First record of March flies (Insecta: Diptera: Bibionidae) from the Miocene Gračanica mine (Bugojno, Bosnia-Herzegovina)

Two fossilMarch flies (Insecta, Diptera, Bibionidae) are recorded from open lake deposits of the middleMiocene Gračanica mine near Bugojno in Bosnia and Herzegovina. These two fossils represent the first insects found in this outcrop. One specimen is described as Plecia sp. indet, the other as Bibio sp. indet. The incomplete preservation does not allow for species descriptions, but the specimens seem not to be conspecific with any known fossil bibionids. The habitus and colouration of the Bibio specimen is very similar to the recent species Bibio hortulanus (Linnaeus). The presence of the genus Plecia is considered to indicate warm climate, while Bibio points more to a temperate climate. But for reliable conclusions on the climate, much more records of bibionid fossils would be needed.


Introduction
The Gračanica mine is situated in the Bugojno Basin in central Bosnia and Herzegovina (Mandic et al. 2016; papers in this special issue). It is a large opencast mine from which mammalian remains and molluscs have been reported earlier. Based on earlier geological investigations (Mandic et al. 2011(Mandic et al. , 2016 and on gastropods (Harzhauser et al. in press, this issue), the age of the deposits was dated as middle Miocene, early Langhian and was interpreted to range between 15.4 and 15.0 Ma, correlating to MN5. Ongoing stratigraphic analyses currently indicate a slightly younger date, somewhere in the range between 15.2 and 14.0 Ma (pers. comm. O. Mandic), correlating with MN5 and early MN6. For a final integrative biostratigraphic and magnetostratigraphic dating of the Gračanica section, see Mandic et al. (in prep., this issue).
In 2017, excavations were conducted in the Gračanica mine by the JURASSICA Museum, Porrentruy (Switzerland) and by the Geological Paleontological Department of the Natural History Museum Vienna (Austria) in cooperation with the Federal Geological Survey of Bosnia and Herzegovina and the mine authorities. During these excavations, the first macroscopic arthropod fossils from this locality have been found, which comprise two insects and a crab. In this paper we report on the two insects.
The two fly fossils (and also the only other arthropod found until now, a crab) were found in the upper half of the section cropping out in the Gračanica coal mine. The upper half of the section (about 20 m in height) are clear light-coloured, laminated sediments (marls, silts, sands), whereas the lower half of the section (also about 20 m) is dominated by interlayers of coaly sediments and lignites (this is where all mammalian remains come from) (see also Mandic et al. 2016, fig. 28, personal communication Ursula Göhlich 2018. These two parts represent different habitats, and the March flies were found in what corresponds to the Blittoralô r the Bprofundal^part of an open lake (personal communication Davit Vasilyan 2017 and 2018). This fits well with other records of fossil bibionid flies which are mostly found in deposits of former lakes.
The bibionid specimens were examined under a Leica MZ12.5 stereomicroscope with an attached Camera lucida for making drawings. Photographs were taken either using a Leica MZ12.5 stereomicroscope and an attached Nikon D300 digital camera or with a Leica M165C stereomicroscope with an attached digital camera PROGRES GRYPHAX® ARKTUR (Jenoptik Optical Systems). Terminology follows Merz and Haenni (2000).   much nearer to origin of R 2 + 3 than to crossvein r-m. Vein R 2 + 3 slightly curved, R 4+5 distally distinctively curved.

Systematic palaeontology
Wing vein measurements (curves are neglected in the measurements, measured on reconstructed wing venation, Fig. 2b): length Rs basal to r-m 1.8 mm; length Rs distal to r-m 1.4 mm; length R 2 + 3 1.2 mm; length R 4+5 2.6 mm; length M distal to r-m 0.5 mm; length mm; length M1 3.1 mm; length M2 2.9 mm; length r-m 0.4 mm; length m-cu 0.4 mm; length CuA1 3.3 mm; length CuA2 2.3 mm. Legs completely black, slender; fore femur length 2.1 mm, fore tibia length 2.3 mm, length of first tarsomere of fore leg 1.6 mm.
Abdomen length ca. 5 mm, width 2.7 mm. Discussion of systematic placement: The partly preserved, flattened antennal flagellomeres are typical for Bibionidae. The bibionid genera Plecia and Penthetria are usually separated using the length of vein R 2 + 3 . This works well with the recent species but frequently becomes difficult when trying to identify fossils, since the length of this vein seems to vary in a rather continuous manner (e.g. Rice 1959). In this specimen, the shape of vein R 2 + 3 is intermediate between typical Plecia and typical Penthetria; however, the generally stocky habitus and particularly the short, inconspicuous cerci of this specimen suggest it could be placed in the genus Plecia reasonably unambiguously. The incomplete preservation of the specimen does not allow a species description.
Insecta, Diptera, Bibionidae Bibio Geoffroy, 1762 Bibio sp. indet. Description and preservation: Female. Body length from thorax to tip of abdomen 10 mm, laterally embedded, strongly compressed. Body with colouration preserved: the mesonotum, and the abdomen have a reddish-brown colour, the rest of the body including the legs is black.
Head badly preserved. Antenna incompletely preserved, seven characteristic flagellomeres preserved. Thorax length and width not measurable. Wings only partly preserved, membrane dark, veins black.
Legs black, femora of all legs appear to be thickened, length not measurable. Slender; fore femur length 2 mm, fore tibia length 2.2 mm, length of first tarsomere of fore leg 1.5 mm. Abdomen length ca. 6 mm, width less than 3 mm. Discussion of systematic placement: The partly preserved, flattened antennal flagellomeres are typical for Bibionidae. The habitus and the thickened shape of the fore tibia support an assignment to the genus Bibio, because there seem to be no rows of spines, but a strong apical spine. This is characteristic for Bibio, but also for the very rare genera Bibiodes, Bibionellus, and Enicoscollus. The incomplete preservation of the specimen does not allow a species description.
The general habitus and the colouration of this specimen are strikingly similar to the recent species Bibio hortulanus (Linnaeus). The head, pronotum, pleura and legs are black, mesonotum, and abdomen brownish-orange (likely orangered in life). The blackish-brown fumose wing is also like in modern B. hortulanus females. Apparently, the specimen also has a crossvein R-M a little less than half the length of basal Rs, and a spoon-shaped spur on the hind tibia, though these characters are more difficult to ascertain in the specimen at hand.
We cannot see that this reasonably well-preserved specimen differs from the recent Bibio hortulanus in any observable character. This raises the question if it is possible that a species from the Miocene may have survived to the present. There is some evidence for the presence of the same species from localities widely separated in time e.g. the Oligocene sites of Céreste and Aix-en-Provence (both France) (Skartveit and Nel 2017) and also Rott (Oligocene, Germany) and Radoboj (Miocene, Croatia) (Skartveit, pers. obs.), suggesting that it may be possible for a bibionid species to survive for 10 Myr or more; however, because of its insufficient preservation, we cannot be sure if this single specimen could be conspecific with B. hortulanus. Finding more specimens, preferably of both sexes is essential if this is to be confirmed. In the literature, there is one record of the recent, oriental species Bibio flavissimus Brunetti from the Miocene of China (Zhang 1989).

Comparison to other Miocene March flies (Bibionidae)
From European Fossil localities, Bibionidae are known from Mókollsdalur in Iceland , from Öhningen in Southern Germany (material revised by Skartveit and Pika 2014), from the Randeck Maar in Germany (no published species, but the material has been studied by JS) and from Radoboj in Croatia (Skartveit and Krizmanić, manuscript in preparation). Recently, a single specimen of Penthetria was recorded from Zillerleite in Germany (Beaury et al. 2017). JS has also seen material from the Miocene localities Andance and Sainte-Reine, both in France and from Bellver in Spain. The specimens were also compared to material from the late Oligocene sites of Aix-en-Provence, France (Skartveit and Nel 2017), Rott (Statz 1943) and Enspel (Wedmann 2000), both in Germany, and Bes-Konak in Turkey (Skartveit and Nel 2012).
Concerning MJSN GRC-127, determined as Plecia sp. indet., there are a few female specimens from Randeck Maar which may habitually resemble this specimen, but it does not resemble any species known from Öhningen (Skartveit and Pika 2014) nor from Radoboj (Skartveit and Krizmanić, manuscript in preparation). There also do not appear to be any species closely resembling this species from the late Oligocene localities Aix-en-Provence or Rott.
The specimen MJSN GRC-126, determined as Bibio sp. indet, does not resemble any species known from Radoboj (Skartveit and Krizmanić, manuscript in preparation) nor from Mókollsdalur . Heer (1849: 223) stated that the nominal species Bibio lividus Heer from Radoboj was coloured like the recent Bibio hortulanus; however, this was not confirmed by the examination of the holotype (Skartveit and Krizmanić, manuscript in preparation). Also, Bibio lividus is considerably larger than the present specimen with a body length of approximately 15 mm. In the key to species from Öhningen (Skartveit and Pika 2014), this species would key out to Bibio brevis Heer, but it differs from this species in numerous characters e.g. the colour of mesonotum and wings. It also does not resemble any species from Randeck Maar (JS unpublished observation). From the Spanish Miocene locality Bellver there are some isolated wings known which appear similar to the present specimen in size and colour (material in Musée National d'Histoire Naturelle, Paris) but no meaningful comparison can be made since the wings of specimen GRA 2017-8 are poorly preserved. There are no species from the late Oligocene sites closely resembling this specimen.

Palaeoecological inferences of the bibionid fossils
Recent species of Plecia are widespread in the tropics and are dominating elements of tropical bibionid faunas, while in temperate climates species of Bibio dominate the faunas (Hardy 1950(Hardy , 1981(Hardy , 1989. For example, in the holarctic region, there are many more species of Bibio than of Plecia. In Europe today, no species of Plecia occur, and there are only two Nearctic species of Plecia (Hardy 1965, Krivosheina 1986). And in the Neotropical, Australasian, Oriental and Afrotropical regions species of Plecia are much more common than Bibio (Fitzgerald 2004).
Based on the distribution of extant genera, fossil March flies have been used to infer temperature changes during the Cenozoic (Gentilini 1991;Wedmann 1998Wedmann , 2000Collomb et al. 2008). Especially Plecia is considered to be a good indicator of warm conditions (Collomb et al. 2008).
In the Gračanica material, the presence of one Plecia and one Bibio could point to a rather warm climate, but of course much more specimens are required for reliable conclusions.