypothesis of vicariance between South American and Afro-Oriental Asthenopodinae was the formation of the Atlantic Ocean during the breakup of Gondwana (Lomolino et al. 2006). But as this event was found here to be more recent than those involving the more basal clades (Asthenopodes, Hubbardipes and Priasthenopus), a more complex taxon history of the group is suggested. As these events (generic and suprageneric divergences) are deeper than the divergence of sister species in Asthenopus mentioned above, marine transgressions (>10 millions years, Ortiz-Jaureguizar and Cladera 2006) become possible scenarios.
Species included. Hubbardipes crenulatus (Molineri, Cruz & Emmerich) comb. n. Diagnosis. Eight autapomorphies define the genus Hubbardipes in our cladistic analysis (Appendix 2), some of them are: 1) male foretibia very wide apically ( Fig. 20D), first tarsal segment partially fused with second tarsal segment (Fig. 7B); 2) thumb (inner basal projection) of penes reduced, indistinguishable (Fig. 7A); 3) apex of penis lobe wider than base and with many marginal spines (Fig. 7A); and 4) nymphal mandibular tusk with smooth to slightly crenulated inner margin, without submedian tubercles (Figs 5B, D). Additionally, the following combination of characters is useful to distinguish Hubbardipes from other genera in Polymitarcyidae: 1) ratio length male FW/foreleg = 1.1; 2) male foreleg with large and apically widening tibia, tarsal segment 1 small and partially fused to tarsal segment 2, tarsal claws slender slightly wider at apex; 3) pronotum width/length ratio: 1.5 (male), 2.2 (female); 4) marginal intercalary veins present on the entire margin of fore and hind wings, generally shorter than distance between longitudinal veins in male, but longer and anastomosed in female; 5) in both sexes FW with 3-4 crossveins between R and M, basally to R fork; 6) basal relation of FW veins IMP-MP 1 variable (IMP joined to MP 1 , or basally free), MP 2 curved toward CuA and fused to CuA and MP 1 by cross veins (forming a characteristic oblique Y) ; 7) median remnant of styliger plate subquadrate and small, pedestals short also subquadrate and relatively small; 8) forceps slender, ratio length/ basal width = 8.5-9.0 (Fig. 7A); 9) penes relatively thin, with many spines near the apex on outer edge, slightly curved inward (Fig. 7A); 10) female abdominal sternum VIII with anteromedian paired sockets on a protruding subcircular structure; 11) eggs with relatively large polar caps (almost as wide as egg, ratio width egg/cap 1.2-1.4), each cap formed by 4-7 threads, chorion completely covered by large disk-like plates and smaller irregular plates (Figs 7D-E); 12) nymphal head with a small median projection on the frons (arrow in Fig. 5C); 13) nymphs with long robust tusks, without inner tubercles, with 2 or 3-pointed apex (right and left mandible respectively) (Figs 5A-B, D); 14) nymphal foretarsal claw with single row of about 14 denticles, denticles are small basally, and larger medially, 3 denticles are closer together near the apex, and the last one is much smaller (Fig. 6D); 15) nymphal dorsal apex of hind femur with ca. 20 stout spines (Fig. 6E).
Nymphs (Fig. 4A). Length (mm): body, 11.0-14.5; cerci, 4.5-5.5; terminal filament, 6.0. Head (Figs 5A, C) subquadrate in dorsal view, smooth (without pilose area), antennae 1.7-2.0 times length of head. Occipital region well developed, convex (Fig. 5A). Head capsule dorsally projected at bases of antennae. Frontal ridge marked only by a dense transversal row of setae; frons acutely projected medially (Fig. 5C); clypeus and labrum membranous and small, labrum densely covered with long setae on dorsum. Mandibular tusks (Figs 5A, B, D) relatively long and slender, similar in length to head capsule, dorso-ventrally flattened, left tusk apically with 3 tubercles (the median is reduced in length), right tusk with 2 tubercles; dorsal surface of tusks wide, with crenulated inner margin bearing long setae; outer margin with a small dorsal tubercle near base ("b" in Figs 5A-B) and densely covered with stout setae along entire margin; the small basal tubercle ("b") forms an additional articulation between mandible and head capsule ("a" in Fig. 5A). Body of mandible: molae and canines present but small, margin between them sharp-edged (acutely protruding in right mandible, Fig. 5D); with basal U-row of long filtering setae in both mandibles. Thorax. Pronotum with short anterior ring (collar), 1/3 the length of posterior ring (length taken at the medio-longitudinal line), anterolateral corners projected, spine-like. Legs (Fig. 6). Leg I (Fig. 6A): femora very wide, well developed, with a double ventro-basal row of long filtering setae; tibio-tarsus (fused) with 3 rows of filtering setae (2 on anterior face and 1 on inner margin), roundly projected apically; tarsal claw long and slender with a row of marginal denticles (Fig.  6D). Leg II (Fig. 6B): smaller, with thinner femora, with scattered long setae, mostly basally and along hind margin; tibia and tarsi with row of long setae on outer (dorsal) margin, ventrally with many stout spines on apical half, with a distal brush of thick setae; tarsal claw weaker, without denticles. Leg III (Fig. 6C): as leg II except larger and with anterior margin of femur densely covered with thick setae, femur distally with a group of acute stout spines (Fig. 6E), tibia without distal brush. Coxae I and II directed ventrally, coxae III directed laterally. Abdomen. Sternite I stronger and partially fused with metasternum. Gill I reduced in size, dark gray, double, both portions of a similar length, but the dorsal is wider (arrow in Fig. 5E). Gills II-VII well developed, ventral portion smaller than dorsal portion. Tergum X with short and blunt, poorly developed posterolateral spine. Cerci slightly shorter than terminal filament, with long setae at joinings.
Etymology. Hubbardipes from "Hubbard" and "pes", Latin, masculine, meaning "foot". We dedicate the genus to Mike Hubbard, mayfly specialist, who devoted many of his works to the Polymitarcyids.
Distribution. Amazonas subregion (Amazonas river in Colombia and Brazil).

Discussion.
Hubbardipes was recovered as sister to a larger clade containing Priasthenopus, Asthenopus, Povilla and Languidipes (see synapomorphies in Appendix 2), not related with Asthenopodes, as previously thought . Hubbardipes shows many differences in the adult stage and, more markedly in the nymph and egg here described for the first time. Male genitalia is unique in form and structure (Molineri et al. 2011), nymphal tusks and microsculpture on disk-like structures of the egg are exceptional as well. Diagnosis. Hubbardipes crenulatus  comb. n. is known from adults of both sexes, eggs and nymphs, and for the moment it is the only known species in the genus. The characters useful to distinguish it from other Asthenopodinae are listed in the generic diagnosis.

Asthenopus crenulatus
Nymphs (Fig. 4A). Length (mm): body, 11.0-14.5; cerci, 4.5-5.5; terminal filament, 6.0. General coloration whitish light brown. Head with a gray band between lateral ocelli and fine netting pattern on occiput (Fig. 5A). Antennae: scape bare, long and slender, pedicel shorter with many dorsal setae, flagellum bare with numerous annuli increasing in length distally. Thorax. Pronotum shaded black on anterior ring and more slightly shaded witn brownish-gray on posterior ring except on a pair of submedian longitudinal pale lines. Meso-and metanotum shaded widely with gray, with dark gray wingbuds, developing veins paler (Fig. 5E). Legs (Fig. 6). All coxae and trochanters shaded with gray, femur I shaded with gray, remaining segments and legs II and III whitish-yellow; foretarsal claw with ca. 20 denticles increasing in size distally. Abdomen. Terga more or less uniformly shaded brownish-gray, except on pale transverse dashes laterally, and pale subcircular submedian marks; tergum X with two submedian pale bands; sterna paler than terga, shaded with gray very slightly, somewhat darker on sterna IX-X. Gill I dark gray, gills II-VII brownish gray, ventral portion paler than dorsal portion. Caudal filaments whitish.
Distribution. Amazonas river, from Leticia (Colombia) to Manaus (Brazil). Discussion. Hubbardipes crenulatus  comb. n. was recently described from male and female adults in the genus Asthenopus, with the knowledge of the nymphs and based on the results of the phylogenetic analysis, it became evident that this species pertain to a distinct group, that we propose here as a new genus. Diagnosis. Priasthenopus gen. n. presents seven autapomorphies in our cladistic analysis (Appendix 2), six are variations in continuous characters (e.g., stoutter penis lobe and larger thumb) and the seventh is the presence of a short closed cleft between the aforementioned structures. This genus can be distinguished from the other by the following combination of characters: 1) ratio length male FW/foreleg = 1.6-2.0; 2) first tarsal segment subquadrate not fused with tibia (Fig. 20E); 3) pronotum width/ length ratio: 2.1-2.5 (male), 2.6 (female); 4) 5-10 marginal intercalary veins present on the margin of FW (Figs 8F, H), about as long as the distance between longitudinal veins in male, slightly longer in female, HW without marginal intercalaries; 5) male FW with 1 cross vein between Rs and MA basal to Rs fork (2 in female); 6) vein MP 1 basally free (types from Uruguay and specimens from Colombia, Fig. 8D) or tending to fuse, although not completely, with MP 2 (in specimens from Bolivia, Fig. 8E), IMP basally free; 7) median remnant of styliger plate rectangular thin and convex posteriorly, pedestals rectangular (Fig. 8A); 8) ratio total length/basal width of forceps 6.3-7.3 (Fig. 8A); 9) penes relatively short with a similar width along their length, strongly curved, without apical spine or spine very slightly marked as a subapical indentation, thumb rounded (Figs 8A-C); 10) female sternum VIII with well distinguishable anteromedian sockets (Figs 8J); 11) eggs with relatively large polar caps formed by 14-16 long threads (Fig. 8K), chorion loosely covered by medium-sized and small circular plates (Fig. 8L).
Etymology. Arbitrary combination of letters. Discussion. Treating this sole species in a new genus, distinct from Asthenopus is justified by its phylogenetic position (sister to the clade Povilla-Asthenopus). The other possibility to fit taxonomy to phylogeny would be to synonymize the entire clade (including Povilla and Languidipes besides gilliesi) in Asthenopus. This is the scheme apparently presented by Kluge (2004) but this seems inadequate to us because of relatively large morphological gaps between the groups (including characters from eggs, nymphs and adults of both sexes). Furthermore, we tested P. gilliesi position considering the hypotethical situation that its nymphs (still unknown) be identical to Asthenopus s.s., since male genitalia of both groups are similar. Even so, there were no changes in the resulting tree. The description of female adults and eggs is an original contribution of this work that gives additional diagnostic characters to genus level. ( Diagnosis. Priasthenopus gilliesi is known from adults of both sexes and eggs, and is the only species known in the genus. The characters useful to distinguish it from other Asthenopodinae are listed in the generic diagnosis.

Priasthenopus gilliesi
Male imago. See generic section above and original description in Domínguez (1988). Female subimago. Length (mm): body, 6.5; FW, 8.9; HW, 3.3; cerci, 1.5. General coloration yellowish white with gray markings. Head cream extensively shaded gray dorsally, the shading is uniform anteriorly but in the form of a fine netted pattern posteriorly to lateral ocelli, occiput with a pair of submedian pale anterior spots and a pair of submedian dark posterior spots; venter of head whitish. Antennae yellowish white shaded with gray on scape and pedicel. Thorax cream. Anterior ring of prothorax very thin, less than 1/4 the dorsal length of posterior ring; ratio width/length: 2.6; pronotum shaded blackish on median area except pale medial line, presternum paler, shaded gray before coxa. Mesonotum shaded very diffusely with gray, darker on longitudinal carinae and between posteroscutal protuberances; mesosternum and pleurae paler, shaded gray on anterior corner of katepisternum. Metanotum shaded gray on posterior half, except on a pale median triangular mark, shaded darker posteriorly to this pale mark; metasternum whitish. Legs whitish except coxae yellowish shaded gray and apex of hind trochanter pointed and yellowish orange. Wings (Figs 8H-I). Membrane of fore and hind wings slightly shaded with light brown, veins translucent shaded with light brown, markedly on larger veins (C, Sc, Rs); marginal intercalaries relatively long (Fig. 8H), MP 1 fused with MP 2 at base; 2 crossveins present between M and R stems basally to Rs fork. Abdomen whitish, shaded with gray dorsally and darkening posteriorly, shading on terga interrupted on medial paler line, this line is wide on terga 1-4, narrows posteriorly on 5-7, thin on 8, and widens posteriorly on 9-10; sterna whitish translucent except gill sclerites yellowish white; sternum 8 with anteromedian pair of relatively large sockets (Fig. 8J). Cerci whitish, about 0.2 the length of FW.
Eggs. See generic description. Distribution. This species presents a wide geographic range that spans from the Amazon River in the North to the Uruguay River in the South, also extending towards the West in Bolivian Chiquitania.
Discussion. Priasthenopus gilliesi male imagos were adequately described by Domínguez (1988), females and eggs are described here for the first time. There are no morphological differences between the male imagos examined from the different localities, except for the penes of the Colombian males are slightly stouter, and those from the Bolivian males are slightly slender than the penes of the Uruguayan types. In the forewings, vein MP 1 is basally free except on Bolivian males where this vein tends to fuse with MP 2 , although not completely. Discussion. In the forewings of Priasthenopus gilliesi, vein MP 1 is basally free except on Bolivian males where this vein tends to fuse with MP 2 , although not completely. This last arrangement of the MP sector is also present in Campsurus paraquarius Navas (1920) and, also coincident, are the length and arrangement of the two imv (intercalary marginal veinlets) figured by Navás, the relatively short ICu veins and the vein AA basally curved to CuP; also the small size of the male described by Navás coincide with Priasthenopus size range. Navás described the color of legs without saying that middle and hind legs are reduced (as Campsurus), so probably they were present and complete as in all Asthenopodinae. Finally Navás stated that the pronotum is wider than long (transverse), feature also present in Priasthenopus and related genera (Asthenopus, Povilla) but not in Campsurus males. Other species of Neotropical Asthenopodinae show shorter intercalary marginal veins (Asthenopus s.s.) or longer ICu veins (Asthenopodes). Lestage (1923) noted the similiraty of Navás species with the genus Asthenopus and Kluge (2004) treated this species in Asthenopus because of the arrangement of veins in the Cu sector of FW. We here coincide with these authors and because of the bad original description by Navás, and the apparent loss of type material, we treat the name paraquarius (Navás 1920) as a NOMEN NUDUM.
Nymphs, nearly mature (Fig. 4B, F). Length (mm): body, 7.8; cerci, 2.0-2.3; terminal filament, 3.1. Head subquadrate in dorsal view, smooth (without pilose areas), antennae subequal in length to head. Occipital region well developed, convex (Figs 9A-B). Head capsule dorsally projecting at bases of antennae. Frontal ridge marked only by a dense transversal row of setae; frons not projecting medially; clypeus and labrum membranous and small, labrum densely covered with long setae on dorsum. Mandibular tusks very short and robust, the part visible in dorsal view ca. 1/3 the length of head capsule; left tusk ( Fig. 9E-H) apically with 3 teeth, the median reduced in length, and the innermost is strongly widened appearing as a ridge with two points (Fig. 9F); right tusk ( Fig. 9C-D, G-H) with 2 distal teeth; inner surface with a small tubercles located distally (in relation to other Asthenopodinae), dorsal surface with a small tubercle ("b" in Fig. 9G) that forms an additional articulation between mandible and head capsule; tusk densely covered with long setae, except at apex. Body of mandible: molae strongly protruded medially, canines present but small, margin between them sharp-edged (acutely protruding in right mandible); with basal U-row of long filtering setae in both mandibles. Thorax. Pronotum with anterior ring (collar) subequal in length to posterior ring (length taken at the medio-longitudinal line), anterolateral corners projecting, spine-like, posterior ring with dense patches of short setae medially (Figs 4F, 9B). Legs (Fig. 10A-D). Leg I (Fig. 10A): femora robust, relatively slender, with a U-shaped ventro-basal row of long filtering setae, distal points of the U almost touching each other; tibio-tarsus (fused) with 2 U-rows of filtering setae: 1 on anterior face (each branch well separated in the base) and 1 on inner margin, with the branches near each other, apex of tibio-tarsus relatively pointed; tarsal claw with 2 rows of marginal denticles (Fig. 10B). Leg II (Fig. 10C): smaller, with thinner femora, with scattered long setae basally and a row of long and short setae along outer margin; tibia and tarsi with row of long setae on outer (dorsal) margin, ventrally with many stout spines on apical half, anterior face of tibia with a distal row of thick setae (at base of tarsus) and with a crown of thick setae at apex; tarsal claw weaker, without denticles. Leg III (Fig. 10D): intermediate in size, outer margin of femur with row of short setae, longer at apex, distal corner of femur densely covered with thick, blunt setae; inner margin of femur, tibia and tarsus densely covered with short setae; margin between tibia and tarsus with row of thick setae; outer margin of tibia and tarsus with row of long pectinated setae. Coxae I and II directed ventrally, coxa III directed postero-laterally. Abdomen. Sternite I longer than the others and partially fused with metasternum. Gill I reduced in size, whitish, single and lanceolate. Gills II-VII well developed, ventral portion smaller than dorsal portion. Tergum X without posterolateral spine (Fig. 10E). Cerci slightly shorter than terminal filament, with long setae at joinings, basal 1/4 with thick blunt setae ventraly (Fig. 10F).
Discussion. Asthenopodes and Asthenopus have been treated as synonyms (Hubbard and Domínguez 1988) after the discovery of Priasthenopus gilliesi (Domínguez, 1988), that somewhat blurred the distinction between both genera. Additionally, Hubbard and Domínguez (1988) based their synonymic proposal in the fact that all known nymphs of Asthenopodinae from South America where indistinguishable and could be classified in a single genus. As the knowledge of this group had largely improved in last years we are proposing here a new rearrengement of supraspecific taxa.
The revalidation of the genus Asthenopodes Ulmer is based not only in the clade that its type species (Asthenopodes picteti) forms with other two new species, but also on the fact that the nymph shows characters considered important at the generic level in the family, mainly the shape of nymphal mandibular tusks, legs and gill I. FW 11.4-11.9 mm; genitalia as in Fig. 12A (Fig. 11K-L), smaller size (FW 12.2-14.8 mm); polar cap of eggs formed by 14-16 long coiled threads (Fig. 13C)  Type material. Type material was not studied; it consists of the holotype male imago, damaged, with many parts missing including the genitalia. It is a pinned specimen deposited at Naturistorisches Museum Wien, Hubbard and Domínguez (1988)  Diagnosis. Asthenopodes picteti, type species of the genus Asthenopodes (Ulmer 1924, Traver 1956, Hubbard 1975 presents 7 autapomorphies (Appendix 2) including: thinner male foretibia and slender penes. This species can be distinguished from the other species of the genus by the following combination of characters: 1) general coloration yellowish white (male), darker in female (yellowish light brown); 2) male FW 11.4-11.9 mm, female FW 14.5-19.0 mm, membrane whitish hyaline tinged with yellowish near hind margin; 3) foreleg length 0.9 times the length of FW; 4) pronotum width/length ratio: 1.6-1.9 (male), 1.5-2.3 (female); 5) 14-17 marginal intercalary veins present on the entire margin of FW (Fig. 11A) and HW (Fig. 11B), miv generally longer than distance between longitudinal veins; 6) male FW with 4 to 6 cross veins between Rs and MA basal to Rs fork; 7) ratio total length/basal width of forceps 6.7 (Fig. 12A); 8) penes long and slender (Fig. 12A-B), male median remnant of styliger plate posterolaterally expanded forming a pair of rounded projections ("Ll" in Fig. 12B); 9) female sternum VIII, anteromedian sockets present but concolorous with sternum, not evident; 10) egg polar caps only slightly thinner than maximum width of the egg, formed by 6 filaments; medium-sized and small chorionic disks-like structures present (Fig. 13A).
Female adult. Length (mm): body, 11.5-12.6; FW, 14.5-19.0; HW, 6.3-9.4. General coloration yellowish light brown. Head dorsally blackish except on median zone, paler; venter of head yellowish white. Antennae light brown, shaded gray on scape. Thorax yellowish brown with blackish membranes, shaded with brownish gray on pronotum and with black on posteromedian marks on meso-and metanotum. Pronotum width/length: 1.5-2.3. Legs whitish yellow shaded brownish on dorsum of leg I and on apex of femur III. Wings (Fig. 11C-D), membrane tinged with light brown, veins yellowish brown. Abdomen. Terga brownish with a pale mediolongitudinal line and paler areas on lateral margins on terga I-VII; sterna whitish yellow; female sternum VIII with anteromedian sockets, small and almost not distinguishable. Terminal filament whitish, shorter than tergum VIII; cerci yellowish brown paler apically, 0.5-0.6 times the length of the abdomen (0.2 the length of FW). Eggs (Fig. 13A). Length, 325−355 μ; width, 215−235 μ. Two large and flat polar caps (maximum width, 155−180 μ), formed by 6 very long coiled threads. Chorionic surface with a fine granulated aspect, with small disk like structures.
Distribution. Uruguay, Argentina. A. picteti is here newly recorded from Argentina. The record from Guyana given by Domínguez et al. (2006, p.562) is no longer valid since this material is now considered a different species (A. chumuco see below).
Discussion. Asthenopodes picteti Hubbard was only partially known from the damaged holotype male from Brazil until Traver (1956) redescribed it (at that time as Asthenopodes albicans) based on a complete male from Uruguay. Later Hubbard (1975) gave a new name to this species (A. picteti). Domínguez (1988) and Hubbard and Domínguez (1988) presented additional discussion and illustrations of the male imago, from type and non-type material (also collected in Uruguay). The description of Traver (1956) coincides with the new material studied, except for the relation of the male foretibia and forefemur. Traver reported that tibia is 1.33 times the length of femur but we found that it is 1.7 times that length. Domínguez (1988) reported 2 female imagos collected at the same time than the males he redescribed, but did not presented a formal description of them. Both females and additional ones from Argentina (see list of material) are here shortly described and figured. It is difficult from the material available to determine if the females are in subimaginal or imaginal stage, so they are here referred as "adults". The females are similar to the sympatric A. traverae but can be separated from this species because A. picteti females are lighter in color, hind femora only shaded black on apex, paired female sockets present but small and hard to distinguish, cerci light colored, abdominal gill sclerites smaller, FW wider with shorter marginal intercalaries joined to main veins and partially anastomosed. The eggs present wider polar caps, only slightly thinner than the egg. Males can be separated from those of A. traverae because the hind femora are only shaded black on apex and the penes are much slender. Diagnosis. Four autapomorphies characterize this species, all are small changes in continuous characters, except the marked shortening of the median remnant of styliger plate (at the middle, since laterally a tong-like projection is present, Fig. 12C). Asthenopodes traverae sp. n. is known in the alate and egg stages, all females are subimagos even those that apparently have laid the eggs (empty females). This species can be distinguished from the other species of the genus by the following combination of characters: 1) males yellowish white, females blackish; 2) FW length male 14.0-14.5 mm, female 15.5-22.5 mm, membrane yellowish hyaline; 3) male foreleg length 0.9-1.0 times the length of FW; 4) pronotum width/length ratio: 1.3-1.9 (male), 2.0 (female); 5) 16-22 marginal intercalary veins present on the entire margin of FW (also present in HW) generally longer than distance between longitudinal veins, fused to main veins, poorly anastomosed ( Fig. 11E-H); 6) male FW with 4 to 5 cross veins between Rs and MA basal to Rs fork; 7) ratio total length/basal width of forceps 4.7-4.9 (Fig. 12C-D); 8) median remnant of styliger plate posterolaterally expanded forming a pair of rounded projections, penes large and robust, sclerotized, curved and twisted ( Fig. 12C-D); 9) female sternum VIII with anteromedian sockets well developed, whitish surrounded by a large brownish area (Fig.  13D); 10) egg caps small, much thinner than maximum width of the egg, formed by 5-6 filaments; medium-sized and small chorionic plates present (Fig. 13B).
Male imago. Length (mm): body, 12.2-13.5; FW, 14.0-14.5; HW, 6.8-7.3; foreleg, 12.3-14.0; cerci, 35.5-38.5. General coloration yellowish white. Head shaded black dorsally almost entirely, with a pair of distinct submedian black marks anteriorly to median ocellus; occipital hind margin with pale median zone; head ventrally pale without markings. Antennae: scape and pedicel short, yellowish on venter of pedicel, both slightly shaded with gray; flagellum very thin, hyaline, similar in length to forefemur. Thorax. Pronotum translucent, shaded slightly with gray except on membranes separating anterior and posterior rings, darker laterally; pronotum width/length: 1.3-1.9. Meso-and metanotum yellowish white with gray markings mainly posteriorly but also on sutures. Thoracic pleurae and sterna yellowish white shaded gray only at base of coxae and wings. Legs. Forelegs: coxa dorsally whitish with a gray mark, ventrally yellowish; femur dorsally yellowish shaded gray on apical third, ventrally whitish; tibia whitish translucent shaded gray mainly on dorsum; tarsi translucent shaded slightly with gray; large claws, stalk with brown inner margin, rest whitish; articulations between femur-tibia and tibia-tarsus very sclerotized, brownish. Middle and hind legs yellowish, shaded with gray from half of femur to apex of leg. Wings (Fig. 11E-H). Membrane hyaline slightly tinged with gray on costal and subcostal sectors (membrane yellowish under transmitted light); all veins light gray, lighter toward hind margin; 5 cross veins between R stem and M sector; long marginal intercalaries on hind margin of both wings. Abdomen yellowish white widely shaded with gray, some darker marks as follows: submedian short anterior dashes and sublateral oblique stripes on terga III-VIII, lateral margins of terga VIII-IX, and median line of IX-X (thinner on X). Abdominal sterna whitish except gill sclerites yellowish white. Genitalia (Fig. 12C-D): lateral margins of sternum IX and pedestals yellowish; median remnant of styliger plate whitish with a gray strip on hind margin between the tongue-like projections ("Ll"Fig. in Fig. 12D); pedestals well separated from each other, relatively long and becoming wider distally; forceps whitish, long and wide; base of penes well developed, subquadrate, whitish; penes strongly sclerotized, orangeish basally but yellowish distally. Cerci whitish, terminal filament reduced to 7-8 thin annuli, straight, whitish.
Etymology. The species is dedicated to the great mayfly specialist Jay R Traver, who visited Uruguay and worked with Mr. C. S. Carbonell's collections at the "Museo de la República" recognizing the females of this species as distinct from A. picteti (also unknown at that time).
Distribution. Parana biogeographic subregion in Argentina, Uruguay and Brazil. Biological remarks. Females subimagos were collected (in Misiones Province) while swarming in compact groups at about 3 m above water in pool areas around sunset. The same behavior was reported by Traver (1956) for the Uruguayan females. Males were caught at light traps during the first hours of dark, so the male flight is unknown.
Discussion. Traver (1956) described females of A. traverae as distinct from A. picteti in spite of the fact that the female of that species was also unknown at that time. Nevertheless Traver realized that they differ from A. picteti males in color and size, and left them unnamed. With the collection of new material from both species (and sexes) in Misiones (Argentina), it became evident that they constitute a new species. Asthenopodes traverae females can be distinguished from A. picteti females by their black general coloration, somewhat slender forewings, femora and cerci widely shaded with black, and enlarged abdominal gill sclerites (remnants of nymphal gill muscles insertions), for other differences see discussion under A. picteti. Very similar in aspect but much smaller are the females of A. chumuco, the allopatric sister species of A. traverae. Diagnosis. Asthenopodes chumuco known from all the stages presents seven autapomorphies, all of them are changes in continuous characters (Appendix 2). This species can be distinguished from the other species of the genus by the following combination of characters: 1) general coloration yellowish white in male, dark brown in female; 2) male FW 7.0-9.0 mm, female FW 12.2-14.8 mm; 3) male, ratio FW/foreleg length 1.4-1.6; 4) pronotum width/length male 1.3-1.4, female 2.0-2.1; 5) 7-14 marginal intercalary veins present on the entire margin of FW (Fig. 11I, K), HW (Fig. 11J, L) with 3-7 marginal intercalary veins, generally longer than distance between longitudinal veins, fused to main veins, poorly to heavily anastomosed; 6) male FW (Fig. 11I) with 3-4 cross veins between Rs and MA basal to Rs fork; 7) ratio total length/basal width of forceps 7.8-9.5 (twisted CUIC slide 6.3-6.7) (Fig. 12E); 8) male median remnant of styliger plate subrectangular, slightly convex (not expanded forming a pair of rounded lateral projections), penes long and slender, acute distally ( Fig. 12E-F), rectangular pedestals strongly enlarged, ½ the length of forceps; 9) female sternum VIII, with small reduced anteromedian female sockets at the base of a median keel; 10) egg caps small, much thinner than maximum width of the egg, formed by 14-16 filaments; large and small chorionic plates present (Fig. 13C).

Asthenopodes chumuco
Male imago. Length (mm): body, 7. 3-8.0; forewing, 7.0-8.8; hind wing, 3.7-4.3; foreleg, 5.0-5.8; cerci, 20.0-23.0. General coloration yellowish white. Head shaded gray dorsally almost entirely, frons with black dot at base of antenna, with medial line and irregular black marks; occipital region with pale median zone; head ventrally pale without markings. Antennae: scape and pedicel short, subequal in length, whitish shaded with purplish; flagellum very thin, hyaline. Thorax. Pronotum whitish with anterior and posterior portion subequal in size, shaded gray in a transverse band between both portions, posterior portion black along hind and lateral margins; pronotum width/length: 1.2-1.3. Mesonotum yellowish white shaded with gray on posterior half of medial line, on area between posterolateral protuberances and on anterior margin of these structures. Metanotum yellowish slighlty shaded with gray medially. Thoracic pleurae and sterna yellowish white shaded gray dorsally and anteriorly to mid coxa. Legs. Forelegs: whitish completely shaded with gray; large claws, apically expanded (Fig. 13E). Middle and hind legs yellowish white, shaded with gray on coxae. Wings (Fig. 11I-J). Membrane hyaline shaded with gray on basal 1/3 of costal and subcostal sectors; all veins translucent, except costal cross veins grayish; 3-4 cross veins between R stem and M sector; long marginal intercalaries on hind margin of both wings. Abdomen whitish shaded with gray on terga, mainly on lateral margin, submedian black dot on terga III-VIII. Abdominal sterna pale. Genitalia (Fig. 12E-F): yellowish white, except for penis apically yellowish; pedestals well separated from each other, very long, ½ length of forceps; median remnant of styliger plate with slightly convex hind margin; base of penis rounded, projecting posterolaterally, lobe of penis long and sclerotized, curved ventromedially, constricted on median length, gonopore well developed. Cerci whitish; terminal filament reduced to 5-7 thin annuli, straight, whitish.
Eggs. Length 210-240 μ, width 180-200 μ. Subovate, yellowish, with two small whitish polar caps (maximum width, 75-85 μ), polar caps much thinner than the egg and formed by 14-16 threads. Under SEM the larger disk-like chorionic structures are surrounded by many smaller ones, which at their time are surrounded by smooth chorion (Fig. 13C).
Nymphs. Length (mm): body, 7.8; cerci, 2.0-2.3; terminal filament, 3.1. General coloration yellowish light gray (Fig. 4B, F). Head with a black band between lateral ocelli and fine netting pattern on occiput (Fig. 4F). Antennae: scape bare, slightly longer than pedicel, pedicel with many dorsal setae, flagellum bare with numerous annuli increasing in length distally. Thorax. Pronotum shaded black on sublateral area of anterior ring and laterally on posterior ring. Meso-and metanotum shaded widely with gray, with dark gray wingbuds, developing veins paler. Legs (Fig. 10A-D). Coxae and trochanters of mid and hind legs slightly shaded with gray, remainder of legs yellowish-white; foretarsal claw with double parallel rows of 15 and 12 denticles each (Fig. 10B). Abdomen. Terga more or less uniformly shaded brownish-gray, except on pale transverse dashes laterally, thin medial line on tergum I-IX becoming wider posteriorly and pale with subcircular submedian sigilla; sterna yellowish. Gill I whitish, gills II-VII purplish gray, ventral portion paler than dorsal portion. Caudal filaments yellowish.
Etymology. "Chumuco" is one of the common names applied to river cormorans in some South American countries. The penis lobe of this new species resembles the neck and head of that bird.
Distribution. Brazil (Amazonas, Espírito Santo), Colombia (Amazonas), Guyana. Discussion. The male imago of this species has been known since Traver (1950) found a group of slides containing some body parts of a missing specimen. She stated that it was surely not the type species of Asthenopodes mainly because its smaller size. Domínguez et al. 2006 (p. 562) treated this specimen in Asthenopus picteti, extending the distribution of this last species to British Guiana. With the discovery of new specimens (male and female adults) we gathered more morphological information and realized that a new species must be described. One of the results is that A. picteti is no longer considered to be in British Guiana. The Colombian females described here as Asthenopodes chumuco are associated with the males, because of the egg morphology, compared with those extracted from a pharate female from São Mateus. Adults show the smallest size of the genus, further differing from the other two species in many morphological aspects. The females are similar to A. picteti in coloration of hind femur and abdomen but wing venation and sockets on sternum VIII are different, also the female cerci of A. chumuco are much less developed. The eggs are similar to those of A. traverae, because of the small polar caps, nevertheless A. chumuco presents much more threads forming each cap (between 14 and 16, but each thread is relatively very thin). Figs 4C-D, 14-18, 20F-J Asthenopus Eaton 1871: 59;Lestage 1922: 142;Traver 1950: 605;Traver 1956b: 7;Hubbard and Domínguez 1988: 209;Domínguez 1988a: 24.
Nymphs. Length (mm): body, 9.7-15.0 mm; cerci, 4.0-7.0; terminal filament, 5.0-5.1. Head suboval in dorsal view, smooth (without pilose area); occipital region well developed, strongly convex (Figs 4C-D, 14I-J). Head capsule with a dorsal spine-like projection at bases of antennae. Antennae 1.1-1.5 times length of head (length of head taken from hind margin to the apex of clypeus); pedicel with tuft of setae on dorsum, flagellum with minute scattered setae; length (mm): scape (0.5), pedicel (0.28), flagellum (2.0). Frons with anterior margin more or less straight (arrow in Fig. 14J), with a small blunt lateral projection ("a" in Fig. 14I), without median projection. Clypeus and labrum small, membranous, with many setae on dorsum of labrum. Mandibular tusks robust, relatively stout, left tusk ( Fig. 14A-C, E) with 3 apical teeth, increasing in size from the median (smallest), inner and outer; inner tooth slightly directed medially, others directed distally; right tusk ( Fig. 14F-G) with 2 teeth, the inner shorter. Inner margin of both tusks with a rounded small tubercle near subapex and a larger and pointed subbasal tubercle (associated with a tuft of rigid setae), this large basal tubercle shows a small basal protuberance (giving the impression of a bifid tubercle but with one of the sides aborted); ventral surface and outer margin of tusks with small rounded protuberances on the extremely hard cuticle; dorsal surface of tusks with numerous setae and with a small basal tubercle; this small dorsal tubercle is easily seen without dissecting the mandible and gives an additional point of articulation between the mandible and the head capsule ("a" and "b" in Fig. 14I). Incisors and prostheca of both mandibles very reduced in size, molae relatively well developed.
Maxillae with a small subtriangular basal membranous "gill" (membranous outgrouth). Thorax. Anterior ring of pronotum (or collar sensu Kluge 2004) short (ca. 1/4 the length of posterior ring), anteriorly projecting as spines on lateral corners; posterior ring longer, ring-like. Legs (Fig. 15A-D, F-G). Leg I (Fig. 15A-B): femora very wide, well developed, with a double ventro-basal row of long filtering setae; tibio-tarsus (fused, but fusion line distinguishable) with 3 rows of filtering setae (2 on dorsal "face" and 1 on inner margin), tarsus slightly and bluntly projecting apically (arrow in Fig. 15B); tarsal claw relatively large and stout with a row of marginal denticles (Fig. 15G). Leg II (Fig. 15C): smaller, with thinner femora, with scattered long setae, mostly basally and along hind margin; tibia and tarsi with row of long setae on outer (dorsal) margin, ventrally with many stout spines on apical half, with a distal brush of thick setae (arrow in Fig. 15C); tarsal claw relatively small, without denticles. Leg III (Fig. 15D, F): as leg II except larger and with anterior margin of femur densely covered with thick setae, and posterior margin roundly expanded at apex bearing a group of stout acute spines (Fig. 15F); tibia without distal brush. Coxae I and II directed ventrally, coxae III directed laterally. Abdomen. Gill I reduced in size, double, both portions subequal in length and width. Gills II-VII well developed, ventral portion smaller than dorsal portion; tergum X with well developed ventral spine on posterior margin (not visible dorsally, Fig. 15E). Caudal filaments short (curved in mature nymphs) with whorls of stout spines and simple setae at joinings.

Key to the species of Asthenopus
Male imago. Length: body,FW,10.0;HW,4.4;foreleg,7.5;cerci,. General coloration yellowish light brown. Head whitish, heavily shaded black dorsally, paler on posteromedian zone of occiput, black shading extending anteriorly on frons as two parallel lines surrounding median ocellus. Antennae pale, slightly shaded gray on dorsum. Thorax. Pronotum yellowish translucent completely shaded gray, darker on anterior ring; paler on two transverse lines, one separating anterior and posterior rings and another more posterior and obliquely transverse; pleurae shaded with black, sternum with a median gray macula. Pronotum width/length ratio: 2.0-2.3. Mesonotum whitish yellow (or brownish in some males) with a black median triangle between posteroscutal protuberances, metanotum similar in color, also shaded black posteromedially; mesopleurae and sterna paler, shaded with black along anterior margin of katepisternum. Legs yellowish white shaded with gray dorsally on all coxae, femora and tibiae; foretarsal segment 1 blackish (Fig. 20G), remaining tarsal segments paler shaded with gray distally, claws grayish thin troughout. Wings ( Fig.  16A-B). Membrane hyaline shaded very slightly with brownish near anterior margin and turning whitish translucent towards apical zone of C-Sc areas; veins translucent shaded with brown. Abdomen yellowish white shaded extensively with grayish brown dorsally, darkening very slightly towards rear segments. Sterna pale very slightly shaded gray, shaded stronger on mediolongitudinal line near anterior margin of sterna VIII-IX, this line is blurred posteriorly; a grayish black triangular mark is present at each side of this line, on anterior margin of sterna VIII-IX; sternum X shaded black except medially. Genitalia (Fig. 17A-B): median remnant of styliger plate and pedestals yellowish, forceps whitish translucent shaded gray along outer margin, penes dark orange with whitish base. Caudal filaments whitish, shaded gray at base of terminal filament.
Mature nymph. Length of male: body, 9.5-9.7; cercus, 7.0; terminal filament, 5.0. Length of female: body, 17.0; cercus, 8.0; terminal filament, 7.0. Only characters that differ from A. angelae are given here, refer to that description for more detailed information. Head (occipital area) dorsally brownish uniformly shaded with gray. Mouthparts. Left mandibular tusks with a relatively shorter space between the large subbasal tubercle and the smaller subdmedian one, this space is somewhat C-shaped (Fig. 14A). Right mandible with distal corner of mola strongly protruding. Thorax. Mesonotum uniformly brownish (cuticular) without strongly gray-shading on carinae (Fig. 4E). Legs and paraprocts identical to those on Fig. 15.
Distribution. Amazonas River from Leticia (Colombia) to Manaus (Brazil). Discussion. Much confusion exists in the literature concerning this species. Many authors mention A. curtus but from missidentified material. For example Ulmer (1942) described and illustrated (as A. curtus) a pair of males of Asthenopus angelae. The material from Ecuador studied by Domínguez (1988) proved to be a different but related species (A. magnus sp. n.). Berner (1978) synonymized A. curtus with A. amazonicus, showing that the differences between both species were only attributable to sexual dimorphism, but he was working with A. angelae males (de Souza and Molineri 2012). Nevertheless, Berner conclusions were correct given that sexual dimorphism in FW venation is present in both species. As it is impossible to assign any specimen to A. amazonicus, we prefer to treat it as synonym of A. curtus, as Berner proposed. Actually, only one specimen from previous works is positively determined as A. curtus: the type, studied by Eaton (1883) and illustrated by Kimmins (1966). We add here some other records from the Amazonas River: a pair of males from Colombia, some reared nymphs from Brazil and Fittkau's slides at FAMU. These male imagos show the characteristic genitalia of the holotype of A. curtus, with extremely wide forceps, long penis lobes and slender and very acute apical spines, and the more or less uniform brownish mesothoracic coloration (an exception of this last character are the males from Colombia-Leticia, much paler). The egg of A. curtus (Fig. 18A) is similar to that of A. hubbardi (Fig. 18D), in the shape and relative large size of the disk-like structures that leaves exposed only a reduced surface of smooth chorion. On the contrary the egg of A. angelae presents smaller disk-like structures with a larger surface of smooth chorion among them (Fig. 18B).
Male imago. Length (mm): body, 9.0-10.5; FW, 9.0-10.1; HW, 4.0-4.8; leg I, 6.5-7.4; cerci, 37.0. General coloration yellowish light brown. Head whitish shaded black dorsally on pale median mark on hind margin and along inner margin of eyes; frons pale except paired submedian black lines. Antennae: whitish shaded diffusely with gray on scape and apex of pedicel; length (mm): scape 2.25, pedicel 1.5, flagellum 7.25. Thorax. Pronotum yellowish translucent shaded with black dorsally except at pale median membrane between both pronotal rings, on mediolongitudinal line and along margins; the black shading presents many scattered and small pale spots. Pronotum width/ length: 1.7-2.4. Meso-and metanotum yellowish shaded with grayish on carinae, pos- teromedian triangular mark (on mesonotum), and scutellum (both). Thoracic pleurae and sterna paler, shaded gray on pleural sclerites. Legs yellowish white, shaded gray on all coxae. Leg I shaded gray almost completely, stronger on femur and tibia, paler on tarsal segments (Fig. 20J). Legs II-III shaded with black dorsally on apex of femora and entire dorsum of tibiae, rest pale. Wings (Fig. 16C-D). Membrane hyaline shaded grayish near costal margin on basal half, more whitish apically; all veins translucent completely shaded gray; 1-3 cross veins between MA and R, basad to R stem. Abdomen yellowish white shaded with gray and black dorsally except on median and lateral zones and intersegmental membranes. Median pale areas on terga II-IX are oval and are surrounded by darker pigments, a delicate mediolongitudinal black line is present but sometimes is only visible on tergum IX. Sterna whitish turning yellowish laterally and on sternum IX, shaded gray on paraproct and basally to terminal filament. Genitalia (Fig. 17G): forceps whitish, penes yellowish white. Cerci whitish very slightly shaded with gray.
Female imago. Length (mm): body, 15.5-19.5; FW, 16.0-17.5; HW, 5.2-7.8; cerci, 3.5-5.0. Pronotum width/length: 2. Morphologically very similar to A. curtus and A. angelae, the last is described elsewhere (de Souza and Molineri 2012). Color pattern similar to male but more strongly marked, exceptions follows: pale median mark on occiput longer, reaching median ocellus; wing membrane tinged with yellowish near costal margin and base, all veins shaded brown; shading on abdominal terga more extended, shaded widely gray except medially, tergum VIII-IX (sometimes also II-VII) with a black line in the pale median zone; medial margins of gill sclerites on abdominal sterna II-VII and lateral margins of sterna VIII-IX grayish; Sternum VIII with keel as in Fig. 18F. Cerci yellowish, 0.3-0.4 the length of FW.
Nymphs. Length of male (mm): body, 10.0-11.0 mm; cerci, 7.0-8.0; terminal filament, 5.0-5.5. Length of female (mm): body, 17.0-20.0 mm; cerci, 4.0-5.0; terminal filament, 5.0. General coloration brownish. Head (Fig. 14I-J) yellowish brown extensively shaded with grayish brown, darker on a band between ocelli, occiput with a profuse netted grayish pattern, except on paler median zone and along inner margin of eyes; paler areas also present basally to antennae and around median ocellus. Antennae yellowish white, length (mm): scape (0.5), pedicel (0.28), flagellum (2.0). Mandibular tusks with relatively large space between large basal tubercle and smaller subdistal tubercle, not C-shaped as in A. curtus but in the form of broad "C" or bracket-shaped (Fig. 14C, E, G). Thorax. Anterior ring of pronotum (collar) blackish; posterior ring shaded gray except on mediolonitudinal line and a pair of sublateral pale marks; pronotal membranes whitish. Mesonotum brownish, lighter toward apex of wingpads. Thorax ventrally whitish. Legs (Fig.  15A-D) whitish yellow shaded with gray on coxae, apex of femora and along tibiae; foretarsal claw with 20 (male) to 31 (female) denticles in a marginal row, increasing in size distally; distal region of hind femur with a group of ca. 60 (male) to ca. 100 (female) stout spines. Abdomen. Terga shaded gray dorsally, except on paler median band, a thin black median line is present inside this pale area on tergum IX-X; lateral zones of terga below gills, pale. Gills whitish almost completely shaded with gray, darker on outer (exposed) zones. Sterna whitish shaded with gray on lateral margins of sterna VIII-IX, and with grayish black on paraprocts (Fig. 15E). Caudal filaments yellowish. Etymology. From Latin "magnus" meaning "large", noun in apposition. The name alludes to the general size of the individuals, mainly the female adults.
Distribution. Only known from the type locality in Napo (Ecuador). Discussion. The type series described here as A. magnus were previously treated as A. curtus (Domínguez 1989: 173;Domínguez et al. 2006: 561), and used to record the latter species in Ecuador. As a result of our study we found some characters distinguishing these specimens as a new species, and thus A. curtus is no more considered to be present in that country. Nymphs and adults of both sexes were associated by nymphal exuviae and adults caught at the moment of emergence. Diagnosis. Asthenopus hubbardi, known from adults of both sexes, can be distinguished from the other species in the genus by the following combination of characters (one autapomorphy is listed in Appendix 2): 1) male FW 7.8-9.2 ( Fig. 16E-F), female FW 13.0; 2) forelegs of male 0.57-0.65 × the length of FW, apex of foretibia with stout spines (Fig. 20I); 3) pronotum width/length ratio: 2.1 (male), 2.7 (female); 4) FW with 4-14 relatively short marginal intercalaries (ca. 20 in female), hind wings without marginal intercalaries (present in all spaces in female); 5) FW with 2-3 cross veins between Rs and MA basal to Rs fork, in both sexes; 6) forceps relatively slender, ratio length/basal width 4.7-6.0 (Fig. 17C); 7) penes tubular and robust, furrow separating penis lobe from thumb well marked, median remnant of styliger plate subrectangular, pedestals subrectangular to subovate, relatively large (Fig. 17C); 8) female sternum VIII with reduced female sockets, anteromedian keel present (Fig. 18H); 9) eggs ratio maximum width of egg/maximum width of PC 1.1-1.3, cap formed by 3-8 filaments, chorionic plates almost contiguous (Fig. 18D).
Male imago. Length (mm): body, 7.0-7.1; FW, 7.8-9.2; HW, 3.3-4.0; leg I, 5.1-5.2; cerci, 22.0-23.0. General color whitish brown. Head whitish shaded black dorsally except thin line along hind margin, with a pair of blackish short lines anteriorly to median ocellus. Antennae whitish shaded gray at margins of scape. Thorax. Pronotum ratio width/length: 2.1. Pronotum yellowish translucent widely shaded black, except on pale transversal line between anterior and posterior rings, and on mediolongitudinal line of posterior ring, and posterolateral oblique dashes. Meso-and metanotum whitish yellow shaded black on carinae and margins, also shaded on black on posteromedian triangular zone. Legs whitish shaded gray on dorsum of foreleg, all coxae, and on legs II-III on apex of tibiae and dorsum of tarsi; foretarsal segment 1 is shown in Fig. 20I. Wings (Fig.  16E-F) membrane hyaline except apically whitish on C and Sc areas, veins translucent yellow turning hyaline distally, except basal 2/3 of veins Sc and R 1 yellowish; 2 cross veins between R and M basally to R stem. Abdomen whitish shaded almost completely with brownish gray except at pale intersegmental membranes, oblique dashes on lateral zones of terga I-VIII, submedian pale spots on anterior margin of terga II-IX, and single posteromedian pale spot on terga III-IX. Genitalia (Fig. 17C) whitish except penes yellowish. Caudal filaments whitish, slightly shaded gray on basal segments of cerci. Female imago. Length (mm): body, 10.5; FW 13.0; HW, 5.0, cerci 3.8. Similar to male, shaded dorsally more uniformly and markedly. Pronotum ratio width/length: 2.7. FW with 4 cross veins between R and M basal to R stem (none of them just below fork). Abdominal sternum VIII with keel as in Fig. 18H. Cerci yellowish turning whitish apically; ratio length cercus /FW = 0.3. Eggs (Fig. 18D). Length, 210-245 μ; width, 135-150 μ. Polar caps (maximum width, 110-135 μ) formed by 3-8 long coiled threads. The chorionic plates are almost contiguous, leaving a reduced smooth chorionic surface among them (Fig. 18D).
Etymology. The species is named for Mike Hubbard who has contributed significantly to the understanding of mayflies throughout the world.
Distribution. Two near localities in the Amazonas River from Colombia. Discussion. This species is very similar to A. angelae, and both were collected in the same lightraps in Colombia, nevertheless they can be separated because A. hubbardi shows translucent veins in the wings (brownish in A. angelae); in most specimens a cross vein is present just below R fork in A. angelae (more basal or distal in A. hubbardi), and the penes are shorter and well separated from the basal thumb in A. hubbardi (similar to A. curtus). Asthenopus hubbardi is further characterized because foretibia (Fig. 20I) presents strong marginal spines (weak or absent in A. curtus). Diagnosis. Asthenopus guarani, known from all stages, can be distinguished from the other species in the genus by the following combination of characters (seven autapomorphies are detailed in Appendix 2): 1) male FW 8.0-9.0 mm (Fig. 16G-H), female FW 16.0 mm; 2) Ratio FW/foreleg length 1.4-1.8; 3) pronotum width/length ratio: 2.15 (male), 3.1 (female); 4) male FW with 4-6 marginal intercalaries (24-26 in female), slightly shorter than the separation of main veins, HW (Fig. 16I) with 2-4 marginal intercalaries (8-9 in female); 5) male FW with 0-2 cross veins between Rs and MA basal to Rs fork (3 in female); 6) forceps relatively slender, ratio length/basal width 4.8-5.8 (Fig.  17E-F); 7) penes tubular and slender, furrow separating penis lobe from thumb well marked; median remnant of styliger plate with lateral rounded lobes as in Fig. 17E). 8) female sternum VIII with reduced, not distinguishable female sockets, but with a long anteromedian keel; 9) eggs (Fig. 18C) ratio maximum width of egg/maximum width of PC 1.2, cap formed by 3-8 filaments, disk-like structures well separated by smooth chorion, with 2-3 small disks beneath each larger disk (Fig. 18C); 10) nymph, ratio total length of mandible/mandibular tusk length 1.4; 11) inner margin of left mandibular tusk with subbasal and submedian tubercles well separated (similar to Fig. 14C, E).
Female subimago. Length (mm): body, 12.5; FW, 16.0; HW, 6.5; cerci broken off and lost. General coloration orangish yellow shaded widely black. Head dorsally blackish except medial line on occiput, and anteriorly to median ocellus. Thorax. Pronotum width 2.5 mm, total length 0.8 mm; cream shaded black on thin anterior ring, with gray on posterior ring, membranes whitish. Mesothorax orangish yellow with gray markings. Wings translucent whitish, veins whitish except basal half of C, Sc and R yellowish. Abdomen uniformely shaded gray dorsally, except on pale medial line. Sternum VIII with long and thin anteromedian keel. Base of caudal filaments whitish (rest broken off and lost).
Nymph (cuticle from reared female described above). Length (mm): body, 16.5 mm; cerci and terminal filament, 7.0 (both broken at apex). Antennae broken off and lost. Mouthparts. Mandibular tusks with relatively large space between the large basal tubercle and smaller subdistal tubercle, not C-shaped (similar to Fig. 14C, E, G). Legs. Foretarsal claw with a row of 26-29 denticles. Apex of hind femur with a group of 90-100 stout acute spines.
Etymology. The name refers to one of the etnic groups inhabiting the area where the specimens were collected.
Distribution. Argentina (Corrientes), Brazil (Sao Paulo), Uruguay (Salto). Discussion. This species is very distinctive, not only by the long and slender penis lobe, but also because of the presence of relatively long marginal intercalary veins (in male FW and HW). The reared female subimago is tentatively associated with the male, because of similarity in coloration and shared distributional range. Eggs extracted from this female show also some differences from the other known in the genus, mainly the larger extent of smooth chorion around the plates and the presence of small disks below the larger ones (Fig. 18C). Diagnosis. Only one autapomorphy was recovered in our analysis for A. angelae, a change in the ratio A (total length forceps)/E(basal width) from 6.2 to 6.5-7.1 (i.e., forceps become slightly slender). This species can be recognized by the following combination of characters: 1) FW size male 7.0-10.0 mm (Fig. 16J-K), female 12.0-17.0 mm; 2) ratio FW/foreleg length 1.4-1.6; 3) pronotum width/length ratio: 2.0-2.2 (male), 2.2-2.9 (female); 4) 5-11 imv present in male FW, relatively short and poorly anastomosed; 5) male FW with 1-4 cross veins between Rs and MA basal to Rs fork; 6) ratio total length/ basal width of forceps 6.5-7.1 (Fig. 17D); 7) penes tubular, with well developed thumb, curved ventro-medially, with apex projecting acutely, furrow separating penial lobe from thumb reduced or absent (Fig. 17D), pedestals subrectangular to subovate, relatively large; 8) female sternum VIII with anteromedian keel and reduced sockets similar to other species; 9) eggs ( Fig. 18B) with 3-5 threads on polar caps, ratio maximum width of egg/maximum width of PC 1.1-1.3, chorion with smooth area around rounded disks; 10) nymph, ratio total length of mandible / mandibular tusk length 1.6; 11) space between the subbasal and the submedian tubercles relatively long and straight (Fig. 14B).

Asthenopus angelae de Souza & Molineri
Distribution. Argentina, Bolivia, Brazil, Colombia and Peru. Discussion. This species was recently described from all the stages (de Souza and Molineri 2012), and in our analysis it appears as sister to A. magnus, but with relatively low support. In the original description, nymphs and female adults were not distinguished from A. curtus, because of lack of characters. We proposed (above in diagnosis) some characters that should be checked and confirmed with the study of more material. We suggest that specific identification of this and the other species of the genus should be confirmed when possible with the study of male genitalia. Peru was mentioned by de Souza and Molineri (2012) in the list of material but not in the distribution, so here it is added in that section.