Supporting Information Renewed diversification is associated with new ecological opportunity in the Neotropical turtle ants Shauna L. Price, Scott Powell, Daniel J. C. Kronauer, Lucy A. P. Tran, Naomi E. Pierce, and R. K. Wayne Appendix S1: Materials and methods Molecular methods Molecular work was conducted at UCLA and Harvard. At UCLA PCR was performed in 25 µL volumes: 16.3 µL ddH2O, 2.5 µL 10x PCR Buffer, 1.5 µL 25 mM MgCl2, 0.5 µL dNTPs (25mM each), 1 µL DMSO, 0.2 µL of each primer (25x), 0.2 µL QIAGEN Taq DNA Polymerase, and 2 µL DNA. PCR cycles were: initial denaturation for 3 min at 94o, followed by 30 cycles of 94o for 30 s, 50-58o for 30 s, and 72o for 45 s, and a final extension of 72o for 5 min. Annealing temperatures depended on the gene segment amplified. PCR products were purified with Exonuclease I and Shrimp Alkaline Phosphatase. At Harvard PCR was performed in 25 µL volumes with the same PCR conditions as UCLA and a cocktail containing 15.05 µL ddH2O, 2.5 µL 10x PCR Buffer, 1 µL 25 mM MgCl2, 0.25 µL dNTPs (25mM each), 2 µL of each primer (10x), 0.2 µL QIAGEN Taq DNA Polymerase, and 2 µL DNA. PCR products were sent to Macrogen for purification. All loci were sequenced in both directions using an ABI 3730 automated sequencer with Big Dye Terminator chemistry (Applied Biosystems Inc) either at Macrogen or the Cornell University Life Sciences Core Laboratories Center. Heterozygous positions were left ambiguous and occurred in less than 0.0007% of nuclear base pairs. 1 Species monophyly We performed Bayesian inference (BI) using 2-4 individuals of 12 widespread turtle ant species in order to test species monophyly (Table S1). We used MrBayes v3.1.2 (Huelsenbeck & Ronquist, 2001; Ronquist & Huelsenbeck, 2003) on the molecular data set. Data were partitioned by nuclear gene segment and each mitochondrial codon position, and we applied a GTR+G model to each partition. We ran this analysis for 20 million generations, sampling every 1000 generations, with a heating parameter of 0.1 to increase mixing between chains. Morphological phylogeny The model-based framework of BI was used to compare results from a previously published morphological phylogeny that was constructed using maximum parsimony (de Andrade & Baroni-Urbani, 1999). We performed BI on the 131 character morphological data set using MrBayes. We ran two independent runs with four chains (one cold, three heated) for ten million generations, sampling every 1000 generations, and applied the Markov model (Lewis, 2001) with a gamma parameter. 2 Appendix S2: Results Species monophyly In analyses including multiple individuals per species, species monophyly was supported in all but two cases (Fig. S2). First, Cephalotes atratus is paraphyletic with respect to its putative sister species, C. marginatus. Second, C. bohlsi is paraphyletic with respect to an undescribed species in the fiebrigi clade. These results suggest that C. marginatus and the undescribed species may be geographic variants of C. atratus and C. bohlsi, respectively. Alternatively, C. atratus or C. bohlsi may actually be two or more species. More work is needed to discern between these possibilities. 3 Figure S1. Bayesian consensus phylogram of Cephalotes using the molecular data set, which includes multiple individuals for 12 species. PP values ≥ 0.95 are given. sp#1 1 P.#carbonarius 1 sp#2 1 P.#hylaeus C.#placidus 1 C.#opacus C.#opacus 1 C.#alfaroi 1 C.#serraticeps 1 C.#atratus C.#atratus 1 C.#atratus C.#marginatus C.#unimaculatus 1 C.#auricomus 1 C.#basalis 1 C.#basalis C.#cordiae 1 1 C.#ramiphilus C.#cordiventris 1 C.#mompox 1 0.99 C.#goniodontus C.#biguttatus 1 C.#multispinosus 1 C.#rowheri 1 C.#hirsutus C.#texanus C.#kukulcan 1 1 C.#scutulatus C.#clypeatus 1 C.#clypeatus 1 C.#clypeatus C.#umbraculatus 1 C.#umbraculatus 0.99 sp3 C.#adolphi 1 C.#targionii 0.98 C.#pilosus 0.95 C.#guayaki 1 C.#liogaster 1 C.#bruchi 0.96 C.#jheringi 0.99 C.#bohlsi# 1 C.#bohlsi sp1 C.#simillimus 1 C.#minutus 1 1 C.#minutus 1 C.#minutus 1 C.#christopherseni C.#laminatus 1 1 C.#spinosus 1 C.#spinosus C.#pusillus 1 C.#pusillus 1 C.#pusillus C.#pellans C.#pallidus 1 1 C.#pallidoides 0.97 C.#pallens C.#varians 1 1 C.#porrasi 1 C.#patellaris sp#2 C.#maculatus C.#maculatus 1 1 1 C.#maculatus 1 C.#grandinosus 1 C.#grandinosus 1 C.#persimplex 1 1 C.#persimilis C.#crenaticeps C.#peruviensis 1 1 C.#setulifer C.#trichophorus 0.99 C.#cristatus 0.98 C.#cordatus 1 C.#cordatus C.#borgmeieri 0.99 C.#betoi 1 C.#eduarduli C.#depressus 1 4 Figure S2. Bayesian consensus phylogram of extant and extinct species (indicated by †) based on molecular and morphological data. This combined analysis was used to determine fossil calibration points for divergence dating analysis. Though many nodes were not well supported, we defined calibration points where extinct species fell within well supported species groups in molecular only analyses. Species used in the calibration are in bold. P. sp1 0.96 P. hylaeus 0.87 P. carbonarius 0.92 P. sp2 C. opacus 1 C. placidus C. alfaroi 1 1 C. serraticeps 1 C. marginatus 1 C. atratus C. unimaculatus 1 C. auricomus C. clypeatus C. crenaticeps 0.79 †C. dieteri 0.79 †C. integerrimus †C. jansei 0.84 †C. serratus C. setulifer †C. bloosi 0.79 C. peruviensis 0.75 0.67 C. trichophorus †C. alveolatus †C. ventriosus 0.56 †C. obscurus 0.88 †C. sucinus C. cristatus C. borgmeieri 0.78 C. cordatus 0.93 0.69 C. betoi 0.8 C. depressus 0.78 0.8 C. eduarduli C. minutus 1 C. simillimus 1 C. pusillus C. christopherseni 1 C. laminatus 0.58 1 C. spinosus C. pellans C. pallidoides 1 1 C. pallidus 0.95 C. pallens C. varians 1 1 C. porrasi 1 C. patellaris C. sp2 †C. maya 0.69 †C. olmecus C. maculatus 0.55 C. grandinosus 0.95 C. persimilis 1 C. persimplex C. umbraculatus C. sp3 0.75 C. adolphi 0.85 C. targionii 0.92 C. pilosus 0.85 C. guayaki 1 C. liogaster 1 C. bruchi 0.97 1 C. jheringi 0.98 C. sp1 1 C. bohlsi †C. brevispineus C. basalis 0.92 C. ramiphilus 0.92 0.96 C. cordiae 0.87 C. mompox 1 C. cordiventris C. goniodontus 0.53 †C. hispaniolicus †C. poinari 0.82 †C. squamosus C. multispinosus 0.99 C. biguttatus 0.53 C. hirsutus 1 C. rowheri 0.58 C. texanus C. kukulcan 0.61 C. scutulatus 0.62 †C. caribicus 5 Figure S3. Primary concordance tree obtained from the Bayesian concordance analysis of the molecular data set. Concordance factors are shown at nodes. P.#hylaeus 0.76 sp1 0.91 sp2 0.78 P.#carbonarius C.#placidus 0.99 C.#opacus C.#marginatus 0.95 0.98 C.#atratus 0.67 C.#alfaroi 0.62 C.#serraticeps C.#unimaculatus 1 C.#auricomus 0.89 C.#basalis C.#ramiphilus 0.59 0.34 C.#cordiae 0.52 C.#mompox 0.85 C.#cordiventris 0.36 C.#goniodontus 0.17 C.#multispinosus 0.34 0.73 C.#biguttatus 0.5 C.#hirsutus 0.74 C.#rowheri C.#texanus 0.24 C.#kukulcan 0.5 0.15 0.47 C.#scutulatus C.#clypeatus C.#crenaticeps C.#peruviensis 0.43 C.#setulifer 0.34 0.24 C.#trichophorus 0.9 C.#minutus 0.27 C.#simillimus C.#christopherseni 0.87 C.#pusillus 0.19 0.55 C.#laminatus 0.27 0.74 C.#spinosus 0.26 C.#umbraculatus sp3 0.25 C.#adolphi 0.61 C.#targionii 0.39 C.#pilosus 0.28 0.11 C.#bruchi 0.48 sp1 0.48 C.#bohlsi 0.25 0.28 C.#liogaster C.#jheringi 0.03 0.27 C.#guayaki C.#cristatus C.#borgmeieri 0.71 0.26 C.#cordatus 0.33 C.#betoi C.#depressus 0.27 0.48 C.#eduarduli 0.24 sp2 C.#maculatus 0.83 C.#grandinosus 0.57 C.#persimilis 0.74 0.81 C.#persimplex 0.22 C.#pellans C.#pallidoides 0.71 0.65 C.#pallidus C.#porrasi 0.27 0.32 C.#patellaris 0.52 C.#varians 0.54 C.#pallens 6 Figure S4. Bayesian consensus phylogram of the 131 character morphological data set for Cephalotes. PP values ≥ 0.90 are given. P.#belli P.#elegans 1 P.#goeldii 0.94 P.#spiniperdus C.#oculatus C.#opacus 1 C.#placidus C.#alfaroi 0.95 C.#serraticeps C.#atratus 0.99 C.#marginatus C.#flavigaster 0.91 C.#taino C.#argentiventris 0.99 C.#auricomus C.#resinae 0.91 C.#hamulus 0.93 C.#unimaculatus 0.91 C.#vinosus 1 C.#duckei C.#solidus C.#ustus C.#clypeatus 1 C.#membranaceus C.#patei C.#emeryi C.#crenaticeps C.#ecuadorialis C.#columbicus 0.94 C.#pusillus C.#minutus 0.96 C.#simillimus C.#christopherseni C.#spinosus 0.92 0.96 C.#inaequalis 0.98 C.#laminatus C.#manni C.#cordiae C.#ramiphilus C.#complanatus 0.94 C.#cordiventris 0.98 0.93 C.#mompox 0.98 C.#femoralis C.#basalis 0.94 C.#inca C.#umbraculatus 1 C.#biguttatus C.#goniodontus 1 C.#multispinosus C.#cristatus C.#palustris C.#pavonii C.#betoi C.#borgmeieri C.#cordatus C.#depressus 0.93 C.#eduarduli C.#coffeae C.#trichophorus 0.92 0.94 C.#peruviensis C.#setulifer C.#bimaculatus C.#rohweri C.#hirsutus 1 C.#insularis C.#toltecus C.#wheeleri C.#texanus C.#curvistriatus C.#auriger 0.91 0.99 C.#chacmul 0.91 C.#lenca 0.91 C.#sobrius C.#conspersus C.#palta C.#targionii C.#angustus C.#frigidus 0.98 C.#adolphi 1 C.#dentidorsum C.#pallidicephalus 0.94 C.#goeldii 0.9 C.#notatus C.#bruchi C.#bivestitus C.#prodigiosus 1 C.#bohlsi C.#jheringi C.#fossithorax C.#supercilii C.#quadratus 0.91 C.#fiebrigi 0.98 C.#guayaki 0.97 C.#pitosus 0.96 C.#lanuginosus 0.96 C.#liogaster C.#varians C.#jamaicensis 1 C.#decolor 0.96 C.#decoloratus 0.92 C.#pellans 0.98 C.#patellaris C.#pallidoides C.#pallidus C.#pallens 0.95 C.#porrasi C.#incertus C.#kukulcan C.#scutulatus 0.97 C.#liepini C.#maculatus C.#nilpiei C.#pilieini C.#pinelii C.#foliaceus C.#grandinosus C.#klugi C.#persimilis C.#persimplex 7 Table S1.