Fossil butterflies, calibration points and the molecular clock (Lepidoptera: Papilionoidea) Author Jong, Rienk De text Zootaxa 2017 4270 1 1 63 journal article 32975 10.5281/zenodo.583183 6c479acc-8b18-4f0b-a6e5-85bcd6d7b6b7 1175-5326 583183 2D00AFF5-4FE2-4EC1-A328-C8670CFB8D6D bosniaskii . Doritites bosniaskii Rebel, 1898 Papilionidae : Parnassiinae : Luehdorfiini . Italy, Gabbro (near Pisa), Italy; Messinian, late Miocene. Depository: NHMW (holotype, 1898/0013/0001; counterpart holotype, 1898/0013/0002). Published figures: Demoulin (1975: Fig 1 ) ; Murata (1998: Figs. 35, 36) ; Ponomarenko & Schultz (1988: Pl. 7 Fig. 1 ) ; Rebel (1898: Pl. I Figs 1–3 ) . Reasonably well preserved butterfly: head, thorax, abdomen, right forewing partly overlapped by right hindwing, basal parts of left wings, all seen from the underside. Forewing radial formula: 1, 2, 3+(4+5). The posterior margin of the cell is reconstructed as quadrifid, a papilionid autapomorphy. A cubital spur (also a papilionid autapomorphy) is not apparent. Further apomorphies of the Papilionidae , such as the second anal vein on the forewing running to the dorsum, and the single anal vein on the hindwing, are not apparent in the fossil. Rebel (1898: 736) remarked: “Die länglichen Hinterflügel zeigen, wie in der Gattung Parnassius Latr. , einen stark ausgeschnittenen Innenrandtheil [...]” ("As in the genus Parnassius Latr. , the oblong hindwings have a dorsum that strongly curves inwards […]"), but the fossil is not decisive here (see also Zeuner 1961 ). At the tip of the abdomen a sphragis is visible, indicating that the fossil is a fertilized female (or at least that it has mated). A sphragis is found in members of the Parnassiinae and some Papilioninae . It could, at most, be an apomorphy at the family level and is plesiomorphic within the subfamilies. Probably as a convergent development, a sphragis also can be found in some Nymphalidae , such as the genus Actinote (see Paluch et al . 2003 ). With its rounded and untailed wings the fossil is similar to species of the genera Parnassius and Archon of the subfamily Parnassiinae . Monophyly of the Parnassinae is only weakly based on characters that are not visible in the fossil ( Ackery et al . 1999 ). Yet, the similarity with species in this subfamily is so striking, that it most probably belongs here. On the basis of recent analyses based on mitochondrial DNA ( Omoto et al . 2004 ) and mitochondrial and nuclear DNA as well as morphology ( Nazari et al . 2007 ), three monophyletic groups can be recognized in the subfamily, that were given tribal rank by Nazari et al . (2007) as follows: Parnassiini +( Luehdorfiini + Zerynthiini ). In the fossil the precostal (=humeral) vein on the hindwing is simple. This is found in Luehdorfiini and Zerynthiini , while in Parnassiini this vein is forked. The unforked condition is supposed to be apomorphic with respect to the forked condition by Hancock (1983) , who based his conclusions on morphological characters only. According to modern analyses based on DNA ( Zakharov et al . 2004 ) or on DNA and morphological characters ( Simonsen et al . 2011 ) the relationship of the subfamilies of the Papilionidae is Baroniinae +( Parnassiinae + Papilioninae ), although Nazari et al . (2007) , who also used molecular as well as morphological data found a sister group relationship between Baroniinae and Parnassiinae . A forked precostal vein is found in Baroniinae , part of the Parnassiinae (Parnassiini) and in most of the Papilioninae . In Pieridae the condition is unforked, if present; in Nymphalidae both conditions occur, in Riodinidae the precostal vein is unforked, in Lycaenidae generally absent. Such a distribution makes it difficult to attach much value to this character. If the unforked condition is apomorphic within the Papilionidae , it evolved in Parnassiini and Papilioninae , and it evolved several times in distantly related groups of Nymphalidae . If so, the simple condition of the precostal vein in Doritites bosniaskii is a plesiomorphy and need not point to a close relationship with Zerynthiini and Luehdorfiini , but does not contradict it either. The forewing has a dark outer border, which according to Rebel (1898: 737) was probably hyaline, but this idea was apparently prompted by the assumption that the fossil was in the direct ancestry of Parnassius . Further, there are four bands, one subapical between cell and apex and extending from costa to halfway between M1 and M2, one at the end of the cell, one across the central part of the cell, and one more basally in the cell. Elements of these markings can be found throughout the Parnassiinae and in some Papilioninae as well (for instance, Iphiclides , some Papilio species), though the exact place may differ a little. It is considered to be part of the original papilionid design ( Schwanwitsch 1943 ), and the fact that it resembles for instance the markings of Parnassius delphius Eversmann (cf. Rebel 1898 ) cannot be considered evidence of close relationship with the latter. The hindwing shows a dark outer border of which the proximal edge is less well defined than in the forewing, and a dark oblique band over the discocellular veins continuing to the costa. In Parnassiinae a similar oblique band is only found in Luehdorfia though elements of such a band can be found in Sericinus (female) and Hypermnestra (underside), and less clearly, in Allancastria and Zerynthia . Some elements of the band, namely a bar or irregular spot over the discocellular veins and a spot on the costa, are probably plesiomorphic within the Parnassiinae . In that case, the well-defined band of Doritites and Luehdorfia could well be synapomorphic. Rebel (1898) thought the fossil to be closely related to the genus Dorites (now Archon ), hence the name Doritites . Munroe (1961) assigned the fossil to Luehdorfia , as did Bryk (1913) half a century earlier. In view of the discussion above, a relationship of the fossil with the tribe Luehdorfiini (consisting of the genera Archon and Luehdorfia ) seems supported by the evidence. Hancock (1983) considered Doritites an offshoot of the Allancastria / Zerynthia ancestral line (i.e. Zerynthiini ), but did not give supporting characters. Nazari et al . (2007) included Doritites in their analysis, where it ended as sister taxon of Archon . At this position they used it as calibration point. As discussed under Praepapilio colorado , such an inclusion in the analysis must be rejected. If used for calibration purposes, the fossil should be placed at the ancestral line of the Luehdorfiini , a closer relationship with Archon not being supported by evidence. The colours and hindwing tail of Doritites depicted by Bryk (1914: Pl. 6 Fig. 46) were invented by him to make the relationship to Luehdorfia seems more plausible. Later ( Bryk 1934: 112 ) he wrote: “[...] ob der abweichende Praecostalsporn des Hinterflügels, der einfach distal gekrümmt und nicht 2ästig ist, tatsächlich so gebaut war oder unrichtig rekonstruiert wurde, sei dahingestellt.” ("[…] whether the deviating precostal thorn of the hindwing really was simply curved and not branched, or was reconstructed incorrectly, remains undecided."). References with regard to this fossil up to 1934 can be found in Bryk (1934) . A single paper may be mentioned here, that of Turati (1918) . This author considered Doritites bosnaskii the ancestor of two subspecies of different Parnassius species: apollo pumilus Stichel and mnemosyne calabricus Turati. More recent opinions can be found in Verity (1947) who, without much comment, assigns the fossil to “Zerynthiinae-Lühdorfiidi”, and in Demoulin (1975) who followed Bryk's (1914) suggestion. With or without valid arguments, recent authors are thus unanimous in assigning the fossil to the Palaearctic tribes Zerynthiini + Luehdorfiini . Larsen’s (1974: 34) remark that “ Parnassius -like butterflies, close to Archon apollinus , have been found in Baltic Amber” and uncritically quoted by Shields (1976) must be based on a misunderstanding. Larsson (1978) , while dealing with all insects in Baltic Amber, does not mention any butterfly. Larsen’s remark must refer to D. bosniaskii or to Thaites ruminiana , the only Parnassius -like fossil butterflies known, but not from amber. With two other papilionid fossils, this fossil was used for calibration by Condamine et al. (2012) , see discussion under Praepapilio colorado .