Dixella spinilobata Greenwalt and Moulton

Greenwalt, D, 2016, The first fossil New World Dixidae with a critical discussion of generic definitions, Palaeontologia Electronica 19 (3), pp. 1-32 : 10-17

Dixella spinilobata Greenwalt and Moulton , new species

Figures 4.3 , 10 , 11

zoobank.org/ E669A17A-4957-483C-ACC4-5E8BDF6CDF7A

Etymology. The specific epithet is derived from a combination of the terms spinosa (L) meaning spiny and lobata (L) meaning lobed and refers to the spiniform shape of both the apical lobe and the tip of the basal lobe of the gonocoxite.

Holotype. Dixella spinilobata Greenwalt and Moulton , male; NMNH, USNM 553522 .

Type Horizon. Middle Eocene Coal Creek member, Kishenehn Formation.

Type Locality. Dakin site, Middle Fork of the Flathead River (Pinnacle, Montana) .

Differential Diagnosis. Distinguished from all other known fossil Dixidae based on one or more of the following characters: first flagellomere filliform, wing length to body length ratio equal to 0.95, R 1 much closer to C than R 2, antennal length (0.61 of its total body length), size of the gonocoxite and shape of the gonostylus.

Description. Adult male ( Figure 10 ), body length 3.5 mm.

Head. Black; eyes dichoptic, large, spherical, diameter 0.222 mm. Maxillary palps long, length 0.34 mm, basal most segments not visible; segments four and five 88 x 40 and 230 x 27 μm (length x width), respectively. Antenna brown, relatively long (2.12 mm; Figure 10 ), thin. Scape and pedicel poorly resolved, pedicel spherical, 73 μm in diameter. Both F1 apparently filiform although partially covered by foreleg; dimensions unknown. Widths F2–F14 gradually decreasing. All flagellomeres setose.

Thorax. Length 0.95 mm, black with dark brown medial, lateral vittae covering scutellum; scutellum light brown. Setae absent.

Wings. Length 3.32 mm, width 1.08 mm. L/W ratio = 3.07. R1 very close to costal vein; vein R 2+3 arched. Vein Sc apex basal to Rs origin (by distance 1.5 x distance separating C, R1). Ratio R 2+3 / R 3 0.59. Wing surface covered with microtrichae; veins, margin covered with long setae approximately 59 μm in length. Haltere dark brown, length 0.55 mm, knob elongate 114 μm x 256 μm.

Legs. All legs setose; hind tibia expanded distally. Foreleg femur, tibia, tarsus lengths 1.26, 1.07, 1.69 mm, respectively. Midleg femur, tibia, tarsus lengths 1.66, 1.42, 1.78 mm, respectively. Hind leg femur, tibia, tarsus lengths 1.56, 1.58, 2.0 mm, respectively; hind leg claws without apparent teeth ( Figure 4.3 ).

Abdomen and genitalia. Abdomen brown, setose, length 2.84 mm (including terminalia). Sternite 8, tergite 9 dark brown/black, more heavily setose. Terminalia width (at widest point of two gonocoxites) 0.419 mm, each gonocoxite approximately 0.202 mm long and 0.139 mm wide; gonostylus elongate, length 0.179 mm, width 43 μm ( Figures 11.1, 11.2 ). Gonocoxite apex with possibly two or three spiniform structures, the largest and basalmost 63 μm long, 15 μm wide at base, not preserved/visible on right gonocoxite. Gonocoxite with large basal lobe, tip spiniform.

Female unknown.

Remarks

Given the demonstrably unreliable criteria for identification of adult Dixidae at the generic level (see Discussion), the option of designating the four fossil specimens described herein as “ Dixidae incertae sedis ” was considered. However, given 1) our desire to distinguish these four species from all fossil dixids described previously and 2) the observation that designations such as “Genus incertae sedis ” or “Genus indeterminate” are often not recorded in commonly used databases, a generic designation has been provided. The four new species described herein are assigned to the genus Dixella , not necessarily based on morphological criteria, but rather on the near-shore lacustrine environment in which they lived and were deposited ( Greenwalt et al., 2015; see Discussion). This of course presumes that genus-specific ecological niches (fast-flowing waters in the mid-stream of waterways or water falls [ Dixa ] vs. still waters of near-shore lake or pond environments [ Dixella ] [ Nowell, 1951]) had already been established 46 Ma. Specific characteristics, in various combinations, used to distinguish the four fossil Nearctic species from all other dixid genera both extant and extinct, other than Dixa and Dixella , are as follows (with number of species given in parentheses): Syndixa (3): R 2 fused with R 1 ( Lukashevich, 1996); Eucorethrina (3): R 1 equidistant from C and R 3 (R 2 in Peters and Cook [1966]) and Sc ending distad of Rs origin ( Lukashevich, 1996); Asiodixa Papp et al., 2006 (2): small size (wings 2.0– 2.5 mm) and Sc ending substantially before the Rs origin, Rs sigmoidal, F1 of male with a ventromedial dilation ( Papp et al., 2006); Neodixa Tonnoir, 1924 (1): R 2+3 unbranched ( Belkin,1968); Nothodixa (10): first flagellomere oval, its length equal to or less than 2.5x its width ( Belkin, 1968); Metadixa Peters and Savary, 1994 (1): known only from a single larval specimen ( Peters and Savary, 1994); Mesodixa Belkin et al., 1970 (1): F1 longer than F2 + F3, gonostylus longer than gonocoxite, curved, spinose and small size (wing 2.5 mm) ( Belkin et al., 1970); Meringodixa Nowell, 1951 (1): large spatulate ventral process of gonocoxite, F1 fusiform and 6x as long as wide, head and thorax brown and brownish yellow, respectively, stem of haltere yellow and dorsum of abdomen black ( Nowell, 1951).

The Cenozoic dixids consist of nine species including the four described herein as well as two designated as incertae sedis ( Table 2). A comparison between these is impeded by deficiencies in the published descriptions of the European fossils. Both Dixa priscula and D. cimbrica Ansorge, 1992 were described from isolated wings and, given the variability in wing venation in this family, can only be identified as members of the family Dixidae (see below). Dixa tertiaria Meunier, 1915 was originally described as a member of the family Limoniidae and, although the fossil is of an intact adult, no relevant information was provided other than a very low-resolution photograph and scale bar ( Meunier, 1915). Theobald (1937) described Dixa hyalipennis Theobald, 1937 as having venation identical to D. tertiaria and synonymized the two species; Evenhuis (1994) synonymized both under D. tertiaria . However, comparison of Theobald’s figure and Meunier’s photograph clearly shows numerous differences in wing venation. Theobald’s depiction of the first abscissa of CuA1 portrays a vein unlike that in any other species of dixid; the distal half of this vein segment is parallel with vein M and forms a smooth continuous curve with the distal portion of CuA1. Meunier’s photograph depicts the first abscissa of CuA 1 as straight and at a distinct angle to the second abscissa, in a fashion similar to all other dixids. Descriptions of genitalia are not provided and, in fact, neither specimen is identified as to its sex. Similarly, the genitalia of Dixa succinea were not visible in the original specimen described by Meunier (1906) although it was described as male (mistakenly identified as female in the figure legend). Because the type specimen of Dixa succinea has been lost, Hennig (1966) designated and described a neotype and figured its genitalia as well as those of Dixella filiforceps , D. distans and Dixa minuta Meunier, 1906 , all males. Unfortunately, Hennig’s descriptions of Dixa succinea , Dixella filiforceps and D. distans were otherwise purposefully very brief – no measurements were made – as he believed that “A detailed description of this and the following species would have little meaning. In order to clarify their position within the Paradixinae ( Dixella ), the extant species must first be carefully worked through.”

All four of the Nearctic fossil dixids are distinguished from Dixa minuta and Dixella distans in that these specimens, both Baltic amber inclusions, are much smaller in size with wing widths less than 0.6 mm ( Cockerell, 1921; Hennig, 1966). Although both Dixa priscula and Dixa cimbrica are herein declared nomina dubia (See Discussion), differences in venation between them and the four Nearctic fossils are as follows: Dixa cimbrica differs from Dixella eomarginata in that Sc originates at the origin of Rs, Rs forks basal of r-m, and m-cu originates basal of r-m; from Dixella spinilobata in that Sc originates at the origin of Rs; from Dixella curvistyla in that M 1+2 is longer than M 1, and m-cu originates basal of r-m; from Dixella intacta in that m-cu originates basal of r-m. Dixa priscula differs from Dixella eomarginata in that Sc originates basal of Rs, and m-cu originates basal of r-m; from Dixella spinilobata in that Sc originates basal of the origin of Rs; from Dixella curvistyla in that M 1+2 is longer than M 1, m-cu originates basal of r-m, and Sc originates basal of the origin of Rs; from Dixella intacta in that m-cu originates basal of r-m and Sc originates basal of the origin of Rs.

Dixella eomarginata and D. intacta differ from all other fossil dixids in being female. Dixella eomarginata differs from all other fossil dixids in having the posterior margin of the wing slightly emarginate between CuA1 and CuA2 – similar to the extant Dixella marginata . The prominent apparent false vein between M and Cu that continues between M 1+2 and CuA 1 in the left wing of D. eomarginata may be an artifact, possibly a result of partial delamination of the wing; it does not appear in the right wing. This character is not common in the ‘Nematocera’ but it is found, between M 2 and CuA 1, in both Scatopsidae ( Coboldia Melander, 1916 ) and several genera of Simuliidae ( McAlpine et al., 1981) . A false vein parallel to and just posterior of Cu and CuA 2 was described in Dixa appalachiensis ( Moulton, 2016) . While the male of Dixella marginata lacks the emarginate (excavate) wing margin of the female ( Peters and Cook, 1966), there is no evidence for or against Dixella eomarginata being conspecific with either D. curvistyla or D. spinilobata . The body and wing lengths of Dixella intacta are 34% and 48% longer than those of D. curvistyla and 29% longer than D. spinilobata , respectively. Males and females of a given species can vary in size, females commonly with larger wing and body lengths, but the relationship is not consistent (e.g., wing lengths in males of both Dixella fraxina Taber, 2010 and D. indiana Dyar, 1925 are often larger than in the female) (Takahasi, 1958; Peters and Cook, 1966; Taber, 2010). Because wing and body lengths within a single sex of a species can vary as much as 50% (Takahasi, 1958; Peters and Cook, 1966), differences in wing and body lengths alone do not preclude Dixella intacta from a conspecific relationship with D. spinilobata or D. curvistyla . However, given the absence of any analysis of female genitalia relative to the definition of the genera Dixa and Dixella , the variability in wing venation patterns between these two genera and within individual species (see below and Discussion), and the dearth of morphological detail available for the European fossils, it would be speculative to depict Dixella intacta as congeneric with any other known fossil male dixid.

The male Nearctic fossil species, Dixella curvistyla and D. spinilobata , are distinguished from those of D. succinea , D. filiforceps , and D. distans based on the shape of the gonocoxite and gonostylus. The gonocoxite of D. succinea is short relative to the gonostylus, the latter tapered distally with the terminal quarter of the stylus bent at a right angle to the base, approximately 150 μm long and less than 20 μm wide at mid-length. The apical lobe of the gonocoxite is long and narrow (about 75 μm x 10 μm). The gonostylus of D. filiforceps is very long and narrow (approximately 225 μm long and less than 20 μm wide), slightly widened at the end, and smoothly curved such that the distal half is at a right angle to the base. The apical lobe of the gonocoxite is long and narrow (about 10 μm x <10 μm). The gonocoxite of D. distans is short relative to the gonostylus and, like D. filiforceps , the gonostylus is long and narrow (approximately 125 μm long x 20 μm wide) and smoothly curved. Unlike in D. filiforceps , the gonostylus is tapered to an asymmetrical point at the end. The apical lobe of the gonocoxite is long and narrow (about 75 μm x <10 μm). These characteristics contrast with D. curvistyla in which the gonostylus is relatively short and wide (approximately 158 μm x 79 μm at its base) and bent at a near right angle and the apical lobe of the gonocoxite is quite wide (approximately 24 μm). In contrast, D. spinilobata has a straight and relatively wider rectangular gonostylus (179 μm x 43 μm) with a relatively shorter apical lobe (0.35 x the length of the gonostylus vs. ratios of 0.5–0.6 for the three specimens described by Hennig (1966)).

Dixa tertiaria is problematic in that the original description ( Meunier, 1915) is essentially devoid of useful information and no redescription has so far been produced. A photograph and a scale bar is provided and enables body (3.2 mm) and wing (2.8–3.2 mm) lengths to be determined, albeit with a degree of uncertainty. These measurements contrast with those of Theobald (1937), who described Dixa hyalipennis with a body length significantly shorter than the wing length (3.7 mm and 4.2 mm respectively). All of the Nearctic fossils have body/ wing length ratios greater than one. As described above, the very peculiar morphology of vein CuA1, as figured by Theobald (1937), also differentiates these two specimens. Without examination of the actual specimens, it is impossible to accurately compare them with the Nearctic dixids.

It is difficult to compare Dixella intacta and D. eomarginata with extant dixids of the genera Dixa and Dixella as there exists neither detailed nor extensive studies of female genitalia with respect to the taxonomy of these two genera. Some keys to females of this family exist (e.g., Disney, 1975 - which does not use genitalic characters to distinguish between genera) but they are of limited use. The structure of the genitalia of D. intacta is very similar to that of Dixa brevis as figured by Nowell (1951). The very broad and acutely curved gonostylus of Dixella curvistyla resembles that of Dixa formosana Papp, 2007 although the latter has a very short apical lobe on the gonocoxite ( Papp, 2007). Dixa neohegemonica , with a sickle/fingershaped gonostylal apex, also resembles D. curvistyla (Moulton, personal obs.); note that this structure is figured as straight in Peters and Cook (1966). The genitalic morphology of Dixella spinilobata is unique amongst extant members of this

Key to the fossil species of Dixidae family although the gonostylus of Dixa pollex Nowell, 1980 also has a small projection near its distal terminus ( Nowell, 1980). A key to all fossils of the family Dixidae , with the exception of Dixa cimbrica , D. priscula and D. tertiaria , which either consist only of a wing or are too poorly described to provide the data required for inclusion in a key, is provided here so as to facilitate identification of new fossil specimens.

See Key to the fossil species of Dixidae below.

Wing Venation Pattern Variability

Examination of the 35 different character states for vein position within extant Dixidae reveals that in only 10 instances are specific character states present in one genus and not the other. For example, state “3” for the position of the terminus of A 1 relative to the origin of Rs is not present in extant Dixella . However, this character state exists in extant Dixa in only one of the 50 species examined. In fact, in each of the 10 cases, the character state is rare in the other genus, occurring only once or twice. Therefore, no single character

1a. R 2+3 not strongly arched............................................ 2 1b. R 2+3 strongly arched............................................... 6 2a. R 1 fused with R 2 ........................................... Syndixa 3 2b. R 1 not fused with R 2 ..................................... Eucorethrina 5 3a. r-m and m-cu crossveins in-line...................................... 4 3b. m-cu distal of r-m.....................................? Syndixa liasina 4a. R 2+3 short, <¼ length of R 2+3 + R 2........ ...................... Syndixa mollis 4b. R 2+3 subequal to R 2 ..................................... Syndixa sibirica 5a. Length of Sc relative to wing length (h – apex) <0.45.... Eucorethrina westwoodi 5b. Length of Sc relative to wing length (h – apex)> 0.45≤ 0.50.. Eucorethrina flexa 5c Length of Sc relative to wing length (h – apex) ≥ 0.50.... Eucorethrina convexa 6a. Wing width <0.6 mm ............................................... 7 6b. Wing width> 1.0 mm............................................... 8 7a. Gonostylus stout, divided apically........................... Dixella minuta 7b. Gonostylus long, slender, not divided apically.................. Dixella distans 8a. Female......................................................... 9 8b. Male.......................................................... 10 9a. Wing margin emarginated between CuA 1 and CuA 2 ......... Dixella eomarginata 9b. Wing margin not emarginated.............................. Dixella intacta 10a. Gonostylus bent at about right angle apically............................ 11 10b. Gonostylus not bent apically........................................ 12 11a. Gonostylus stout, with broad base....................... Dixella curvistylus 11b. Gonostylus long, slender................................ Dixella succinea 12a. Gonostylus moderately long, L/W <5..................... Dixella spinilobata 12b. Gonostylus very long, slender, L/W> 10.................... Dixella filiforceps state is characteristic of (i.e., both universally present in and restricted to) either genus. Similarly, while there are combinations of specific character states that occur in only one of the two genera, they occur only rarely (e.g., the Dixa aliciae Johannsen, 1924 pattern occurs only once within the 50 extant Dixa species examined). There are also instances in which a specific character state or combination of character states is much more likely to occur in one genus. For example, the character state combination m-cu/r-m as “4” and r-m/Rs as “2” is 2-fold more common in Dixa than Dixella , but occurs in less than one in four of the Dixa species examined. Similarly the position of the bifurcation of R2+3 relative to that of M is three times more likely to be “3” or “4” in Dixa than in Dixella , but that combination occurs in a minority of the Dixa species. When the statistically most common combination of character states for each character is calculated for each genus, they are essentially identical: 0, 4/2, 4, 4, 4, 0, 4/0 for Dixa , and 0, 4, 4, 4, 4, 0, 0 for Dixella . Obviously, however, in order for a character state to define a clade, in this case either Dixa or Dixella , that character state must not simply occur once or even statistically most commonly, it must be invariant within that genus. Given these data and the fact that an examination of the holotype of Dixa priscula clearly shows that, in contrast to Cockerell’s original description, the crossveins r-m and m-cu are not preserved, D. priscula is reassigned to Dixidae incertae sedis. Similarly, although the venation pattern (2, 4, 0, 4, 4, 2, 0) of Dixa cimbrica is unique relative to the 92 species examined in the current study, it too is reassigned to Dixidae incertae sedis.

Figure 12.1-8 depict four examples of intraspecific venation pattern variability. A total of seven different species were examined: photographs of wings from different individuals of Dixa nova Walker, 1848 (Now classified as Dixella [ Pape and Thompson, 2013]), Dixa terna Loew, 1863 , D. rhathyme Dyar and Shannon, 1924 and Dixella obscura Loew, 1849 , for which 16, 14, seven and four specimens were examined, respectively, are shown. Dixella torrentia Lane, 1939 and D. solomonis Belkin, 1962 , for which six and seven specimens were examined, respectively, displayed variability but to a degree less than that found in the figured species. Dixella cornuta Johannsen, 1923 , for which only six specimens were examined, exhibited no variability. For four of the seven species examined, all specimens were from the same locality. Specimens of Dixa terna were from Pennsylvania and Virginia, Dixella obscura from Arizona and Alaska and Dixella cornuta from Delaware and Idaho. In all four figured species, the relative position of the crossvein r-m varies from even (2) to apical (4) of the Rs fork. The relative position of Sc differs in two of the four figured species, ranging from even to basal relative to the origin of Rs in Dixella obscura and from apical to basal in Dixa rhathyme . The position of m-cu relative to r-m is apical in one of the two specimens of Dixella obscura and basal in the other.

Figures

FIGURE 4. Tarsal claw teeth of the Kishenehn Dixella. 1, Dixella intacta left metatarsal claw; 2, D. curvistyla left metatarsal claw; 3, D. spinilobata right metatarsal claw. Arrows point to teeth. Scale bars equals 50 μm.

FIGURE 10. Habitus of Dixella spinilobata, holotype. Scale bar equals 3 mm.

FIGURE 11. Morphological features of Dixella spinilobata. 1, Photograph of the terminalia; spiniform apical lobe marked with an arrow; 2, Line drawing of a partial reconstruction of the right portion of the terminalia (dorsal view). Scale bar equals 0.25 mm.

FIGURE 12. Intraspecific wing venation pattern variability in Dixidae. 1-2, Wings from two different individuals of Dixella nova (USNM01193745, USNM01193785, ♀, ♂, respectively; NY). 3-4, Wings from two different individuals of Dixa terna (USNM01193791, USNM01193792, ♂, ♂, respectively; PA & VA). 5-6, Wings from two different individuals of Dixa rhathyme (USNM01193798, USNM01193868,?, ♀, respectively; CA). 7-8, Wings from two different individuals of Dixella obscura (USNM01193908, USNM01193974,?, ♀, respectively; AZ & AK). Collection localities are in parentheses. Arrows and arrowheads denote specific points of variation. Abbreviations: AK = Alaska, AZ = Arizona, CA = California, NY = New York, PA = Pennsylvania, VA = Virginia, USA.