Neurolyga magnifica, Nel & Prokop, 2006

Nel, André & Prokop, Jakub, 2006, New fossil gall midges from the earliest Eocene French amber (Insecta, Diptera, Cecidomyiidae), Geodiversitas 28 (1), pp. 37-54 : 43-47

Neurolyga magnifica n. sp. ( Fig. 3 )

TYPE MATERIAL. — Holotype specimen PA 1678 2/3 (female).

ETYMOLOGY. — After the wonderful state of preservation of the fossil.

DIAGNOSIS (FEMALE). — Antenna 14-segmented, flagellomeres without distal sensory collar but with numerous long setae arranged in three rows, palpus with four visible long segments, compound eyes completely divided into two lateral portions and one dorsal portion, ocelli three, tibiae without disto-ventral spines and spurs, empodia well developed and very setose, macropterous; cubital fork forming almost a right angle, vein M1+2 ending on posterior wing margin, M3+4 very faint, appearing as a fold distally vanishing, M1+2 simple, R5 as long as wing, R1 more than three times length of Rs basal of r-m, cubital fork forming almost a right angle, strong anterior portion of Costa extending well beyond R5 for half distance between R5 and M1+2, along wing margin.

DESCRIPTION

Head 0.9 mm long, 0.3 mm wide, 0.33 mm high; eyes very broad, separated by a distinct space (0.04 mm long) into two large lateral eyes, 0.1 mm long, 0.25 mm wide, and a upper eye, 0.07 mm long, 0.09 mm wide; three ocelli well developed just above the upper eye; antenna 0.71 mm long, scape 0.03 mm long, 0.03 mm wide, pedicel 0.05 mm long, 0.04 mm wide, 12 long flagellomeres, of nearly the same sizes, all longer than wide (0.06 mm long, 0.07 mm wide), and with a distinct neck, but nodes of flagellomeres distinctly longer than necks (ratio length of node/length of neck = 3), last flagellomere 0.06 mm long, three complete rows of rather long setae on flagellomeres but no collar-shaped sensoria; palpus not very long, 0.18 mm long, with four long visible segments, with palpal segment 1 short and stout, palpal segment 2 0.05 mm long, palpal segment 3 0.04 mm long, palpal segment 4 rather long, but its apex is hidden by the fore tibia.

Thorax 0.48 mm long, 0.35 mm high, 0.40 mm wide, longer than broad; with numerous setae on scutum.

Wing 2.0 mm long, 0.94 mm wide, hyaline, with microtrichia and few macrotrichia on membrane; veins of posterior part of wing less strong than those of anterior part; strong anterior portion of Costa not ending at apex of R5 but continuing for 0.24 mm, close to apex of M1+2; humeral vein absent; subcostal vein very close to R, incomplete, 0.4 mm long, not reaching wing margin; vein R1 1.15 mm long, five times longer than basal part of Rs, reaching anterior wing margin 2.1 mm basal of apex of R5; R5 ending close to wing apex; vein Rs emerging from R 10.4 mm from wing base; Rs basal of cross-vein r-m very short (0.16 mm long); cross-vein r-m longer than Rs, 0.18 mm long; vein M1+2 not forked into M1 and M2, and distinctly weaker than radius, almost vanishing near its apex; vein M3+4 absent, replaced by a basal fold; CuA forked 0.72 mm from wing base, curved, fork very opened with branches of CuA nearly at right angle; CuP absent; A1 and A2 absent; anal area weakly convex; presence of at least one pore (“Sinnesporen” sensu Jaschhof 1998) on R1, two on basal part of Rs, two on part of Rs just distal r-m, and three on CuA basal of its fork.

Haltere with large knob, 0.05 mm wide and 0.18 mm long, and stem 0.16 mm long, bare.

Legs slender, without visible scales but with numerous short setae; fore femur 0.45 mm long, tibia 0.46 mm long, tarsi 0.61 mm long; mid femur 0.45 mm long, tibia 0.5 mm long, tarsi 0.6 mm long; hind femur 0.58 mm long, tibia 0.57 mm long, tarsi 0.85 mm long; femora and tibiae not especially swollen; tibial spurs absent; empodia well developed and very setose; all legs with five tarsomes, with tarsomere 1 twice as long as 2.

Abdomen (female) 1.10 mm long, 0.32 mm wide, bearing numerous long setae; sternite 8 0.09 mm long, tergite 9 0.05 mm long, cerci 2-segmented, 0.08 mm long, with apex rounded and with few short setae, shorter than those of sternite 8; spermathecae not visible.

DISCUSSION

Following the keys to Nearctic genera of Gagné (1981), Neurolyga magnifica n. sp. falls in the Micromyini subgroup of genera Campylomyza Meigen, 1818 , Neurolyga Rondani, 1840 (under the name Cordylomyia Felt, 1911 as junior synonym, see Jaschhof 1997), and Corinthomyia Felt, 1907 (all in Campylomyzini ), because of its CuA forked, M3+4 very faint, appearing as a fold distally vanishing, R5 as long as wing, tibiae without disto-ventral spines, C extending well beyond R5 for half distance between R5 and M1+2, along wing margin, R1 more than three times length of Rs, empodia broad. The new taxon has the following synapomorphies of the clade Micromyini , as outlined by Jaschhof (1998: fig. 236): anal area weakly convex, veins h, sc-r, CuP, and A absent, M1+2 simple, M3+4 reduced to a fold between M1+2 and CuA, density of macrotrichia reduced on wing membrane. Following the key to European genera of Skuhravá (1997), N. magnifica n. sp. falls in the Campylomyzini because of the following characters: ocelli three, M1+2 simple, R5 as long as wing, few macrotrichia on wing membrane, eye bridge narrowed laterally, forming medially the apparent third eye, empodia well developed and very setose. Furthermore, N. magnifica n. sp. has not the synapomorphies of the Acoenoniini , sister group of Campylomizini, viz. tibia with distal rigid and sharp setae, shortening of R5 and of distal part of C, “Sinnesporen” few ( Meyer & Spungis 1994; Jaschhof 1998).

Following the key to Campylomizini genera of Skuhravá (1997), N. magnifica n. sp. would fall near the genus Corinthomyia because of the following characters: female flagellomeres without distal sensory collar (the presence of a sensory collar is a synapomorphy of the group of genera Excrescentia + Micropteromyia + Campylomyza ); macropterous; cubital fork forming almost a right angle, more acute in Neurolyga ( Edwards 1938c) . This character is probably subject to variation and of little value. Other differences between Neurolyga and Corinthomyia concern male structures and female spermathecae, not available in our fossil ( Jaschhof 1998).

More precisely, N. magnifica n. sp. differs from Micropteromyia Mamaev, 1960 in its antennae 14- instead of 12-segmented and palpus 4-segmented ( Mamaev 1960, 1963; Jaschhof 1998). The female of Excrescentia Mamaev & Berest, 1991 has 14-segmented antennae, but most of them with a well developed collar-shaped sensoria, unlike in N. magnifica n. sp. ( Mamaev & Berest 1991; Økland 1995). The female Campylomyza have also a collar-shaped sensoria. Among the non Holarctic faunas, the Neotropical genus Psadaria Enderlein, 1940 has a female identical to those of Corinthomyia brevicornis , after Freeman (1953). Gagné (1994) indicated that it could be a junior synonym of this last genus but after the revision of Lagrange (1980: fig. 2), the wing venation of Psadaria pallida Enderlein, 1940 strongly differs from that of C. brevicornis and of N. magnifica n. sp. in its reduced vein M1+2, and veins R1, R5 and CuA not reaching wing margin. Other known Neotropical taxa correspond to genera recorded from the Holarctic. Kieffer (1919) described the genus African Tricampylomyza ( Cameroun) in the tribe “Campylomyzariae”. It has antennae with eight flagellomeres and a one-segmented palp. The Asiatic and Australian faunas are very incompletely known with species belonging to Holarctic genera.

The type female of N. magnifica n. sp. differs from the female of the recent unique species Corinthomyia brevicornis (Felt, 1907) in the following characters: 14 antennomeres instead of 11; presence of numerous long setae on flagellomeres disposed in three rows, instead of only a “subapical band of very numerous short sensory spines and some a little longer” ( Edwards 1938c: 203, fig. 13b). Thus, we can exclude the genus Corinthomyia . We tentatively attribute this fossil to the genus Neurolyga , but only the discovery of the male will allow to precise its exact position.

One “spectacular” character of N. magnifica n. sp. is its composed eyes completely divided into two lateral eyes and one dorsal eye, unlike Corinthomyia , in which the eyes are 2-3 facets wide at narrowest point ( Edwards 1938b). The eyes are completely divided into three parts in Campylomyza ( Edwards 1938c) and Tricampylomyza ( Kieffer 1919) . The situation is variable (divided or not) in the different Neurolyga species ( Edwards 1938c) , and the eyes are not divided in Excrescentia (after Mamaev & Berest 1991). Thus, this character alone is not sufficient to correctly separate the new species N. magnifica n. sp. from other Campylomizini.

Gagné (1977) described the Cretaceous genus and species Cretocordylomyia quadriseries from the Cedar Lake amber ( Canada). He considered that it is “close to Cordylomyia and Campylomyza ” on the basis of the “long, broad empodium and long R1”. Gagné indicated that the female flagellomeres of Cretocordylomyia are “similar to those of Cordylomyia spp. ” but, after his figure 7, they bear a row of long basal setae, absent in recent Neurolyga (see Edwards 1938c: figs 11, 12). Nevertheless, Cretocordylomyia differs from N. magnifica n. sp. in the presence of only one row of long setae on female flagellomeres, its fork of CuA forming an acute angle, and its vein M1+2 vanishing in wing membrane and not ending on posterior wing margin.

It is nearly impossible at this stage to be precise about the phylogenetic affinities of Cretocordylomyia within the Campylomyzini because of the following problems: the Neotropical and African genera were not introduced in the analysis of Jaschhof; our knowledge is still very incomplete, as the Asiatic and Australasian fauna are poorly known, with probably new taxa to discover. Nevertheless, the presence of fossil representatives of this clade in the Late Cretaceous and Early Tertiary demonstrates its great antiquity.

Campylomyza crassitarsis Meunier, 1904 , from the Eocene Baltic amber, has 12 flagellomeres with row(s) of long setae. Meunier (1904: figs 16, 18-20) did not figure any collar-shaped sensoria, unlike in the recent Campylomyza species. The female antennae of C. crassitarsis look very similar to those of N. magnifica n. sp., in number and ornamentation of flagellomeres. Meunier (1904) indicated nearly nothing on the wing venation and eye structure of C. crassitarsis . Thus, the exact generic position of this species is uncertain and this species should be revised. C. crassitarsis differs from N. magnifica n. sp. in its pilose wings (“ailes bien poilues”, after Meunier 1904: 29). Fedotova (in Fedotova & Perkovsky 2004) described Campylomyza falciformis , C. superposita , and Neurolyga declinata from the Eocene Rovno amber, without indicating or figuring anything on the collar-shaped sensoria of flagellomeres, the female spermathecae, and their eyes. Thus, their exact affinities are questionable. They remain difficult to compare to N. magnifica n. sp. Nevertheless, C. superposita has wings 0.9 mm long, C. falciformis has wings 1.1 mm long, instead of 2.0 in N. magnifica n. sp. Neurolyga declinata is based on a poorly preserved male specimen, apparently smaller than N. magnifica n. sp.

Figures

FIG. 3. — Neurolyga magnifica n. sp., holotype (PA 1678 2/3),general habitus:A, drawing; B, photograph of lateral view;C, photograph of dorsal view. Scale bars: 1 mm.