Exploring COI variation in Ecuadorian butterflies: The first step in an integrative approach to resolving species limits in this widely distributed & cryptic genus Denise Tan & Keith R. Willmott McGuire Center of & Biodiversity, Museum of Natural History & Dept. of Entomology & Nematology, University of Florida

OUTGROUPS GMYC bPTP BOTH

1 Introduction 1 Results & Discussion H. fallax 1 - 13

0.99

The Lepidopteran genus Hermeuptychia 1 H. sp n 2 1 • 234 Ecuadorian samples have been sequenced, with H. sp n 2 2 0.98 H. sp NS (: Satyrinae: Euptychiina) was 1 H. intricata 1 - 5 pairwise genetic distances of 0% – 5.46% 0.79 0.95 1 1 H. gisella 1 first proposed by Forster in 1964 and has an H. atalanta 1 0.94 • 225 of the 520 sequences (combined dataset) were unique 0.84 H. gisella 2 - 6 extensive distribution from south eastern USA LEP 18150 1 LEP 04311 LEP 18153 2 0.72 LEP 04346 to northern Argentina . LEP 18116 • The resulting phylogeny is highly structured with moderate 1 LEP 18131 0.98 LEP 18147 LEP 18133 LEP 18120 support (Fig. 3) LEP 17667 1 1 H. cucullina 1 - 3 Despite most taxa being relatively common, species 1 H. gisella 7 LEP 18473 LEP 17661 LEP 17685 H. gisella 8 delimitation for the group is particularly challenging for 1 LEP 18107 There are currently 10 recognized species 0.88 LEP 18106 H. sosybius 1 2 0.65 1 H. hermybius 1 8&3 several reasons : 0.96 1 H. hermybius 2 in the genus Hermeuptychia 0.93

H. sosybius 2 - 10 0.96 1 • These are small with very similar external H. pimpla 1 • For this dataset, bPTP identified 68 ‘evolutionary significant 1 H. pimpla 2 1 LEP 14833 0.71 LEP 14842 LEP 11189 morphologies (few informative characters, Fig. 1) LEP 15311 units’ or ‘genetic species’ whereas GMYC identified 45 LEP 14843 1 H. pimpla 3 0.99 LEP 14838 0.97 H. pimpla 4 1 LEP 04299 (Fig. 3; 24 were recovered by both methods) LEP 14832 H. harmonia 1 1 1 LEP 18251 LEP 18464 LEP 18249 1 H. harmonia 2 0.96 H. harmonia 3 1 LEP 18236 • Only six of these potential species are already associated LEP 18422 LEP 18234 LEP 18233 LEP 18154 H. harmonia 4 with species names (i.e. these are clustered with previously LEP 11191 1 H. harmonia 5 1 LEP 18142 2 LEP 18127 H. harmonia 6 published sequences ). 0.9 LEP 11195 LEP 18138 H. harmonia 7 H. harmonia 8 LEP 15307 LEP 04341 H. hermes 1 • The others may well represent taxa that are new to science! 1 LEP 14865 LEP 10383 1 LEP 18115 1 H. hermes 2 LEP 18112 H. hermes 3 0.86 LEP 18104 0.74 LEP 17697 LEP 18113 Data suggests that diversity of Hermeuptychia is 0.98 H. hermes 4 - 8 severely underestimated 0.8 1 LEP 18486 LEP 17687 LEP 17691 LEP 17699 H.#sosybius# H.#intricata# H.#hermybius# LEP 17689 Figure 1. Left: A magnified image of prominent eyespots (yellow rim, black-center pupiled with 0.77 H. hermes 9 - 14 LEP 18243 pale-blue scales). Right: Images for Hermeuptychia taxa found in USA. Dorsal wings are brown LEP 04325 0.98 0.7 H. maimoune 1 1 LEP 17690 0.77 1 LEP 17658 without obvious markings. Ventral wings are a paler brown with a series of ocelli along wing margin. H. maimoune 2 What’s Next? H. maimoune 3 LEP 17692 LEP 04339 1 LEP 04342 H. maimoune 4 H. maimoune 5 • High intra-specific variation in wing ocelli patterns 0.99 LEP 18152 LEP 04386 1 H. maimoune 6 1. There is urgent need to verify species boundaries as indicated H. maimoune 7 1 LEP 14800 0.92 0.94 0.72 LEP 14798 by these model-based methods via the identification of • Many type specimens are old, damaged or missing 0.75 H. maimoune 8 - 10 0.94 LEP 14855 0.65 LEP 04330 LEP 18144 1 LEP 14849 corroborating differences in other data types, such as: LEP 04326 LEP 18149 LEP 18146 • Poor taxonomic knowledge has led to default assignment of 0.94 LEP 18143 1 LEP 18412 1 LEP 17659 0.73 LEP 04305 A. C. LEP 18088 Considering Testing for many specimens to type species, Hermeuptychia hermes LEP 04310 0.71 LEP 18087 1 LEP 18086 0.93 LEP 18158 needs verification! LEP 04290 distribution reproductive 1 LEP 18156 0.91 1 0.98 0.97 compatibility 0.99 Therefore, our eventual research aims are: 1 H. atalanta 1 - 13 B. 1 Identifying 0.99 morphological D. 1 LEP 18122 Characterising 1. To determine species boundaries in the genus Hermeuptychia 1 LEP 18151 1 LEP 18134 LEP 18126 0.94 LEP 18129 differences life history traits

2. To better understand processes that drive diversification in 1 Hermeuptychia butterflies Previously published sequences H. atalanta 14 - 25 2. Additionally, type specimens should be

LEP 18128 Sequences generated in this study H. atalanta voucher 26 - 28 barcoded, whenever possible, for the Figure 4. Hermeuptychia sosybius 3. To develop a framework for addressing similar issues in the LEP 04319 LEP 18139 clarification & proper assignment of (Carolina satyr) pair in copulo. 0.88 subtribe Euptychiina (where extensive taxonomic revisions species names are in progress) H. atalanta 29 - 43 3. Genome-wide variation will be sampled using double digest 0.01 restriction-site associated DNA sequencing (ddRADseq; a Figure 3. Phylogenetic tree based on Bayesian inference analysis (posterior probability values > 0.6 shown). Species boundaries indicated by GMYC (black & white) & bPTP (green & white) reduced-representation NGS method) for two purposes: Objectives are illustrated as colored bars on the side. Arrows indicate groups recovered by both methods. • To characterize intra-specific gene flow & genetic structuring Work presented here is a vital first step towards our broader (e.g. evidence of admixture, hybridization or introgression) goals. In order to identify distinct ‘genetic species groups’ for which Methods species hypotheses can be further refined by incorporating various • To test species limits inferred from DNA barcoding data types, COI variation is broadly surveyed using specimens Genomic DNA was extracted (using two legs or whole (a direct comparison using the same specimens) thorax) using Qiagen’s DNeasy Blood & Tissue Kit. collected from numerous localities in Ecuador (Fig. 2). BROADER IMPACTS Nymphalidae-specific primers4 were used for PCR. Even as NGS technologies become more affordable, barcoding Ecuador’s rich diversity, remains as the cheaper, simpler (bioinformatics analyses) & more extraordinary range of habitats & unique • Phylogenetic analysis: 250 previously published2&3 accessible (less stringent requirements for DNA quantity & quality) geographical landscape makes it an ideal sequences & 13 outgroup sequences were included. approach for rapidly & extensively assessing diversity. study site for: Identical haplotypes were removed before a Bayesian Knowing how COI variation ‘measures up’ to the more rigorous inference analysis was performed using BEAST 25 1. Testing if species diversity is approach is crucial for developing suitable strategies for resolving (partitioned by codon position, BEAST modeltest, underestimated in Hermeuptychia similar inquiries in related groups (barcoding data already available). duplicate runs, 1X108 generations, tree sampled every 2. Examining the extent to which References 5000 generations, 10% burn-in). [1] Forster, W. (1964) Zoolog. Staatssamml. 8: 51-188 [2] Seraphim, N., et al. (2014) Mol. Ecol. Resour. 14: 39–49 [3] Cong, geographic isolation drives speciation Q. & Grishin, N. V. (2014) ZooKeys 379: 43-91 [4] Nakahara, S., et al. (2015) Zootaxa 4012(3): 525-541 [5] Bouckaert, R., et • Species richness estimation: Two model-based al. (2014) PLoS Comput. Biol. 10(4): e1003537. [6] Zhang, J., et al. (2013) Bioinformatics 29(22): 2869-2876. [7] Fujisawa, T. & Barraclough, T. G. (2013) Syst. Biol. 62: 707-724. [8] Lamas, G. (2004) Atlas of Neotropical Lepidoptera 5: 261-274 In previous studies, samples were mainly approaches were utilized. The bPTP6 analysis was 2&3 Acknowledgements from Brazil, Colombia & USA . Therefore, 7 I would like to thank Anamaria Parus, Michelle Dunbar & Marianne Espeland for their involvement in this project. Special Figure 2. Map of Ecuador with red implemented via its web server & GMYC testing was thanks also to my committee members: Andrea Lucky, Akito Kawahara, Jacqueline Miller & Colette St Mary. Image credits: phylogenetic relationships are reassessed dots indicating collection localities. performed using the SPLITS package for R. Jonathan S. Bremer, Nick V. Grishin, Denise Tan & Keith R. Willmott. This research is supported by the McGuire Center for Lepidoptera & Biodiversity (Florida Museum of Natural History), Department of Entomology & Nematology (University of after broadening geographic sampling. Florida) and the National Science Foundation (DEB-1256742).