Logan J. Mullen & Derek S. Sikes

A preliminary morphological and molecular phylogeny of the rove beetle genus Phlaeopterus ! (Coleoptera: Staphylinidae: Omaliinae: Anthophagini)

Logan J. Mullen & Derek S. Sikes University of Alaska Museum, University of Alaska Fairbanks, Fairbanks AK, 99775 email: [email protected] a b 2 mm

Introduction Figure 2. Morphology: 50% majority 0.5 mm

rule consensus phylogram of 1,500 post 56 Phlaeopterus_hatchi

The omaliine rove beetle genus Phlaeopterus Motschulsky 1853 contains 15 species, which are burn-in trees using the Mkv+G model in 93 Phlaeopterus_elongatus filicornis group MrBayes 3.2.0 with all 18 Phlaeopterus known from the northwestern United States, western provinces of Canada, Alaska, and Siberia. Phlaeopterus_filicornis species. Posterior probabilities are Phlaeopterus_cavicollis These beetles range in size from 3-10 mm in length, and seem to cluster in phylogenetically 95 given above branches. Colored Subgenus Phlaeopterus: all Phlaeopterus_bakerensis cavicollis group

informative large- and small-bodied groups. Although Phlaeopterus has never been analyzed with brackets indicate clades corresponding Phlaeopterus species greater Phlaeopterus_smetanai to species group hypotheses. All labeled than 5 mm in length modern phylogenetic methods, a taxonomic revision of the genus including 8 new species and the 96 Phlaeopterus_castaneus subgenera and species groups are castaneus group Phlaeopterus_kavanaughi transfer of two Phlaeopterus species to the genus Unamis was drafted, but never published, by sensu Campbell (unpublished). 98 Phlaeopterus_olympicus 84

J.M. Campbell in the 1980’s. Most recently a morphologically unique species from Siberia, Phlaeopterus_loganensis fusconiger group Lesteva czerskyi Shavrin 2001, was moved into Phlaeopterus by Shavrin & Mullen (2015). Phlaeopterus_occidentalis

All Phlaeopterus species Phlaeopterus_fusconiger

less than 5 mm in length Phlaeopterus_frosti The objectives of this study are to test these published and unpublished taxonomic 51 Figure 5. 99 Phlaeopterus_longipennis hypotheses of Phlaeopterus and develop a preliminary phylogeny of the genus using Subgenus Vellica a) Phlaeopterus n.sp. 74 Phlaeopterus_obsoletus bakerensis, the largest Bayesian analyses of morphological and molecular data. 92 Phlaeopterus_houkae Subgenus Vellicoides Phlaeopterus species: length = Phlaeopterus_lagrandeuri Subgenus Phlaeopteroides ~10 mm Phlaeopterus_czerskyi b) Phlaeopterus n. sp.

67 Unamis_truncata obsoletus, the smallest Research Questions 79 Unamis_stacesmithi Phlaeopterus species length = Unamis_kootenayensis ~3 mm. Phlaeopterus n.sp. bakerensis Phlaeopterus n.sp. obsoletus 1.) Do the small (<5 mm in length) and large (>5 mm in length) Phlaeopterus species Lesteva_cribratula

0.04 represent diagnosable monophyletic groups? Table 1. Testing previous taxonomic hypotheses of the genus Phlaeopterus using Bayesian morphological (Fig. 1), P.fus.WA1.40 P.fus.AK11.80 P.fus.AK8.58 Figure 3. COI: 50% majority rule P.fus.AK8.54 molecular (Fig. 2), and concatenated (Fig. 3) analyses. Posterior probabilities are given as percent values. g = genus, subg P.fus.AK2.5 P.fus.AK2.2 2.) Are the 8 new species proposed in Campbell’s unpublished revision supported? consensus phylogram of 1,500 post P.fus.AK3.10 P.fus.AK8.45 = subgenus, sg = informal species group, - = untested due to monotypy, and NA = untested due to lack of molecular data. P.fus.AK.Juneau.UAMIC1121.13 P.fus.AK.Juneau.UAMIC1120.13 burn-in trees using the GTR+I+G model P.fus.AK3.11 P.fus.AK3.19 P.fus.AK3.17 !"#$%&'()( *)%+%)$, -$(%'.)$./#.$0+0)1)%"/ -$(%'.)$./#.$0+0)1)%"/ -$(%'.)$./#.$0+0)1)%"/ in MrBayes 3.2.0 with 9 Phlaeopterus P.fus.AK3.16 3.) Is Campbell’s proposed transfer of Phlaeopterus kootenayensis and Phlaeopterus P.fus.AK3.15 P.fus.AK3.13 2$.#&$1$3" *45 2$.#&$1$3"/6/*45 P.fus.AK3.8 species. Posterior probabilities are P.fus.AK5.23 P.fus.AK2.3 65 P.fus.AK3.12 stacesmithi to the genus Unamis supported? P.fus.AK3.18 !"#$%&'(%)&*+%,!"#$%!&'()(*+,-./ 01',$233!&#),#$3-"425/ 6 6 6 given above branches. Colored P.fus.AK2.4 P.fus.AK8.49 P.fus.AK8.57 -%##*.&*+%,!"#$%!&'()(*+,-./ 01',$233!&#),#$3-"425/ 6 6 6 P.fus.AK3.7 brackets indicate clades corresponding P.fus.AK3.6 P.fus.AK3.9 P.fus.AK3.14 -%##*.$!"#$% 01',$233!&#),#$3-"425/ 77 89 89 4.) How does the recently described and morphologically unique Siberian species to subgenus and species group 69 P.fus.AK8.46 P.fus.AK8.48 P.fus.AK8.59 !"#$%&'(%)/,!"#$% 01',$233!&#),#$3-"425/ 7: ;<= >== P.fus.AK8.47 hypotheses from Figure 2. All labeled P.fus.AK8.55 Phlaeopterus czerskyi fit into the phylogeny of Phlaeopterus? P.fus.AK8.50 96 P.frosti.CNCCJ982.13 .$,($0%/,!%?(#,!"% 01',$233!&#),#$3-"425/ 7@ ;<= 77 subgenera and species groups are P.fusconiger.CNCCJ979.13 57 P.frosti.CNCCJ981.13 P.fusconiger.CNCCJ978.13 .$1*.&##*,!%?(#,!"% 01',$233!&#),#$3-"425/ 7< 89 89 P.fusconiger.CNCCJ976.13 sensu Campbell (unpublished). P.fusconiger.CNCCJ977.13 P.frosti.CNCCJ980.13 P.cav.WA1.41 2/,.&0*3%)!%?(#,!"% 01',$233!&#),#$3-"425/ ;<= ;<= ;<= P.cav.WA1.42 P.cavicollis.CNCCJ970.13 P.cav.AK4.27 2*#*.&)0*,!"% 01',$233!&#),#$3-"425/ 7A 6 6 P.cav.AK4.22 P.cav.AK6.24 P.cav.CA2.39 P.cav.AK4.20 !"#$%&'(%)/,!%!&-).3#5-)%!.B2?"C+-/! D41E?-)!F!G#332)!&H=>5 mm in length) Phlaeopterus 98 P.loganensis.CNCCJ975.13 P.loganensis.CNCCJ974.13 100 P.loganensis.CNCCJ973.13 P.loganensis.CNCCJ972.13 species is supported in morphological and concatenated analyses. Relationships of small and 100 P.cas.AK1.1 P.cas.AK.Haines.UAMIC1108.13 100 P.cast.BC1.32 P.kav.CA.1.34 large Phlaeopterus species, P.elo.AK10.79 P.elo.AK9.62 P.elo.AK10.75 2.) One of the 2 proposed new species of Campbell (unpublished) for which molecular data are and the genera Unamis and P.elo.AK10.81 99 P.elo.AK9.61 P.elo.AK10.78 P.elo.AK9.60 Lesteva unresolved 57 P.elo.AK10.71 available, Phlaeopterus n. sp. elongatus, is supported by the mtDNA and concatenated P.elo.AK10.70 P.elo.AK9.66 100 P.elo.AK10.72 P.lagrandeuri.CNCCJ957.13 phylogeny. A second, Phlaeopterus n. sp. kavanaughi, is represented by a single specimen that Lesteva.longoelytrata.KM442270.1 100 Lesteva.longoelytrata.KM440210.1 100 100 Lesteva.monticola.KJ964702.1 99 100 Lesteva.monticola.KJ964422.1 Lesteva.pubescens.KM445870.1 shows a large enough genetic distance to its nearest relatives to support its species status. Lesteva.pallipes.CNCCJ949.13 P.lag.UAMIC1845.14 P.lagrandeuri.CNCCJ956.13 100 P.lag.UAMIC633.13 Subgenus Phlaeopteroides P.lagrandeuri.CNCCJ955.13 P.lag.BC1.38 3.) The transfer of Phlaeopterus kootenayensis to the genus Unamis of Campbell (unpublished) is P.lag.UAMIC632.13 97 Unamis.kootenayensis.CNCCJ952.13 Unamis.kootenayensis.CNCCJ953.13 Unamis.kootenayensis.CNCCJ951.13 P.houkae.CNCCJ962.13 supported in the morphological and concatenated analyses. The transfer of Phlaeopterus 100 P.houkae.CNCCJ960.13 Subgenus Vellicoides 100 P.houkae.CNCCJ959.13 L.nsp.RU1.43 Arpedium.quadrum.KJ964193.1 Acidota.crenata.KJ962832.1 stacesmithi to the genus Unamis is supported in the morphological analysis. Figure 1. Distribution of the genus Phlaeopterus generated using all georeferenced records for the genus 0.02 4.) The Siberian species Phlaeopterus czserskyi, recently transferred to Phlaeopterus, appears to in the Arctos database and visualized with P.fus.AK2.2 represent a basal lineage. BerkeleyMapper. Red circles represent error. Note the P.fus.AK3.10 P.fus.AK.Juneau.UAMIC1121.13 P.fus.AK.Juneau.UAMIC1120.13 single record for Phlaeopterus czerskyi in Siberia. Figure 4. Morphology + COI: 50% P.fus.AK3.11 P.fus.AK3.19 P.fus.AK3.17 majority rule consensus phylogram of P.fus.AK3.16 P.fus.AK3.15 P.fus.AK3.13 Discussion 1,500 post burn-in trees of a P.fus.AK3.8 P.fus.AK3.12 P.fus.AK3.18 concatenated analysis using the Mkv+G P.fus.WA1.40 P.fus.AK11.80 The COI only phylogeny (Fig. 2) is less well resolved than the morphology or concatenated model for the morphology partition and P.fus.AK8.58 P.fus.AK8.54 P.fus.AK2.5 Methods the GTR+I+G model for the COI P.frosti.CNCCJ982.13 phylogenies, and has a 6-branch polytomy at the base of the Phlaeopterus clade. This could be P.frosti.CNCCJ981.13 P.fus.AK8.45 partition in MrBayes 3.2.0 with 9 P.fus.AK5.23 P.fus.AK2.3 due to saturation, which often confounds mtDNA analysis at deep branches (Xia et al. 2003). The P.fus.AK2.4 Morphological data Phlaeopterus species. Posterior P.fus.AK8.49 58 P.fus.AK8.57 superior resolution of the concatenated phylogeny highlights the value of integrated taxonomy P.fus.AK3.7 probabilities are given above branches. P.fus.AK3.6 • 50 morphological characters were mined from Hatch (1957), Moore & Legner (1979), Campbell P.fus.AK3.9 P.fus.AK3.14 over single data type approaches. Colored brackets indicate clades P.fus.AK8.46 (unpublished), and Newton et al. (2000), and coded using MacClade 4.04 (Maddison & Maddison 2000) 59 P.fus.AK8.48 corresponding to species group P.fus.AK8.59 P.fus.AK8.47 P.fus.AK8.55 Molecular data are currently available for only 10 of the 17 potential Phlaeopterus species, and P.fus.AK8.50 hypotheses as labeled in Figure 2. All 66 P.fusconiger.CNCCJ979.13 Molecular data 97 P.fusconiger.CNCCJ978.13 P.fusconiger.CNCCJ976.13 only a single locus, COI, has been sequenced so far. In order to increase the resolution of our labeled subgenera and species groups P.fusconiger.CNCCJ977.13 • A 658 bp region of the mtDNA gene COI was sequenced for 66 specimens representing 10 species P.frosti.CNCCJ980.13 are sensu Campbell (unpublished). P.cav.WA1.41 P.cav.WA1.42 analyses, we plan to sequence COI as well the nuclear locus CAD for all 17 potential species. P.cavicollis.CNCCJ970.13 collected by the first author, donated by collaborators, or belonging to University of Alaska Museum P.cav.AK4.27 P.cav.AK4.22 P.cav.AK6.24 P.cav.CA2.39 Insect Collection P.cav.AK4.20 P.cav.AK4.28 100 P.cav.AK4.30 • Sequences were translated to amino acids, checked for stop codons and frameshifts, and aligned by eye P.cav.AK4.21 Literature Cited P.cav.AK4.29 P.cav.AK4.82 Subgenus Phlaeopterus: all P.cav.AK7.26 using MacClade 4.04 and a reference sequence (Lunt et al. 1996) P.cav.AK7.25 Casey, T.L. 1894. Ann. N.Y. Acad. Sci. 7: 281-606. Hatch, M.H. 1957. University of Washington Publications P.cavicollis.CNCCJ969.13 large species, greater than 5 100 P.cavicollis.CNCCJ968.13 P.cav.CA1.36 in Biology 16: 1-384. Kass, R.E. Raftery, A.E. 1995. J. Amer. Statist. Assoc. 90: 773-795. Lewis, P.O. 2001. • 38 additional sequences were downloaded from GenBank and BOLD P.cav.CA1.33 mm in length P.cav.CA1.35 P.cav.CA1.31 Syst. Biol. 50(6): 913-925. Lunt, D.H. Zhang, D.X. Szymura J.M. & G.M. Hewitt. 1996 Insect Molecular P.loganensis.CNCCJ975.13 P.loganensis.CNCCJ974.13 Model selection P.loganensis.CNCCJ973.13 73 P.loganensis.CNCCJ972.13 Biology 5(3):153-165. Maddison D.R. & W.P. Maddison. 2000. Sinauer Associates, Sunderland 97 P.log.BBCCN358.10 P.castaneus.CNCCJ965.13 • Morphology: Bayes Factors (Kass & Raftery 1995) were used to determine that gamma correction for 99 P.castaneus.CNCCJ967.13 Massachusetts. Motschulsky, V. 1853. Etudes Entomologiques 1: 77-80. Rambaut, A. Suchard, M.A. Xie, P.castaneus.CNCCJ966.13 100 P.castaneus.CNCCJ964.13 P.cas.BBCCN675.10 D. & A.J. Drummond 2014. http://beast.bio.ed.ac.uk/Tracer. Ronquist, F., Huelsenbeck, J.P., & F. among character rate heterogeneity was preferred with the Mkv model (Lewis 2001) 100 100 P.cas.AK1.1 P.cas.AK.Haines.UAMIC1108.13 P.cast.BC1.32 Ronquist. 2001. Bioinformatics 17: 754-755. Shavrin, A.V. 2001. Zoosystematica Rossica. 9:189-193. • COI: Automated model selection in PAUP*4.0a136 (Swofford 2003) chose the general time reversal + P.kav.CA.1.34 P.elo.AK10.79 P.elo.AK9.62 All Phlaeopterus species P.elo.AK10.75 Shavrin, A.V & L.J. Mullen 2015. Zootaxa 1: 121-128. Swofford, D.L. 2003. Sinauer Associatiates, proportion of invariable sites + gamma distribution (GTR+I+G) model under all criteria (AIC, AICc, BIC) 100 P.elo.AK10.81 100 P.elo.AK9.61 less than 5 mm in length P.elo.AK10.78 Sunderland Massachusetts. Xia, X. et al. 2003. Mol. Phylogenet. Evol. 26: 1-7. P.elo.AK9.60 P.elo.AK10.71 Tree searching P.elo.AK10.70 P.elo.AK9.66 P.elo.AK10.72 P.lag.UAMIC1845.14 • Morphology, COI, and concatenated (morphology + COI) data sets were analyzed using MrBayes 3.2.0 100 100 P.lagrandeuri.CNCCJ956.13 P.lag.UAMIC633.13 Acknowledgments 100 P.lagrandeuri.CNCCJ955.13 Subgenus Phlaeopteroides (Ronquist et al. 2001), each retaining their respective best-fitting models P.lag.BC1.38 92 P.lag.UAMIC632.13 We thank our collaborators who have collected and donated specimens to this study: Dave Kavanaugh, P.lagrandeuri.CNCCJ957.13 P.houkae.CNCCJ962.13 • MCMCMC chains were run for 2,000,000 steps sampled every 1,000 steps 100 P.houkae.CNCCJ960.13 P.houkae.CNCCJ959.13 Subgenus Vellicoides Vladimir Gusarov, Link Olson, Jessica Rykken, and Jim LeBonte. We thank Steffi Ickert-Bond for hand- 100 Unamis.kootenayensis.CNCCJ952.13 Unamis.kootenayensis.CNCCJ953.13 • First 25% of samples were discarded as burn-in for each run, resulting in 1,500 trees sampled Unamis.kootenayensis.CNCCJ951.13 carrying valuable type specimens from the NMNH. Particular thanks are due to Megan McHugh and Adam Lesteva.KM442270.1 100 Lesteva.KM440210.1 • Tracer v1.6 (Rambaut et al 2014) was used to determine that runs had reached stationary (ESS values 100 Lesteva.KJ964702.1 Haberski for their hard work in the UAM Entomology Lab. Funding was provided by the University of Alaska 100 Lesteva.KJ964422.1 Lesteva.KM445870.1 100 Lesteva.CNCCJ949.13 Fairbanks, University of Oslo, Arctic Audubon Society, and the Society of Systematic Biologists. We are >1000 for all parameters) Lesteva.RU1.43 grateful to the 37 institutions that have loaned us specimens, and are especially grateful to the Canadian 0.02 National Collection of Insects, Arachnids and Nematodes for extensive specimen loans and assistance. Finally, this study relies heavily on the unpublished taxonomic work of J.M. Campbell.