Phylogeny and biogeography of the dung beetle genus ...

Systematic Entomology (2009), 34, 137–150 DOI: 10.1111/j.1365-3113.2008.00443.x

Phylogeny and biogeography of the dung beetle genusPhanaeus (Coleoptera: Scarabaeidae)

DANA L . P R I C EDepartment of Ecology, Evolution and Natural Resources, Rutgers University, New Brunswick, New Jersey, U.S.A.

Abstract. The dung beetle genus Phanaeus as currently recognized by Edmonds(1994) consists of 51 species placed in 13 species groups and two subgenera. Here, Iexamine the phylogeny and biogeography of this genus by analysing themitochondrial cytochrome oxidase subunit I (530 bp), nuclear large subunitribosomal RNA (28S, D2 region), and 67 morphological characters for 28 speciesof Phanaeus. Both maximum parsimony and Bayesian analyses from the combineddata yielded well-resolved trees, although low bootstrap and posterior probabilitysupport were found for basal nodes. The phylogenetic hypotheses presented heresuggest that the subgenera Phanaeus s.str. and Notiophanaeus should each beelevated to the status of full genus. With the exception of the eucraniine outgroups,the paleano species group of the genus Phanaeus is recovered as sister to all othertaxa, including the outgroups Oxysternon, Sulcophanaeus and Coprophanaeus.High bootstrap values and posterior probabilities supported the species groupsendymion, tridens and vindex. Biogeographical analyses suggest an ancestraldistribution for Phanaeus in the Andes in South America, although numerousdispersal events evidently have produced a complicated biogeographical history.

Introduction

The genus Phanaeus Macleay, 1819 comprises a group oftunnelling (as opposed to rolling) dung beetles distributed in

the Neotropical and Nearctic regions. Most of these beetlesare preferentially coprophagous, feeding on large herbivoreand omnivore excrement. During the breeding season, themale and female may cooperate in provisioning a nesting

gallery, and the female subsequently uses the stored food inconstructing brood balls (Edmonds, 1994). Phanaeus are ofbiological interest for their ecological importance and their

behaviours (male–male competition, bisexual cooperation andnidification, etc.) and regarding the evolution of their metalliccolours and variably developed cephalic and pronotal horns.

Much of the previous taxonomy of Phanaeus has beenbased on secondary sexual features, especially the shape ofthe cephalic horn and pronotum of the males, although theoften extreme intraspecific variation in shape has caused

some difficulties (Edmonds, 1994; Price, 2007). Based on

morphological and biogeographical characters, Edmonds

(1994) split Phanaeus into two subgenera: (i)Notiophanaeus,comprising 15 mostly South American species in five speciesgroups; and (ii) Phanaeus s.str., including 27 species (þ four

subspecies) in eight species groups, mostly in CentralAmerica and Mexico. Since 1994, nine new species havebeen described (Delgado-Castillo, 1991 [not reported inEdmonds’s 1994 revision]; Arnaud, 1997, 2000, 2001;

Kohlmann & Solis, 2001; Edmonds, 2004, 2006). In a phy-logenetic analysis of the genus based on 67 morphologicaland one biogeographical character, and including 49 of the

51 Phanaeus species, Price (2007) recovered a monophyleticPhanaeus if the genus Oxysternon is included within it.Phanaeus dung beetles are restricted to the New World,

from northern Argentina throughout much of South andCentral America, Mexico and the United States, northwardsand eastwards to Massachusetts. Most tropical Phanaeusspecies are mostly stenotopic (Edmonds, 1994), with distri-

butions determined by ground cover and prevailing climate,especially the amount and timing of warm-season rains.Geographic distributions can be extremely patchy in regions

such as Mexico, where physiography and habitats are highlyfragmented. Temperate zone species are far fewer and aredistributed across a wider range of habitat and area. The

Correspondence: DanaL.Price,DepartmentofBiological Sciences,

Salisbury University, 1101 Camden Avenue, Salisbury, MD 21801,

U.S.A. E-mail: [email protected]

# 2008 The AuthorJournal compilation # 2008 The Royal Entomological Society 137

SystematicEntomology

present distributions of the more stenotopic, tropical speciesundoubtedly have been strongly affected by human activi-

ties (Bennett, 1969; Edmonds, 1994).Edmonds (1994) hypothesized a South American origin for

Phanaeus as a whole;Notiophanaeus species were presumed to

be descendants of ancestral Phanaeus and to represent anearly radiation in South America, with the endymion grouphaving recently entered Central America, as have otherPhanaeini genera, Sulcophanaeus and Coprophanaeus, and

other scarabaeines (Edmonds, 1972; Kohlmann & Halffter,1988, 1990). Phanaeus s.str. was thought to have originated inMesoamerica (the region from northwest SouthAmerica west

and north of the Andes to the extreme southwestern U.S.)and to consist of three or perhaps four evolutionary lines(Edmonds, 1994): (i) a lineage including the hermes, triangu-

laris and tridens groups that expanded both northwards andsouthwards, ultimately entering northwest South Americaduring the Pleistocene; (ii) a lineage comprising the amethys-tinus and quadridens groups that diversified in Mesoamerica

montane habitats; (iii) the beltianus and mexicanus groups,which diversified extensively throughout Mesoamerica inwarmer, open habitats (i.e. intermediate between montane

and forest zones); and (iv) the vindex group, which mayrepresent a fourth lineage that invaded north temperateNorth America.

The purpose of this study is to analyse the phylogeneticrelationships within Phanaeus using the mitochondrialcytochrome oxidase subunit I (COI), 28S rRNA (D2

region) and morphological characters. The results ofa previous morphological study (Price, 2007) indicatedthat the examination of hypothesized sister group relation-ships is needed as well. This study presents the first

phylogenetic hypotheses for Phanaeus species using molec-ular data. Relationships with other Phanaeini generaare also discussed. The historical biogeography of Pha-

naeus will also be analysed and discussed using the phy-logenetic trees based on combined molecular andmorphological data.

Materials and methods

Taxon sampling

Twenty-eight species of Phanaeuswere sampled (Table 1),

including at least one species from each of the 13 speciesgroups of Edmonds (1994). When possible, several speci-mens of the same species were examined in order to account

for intraspecific variation. Outgroup representatives werechosen according to a recent phylogenetic analysis of thetribe Phanaeini (Philips et al., 2004a). Sequences for Anomi-

opsoides heteroclyta (Blanchard, 1845) and Glyphoderussterquilinus (Westwood, 1837), both from the Eucraniini,thought to be the sister to Phanaeini (Philips et al., 2004;Ocampo & Hawks, 2006), were generously supplied by

Federico Ocampo (University of Nebraska State Museum).Included taxa, voucher location and collection data areprovided in Table 1.

Morphology

For this study, 67 morphological characters from Price(2007) were used, including characters of the antennae, legs,labrum, pygidium and genitalia, as well as re-evaluated

characters from Edmonds (1994). Emphasis was placed onlarge male secondary sexual characters. For the purposes ofthis paper, morphological data are used only in combinedanalyses with molecular data.

DNA extraction

Except for 30 dry pinned museum specimens, all speci-mens were stored at �208C in 95% ethanol. DNA was

extracted from a leg or the thoracic muscles using a DNeasytissue kit (Qiagen). Tissues were incubated at 558C for 24 to72 h in 180 mL of ATL buffer and 20–40 mL of Proteinase

K. After incubation, standard protocols following theDNeasy Tissue Handbook (Qiagen) for DNA extractionof animal tissue were used.

Polymerase chain reaction amplification and DNA sequencing

The primers used for polymerase chain reaction (PCR)amplification are shown in Table 2. These primers were usedto amplify a 530-bp portion of mtDNA (COI) and a 625-bp

portion of 28S rRNA (D2 region). PCR was performed usinga GeneAmp PCR System 9700 (Applied Biosystems, FosterCity, CA) in 25 or 50 mL reaction volumes using either Taq

PCR Core Kits (Qiagen) or Taq PCR Master Mix Kits(Qiagen). PCR cycle conditions were as follows: 94–968C for3 min; 40–45 cycles of 948C for 30 s, 46–588C for 30–45 s,

728C for 30–60 s; and 728C for 10 min. Amplified PCRproduct was visualized on a 2% high-melting agarose gel(Fisher Biosciences). PCR products were purified using the

QIAquick PCR Purification Kit (Qiagen) and cycle-sequenced with the BigDye Terminator v3.1 Cycle SequencingKit (Applied Biosystems) on an ABI 3100 capillary sequencer.Each product was sequenced in both directions. The multiple

sequence alignment software Sequence Navigator (AppliedBiosystems) was used to compare and edit sequences.

Alignment

For a first approximation, sequences were aligned withCLUSTALX (Thompson et al., 1997). The CLUSTALX alignment

was sufficient for the alignment of COI, because there was nolength variation among sequences. The computer alignmentsof 28S rRNA sequences were manually adjusted according to

the criteria in Kjer (1995). The published secondary structureof the 28S rRNAwas used as a reference, using the secondarystructural model of Gillespie et al. (2004) for the 28S rRNAexpansion segment D2 from chrysomelid and related leaf

beetles. Insertions and deletions in regions for which thealignment was unambiguous, but which possessed shortindels (1–2 nucleotides long), were treated as a fifth character

138 D. L. Price

# 2008 The AuthorJournal compilation # 2008 The Royal Entomological Society, Systematic Entomology, 34, 137–150

Table 1. Species, specimen code, voucher location, year collected, country localities and GenBank accession number for specimens and

sequences used in the phylogenetic analysis. WDE, collection of W.D. Edmonds; CMN, collection of the Canadian Museum of Nature;

DLP, collection of Dana L. Price.

Species

Specimen

code

Voucher

location Collector

Year

collected Country

COI

GenBank D2 GenBank

Phanaeus achilles 34 WDE B. Streit 2003 Ecuador EU477300 EU432227

P. achilles 35 WDE B. Streit 2003 Ecuador EU477301 EU432228

P. achilles 180 DLP B. Streit 2006 Ecuador EU477302

P. achilles 181 DLP B. Streit 2006 Ecuador EU477303 EU432229

P. alvarengai 120 DLP S. Spector 1999 Bolivia EU477307 EU432232

P. a. amethystinus 261 CMN F. Genier 1993 Guatemala EU477339 EU432258

P. a. guatemalensis 260 CMN F. Genier 1993 Guatemala EU477340 EU432257

P. amithaon 177 DLP B. Streit 2004 USA EU477333 EU432251



P. bispinus 101 DLP S. Spector 1999 Bolivia EU477308 EU432233

P. chalcomelas 202 DLP T. Larsen 2004 Peru EU477298

P. chalcomelas 303 DLP D.L. Price 2005 French Guiana EU477299 EU432226

P. dejeani 255 CMN F. Genier 2000 Brazil EU477292 EU432223

P. endymion 15 WDE W.D. Edmonds 1997 Bolivia EU477315 EU432237

P. endymion 250 CMN B. Gill 1990 Mexico EU477316 EU432238

P. endymion 313 DLP C. Gillett 2006 Belize EU477317

P. furiosus 21 WDE W.D. Edmonds 1999 Mexico EU477324 EU432243

P. furiosus 274 DLP B. Streit 2005 Mexico EU477325 EU432244

P. furiosus 275 DLP B. Streit 2005 Mexico EU477326 EU432245

P. haroldi 179 DLP B. Streit 2005 Ecuador EU477293 EU432219

P. haroldi 185 DLP B. Streit 2006 Ecuador EU477294 EU432220

P. haroldi 254 CMN F. Genier 1994 Ecuador EU477295 EU432221

P. howdeni 257 CMN Gillogly-Stockwell 1996 Panama EU477337 EU432257

P. igneus 301 DLP K. Beucke 2005 USA EU477345 EU432262

P. igneus 302 DLP C. Marshall 2005 USA EU477346 EU432263

P. kirbyi 10 WDE S. Spector 1999 Bolivia EU477309 EU432234

P. kirbyi 102 DLP S. Spector 2000 Bolivia EU477102

P. kirbyi 117 DLP S. Spector 1999 Bolivia EU477117

P. lecourti 204 DLP T. Larsen 2004 Peru EU477306 EU432231

P. lunaris 182 DLP B. Streit 2006 Ecuador EU477329 EU432248



P. meleagris 205 DLP T. Larsen 2004 Peru EU477304

P. meleagris 256 CMN F. Genier 1999 Bolivia EU477305 EU432230

P. melibaeus 5 WDE S. Spector 1999 Bolivia EU477296 EU432224

P. melibaeus 6 WDE S. Spector 1999 Bolivia EU477297 EU432225

P. nimrod 18 WDE W.D. Edmonds & Reyes 2003 Mexico EU477322 EU432241

P. nimrod 19 WDE W.D. Edmonds & Reyes 2003 Mexico EU477323 EU432242

P. paleano 8 WDE E. Cororo 2001 Paraguay EU477312 EU432235

P. paleano 104 DLP S. Spector 1999 Bolivia EU477313

P. paleano 258 CMN F. Genier & F. Vaz de Mello 2001 Brazil EU477314 EU432236

P. prasinus 16 WDE 2000 Venezuela EU477320

P. prasinus 252 CMN S. Peck 1993 Trinidad EU477321 EU432240

P. pyrois 13 WDE Olaya & Mosquero 2003 Colombia EU477318

P. pyrois 251 CMN J.S. Ashe & A.K. Ashe 1993 Costa Rica EU477319 EU432239

P. quadridens 22 WDE W.E. Edmonds & Reyes 1996 Mexico EU477341

P. quadridens 271 DLP B. Streit & Cunningham 2005 Mexico EU477342 EU432261

P. quadridens 175 DLP B. Streit 2004 Mexico EU477343 EU432259

P. quadridens 178 DLP B. Streit 2004 USA EU477344 EU432260

P. sallei 311 DLP C. Gillett 2006 Belize EU477311 EU432256

P. splendidulus 253 CMN F. Genier 2000 Brazil EU477291 EU432222

P. triangularis texensis 32 WDE W.D. Edmonds 2002 USA EU477327 EU432246

P. triangularis texensis 33 WDE W.D. Edmonds 2002 USA EU477328 EU432247

P. vindex 200 DLP D.L. Price 2003 USA EU477347

Phanaeus phylogeny and biogeography 139


state. For two larger insertions and deletions, differentapproaches were used. In order to retain as much informationas possible, the first region of 14 nucleotides was aligned usinga guide tree created from the remaining nucleotides in

CLUSTALX. A second ambiguous region of 34 nucleotideswas coded as multistate characters, and the nucleotides wereexcluded from the analysis. All molecular data have been

submitted to GenBank under the accession numbersEU432219–EU432272 (D2) and EU477300–EU477363(COI) (see Table 1).

Phylogenetic analysis

Datasets were analysed individually and in combination(i.e. COI, D2, COI þ D2 and COI þ D2 þ morphology)

using parsimony and Bayesian analyses. All parsimony-

based analyses were performed with PAUP* 4.0b 10 (Swofford,1999). Trees were estimated using heuristic searches with10 000 random addition replicates and tree bisection–reconnection (TBR) branch swapping, and branches were

collapsed if the minimum length was zero. Branch supportin the resulting cladograms was assessed using bootstrapanalysis (Felsenstein, 1985). One thousand replicates were

implemented, with 10 random addition sequences perreplicate. For the COI data, homogeneity of third-codonbase frequencies across taxa was evaluated with a chi-square

test using PAUP*. In order to examine Edmonds’ (1994)hypothesis regarding the evolutionary lines based on theecogeographic regions of Phanaeus s.str. species groups,

a constraint analysis using parsimony was conducted inPAUP* as follows {Outgroups (Notiophanaeus) [(triangularisgr., tridens gr., hermes gr.)(quadridens gr., amethystinusgr.)(mexicanus gr., beltianus gr.)(vindex gr.)]}. Because of

uncertainty in the placement of the vindex species group,three additional analyses were conducted with vindex gr.constrained within each of the above lineages of the sub-

genus Phanaeus s.str.Prior to Bayesian analyses, MODELTEST 3.06 Akaike

weights (Posada & Crandall, 1998; Posada & Buckley,

2004) and DT-MODSEL (Minin et al., 2003) were used toselect an appropriate model of evolution for each of the twoindependent gene fragments. Both programs suggestedGTR þ I þ G for both COI and D2 (Yang, 1994; Yang

et al., 1994; Gu et al., 1995). Bayesian analyses wereperformed on both molecular datasets as well as fora combined analysis (D2 þ COI þ morphology) using the

program MRBAYES 3.1.2 (Hulsenbeck & Ronquist, 2001).Bayesian analyses used a mixed model with three partitions,namely COI, D2 and morphology. Bayesian analyses for

COI data were run twice for 5 million generations, with

Table 2. Primers used for polymerase chain reaction amplification

and sequencing of the cytochrome oxidase subunit I mtDNA gene

fragment and the 28S rRNA (D2) fragment.

Primer Sequence

C1-J-1718* GGAGGATTTGGAAATTGATTAGTTCC

C1-J-1751* GGATCACCTGATATAGCATTCCC

C1-N-2191* CCCGGTAAAATTAAAATATAAACTTC

C1-N-2329* ACTGTAAATATATGATGAGCTCA

D2-J-4 AGTCGTGTTGCTTGATAGTGCAG

D2-J-6TR GGTAAACTCCATCTAAGGCTAA

D2-N-B TTGGTCCGTGTTTCAAGACGG

D2-N-BJ CTTTGGTCCGTGTTTCAAGAC

J is for sense and N is for antisense strands.Primers marked with an asterisk were compiled from Simon et al. (1994).D2 primers were designed in Kjers laboratory at Rutgers University.

Table 1. Continued.

Species

Specimen

code

Voucher

location Collector

Year

collected Country

COI

GenBank D2 GenBank

P. vindex 270 DLP D.L. Price 2004 USA EU477348 EU432264

P. wagneri wagneri 176 DLP B. Streit 2004 Costa Rica EU477332

P. yecoraensis 276 DLP B. Streit 2005 Mexico EU477336 EU432254

Outgroups

Coprophanaeus ignecinctus 207 DLP T. Larsen 2004 Peru EU477353 EU432267

C. pluto 44 WDE W.D. Edmonds 1999 Mexico EU477349 EU432265

C. pluto 45 WDE W.D. Edmonds 1999 Mexico EU477350 EU432266

C. telamon 208 DLP T. Larsen 2004 Peru EU477351

C. telamon 312 DLP C. Gillett 2006 Belize EU477352

Oxysternon conspicillatum 40 WDE Olaya & Mosquero 2003 Colombia EU477359 EU432270

O. conspicillatum 211 DLP T. Larsen 2004 Peru EU477360

O. durantoni 304 DLP D.L. Price 2005 French Guiana EU477358 EU432273

O. festivum 277 DLP D.L. Price 2005 French Guiana EU477357 EU432272

O. silenus 41 WDE Olaya & Mosquero 2003 Colombia EU477361

O. silenus 42 WDE Olaya & Mosquero 2003 Colombia EU477362 EU432271

O. spiniferum spiniferum 262 CMN F. Genier 1999 Bolivia EU477363

Sulcophanaeus auricollis 46 WDE Olaya & Mosquero 2003 Colombia EU477354 EU432268

S. faunus 113 DLP S. Spector 1999 Bolivia EU477355

S. faunus 278 DLP D.L. Price 2005 French Guiana EU477356 EU432269

140 D. L. Price


a sampling frequency of 500 generations. Analyses ofcombined data were run twice for 7 million and 5 million

generations, respectively. The first 200 trees in each file werediscarded as ‘burn-in’. The remaining trees were pooled andused for estimating Bayesian posterior probabilities by

a majority-rule consensus procedure in PAUP. Bayesian treesfrom the combined data were also loaded into PAUP*, and alltrees supporting the monophyly of Phanaeus were filteredusing the ‘filter’ command under the ‘trees’ menu and the

following constraint [eucraniines, Coprophanaeus, Sulcopha-naeus, Oxysternon (Phanaeus)].

Biogeography

Distribution data were obtained from the literature(Edmonds, 1994, 2000; Arnaud, 2002; Edmonds & Zidek,2004; Ocampo, 2004, 2005) and from specimens used in themolecular analyses.

The distribution ranges of ancestral nodes were inferredusing the program DIVA (Ronquist, 1996, 1997), whichestimates ancestral distributions and differentiates puta-

tive dispersal and vicariance events. The data matrix wasconstructed by scoring the taxa for presence or absence in

12 areas (12 characters). Because DIVA requires a completelybifurcated tree, the tree with the highest likelihood scorerecovered from the combined Bayesian analysis was used

for these analyses. DIVA optimizations were then conductedeither with an unrestricted number of areas assigned toeach node or with areas per node restricted to two, three orfour.

The literature used to help with the establishment ofgeographic regions (Fig. 1) included Hooghiemstra et al.(2006), Metcalfe (2006) and Perret et al. (2006). The geo-

graphic areas defined for these analyses were: (A) portionsof the Guyana shield ranging from the Rio Orinocowestwards and from the Parque Nacional do Pico Neblina

eastwards; (B) Amazonia, including northern Colombia andVenezuela; (C) the caatingas of northeastern Brazil west-wards, the cerrado, and southwards including the camposrupestre (North Bahia southwards to Serra do Ouro Bran-

co), distributed east of the Sao Francisco river southwardsto the Parana river; (D) northern Argentina eastwards to theParana River; (E) the Andes; (F) west of the Andes from

Fig. 1. Biogeographical areas used to

delimit Phanaeus species distributions in

South and Central America, Mexico and

the United States.



southern Ecuador, northwards through Colombia and intoPanama; (G) central Panama northwards to Costa Rica,

northwards through Honduras, Nicaragua, Guatemala(except the northern Yucatan region), southern Belizeincluding the Maya Mountains and into Mexico (including

the Chiapas and the easternmost portion of Oaxaca includ-ing Tehuantepec); (H) northern Belize, northern Guatemala,Mexico (including Yucatan, Quintana Roo, Campeche,Tabasco, Veracruz, Tamaulipas and Nuevo Leon) and

southern Texas; (I) Mexico (southern Mexico, the SierraMadre Oriental and the Sierra Madre Occidental); (J)Mexico east of the Sierra Madre Occidental, including the

Sonoran desert of Arizona; (K) the western United States(encompassing the southern tip of Texas to an arbitrarylocation just east of Austin, northwards into eastern por-

tions of Oklahoma, Kansas and Nebraska); (L) the easternUnited States.

Results

Parsimony analyses

A 530-bp region of the mtDNA COI gene from 28ingroup taxa (58 individuals) and 12 outgroup taxa

(17 individuals) contained 170 parsimony-informative char-acters. Parsimony analyses of COI recovered 38 most-parsimonious trees of 1319 steps (CI ¼ 0.32, RI ¼ 0.63).

Nucleotide frequency at third codon positions showedconserved homogeneity of base composition across taxa(x2 ¼ 218.03, d.f. ¼ 222, P ¼ 0.56). Mean base composi-tion overall showed a bias towards A and T nucleotides

(A ¼ 30.4%; C ¼ 16.7%; G ¼ 15.3%; T ¼ 37.6%). Pha-naeus was recovered as monophyletic only if defined toinclude Oxysternon, Coprophanaeus and Sulcophanaeus

auricollis (not S. faunus). The strict consensus tree placedthe paleano group, splendidulus group and P. alvarengai (ofthe bispinus group) in a polytomy, sister to all other

Phanaeini except S. faunus (Fig. 2).A 625-bp region of the 28S rRNA (D2) from 27 ingroup

taxa (46 individuals; Phanaeus w. wagneri could not be ampli-

fied) and ten outgroup taxa (11 individuals) contained 100parsimony-informative characters. The resulting strict con-sensus tree (not shown) based on 68 987 most-parsimonioustrees (CI ¼ 0.76, RI ¼ 0.85) included a large polytomy with

only a few species groups recovered as monophyletic andgenerally with low bootstrap support. Phanaeus melibaeusplus P. haroldiwere recovered as sister to all other taxa except

the eucraniines. A clade including P. yecoraensis, P. quad-ridens, plus the tridens group, and also subclade P. lunaris þP. howdeni þ P. sallei þ all Coprophanaeus species were

recovered with bootstrap values of 4 and 34, respectively.Also grouped together were Oxysternon silenus and P.alvarengai with a bootstrap value of 56%, although onlyone character was without homoplasy.

Parsimony analysis of the combined molecular data (notshown) showed little resolution. Phanaeus s.str. was recov-ered in a terminal polytomy including chalcomelas, hermes

and endymion species groups. The splendidulus, paleano andbispinus groups and P. achilles, along with Oxysternon,

Coprophanaeus and S. auricollis, formed another polytomy,sister to Phanaeus s.str. The eucraniines and S. faunus wereoutside the latter assemblage.

The results from the parsimony analysis of the combinedmolecular and morphology data are shown in Fig. 2. Onemost-parsimonious tree of 1865 steps (CI ¼ 0.38, RI ¼0.68) was recovered. Three hundred and thirty-six charac-

ters were parsimony-informative. The resulting phylogenyrecovered the paleano group as sister to all other phanaeinetaxa. The strict consensus yielded a well-resolved tree,

although bootstrap values are low for most of the earlybranches. Because the support values are so low, therelationships among Phanaeus, Oxysternon, Sulcophanaeus

and Coprophanaeus are considered unresolved.Parsimony analyses constrained to recover the principal

clades in Edmonds’ (1994) hypotheses of Phanaeus s.str.evolution recovered four most-parsimonious trees of 1887

steps (not shown). The final strict consensus, with speciesgroups indicated, was as follows (mexicanus gr. þ beltianusgr. <amethystinus gr. þ quadridens gr. {vindex gr. [triangu-

laris gr. (hermes gr. þ tridens gr.)]}>. The chalcomelas andendymion species groups were recovered in a polytomy withNotiophanaeus (bispinus gr. and splendidulus gr.). Additional

analyses with vindex gr. constrained to each of the Phanaeuss.str. lineages resulted in trees of lengths equal to or greaterthan 1887.

Bayesian analyses

Bayesian analysis of COI data recovered P. melibaeus andP. haroldi together as sister to all other taxa except S. faunusand the eucraniines (Fig. 3). Species groups recovered as

monophyletic included the beltianus, endymion, paleano,tridens and vindex groups. Support for the relationshipsamong species groups is relatively weak, except for identi-

fication of the triangularis and quadridens groups as sisterclades (100% posterior probability). Phanaeus s.str. appearsas monophyletic (with the exception of P. achilles), although

the relationships within the subgenus are tenuous.The D2 Bayesian analysis recovered a tree similar to

that of the parsimony analysis. Phanaeus melibaeus plusP. haroldi were recovered as sister to all other taxa except

the eucraniines, and the majority of the nodes were sup-ported with low posterior probabilities. The clade includingP. yecoraensis, P. quadridens, plus the tridens group, and

also subclade P. lunaris þ P. howdeni þ P. sallei þ allCoprophanaeus species was again recovered with a 96%posterior probability. The members of the P. lunaris sub-

clade were grouped together with 97% posterior probabil-ity, based on three characters, although only one waswithout homoplasy. As in the parsimony analysis, Oxy-sternon silenus and P. alvarengai were grouped together with

a 59% posterior probability.Bayesian analysis of the combined molecular and mor-

phological data recovered a well-resolved tree (Fig. 3), with

142 D. L. Price


Gly

phod

erus

ste

rqui

linus

Ano

mio

psoi

des

hete

rocl

yta

Sul

coph

anae

us fa

unus

S. f

aunu

s

P. a

lvar

enga

i

P. s

plen

didu

lus

P. d

ejea

niP

. mel

ibae

usP

. mel

ibae

usP

. har

oldi

P. h

arol

diP

. har

oldi

Pha

naeu

s ki

rbyi

P. k

irbyi

P. k

irbyi

P. p

alea

noP

. pal

eano

P. p

alea

no

P. b

ispi

nus

Cop

roph

anae

us p

luto

C. p

luto

C. i

gnec

inct

usC

. tel

amon

C. t

elam

onS

. aur

icol

lis

O. s

ilenu

sO

. sile

nus

O. f

estiv

umO

. dur

anto

niO

. con

spic

illat

umO

. con

spic

illat

umP

. ach

illes

P. a

chill

esP

. ach

illes

P. a

chill

esP

. end

ymio

nP

. end

ymio

nP

. end

ymio

nP

. pyr

ois

P. p

yroi

s

P. h

owde

niP

. sal

lei

P. l

unar

isP

. lun

aris

P. l

unar

isP

. tria

ngul

aris

texe

nsis

P. t

. tex

ensi

sP

. qua

drid

ens

P. q

uadr

iden

sP

. qua

drid

ens

P. q

uadr

iden

sP

. yec

orae

nsis

P. w

. wag

neri

P. a

mith

aon

P. a

mith

aon

P. a

mith

aon

P. p

rasi

nus

P. p

rasi

nus

P. l

ecou

rti

P. m

elea

gris

P. m

elea

gris

P. n

imro

dP

. nim

rod

P. f

urio

sus

P. f

urio

sus

P. f

urio

sus

P. i

gneu

sP

. ign

eus

P. v

inde

xP

. vin

dex

P. a

. am

ethy

stin

usP

. a. g

uate

mal

ensi

s

100*

100

70

9965

80*

59

88*

99*

92*

55

100

*

72

55*

8287

6710

0

77

100

9558

71 100 70

99*

100

9858

9458

93

100*

63

85

9886

82

100

67 100

100

Gly

phod

erus

ste

rqui

linus

Ano

mio

psoi

des

hete

rocl

yta

Pha

naeu

s ki

rbyi

P. k

irbyi

P. k

irbyi

P. p

alea

noP

. pal

eano

P. p

alea

noC

opro

phan

aeus

plu

toC

. plu

toC

. ign

ecin

ctus

C. t

elam

onC

. tel

amon

S. a

uric

ollis

Sul

coph

anae

us fa

unus

S. f

aunu

sO

xyst

erno

n s.

spi

nife

rum

O. s

ilenu

sO

. sile

nus

O. f

estiv

umO

. dur

anto

niO

. con

spic

illat

umO

. con

spic

illat

umP

. ach

illes

P. a

chill

esP

. ach

illes

P. a

chill

esP

. bis

pinu

sP

. spl

endi

dulu

sP

. dej

eani

P. a

lvar

enga

i

P. m

elib

aeus

P. m

elib

aeus

P. h

arol

diP

. har

oldi

P. h

arol

di

P. p

rasi

nus

P. p

rasi

nus

P. c

halc

omel

asP

. cha

lcom

elas

P. l

ecou

rti

P. m

elea

gris

P. m

elea

gris

P. e

ndym

ion

P. e

ndym

ion

P. e

ndym

ion

P. p

yroi

sP

. pyr

ois

P. a

. am

ethy

stin

usP

. a. g

uate

mal

ensi

s

P. w

. wag

neri

P. a

mith

aon

P. a

mith

aon

P. a

mith

aon

P. h

owde

niP

. sal

lei

P. l

unar

isP

. lun

aris

P. l

unar

is

P. i

gneu

sP

. ign

eus

P. v

inde

xP

. vin

dex

P. y

ecor

aens

isP

. nim

rod

P. n

imro

dP

. fur

iosu

sP

. fur

iosu

sP

. fur

iosu

s

P. t

riang

ular

is te

xens

isP

. t. t

exen

sis

P. q

uadr

iden

sP

. qua

drid

ens

P. q

uadr

iden

sP

. qua

drid

ens

99

100

100

96

100

97100

100

94

77

6310

0100

100

9857

100 63

99

66

8391

49

71

8258

55

100 97

5810

010

0

67

976395

75

100

87

100

8566

8810

0

100

100

7370

100

P. c

halc

omel

asP

. cha

lcom

elas

95

AB

Pal

ean

o

Ch

alco

mel

as

Bis

pin

us

Sp

len

did

ulu

s

Her

mes Ch

alco

mel

as

En

dym

ion Tr

ian

gu

lari

s

Qu

adri

den

s

Trid

ens

Mex

ican

us

Vin

dex

Bel

tian

us A

met

hys

tin

us

Ou

tgro

up

s

Mex

ican

us

Oxy

ster

non

s. s

pini

feru

m

Ou

tgro

up

s

Pal

ean

o

Ch

alco

mel

as

Bis

pin

us

Sp

len

did

ulu

sB

isp

inu

s

Sp

len

did

ulu

s

Her

mes C

hal

com

elas

En

dym

ion Tr

ian

gu

lari

s

Qu

adri

den

s

Trid

ens

Vin

dex M

exic

anu

s

Bel

tian

us A

met

hys

tin

us

Mex

ican

us

Ou

tgro

up

sO

utg

rou

ps

Fig.2.(A

)PhylogenetichypothesisforPhanaeusresultingfrom

parsim

onyanalysisofcytochromeoxidase

subunitIdata.Strictconsensusof38most-parsim

onioustreesof1319

steps.Numbersabovebranches

representbootstrapvalues

>50%.Branches

withanasteriskindicate

congruence

withD2parsim

onyanalysis.(B)Phylogenetic

hypothesisfor

Phanaeusinferred

from

aparsim

onyanalysisofthecombined

molecularandmorphologicaldata.Numbersabovebranches

representbootstrapvalues

>50%.



Gly

phod

erus

ste

rqul

inus

Ano

mio

psoi

des

hete

rocl

yta

Sul

coph

anae

us fa

unus

S. f

aunu

sP

hana

eus

mel

ibae

usP

. mel

ibae

usP

. har

oldi

P. h

arol

diP

. har

oldi

P. a

lvar

enga

iP

. bis

pinu

sP

. spl

endi

dulu

sP

. dej

eani

P. a

chill

esP

. ach

illes

P. a

chill

esP

. ach

illes

S. a

uric

ollis

Oxy

ster

non

s. s

pini

feru

mO

. sile

nus

O. s

ilenu

sO

. fes

tivum

O. d

uran

toni

O. c

onsp

icill

atum

O. c

onsp

icill

atum

Cop

roph

anae

us p

luto

C. p

luto

C. i

gnec

inct

usC

. tel

amon

C. t

elam

on

P. k

irbyi

P. k

irbyi

P. k

irbyi

P. p

alea

noP

. pal

eano

P. p

alea

no

100*

100

100

99*

100*

96*

77

81

94*

100

7155

9088

100

8498 10

0

81*

66

5999

100*

100

83

93

P. p

yroi

sP

. pyr

ois

P. e

ndym

ion

P. e

ndym

ion

P. e

ndym

ion

9659

80

P. c

halc

omel

asP

. cha

lcom

elas

P. l

ecou

rti

P. m

elea

gris

P. m

elea

gris

5010

0

7999

P. p

rasi

nus

P. p

rasi

nus

100

85

P. a

. am

ethy

stin

usP

. a. g

uate

mal

ensi

s10

0

67

Gly

phod

erus

ste

rqui

linus

Ano

mio

psoi

des

hete

rocl

yta

Pha

naeu

s ki

rbyi

P. k

irbyi

P. k

irbyi

P. p

alea

noP

. pal

eano

P. p

alea

noO

xyst

erno

n fe

stiv

umO

. dur

anto

niO

.con

spic

illat

umO

. con

spic

illat

umP

. bis

pinu

sP

. spl

endi

dulu

sP

. dej

eani

P. a

lvar

enga

iP

. mel

ibae

usP

. mel

ibae

usP

. har

oldi

P. h

arol

diP

. har

oldi

Sul

coph

anae

us a

uric

ollis

S. f

aunu

sS

. fau

nus

O. s

. spi

nife

rum

O. s

ilenu

sO

. sile

nus

Cop

roph

anae

us ig

neci

nctu

sC

. plu

toC

. plu

toC

. tel

amon

C. t

elam

onP

. end

ymio

nP

. end

ymio

nP

. end

ymio

nP

. pyr

ois

P. p

yroi

s

P. p

rasi

nus

P. p

rasi

nus

P. i

gneu

sP

. ign

eus

P. v

inde

xP

. vin

dex

P. t

riang

ular

is te

xens

isP

. t. t

exen

sis

P. q

uadr

iden

sP

. qua

drid

ens

P. q

uadr

iden

sP

. qua

drid

ens

P. a

met

hyst

inus

am

ethy

stin

usP

. a. g

uate

mal

ensi

s

P. w

agne

ri w

agne

riP

. am

ithao

nP

. am

ithao

nP

. am

ithao

n

P. s

alle

iP

. how

deni

P. l

unar

isP

. lun

aris

P. l

unar

is

100

8010

086

8550

67

53

100

100

100

9080

99

9410

010

0

100

61

6752

99

100

100

69

100

6310

0

96

89

100

70

78

100

7768

69

100

100

100

100

100

57

100

9497

78

9172

9972

P. a

chill

esP

. ach

illes

P. a

chill

esP

. ach

illes

P. c

halc

omel

asP

. cha

lcom

elas

P. l

ecou

rti

P. m

elea

gris

P. m

elea

gris

98

100

63

5910

0

9310

0

P. s

alle

iP

. how

deni

57

P. i

gneu

sP

. ign

eus

P. v

inde

xP

. vin

dex

72 100

P. t

riang

ular

is te

xens

is

P. t

. tex

ensi

sP

. qua

drid

ens

P. q

uadr

iden

sP

. qua

drid

ens

P. q

uadr

iden

s

100*

50

100

100

P. l

unar

isP

. lun

aris

P. l

unar

is

9784

P. w

. wag

neri

P. a

mith

aon

P. a

mith

aon

P. a

mith

aon

9997

84

P. y

ecor

aens

is

P. n

imro

dP

. nim

rod

P. f

urio

sus

P. f

urio

sus

P. f

urio

sus

100*

9554

P. y

ecor

aens

is

P. n

imro

dP

. nim

rod

P. f

urio

sus

P. f

urio

sus

P. f

urio

sus

100

100

9864

AB

Ou

tgro

up

s

Pal

ean

o Ch

alco

mel

as

Bis

pin

us

Sp

len

did

ulu

s

Her

mes

En

dym

ion Tr

ian

gu

lari

s

Qu

adri

den

s

Trid

ens

Bel

tian

us A

met

hys

tin

us

Ou

tgro

up

s

Mex

ican

us

Sp

len

did

ulu

s

Vin

dex

Ou

tgro

up

s

Pal

ean

o

Bis

pin

us

Sp

len

did

ulu

sB

isp

inu

s

Sp

len

did

ulu

s

Her

mesCh

alco

mel

as

En

dym

ion Tr

ian

gu

lari

s

Qu

adri

den

s

Trid

ens

Vin

dex

Mex

ican

us

Bel

tian

us

Ou

tgro

up

s

Ou

tgro

up

s

Am

eth

ysti

nu

s

Fig.3.(A

)Majority

rule

consensusoftheBayesianmaxim

um

likelihoodanalysisfrom

cytochromeoxidase

subunitIdata.Numbersrepresentposteriorprobabilities.Branches

withanasteriskindicate

congruence

withD2Bayesiananalysis.

(B)Majority

rule

consensusoftheBayesianmaxim

um

likelihoodanalysisfrom

combined

molecularand

morphologicaldata

(mixed

model).Numbersrepresentposteriorprobabilities.

144 D. L. Price


the paleano group as sister to all other Phanaeini taxa(Phanaeus, and outgroups Oxysternon, Sulcophanaeus and

Coprophanaeus). Oxysternon festivum, O. durantoni and O.conspicillatum were recovered as sister clade to the bispinusand splendidulus species groups. Phanaeus bispinus is sister

to the other species (90% posterior probability), and P.alvarengai is the sister taxon to P. melibaeus and P. haroldi(94% posterior probability). Although not supported in anyother analyses, P. achilles is recovered here with the

chalcomelas group with 98% posterior probability. As inthe combined parsimony tree, the mexicanus species group isfound to be polyphyletic with P. yecoraensis as sister taxon to

P. nimrod and P. furiosus, and P. lunaris recovered within thebeltianus species group. Species groups (with two or morespecies) recovered as monophyletic include the paleano,

endymion, chalcomelas, vindex, tridens and beltianus groups.Within Phanaeus s.str. there appear to be five evolutionarylineages as follows: the endymion group; the hermes andchalcomelas groups; the triangularis, quadridens and vindex

groups; the tridens group; and the mexicanus, beltianus andamethystinus groups. When Bayesian trees were filtered forPhanaeus monophyly no trees were recovered.

Biogeography

The DIVA analysis resulted in multiple combinations of

optimal reconstructions. The ancestral distribution for theroot of the tree is uncertain, but constraining the areas totwo resulted in an ancestral distribution in the Andes (as

opposed to the widespread ancestral distribution when noconstraints were used). When no restrictions were used, andwhen the maximum number of ancestral nodes wasrestricted to three (Fig. 4), DIVA identified a total of 55

dispersal events. Fifty-four dispersal events were identifiedwhen the maximum number of ancestral nodes wasrestricted to four. Ancestors of Notiophanaeus, as limited

here (bispinus and splendidulus species groups), are sup-ported as having an origin in Amazonia. Biogeographicalscenarios are shown in Fig. 5.

Oxysternon, Sulcophanaeus and Coprophanaeus, the pa-leano species group, Notiophanaeus as here restricted, andthe chalcomelas species group are largely restricted to the

Andes and regions east of the Andes. Only five species havepenetrated northwest South America and Central America:P. pyrois, O. conspicillatum, O. silenus, C. pluto and C.telamon. The tridens, triangularis, quadridens and vindex

species groups and P. yecoraensis make up a clade thatoriginated in Mexico and have since dispersed throughoutthat country and much of the United States. Themexicanus,

beltianus and amethystinus species groups form a clade thathas dispersed back into Central and South America.

Discussion

In this study, the systematics of the genus Phanaeus wasanalysed using the subunit cytochrome oxidase I (COI), 28S

rRNA (D2 region), and 67 morphological characters fromPrice (2007). COI data are confirmed as a good data source

to resolve phylogenetic questions at this level. By contrast,D2 has evolved much more slowly, and does not give muchinformation about Phanaeus relationships. Although some

topological differences exist between the two genes, most ofthe differences are not supported by high bootstrap values.The phylogenies produced from these molecular data,although not in complete agreement with morphological

data (Price, 2007), are congruent in most respects. Conse-quently, phylogenetic analyses were also performed ona combined dataset.

Phanaeus

Based on combined analyses of COI, D2 and morpho-logical data, the paleano species group is sister to all otherphanaeine taxa. By contrast, its position in previous mor-

phological analyses (Price, 2007) suggested that the paleanospecies group was nested within the subgenus Notiopha-naeus. Phanaeus s.str. appears to represent a clade if the

endymion and chalcomelas groups are included (89% poste-rior probability). This, however, contradicts Edmonds’(1994) placement of the chalcomelas and endymion groups

within Notiophanaeus.Edmonds (1994) placed the chalcomelas and endymion

groups in Notiophanaeus based on the texture of their

pronotum. His primary distinction between Notiophanaeusand Phanaeus s.str. was that in Notiophanaeus the pronotaof both sexes are minutely punctate, although they appearglassy smooth to the unaided eye, in contrast to the rugose

to punctatorugose sculpturing of Phanaeus s.str. However,Price (2007) questioned the position of the chalcomelas andhermes groups. Here I suggest that the shape of the

pronotum is perhaps more important for separating Notio-phanaeus and Phanaeus s.str. Based on the present study, allNotiophanaeus species possess spiniform projections on their

pronota. Pronota of Phanaeus s.str. species, on the otherhand, are triangular to heart-shaped, without spiniformprojections, and are comparatively flat, except in the tridens

group and in some species of the mexicanus group. In thepresent study, Notiophanaeus consists of the bispinus andsplendidulus groups (67% parsimony bootstrap; 90% pos-terior probability), although relationships with outgroup

genera can be proposed only tentatively owing to lowsupport values. The data presented here support the mono-phyly of Phanaeus s.str.; Notiophanaeus, in the restricted

sense proposed here, probably deserves genus status, as allof the current analyses recover this clade and suggest that itis not sister to Phanaeus s.str. (see Table 3 for a breakdown

of the past and currently proposed taxonomy).With the exception of the bispinus, chalcomelas and

mexicanus groups, Edmonds’ (1994) species groups arerecovered as monophyletic. In several analyses, Price

(2007) recovered the bispinus group either in a polytomywith the splendidulus group, or inside the splendidulus group(all with less than 50% bootstrap support). However, the



bispinus group is clearly distinct morphologically from thesplendidulus group in the shape of the male pronotum: males

of both P. bispinus and P. alvarengai have convex pronotawith two short medial spiniform processes projectingtowards the head. The splendidulus group, by contrast, hasconcave pronota with lateral spiniform processes.

The mexicanus group is morphologically and geograph-ically similar to the tridens group (Edmonds, 1994). In boththe parsimony and the Bayesian combined analyses,

P. yecoraensis (mexicanus group), a recently describedspecies (Edmonds, 2004), is recovered with the tridensgroup. Furthermore, P. lunaris, a species Edmonds (1994)

regarded as taxonomically isolated from other species in the

mexicanus group, is recovered as closest to the beltianusgroup.

Edmonds (1994) did not present a phylogeny, but hisevolutionary, ecogeographic hypotheses clearly imply thefollowing relationships: {Outgroups (Notiophanaeus) [(tri-angularis gr., tridens gr., hermes gr.)(quadridens gr., ame-

thystinus gr.)(mexicanus gr., beltianus gr.)(vindex gr.)]}. Asalready mentioned, the data presented in this study suggestthe following relationships (hermes gr., chalcomelas gr.)((en-

dymion gr.)(((vindex gr.)((triangularis gr., quadridens gr.)(tri-dens gr.)))((amethystinus gr.)((mexicanus gr., beltianusgr.)))). Within Phanaeus s.str., Edmonds’ hypotheses re-

garding themexicanus and beltianus groups are in agreement

Glyphoderus sterqulinus

Anomiopsoides heteroclyta

Phaneaus paleano

P. kirbyi

Oxysternon conspicillatum

O. festivum

O. durantoni

P. bispinus

P. splendidulus

P. dejeani

P. alvarengai

P. haroldi

P. melibaeus

Sulcophanaeus faunus

S. auricollis

O. silenus

O. s. spiniferum

Coprophanaeus pluto

C. telamon

C. ignecinctus

P. endymion

P. pyrois

P. prasinus

P. achilles

P. chalcomelas

P. meleagris

P. lecourti

FI J

GI/GI J

EE

E

EI

EEG/EH

E

EE

E

E

C

BB

BC

B

BB

B

BE

E

E/BE/CE

BCD/DE/CDE

D

I

E

G/EG/EH/EFH/EGH/EI/EFI/EGI/EHI

E/EG

P. a. amethystinus

P. a guatemalensis

P. amithaon

P. w. wagneri

P. lunaris

P. howdeni

P. salleiFH/FGH

FF

D

D

BCE

CE

ABCEFAB

AB

ABE

C

C

B

ABE

BC

ABCE

E

ABEF

ABE

GH

BEFGH

B

GHI

EFG

BE

E

ABE

BE

BE

I JFG

EF

FGH

G G

G

Outgroups

paleano

Outgroups

bispinus

splendidulus

bispinus

splendidulus

hermes

chalcomelas

endymion

mexicanus

beltianus

amethystinus

Outgroups

EGI

EGH

EFG

I

FG

FGIFI

BCE

BDE

P. triangularis texensis

P. quadridens

P. igneus

P. vindexL/IL/I JL

IL

I IKL

I

L

I JKL

triangularisquadridens

vindex

I

IKL

P. yecoraensis

P. nimrod

P. furiosusI

I

Itridens

mexicanusI

I J

1a

2

1b

3a

3b

3c

4

Fig. 4. Ancestral area reconstruction from the DIVA analysis mapped onto the most likely tree (�9667.446) recovered from the Bayesian

analysis of combined molecular and morphological data. Letters above the branches are optimized ancestral areas when the maximum number

of areas is constrained to three. Letters below the branches represent alternative, equally parsimonious scenarios. Letters in square boxes

correspond to the biogeographical scenarios shown in Fig. 5.

146 D. L. Price


with the present study. In this study, the ancestral distribu-tion for the hermes group is in the Andes, and this group is

recovered as sister to the chalcomelas group, also an Andeanspecies group. Phanaeus quadridens (quadridens group) isrecovered as the sister taxon to P. triangularis texensis

(triangularis group), both of which have ancestral distribu-tions in Mexico. Phanaeus amethystinus (amethystinusgroup), by contrast, has an ancestral distribution in Central

America, and is sister to the mexicanus and beltianus groups(Edmonds’ third lineage). The phylogenetic hypothesesfrom this study provide evidence against Edmonds’s ecogeo-

graphic hypotheses.

Oxysternon, Sulcophanaeus and Coprophanaeus

Because no trees were recovered when the Bayesian treeswere filtered for trees supporting the monophyly ofPhanaeus,

these analyses suggest that Phanaeus is not monophyleticunless it is defined to includeOxysternon, Coprophanaeus andat least some Sulcophanaeus. In contrast to these findings,

recent studies that have examined the phylogeny of Phanaeini(Philips et al., 2004a) and the phylogeny of Scarabaeinae(Philips et al., 2004b; Monaghan et al., 2007) suggest that

Phanaeus is monophyletic and sister to Oxysternon.

Bayesian analyses of COI and of the combined molecularand morphological data both suggest that Oxysternon is

a polyphyletic genus with two clades: one consisting of O.festivum,O. durantoni and O. conspicillatum; and the secondconsisting of O. spiniferum spiniferum and O. silenus nested

inside Sulcophanaeus. Parsimony analyses, however, recov-ered a monophyletic (COI) or paraphyletic (combined data)Oxysternon. According to Edmonds (1972) and Edmonds &

Zidek (2004), Oxysternon is presumed to be monophyleticbased on three synapomorphic morphological characters(posterior median angle of pronotum acutely produced

between basal angles of elytra, long spiniform extension ofthe anterior angle of the metasternum, fifth abdominalsternum impressed). Price (2007) also found Oxysternon tobe monophyletic, although nested well within Phanaeus.

Sulcophanaeus is recovered as polyphyletic in all COIanalyses, although D2 analyses recovered a monophyleticSulcophanaeus. According to Edmonds (2000), no identified

synapomorphies define this genus, and Sulcophanaeus is bestinterpreted as a polyphyletic grouping of survivors placed inat least four clades. Philips et al. (2004a) further confirmed

that Sulcophanaeus is not monophyletic. Although Sulco-phanaeus is represented by only two species in the presentstudy, this research indicates that COI and D2 may have

different evolutionary histories. Further molecular evidence

Fig. 5. Biogeographical scenarios based on the DIVA analysis when the maximum number of areas is restricted to three. Top left figure is

coordinated with nodes 1A and B in Fig. 4; top right – node 2; bottom left – nodes 3A, B and C; bottom right – node 4.



using several genes of different evolutionary rates may beuseful in clarifying relationships within the genus (Philipset al., 2004a).Coprophanaeus, a monophyletic group in these analyses,

was recovered as sister taxon to Phanaeus s.str. in theBayesian analyses of combined data. In parsimony analyses,however, Coprophanaeus was recovered as sister to Sulco-

phanaeus. Interestingly, in a monograph on the phanaeinespecies, Olsoufieff (1924) classified Sulcophanaeus and Cop-rophanaeus as subgenera within Phanaeus, although, upon

later examination, Edmonds (1972) raised these taxa togenus level. Edmonds (1972) was the first to publisha topology of hypothesized generic relationships within

the phanaeines, although his hypothesis suggested a basalpolytomy as follows [(Oxysternon, Phanaeus)(Sulcopha-naeus)(Diabroctis)(Coprophanaeus, Dendropaemon, Tetra-meria, Megatharsis)]. In their examination of the

Phanaeini, Philips et al. (2004a) recovered an equallyweighted tree with Bolbites, Diabroctis, Oxysternon, Pha-naeus and Sulcophanaeus in a polytomy with another clade

including Coprophanaeus. Additional weighted schemesrecovered Sulcophanaeus and Oxysternon þ Phanaeus asmonophyletic. The evolutionary relationships among the

Phanaeini require further examination with more data andseveral species from each of the 12 genera, including speciesfrom each of the proposed species groups.

Biogeography

The age of the common ancestor of Phanaeini isspeculative, but phanaeines probably evolved in SouthAmerica after the late Mesozoic separation from Africa

(Philips et al., 2004). Fossilized brood balls attributed toscarabs, including phanaeines, have been found in variousTertiary deposits in southern Argentina (cited in Frenguel-li, 1938, 1939; Edmonds, 1972). These balls indicate the

presence of phanaeines or their ancestors in southernSouth America at least 28.5 million years ago (Ma)(F. C. Ocampo, personal communication, 2006). Extant

species of Phanaeus are distributed in three main geo-graphic areas. Dispersal-vicariance reconstruction indi-cates that the Andes are probably the ancestral area of

the genus Phanaeus. Dispersal from the Andes into Ama-zonia was followed by a vicariance event. Within Notio-phanaeus (as presented here), further dispersal occurredfrom Amazonia into southern Brazil and northwards into

the Guianas region. Oxysternon and Sulcophanaeus haveundergone similar expansions. Coprophanaeus, by con-trast, moved northwestwards into Central America.

The earliest representatives of Phanaeus s.str. probablyarrived inMesoamerica during theMiocene (23–5 Ma), whenthe physiographic diversification of Central America and

Mexicowas just beginning (Edmonds, 1994). Studies on otherdung beetle genera (i.e. Ateuchus, Canthon, Boreocanthon,Melanocanthon) suggest that there have been two main incur-

sions of scarabaeines from South America intoMesoamerica:one during the Miocene and the other following the finalestablishment of the Central America landbridge during thelate Pliocene and early Pleistocene (Kohlmann & Halffter,

1988, 1990). Portions of the landbridge between SouthAmerica and Mexico were present during the middle of theMiocene (15–16 Ma), although the northwest portion of

South America (biogeographical region F), as indicated inFig. 1, was still probably deep ocean. By the late Pliocene(c. 3 Ma), much of the landbridge was present, including

portions of northwestern SouthAmerica and Panama (Cox&Moore, 2000). Fossil Phanaeus are limited to two species:Phanaeus antiquus Horn, 1876, and Phanaeus labreae Pierce,

1946 (Krell, 2007). These records are in agreement with theevolutionary hypotheses presented in this manuscript.Edmonds (1994) hypothesized that the endymion group

represents a recent invasion of Notiophanaeus into Middle

America. The scenario presented here suggests that ancestorsof the endymion group were the first of the Phanaeus s.str. toinvade Middle America. Ancestors of the tridens (including

P. yecoraensis from Edmonds’mexicanus group), triangularis,quadridens and vindex groups represent the second wave ofMesoamerican invaders, and most extant species are now

endemic toMexico and the United States. Species groups thathave dispersed back into Central and South America includethe mexicanus, beltianus and amethystinus groups.Tropical South America is home to a rich fauna of dung

beetles, including a number of large species with ecologicalrequirements similar to those of Notiophanaeus. Hence, theextensive radiation of Phanaeus s.str. in Mesoamerica could

Table 3. Previous taxonomy of species groups versus current

hypotheses.

Edmonds’s (1994)

hypothesis Price (2007) Present study

Outgroups Outgroups Outgroups

Anomiopsoides Anomiopsoides Anomiopsoides

Glyphoderus Glyphoderus Glyphoderus

Coprophanaeus Coprophanaeus paleano species gr.a

Sulcophanaeus Sulcophanaeus Oxysternonb

Oxysternon

Phanaeus (genus) Phanaeus (genus) Notiophanaeus (genus)

Notiophanaeus

(subgenus)

Notiophanaeus

(subgenus)

bispinus gr.

splendidulus gr.

splendidulus gr. Oxysternon Sulcophanaeusb

chalcomelas gr. splendidulus gr. Coprophanaeus

bispinus gr. bispinus gr. Phanaeus (genus)

paleano gr. paleano gr. endymion gr.

endymion gr. chalcomelas gr. hermes gr.

Phanaeus s.str.

(subgenus)

endymion gr.

hermes gr.

chalcomelas gr.

triangularis gr.

hermes gr.

tridens gr.

Phanaeus s.str.

(subgenus)

quadridens gr.

vindex gr.

triangularis gr. tridens gr. tridens gr.

mexicanus gr. triangularis gr. mexicanus gr.

beltianus gr. mexicanus gr. beltianus gr.

amethystinus gr. beltianus gr. amethystinus gr.

quadridens gr. amethystinus gr.

vindex gr. quadridens gr.

vindex gr.

aRequires new generic name if position confirmed.bPossibly polyphyletic.

148 D. L. Price


have been favoured in great part by the scarcity of potentialcompetitors. The rich Miocene–Pliocene fauna of large

mammals is presumed to have supported the originalinvasion of Phanaeus (Edmonds, 1994).The mountains of Mexico and Central America are cur-

rently recognized as regions of high levels of diversificationand endemism for flora and fauna, including insects,herpetofauna, birds and mammals (Leon-Paniagua et al.,2007). Halffter (1987) attributes this rich diversity partly to

the great variety of environments and ecological refugesavailable in the zone and partly to adequate routes fordispersal of faunas of different origin, ranging from cold-

temperate mountain to humid tropical corridors.Currently, the extant fauna of Phanaeus is richer in North

and Central America (32 species) than it is in South America

(19 species). The survival of Phanaeus during the Pleistoceneextinctions may be attributable to their generalist feedinghabits and to their ability to reproduce in a variety ofecological habitats. Edmonds (1994) argued that the mod-

ern distribution of Phanaeus has been strongly influenced byhumans. In Mexico, the presence, and increasing popula-tion, of humans (and livestock) as early as 650 AD was

important not only as a source of food but also as the causeof habitat changes that favoured the expansion of speciespreferring open habitats. The phylogenies presented here

support Edmonds’ (1994) idea that species groups found inMexico and Central America are actively evolving. Addi-tional molecular studies focusing on the Pliocene–Pleisto-

cene radiation in Mexico may provide insight into, forexample, how and why some species in the mexicanus grouphave migrated back into Central and South America, andinto their morphological evolution (for example, some

species within the mexicanus group have male pronotalfeatures that are not typical of Phanaeus s.str.).

Acknowledgements

I would like to thank Michael May, Karl Kjer, Lena Struwe

and Doug Tallamy for discussions and comments on thispaper. I also thank W.D. Edmonds for discussions andcomments. I thank several people who have helped me along

the way, including Jessica Ware (Rutgers University), JohnLaPolla (Towson University) and Jeremy Huff (AMNH). Iam grateful to the following people who provided me with

specimens for sequencing: W.D. Edmonds (Marfa, TX),Francois Genier (CMN, Ottawa, Canada), Sacha Spector(AMNH, New York, NY), Trond Larsen (Princeton Univer-

sity, Princeton, NJ), Kevina Vulinec (Delaware State Univer-sity, Dover, DE), Barney Streit (Tucson, AZ), Conrad Gillett(NHM, London, U.K.), Kyle Beucke (University of Flori-da, FL) and Christopher Marshall (Oregon State Univrsity,

OR). I also thank Federico Ocampo for providing me witheucraniine sequences, Scott Haag (Rutgers University) forcreating and providing a biogeography map, and Frederick

Ronquist for providing help with DIVA. Lastly, I thankMichael May for providing the funds for this project.

References

Arnaud, P. (1997) Description d’une nouvelle espece de Phanaeus

(Col. Scarabaeidae). Besoiro, 3, 6–7.

Arnaud, P. (2000) Descriptions de nouvelles especes de Phanaeini.

Besoiro, 5, 6–8.

Arnaud, P. (2001) Description de nouvelles espece de Phanaeides.

Besoiro, 6, 2–8.

Arnaud, P. (2002) The Beetles of the World, Vol. 28: Phanaeini.

Hillside Books, Canterbury.

Bennett, C.F. (1969) Human Influences on the Zoogeography of

Panama. University of California Press, Berkeley, CA.

Blanchard, C.E. (1845) Insectes, Voyage dans I’Amerique Meridionale

(ed. by A. d’Orbigny), vol. 6, pt. 2, pp. 61–222. Paris & Strasbourg.

Cox, C.B. & Moore, P.D. (2000) Biogeography an Ecological and

Evolutionary Approach, 6th edn. Blackwell Science, London.

Delgado-Castillo, L. (1991) A new Mexican species of Phanaeus.

Opscula Zoologica Fluminensia, 68, 1–6.

Edmonds, W.D. (1972) Comparative skeletal morphology, system-

atics, and evolution of the Phanaeine dung beetles (Coleoptera:

Scarabaeidae). The University of Kansas Bulletin, 49, 731–874.

Edmonds, W.D. (1994) Revision of PhanaeusMacleay, a new world

genus of scarabaeine dung beetles (Coleoptera: Scarabaeidae,

Scarabaeinae). Contributions in Science, Natural History Museum

of Los Angeles County, 443, 1–105.

Edmonds, W.D. (2000) Revision of the Neotropical dung beetle

genus Sulcophanaeus (Coleoptera: Scarabaeidae: Scarabaeinae).

Folia Heyrovskyana, 6, 1–60.

Edmonds, W.D. (2004) A new species of Phanaeus Macleay

(Coleoptera: Scarabaeidae, Scarabaeinae) from Sonora, Mexico.

The Coleopterists Bulletin, 58, 119–224.

Edmonds, W.D. (2006) A new species of Phanaeus Macleay

(Coleoptera: Scarabaeidae: Scarabaeinae: Phanaeini) from Oaxa-

ca, Mexico. Zootaxa, 1171, 31–37.

Edmonds, W.D. & Zidek, J. (2004) Revision of the Neotropical

dung beetle genus Oxysternon (Coleoptera: Scarabaeinae: Pha-

naeini). Folia Heyrovskyana, 11, 1–58.

Felsenstein, J. (1985) Confidence limits on phylogenies: an

approach using the bootstrap. Evolution, 39, 783–791.

Frenguelli, J. (1938) Bolas de escarabaeoides y nidos de vespidos

fosiles. Physis, 12, 348–352.

Frenguelli, J. (1939)Nidos fosiles de insectos en el Tercario deNeuquen

y Rio Negro. Notas Museo La Plata, Paleontologia, 18, 379–393.

Gillespie, J., Cannone, J., Gutell, R. & Cognato, A. (2004) A

secondary structural model of the 28S rRNA expansion segments

D2 and D3 from rootworms and related leaf beetles (Coleoptera:

Chrysomeldae; Galerucinae). Insect Molecular Biology, 13,

495–518.

Gu, X., Fu, Y.-X. & Li, W.-H. (1995) Maximum likelihood estima-

tion of the heterogeneity of substitution rate among nucleotide

sites. Molecular Biology and Evolution, 12, 546–557.

Halffter, G. (1987) Biogeography of the montane entomofauna of

Mexico and Central America. Annual Review of Entomology, 32,

95–114.

Hooghiemstra, H., Wijninga, V.M. & Cleef, A.M. (2006) The paleo-

botanical record of Colombia: implications for biogeography and

biodiversity. Annals of the Missouri Botanical Garden, 93, 297–324.

Hulsenbeck, J.P. & Ronquist, F. (2001) MRBAYES 3: Bayesian

inference of phylogeny. Bioinformatics, 17, 754–755.

Kjer, K.J. (1995) Use of rRNA secondary structure in phylogenetic

studies to identify homologous positions: an example of align-

ment and data presentation from the frogs. Molecular Phyloge-

netics and Evolution, 4, 314–330.



Kohlmann, B. & Halffter, G. (1988) Cladistic and biogeographical

analysis of Ateuchus (Coleoptera: Scarabaeidae) of Mexico and

the United States. Folia Entomologica Mexicana, 74, 109–130.

Kohlmann, B. & Halffter, G. (1990) Reconstruction of a specific

example of insect invasion waves: the cladistic analysis of

Canthon (Coleoptera: Scarabaeidae) and related genera in North

America. Quaestiones Entomologicae, 26, 1–20.

Kohlmann, B. & Solis, A. (2001) A new species of Phanaeus

Macleay from Costa Rica and Panama. Besoiro, 6, 9–11.

Krell, F.T. (2007) Catalogue of Fossil Scarabaeoidea (Coleoptera:

Polyphaga) of the Mesozoic and Tertiary, Version 2007. Denver

Museum of Nature and Science Technical Report No. 2007-8, 1–79.

Leon-Paniagua,L.,Navarro-Siguenza,A.G.,Hernandez-Banos,B.E.&

Morales, J.C. (2007) Diversification of the arboreal mice of the genus

Habromys (Rodentia: Cricetidae: Neotominae) in theMesoamerican

highlands.Molecular Phylogenetics and Evolution, 42, 653–664.

Metcalfe, S.E. (2006) Late quaternary environments of the northern

deserts and central transvolcanic belt of Mexico. Annals of the

Missouri Botanical Garden, 93, 258–273.

Minin, V., Abdo, Z., Joyce, P. & Sullivan, J. (2003) Perform-

ance-based selection of likelihood models for phylogeny estima-

tion. Systematic Biology, 52, 674–683.

Monaghan, M.T., Inward, D.J.G., Hunt, T. & Vogler, A.P. (2007)

A molecular phylogenetic analysis of the Scarabaeinae (dung

beetles). Molecular Phylogenetics and Evolution, 45, 674–692.

Ocampo, F.C. (2004) Food relocation behavior and synopsis of the

southern South America genus Glyphoderus Westwood (Scara-

baeidae: Scarabaeinae: Eucraniini). The Coleopterists Bulletin, 58,

295–305.

Ocampo, F.C. (2005) Revision of the southern South American

endemic genus Anomiopsoides Blackwelder, 1944 (Coleoptera:

Scarabaeidae: Scarabaeinae: Eucraniini) and description of its food

relocation behaviour. Journal of Natural History, 39, 2537–2557.

Ocampo, F.C. & Hawks, D. (2006) Molecular phylogenetics and

evolution of the food relocation behavior of the dung beetle tribe

Eucraniini (Coleoptera: Scarabaeidae: Scarabaeinae). Inverte-

brate Systematics, 20, 557–570.

Olsoufieff, G. (1924) Les Phanaeides (Col. Lamel.), Famille Scar-

abaeidae – Tr Coprini. Insecta, 13, 4–172.

Perret, M., Chautems, A. & Spichiger, R. (2006) Dispersal–

vicariance analyses in the Tribe Sinningieae (Gesneriaceae): a clue

to understanding biogeographical history of the Brazilian

Atlantic Forest. Annals of the Missouri Botanical Garden, 93,

340–358.

Philips, T.K., Edmonds, W.D. & Scholtz, C.H. (2004a) A phyloge-

netic analysis of the New World tribe Phanaeini (Coleoptera:

Scarabaeidae: Scarabaeinae): hypotheses on relationships and

origins. Insect Systematics and Evolution, 35, 43–63.

Philips, T.K., Pretorius, E. & Scholtz, C.H. (2004b) A phylogenetic

analysis of dung beetles (Scarabaeinae: Scarabaeidae): unrolling

an evolutionary history. Invertebrate Systematics, 18, 53–88.

Posada, D. & Buckley, T. (2004) Model selection and model

averaging in phylogenetics: advantages of Akaike information

criterion and Bayesian approaches over likelihood ratio tests.

Systematic Biology, 53, 793–808.

Posada, D. & Crandall, K.A. (1998) Modeltest: testing the model of

DNA substitution. Bioinformatics, 124, 817–818.

Price, D.L. (2007) A phylogenetic analysis of the dung beetle genus

Phanaeus (Coleoptera: Scarabaeidae) based on morphological

data. Insect Systematics and Evolution, 38, 1–18.

Ronquist, F. (1996) Reconstructing the history of host-parasite

associations using generalised parsimony. Cladistics, 11, 73–89.

Ronquist, F. (1997) Dispersal-vicariance analysis: A new approach

to the quantification of historical biogeography. Systematic

Biology, 46, 195–203.

Simon, C., Frati, F., Beckenbach, A., Crespi, B., Liu, H. & Flook,

P. (1994) Evolution, weighting, and phylogenetic utility of

mitochondrial gene sequences and a compilation of conserved

polymerase chain reaction primers. Annals of the Entomological

Society of America, 87, 651–701.

Swofford, D.L. (1999) PAUP*: Phylogenetic Analysis using Parsi-

mony, Version 4. Sinauer, Sunderland, MA.

Thompson, J.D., Gibson, T.J., Plewniak, F., Jeanmougin, F. &

Higgins, D.G. (1997) The ClustalX windows interface: flexible

strategies for multiple sequence alignment aided by quality

analysis tools. Nucleic Acids Research, 24, 4876–4882.

Westwood, J. O. (1837) Illustrations of Exotic Entomology,

Containing Upwards of Six Hundred and Fifty Figures and Descrip-

tions of Foreign Insects, Interspersed with Remarks and Reflections

on their Nature and Properties, Vol. 1. H.G. Bohn, London.

Yang, Z. (1994) Estimating the pattern of nucleotide substitution.

Journal of Molecular Evolution, 39, 105–111.

Yang, Z., Goldman, N. & Friday, A.E. (1994) Comparison of

models for nucleotide substitution used in maximum estimation.

Molecular Biology and Evolution, 11, 316–324.

Accepted 15 May 2008

First published online 18 December 2008

150 D. L. Price


Phylogeny and biogeography of the dung beetle genus ...

Documents

Transcript of Phylogeny and biogeography of the dung beetle genus ...