A conspectus on the Canacidae (Diptera) of Brazil

Abstract Species of Canacidae sensu lato of Brazil are reviewed, including the subfamilies Canacinae and Tethininae. Included are seven species in five genera with two species, Nocticanace austra and Nocticanace packhamorum, from southern Brazil being newly described. To facilitate identification, we have included keys and diagnose to taxa at all levels.


Introduction
The Canacidae of Brazil have never been treated comprehensively even though specimens are often abundant and species are relatively diverse on beaches of this large Neotropical and biologically diverse country where life on the beach is often a way of life. This deficiency is not uncommon, however, and characterizes many insect families occurring on beaches, especially groups that have relatively few species, that are collected infrequently despite being common locally, and that have no species of known economic importance. Although the Canacidae lack pestiferous species, study of the family is warranted, as its species comprise an important component of the beach fauna. Beyond satisfying the immediate objective--a taxonomic conspectus of the Brazilian fauna--we are also seeking to discover and contribute toward other aspects of their natural history, such as their distribution, historical biogeography, ecology, behavior, and biodiversity. The underlying basis for all of these studies, however, is the taxonomy of the included taxa.
True flies of the family Canacidae occur in cool-temperate and tropical zones of the world, primarily on or near seashores with oceanic climates. A few species are found inland, usually in saline or alkaline environments, but occasionally in meadow-like habitats or in freshwater, such as the streams of Hawaii. Worldwide there are 308 species in the family (6 subfamilies, 27 genera) Munari and Stuke 2011), and from Brazil, we now have records of seven species and others are likely to be discovered here. The species of Brazil are in two subfamilies and five genera. Although the Canacidae of Brazil have never been treated comprehensively until now, the research published here had its genesis and basis in the works of others, as noted in the synonymy for the taxonomic categories from family to species that are included here. The classification adopted here is intended to provide perspective for this faunistic study and to serve as the organizational structure for this paper.
The historical record concerning Canacidae from Brazil began slightly more than a century ago when Williston (1896) described two species (Anthomyza cinerea (= Tethina willistoni (Melander)) and Rhicnoessa xanthopoda) from specimens collected on the island of St. Vincent (Caribbean). During the intervening 100 years, another species, Tethina albula (Loew), had been reported from Brazil Mathis and Munari 1996) but was apparently based on a misidentification. We have not examined any specimens of T. albula, and all specimens that are light colored, including mostly pale setae, are T. willistoni. Herein we review seven species in five genera that occur in Brazil. Three of these genera are reported for the first time from Brazil, as are four of the species. Two species, Nocticanace austra and N. packhamorum, are new to science and are described in this paper.
Because many species of Canacidae are widespread, especially those that occur in coastal marine habitats, we have examined most New World species, including primary types, to determine the correct identifications and valid names for the included species.

Materials and methods
The descriptive terminology for external structures and many internal structures follows that published in the Manual of Nearctic Diptera (J. F. McAlpine 1981). For structures of the male terminalia, however, we have adopted the terminology that Cumming et al. (1995) have suggested. Because specimens are small, usually less than 5.0 mm in length, study and illustration required use of dissecting and compound microscopes. Two wing ratios used in the descriptions are defined below (ratios are averages of three specimens: the largest, smallest, and one other).
1. Costal section ratios are the relative straight line distances between the apices of the subcosta and vein R 1 : apices of R 1 and R 2+3 ; and apices R 2+3 and R 4+5 . 2. M vein ratio: the straight line distance along M between crossveins (r-m and dmcu)/distance apicad of crossvein dm-cu. Label data from each specimen were recorded and listed alphabetically according to country, state or province, county, and specific locality, such as city. As available, date of collection, collector, sex, and specimen location were listed. Label data from holotype specimens were recorded exactly, and clarifying information, such as script style and label color, is enclosed within brackets.
Dissections of male and female genitalia and descriptions were performed using the method of Clausen and Cook (1971) and Grimaldi (1987). Microforceps were used to remove abdomens, which were macerated in a potassium or sodium hydroxide solution. Cleared genitalia were rinsed in distilled water and 70% ethanol and then transferred to glycerin for observation. If necessary for proper orientation, the genitalia were transferred from glycerin to glycerin jelly. The glycerin jelly was heated, and the genitalia appropriately oriented. After cooling, the embedded specimen became immobilized. Abdomens were placed in an attached plastic microvial filled with glycerin and attached to the pin supporting the remainder of the insect from which it was removed. For freshly caught specimens, we recommend that the epandrium and associated structures of the male terminalia be teased open, thus allowing examination of these structures and identification of the species without need of dissection.
Species' descriptions are composite and not based solely on the holotypes, and paired structures are described in the singular except where the context makes this inappropriate.
Although most specimens for this study are in the Department of Zoology, Universidade Federal do Paraná, Curitiba, Brazil (DZUP) and the National Museum of Natural History, Smithsonian Institution, Washington, D. C. (USNM), we also studied numerous specimens that were borrowed and are deposited elsewhere. These include (acronyms that are used in the text are noted first): ; posterior surstylus partially articulated or fused with epandrium; inner basal corner of surstylus connected to broad interparameral sclerite; cercus very short to exceptionally developed (Horaismopterinae); postabdomen of female more or less telescopically retractile; 2 sclerotized spermathecae variable in shape, below with a nar-rower cylindrical extension into the spermathecal duct; cercus subcylindrical to compressed, or even tapered distally, sometimes bearing stout to pointed, spinelike setulae. Discussion. Our concept of Canacidae includes what had been considered as two families, the Canacidae and Tethinidae. At the familial level, J. F. McAlpine (1989McAlpine ( : 1472 identified five synapomorphies that link Canacidae with Tethinidae and noted that "...these are clear indications of a sister-group relationship between them ... and may even indicate that they are subgroups of a single family." Other authors (Hennig 1958;Griffiths 1972;McAlpine 1982;Freidberg 1995) have also suggested a relationship with the family Tethinidae, and Griffiths (1972) further noted some affinities with the Chloropidae and Milichiidae. According to J. F. McAlpine's (1989) cladogram, which included an analysis of 25 characters for the families Canacidae and Tethinidae, the superfamily Carnoidea (= Chloropoidea) comprises the families with the following relationships in parenthetic notation: ((Australimyzidae, Braulidae) Carnidae)((Tethinidae, Canacidae)((Milichiidae, Risidae) ((Cryptochetidae, Chloropidae)))).

BPBM
More recently, Buck (2006) andD. K. McAlpine (2007) provided rather compelling character evidence, substantiating that these two families are closely associated, and more specifically that the Canacidae sensu stricto are an included lineage within the Tethinidae. Thus, not to include the Canacidae within the Tethinidae would render the Tethinidae as a paraphyletic family. Buck and D. K. McAlpine cited ten synapomorphies that corroborate the monophyly of the family Canacidae sensu lato (the family-group name Canacidae is older than Tethinidae). These synapomorphies are (only derived state cited): (1) Precoxal bridge present; (2) anepisternum with enlarged, dorsally curved setae at posteroventral corner; (3) vein A 2 long, present as a fold; (4) male sternite 6 reduced and divided medially; (5) male tergite 6 fused with sternite 8, forming a symmetrical pregenital sclerite; (6) male sternite 7 lost; (7) postgonites firmly connected laterally to base of phallapodeme, distinctly anterior to basiphallus; (8) hypandrium forming a sheath or phallic mantle around the postgonite and basiphallus; (9) cuticle of larva with covering of fine spicules, and (10) halobiontic in habitat preference, secondarily in freshwater habitats. Buck (2006) further suggested that the sister group to Canacinae sensu stricto is the subfamily Apetaeninae and not Zaleinae and provided four characters as corroborative evidence for this relationship: (1) antennae broadly separated, inserted more or less on protuberant facial tubercles; (2) clypeus distinctly enlarged and produced anteriorly; (3) prementum distinctly emarginated apically; and (4) tentorial arms of head capsule enormously developed and strongly sclerotized.
Puparium. Brown, similar in size and form to third-instar larva, rather spindleshaped, curved at each end; integumental spinules more prominent than on larva and anterior respiratory processes fully extended.
Biology. All Canacinae from the New World occur in intertidal habitats and are sometimes called surf flies. Although the natural history of the subfamily is poorly known, the larvae and adults are probably grazers on algae or are saprophytic in both saline and freshwater habitats. In Brazil, all species of the subfamily Canacinae occur in the littoral biotic region.
Discussion. Adult of Canacinae are similar and sometimes confused with shore flies (Ephydridae) and most species described in the 19 th century were placed in the Ephydridae. Canacids are distinguished by the wing venation (cells bm and cup complete) and by the additional abdominal segments (5 in ephydrid males, 6 in canacids), which in females terminate as an elongate and fused epiproct+cercus that bears enlarged, apical setae.
The Canacinae now include 122 valid species that are placed in 11 genera (Wirth 1951;Mathis 1992;Munari and Mathis 2010). The New World fauna comprises five genera and 35 species (Wirth 1965(Wirth , 1975(Wirth , 1987Mathis 1992). No fossils are known. Mathis' catalog (1992) included all species then known plus references to papers containing keys and illustrations. The recent catalog of Munari and Mathis (2010) is a complete updating, including keys to all known genera. In the New World, Mathis (1989Mathis ( , 1997 reviewed the surf-fly fauna for the Caribbean and Gulf of Mexico. Mathis (1982) proposed a classification for the Canacinae sensu stricto that should be revised. The subfamily includes two tribes, Canacini and Nocticanacini. The Canacini are represented in the New World by a single genus, Canacea Cresson, which belongs to the subtribe Dynomiellina. The Nocticanacini are represented by three genera in the New World, Canaceoides Cresson, Nocticanace, and Paracanace. Procanace, the fifth New World genus, was initially placed in Nocticanacini, but it is now evident that this genus is the sister group to all other genera of the subfamily Canacinae. Diagnosis. Small to medium-sized beach flies, body length 1.80-3.70 mm; general coloration grayish black to black. Head: Interfrontal setae 1 pair; postocellar setae either absent or much reduced, less than 1/4 length of ocellar setae; ocelli arranged to form an isosceles triangle, distance between posterior ocelli greater than that between either posterior ocellus and the anterior ocellus. Two-3 long dorsoclinate genal setae; anteroclinate genal setae moderately well developed, at least 1/2 length of larger dorsoclinate genal setae. Epistomal margin sinuous; clypeus low, width subequal to length of antenna. Palpus grayish black, bearing 1 to several long setae, each seta 2-3 times greatest width of palpus. Thorax: Anepisternum with scattered setulae; proepisternal seta absent; katepisternal seta present, well developed. Legs entirely dark colored, grayish black; forefemur bearing 4-6 long and evenly spaced setae along posteroventral margin, length of setae at least equal to and usually greater than width of femur. Discussion. This is the most species-rich genus of surf flies (Canacinae; 35 species) and has greatest species diversity in the Old World (Mathis 1992). The New World fauna now comprises 14 species. The species known from Brazil belong to the pacifica, and galapagensis groups.  Mathis & Marinoni, DZUP [red]."The holotype is double mounted (minuten in a block of plastic), is in excellent condition, and is deposited in DZUP. Seventeen paratypes (13♂, 4♀; DZUP, USNM) bear the same label data as the holotype.

Annotated Key to Species Groups of the Genus Nocticanace
Distribution. Neotropical: Brazil (Santa Catarina). Etymology. The specific epithet, packhamorum, is a Latin genitive patronym to recognize and honor Dean and Ieda Packham, who guided us to the type locality and offered hospitality.

Remarks.
Finding a species of the pacifica group along the Atlantic beaches of southern Brazil was unanticipated.
The galapagensis Group Diagnosis. Thorax: Acrostichal setae lacking; apical scutellar setae nearly straight in lateral view, slightly convergent in dorsal view, but not distinctly curved dorsally; anterior notopleural seta present but weaker than posterior seta; proepisternal seta(e) present; midfemur of male lacking comblike row of setae; hind basitarsomere lacking spinelike basoventral setae. Wing with length of apical section of vein CuA 1 long, length nearly twice that of crossvein dm-cu; M vein index 0.42-0.49.
Discussion. The galapagensis group now comprises nine species with the addition of the new species described below. Previously, there were eight species (Nocticanace arnaudi Wirth, N. cancer Wirth, N. curioi Wirth, N. darwini Wirth, N. galapagensis (Curran), N. scapanius Wirth, N. spinicosta Wirth, and N. usingeri Wirth) that were only known from the Galápagos Archipelago and southwestern Nearctic Region. The discovery of N. austra from southern Brazil is a major range extension for this species group and perhaps indicates a more extensive distribution in southern South America for the group. Better sampling in southern South America is urgently needed to test this possibility. Diagnosis. As in the species group diagnosis with the following additions: Small to moderately small beach flies, body length 1.80-2.40 mm, of the galapagensis group (see key to species groups). Head (Figs 3-4): 3 large dorsoclinate and 1 inclinate genal setae. Thorax: Scutellar disc with 1 pair of setae, apical scutellar setae very shallowly curved, not distinctly oriented dorsally compared with lateral scutellar setae. Legs generally gray, with basitarsomeres blackish gray dorsally. Abdomen: Tergites generally gray or slightly brownish gray medially. Male terminalia as follows (Figs 5-6): Epandrium in posterior view bearing long setulae on dorsal half, with medial projection at level of dorsal 1/3 from each lateral arm, forming a cercal cavity, but cerci not evident; medial margin thereafter ventrally forming a wide cavity that narrows ventrally because of medially directly surstyli; surstylus broadly attached or fused to ventral margin of epandrium, in lateral view only slightly narrower than ventral portion of epandrium, essentially an extension of epandrium, slightly swollen posteroventrally, bearing numerous short setulae along posterior margin, ventral margin shallowly bifurcate, forming posterior and anterior lobes, posterior lobe slightly shorter than anterior lobe, gently rounded; anterior lobe more robustly developed than posterior lobe, bluntly rounded to truncate apically, very slightly produced anteroventrally as a shallow, obtuse, point, in posterior view with posterior lobe of surstylus extended medially, pointed apically, anterior lobe more broadly developed apically. Etymology. The specific epithet, austra, is of Latin derivation and means southern, referring to the distribution of this species in the Southern Hemisphere.
Remarks. This species differs from congeners in the galapagos group in structures of the male terminalia, especially the shape of the surstylus (see figures and description above). The surstylus has a shallow, ventral bifurcation, somewhat like N. wirthi, but is more narrowly developed, like N. panamensis. The anteroventral surstylar lobe is slightly longer than the posterior lobe. Diagnosis. Small to moderately small beach flies, body length 1.40-2.60 mm; generally densely microtomentose, gray, with face and gena usually whitish gray, frons light brown, mesonotum with some brown coloration. Head: Interfrontal setae 2; postocellar seta well developed, proclinate and very slightly divergent, subequal in length to interfrontal setae; ocelli arranged in isosceles triangle, with greater distance between posterior ocelli. Two to 3 large dorsoclinate genal setae; anteroclinate genal seta well developed, subequal in length to larger dorsoclinate genal setae; epistomal margin sinuous; clypeus low, width more than 4× height; palpus yellowish. Thorax: Mesonotum darker than pleural areas, usually light to blackish brown, becoming lighter laterally. Acrostichal setulae in 2-4 irregular rows, with a distinctly larger prescutellar pair; scutellar disc lacking setulae; apical scutellar setae not oriented dorsally; anterior notopleural seta usually present (very weak or absent in one species); proepisternal seta(e) present; anepisternum with scattered setulae; katepisternal seta present. Femora and tibiae gray to blackish gray; tarsomeres yellow to dark brown, apical 2-3 tarsomeres darker; midfemur of male bearing comblike row of setae along posteroventral surface; midtibia bearing short evenly spaced setulae along ventral surface; hindtibia lacking spinelike setae apically. Wing with length of apical section of vein CuA 1 twice or more that of crossvein dm-cu; M vein ratio 0.35-0.45. Abdomen: Male terminalia: Surstylus a simple, narrow, posteriorly shallowly curved, setulose process extended from ventral margin of epandrium.

Genus Paracanace Mathis and Wirth (8 species in the New
Discussion. Like Canacea, all of the described species of Paracanace occur in the New World, with primarily tropical or subtropical distributions (Mathis and Wirth 1978).
Although two species groups are recognized in the key to species within Paracanace, adhering to the cladogram for the species of this genus (Mathis and Wirth 1978: 535), these groups are mostly for convenience and no phylogenetic signal should be attributed. Diagnosis. This species is similar to other species of the maritima group but can be distinguished from other congeners by the following combination of characters: As in generic descriptions and key to species with the following details: Generally appearing setulose, although less so than P. maritima; body length 1.54-2.05 mm. Head: Frons moderately densely golden brown to brownish tan microtomentose; face microtomentose, mostly silvery white, with some faint grayish blue near middle; gena similar in coloration and vestiture to face but more silvery white, with some gray adjacent to anteroventral margin of eye; middle dorsoclinate genal seta subequal in length to setae on either side. Thorax: Mesonotum tan to brown, becoming more grayish brown toward lateral margins and posteriorly; acrostichal setulae in 2 rows, posterior pair longer; scutellum gray; pleural area pale gray with some faint bluish coloration. Wing evenly faintly infumate, pale grayish brown; spinelike setulae along costal margin short, length less than half width of 1st costal cell; costal vein ratio 0.13-0.20; M vein ratio 0.37-0.40. Femora and tibiae gray with some darker coloration dorsally; basal 3 tarsomeres yellow, apical 2 yellowish brown to brown; long setae along posteroventral margin of forefemur with apical 1-2 black, others pale. Abdomen: Generally gray, dorsum darker, somewhat shiny, with faint metallic reflections, lateral margins dull. Male terminalia (Fig. 8)  Remarks. This species is similar and evidently closely related to P. aicen Mathis and Wirth from the West Indies, and these two species have been confused. Wirth's original description and illustration of P. oliveirai, for example, included specimens of both species in the type series, and Wirth's illustration, which is based on a specimen from the Dominican Republic, is actually P. aicen (Fig. 7), not P. oliveirai (Fig. 8). Because these two species have been confused, we present here comparable lateral views of the respective epandrium, surstylus, and cercus for both species to facilitate their identification. The illustration of P. oliveirai is the first for that species. Please note that the lateral view of the fused surstylus of P. oliveirai (Fig. 8) is more rectangular than the more elliptical shape of the comparable structure of P. aicen (Fig. 7) Genus Procanace Hendel ( Diagnosis. General coloration whitish gray, olivaceous, to blackish brown. Head: Interfrontal setae absent, but with a few setulae inserted anteriorly; fronto-orbital setae 3; ocelli arranged to form equilateral or isosceles triangle, if isosceles, the greater distance is between posterior ocelli. Arista pubescent over entire length. Two large dorsoclinate genal setae; anteroclinate genal seta moderately well developed. Palpus not bearing long setae. Epistomal margin, in lateral view, more or less horizontal. Thorax: Acrostichal setae, especially a prescutellar pair of large setae, usually lacking (setulae present in species of the williamsi group); scutellar disc lacking setae (1-2 pairs of scutellar disc setulae occur in P. nakazatoi Miyagi of the williamsi group); 2 pairs of marginal scutellar setae, apical pair not dorsoclinate; anterior and posterior notopleural setae present, length of both subequal; anepisternum with scattered setulae. Katepisternal seta usually present (lacking in species of the grisescens group). Hindtibia lacking spine-like setae apically. Abdomen: Male genitalia as follows: Epandrium in posterior view wider than high; cerci reduced, poorly sclerotized; surstylus with an anterior and posterior lobe, the latter larger, sometimes markedly so and shape unique to species. Discussion. Mathis (1988) first reported the occurrence of Procanace in the New World from specimens collected along the tidal shores of the Potomac River in Virginia. This species is now known from coastal habitats on Bermuda and from Virginia south through the West Indies to Brazil. Whether this species is adventive to the New World is unknown but likely.

Key to Species of Paracanace
The only species known from the New World is P. dianneae, which is in the cressoni group of Procanace (Mathis 1988). The cressoni group is diagnosed by the following combination of external characters: Head: Postocellar setae present, subequal to length of ocellar seta; clypeus low, width at least 4X height; palpus yellowish. Thorax: Acrostichal setulae lacking; proepisternal seta(e) present; katepisternal seta present.  11) narrowly rectangular, over 2X as long as wide; sternite 5 ( Fig. 11) wider than long, width of anterior margin subequal to that of sternite 4, becoming wider posteriorly, lateral margins irregular, widest at posterior margin, bearing a short process posterolaterally; epandrium wider than high in posterior view, bearing numerous setae, in lateral view (Fig. 10) posterodorsal margin broadly rounded, ventral margin nearly flat, anterior margin nearly straight except for anteroventral prong and irregular dorsal 1/3; surstylus (Figs 9-10) as 2 processes, anterior one much smaller, digitiform, bearing several setulae preapically and apically, posterior process much larger, length nearly equal to that of epandrium and equally as wide, in lateral view with posterior margin irregularly arched, anteroventral process very angulate in lateral view and spatulate in posterior view. Natural History. All specimens of the type series were collected from the shoreline of the tidal portion of the Potomac River at Westmoreland State Park (Virginia, United States). At the park, the river is over a mile wide, and the water is slightly brackish due largely to the tidal influence. The shore is either almost entirely sandy, the bathing area of the beach, or a combination of sand, considerable gravel, and some cobble and large rocks. In the latter habitat, the shore is quite narrow, at most two to three meters, and immediately adjacent to the shore is a cliff. In the sandy area, specimens occurred along the protected sides of narrow, wooden jetties that were installed perpendicular to the shoreline to break up the action of waves and prevent erosion of the beach. In the sand/ cobble/rock habitat, specimens were found only on rocks and were easily collected by sweeping immediately over and between the rocks. Most of the rocks and jetties were covered in part with algae, and we suspect that the larvae of this species were feeding on them.

Procanace dianneae
Remarks. Much of the temperate and tropical Atlantic Coast of the New World has some of the busiest commercial waterways in the world, and we do not dismiss the possibility that this species was introduced in conjunction with the large volume of traffic on these waters. This species has a demonstrated ability to disperse well. Although initially discovered in Virginia, where it occurs widely along the state's maritime coast, the species has now been found from Delaware south to Florida, along the Gulf Coast (Alabama and Mississippi), and into the Neotropics (Cuba and Brazil). The records from the state of Paraná are the southernmost thus far.

Diagnosis.
Adult. Small to moderately large flies, body length 1.43-3.66 mm; frequently invested with pale yellowish to brown microtomentum. 3-4 lateroclinate fronto-orbital setae, 3 inclinate frontal setae; postocellar seta convergent. Face sometimes characterized by 2 shiny protuberances laterad of facial cavity above vibrissal pore (Tethina); vibrissal seta variable, if present usually weak. 1 proepisternal seta; 1 proepimeral seta (sometimes absent in the genus Tethina); anepisternum with 1 or more setae and some setulae posteriorly. Precoxal bridge present. Wing hyaline to infuscate or pale yellow or even patterned (Tethina pictipennis Freidberg and Beschovski and T. lusitanica Munari, Almeida and Andrade); C with Sc break only; cell cup present but small; A 1 weakened apically, not reaching wing margin. Tibiae lacking preapical dorsal seta. Male epandrium bearing 2 lobes ventrally (the lobe that articulates dorsally with the subepandrial sclerite is considered to be the true surstylus while the anterior surstylar-like lobe may or may not be surstylar in origin); the true surstylus is generally strongly setulose; aedeagal apodeme long, slender; ejaculatory apodeme usually large; aedeagus usually elongate, ribbonlike, sinuous, subcylindrical, with a more or less dense ventral pubescence, often with several microscopic papillae. Female with 2 sclerotized spermathecae; cercus subcylindrical or compressed, sometimes bearing strong, spinelike setulae (pseudacanthophorites); tergites 7-8 mostly with characteristic pigmented areas; epiproct generally small, bearing a pair of setulae dorsally on apical third; hypoproct large. Natural History. Tethininae are mostly halobiont/thalassophiles, occurring in coastal marine habitats. Adults of thalassophilous species are commonly found in coastal marine habitats (Karl 1930;Munari and Vanin 2007), including the intertidal zone, wrack heaps (usually brown algae that are most abundant along temperate seashores bathed by cold currents), salt marshes, dune vegetation, and on salty soils or bare sand. We have also observed adults often in large numbers on carcasses of marine animals on beaches.
The immature stages of the subfamily are incompletely known. Ferrar (1987) provided some observations on the puparia of Tethina grisea (Fallén). Gorczytza (1988) reported on the spatial and seasonal distribution of some European species (Pelomyiella mallochi (Sturtevant), Tethina albosetulosa (Strobl), T. illota Haliday, T. flavigenis (Hendel), and T. grisea (Fallén)) from a study using color traps on the Frisian Islands of Mellum and Memmert. In nature, an abundance of individuals and a paucity of species sometimes characterize sandy sites where tethinids occur. Diagnosis. Dasyrhicnoessa is distinguished from other genera of the family by the following combination of characters: Head: Frons bearing some setulae in addition to larger setae; fronto-orbital and orbital setae usually with similar orientation, mostly reclinate or lateroclinate; fronto-orbital setae 3-4; paravertical setae more or less convergent. Head: Face lacking shiny tubercle above vibrissal pore; vibrissal seta present on apex of vibrissal angle. Eye mostly densely covered with small, pale, interfacetal setulae. Gena bare except for a ventral or nearly ventral row of setae (peristomal setae); gena narrow, about 1/8-1/3 eye height. Palpus and proboscis usually normally developed; clypeus small, if exposed not protruding anteriad beyond oral margin. Thorax: Scutum with numerous rows of coarse setulae arising from punctures; scutellar disc bare; postpronotum with 3 main setae, ventral seta curved upward; acrostichal setulae in two or more complete or nearly complete rows; prescutellar acrostichal setae present; scutellar disc bare except for marginal setae. Wing with costa not spinose; vein A 1 +CuA 2 short, much shorter than discal cell; wing usually short, about twice as long as wide (less often 2.5-3.0 times); cell bm and discal cell distinct. Forefemur generally bearing an anteroventral ctenidial comb on distal third; mid and hind tibiae evenly setulose, lacking anterodorsal or posterodorsal setae. Abdomen: Tergites wider than long; tergite 6 well differentiated from short syntergosternite 7+8, the latter forming a dorsal pregenital sclerite. Male terminalia: Epandrium with a posterior (true) surstylus, articulating with sternite 10. In some species, articulating broadly with ventral margin of epandrium, in others, reduced and positioned more dorsad, along posterior margin of epandrium. Anterior process a surstylarlike lobe, not articulating with sternite 10 but only with anterior margin of epandrium. This lobe much reduced in some species (absent in D. platypes Sasakawa) and positioned more or less medially along anterior margin of epandrium. Aedeagus long, sinuous, ribbonlike. Discussion. In the New World, a single species, D. insularis (Aldrich), is known, and was probably introduced through human commerce. Woodley and Hilburn (1994) and Mathis and Munari (1996) first recorded this genus from the New World (as D. ferruginea (Lamb)), and here we provide detailed locality data and descriptive documentation for the genus and the only known species that occurs there. We first discovered the genus and species on barrier islands off the coast of Belize and at the western margin of the Caribbean. Since then, we have found it in the United States (Florida), Mexico (Tabasco), on the Lesser Antilles (Dominica, St. Lucia, St. Vincent), and Bermuda in the western North Atlantic. The genus was probably introduced through human commerce and is now widespread throughout the Caribbean Region and perhaps beyond. Elsewhere, the genus occurs primarily within the Pacific and Indian Ocean basins where 25 species have been described thus far .

Diagnosis.
Tethina is distinguished from other genera of the subfamily Tethininae by the following combination of characters: Head: Frons bearing some setulae in addition to larger setae; fronto-orbital and orbital setae usually with similar orientation, mostly reclinate or lateroclinate; fronto-orbital setae 3-4; postocellar setae more or less convergent (lacking in T. lusitanica). Face with shiny tubercle above vibrissal pore. Eye appearing bare, setulae very sparse or lacking. Gena bare (except for Tethina pictipennis and T. lusitanica, which have scattered, inconspicuous setulae) except for a ventral or nearly ventral row of setulae; gena high in many species, height 0.50-0.75 that of eye height. Palpus and proboscis usually normally developed; clypeus small, if exposed not protruding anteriad beyond oral margin. Thorax: Scutum generally with more or less numerous rows of coarse setulae arising from punctures; scutellar disc bare; postpronotum with 3 or more setae, ventral seta curved upward; acrostichal setulae in two or more complete or nearly complete rows (lacking in T. lusitanica); prescutellar acrostichal setae present (lacking in T. lusitanica). Wing with costa not spinose; vein A 1 +CuA 2 short, much shorter than discal cell; wing usually shorter, about twice as long as wide (less often 2.5-3.0 times); cell bm and discal cell distinct. Mid and hind tibiae evenly setulose, lacking anterodorsal or posterodorsal setae. Abdomen: Tergites wider than long; tergite 6 well differentiated from short syntergosternite 7+8, the latter forming a dorsal pregenital sclerite. Male terminalia: Surstylus positioned at ventral margin of epandrium, usually broadly articulated externally with epandrium, internally with subepandrial sclerite; aedeagus usually very long and sinuous, either thick and straplike or narrow and ribbonlike; aedeagus micropubescent dorsally.
Discussion. Worldwide among genera of Tethininae, Tethina has more than half of the described species (77 of 115) (Munari 2002). Two species occur in the study area and a third, T. albula (Loew), has been reported  but not seen as part of this study. Since T. albula has been reported from Brazil, and as there is the possibility of its occurrence there, we have included it in the key to species. The included species of Tethina occur along maritime beaches of the littoral biotic region. Specimens are sometimes abundant, especially on fresh and decomposing wrack. Diagnosis. This species is distinguished from congeners by the following combination of characters: Body length 1.65-3.00 mm; body generally whitish gray to gray, microtomentose; setae generally white to slightly off white but sometimes with all setae black. Head (Fig. 15): Gena high, greater than 0.5 eye height. Thorax: 4 irregular rows of acrostichal setulae; scutellum uniformly gray; femora mostly yellow to mostly gray; hindfemora of male similar to or only slightly more swollen than fore-and midfemora; tibiae yellow; basal 4 tarsomeres yellow, apical tarsomere brown. Abdomen: Male terminalia (Figs 16-17): Surstylus articulated with and broadly attached to epandrium, broadly spatulate/triangular in posterior view, length 2-3× width, apex broadly rounded; medial margin bearing numerous short, stout setulae along entire length; surstylus in lateral view narrow, tapered to apical point, posterior margin almost straight; basal portion produced anteriorly as a broadly rounded lateral lobe bearing several short setulae medially; aedeagus thick, straplike. these two extremes, making it virtually impossible to distinguish between T. willistoni and other species on the basis of external structures alone. A very similar chaetochromatic variation is also found in the Western Palearctic Tethina albosetulosa (Strobl) (Munari and Canzoneri 1992;Munari and Vanin 2007). ( Diagnosis. This species is distinguished from congeners by the following combination of characters: Body length 1.70-3.10 mm; body with gray microtomentum; setae generally black. Head (Fig. 18): Gena short, less than 0.5 eye height. Thorax: 4 somewhat irregular rows of acrostichal setulae; apex of scutellum with yellowish to reddish spot (sometimes variable in size but always obvious); femora yellow; hindfemora of male similar to or only slightly more swollen than fore-and midfemora; tibiae and basal 4 tarsomeres yellow, apical tarsomere brown. Abdomen: Male terminalia (Figs 19-20): surstylus articulated with and broadly attached to epandrium, broadly spatulate in posterior view, length less than twice width, median margin bearing dense patch of robust setulae along entire length, apex broadly rounded; surstylus in lateral view broadly developed, lateral margin only slightly narrowed posteriorly, apex broadly rounded, lateral surface mostly bare, basal portion only slightly produced anteriorly, bearing moderately dense patch of setulae; aedeagus narrow, ribbonlike. Remarks. This widespread species can easily be distinguished from T. cohiba (often collected at the same locality) in having an obvious reddish yellow spot on the apex of the scutellum. Some specimens must be examined with the scutellum oriented to be directly viewed from behind and with good lighting. In most specimens, however, the spot is immediately obvious. Additional external characters include the mostly yellow femora, which are moderately swollen, as in T. cohiba.

Tethina xanthopoda
Illustrations of the new species were expertly produced by Mirian Nunes Morales, who is gratefully acknowledged and thanked. Other illustrations were carefully inked by Young T. Sohn from pencil drawings. For reviewing a draft of this paper we thank Lorenzo Munari and Oliver S. Flint, Jr. The plates were assembled by Michael Biondi and Karolyn Darrow. The latter also expertly photographed and assembled the plate of head photos of N. austra (Figs 3-4) in addition to general enhancement of all plates.
Recent field work in Brazil (December 2009-June 2010) that resulted in the vast majority of specimens studied in this paper was supported by a grant from CNPq (Visiting Researcher/Process number 401609/2009-0), which we gratefully acknowledge and thank. We thank Dianne Mathis for helping with all aspects of the production of this paper, especially the field work in Brazil.