Estimating Time and Space in the Evolution of the Lepidoptera
Total Page：16
File Type：pdf, Size：1020Kb
Estimating time and space in the evolution of the Lepidoptera Rienk de Jong Several aspects of estimating what happened when and where in the evolution of the Lepidoptera are discussed. Because of their scarcity and often poor preservation, fossils are not very helpful, but at least they demonstrate that, in the Oligocene some taxa of butterflies, perhaps at tribal level or higher, did occur in the Northern as well as in the Southern Hemisphere. The concept of a molecular clock is seen as a most needed test for vicariance explanations of disjunct distributions. Special emphasis is laid on the importance of calibration of the clock. The use of geological vicariance events as calibration points is rejected, because of circularity when vicariance explanations are to be tested. Fossils as calibration points should ideally be replaced by the minimum age of an apomorphous character state demonstrated by the fossil rather than a supposed identity on the basis of overall similarity. Some conditions that directed the evolution of Lepidoptera (called constraints here) are discussed for their possible use as calibration points. Estimation of the evolution in space through time (palaeobiogeography) as found in recent literature is discussed, the often supposed role (of the break-up) of Gondwana in the evolution of the butterflies is challenged, and an alternative hypothesis is given. R. de Jong, National Museum of Natural History, P.O. Box 9517, 2300 RA Leiden, The Netherlands. E-mail: [email protected] Introduction extrapolate the data to far back in history and to Estimating time and space of events in the evolu- make use of (reputed) circumstantial evidence. tion of life has been an important issue ever since the Thus, Forbes (1932), starting from a well pre- acceptance of the concept of evolution. A fair idea served butterfly fossil (Prodryas persephone Scudder, of these events is crucial in understanding how the 1878, about 20 Ma old), concluded that “there is evolution of a group of organisms has been shaped a slight weight of probability that the Lepidoptera not only by intrinsic factors (mutation, population arose in the late Carboniferous or early Permian pe- growth, food preference, etc.), but by extrinsic, en- riod ...” (i.e. some 280–300 million years ago). Not vironmental factors as well. Initially, and for a very only fossils, also the study of recent taxa led some long time, fossils formed the only clue to what had authors to unwarranted extrapolations and specu- happened in deep history. Although the first fossil lations on the origin of higher taxa. In a study of lepidopteron was described well before the idea of American Copper butterflies (Lycaeninae) Miller & evolution had clearly been formulated (Charpentier Brown (1979) speculated that the basal split in the 1843; described as Sphinx atava [Sphingidae], moved Lycaeninae took place over 100 million years ago to Nymphalidae by later authors), fossil Lepidoptera and, thus, the origin of the “Lycaenoid root” should remained too rare to play an important role in elu- be placed much further back into the Mesozoic. cidating the evolutionary history of the Lepidoptera, More generally, Brown (1987), while discussing the both in time and space. Every attempt to estimate evolution of Neotropical butterflies, suggested that the origin of the Lepidoptera from fossils had to pantropical groups at the level of subfamily and Tijdschrift voor Entomologie 150: 319–346, Figs 1–7. [ISSN 0040–7496]. http://www.nev.nl/tve © 2007 Nederlandse Entomologische Vereniging. Published 1 December 2007. Downloaded from Brill.com09/24/2021 11:27:01AM via free access 320 Tijdschrift voor Entomologie, volume 150, 2007 tribe originated in the Mesozoic and the early super- both described by Durden & Rose (1978) from the families of the Lepidoptera in the Palaeozoic. Several (middle Eocene) Florissant fossil beds in Colorado. developments have demonstrated the untenability of The fossils demonstrate two papilionid apomorphies such broad speculations. The rise of cladistics, fast in the venation of the forewing, viz. the presence computer programs and molecular techniques in the of a “basal spur” (term by Miller 1987; a cross-vein second half of the 20th century have largely taken between Cu and 1A), and vein 2A curves to the over the role of fossils in evolutionary studies. The hind margin instead of to 1A. Two other promi- same developments, enhanced by an incomparably nent apomorphies of the papilionids, viz., in forew- better fossil record than found for Lepidoptera, have ing cubitus and lower discocellular vein in line (i.e., also led to a much better understanding of the evolu- cubitus seemingly quadrifid), and in hindwing only tion of the angiosperms (Soltis et al. 2005, Wikström a single anal vein (only in Baronia there are two et al. 2004), the foodplants of all Lepidoptera except anal veins), are not present in Praepapilio, i.e. these the basal-most lineages and some secondary changes conditions are in a plesiomorphous state. As a con- to other food sources. The dependence on food- sequence, Praepapilio does not fit in our system of plants set constraints on the estimation of the age of Papilionidae and the only logical place is at the root the Lepidoptera. Also other environmental factors, of the Papilionidae. leading to particular adaptations, may help to date Notwithstanding the limitations, a fossil fixes a point evolutionary events in Lepidoptera, if we know the in the evolution of a taxon in time and space, the age of the environmental factor. aspects we are interested in, in this contribution. In the present contribution several aspects of estimat- Unfortunately, fossil Lepidoptera are extremely rare. ing divergence times in Lepidoptera are discussed, Kristensen & Skalski (1998; see also many refer- aiming at a better understanding of the interplay of ences therein) estimate the total number known at evolutionary and distributional changes. The survey some 600–700 specimens, covering about 200 mil- is not exhaustive, and emphasis is laid on the evolu- lion years (the oldest known fossil is Archaeolepis tion of skippers and butterflies. mane, described by Whalley (1985), from the Lower Lias of Dorset, England, based on a wing fragment with scales which show a similarity to the scales of Estimating time the most primitive extant Lepidoptera, Micropter- Fossils igidae). From the whole of the Mesozoic era only 19 specimens are known that are considered lepidop- Traditionally fossils are considered the key to the terous (Whalley 1986), all other Lepidoptera fossils past. Indeed, without fossils we would have no idea are from the Tertiary, and a few from the Pleistocene. of the enormous diversity of dinosaurs or extinct Remarkably, the proportion of butterflies among the mammals. It is difficult to assess, however, how far fossils in number of specimens (some 80 specimens fossils have contributed to our understanding of known) is close to the proportions among extant the phylogeny of the groups concerned and have Lepidoptera in species. changed patterns of relationship that had been de- Because of their scarcity, and also their often frag- rived from analysis of extant organisms. The find- mentary nature, fossil Lepidoptera are not very help- ing of a new fossil is not different from the finding ful in estimating time and space in the evolution of of a new extant species when it comes to allocation the Lepidoptera, and conclusions are surrounded in an existing system, but while, in an extant spe- by many “ifs”. Recently, Hermsen & Hendricks cies, all characters used to build up the system can (2007) described a method for constraining the age be studied and analysed in a cladistic way, fossils are of origination of derived characters with the help of usually very fragmentary. In most cases, their place fossils. However, for that method fossil taxa must in the system is more determined by overall similar- be incorporated in a phylogenetic analysis of extant ity than by a critical evaluation of apomorphous and taxa. The authors recognize and discuss the problems plesiomorphous character states, as correctly stated associated with fragmentary and missing characters. by Kristensen & Skalski (1998). If a fossil shares an It would seem that the lepidopterous fossils simply apomorphous state of a character with a group of ex- are too badly preserved for that purpose. But surely, tant taxa, but has retained the plesiomorphous state as far as apomorphous character states can be recog- of one or some characters that have a derived state nized in a fossil, the minimum age of that character in the extant group, the fossil is placed at the root of state can be plotted on the tree of extant taxa. This that group. An example is provided by the famous is actually what is done (or should be done) when butterfly fossils Praepapilio colorado and gracilis, fossils are used to calibrate the molecular clock (see Downloaded from Brill.com09/24/2021 11:27:01AM via free access De Jong: Time and space in Lepidoptera 321 Calibration by fossils below), divergence being nothing else but the origi- nation of a derived character state in one of two sister As stated above, butterfly fossils are rare and usually species. difficult to assign to a taxon rank below subfamily. A critical review of all known butterfly fossils will be Therefore, their application as calibration points published separately. In the present study only those (minimum age) must be done with great care, and fossils that have been used in literature for calibra- fossils should not be identified on the basis of simi- tion of the molecular clock are discussed. The list larity only, but on apomorphous characters. provided by Braby et al. (2005) is incomplete (also Braby at al. (2006) used four fossil pierids to calibrate for the time frame chosen) and uncritical, and the the molecular clock for their tree of the Pieridae.