This is a repository copy of Shifts in hexapod diversification and what Haldane could have said. White Rose Research Online URL for this paper: https://eprints.whiterose.ac.uk/1280/ Article: Mayhew, P J orcid.org/0000-0002-7346-6560 (2002) Shifts in hexapod diversification and what Haldane could have said. Proceedings of the Royal Society B: Biological Sciences. pp. 969-974. ISSN 1471-2954 https://doi.org/10.1098/rspb.2002.1957 Reuse Items deposited in White Rose Research Online are protected by copyright, with all rights reserved unless indicated otherwise. They may be downloaded and/or printed for private study, or other acts as permitted by national copyright laws. The publisher or other rights holders may allow further reproduction and re-use of the full text version. This is indicated by the licence information on the White Rose Research Online record for the item. Takedown If you consider content in White Rose Research Online to be in breach of UK law, please notify us by emailing [email protected] including the URL of the record and the reason for the withdrawal request. [email protected] https://eprints.whiterose.ac.uk/ Received 21 November 2001 Accepted 11 January 2002 Published online 22 April 2002 Shifts in hexapod diversification and what Haldane could have said Peter J. Mayhew Department of Biology, University of York, PO Box 373, York YO10 5YW, UK ([email protected]) Data on species richness and taxon age are assembled for the extant hexapod orders (insects and their six-legged relatives). Coupled with estimates of phylogenetic relatedness, and simple statistical null mod- els, these data are used to locate where, on the hexapod tree, significant changes in the rate of cladogenesis (speciation-minus-extinction rate) have occurred. Significant differences are found between many success- ive pairs of sister taxa near the base of the hexapod tree, all of which are attributable to a shift in diversifi- cation rate after the origin of the Neoptera (insects with wing flexion) and before the origin of the Holometabola (insects with complete metamorphosis). No other shifts are identifiable amongst supraordi- nal taxa. Whilst the Coleoptera have probably diversified faster than either of their putative sister lineages, they do not stand out relative to other closely related clades. These results suggest that any Creator had a fondness for a much more inclusive clade than the Coleoptera, definitely as large as the Eumetabola (Holometabola plus bugs and their relatives), and possibly as large as the entire Neoptera. Simultaneous, hence probable causative events are discussed, of which the origin of wing flexion has been the focus of much attention. Keywords: adaptive radiation; extinction; Insecta; macroevolution; speciation; tree balance 1. INTRODUCTION These have shown that the origin of novel oviposition sub- strates (Zeh et al. 1989), of phytophagy (Mitter et al. In a famous, yet possibly apocryphal event, the biologist 1988) and associations with angiosperms (Farrell 1998), J. B. S. Haldane remarked on the Creator’s ‘inordinate and the opportunity for sexual conflict (Arnqvist et al. fondness for beetles’ (Hutchinson 1959; Williamson 2000) have promoted diversification, whilst carnivorous 1992). In a macroevolutionary context, Haldane’s remark parasitism (Wiegmann et al. 1993) and the leaf mining implies that the hexapod order Coleoptera contains an habit (Connor & Taverner 1997) have not. Unique events unexpectedly large number of species. The biological have not been investigated with the same earnestness, but literature is replete with such cherished statements several such events are commonly implicated as major describing taxa, or their characteristics, that are reputed contributors to hexapod diversification. In addition to to have filled more than their fair share of our planet’s Haldane’s famous statement about Coleoptera, the origin biodiversity (Heard & Hauser 1995). In recent years, the of four taxa and their associated unique characteristics are molecular and cladistic revolutions, coupled with commonly implicated (figure 1): (i) the Insecta, possessing improved techniques for extracting information from a suite of novel characteristics often described as the insect phylogenetic trees (Mooers & Heard 1997), have made it body ground plan, (ii) the Pterygota, possessing wings, possible to re-examine such macroevolutionary hypoth- (iii) the Neoptera, possessing wing flexion, and (iv) the eses (Purvis 1996; Barraclough et al. 1999). I examine the Holometabola, possessing complete metamorphosis hypotheses relating to the diversification of hexapods, (Evans 1984; Carpenter & Burnham 1985; Carpenter probably the most species-rich class of organisms. 1992; Gullan & Cranston 2000). I use data on species The causes of macroevolutionary diversity may be richness, taxon age and phylogenetic relatedness of the environmental, or they may be novel characteristics of taxa extant hexapod orders, combined with simple statistical (key innovations) (Stanley 1979; Heard & Hauser 1995). null models to ask which, if any, of these taxa have diversi- Two general approaches allow them to be identified. The fied more rapidly than expected given their phylogenetic first is to search for characteristics that have arisen repeat- position. I then discuss what this may imply about the edly in different taxa, and to ask if taxa sharing the feature causes of hexapod diversification. are more diverse than their sister clades, which do not. However, the events that have promoted diversification 2. MATERIAL AND METHODS may also be unique, and for these a second approach is required. Taxa that have unusual rates of diversification (a) Species richness must first be correctly identified. Following this, charac- The data on species richness for each hexapod order were teristics or events associated with the origins of those taxa taken mostly from Parker (1982), which is the most recently must be examined to determine which is the most plaus- completed concurrent inventory of all major living taxa. The ible culprit. data are on number of described species only, and were com- The hexapods, particularly the insects, have been a piled by summing the estimates for each individual family com- focus of speculation and research into macroevolution for prising the order. In a few instances, family-level estimates were many decades. In recent years, several studies have not given but an estimate of the order as a whole was, in which addressed the role of repeatedly evolved characteristics. case the latter was taken. For the Diptera, a precise estimate was Proc. R. Soc. Lond. B (2002) 269, 969–974 969 2002 The Royal Society DOI 10.1098/rspb.2002.1957 970 P. J. Mayhew Hexapod diversification Collembola (springtails) Ellipura Protura (proturans) Entognatha Diplura (diplurans) Archaeognatha (bristletails) Thysanura Zygentoma (silverfish) Insecta Ephemerida (mayflies) Palaeoptera Odonata (dragonflies, damselflies) Dicondylia Plecoptera (stoneflies) Pterygota Embiidina (webspinners) Orthoptera (grasshoppers, crickets) Phasmida (walking-sticks) Metapterygota Polyneoptera Zoraptera (zorapterans) Dermaptera (earwigs) Grylloblattaria (ice-crawlers) Isoptera (termites) Dictyoptera Mantodea (mantids) Neoptera Blattaria (cockroaches) Hemiptera (true bugs) Paraneoptera Thysanoptera (thrips) Psocoptera (booklice) Psocodea Phthiraptera (lice) Eumetabola Neuropteroidea (lacewings, antlions, alderflies, snakeflies) Coleoptera (beetles) Hymenoptera (sawflies, wasps, ants, bees) Holometabola Trichoptera (caddisflies) Lepidoptera (butterflies, moths) Mecopteroidea Siphonaptera (fleas) Mecoptera (scorpionflies, hangingflies) Strepsiptera (strepsipterans) Halteria Diptera (true flies) Figure 1. One putative set of relationships between extant hexapod orders, showing higher taxa mentioned in the text. Fig. 20 in Wheeler et al. (2001), on which the initial analyses here were based, differs only in assuming a monophyletic Entognatha. Monophyly of the Thysanura, paraphyly of the Polyneoptera, and a sister grouping of Coleoptera and Strepsiptera are major alternatives also considered. not given for one species-rich family, and the ordinal richness conclusions must be restricted to the data that described species given was also very imprecise (100 000–150 000). For this present, with all appropriate caution. group, other concurrent texts (e.g. Richards & Davies 1978) gave estimates slightly below the lower estimate in Parker (b) Taxon age (1982), whilst those family-level estimates given in Parker Data on the age of each extant order, and appropriate higher summed close to the lower ordinal level estimate given. I thus taxa, were estimated by first compiling the age of the oldest fossil used 100 000 as the estimate for this order. Estimates were not definitely attributable to the taxon, from Ross & Jarzembowski given for the Odonata, nor for the non-insect hexapod groups, (1993), using the midpoint of the age span of the stratum con- for which I used other well known and, where possible, contem- cerned. Ignoring earlier but more doubtful fossils is conservative poraneous sources (Richards & Davies 1978; Davies & Tobin because it may underestimate taxon age and thus make signifi- 1985; Hopkin 1997). Whilst ignorance of the total (i.e. cant differences more difficult to detect (see § 2(d), below). described ϩ undescribed) species richness of each order is I modified the estimates of taxon age by making a further logi- frustrating, extrapolating beyond the present data introduces cal step based on