Copyedited by: YS MANUSCRIPT CATEGORY: Systematic Biology Spotlight Syst. Biol. 69(5):813–819, 2020 © The Author(s) 2020. Published by Oxford University Press, on behalf of the Society of Systematic Biologists. All rights reserved. For permissions, please email: [email protected] DOI:10.1093/sysbio/syaa029 Advance Access publication April 7, 2020 Comparative Phylogenetics of Papilio Butterfly Wing Shape and Size Demonstrates Independent Hindwing and Forewing Evolution , ,∗ H. L. OWENS1 2 ,D.S.LEWIS3,F.L.CONDAMINE4,A.Y.KAWAHARA2, AND R. P. GURALNICK2 1Center for Macroecology, Evolution, and Climate, University of Copenhagen, Universitetsparken 15, 2100 Copenhagen Ø, Denmark; 2Florida Museum of 3 Natural History, University of Florida, 1659 Museum Rd, Gainesville, FL 32611, USA; Department of Biology, Burman University, 6730 University Drive, Downloaded from https://academic.oup.com/sysbio/article/69/5/813/5817324 by Royal Library Copenhagen University user on 11 September 2020 Lacombe, Alberta, Canada T4L 2E5; and 4CNRS, Institut des Sciences de l’Evolution de Montpellier, Place Eugène Bataillon, 34095 Montpellier, France ∗ Correspondence to be sent to: Center for Macroecology, Evolution, and Climate, University of Copenhagen, Universitetsparken 15, 2100 Copenhagen Ø, Denmark; E-mail: [email protected]. Received 16 December 2019; reviews returned 27 March 2020; accepted 30 March 2020 Associate Editor: April Wright Abstract.—The complex forces that shape butterfly wings have long been a subject of experimental and comparative research. Butterflies use their wings for flight, camouflage, mate recognition, warning, and mimicry. However, general patterns and correlations among wing shape and size evolution are still poorly understood. We collected geometric morphometric measurements from over 1400 digitized museum specimens of Papilio swallowtails and combined them with phylogenetic data to test two hypotheses: 1) forewing shape and size evolve independently of hindwing shape and size and 2) wing size evolves more quickly than wing shape. We also determined the major axes of wing shape variation and discovered that most shape variability occurs in hindwing tails and adjacent areas. We conclude that forewing shape and size are functionally and biomechanically constrained, whereas hindwings are more labile, perhaps in response to disruptive selective pressure for Batesian mimicry or against predation. The development of a significant, re-usable, digitized data resource will enable further investigation on tradeoffs between flight performance and ecological selective pressures, along with the degree to which intraspecific, local-scale selection may explain macroevolutionary patterns. [Batesian mimicry; Lepidoptera; geometric morphometrics; museum specimens.] “…[O]n these expanded membranes Nature writes, as Still, experimental manipulations (e.g. Jantzen on a tablet, the story of the modifications of species, and Eisner 2008) cannot characterize processes so truly do all changes of the organization register at evolutionary time scales and across lineages. themselves thereon.”—Henry Walter Bates on butterfly Comparative analysis of data from natural history wings, 1863 collections may ameliorate this shortcoming by bridging the gap between experimental manipulation For decades, researchers have examined butterfly wing and observed macroevolutionary patterns. Strauss diversity through lenses of functional adaptation, (1990) quantified variation in wing morphology in evolutionary history, and development. For nearly all select heliconiine and ithomiine butterflies and found Lepidoptera species, wings power flight to search hindwings were much more variable than forewings, for larval host plants, nectar sources, mates, and providing a tantalizing link between functional studies new territory (Scoble 1992). The physical requirements and the impact of aerodynamic constraints on wing for powered flight are thought to exert natural shape evolution. In contrast, a recent study of Morpho selective pressure on lepidopteran wing size and shape; butterflies found a strong correlation between forewing indeed, artificial selection experiments on wing and and hindwing sizes as well as shapes (Chazot et al. body size allometries have demonstrated significant 2016). Such data sets can also be used to identify fitness advantages for wild-type males compared morphological “paths of least resistance,” axes along those selectively bred for alternative allometries which diversification happens most quickly (Schluter (Frankino et al. 2005). However, evidence suggests that 1996). Comparative studies of Myotis bat skulls forewings and hindwings unequally contribute to flight (Dzeverin 2008) and whole Pheidole ants (Pie and Tschá performance: in a study of 19 species of butterflies 2013) found size evolved more quickly than shape, but and 25 species of moths, all could fly with their size variation as an evolutionary path of least resistance hindwings removed, although at the cost of speed and remains untested in Lepidoptera. maneuverability (Jantzen and Eisner 2008). Therefore, We built on this groundwork to test two predictions forewing shape and size may result from stabilizing via examination of swallowtail butterflies in one clade of selection imposed by the biomechanical requirements of the genus Papilio (subgenera Agehana, Alexanoria, Chilasa, flight, whereas hindwing shape and size may respond Heraclides, and Pterourus, hereafter “swallowtails in the more readily to neutral or selective processes such as clade of interest”): 1) forewing shape and size evolve sexual selection (Chazot et al. 2016), and predation independently of hindwing shape and size and 2) wing pressure (Sourakov 2013, Barber et al. 2015, Willmott et al. size evolves more quickly than wing shape (Table 1). 2017, Rubin et al. 2018). Our first prediction is based on the presumption that 813 [13:01 13/8/2020 Sysbio-OP-SYSB200029.tex] Page: 813 813–819 Copyedited by: YS MANUSCRIPT CATEGORY: Systematic Biology 814 SYSTEMATIC BIOLOGY VOL. 69 TABLE 1. Hypotheses examined in this study, with predictions Geometric Phylogenetic Morphometric Relationships regarding phylogenetic signal and evolutionary rate. Landmarks F1 Hypothesis Predictions F2 > Forewing shape is evolving Forewing Kmult Hindwing Kmult F3 2 2 > independently of hindwing (Forewing / Hindwing ) 1 F4 shape Forewing and hindwing r <1 PLS F5 Heraclides > F6 Forewing size is evolving Forewing Kmult Hindwing Kmult 2 2 > F7 independently of hindwing (Forewing / Hindwing ) 1 F11 size Forewing and hindwing R2 <1 F8 F9 Downloaded from https://academic.oup.com/sysbio/article/69/5/813/5817324 by Royal Library Copenhagen University user on 11 September 2020 Shape is evolving more Forewing shape 2 < Forewing size 2 quickly than size Hindwing shape 2 < Hindwing size 2 F10 H1 H2 Pterourus H3 H12 H4 Agehana the forewing is functionally constrained whereas other H5 H7 Pterourus selective pressures (e.g. predation, sexual selection) H11 H9 H6 operate on hindwing shape. The second is based H8 H10 Chilasa on the presumption that overall size change is an evolutionary path of least resistance. To test these Alexanoria hypotheses, we took morphometric measurements from FIGURE 1. Geometric morphometric landmarks and phylogeny digitized museum specimens and analyzed them in used for analysis. Phylogeny shows nonmonophyly of New World a comparative phylogenetic framework with a well- swallowtails (Heraclides + Pterourus) and subgenus Pterourus. Shape sampled and resolved species-level phylogeny (Owens landmarks, indicated by dots, adapted from Lewis et al. (2015). et al. 2017). Phylogenetic relationships from Owens et al. (2017) with bars indicating currently recognized subgenera; bar colors correspond with subsequent figure plots. Images depict species corresponding to each labeled clade. MATERIALS AND METHODS Morphometrics https://pj.fieldmuseum.org/event/626b0f98-98e7-49c8 Standardized dorsal and ventral images of Papilio -903e-d67017fe2356; MGCL: https://scan-bugs.org/ butterfly specimens with scale and color bars portal/collections/list.php?taxa=Papilio&type=1&has were obtained from four natural history museums images=1&db=70;&page=1; NMNH: https://collect (Supplementary Fig. S1 available on Dryad at ions.nmnh.si.edu/search/ento/). https://doi.org/10.5061/dryad.p2ngf1vmn): the Landmarks for morphometric measurement were American Museum of Natural History (AMNH), Field based on previous morphological work on Heraclides Museum (FMNH), Florida Museum of Natural History, swallowtails (Lewis et al. 2015; Fig. 1). One forewing McGuire Center for Lepidoptera and Biodiversity landmark (F1 in Fig. 1) was removed from final (MGCL), and the Smithsonian Institution National analysis due to particularly high rate of measurement Museum of Natural History (NMNH; Supplementary error; this was largely due to curling of the anterior Table S1 available on Dryad). Images were taken with wing margin in many specimens. To allow full view a NIKON D300S with an AF Micro-Nikkor 60mm of otherwise overlapping wing elements, we used f/2.8D lens (AMNH), Canon EOS 70D with a Canon EF 10 forewing landmarks from dorsal images and 12 50mm f/1.4 USM lens (FMNH), Canon EOS 70D with a hindwing landmarks from ventral images (Fig. 1). Canon EF-S 60mm f/2.8 Macro USM lens (MGCL), or a Landmark and 1-cm scale bar coordinates were Canon EOS 6D with a Canon EF 28-80mm f/3.5-5.6 lens collected in ImageJ 1.49 (https://imagej.nih.gov/ij/) (NMNH) mounted on either a copy stand or tripod and using the