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 colorado . Praepapilio colorado Durden & Rose, 1978 Papilionidae . USA , Colorado , Ray Dome , Rio Blanco Co. ; early Lutetian, middle Eocene. Depository : private collection Hugh Rose , New Hampshire , USA ( holotype , 1). Published figures: Durden & Rose (1978: Figs 1 , 6A–B) ; Murata (1998: Figs 1 , 2 , 5 ) . Rather well preserved, seen from above, with wings spread. Length of forewing 37 mm . Two papilionid autapomorphies are present, both in the forewing: crossvein Cu-1A (Durden & Rose: Cu-V; Miller 1987 : basal spur) present, and 2A (Durden & Rose: V3 ) terminates on dorsum. In the hindwing there are two anal veins, a primitive condition, which in the Papilionidae is found in Baronia (Baroniinae) only. The posterior margin of the forewing cell is strongly trifid, a plesiomorphic condition not found in extant Papilionidae . In the forewing CuP seems to be present (interpreted by Durden & Rose as a sclerotized crease), a primitive character not found in other Papilionoidea, although the course of the lost vein can be visible in extant Hesperiidae by separate spots in the upper and lower part of the area between CuA2 and 1A+2A. Durden & Rose also interpreted a possible crease in the hindwing as CuP, but this seems unlikely, since it seems to originate from the base of 1A+2A, and it may be an artefact. Contrary to what is found in Papilioninae , there are only four radial branches in the forewing (formula: 1, 2, 3+5 (according to Miller 1987 , R4 is missing); R2 originates at upper corner of cell. Baronia also has a 4- branched radius in the forewing, but so do Parnassius and Hypermnestra , as well as species of Pieridae (where even more radial veins may be lost), Lycaenidae , Riodinidae and Nymphalidae , suggesting that the loss of a radial branch occurred more than once in the evolution of the butterflies. Therefore, this character state in Praepapilio cannot be seen as a measure of close affinity with Baronia . According to Miller (1987) the crossvein Cu-1A is an apomorphy of the subfamily Papilioninae . Durden & Rose (1978) state that the patagia have a narrow elongate sclerotization. This is an apomorphy of Baroniinae ( Ackery et al . 1999 ), but these authors indicate that the sclerotization is variable in butterflies, although generally not so much reduced as in Baronia . It is unlikely that this character is well visible in the fossil. It is even more unlikely that Durden & Rose have actually seen a narrow sclerotic band uniting the cervical sclerites beneath the neck, as they state in the list of characters agreeing with the Papilionidae , in a fossil that is seen from above. The remark by Durden & Rose (1978: 6) that "Based on wing shape, abdominal dimensions preserved as deformation of matrix, and position of abdominal margin of hindwing, this specimens is a female", must be considered speculative, and their remark ( Durden & Rose 1978: 8 ), based on a single specimen, that "Range of individual variation within sex, between sexes, and within species may be comparable to that found in the modern Baronia brevicornis …" lacks evidence. Miller's (1987) remark that "the discovery of Praepapilio confirms that genera within the Papilionidae are at least 48 million years old", is not particularly meaningful, since the recognition of a supraspecific taxon is subjective and all extant genera could well be younger. Apart from Praepapilio many other genera may have existed that presently no longer occur, so why would extant genera be of such antiquity? The discovery of the fossil does confirm that some of the apomorphies of the Papilionidae , Baroniinae and Papilioninae are at least the age of the fossil. The genus Praepapilio and the subfamily Praepapilioninae , erected by Durden & Rose (1978) for this species and its supposed fossil congener P. gracilis (see below), are not based on autapomorphic character states, but on the absence of apomorphies of other Papilionidae . In recent analyses (e.g. Zakharov et al . 2004 ; Simonsen et al . 2011 ) the extant Papilionidae are subdivided in Baroniinae , Parnassiinae and Papilioninae , which are interrelated as follows: Baroniinae +( Parnassiinae + Papilioninae ). (The Baroniinae is monotypic, consisting solely of Baronia brevicornis Salvin, 1893 .) An attempt to find the taxonomic position of Praepapilio in the phylogeny of the Papilionidae is hampered by the fact that only venational characters of the fossil are helpful. On the basis of the trifid condition of the posterior margin of the forewing cell Praepapilio is more primitive than the extant Papilionidae . The primitive condition agrees with the apparent presence of (a remnant of) CuP in the forewing. Moreover, it shares with Baronia another plesiomorphic trait, the presence of two anal veins in the hindwing. The the basal spur in the forewing is a character only found in Papilionidae , but it is absent in Baroniinae and many Parnassiinae as well as in the genus Teniopalpus of the Papilioninae ( Ackery et al. 1999 ), indicating that it was probably lost several times. Alternatively, it could quite well be a remnant of the otherwise lost CuP, and without further evidence its presence is no criterion for close relationship within Papilionidae . Summing up this evidence, if we want to use Praepapilio as a calibration point, it is best placed at the root of the Papilionidae and not higher up. Placing Praepapilio at the stem node of Parnassiinae + Papilioninae , Simonsen et al. (2011) found an estimated time of origin of the Papilionidae at 68 (53–87) Ma. Nazari et al . (2007) included the fossil as well as two other fossil Papilionidae , in their analysis of extant Papilionidae , making use of morphological and molecular characters, even though the fossils obviously lacked most characters. In this analysis Praepapilio turned up as sister to the genus Papilio . Using the fossils as well as some other information for calibration, Nazari et al . (2007) found an estimated age for the Papilionidae of c. 100 million years, strongly contrasting with the age as found by Simonsen et al . (2011) on the basis of a more realistic position of Praepapilio . The reason for this large difference is aptly discussed by Simonsen et al. (2011) . Using Praepapilio as well as Thaites ruminiana and Doritites bosnaskii for calibration, Condamine et al . (2012) estimated the earliest split in the papilionid lineage at about 52 Ma, agreeing with an age for the family not very different from Simonsen’s et al . (2011) conclusion.