BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors nonprofitpublishers academic institutions research libraries and research funders in the common goal of maximizing access tocritical research

Mate Recognition Differences Among Allopatric Populationsof the Scarab Canthon cyanellus cyanellus (ColeopteraScarabaeidae)Author(s) Maribel Ortiacutez-domiacutenguez Mario E Favila Maria R Mendoza-loacutepezSource Annals of the Entomological Society of America 99(6)1248-12562006Published By Entomological Society of AmericaDOI httpdxdoiorg1016030013-8746(2006)99[1248MRDAAP]20CO2URL httpwwwbiooneorgdoifull1016030013-874628200629995B12483AMRDAAP5D20CO3B2

BioOne (wwwbiooneorg) is a nonprofit online aggregation of core research in thebiological ecological and environmental sciences BioOne provides a sustainable onlineplatform for over 170 journals and books published by nonprofit societies associationsmuseums institutions and presses

Your use of this PDF the BioOne Web site and all posted and associated contentindicates your acceptance of BioOnersquos Terms of Use available at wwwbiooneorgpageterms_of_use

Usage of BioOne content is strictly limited to personal educational and non-commercialuse Commercial inquiries or rights and permissions requests should be directed to theindividual publisher as copyright holder

BEHAVIOR

Mate Recognition Differences Among Allopatric Populations of theScarab Canthon cyanellus cyanellus (Coleoptera Scarabaeidae)

MARIBEL ORTIZ-DOMINGUEZ12 MARIO E FAVILA1 AND MARIA R MENDOZA-LOPEZ3

Ann Entomol Soc Am 99(6) 1248ETH1256 (2006)

ABSTRACT We analyzed mate recognition and the cooperative behavior of maleETHfemale pairs inCanthon cyanellus cyanellus LeConte (Coleoptera Scarabaeidae) a carrion-feeding scarab duringfood ball rolling among THORNve populations found in tropical forest fragments on the eastern coastal plainsof Mexico Sexual recognition in this species is mostly mediated by chemical cues Intrapopulation andmaleETHfemale pairs formed with individual from populations 400 km apart showed cooperativebehavior (joint rolling of the food ball by both members) However maleETHfemale pairs from popu-lations 400 km apart showed no cooperative behavior (individual food ball rolling by one of thepartners and THORNghts) although they also rolled food balls together In maleETHfemale pairs from the moredistant populations (600 km apart) 50 fought for food balls Brood ball production by femalesand the number of eclosed beetles from the northern population and from the crosses betweenindividuals from the most distant populations were lower than those of other intra- or interpopulationcombinations Gas chromatograph mass spectroscopy of cuticular extract showed that the southernpopulation had 48 surface compounds 18 of them not present in the northern or intermediatepopulations The intermediate population had 41 compounds 11 of them exclusive However thenorthern population had 24 compounds and only one was exclusive The lack of cooperative inter-action among maleETHfemale pairs from distant populations along with their reduced reproductivesuccess and the differences in cuticular composition between populations suggest that extremepopulations of Canthon c cyanellus are in a process of incipient speciation

KEY WORDS ecological speciation Canthon cyanellus prezygotic isolation sexual recognition

Recent studies of the evolutionary forces involved inspeciation suggest that only a few mechanisms areresponsible for this phenomenon (Schluter 2001) Themost widely accepted idea is that reproductive isola-tion is a consequence of adaptive divergence amongpopulations (Mendelson 2003 Vines and Schluter2006) Isolation offers the potential for adaptation tothe local environment Spatially separated populationsfollow different sexual selection processes and theirseparation favors reproductive barriers (Boake et al1998 Del Campo et al 2003) Reproductive isolationfrom other species or from allopatric populations ofthe same species can occur at the prezygotic or post-zygotic level (Panhuis et al 2001)

Geographical ecological seasonal morphologicalphysiological orethological factors are responsible forprezygotic isolation Sexual isolation occurs when geo-graphical separation produces different modes ofcourtship between males and females includingchemical visual tactile and auditory signals (Paillette

et al 1997 Ritchie et al 1999 Rybak et al 2002) Thesesignals may be exclusive to each species and can varyamong populations (Paillette et al 1997) Thus onanalyzing sexual isolation among populations it is im-portant to distinguish between the effects of naturalselection and sexual selection forces Natural selectionoperates independently of mating system (Miller et al1998 Noor et al 2000) Subspecies or new species areformed after natural selection has led to differences insexual recognition between individuals of differentpopulations (Michalak et al 1997 Schluter 2001 Etges2002)

The formative phase in the history of species isfrequently extrapolated to compare closely relatedspecies which because of their relatedness are sup-posed to have speciated recently A more difTHORNcultapproach is to attempt to identify incipient speciationwithin what seems to be by most criteria a singlespecies (Schluter 2000) Canthon cyanellus LeConte(Coleoptera Scarabaeidae) is a carrion ball rollerscarab found from Brazil to southern Texas Owing toits large distribution several subspecies have beenproposed according their coloration (Halffter 1961)In Mexico Canthon cyanellus cyanellus lives in thetropical forest of the south and on the eastern andwestern coastal plains (Halffter et al 1992) The re-productive period of this subspecies is from May to

1 Instituto de Ecologotildea AC Km 25 Antigua Carretera a CoatepecNo 351 Congregacion el Haya AP 63 Xalapa CP 91070 VeracruzMexico

2 Corresponding author e-mail mariofavilainecoledumx3 Universidad Veracruzana Unidad de Servicios de Apoyo a la

Resolucion Analotildetica Dr Luis Castelazo Ayala SN col IndustrialAnimas Xalapa CP 91190 Veracruz Mexico

0013-8746061248ETH1256$04000 2006 Entomological Society of America

September the rainy season (Halffter et al 1983Favila and Dotildeaz 1996 Villalobos et al 1998) Individ-uals locate vertebrate carcasses by their odor andinitial sexual interactions occur at the carcass Sexualrecognition is mostly mediated by cuticular com-pounds putatively hydrocarbons (Ortiz-Domotildenguezet al 2006) When same-sex or opposite-sex individ-uals meet they make head-to-head contact extendingtheir antennae and maxillary palpi or even touch thebody of the other individual with the antennae Same-sex individuals THORNght THORNercely for the food ball but if themeeting is between a male and a female cooperativelythey will cut and roll a food ball for nesting (Favila1988a Favila and Dotildeaz 1996) MEF (unpublisheddata) has observed the behavior of males and femalescollected in tropical forests from the east coast ofMexico but from sites 600 km apart maleETHfemale pairsfrom the same population cooperatively rolled a foodball but the female from the northern population(Gomez Farotildeas Mexico) always refused the male fromthe southern population (Los Tuxtlas Mexico) Suchobservations suggested that there was either difTHORNcultyin the sexual recognition of or discrimination againstgeographically separated potential mates among indi-viduals from distant populations

In this study we analyze premating behavior duringsexual encounters and food ball rolling of male andfemale pairs of individuals belonging to THORNve separatedpopulations of Canthon c cyanellus found over a lat-itudinal gradient on the eastern coastal plains of Mex-

icoBecause sexual recognition in this species ismostlymediated by cuticular compounds cuticular chemicalcomposition was compared among the populations Toevaluate potential postmating reproductive isolationreproductive success (measured as brood ball produc-tion by the female and survival of eclosed beetles) wascompared in maleETHfemale pairs belonging to the samepopulation and for pairs in which each sex belongedto a different population

Materials and Methods

The accelerated process of deforestation in Mexicohas produced a landscape dominated by pastureswhere forest fragments are maintained to provideshade for cattle This is particularly true of the easterncoastal plain where the original continuous distribu-tion of C c cyanellus on this plain is now mostlyrestricted to these forest fragments (Halffter et al1992)

To adequately represent the separate populationsfrom along a latitudinal gradient on the eastern coastalplains of Mexico pitfall traps were baited with THORNsh andset in the soil to capture carrion beetles at the follow-ing localities (Fig 1) Los Tuxtlas Veracruz (18 35 N95 04 W) Palma Sola Veracruz (19 46 N 95 26 W)Papantla Veracruz (20 25 N 97 27 W) TancocoVeracruz (21 19 N 97 50 W) and Gomez FarotildeasTampaulipas (23 03 N 99 00 W) Beetles werealways collected from within the forest although they

Fig 1 Location of the THORNve allopatric populations of Canthon c cyanellus in Mexico

November 2006 ORTIZ-DOMINGUEZ ET AL MATE RECOGNITION IN ALLOPATRIC SCARABS 1249

are also found in pastures near the fragments (Halffteret al 1992) The THORNrst locality was in the OcircLos TuxtlasOtildebiological research station (National AutonomousUniversity of Mexico) located within the Los TuxtlasBiosphere Reserve and the last site is located in the OcircElCieloOtilde Biosphere Reserve (University of Tamaulipas)There are visible differences between some popu-lations in the morphology of their cuticular colorand texture We focused our attention on elytral vari-ation in three of the populations analyzed specimensfrom Los Tuxtlas the southern site have shiny greenelytra with a smooth interpunctual surface and feeblypunctuate elytral striae The elytra of the Papantlaspecimens the intermediate site are blue or mattegreen but the interpunctual surface is generally THORNneshagreening with punctuate elytral striae The elytraof the Gomez Farotildeas specimens the northern site areshiny green with a smooth interpunctual surface andpunctuate elytral striae Specimens from each popu-lation were put in containers with moist soil in arearing room at 27C 70 10 RH and a photoperiodof 1212 (LD) h We followed the rearing protocol ofFavila (1993)

Experiments were conducted in another insec-tarium under the same conditions as those of therearing room Because it is not possible to distinguishthe sex of these beetles with the naked eye sex iden-tiTHORNcation was done beforehand under the microscopeThen in each experiment one elytrum of one memberof the pair was marked with a small spot of latex paint(known to not affect behavior Favila 1988b) Exper-

iments were conducted from 0900 to 1300 hours themost active rolling period of the day for this species(Favila and Dotildeaz 1996)Intra- and Intersite Sexual Recognition MaleETHfe-

male pairs were randomly selected from each site andeach one pair was placed in an observation arena (10-by 2-cm petri dish lined with THORNlter paper and a 2 g ofTHORNsh meat ball) As we were interested in observing thebehavior of an individual of each sex in the presenceof an individual of the opposite sex from the same oranother site we made 25 combinations (male frompopulation A with female from population B and viceversa male from C and vice versa and so on) includ-ing intra- and interpopulation combinations (unequaln for each combination is indicated in Table 1) Timeuntil interaction was measured as the elapsed time (inseconds) between introducing the pair into the arenaand the beginning of maleETHfemale interaction on thefood ball Time to rolling was measured as the numberof seconds that elapsed between the beginning of themaleETHfemale pair interaction on the food ball and ballrolling by one or both partners The sex of the indi-vidual that rolled the food ball was recorded Weconsidered the rolling of the food ball by both indi-viduals as cooperative behavior but always during thisjoint rolling one of the partners usually the male rollsthe ball while the other partner is transported on theball The frequency of the following noncooperativebehaviors also was analyzed 1) avoidance when oneindividual rolled the food ball alone and did not allowthe other member to climb on the ball 2) pushing

Table 1 Latency in initial interaction and rolling and proportion of noncooperative behavior (individual rolling of the food ball andfights) observed between malendashfemale pairs of Canthon c cyanellus from five populations on the east coast of Mexico and in theircombinations

Male pop Female popDistance

(km)n

Time to interaction(25ETH75)

Time to rolling(25ETH75)

Noncooperativebehavior ()

Papantla Papantla 0 30 66 (3225ETH165) 14 (60ETH27)a 0Palma Sola Palma Sola 0 29 39 (2825ETH11175) 12 (675ETH285)bc 3Gomez Farotildeas Gomez Farotildeas 0 26 33 (16ETH75) 26 (65ETH405) 0Tancoco Tancoco 0 26 46 (25ETH65) 14 (60ETH40) 0Tuxtlas Tuxtlas 0 30 38 (15ETH92) 255 (100ETH50) 3Papantla Tancoco 107 28 605 (275ETH81) 23 (1125ETH10625) 4Tancoco Papantla 107 30 505 (19ETH123) 295 (60ETH45) 7Papantla Palma Sola 122 31 365 (20ETH93) 20 (975ETH505) 13Palma Sola Papantla 122 30 46 (23ETH13275) 24 (1425ETH3675) 17Palma Sola Tuxtlas 195 23 40 (2425ETH7875) 20 (95ETH51) 9Tuxtlas Palma Sola 195 15 40 (15ETH745) 15 (1075ETH2525) 7Gomez Farotildeas Tancoco 223 19 705 (14ETH94) 21 (85ETH7125) 5Tancoco Gomez Farotildeas 223 20 74 (35ETH93) 28 (10ETH565) 0Palma Sola Tancoco 228 30 65 (35ETH127) 615 (25ETH144)ac 10Tancoco Palma Sola 228 30 595 (38ETH146) 325 (13ETH60) 23Papantla Tuxtlas 317 29 67 (1475ETH9425) 345 (14ETH93) 14Tuxtlas Papantla 317 30 58 (36ETH75) 15 (65ETH3875) 17Papantla Gomez Farotildeas 331 26 60 (2875ETH805) 56 (1875ETH12075)b 19Gomez Farotildeas Papantla 331 26 41 (24ETH925) 385 (12ETH785) 19Tancoco Tuxtlas 424 25 116 (45ETH18925) 24 (165ETH155) 20Tuxtlas Tancoco 424 27 81 (3525ETH117) 26 (1425ETH4975) 4Palma Sola Gomez Farotildeas 447 31 36 (20ETH6475) 28 (105ETH6725) 6Gomez Farotildeas Palma Sola 447 20 99 (3575ETH16625) 28 (115ETH385) 70a

Gomez Farotildeas Tuxtlas 643 31 53 (2525ETH1075) 345 (15ETH124) 55a

Tuxtlas Gomez Farotildeas 643 24 575 (375ETH715) 32 (1225ETH9825) 50a

Similar letters indicate signiTHORNcant differences between maleETHfemale pair combinations according to DunnOtildes test (P 005) Q values a 44 b 39 and c 47a G 10947 df 24 P 005

1250 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 99 no 6

when one individual attempted to climb on the foodball and the other pushed it back by ldquobuttingrdquo it and3) THORNghts both male and female fought for the ball onsome occasions cutting it so that each beetle rolled anindependent fragment of food Once the pair hadmade contact on the food ball the total observationtime lasted 5 min An analysis of variance (ANOVA)was used to analyze time until interaction and rollingoccurred both among populations and for combina-tions The frequency of each noncooperative interac-tion was analyzed with the G test (Sokal and Rohlf1981)Reproductive Success among Populations We did

not know the age and reproductive condition of spec-imens collected in the THORNeld Therefore we allowedpairs of the F1 of the laboratory strains originallycollected from the THORNve sites and more distant combi-nations (where noncooperative behaviors were morefrequently observed see Results) to mate MaleETHfe-male pairs were 20ETH30 d old the optimal age fornesting in this species (Favila 1993) Each maleETHfe-male pair was placed in a plastic nesting box (9 by 9by 8 cm) containing 25 cm of moist soil and a pieceof THORNsh suitable for reproduction (1 g) Pairs were fedTHORNsh every other day and brood ball production wasfollowed for 40 d the most active period in the re-productive life of a female under laboratory condi-tions After this period females tend to produce broodballs sporadically and offspring survival is low (Favila1993) We measured the elapsed time between theintroduction of the pair to the nesting box and theconstruction of the THORNrst brood ball The total numberof brood balls produced by a female was considered ameasure of fecundity and the number of beetleseclosed from the brood balls as a measure of fertilityEvery variable was analyzed using a KruskalETHWallistestChemical Analysis Elytra from THORNve males and THORNve

females of the populations that were furthest apart(Gomez Farotildeas and Los Tuxtlas) and one intermediatepopulation (Papantla) were individually put intoWheaton vials containing 1 ml of hexane (HPLCgrade Sigma-Aldrich St Louis MO) Every vial wasgently shaken on a shaker table for 1 min and left at 4Cfor 24 h Next the elytra were removed from thevials and the solvent was evaporated using N2 Then50 l of hexane was added to each vial along withpentadecane (n-C15) as an internal standard A 2-lsample was injected into a HewlettETHPackard G1800BGC equipped with a fused silica capillary column(HP-5) cross-linked with 5 methylsiloxane (025mm 30 m) and 025-m THORNlm thickness supplied byAgilent (Mexico DF) with helium as the carrier gasat a szligow rate of 1 mlmin We optimized the separa-tion of extracted compounds by using a column tem-perature proTHORNle in which the analysis began at 70Cwas increased 20Cmin to 190C increased again10Cmin to 250C and then increased 30Cmin to280C and that temperature was held for 10 minWith the presenceETHabsence matrix of compounds weapplied the simple matching coefTHORNcient The relation-ship among individuals belonging to the three popu-

lations was represented using average linkage clusteranalysis (unweighted pair-group method with arith-metic average)

Results

Intra- and Interpopulation Sexual RecognitionWedid not THORNnd signiTHORNcant differences in the time untilinteraction occurred among the maleETHfemale com-binations (F 124 df 24 P 020) Howevertime to rolling was signiTHORNcantly different amongtreatments (Table 1 H 5636 df 24 P 0001)Pairs from Palma Sola began to roll the food ball fasterthan pairs belonging to the Papantla-Gomez Farotildeasand Palma Sola-Tancoco combinations Pairs fromPapantla began to roll faster than pairs from thePalma Sola-Tancoco combination MaleETHfemale pairsfrom the same populations always showed joint rollingbehavior only for one pair from Los Tuxtlas and onepair from Palma Sola was pushing (a noncooperativebehavior) observed during joint rolling (Table 1) inboth cases the female pushed the male

Joint rolling was observed in all the pairs from in-terpopulation combinations but there was also non-cooperative behavior in 19 of these combinationsThere was a clear correlation between the frequencyof noncooperative behavior and the geographic dis-tance between populations (Pearson coefTHORNcient r 0727 P 0001) However only in three of thesecombinations (male from Gomez Farotildeasfemale fromPalma Sola male from Gomez Farotildeasfemale from LosTuxtlas and male from Los Tuxtlasfemale fromGomez Farotildeas) was the frequency of noncooperativebehaviors signiTHORNcantly higher than in pairs from theother combinations (Table 1G 10947 df 24 P005) Individuals in these combinations belonged tothe most distant populations For the combinationwith the male from Gomez Farotildeas and the female fromPalma Sola in 45 of the pairs formed females rolledthe food ball alone preventing the males from climb-ing onto the food ball by pushing them off with theirheads In 25 of these combinations THORNghts for the foodball were observed However for the inverse combi-nation (male from Palma Solafemale from GomezFarotildeas) joint rolling was frequently observed (Table1) except in two pairs (6) where the females rolledthe ball alone and prevented the males from climbingonto the food ball In contrast for both combinationsfrom Los Tuxtlas and Gomez Farotildeas a signiTHORNcant num-ber of noncooperative behaviors were observed Inthe male from Gomez Farotildeasfemale from Los Tuxtlascombination males exhibited signiTHORNcantly more non-cooperative behavior than females (67 versus 8 re-spectively Fisher test 005) However in the malefrom Los Tuxtlasfemale from Gomez Farotildeas combi-nation 70 of the females exhibited noncooperativebehavior with males (Fisher test 002) whereasmales did not show any noncooperative behavior Ingeneral individuals from Gomez Farotildeas showed morenoncooperative behaviors toward individuals fromthe opposite sex of Los Tuxtlas than vice versa (69 and5 respectively Fisher test 00015)

November 2006 ORTIZ-DOMINGUEZ ET AL MATE RECOGNITION IN ALLOPATRIC SCARABS 1251

Reproductive Success among Populations Fecun-dity was different among populations (Fig 2 H 2177 df 5 P 005) The fecundity of femalesfrom Gomez Farotildeas and that of the Los Tuxtlas-GomezFarotildeas combination was signiTHORNcantly lower than thoseof Los Tuxtlas Palma Sola and Papantla (Fig 2Q testP 005 in all three cases) Female fertility was alsosigniTHORNcantly different among the populations (Fig 3H 1976 df 5 P 0001) The fertility of femalesfrom Los Tuxtlas was higher than fertility from GomezFarotildeas and Palma Sola (Fig 3 Q 36 P 005)Although the fecundity of females belonging to the

Los Tuxtlas-Gomez Farotildeas combination was the lowest(maximum four brood balls per female) their fertilitywas high and not signiTHORNcantly different from that ofthe other populations (80 offspring survival)Chemical Analysis The southern population of Los

Tuxtlas had 48 cuticular compounds Twenty-onecompounds were exclusive to females but only twocompounds were exclusive to males The other 25compounds were shared by individuals of both sexesThe intermediate population (Papantla) had 41 com-pounds nine exclusive to females and three exclusiveto males Both sexes shared 29 compounds The north-

Fig 2 Brood balls produced by Canthon c cyanellus females from THORNve populations located along a latitudinal gradienton the east coast of Mexico (see Fig 1) and brood ball production by females from Gomez Farotildeas mated with males fromLos Tuxtlas (GF-TX) The other abbreviations are Papantla (PP) Palma Sola (PS) and Tancoco (TN) Brood ball productionby females fromtheGFpopulationand fromtheGFETHTXcombinationwas signiTHORNcantly lower than thatof theotherpopulations(DunnOtildes method P 005)

Fig 3 Survival of eclosed beetles from THORNve populations located along a latitudinal gradient on the east coast of Mexicoand beetles eclosed from brood balls produced by females from Gomez Farotildeas mated with males from Los Tuxtlas (GF-TX)Survival was higher in the Los Tuxtlas population than in the Gomez Farotildeas and Palma Sola populations (DunnOtildes methodP 005) Abbreviations as in Fig 2

1252 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 99 no 6

Table 2 Cuticular compounds found in the elytra of Canthon c cyanellus males and females from Los Tuxtlas Papantla and GoacutemezFariacuteas

Retentiontime

Tuxtlasfemale

Tuxtlasmale

Papantlafemale

Papantlamale

Gomez Farotildeasfemale

Gomez Farotildeasmale

Sharedcompounds

1041 014 (086) 006 (015) 004 (080) 51061 003 (028) 046 (333) 017 (018ETH025) 008 (080) 11128 052 (478) 015 (033) 006 (090) 51199 007 (040) 056 (348) 013 (011ETH022) 009 (040) 11229 006 (058) 047 (332)1258 21 (1868)1272 005 (027)1279 004 (042) 004 (023)1309 01 (052) 027 (087)1321 032 (197) 2 (12) 055 (399) 033 (194) 006 (031) 23135 005 (029) 006 (062) 3145 015 (075) 004 (023)1464 042 (423) 116 (695) 055 (058) 021 (212) 002 (190) 04 (130) 121482 028 (177) 019 (095) 006 (056) 61507 008 (034)1529 083 (342) 116 (256) 045 (176) 043 (233) 046 (068) 016 (048) 121567 013 (129) 039 (252)1585 093 (274) 029 (173) 02 (069) 009 (040) 371595 157 (535) 009 (055)1658 007 (074) 008 (030)1679 009 (067)1702 038 (191) 082 (42)1715 581 (074ETH1372) 332 (028ETH1667) 235 (602) 263 (629) 014 (030) 05 (146) 121733 007 (065)1769 141 (051ETH296) 124 (040ETH457) 062 (108) 054 (101) 019 (077) 231789 16 (634)1803 011 (160)1815 263 (679) 053 (320) 036 (254) 152 (445) 009 (050) 231830 55 (136ETH1275 344 (053ETH1358 505 (1729 393 (933) 038 (135) 076 (024ETH217) 121851 323 (838) 061 (368) 076 (305) 148 (511) 007 (034) 231865 06 (299) 144 (566) 001 (060) 81879 184 (545) 092 (441) 31885 256 (1572) 424 (1711) 011 (052) 81892 13 (753) 05 (299)1900 052 (129) 048 (110) 075 (343) 003 (032) 005 (026) 03 (136) 121905 047 (254) 093 (666) 16 (529) 31924 604 (2447) 077 (469) 22 (884) 344 (1547) 361943 054 (227)196 351 (1128) 062 (373) 006 (030)) 104 (017ETH41) 471967 103 (419) 039 (234) 1472 (9678) 1848 (4854) 361979 073 (279) 029 (084) 359 (1128) 174 (473) 016 (079) 037 (172) 122007 84 (2859) 155 (928) 201 (593) 372 (957) 02 (098) 232025 016 (098) 019 (19) 04 (32) 62047 09 (438) 213 (695)2056 029 (287) 039 (233) 034 (092) 09 (349) 472062 97 (3886) 1684 (380)2074 021 (10) 163 (1075) 106 (45) 039 (145) 382084 309 (862) 05 (298) 359 (1396) 782 (1839) 362117 396 (1353)2129 016 (162)2146 102 (553) 179 (529)2167 782 (2483) 151 (978) 056 (247) 72173 67 (3218) 1085 (3574)2183 104 (1044)219 006 (030) 019 (093)2235 459 (1964) 082 (494) 118 (347) 259 (875) 362273 166 (1662)2278 176 (665)2306 061 (312) 162 (669)2323 013 (312) 009 (093)2348 028 (277) 098 (376)2442 1527 (7818) 46 (2760)2576 123 (529)2641 4 (529)

Mean concentration in micrograms per milliliter (minimum and maximum value) Minimum value is 0 except where another value isindicated

Italics indicate compounds exclusive to sex for a given population as follows 1 females from all the populations 2 males from all thepopulations 3 females from Los Tuxtlas and Papantla 4 females from Los Tuxtlas and Gomez Farotildeas 5 females from Papantla and GomezFarotildeas 6 males from Los Tuxtlas and Papantla 7 Males from Los Tuxtlas and Gomez Farotildeas 8 males from Papantla and Gomez Farotildeas

November 2006 ORTIZ-DOMINGUEZ ET AL MATE RECOGNITION IN ALLOPATRIC SCARABS 1253

ern population of Gomez Farotildeas had 24 compoundsNo compounds were exclusive to females whereas 11compounds were exclusive to males Thirteen com-pounds were shared (Table 2)

The comparison among populations showed thatindividuals of Los Tuxtlas had 18 compounds notpresent in individuals of Papantla or Gomez FarotildeasIndividuals of Gomez Farotildeas had only one com-pound shared by both sexes that was not present inthe Los Tuxtlas or Papantla populations Individualsfrom Papantla had 11 compounds not present in theGomez Farotildeas or Los Tuxtlas populations In generalboth sexes from the three populations shared six cu-ticular compounds The chemical analysis showed noquantitative differences in shared compounds be-tween the populations or between the sexes of eachpopulation (Table 2)

Two groups were formed with the Simple MatchingCoefTHORNcient (Fig 4) One group has females from LosTuxtlas The other group has two subgroups one withmales and females from Papantla and the other withmales and females from Gomez Farotildeas that were morerelated to males from Los Tuxtlas (Fig 4)

Discussion

In C c cyanellus the behavior during sexual en-counter and food ball rolling by intrapopulation pairswas mostly cooperative However although most ofthe pairs of individuals belonging to different localitiesalso were able to roll cooperatively noncooperativebehaviors increased signiTHORNcantly with distance sug-gesting that individuals prefer similar mates to matesfrom other populations This was very clear in maleETHfemale pairs of individuals from 600 km apart thatwere unwilling to roll a food ball with a partner eitherkeeping it away from the partner or THORNghting insteadof rolling the food ball cooperatively for nestingThese noncooperative behaviors are similar to those

observed between individuals of the same sex (Favila1988a) but in this case they act as behavioral precop-ulatory barriers The increased conszligict during sexualencounters between males and females from distantpopulations suggests that an incipient process of spe-ciation is occurring in C c cyanellus

Reproductive success was low in females fromGomez Farotildeas and in Gomez Farotildeas-Los Tuxtlas com-binations Pairs were fed THORNsh because this type offood has been successfully used in our C c cyanellusbreeding program (Favila 1993) However when pairsfrom Gomez Farotildeas and those from the Gomez Farotildeas-Los Tuxtlas combinations were feeding we observedthat after two to eight brood balls had been made thepair did not use the food for nesting thus reducingtheir fecundity In general C c cyanellus breed usingsmall vertebrate corpses (Halffter et al 1983 Favila andDotildeaz 1996) However Villalobos et al (1998) foundthat individuals from Gomez Farotildeas are able to eat andbreed by using carcasses ofOrthoporus ornatosGirard(Spirostreptida Spirostreptidae) a common diplopodin this region This behavior together with our resultssuggests that individuals from Gomez Farotildeas may havebecome very specialized in eating and nesting withDiplopoda as a local adaptation However althoughfecundity was low for the females from Gomez Farotildeasand the Gomez Farotildeas-Los Tuxtlas combinations off-spring survival was high so there is no evidence ofpostmating isolation in the interpopulation crossesUnfortunately the low number of eclosed beetles ob-tained in these combinations did not allow us to crossthe F2 generation

According to the ecological theory of speciationreproductive isolation ultimately evolves as a conse-quence of divergent natural selection on traits be-tween environments (Schluter 2001) If intrademicmate recognition systems are similar to species rec-ognition systems divergence in signaling systems canlead to premating isolation when geographically iso-

Fig 4 Tree representing cluster analysis for Canthon c cyanellus males and females in three populations from the eastcoast of Mexico on the basis of cuticular compounds TX Tuxtlas PP Papantla GF Gomez Farotildeas M male F female

1254 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 99 no 6

lated populations have secondary contact (Blows andAllan 1998 Greenberg et al 2003 Nosil 2004) Forexample hydrocarbons are involved in the mate rec-ognition system of Drosophila mojavensis Paterson(Diptera Drosophilidae) (Etges 1998) but this spe-cies shows regional variation associated with switch-ing hosts This has led to premating isolation amonggeographically isolated populations (Etges and Ahrens2001) In C c cyanellus the most important signalingsystem during sexual recognition is chemical via cu-ticular compounds (Ortiz-Domotildenguez et al 2006)Males and females from Gomez Farotildeas had the lowestnumber of cuticular compounds and were most ag-gressive with partners from the most distant popula-tion (Los Tuxtlas) which had the highest number ofcompounds The reduction in compounds could limitthe beetleOtildes ability to recognize individuals of theopposite sex from Los Tuxtlas which have the mostcomplex blend of cuticular compounds We hypoth-esize that switching in food preferences from car-casses of small vertebrates to Diplopoda carcasses inindividuals from the Gomez Farotildeas population couldmodify the quality and quantity of their cuticular com-pounds and produce the difTHORNculties observed in ac-cepting individuals from Los Tuxtlas and Palma SolaBecause Canthon c cyanellusmales recognize femalesby odor but females recognize individuals of the op-posite sex by other unknown behavioral cues as well(Ortotildez-Dominguez et al 2006) we do not discard thepossibility that another system of signals also could beinvolved in sexual recognition and these signals mayhave affected our results

Solotildes and Kohlmann (2002) suggest that allCanthoncyanellus subspecies from South America to Texasshould disappear However our results show that forthe subspecies C c cyanellus prezygotic isolationcould occur between distant populations when thereis secondary contact Owing to the ample distributionof C cyanellus in the Americas it is possible that thesouthern populations will have greater difTHORNculties insexual recognition if they come into contact withnorthern populations So we think that at presentit is not convenient to eliminate the subspecies ofC cyanellus but rather to reanalyze them in the con-text of Avise and Ball (1990) who consider subspeciesas members who share a unique geographic locale aset of phylogenetically concordant phenotypic char-acters and a unique natural history relative to othersubdivisions of the species but that are reproductivelycompatible with other such groups Phylogenetic dis-tinction must be based on the distributions of severalindependent genetically based traits (Miththapala etal 1996) not only on morphological traits as has beendone with C cyanellus

In summary the experiments here presented sug-gest that premating isolation is result of local adapta-tion related to a switch in food preferences If that isthe case future studies should analyze the factorsrelated with this switch Comparison with other south-ern populations will be crucial to determiningwhether this species is still cohesive or is in a processof THORNssion or whether there are even a complex of

cryptic species that are difTHORNcult to distinguish mor-phologically

Acknowledgments

We are grateful to Ingrid Marquez and Bianca Delfosse forlinguistic corrections and also to Carlos Fragoso for helpfulcomments and suggestions on the manuscript This study wassupported by Consejo Nacional de Ciencia y Tecnologia-Mexico Grants 4161P-N9607 35125-V and 49472-Q

References Cited

Avise J C and R M Ball 1990 Principles of genealogicalconcordance in species concepts and biological taxon-omy Oxford Surv Evol Biol 7 45ETH67

Blows M W and R A Allan 1998 Levels of mate recog-nition within and between two Drosophila species andtheir hybrids Am Nat 152 826ETH837

Boake CRB D K Price and D K Andreadis 1998 In-heritance of behavioral differences between two infer-tile sympatric species Drosophila silvestris and D het-eroneura Heredity 80 642ETH650

Del Campo M L S Via and M C Caillaud 2003 Rec-ognition of host-speciTHORNc chemical stimulants in two sym-patric host races of the pea aphid Acyrthosiphon pisumEcol Entomol 28 397ETH404

Etges W J 1998 Premating isolation is determined by lar-val-rearing substrates in cactophilic Drosophila mojaven-sis IV Correlated response in behavioral isolation toartiTHORNcial selection on a life-history trait Am Nat 152129ETH144

Etges W J 2002 Divergence in mate choice systems doesevolution play by rules Genetica 116 151ETH166

Etges W J and M A Ahrens 2001 Premating isolation isdetermined by larval-rearing substrates in cactophilicDrosophila mojavensis V Deep geographic variation inepicuticular hydrocarbons among isolated populationsAm Nat 158 585ETH598

Favila M E 1988a Comportamiento durante el periodode maduracion gonadica en un escarabajo rodador(Coleoptera Scarabaeidae Scarabaeinae) Folia EntomolMex 76 55ETH64

Favila M E 1988b Un metodo sencillo para marcar escar-abajos Folia Entomol Mex 75 117ETH118

Favila M E 1993 Some ecological factors affecting thelife-style of Canthon cyanellus cyanellus (ColeopteraScarabaeidae) an experimental approach Ethol EcolEvol 5 319ETH328

FavilaM E andADıaz 1996 Canthon cyanellus cyanellusLeConte (Coleoptera Scarabaeidae) makes a nest in theTHORNeld with several brood balls Coleopt Bull 50 52ETH60

GreenbergA J J RMoran J ACoyne andC IWu 2003Ecological adaptation during incipient speciation re-vealed by precise gene replacement Science (WashDC) 302 1754ETH1757

Halffter G 1961 Monografotildea de las especies norteameri-canas del genero Canthon Hoffsg (Coleopt Scarab)Ciencia 20 225ETH320

Halffter G M E Favila and V Halffter 1992 A compar-ative study of the structure of the scarab guild in Mexicantropical rain forest and derived ecosystems Folia Ento-mol Mex 84 131ETH156

Halffter G V Halffter and C Huerta 1983 Comporte-ment sexual et nidiTHORNcation chez Canthon cyanellus cya-nellus LeConte Bull Soc Entomol Mex 88 585ETH594

November 2006 ORTIZ-DOMINGUEZ ET AL MATE RECOGNITION IN ALLOPATRIC SCARABS 1255

Mendelson T C 2003 Sexual isolation evolves faster thanhybrid inviability in a diverse and sexually domorphicgenus of THORNsh (Percidae Etheostoma) Evolution 57 317ETH327

Michalak P JGrzesik and J Rafinski 1997 Test for sexualincompatibility between two new species Triturus vul-garis and Triturus montandoni no choice mating designEvolution 51 2045ETH2050

MillerGLGE StrattonPRMiller andEHebets 1998Geographical variation in male courtship behavior andsexual isolation in wolf spiders of the genus SchizocosaAnim Behav 56 937ETH951

Miththapala S J Seidensticker and S J OrsquoBrien 1996Phylogeographic subspecies recognition in leopards(Panthera pardus) molecular genetic variation ConservBiol 10 1115ETH1132

Noor MAF M A Williams D Alvarez and M Ruiz-Garcıa 2000 Lack of evolution divergence in courtshipsongs of Drosophila pseudoobscura subspecies J InsectBehav 13 255ETH262

Nosil P 2004 Reproductive isolation caused by visual pre-dation on migrants between divergent environmentsProc R Soc Lond B 271 1521ETH1528

Ortız-Domınguez M M E Favila M R Mendoza-LopezO Garcıa-Barradas and J Cruz-Sanchez 2006 The re-lationship between epicuticular compounds behavioralcues and sexual recognition in the ball roller scarabCanthon c cyanellus Entomol Exp Appl 119 23ETH27

Paillette M N Bizat and D Joly 1997 Differentiation ofdialects and courtship strategies in allopatric populationsof Drosophila teissieri J Insect Physiol 43 809ETH814

Panhuis T M R Butlin M Zuk and T Tregenza 2001Sexual selection and speciation Trends Ecol Evol 16364ETH371

Ritchie M G E J Halsey and M G Gleason 1999 Dro-sophila song as a species-speciTHORNc mating signal and thebehavioral importance of Kyriacou and Hall cycles inD melanogaster song Anim Behav 58 649ETH657

Rybak F G Sureau and T Aubin 2002 Functional cou-pling of acoustic and chemical signals in the courtshipbehavior of the male Drosophila melanogaster Proc RSoc Lond B 269 695ETH701

Schluter D 2000 The ecology of adaptive radiationOxford University Press Oxford United Kingdom

Schluter D 2001 Ecology and the origin of the speciesTrends Ecol Evol 16 372ETH380

Sokal R R and F J Rohlf 1981 Biometry 2nd ed W HFreeman San Francisco CA

Solıs A and B Kohlmann 2002 El genero Canthon (Co-leoptera Scarabaeidae) en Costa Rica G Ital Entomol10 11ETH35

Villalobos F J ADiaz andMEFavila 1998 Two speciesof Canthon Hoffmannsegg (Coleoptera Scarabaeidae)feed on dead and live invertebrates Coleopt Bull 52101ETH104

Vines T H and D Schluter 2006 Strong assortative mat-ing between allopatric sticklebacks as a by-product ofadaptation to different environments Proc Soc B 273911ETH916

Received 7 December 2005 accepted 2 August 2006

1256 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 99 no 6

September the rainy season (Halffter et al 1983Favila and Dotildeaz 1996 Villalobos et al 1998) Individ-uals locate vertebrate carcasses by their odor andinitial sexual interactions occur at the carcass Sexualrecognition is mostly mediated by cuticular com-pounds putatively hydrocarbons (Ortiz-Domotildenguezet al 2006) When same-sex or opposite-sex individ-uals meet they make head-to-head contact extendingtheir antennae and maxillary palpi or even touch thebody of the other individual with the antennae Same-sex individuals THORNght THORNercely for the food ball but if themeeting is between a male and a female cooperativelythey will cut and roll a food ball for nesting (Favila1988a Favila and Dotildeaz 1996) MEF (unpublisheddata) has observed the behavior of males and femalescollected in tropical forests from the east coast ofMexico but from sites 600 km apart maleETHfemale pairsfrom the same population cooperatively rolled a foodball but the female from the northern population(Gomez Farotildeas Mexico) always refused the male fromthe southern population (Los Tuxtlas Mexico) Suchobservations suggested that there was either difTHORNcultyin the sexual recognition of or discrimination againstgeographically separated potential mates among indi-viduals from distant populations

In this study we analyze premating behavior duringsexual encounters and food ball rolling of male andfemale pairs of individuals belonging to THORNve separatedpopulations of Canthon c cyanellus found over a lat-itudinal gradient on the eastern coastal plains of Mex-

icoBecause sexual recognition in this species ismostlymediated by cuticular compounds cuticular chemicalcomposition was compared among the populations Toevaluate potential postmating reproductive isolationreproductive success (measured as brood ball produc-tion by the female and survival of eclosed beetles) wascompared in maleETHfemale pairs belonging to the samepopulation and for pairs in which each sex belongedto a different population

Materials and Methods

The accelerated process of deforestation in Mexicohas produced a landscape dominated by pastureswhere forest fragments are maintained to provideshade for cattle This is particularly true of the easterncoastal plain where the original continuous distribu-tion of C c cyanellus on this plain is now mostlyrestricted to these forest fragments (Halffter et al1992)

To adequately represent the separate populationsfrom along a latitudinal gradient on the eastern coastalplains of Mexico pitfall traps were baited with THORNsh andset in the soil to capture carrion beetles at the follow-ing localities (Fig 1) Los Tuxtlas Veracruz (18 35 N95 04 W) Palma Sola Veracruz (19 46 N 95 26 W)Papantla Veracruz (20 25 N 97 27 W) TancocoVeracruz (21 19 N 97 50 W) and Gomez FarotildeasTampaulipas (23 03 N 99 00 W) Beetles werealways collected from within the forest although they

Fig 1 Location of the THORNve allopatric populations of Canthon c cyanellus in Mexico

November 2006 ORTIZ-DOMINGUEZ ET AL MATE RECOGNITION IN ALLOPATRIC SCARABS 1249

are also found in pastures near the fragments (Halffteret al 1992) The THORNrst locality was in the OcircLos TuxtlasOtildebiological research station (National AutonomousUniversity of Mexico) located within the Los TuxtlasBiosphere Reserve and the last site is located in the OcircElCieloOtilde Biosphere Reserve (University of Tamaulipas)There are visible differences between some popu-lations in the morphology of their cuticular colorand texture We focused our attention on elytral vari-ation in three of the populations analyzed specimensfrom Los Tuxtlas the southern site have shiny greenelytra with a smooth interpunctual surface and feeblypunctuate elytral striae The elytra of the Papantlaspecimens the intermediate site are blue or mattegreen but the interpunctual surface is generally THORNneshagreening with punctuate elytral striae The elytraof the Gomez Farotildeas specimens the northern site areshiny green with a smooth interpunctual surface andpunctuate elytral striae Specimens from each popu-lation were put in containers with moist soil in arearing room at 27C 70 10 RH and a photoperiodof 1212 (LD) h We followed the rearing protocol ofFavila (1993)

Experiments were conducted in another insec-tarium under the same conditions as those of therearing room Because it is not possible to distinguishthe sex of these beetles with the naked eye sex iden-tiTHORNcation was done beforehand under the microscopeThen in each experiment one elytrum of one memberof the pair was marked with a small spot of latex paint(known to not affect behavior Favila 1988b) Exper-

iments were conducted from 0900 to 1300 hours themost active rolling period of the day for this species(Favila and Dotildeaz 1996)Intra- and Intersite Sexual Recognition MaleETHfe-

male pairs were randomly selected from each site andeach one pair was placed in an observation arena (10-by 2-cm petri dish lined with THORNlter paper and a 2 g ofTHORNsh meat ball) As we were interested in observing thebehavior of an individual of each sex in the presenceof an individual of the opposite sex from the same oranother site we made 25 combinations (male frompopulation A with female from population B and viceversa male from C and vice versa and so on) includ-ing intra- and interpopulation combinations (unequaln for each combination is indicated in Table 1) Timeuntil interaction was measured as the elapsed time (inseconds) between introducing the pair into the arenaand the beginning of maleETHfemale interaction on thefood ball Time to rolling was measured as the numberof seconds that elapsed between the beginning of themaleETHfemale pair interaction on the food ball and ballrolling by one or both partners The sex of the indi-vidual that rolled the food ball was recorded Weconsidered the rolling of the food ball by both indi-viduals as cooperative behavior but always during thisjoint rolling one of the partners usually the male rollsthe ball while the other partner is transported on theball The frequency of the following noncooperativebehaviors also was analyzed 1) avoidance when oneindividual rolled the food ball alone and did not allowthe other member to climb on the ball 2) pushing

Table 1 Latency in initial interaction and rolling and proportion of noncooperative behavior (individual rolling of the food ball andfights) observed between malendashfemale pairs of Canthon c cyanellus from five populations on the east coast of Mexico and in theircombinations

Male pop Female popDistance

(km)n

Time to interaction(25ETH75)

Time to rolling(25ETH75)

Noncooperativebehavior ()

Papantla Papantla 0 30 66 (3225ETH165) 14 (60ETH27)a 0Palma Sola Palma Sola 0 29 39 (2825ETH11175) 12 (675ETH285)bc 3Gomez Farotildeas Gomez Farotildeas 0 26 33 (16ETH75) 26 (65ETH405) 0Tancoco Tancoco 0 26 46 (25ETH65) 14 (60ETH40) 0Tuxtlas Tuxtlas 0 30 38 (15ETH92) 255 (100ETH50) 3Papantla Tancoco 107 28 605 (275ETH81) 23 (1125ETH10625) 4Tancoco Papantla 107 30 505 (19ETH123) 295 (60ETH45) 7Papantla Palma Sola 122 31 365 (20ETH93) 20 (975ETH505) 13Palma Sola Papantla 122 30 46 (23ETH13275) 24 (1425ETH3675) 17Palma Sola Tuxtlas 195 23 40 (2425ETH7875) 20 (95ETH51) 9Tuxtlas Palma Sola 195 15 40 (15ETH745) 15 (1075ETH2525) 7Gomez Farotildeas Tancoco 223 19 705 (14ETH94) 21 (85ETH7125) 5Tancoco Gomez Farotildeas 223 20 74 (35ETH93) 28 (10ETH565) 0Palma Sola Tancoco 228 30 65 (35ETH127) 615 (25ETH144)ac 10Tancoco Palma Sola 228 30 595 (38ETH146) 325 (13ETH60) 23Papantla Tuxtlas 317 29 67 (1475ETH9425) 345 (14ETH93) 14Tuxtlas Papantla 317 30 58 (36ETH75) 15 (65ETH3875) 17Papantla Gomez Farotildeas 331 26 60 (2875ETH805) 56 (1875ETH12075)b 19Gomez Farotildeas Papantla 331 26 41 (24ETH925) 385 (12ETH785) 19Tancoco Tuxtlas 424 25 116 (45ETH18925) 24 (165ETH155) 20Tuxtlas Tancoco 424 27 81 (3525ETH117) 26 (1425ETH4975) 4Palma Sola Gomez Farotildeas 447 31 36 (20ETH6475) 28 (105ETH6725) 6Gomez Farotildeas Palma Sola 447 20 99 (3575ETH16625) 28 (115ETH385) 70a

Gomez Farotildeas Tuxtlas 643 31 53 (2525ETH1075) 345 (15ETH124) 55a

Tuxtlas Gomez Farotildeas 643 24 575 (375ETH715) 32 (1225ETH9825) 50a

Similar letters indicate signiTHORNcant differences between maleETHfemale pair combinations according to DunnOtildes test (P 005) Q values a 44 b 39 and c 47a G 10947 df 24 P 005

1250 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 99 no 6

when one individual attempted to climb on the foodball and the other pushed it back by ldquobuttingrdquo it and3) THORNghts both male and female fought for the ball onsome occasions cutting it so that each beetle rolled anindependent fragment of food Once the pair hadmade contact on the food ball the total observationtime lasted 5 min An analysis of variance (ANOVA)was used to analyze time until interaction and rollingoccurred both among populations and for combina-tions The frequency of each noncooperative interac-tion was analyzed with the G test (Sokal and Rohlf1981)Reproductive Success among Populations We did

not know the age and reproductive condition of spec-imens collected in the THORNeld Therefore we allowedpairs of the F1 of the laboratory strains originallycollected from the THORNve sites and more distant combi-nations (where noncooperative behaviors were morefrequently observed see Results) to mate MaleETHfe-male pairs were 20ETH30 d old the optimal age fornesting in this species (Favila 1993) Each maleETHfe-male pair was placed in a plastic nesting box (9 by 9by 8 cm) containing 25 cm of moist soil and a pieceof THORNsh suitable for reproduction (1 g) Pairs were fedTHORNsh every other day and brood ball production wasfollowed for 40 d the most active period in the re-productive life of a female under laboratory condi-tions After this period females tend to produce broodballs sporadically and offspring survival is low (Favila1993) We measured the elapsed time between theintroduction of the pair to the nesting box and theconstruction of the THORNrst brood ball The total numberof brood balls produced by a female was considered ameasure of fecundity and the number of beetleseclosed from the brood balls as a measure of fertilityEvery variable was analyzed using a KruskalETHWallistestChemical Analysis Elytra from THORNve males and THORNve

females of the populations that were furthest apart(Gomez Farotildeas and Los Tuxtlas) and one intermediatepopulation (Papantla) were individually put intoWheaton vials containing 1 ml of hexane (HPLCgrade Sigma-Aldrich St Louis MO) Every vial wasgently shaken on a shaker table for 1 min and left at 4Cfor 24 h Next the elytra were removed from thevials and the solvent was evaporated using N2 Then50 l of hexane was added to each vial along withpentadecane (n-C15) as an internal standard A 2-lsample was injected into a HewlettETHPackard G1800BGC equipped with a fused silica capillary column(HP-5) cross-linked with 5 methylsiloxane (025mm 30 m) and 025-m THORNlm thickness supplied byAgilent (Mexico DF) with helium as the carrier gasat a szligow rate of 1 mlmin We optimized the separa-tion of extracted compounds by using a column tem-perature proTHORNle in which the analysis began at 70Cwas increased 20Cmin to 190C increased again10Cmin to 250C and then increased 30Cmin to280C and that temperature was held for 10 minWith the presenceETHabsence matrix of compounds weapplied the simple matching coefTHORNcient The relation-ship among individuals belonging to the three popu-

lations was represented using average linkage clusteranalysis (unweighted pair-group method with arith-metic average)

Results

Intra- and Interpopulation Sexual RecognitionWedid not THORNnd signiTHORNcant differences in the time untilinteraction occurred among the maleETHfemale com-binations (F 124 df 24 P 020) Howevertime to rolling was signiTHORNcantly different amongtreatments (Table 1 H 5636 df 24 P 0001)Pairs from Palma Sola began to roll the food ball fasterthan pairs belonging to the Papantla-Gomez Farotildeasand Palma Sola-Tancoco combinations Pairs fromPapantla began to roll faster than pairs from thePalma Sola-Tancoco combination MaleETHfemale pairsfrom the same populations always showed joint rollingbehavior only for one pair from Los Tuxtlas and onepair from Palma Sola was pushing (a noncooperativebehavior) observed during joint rolling (Table 1) inboth cases the female pushed the male

Joint rolling was observed in all the pairs from in-terpopulation combinations but there was also non-cooperative behavior in 19 of these combinationsThere was a clear correlation between the frequencyof noncooperative behavior and the geographic dis-tance between populations (Pearson coefTHORNcient r 0727 P 0001) However only in three of thesecombinations (male from Gomez Farotildeasfemale fromPalma Sola male from Gomez Farotildeasfemale from LosTuxtlas and male from Los Tuxtlasfemale fromGomez Farotildeas) was the frequency of noncooperativebehaviors signiTHORNcantly higher than in pairs from theother combinations (Table 1G 10947 df 24 P005) Individuals in these combinations belonged tothe most distant populations For the combinationwith the male from Gomez Farotildeas and the female fromPalma Sola in 45 of the pairs formed females rolledthe food ball alone preventing the males from climb-ing onto the food ball by pushing them off with theirheads In 25 of these combinations THORNghts for the foodball were observed However for the inverse combi-nation (male from Palma Solafemale from GomezFarotildeas) joint rolling was frequently observed (Table1) except in two pairs (6) where the females rolledthe ball alone and prevented the males from climbingonto the food ball In contrast for both combinationsfrom Los Tuxtlas and Gomez Farotildeas a signiTHORNcant num-ber of noncooperative behaviors were observed Inthe male from Gomez Farotildeasfemale from Los Tuxtlascombination males exhibited signiTHORNcantly more non-cooperative behavior than females (67 versus 8 re-spectively Fisher test 005) However in the malefrom Los Tuxtlasfemale from Gomez Farotildeas combi-nation 70 of the females exhibited noncooperativebehavior with males (Fisher test 002) whereasmales did not show any noncooperative behavior Ingeneral individuals from Gomez Farotildeas showed morenoncooperative behaviors toward individuals fromthe opposite sex of Los Tuxtlas than vice versa (69 and5 respectively Fisher test 00015)

November 2006 ORTIZ-DOMINGUEZ ET AL MATE RECOGNITION IN ALLOPATRIC SCARABS 1251

Reproductive Success among Populations Fecun-dity was different among populations (Fig 2 H 2177 df 5 P 005) The fecundity of femalesfrom Gomez Farotildeas and that of the Los Tuxtlas-GomezFarotildeas combination was signiTHORNcantly lower than thoseof Los Tuxtlas Palma Sola and Papantla (Fig 2Q testP 005 in all three cases) Female fertility was alsosigniTHORNcantly different among the populations (Fig 3H 1976 df 5 P 0001) The fertility of femalesfrom Los Tuxtlas was higher than fertility from GomezFarotildeas and Palma Sola (Fig 3 Q 36 P 005)Although the fecundity of females belonging to the

Los Tuxtlas-Gomez Farotildeas combination was the lowest(maximum four brood balls per female) their fertilitywas high and not signiTHORNcantly different from that ofthe other populations (80 offspring survival)Chemical Analysis The southern population of Los

Tuxtlas had 48 cuticular compounds Twenty-onecompounds were exclusive to females but only twocompounds were exclusive to males The other 25compounds were shared by individuals of both sexesThe intermediate population (Papantla) had 41 com-pounds nine exclusive to females and three exclusiveto males Both sexes shared 29 compounds The north-

Fig 2 Brood balls produced by Canthon c cyanellus females from THORNve populations located along a latitudinal gradienton the east coast of Mexico (see Fig 1) and brood ball production by females from Gomez Farotildeas mated with males fromLos Tuxtlas (GF-TX) The other abbreviations are Papantla (PP) Palma Sola (PS) and Tancoco (TN) Brood ball productionby females fromtheGFpopulationand fromtheGFETHTXcombinationwas signiTHORNcantly lower than thatof theotherpopulations(DunnOtildes method P 005)

Fig 3 Survival of eclosed beetles from THORNve populations located along a latitudinal gradient on the east coast of Mexicoand beetles eclosed from brood balls produced by females from Gomez Farotildeas mated with males from Los Tuxtlas (GF-TX)Survival was higher in the Los Tuxtlas population than in the Gomez Farotildeas and Palma Sola populations (DunnOtildes methodP 005) Abbreviations as in Fig 2

1252 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 99 no 6

Table 2 Cuticular compounds found in the elytra of Canthon c cyanellus males and females from Los Tuxtlas Papantla and GoacutemezFariacuteas

Retentiontime

Tuxtlasfemale

Tuxtlasmale

Papantlafemale

Papantlamale

Gomez Farotildeasfemale

Gomez Farotildeasmale

Sharedcompounds

1041 014 (086) 006 (015) 004 (080) 51061 003 (028) 046 (333) 017 (018ETH025) 008 (080) 11128 052 (478) 015 (033) 006 (090) 51199 007 (040) 056 (348) 013 (011ETH022) 009 (040) 11229 006 (058) 047 (332)1258 21 (1868)1272 005 (027)1279 004 (042) 004 (023)1309 01 (052) 027 (087)1321 032 (197) 2 (12) 055 (399) 033 (194) 006 (031) 23135 005 (029) 006 (062) 3145 015 (075) 004 (023)1464 042 (423) 116 (695) 055 (058) 021 (212) 002 (190) 04 (130) 121482 028 (177) 019 (095) 006 (056) 61507 008 (034)1529 083 (342) 116 (256) 045 (176) 043 (233) 046 (068) 016 (048) 121567 013 (129) 039 (252)1585 093 (274) 029 (173) 02 (069) 009 (040) 371595 157 (535) 009 (055)1658 007 (074) 008 (030)1679 009 (067)1702 038 (191) 082 (42)1715 581 (074ETH1372) 332 (028ETH1667) 235 (602) 263 (629) 014 (030) 05 (146) 121733 007 (065)1769 141 (051ETH296) 124 (040ETH457) 062 (108) 054 (101) 019 (077) 231789 16 (634)1803 011 (160)1815 263 (679) 053 (320) 036 (254) 152 (445) 009 (050) 231830 55 (136ETH1275 344 (053ETH1358 505 (1729 393 (933) 038 (135) 076 (024ETH217) 121851 323 (838) 061 (368) 076 (305) 148 (511) 007 (034) 231865 06 (299) 144 (566) 001 (060) 81879 184 (545) 092 (441) 31885 256 (1572) 424 (1711) 011 (052) 81892 13 (753) 05 (299)1900 052 (129) 048 (110) 075 (343) 003 (032) 005 (026) 03 (136) 121905 047 (254) 093 (666) 16 (529) 31924 604 (2447) 077 (469) 22 (884) 344 (1547) 361943 054 (227)196 351 (1128) 062 (373) 006 (030)) 104 (017ETH41) 471967 103 (419) 039 (234) 1472 (9678) 1848 (4854) 361979 073 (279) 029 (084) 359 (1128) 174 (473) 016 (079) 037 (172) 122007 84 (2859) 155 (928) 201 (593) 372 (957) 02 (098) 232025 016 (098) 019 (19) 04 (32) 62047 09 (438) 213 (695)2056 029 (287) 039 (233) 034 (092) 09 (349) 472062 97 (3886) 1684 (380)2074 021 (10) 163 (1075) 106 (45) 039 (145) 382084 309 (862) 05 (298) 359 (1396) 782 (1839) 362117 396 (1353)2129 016 (162)2146 102 (553) 179 (529)2167 782 (2483) 151 (978) 056 (247) 72173 67 (3218) 1085 (3574)2183 104 (1044)219 006 (030) 019 (093)2235 459 (1964) 082 (494) 118 (347) 259 (875) 362273 166 (1662)2278 176 (665)2306 061 (312) 162 (669)2323 013 (312) 009 (093)2348 028 (277) 098 (376)2442 1527 (7818) 46 (2760)2576 123 (529)2641 4 (529)

Mean concentration in micrograms per milliliter (minimum and maximum value) Minimum value is 0 except where another value isindicated

Italics indicate compounds exclusive to sex for a given population as follows 1 females from all the populations 2 males from all thepopulations 3 females from Los Tuxtlas and Papantla 4 females from Los Tuxtlas and Gomez Farotildeas 5 females from Papantla and GomezFarotildeas 6 males from Los Tuxtlas and Papantla 7 Males from Los Tuxtlas and Gomez Farotildeas 8 males from Papantla and Gomez Farotildeas

November 2006 ORTIZ-DOMINGUEZ ET AL MATE RECOGNITION IN ALLOPATRIC SCARABS 1253

ern population of Gomez Farotildeas had 24 compoundsNo compounds were exclusive to females whereas 11compounds were exclusive to males Thirteen com-pounds were shared (Table 2)

The comparison among populations showed thatindividuals of Los Tuxtlas had 18 compounds notpresent in individuals of Papantla or Gomez FarotildeasIndividuals of Gomez Farotildeas had only one com-pound shared by both sexes that was not present inthe Los Tuxtlas or Papantla populations Individualsfrom Papantla had 11 compounds not present in theGomez Farotildeas or Los Tuxtlas populations In generalboth sexes from the three populations shared six cu-ticular compounds The chemical analysis showed noquantitative differences in shared compounds be-tween the populations or between the sexes of eachpopulation (Table 2)

Two groups were formed with the Simple MatchingCoefTHORNcient (Fig 4) One group has females from LosTuxtlas The other group has two subgroups one withmales and females from Papantla and the other withmales and females from Gomez Farotildeas that were morerelated to males from Los Tuxtlas (Fig 4)

Discussion

In C c cyanellus the behavior during sexual en-counter and food ball rolling by intrapopulation pairswas mostly cooperative However although most ofthe pairs of individuals belonging to different localitiesalso were able to roll cooperatively noncooperativebehaviors increased signiTHORNcantly with distance sug-gesting that individuals prefer similar mates to matesfrom other populations This was very clear in maleETHfemale pairs of individuals from 600 km apart thatwere unwilling to roll a food ball with a partner eitherkeeping it away from the partner or THORNghting insteadof rolling the food ball cooperatively for nestingThese noncooperative behaviors are similar to those

observed between individuals of the same sex (Favila1988a) but in this case they act as behavioral precop-ulatory barriers The increased conszligict during sexualencounters between males and females from distantpopulations suggests that an incipient process of spe-ciation is occurring in C c cyanellus

Reproductive success was low in females fromGomez Farotildeas and in Gomez Farotildeas-Los Tuxtlas com-binations Pairs were fed THORNsh because this type offood has been successfully used in our C c cyanellusbreeding program (Favila 1993) However when pairsfrom Gomez Farotildeas and those from the Gomez Farotildeas-Los Tuxtlas combinations were feeding we observedthat after two to eight brood balls had been made thepair did not use the food for nesting thus reducingtheir fecundity In general C c cyanellus breed usingsmall vertebrate corpses (Halffter et al 1983 Favila andDotildeaz 1996) However Villalobos et al (1998) foundthat individuals from Gomez Farotildeas are able to eat andbreed by using carcasses ofOrthoporus ornatosGirard(Spirostreptida Spirostreptidae) a common diplopodin this region This behavior together with our resultssuggests that individuals from Gomez Farotildeas may havebecome very specialized in eating and nesting withDiplopoda as a local adaptation However althoughfecundity was low for the females from Gomez Farotildeasand the Gomez Farotildeas-Los Tuxtlas combinations off-spring survival was high so there is no evidence ofpostmating isolation in the interpopulation crossesUnfortunately the low number of eclosed beetles ob-tained in these combinations did not allow us to crossthe F2 generation

According to the ecological theory of speciationreproductive isolation ultimately evolves as a conse-quence of divergent natural selection on traits be-tween environments (Schluter 2001) If intrademicmate recognition systems are similar to species rec-ognition systems divergence in signaling systems canlead to premating isolation when geographically iso-

Fig 4 Tree representing cluster analysis for Canthon c cyanellus males and females in three populations from the eastcoast of Mexico on the basis of cuticular compounds TX Tuxtlas PP Papantla GF Gomez Farotildeas M male F female

1254 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 99 no 6

lated populations have secondary contact (Blows andAllan 1998 Greenberg et al 2003 Nosil 2004) Forexample hydrocarbons are involved in the mate rec-ognition system of Drosophila mojavensis Paterson(Diptera Drosophilidae) (Etges 1998) but this spe-cies shows regional variation associated with switch-ing hosts This has led to premating isolation amonggeographically isolated populations (Etges and Ahrens2001) In C c cyanellus the most important signalingsystem during sexual recognition is chemical via cu-ticular compounds (Ortiz-Domotildenguez et al 2006)Males and females from Gomez Farotildeas had the lowestnumber of cuticular compounds and were most ag-gressive with partners from the most distant popula-tion (Los Tuxtlas) which had the highest number ofcompounds The reduction in compounds could limitthe beetleOtildes ability to recognize individuals of theopposite sex from Los Tuxtlas which have the mostcomplex blend of cuticular compounds We hypoth-esize that switching in food preferences from car-casses of small vertebrates to Diplopoda carcasses inindividuals from the Gomez Farotildeas population couldmodify the quality and quantity of their cuticular com-pounds and produce the difTHORNculties observed in ac-cepting individuals from Los Tuxtlas and Palma SolaBecause Canthon c cyanellusmales recognize femalesby odor but females recognize individuals of the op-posite sex by other unknown behavioral cues as well(Ortotildez-Dominguez et al 2006) we do not discard thepossibility that another system of signals also could beinvolved in sexual recognition and these signals mayhave affected our results

Solotildes and Kohlmann (2002) suggest that allCanthoncyanellus subspecies from South America to Texasshould disappear However our results show that forthe subspecies C c cyanellus prezygotic isolationcould occur between distant populations when thereis secondary contact Owing to the ample distributionof C cyanellus in the Americas it is possible that thesouthern populations will have greater difTHORNculties insexual recognition if they come into contact withnorthern populations So we think that at presentit is not convenient to eliminate the subspecies ofC cyanellus but rather to reanalyze them in the con-text of Avise and Ball (1990) who consider subspeciesas members who share a unique geographic locale aset of phylogenetically concordant phenotypic char-acters and a unique natural history relative to othersubdivisions of the species but that are reproductivelycompatible with other such groups Phylogenetic dis-tinction must be based on the distributions of severalindependent genetically based traits (Miththapala etal 1996) not only on morphological traits as has beendone with C cyanellus

In summary the experiments here presented sug-gest that premating isolation is result of local adapta-tion related to a switch in food preferences If that isthe case future studies should analyze the factorsrelated with this switch Comparison with other south-ern populations will be crucial to determiningwhether this species is still cohesive or is in a processof THORNssion or whether there are even a complex of

cryptic species that are difTHORNcult to distinguish mor-phologically

Acknowledgments

We are grateful to Ingrid Marquez and Bianca Delfosse forlinguistic corrections and also to Carlos Fragoso for helpfulcomments and suggestions on the manuscript This study wassupported by Consejo Nacional de Ciencia y Tecnologia-Mexico Grants 4161P-N9607 35125-V and 49472-Q

References Cited

Avise J C and R M Ball 1990 Principles of genealogicalconcordance in species concepts and biological taxon-omy Oxford Surv Evol Biol 7 45ETH67

Blows M W and R A Allan 1998 Levels of mate recog-nition within and between two Drosophila species andtheir hybrids Am Nat 152 826ETH837

Boake CRB D K Price and D K Andreadis 1998 In-heritance of behavioral differences between two infer-tile sympatric species Drosophila silvestris and D het-eroneura Heredity 80 642ETH650

Del Campo M L S Via and M C Caillaud 2003 Rec-ognition of host-speciTHORNc chemical stimulants in two sym-patric host races of the pea aphid Acyrthosiphon pisumEcol Entomol 28 397ETH404

Etges W J 1998 Premating isolation is determined by lar-val-rearing substrates in cactophilic Drosophila mojaven-sis IV Correlated response in behavioral isolation toartiTHORNcial selection on a life-history trait Am Nat 152129ETH144

Etges W J 2002 Divergence in mate choice systems doesevolution play by rules Genetica 116 151ETH166

Etges W J and M A Ahrens 2001 Premating isolation isdetermined by larval-rearing substrates in cactophilicDrosophila mojavensis V Deep geographic variation inepicuticular hydrocarbons among isolated populationsAm Nat 158 585ETH598

Favila M E 1988a Comportamiento durante el periodode maduracion gonadica en un escarabajo rodador(Coleoptera Scarabaeidae Scarabaeinae) Folia EntomolMex 76 55ETH64

Favila M E 1988b Un metodo sencillo para marcar escar-abajos Folia Entomol Mex 75 117ETH118

Favila M E 1993 Some ecological factors affecting thelife-style of Canthon cyanellus cyanellus (ColeopteraScarabaeidae) an experimental approach Ethol EcolEvol 5 319ETH328

FavilaM E andADıaz 1996 Canthon cyanellus cyanellusLeConte (Coleoptera Scarabaeidae) makes a nest in theTHORNeld with several brood balls Coleopt Bull 50 52ETH60

GreenbergA J J RMoran J ACoyne andC IWu 2003Ecological adaptation during incipient speciation re-vealed by precise gene replacement Science (WashDC) 302 1754ETH1757

Halffter G 1961 Monografotildea de las especies norteameri-canas del genero Canthon Hoffsg (Coleopt Scarab)Ciencia 20 225ETH320

Halffter G M E Favila and V Halffter 1992 A compar-ative study of the structure of the scarab guild in Mexicantropical rain forest and derived ecosystems Folia Ento-mol Mex 84 131ETH156

Halffter G V Halffter and C Huerta 1983 Comporte-ment sexual et nidiTHORNcation chez Canthon cyanellus cya-nellus LeConte Bull Soc Entomol Mex 88 585ETH594

November 2006 ORTIZ-DOMINGUEZ ET AL MATE RECOGNITION IN ALLOPATRIC SCARABS 1255

Mendelson T C 2003 Sexual isolation evolves faster thanhybrid inviability in a diverse and sexually domorphicgenus of THORNsh (Percidae Etheostoma) Evolution 57 317ETH327

Michalak P JGrzesik and J Rafinski 1997 Test for sexualincompatibility between two new species Triturus vul-garis and Triturus montandoni no choice mating designEvolution 51 2045ETH2050

MillerGLGE StrattonPRMiller andEHebets 1998Geographical variation in male courtship behavior andsexual isolation in wolf spiders of the genus SchizocosaAnim Behav 56 937ETH951

Miththapala S J Seidensticker and S J OrsquoBrien 1996Phylogeographic subspecies recognition in leopards(Panthera pardus) molecular genetic variation ConservBiol 10 1115ETH1132

Noor MAF M A Williams D Alvarez and M Ruiz-Garcıa 2000 Lack of evolution divergence in courtshipsongs of Drosophila pseudoobscura subspecies J InsectBehav 13 255ETH262

Nosil P 2004 Reproductive isolation caused by visual pre-dation on migrants between divergent environmentsProc R Soc Lond B 271 1521ETH1528

Ortız-Domınguez M M E Favila M R Mendoza-LopezO Garcıa-Barradas and J Cruz-Sanchez 2006 The re-lationship between epicuticular compounds behavioralcues and sexual recognition in the ball roller scarabCanthon c cyanellus Entomol Exp Appl 119 23ETH27

Paillette M N Bizat and D Joly 1997 Differentiation ofdialects and courtship strategies in allopatric populationsof Drosophila teissieri J Insect Physiol 43 809ETH814

Panhuis T M R Butlin M Zuk and T Tregenza 2001Sexual selection and speciation Trends Ecol Evol 16364ETH371

Ritchie M G E J Halsey and M G Gleason 1999 Dro-sophila song as a species-speciTHORNc mating signal and thebehavioral importance of Kyriacou and Hall cycles inD melanogaster song Anim Behav 58 649ETH657

Rybak F G Sureau and T Aubin 2002 Functional cou-pling of acoustic and chemical signals in the courtshipbehavior of the male Drosophila melanogaster Proc RSoc Lond B 269 695ETH701

Schluter D 2000 The ecology of adaptive radiationOxford University Press Oxford United Kingdom

Schluter D 2001 Ecology and the origin of the speciesTrends Ecol Evol 16 372ETH380

Sokal R R and F J Rohlf 1981 Biometry 2nd ed W HFreeman San Francisco CA

Solıs A and B Kohlmann 2002 El genero Canthon (Co-leoptera Scarabaeidae) en Costa Rica G Ital Entomol10 11ETH35

Villalobos F J ADiaz andMEFavila 1998 Two speciesof Canthon Hoffmannsegg (Coleoptera Scarabaeidae)feed on dead and live invertebrates Coleopt Bull 52101ETH104

Vines T H and D Schluter 2006 Strong assortative mat-ing between allopatric sticklebacks as a by-product ofadaptation to different environments Proc Soc B 273911ETH916

Received 7 December 2005 accepted 2 August 2006

1256 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 99 no 6

are also found in pastures near the fragments (Halffteret al 1992) The THORNrst locality was in the OcircLos TuxtlasOtildebiological research station (National AutonomousUniversity of Mexico) located within the Los TuxtlasBiosphere Reserve and the last site is located in the OcircElCieloOtilde Biosphere Reserve (University of Tamaulipas)There are visible differences between some popu-lations in the morphology of their cuticular colorand texture We focused our attention on elytral vari-ation in three of the populations analyzed specimensfrom Los Tuxtlas the southern site have shiny greenelytra with a smooth interpunctual surface and feeblypunctuate elytral striae The elytra of the Papantlaspecimens the intermediate site are blue or mattegreen but the interpunctual surface is generally THORNneshagreening with punctuate elytral striae The elytraof the Gomez Farotildeas specimens the northern site areshiny green with a smooth interpunctual surface andpunctuate elytral striae Specimens from each popu-lation were put in containers with moist soil in arearing room at 27C 70 10 RH and a photoperiodof 1212 (LD) h We followed the rearing protocol ofFavila (1993)

Experiments were conducted in another insec-tarium under the same conditions as those of therearing room Because it is not possible to distinguishthe sex of these beetles with the naked eye sex iden-tiTHORNcation was done beforehand under the microscopeThen in each experiment one elytrum of one memberof the pair was marked with a small spot of latex paint(known to not affect behavior Favila 1988b) Exper-

iments were conducted from 0900 to 1300 hours themost active rolling period of the day for this species(Favila and Dotildeaz 1996)Intra- and Intersite Sexual Recognition MaleETHfe-

male pairs were randomly selected from each site andeach one pair was placed in an observation arena (10-by 2-cm petri dish lined with THORNlter paper and a 2 g ofTHORNsh meat ball) As we were interested in observing thebehavior of an individual of each sex in the presenceof an individual of the opposite sex from the same oranother site we made 25 combinations (male frompopulation A with female from population B and viceversa male from C and vice versa and so on) includ-ing intra- and interpopulation combinations (unequaln for each combination is indicated in Table 1) Timeuntil interaction was measured as the elapsed time (inseconds) between introducing the pair into the arenaand the beginning of maleETHfemale interaction on thefood ball Time to rolling was measured as the numberof seconds that elapsed between the beginning of themaleETHfemale pair interaction on the food ball and ballrolling by one or both partners The sex of the indi-vidual that rolled the food ball was recorded Weconsidered the rolling of the food ball by both indi-viduals as cooperative behavior but always during thisjoint rolling one of the partners usually the male rollsthe ball while the other partner is transported on theball The frequency of the following noncooperativebehaviors also was analyzed 1) avoidance when oneindividual rolled the food ball alone and did not allowthe other member to climb on the ball 2) pushing

Table 1 Latency in initial interaction and rolling and proportion of noncooperative behavior (individual rolling of the food ball andfights) observed between malendashfemale pairs of Canthon c cyanellus from five populations on the east coast of Mexico and in theircombinations

Male pop Female popDistance

(km)n

Time to interaction(25ETH75)

Time to rolling(25ETH75)

Noncooperativebehavior ()

Papantla Papantla 0 30 66 (3225ETH165) 14 (60ETH27)a 0Palma Sola Palma Sola 0 29 39 (2825ETH11175) 12 (675ETH285)bc 3Gomez Farotildeas Gomez Farotildeas 0 26 33 (16ETH75) 26 (65ETH405) 0Tancoco Tancoco 0 26 46 (25ETH65) 14 (60ETH40) 0Tuxtlas Tuxtlas 0 30 38 (15ETH92) 255 (100ETH50) 3Papantla Tancoco 107 28 605 (275ETH81) 23 (1125ETH10625) 4Tancoco Papantla 107 30 505 (19ETH123) 295 (60ETH45) 7Papantla Palma Sola 122 31 365 (20ETH93) 20 (975ETH505) 13Palma Sola Papantla 122 30 46 (23ETH13275) 24 (1425ETH3675) 17Palma Sola Tuxtlas 195 23 40 (2425ETH7875) 20 (95ETH51) 9Tuxtlas Palma Sola 195 15 40 (15ETH745) 15 (1075ETH2525) 7Gomez Farotildeas Tancoco 223 19 705 (14ETH94) 21 (85ETH7125) 5Tancoco Gomez Farotildeas 223 20 74 (35ETH93) 28 (10ETH565) 0Palma Sola Tancoco 228 30 65 (35ETH127) 615 (25ETH144)ac 10Tancoco Palma Sola 228 30 595 (38ETH146) 325 (13ETH60) 23Papantla Tuxtlas 317 29 67 (1475ETH9425) 345 (14ETH93) 14Tuxtlas Papantla 317 30 58 (36ETH75) 15 (65ETH3875) 17Papantla Gomez Farotildeas 331 26 60 (2875ETH805) 56 (1875ETH12075)b 19Gomez Farotildeas Papantla 331 26 41 (24ETH925) 385 (12ETH785) 19Tancoco Tuxtlas 424 25 116 (45ETH18925) 24 (165ETH155) 20Tuxtlas Tancoco 424 27 81 (3525ETH117) 26 (1425ETH4975) 4Palma Sola Gomez Farotildeas 447 31 36 (20ETH6475) 28 (105ETH6725) 6Gomez Farotildeas Palma Sola 447 20 99 (3575ETH16625) 28 (115ETH385) 70a

Gomez Farotildeas Tuxtlas 643 31 53 (2525ETH1075) 345 (15ETH124) 55a

Tuxtlas Gomez Farotildeas 643 24 575 (375ETH715) 32 (1225ETH9825) 50a

Similar letters indicate signiTHORNcant differences between maleETHfemale pair combinations according to DunnOtildes test (P 005) Q values a 44 b 39 and c 47a G 10947 df 24 P 005

1250 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 99 no 6

when one individual attempted to climb on the foodball and the other pushed it back by ldquobuttingrdquo it and3) THORNghts both male and female fought for the ball onsome occasions cutting it so that each beetle rolled anindependent fragment of food Once the pair hadmade contact on the food ball the total observationtime lasted 5 min An analysis of variance (ANOVA)was used to analyze time until interaction and rollingoccurred both among populations and for combina-tions The frequency of each noncooperative interac-tion was analyzed with the G test (Sokal and Rohlf1981)Reproductive Success among Populations We did

not know the age and reproductive condition of spec-imens collected in the THORNeld Therefore we allowedpairs of the F1 of the laboratory strains originallycollected from the THORNve sites and more distant combi-nations (where noncooperative behaviors were morefrequently observed see Results) to mate MaleETHfe-male pairs were 20ETH30 d old the optimal age fornesting in this species (Favila 1993) Each maleETHfe-male pair was placed in a plastic nesting box (9 by 9by 8 cm) containing 25 cm of moist soil and a pieceof THORNsh suitable for reproduction (1 g) Pairs were fedTHORNsh every other day and brood ball production wasfollowed for 40 d the most active period in the re-productive life of a female under laboratory condi-tions After this period females tend to produce broodballs sporadically and offspring survival is low (Favila1993) We measured the elapsed time between theintroduction of the pair to the nesting box and theconstruction of the THORNrst brood ball The total numberof brood balls produced by a female was considered ameasure of fecundity and the number of beetleseclosed from the brood balls as a measure of fertilityEvery variable was analyzed using a KruskalETHWallistestChemical Analysis Elytra from THORNve males and THORNve

females of the populations that were furthest apart(Gomez Farotildeas and Los Tuxtlas) and one intermediatepopulation (Papantla) were individually put intoWheaton vials containing 1 ml of hexane (HPLCgrade Sigma-Aldrich St Louis MO) Every vial wasgently shaken on a shaker table for 1 min and left at 4Cfor 24 h Next the elytra were removed from thevials and the solvent was evaporated using N2 Then50 l of hexane was added to each vial along withpentadecane (n-C15) as an internal standard A 2-lsample was injected into a HewlettETHPackard G1800BGC equipped with a fused silica capillary column(HP-5) cross-linked with 5 methylsiloxane (025mm 30 m) and 025-m THORNlm thickness supplied byAgilent (Mexico DF) with helium as the carrier gasat a szligow rate of 1 mlmin We optimized the separa-tion of extracted compounds by using a column tem-perature proTHORNle in which the analysis began at 70Cwas increased 20Cmin to 190C increased again10Cmin to 250C and then increased 30Cmin to280C and that temperature was held for 10 minWith the presenceETHabsence matrix of compounds weapplied the simple matching coefTHORNcient The relation-ship among individuals belonging to the three popu-

lations was represented using average linkage clusteranalysis (unweighted pair-group method with arith-metic average)

Results

Intra- and Interpopulation Sexual RecognitionWedid not THORNnd signiTHORNcant differences in the time untilinteraction occurred among the maleETHfemale com-binations (F 124 df 24 P 020) Howevertime to rolling was signiTHORNcantly different amongtreatments (Table 1 H 5636 df 24 P 0001)Pairs from Palma Sola began to roll the food ball fasterthan pairs belonging to the Papantla-Gomez Farotildeasand Palma Sola-Tancoco combinations Pairs fromPapantla began to roll faster than pairs from thePalma Sola-Tancoco combination MaleETHfemale pairsfrom the same populations always showed joint rollingbehavior only for one pair from Los Tuxtlas and onepair from Palma Sola was pushing (a noncooperativebehavior) observed during joint rolling (Table 1) inboth cases the female pushed the male

Joint rolling was observed in all the pairs from in-terpopulation combinations but there was also non-cooperative behavior in 19 of these combinationsThere was a clear correlation between the frequencyof noncooperative behavior and the geographic dis-tance between populations (Pearson coefTHORNcient r 0727 P 0001) However only in three of thesecombinations (male from Gomez Farotildeasfemale fromPalma Sola male from Gomez Farotildeasfemale from LosTuxtlas and male from Los Tuxtlasfemale fromGomez Farotildeas) was the frequency of noncooperativebehaviors signiTHORNcantly higher than in pairs from theother combinations (Table 1G 10947 df 24 P005) Individuals in these combinations belonged tothe most distant populations For the combinationwith the male from Gomez Farotildeas and the female fromPalma Sola in 45 of the pairs formed females rolledthe food ball alone preventing the males from climb-ing onto the food ball by pushing them off with theirheads In 25 of these combinations THORNghts for the foodball were observed However for the inverse combi-nation (male from Palma Solafemale from GomezFarotildeas) joint rolling was frequently observed (Table1) except in two pairs (6) where the females rolledthe ball alone and prevented the males from climbingonto the food ball In contrast for both combinationsfrom Los Tuxtlas and Gomez Farotildeas a signiTHORNcant num-ber of noncooperative behaviors were observed Inthe male from Gomez Farotildeasfemale from Los Tuxtlascombination males exhibited signiTHORNcantly more non-cooperative behavior than females (67 versus 8 re-spectively Fisher test 005) However in the malefrom Los Tuxtlasfemale from Gomez Farotildeas combi-nation 70 of the females exhibited noncooperativebehavior with males (Fisher test 002) whereasmales did not show any noncooperative behavior Ingeneral individuals from Gomez Farotildeas showed morenoncooperative behaviors toward individuals fromthe opposite sex of Los Tuxtlas than vice versa (69 and5 respectively Fisher test 00015)

November 2006 ORTIZ-DOMINGUEZ ET AL MATE RECOGNITION IN ALLOPATRIC SCARABS 1251

Reproductive Success among Populations Fecun-dity was different among populations (Fig 2 H 2177 df 5 P 005) The fecundity of femalesfrom Gomez Farotildeas and that of the Los Tuxtlas-GomezFarotildeas combination was signiTHORNcantly lower than thoseof Los Tuxtlas Palma Sola and Papantla (Fig 2Q testP 005 in all three cases) Female fertility was alsosigniTHORNcantly different among the populations (Fig 3H 1976 df 5 P 0001) The fertility of femalesfrom Los Tuxtlas was higher than fertility from GomezFarotildeas and Palma Sola (Fig 3 Q 36 P 005)Although the fecundity of females belonging to the

Los Tuxtlas-Gomez Farotildeas combination was the lowest(maximum four brood balls per female) their fertilitywas high and not signiTHORNcantly different from that ofthe other populations (80 offspring survival)Chemical Analysis The southern population of Los

Tuxtlas had 48 cuticular compounds Twenty-onecompounds were exclusive to females but only twocompounds were exclusive to males The other 25compounds were shared by individuals of both sexesThe intermediate population (Papantla) had 41 com-pounds nine exclusive to females and three exclusiveto males Both sexes shared 29 compounds The north-

Fig 2 Brood balls produced by Canthon c cyanellus females from THORNve populations located along a latitudinal gradienton the east coast of Mexico (see Fig 1) and brood ball production by females from Gomez Farotildeas mated with males fromLos Tuxtlas (GF-TX) The other abbreviations are Papantla (PP) Palma Sola (PS) and Tancoco (TN) Brood ball productionby females fromtheGFpopulationand fromtheGFETHTXcombinationwas signiTHORNcantly lower than thatof theotherpopulations(DunnOtildes method P 005)

Fig 3 Survival of eclosed beetles from THORNve populations located along a latitudinal gradient on the east coast of Mexicoand beetles eclosed from brood balls produced by females from Gomez Farotildeas mated with males from Los Tuxtlas (GF-TX)Survival was higher in the Los Tuxtlas population than in the Gomez Farotildeas and Palma Sola populations (DunnOtildes methodP 005) Abbreviations as in Fig 2

1252 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 99 no 6

Table 2 Cuticular compounds found in the elytra of Canthon c cyanellus males and females from Los Tuxtlas Papantla and GoacutemezFariacuteas

Retentiontime

Tuxtlasfemale

Tuxtlasmale

Papantlafemale

Papantlamale

Gomez Farotildeasfemale

Gomez Farotildeasmale

Sharedcompounds

1041 014 (086) 006 (015) 004 (080) 51061 003 (028) 046 (333) 017 (018ETH025) 008 (080) 11128 052 (478) 015 (033) 006 (090) 51199 007 (040) 056 (348) 013 (011ETH022) 009 (040) 11229 006 (058) 047 (332)1258 21 (1868)1272 005 (027)1279 004 (042) 004 (023)1309 01 (052) 027 (087)1321 032 (197) 2 (12) 055 (399) 033 (194) 006 (031) 23135 005 (029) 006 (062) 3145 015 (075) 004 (023)1464 042 (423) 116 (695) 055 (058) 021 (212) 002 (190) 04 (130) 121482 028 (177) 019 (095) 006 (056) 61507 008 (034)1529 083 (342) 116 (256) 045 (176) 043 (233) 046 (068) 016 (048) 121567 013 (129) 039 (252)1585 093 (274) 029 (173) 02 (069) 009 (040) 371595 157 (535) 009 (055)1658 007 (074) 008 (030)1679 009 (067)1702 038 (191) 082 (42)1715 581 (074ETH1372) 332 (028ETH1667) 235 (602) 263 (629) 014 (030) 05 (146) 121733 007 (065)1769 141 (051ETH296) 124 (040ETH457) 062 (108) 054 (101) 019 (077) 231789 16 (634)1803 011 (160)1815 263 (679) 053 (320) 036 (254) 152 (445) 009 (050) 231830 55 (136ETH1275 344 (053ETH1358 505 (1729 393 (933) 038 (135) 076 (024ETH217) 121851 323 (838) 061 (368) 076 (305) 148 (511) 007 (034) 231865 06 (299) 144 (566) 001 (060) 81879 184 (545) 092 (441) 31885 256 (1572) 424 (1711) 011 (052) 81892 13 (753) 05 (299)1900 052 (129) 048 (110) 075 (343) 003 (032) 005 (026) 03 (136) 121905 047 (254) 093 (666) 16 (529) 31924 604 (2447) 077 (469) 22 (884) 344 (1547) 361943 054 (227)196 351 (1128) 062 (373) 006 (030)) 104 (017ETH41) 471967 103 (419) 039 (234) 1472 (9678) 1848 (4854) 361979 073 (279) 029 (084) 359 (1128) 174 (473) 016 (079) 037 (172) 122007 84 (2859) 155 (928) 201 (593) 372 (957) 02 (098) 232025 016 (098) 019 (19) 04 (32) 62047 09 (438) 213 (695)2056 029 (287) 039 (233) 034 (092) 09 (349) 472062 97 (3886) 1684 (380)2074 021 (10) 163 (1075) 106 (45) 039 (145) 382084 309 (862) 05 (298) 359 (1396) 782 (1839) 362117 396 (1353)2129 016 (162)2146 102 (553) 179 (529)2167 782 (2483) 151 (978) 056 (247) 72173 67 (3218) 1085 (3574)2183 104 (1044)219 006 (030) 019 (093)2235 459 (1964) 082 (494) 118 (347) 259 (875) 362273 166 (1662)2278 176 (665)2306 061 (312) 162 (669)2323 013 (312) 009 (093)2348 028 (277) 098 (376)2442 1527 (7818) 46 (2760)2576 123 (529)2641 4 (529)

Mean concentration in micrograms per milliliter (minimum and maximum value) Minimum value is 0 except where another value isindicated

Italics indicate compounds exclusive to sex for a given population as follows 1 females from all the populations 2 males from all thepopulations 3 females from Los Tuxtlas and Papantla 4 females from Los Tuxtlas and Gomez Farotildeas 5 females from Papantla and GomezFarotildeas 6 males from Los Tuxtlas and Papantla 7 Males from Los Tuxtlas and Gomez Farotildeas 8 males from Papantla and Gomez Farotildeas

November 2006 ORTIZ-DOMINGUEZ ET AL MATE RECOGNITION IN ALLOPATRIC SCARABS 1253

ern population of Gomez Farotildeas had 24 compoundsNo compounds were exclusive to females whereas 11compounds were exclusive to males Thirteen com-pounds were shared (Table 2)

The comparison among populations showed thatindividuals of Los Tuxtlas had 18 compounds notpresent in individuals of Papantla or Gomez FarotildeasIndividuals of Gomez Farotildeas had only one com-pound shared by both sexes that was not present inthe Los Tuxtlas or Papantla populations Individualsfrom Papantla had 11 compounds not present in theGomez Farotildeas or Los Tuxtlas populations In generalboth sexes from the three populations shared six cu-ticular compounds The chemical analysis showed noquantitative differences in shared compounds be-tween the populations or between the sexes of eachpopulation (Table 2)

Two groups were formed with the Simple MatchingCoefTHORNcient (Fig 4) One group has females from LosTuxtlas The other group has two subgroups one withmales and females from Papantla and the other withmales and females from Gomez Farotildeas that were morerelated to males from Los Tuxtlas (Fig 4)

Discussion

In C c cyanellus the behavior during sexual en-counter and food ball rolling by intrapopulation pairswas mostly cooperative However although most ofthe pairs of individuals belonging to different localitiesalso were able to roll cooperatively noncooperativebehaviors increased signiTHORNcantly with distance sug-gesting that individuals prefer similar mates to matesfrom other populations This was very clear in maleETHfemale pairs of individuals from 600 km apart thatwere unwilling to roll a food ball with a partner eitherkeeping it away from the partner or THORNghting insteadof rolling the food ball cooperatively for nestingThese noncooperative behaviors are similar to those

observed between individuals of the same sex (Favila1988a) but in this case they act as behavioral precop-ulatory barriers The increased conszligict during sexualencounters between males and females from distantpopulations suggests that an incipient process of spe-ciation is occurring in C c cyanellus

Reproductive success was low in females fromGomez Farotildeas and in Gomez Farotildeas-Los Tuxtlas com-binations Pairs were fed THORNsh because this type offood has been successfully used in our C c cyanellusbreeding program (Favila 1993) However when pairsfrom Gomez Farotildeas and those from the Gomez Farotildeas-Los Tuxtlas combinations were feeding we observedthat after two to eight brood balls had been made thepair did not use the food for nesting thus reducingtheir fecundity In general C c cyanellus breed usingsmall vertebrate corpses (Halffter et al 1983 Favila andDotildeaz 1996) However Villalobos et al (1998) foundthat individuals from Gomez Farotildeas are able to eat andbreed by using carcasses ofOrthoporus ornatosGirard(Spirostreptida Spirostreptidae) a common diplopodin this region This behavior together with our resultssuggests that individuals from Gomez Farotildeas may havebecome very specialized in eating and nesting withDiplopoda as a local adaptation However althoughfecundity was low for the females from Gomez Farotildeasand the Gomez Farotildeas-Los Tuxtlas combinations off-spring survival was high so there is no evidence ofpostmating isolation in the interpopulation crossesUnfortunately the low number of eclosed beetles ob-tained in these combinations did not allow us to crossthe F2 generation

According to the ecological theory of speciationreproductive isolation ultimately evolves as a conse-quence of divergent natural selection on traits be-tween environments (Schluter 2001) If intrademicmate recognition systems are similar to species rec-ognition systems divergence in signaling systems canlead to premating isolation when geographically iso-

Fig 4 Tree representing cluster analysis for Canthon c cyanellus males and females in three populations from the eastcoast of Mexico on the basis of cuticular compounds TX Tuxtlas PP Papantla GF Gomez Farotildeas M male F female

1254 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 99 no 6

lated populations have secondary contact (Blows andAllan 1998 Greenberg et al 2003 Nosil 2004) Forexample hydrocarbons are involved in the mate rec-ognition system of Drosophila mojavensis Paterson(Diptera Drosophilidae) (Etges 1998) but this spe-cies shows regional variation associated with switch-ing hosts This has led to premating isolation amonggeographically isolated populations (Etges and Ahrens2001) In C c cyanellus the most important signalingsystem during sexual recognition is chemical via cu-ticular compounds (Ortiz-Domotildenguez et al 2006)Males and females from Gomez Farotildeas had the lowestnumber of cuticular compounds and were most ag-gressive with partners from the most distant popula-tion (Los Tuxtlas) which had the highest number ofcompounds The reduction in compounds could limitthe beetleOtildes ability to recognize individuals of theopposite sex from Los Tuxtlas which have the mostcomplex blend of cuticular compounds We hypoth-esize that switching in food preferences from car-casses of small vertebrates to Diplopoda carcasses inindividuals from the Gomez Farotildeas population couldmodify the quality and quantity of their cuticular com-pounds and produce the difTHORNculties observed in ac-cepting individuals from Los Tuxtlas and Palma SolaBecause Canthon c cyanellusmales recognize femalesby odor but females recognize individuals of the op-posite sex by other unknown behavioral cues as well(Ortotildez-Dominguez et al 2006) we do not discard thepossibility that another system of signals also could beinvolved in sexual recognition and these signals mayhave affected our results

Solotildes and Kohlmann (2002) suggest that allCanthoncyanellus subspecies from South America to Texasshould disappear However our results show that forthe subspecies C c cyanellus prezygotic isolationcould occur between distant populations when thereis secondary contact Owing to the ample distributionof C cyanellus in the Americas it is possible that thesouthern populations will have greater difTHORNculties insexual recognition if they come into contact withnorthern populations So we think that at presentit is not convenient to eliminate the subspecies ofC cyanellus but rather to reanalyze them in the con-text of Avise and Ball (1990) who consider subspeciesas members who share a unique geographic locale aset of phylogenetically concordant phenotypic char-acters and a unique natural history relative to othersubdivisions of the species but that are reproductivelycompatible with other such groups Phylogenetic dis-tinction must be based on the distributions of severalindependent genetically based traits (Miththapala etal 1996) not only on morphological traits as has beendone with C cyanellus

In summary the experiments here presented sug-gest that premating isolation is result of local adapta-tion related to a switch in food preferences If that isthe case future studies should analyze the factorsrelated with this switch Comparison with other south-ern populations will be crucial to determiningwhether this species is still cohesive or is in a processof THORNssion or whether there are even a complex of

cryptic species that are difTHORNcult to distinguish mor-phologically

Acknowledgments

We are grateful to Ingrid Marquez and Bianca Delfosse forlinguistic corrections and also to Carlos Fragoso for helpfulcomments and suggestions on the manuscript This study wassupported by Consejo Nacional de Ciencia y Tecnologia-Mexico Grants 4161P-N9607 35125-V and 49472-Q

References Cited

Avise J C and R M Ball 1990 Principles of genealogicalconcordance in species concepts and biological taxon-omy Oxford Surv Evol Biol 7 45ETH67

Blows M W and R A Allan 1998 Levels of mate recog-nition within and between two Drosophila species andtheir hybrids Am Nat 152 826ETH837

Boake CRB D K Price and D K Andreadis 1998 In-heritance of behavioral differences between two infer-tile sympatric species Drosophila silvestris and D het-eroneura Heredity 80 642ETH650

Del Campo M L S Via and M C Caillaud 2003 Rec-ognition of host-speciTHORNc chemical stimulants in two sym-patric host races of the pea aphid Acyrthosiphon pisumEcol Entomol 28 397ETH404

Etges W J 1998 Premating isolation is determined by lar-val-rearing substrates in cactophilic Drosophila mojaven-sis IV Correlated response in behavioral isolation toartiTHORNcial selection on a life-history trait Am Nat 152129ETH144

Etges W J 2002 Divergence in mate choice systems doesevolution play by rules Genetica 116 151ETH166

Etges W J and M A Ahrens 2001 Premating isolation isdetermined by larval-rearing substrates in cactophilicDrosophila mojavensis V Deep geographic variation inepicuticular hydrocarbons among isolated populationsAm Nat 158 585ETH598

Favila M E 1988a Comportamiento durante el periodode maduracion gonadica en un escarabajo rodador(Coleoptera Scarabaeidae Scarabaeinae) Folia EntomolMex 76 55ETH64

Favila M E 1988b Un metodo sencillo para marcar escar-abajos Folia Entomol Mex 75 117ETH118

Favila M E 1993 Some ecological factors affecting thelife-style of Canthon cyanellus cyanellus (ColeopteraScarabaeidae) an experimental approach Ethol EcolEvol 5 319ETH328

FavilaM E andADıaz 1996 Canthon cyanellus cyanellusLeConte (Coleoptera Scarabaeidae) makes a nest in theTHORNeld with several brood balls Coleopt Bull 50 52ETH60

GreenbergA J J RMoran J ACoyne andC IWu 2003Ecological adaptation during incipient speciation re-vealed by precise gene replacement Science (WashDC) 302 1754ETH1757

Halffter G 1961 Monografotildea de las especies norteameri-canas del genero Canthon Hoffsg (Coleopt Scarab)Ciencia 20 225ETH320

Halffter G M E Favila and V Halffter 1992 A compar-ative study of the structure of the scarab guild in Mexicantropical rain forest and derived ecosystems Folia Ento-mol Mex 84 131ETH156

Halffter G V Halffter and C Huerta 1983 Comporte-ment sexual et nidiTHORNcation chez Canthon cyanellus cya-nellus LeConte Bull Soc Entomol Mex 88 585ETH594

November 2006 ORTIZ-DOMINGUEZ ET AL MATE RECOGNITION IN ALLOPATRIC SCARABS 1255

Mendelson T C 2003 Sexual isolation evolves faster thanhybrid inviability in a diverse and sexually domorphicgenus of THORNsh (Percidae Etheostoma) Evolution 57 317ETH327

Michalak P JGrzesik and J Rafinski 1997 Test for sexualincompatibility between two new species Triturus vul-garis and Triturus montandoni no choice mating designEvolution 51 2045ETH2050

MillerGLGE StrattonPRMiller andEHebets 1998Geographical variation in male courtship behavior andsexual isolation in wolf spiders of the genus SchizocosaAnim Behav 56 937ETH951

Miththapala S J Seidensticker and S J OrsquoBrien 1996Phylogeographic subspecies recognition in leopards(Panthera pardus) molecular genetic variation ConservBiol 10 1115ETH1132

Noor MAF M A Williams D Alvarez and M Ruiz-Garcıa 2000 Lack of evolution divergence in courtshipsongs of Drosophila pseudoobscura subspecies J InsectBehav 13 255ETH262

Nosil P 2004 Reproductive isolation caused by visual pre-dation on migrants between divergent environmentsProc R Soc Lond B 271 1521ETH1528

Ortız-Domınguez M M E Favila M R Mendoza-LopezO Garcıa-Barradas and J Cruz-Sanchez 2006 The re-lationship between epicuticular compounds behavioralcues and sexual recognition in the ball roller scarabCanthon c cyanellus Entomol Exp Appl 119 23ETH27

Paillette M N Bizat and D Joly 1997 Differentiation ofdialects and courtship strategies in allopatric populationsof Drosophila teissieri J Insect Physiol 43 809ETH814

Panhuis T M R Butlin M Zuk and T Tregenza 2001Sexual selection and speciation Trends Ecol Evol 16364ETH371

Ritchie M G E J Halsey and M G Gleason 1999 Dro-sophila song as a species-speciTHORNc mating signal and thebehavioral importance of Kyriacou and Hall cycles inD melanogaster song Anim Behav 58 649ETH657

Rybak F G Sureau and T Aubin 2002 Functional cou-pling of acoustic and chemical signals in the courtshipbehavior of the male Drosophila melanogaster Proc RSoc Lond B 269 695ETH701

Schluter D 2000 The ecology of adaptive radiationOxford University Press Oxford United Kingdom

Schluter D 2001 Ecology and the origin of the speciesTrends Ecol Evol 16 372ETH380

Sokal R R and F J Rohlf 1981 Biometry 2nd ed W HFreeman San Francisco CA

Solıs A and B Kohlmann 2002 El genero Canthon (Co-leoptera Scarabaeidae) en Costa Rica G Ital Entomol10 11ETH35

Villalobos F J ADiaz andMEFavila 1998 Two speciesof Canthon Hoffmannsegg (Coleoptera Scarabaeidae)feed on dead and live invertebrates Coleopt Bull 52101ETH104

Vines T H and D Schluter 2006 Strong assortative mat-ing between allopatric sticklebacks as a by-product ofadaptation to different environments Proc Soc B 273911ETH916

Received 7 December 2005 accepted 2 August 2006

1256 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 99 no 6

when one individual attempted to climb on the foodball and the other pushed it back by ldquobuttingrdquo it and3) THORNghts both male and female fought for the ball onsome occasions cutting it so that each beetle rolled anindependent fragment of food Once the pair hadmade contact on the food ball the total observationtime lasted 5 min An analysis of variance (ANOVA)was used to analyze time until interaction and rollingoccurred both among populations and for combina-tions The frequency of each noncooperative interac-tion was analyzed with the G test (Sokal and Rohlf1981)Reproductive Success among Populations We did

not know the age and reproductive condition of spec-imens collected in the THORNeld Therefore we allowedpairs of the F1 of the laboratory strains originallycollected from the THORNve sites and more distant combi-nations (where noncooperative behaviors were morefrequently observed see Results) to mate MaleETHfe-male pairs were 20ETH30 d old the optimal age fornesting in this species (Favila 1993) Each maleETHfe-male pair was placed in a plastic nesting box (9 by 9by 8 cm) containing 25 cm of moist soil and a pieceof THORNsh suitable for reproduction (1 g) Pairs were fedTHORNsh every other day and brood ball production wasfollowed for 40 d the most active period in the re-productive life of a female under laboratory condi-tions After this period females tend to produce broodballs sporadically and offspring survival is low (Favila1993) We measured the elapsed time between theintroduction of the pair to the nesting box and theconstruction of the THORNrst brood ball The total numberof brood balls produced by a female was considered ameasure of fecundity and the number of beetleseclosed from the brood balls as a measure of fertilityEvery variable was analyzed using a KruskalETHWallistestChemical Analysis Elytra from THORNve males and THORNve

females of the populations that were furthest apart(Gomez Farotildeas and Los Tuxtlas) and one intermediatepopulation (Papantla) were individually put intoWheaton vials containing 1 ml of hexane (HPLCgrade Sigma-Aldrich St Louis MO) Every vial wasgently shaken on a shaker table for 1 min and left at 4Cfor 24 h Next the elytra were removed from thevials and the solvent was evaporated using N2 Then50 l of hexane was added to each vial along withpentadecane (n-C15) as an internal standard A 2-lsample was injected into a HewlettETHPackard G1800BGC equipped with a fused silica capillary column(HP-5) cross-linked with 5 methylsiloxane (025mm 30 m) and 025-m THORNlm thickness supplied byAgilent (Mexico DF) with helium as the carrier gasat a szligow rate of 1 mlmin We optimized the separa-tion of extracted compounds by using a column tem-perature proTHORNle in which the analysis began at 70Cwas increased 20Cmin to 190C increased again10Cmin to 250C and then increased 30Cmin to280C and that temperature was held for 10 minWith the presenceETHabsence matrix of compounds weapplied the simple matching coefTHORNcient The relation-ship among individuals belonging to the three popu-

lations was represented using average linkage clusteranalysis (unweighted pair-group method with arith-metic average)

Results

Intra- and Interpopulation Sexual RecognitionWedid not THORNnd signiTHORNcant differences in the time untilinteraction occurred among the maleETHfemale com-binations (F 124 df 24 P 020) Howevertime to rolling was signiTHORNcantly different amongtreatments (Table 1 H 5636 df 24 P 0001)Pairs from Palma Sola began to roll the food ball fasterthan pairs belonging to the Papantla-Gomez Farotildeasand Palma Sola-Tancoco combinations Pairs fromPapantla began to roll faster than pairs from thePalma Sola-Tancoco combination MaleETHfemale pairsfrom the same populations always showed joint rollingbehavior only for one pair from Los Tuxtlas and onepair from Palma Sola was pushing (a noncooperativebehavior) observed during joint rolling (Table 1) inboth cases the female pushed the male

Joint rolling was observed in all the pairs from in-terpopulation combinations but there was also non-cooperative behavior in 19 of these combinationsThere was a clear correlation between the frequencyof noncooperative behavior and the geographic dis-tance between populations (Pearson coefTHORNcient r 0727 P 0001) However only in three of thesecombinations (male from Gomez Farotildeasfemale fromPalma Sola male from Gomez Farotildeasfemale from LosTuxtlas and male from Los Tuxtlasfemale fromGomez Farotildeas) was the frequency of noncooperativebehaviors signiTHORNcantly higher than in pairs from theother combinations (Table 1G 10947 df 24 P005) Individuals in these combinations belonged tothe most distant populations For the combinationwith the male from Gomez Farotildeas and the female fromPalma Sola in 45 of the pairs formed females rolledthe food ball alone preventing the males from climb-ing onto the food ball by pushing them off with theirheads In 25 of these combinations THORNghts for the foodball were observed However for the inverse combi-nation (male from Palma Solafemale from GomezFarotildeas) joint rolling was frequently observed (Table1) except in two pairs (6) where the females rolledthe ball alone and prevented the males from climbingonto the food ball In contrast for both combinationsfrom Los Tuxtlas and Gomez Farotildeas a signiTHORNcant num-ber of noncooperative behaviors were observed Inthe male from Gomez Farotildeasfemale from Los Tuxtlascombination males exhibited signiTHORNcantly more non-cooperative behavior than females (67 versus 8 re-spectively Fisher test 005) However in the malefrom Los Tuxtlasfemale from Gomez Farotildeas combi-nation 70 of the females exhibited noncooperativebehavior with males (Fisher test 002) whereasmales did not show any noncooperative behavior Ingeneral individuals from Gomez Farotildeas showed morenoncooperative behaviors toward individuals fromthe opposite sex of Los Tuxtlas than vice versa (69 and5 respectively Fisher test 00015)

November 2006 ORTIZ-DOMINGUEZ ET AL MATE RECOGNITION IN ALLOPATRIC SCARABS 1251

Reproductive Success among Populations Fecun-dity was different among populations (Fig 2 H 2177 df 5 P 005) The fecundity of femalesfrom Gomez Farotildeas and that of the Los Tuxtlas-GomezFarotildeas combination was signiTHORNcantly lower than thoseof Los Tuxtlas Palma Sola and Papantla (Fig 2Q testP 005 in all three cases) Female fertility was alsosigniTHORNcantly different among the populations (Fig 3H 1976 df 5 P 0001) The fertility of femalesfrom Los Tuxtlas was higher than fertility from GomezFarotildeas and Palma Sola (Fig 3 Q 36 P 005)Although the fecundity of females belonging to the

Los Tuxtlas-Gomez Farotildeas combination was the lowest(maximum four brood balls per female) their fertilitywas high and not signiTHORNcantly different from that ofthe other populations (80 offspring survival)Chemical Analysis The southern population of Los

Tuxtlas had 48 cuticular compounds Twenty-onecompounds were exclusive to females but only twocompounds were exclusive to males The other 25compounds were shared by individuals of both sexesThe intermediate population (Papantla) had 41 com-pounds nine exclusive to females and three exclusiveto males Both sexes shared 29 compounds The north-

Fig 2 Brood balls produced by Canthon c cyanellus females from THORNve populations located along a latitudinal gradienton the east coast of Mexico (see Fig 1) and brood ball production by females from Gomez Farotildeas mated with males fromLos Tuxtlas (GF-TX) The other abbreviations are Papantla (PP) Palma Sola (PS) and Tancoco (TN) Brood ball productionby females fromtheGFpopulationand fromtheGFETHTXcombinationwas signiTHORNcantly lower than thatof theotherpopulations(DunnOtildes method P 005)

Fig 3 Survival of eclosed beetles from THORNve populations located along a latitudinal gradient on the east coast of Mexicoand beetles eclosed from brood balls produced by females from Gomez Farotildeas mated with males from Los Tuxtlas (GF-TX)Survival was higher in the Los Tuxtlas population than in the Gomez Farotildeas and Palma Sola populations (DunnOtildes methodP 005) Abbreviations as in Fig 2

1252 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 99 no 6

Table 2 Cuticular compounds found in the elytra of Canthon c cyanellus males and females from Los Tuxtlas Papantla and GoacutemezFariacuteas

Retentiontime

Tuxtlasfemale

Tuxtlasmale

Papantlafemale

Papantlamale

Gomez Farotildeasfemale

Gomez Farotildeasmale

Sharedcompounds

1041 014 (086) 006 (015) 004 (080) 51061 003 (028) 046 (333) 017 (018ETH025) 008 (080) 11128 052 (478) 015 (033) 006 (090) 51199 007 (040) 056 (348) 013 (011ETH022) 009 (040) 11229 006 (058) 047 (332)1258 21 (1868)1272 005 (027)1279 004 (042) 004 (023)1309 01 (052) 027 (087)1321 032 (197) 2 (12) 055 (399) 033 (194) 006 (031) 23135 005 (029) 006 (062) 3145 015 (075) 004 (023)1464 042 (423) 116 (695) 055 (058) 021 (212) 002 (190) 04 (130) 121482 028 (177) 019 (095) 006 (056) 61507 008 (034)1529 083 (342) 116 (256) 045 (176) 043 (233) 046 (068) 016 (048) 121567 013 (129) 039 (252)1585 093 (274) 029 (173) 02 (069) 009 (040) 371595 157 (535) 009 (055)1658 007 (074) 008 (030)1679 009 (067)1702 038 (191) 082 (42)1715 581 (074ETH1372) 332 (028ETH1667) 235 (602) 263 (629) 014 (030) 05 (146) 121733 007 (065)1769 141 (051ETH296) 124 (040ETH457) 062 (108) 054 (101) 019 (077) 231789 16 (634)1803 011 (160)1815 263 (679) 053 (320) 036 (254) 152 (445) 009 (050) 231830 55 (136ETH1275 344 (053ETH1358 505 (1729 393 (933) 038 (135) 076 (024ETH217) 121851 323 (838) 061 (368) 076 (305) 148 (511) 007 (034) 231865 06 (299) 144 (566) 001 (060) 81879 184 (545) 092 (441) 31885 256 (1572) 424 (1711) 011 (052) 81892 13 (753) 05 (299)1900 052 (129) 048 (110) 075 (343) 003 (032) 005 (026) 03 (136) 121905 047 (254) 093 (666) 16 (529) 31924 604 (2447) 077 (469) 22 (884) 344 (1547) 361943 054 (227)196 351 (1128) 062 (373) 006 (030)) 104 (017ETH41) 471967 103 (419) 039 (234) 1472 (9678) 1848 (4854) 361979 073 (279) 029 (084) 359 (1128) 174 (473) 016 (079) 037 (172) 122007 84 (2859) 155 (928) 201 (593) 372 (957) 02 (098) 232025 016 (098) 019 (19) 04 (32) 62047 09 (438) 213 (695)2056 029 (287) 039 (233) 034 (092) 09 (349) 472062 97 (3886) 1684 (380)2074 021 (10) 163 (1075) 106 (45) 039 (145) 382084 309 (862) 05 (298) 359 (1396) 782 (1839) 362117 396 (1353)2129 016 (162)2146 102 (553) 179 (529)2167 782 (2483) 151 (978) 056 (247) 72173 67 (3218) 1085 (3574)2183 104 (1044)219 006 (030) 019 (093)2235 459 (1964) 082 (494) 118 (347) 259 (875) 362273 166 (1662)2278 176 (665)2306 061 (312) 162 (669)2323 013 (312) 009 (093)2348 028 (277) 098 (376)2442 1527 (7818) 46 (2760)2576 123 (529)2641 4 (529)

Mean concentration in micrograms per milliliter (minimum and maximum value) Minimum value is 0 except where another value isindicated

Italics indicate compounds exclusive to sex for a given population as follows 1 females from all the populations 2 males from all thepopulations 3 females from Los Tuxtlas and Papantla 4 females from Los Tuxtlas and Gomez Farotildeas 5 females from Papantla and GomezFarotildeas 6 males from Los Tuxtlas and Papantla 7 Males from Los Tuxtlas and Gomez Farotildeas 8 males from Papantla and Gomez Farotildeas

November 2006 ORTIZ-DOMINGUEZ ET AL MATE RECOGNITION IN ALLOPATRIC SCARABS 1253

ern population of Gomez Farotildeas had 24 compoundsNo compounds were exclusive to females whereas 11compounds were exclusive to males Thirteen com-pounds were shared (Table 2)

The comparison among populations showed thatindividuals of Los Tuxtlas had 18 compounds notpresent in individuals of Papantla or Gomez FarotildeasIndividuals of Gomez Farotildeas had only one com-pound shared by both sexes that was not present inthe Los Tuxtlas or Papantla populations Individualsfrom Papantla had 11 compounds not present in theGomez Farotildeas or Los Tuxtlas populations In generalboth sexes from the three populations shared six cu-ticular compounds The chemical analysis showed noquantitative differences in shared compounds be-tween the populations or between the sexes of eachpopulation (Table 2)

Two groups were formed with the Simple MatchingCoefTHORNcient (Fig 4) One group has females from LosTuxtlas The other group has two subgroups one withmales and females from Papantla and the other withmales and females from Gomez Farotildeas that were morerelated to males from Los Tuxtlas (Fig 4)

Discussion

In C c cyanellus the behavior during sexual en-counter and food ball rolling by intrapopulation pairswas mostly cooperative However although most ofthe pairs of individuals belonging to different localitiesalso were able to roll cooperatively noncooperativebehaviors increased signiTHORNcantly with distance sug-gesting that individuals prefer similar mates to matesfrom other populations This was very clear in maleETHfemale pairs of individuals from 600 km apart thatwere unwilling to roll a food ball with a partner eitherkeeping it away from the partner or THORNghting insteadof rolling the food ball cooperatively for nestingThese noncooperative behaviors are similar to those

observed between individuals of the same sex (Favila1988a) but in this case they act as behavioral precop-ulatory barriers The increased conszligict during sexualencounters between males and females from distantpopulations suggests that an incipient process of spe-ciation is occurring in C c cyanellus

Reproductive success was low in females fromGomez Farotildeas and in Gomez Farotildeas-Los Tuxtlas com-binations Pairs were fed THORNsh because this type offood has been successfully used in our C c cyanellusbreeding program (Favila 1993) However when pairsfrom Gomez Farotildeas and those from the Gomez Farotildeas-Los Tuxtlas combinations were feeding we observedthat after two to eight brood balls had been made thepair did not use the food for nesting thus reducingtheir fecundity In general C c cyanellus breed usingsmall vertebrate corpses (Halffter et al 1983 Favila andDotildeaz 1996) However Villalobos et al (1998) foundthat individuals from Gomez Farotildeas are able to eat andbreed by using carcasses ofOrthoporus ornatosGirard(Spirostreptida Spirostreptidae) a common diplopodin this region This behavior together with our resultssuggests that individuals from Gomez Farotildeas may havebecome very specialized in eating and nesting withDiplopoda as a local adaptation However althoughfecundity was low for the females from Gomez Farotildeasand the Gomez Farotildeas-Los Tuxtlas combinations off-spring survival was high so there is no evidence ofpostmating isolation in the interpopulation crossesUnfortunately the low number of eclosed beetles ob-tained in these combinations did not allow us to crossthe F2 generation

According to the ecological theory of speciationreproductive isolation ultimately evolves as a conse-quence of divergent natural selection on traits be-tween environments (Schluter 2001) If intrademicmate recognition systems are similar to species rec-ognition systems divergence in signaling systems canlead to premating isolation when geographically iso-

Fig 4 Tree representing cluster analysis for Canthon c cyanellus males and females in three populations from the eastcoast of Mexico on the basis of cuticular compounds TX Tuxtlas PP Papantla GF Gomez Farotildeas M male F female

1254 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 99 no 6

lated populations have secondary contact (Blows andAllan 1998 Greenberg et al 2003 Nosil 2004) Forexample hydrocarbons are involved in the mate rec-ognition system of Drosophila mojavensis Paterson(Diptera Drosophilidae) (Etges 1998) but this spe-cies shows regional variation associated with switch-ing hosts This has led to premating isolation amonggeographically isolated populations (Etges and Ahrens2001) In C c cyanellus the most important signalingsystem during sexual recognition is chemical via cu-ticular compounds (Ortiz-Domotildenguez et al 2006)Males and females from Gomez Farotildeas had the lowestnumber of cuticular compounds and were most ag-gressive with partners from the most distant popula-tion (Los Tuxtlas) which had the highest number ofcompounds The reduction in compounds could limitthe beetleOtildes ability to recognize individuals of theopposite sex from Los Tuxtlas which have the mostcomplex blend of cuticular compounds We hypoth-esize that switching in food preferences from car-casses of small vertebrates to Diplopoda carcasses inindividuals from the Gomez Farotildeas population couldmodify the quality and quantity of their cuticular com-pounds and produce the difTHORNculties observed in ac-cepting individuals from Los Tuxtlas and Palma SolaBecause Canthon c cyanellusmales recognize femalesby odor but females recognize individuals of the op-posite sex by other unknown behavioral cues as well(Ortotildez-Dominguez et al 2006) we do not discard thepossibility that another system of signals also could beinvolved in sexual recognition and these signals mayhave affected our results

Solotildes and Kohlmann (2002) suggest that allCanthoncyanellus subspecies from South America to Texasshould disappear However our results show that forthe subspecies C c cyanellus prezygotic isolationcould occur between distant populations when thereis secondary contact Owing to the ample distributionof C cyanellus in the Americas it is possible that thesouthern populations will have greater difTHORNculties insexual recognition if they come into contact withnorthern populations So we think that at presentit is not convenient to eliminate the subspecies ofC cyanellus but rather to reanalyze them in the con-text of Avise and Ball (1990) who consider subspeciesas members who share a unique geographic locale aset of phylogenetically concordant phenotypic char-acters and a unique natural history relative to othersubdivisions of the species but that are reproductivelycompatible with other such groups Phylogenetic dis-tinction must be based on the distributions of severalindependent genetically based traits (Miththapala etal 1996) not only on morphological traits as has beendone with C cyanellus

In summary the experiments here presented sug-gest that premating isolation is result of local adapta-tion related to a switch in food preferences If that isthe case future studies should analyze the factorsrelated with this switch Comparison with other south-ern populations will be crucial to determiningwhether this species is still cohesive or is in a processof THORNssion or whether there are even a complex of

cryptic species that are difTHORNcult to distinguish mor-phologically

Acknowledgments

We are grateful to Ingrid Marquez and Bianca Delfosse forlinguistic corrections and also to Carlos Fragoso for helpfulcomments and suggestions on the manuscript This study wassupported by Consejo Nacional de Ciencia y Tecnologia-Mexico Grants 4161P-N9607 35125-V and 49472-Q

References Cited

Avise J C and R M Ball 1990 Principles of genealogicalconcordance in species concepts and biological taxon-omy Oxford Surv Evol Biol 7 45ETH67

Blows M W and R A Allan 1998 Levels of mate recog-nition within and between two Drosophila species andtheir hybrids Am Nat 152 826ETH837

Boake CRB D K Price and D K Andreadis 1998 In-heritance of behavioral differences between two infer-tile sympatric species Drosophila silvestris and D het-eroneura Heredity 80 642ETH650

Del Campo M L S Via and M C Caillaud 2003 Rec-ognition of host-speciTHORNc chemical stimulants in two sym-patric host races of the pea aphid Acyrthosiphon pisumEcol Entomol 28 397ETH404

Etges W J 1998 Premating isolation is determined by lar-val-rearing substrates in cactophilic Drosophila mojaven-sis IV Correlated response in behavioral isolation toartiTHORNcial selection on a life-history trait Am Nat 152129ETH144

Etges W J 2002 Divergence in mate choice systems doesevolution play by rules Genetica 116 151ETH166

Etges W J and M A Ahrens 2001 Premating isolation isdetermined by larval-rearing substrates in cactophilicDrosophila mojavensis V Deep geographic variation inepicuticular hydrocarbons among isolated populationsAm Nat 158 585ETH598

Favila M E 1988a Comportamiento durante el periodode maduracion gonadica en un escarabajo rodador(Coleoptera Scarabaeidae Scarabaeinae) Folia EntomolMex 76 55ETH64

Favila M E 1988b Un metodo sencillo para marcar escar-abajos Folia Entomol Mex 75 117ETH118

Favila M E 1993 Some ecological factors affecting thelife-style of Canthon cyanellus cyanellus (ColeopteraScarabaeidae) an experimental approach Ethol EcolEvol 5 319ETH328

FavilaM E andADıaz 1996 Canthon cyanellus cyanellusLeConte (Coleoptera Scarabaeidae) makes a nest in theTHORNeld with several brood balls Coleopt Bull 50 52ETH60

GreenbergA J J RMoran J ACoyne andC IWu 2003Ecological adaptation during incipient speciation re-vealed by precise gene replacement Science (WashDC) 302 1754ETH1757

Halffter G 1961 Monografotildea de las especies norteameri-canas del genero Canthon Hoffsg (Coleopt Scarab)Ciencia 20 225ETH320

Halffter G M E Favila and V Halffter 1992 A compar-ative study of the structure of the scarab guild in Mexicantropical rain forest and derived ecosystems Folia Ento-mol Mex 84 131ETH156

Halffter G V Halffter and C Huerta 1983 Comporte-ment sexual et nidiTHORNcation chez Canthon cyanellus cya-nellus LeConte Bull Soc Entomol Mex 88 585ETH594

November 2006 ORTIZ-DOMINGUEZ ET AL MATE RECOGNITION IN ALLOPATRIC SCARABS 1255

Mendelson T C 2003 Sexual isolation evolves faster thanhybrid inviability in a diverse and sexually domorphicgenus of THORNsh (Percidae Etheostoma) Evolution 57 317ETH327

Michalak P JGrzesik and J Rafinski 1997 Test for sexualincompatibility between two new species Triturus vul-garis and Triturus montandoni no choice mating designEvolution 51 2045ETH2050

MillerGLGE StrattonPRMiller andEHebets 1998Geographical variation in male courtship behavior andsexual isolation in wolf spiders of the genus SchizocosaAnim Behav 56 937ETH951

Miththapala S J Seidensticker and S J OrsquoBrien 1996Phylogeographic subspecies recognition in leopards(Panthera pardus) molecular genetic variation ConservBiol 10 1115ETH1132

Noor MAF M A Williams D Alvarez and M Ruiz-Garcıa 2000 Lack of evolution divergence in courtshipsongs of Drosophila pseudoobscura subspecies J InsectBehav 13 255ETH262

Nosil P 2004 Reproductive isolation caused by visual pre-dation on migrants between divergent environmentsProc R Soc Lond B 271 1521ETH1528

Ortız-Domınguez M M E Favila M R Mendoza-LopezO Garcıa-Barradas and J Cruz-Sanchez 2006 The re-lationship between epicuticular compounds behavioralcues and sexual recognition in the ball roller scarabCanthon c cyanellus Entomol Exp Appl 119 23ETH27

Paillette M N Bizat and D Joly 1997 Differentiation ofdialects and courtship strategies in allopatric populationsof Drosophila teissieri J Insect Physiol 43 809ETH814

Panhuis T M R Butlin M Zuk and T Tregenza 2001Sexual selection and speciation Trends Ecol Evol 16364ETH371

Ritchie M G E J Halsey and M G Gleason 1999 Dro-sophila song as a species-speciTHORNc mating signal and thebehavioral importance of Kyriacou and Hall cycles inD melanogaster song Anim Behav 58 649ETH657

Rybak F G Sureau and T Aubin 2002 Functional cou-pling of acoustic and chemical signals in the courtshipbehavior of the male Drosophila melanogaster Proc RSoc Lond B 269 695ETH701

Schluter D 2000 The ecology of adaptive radiationOxford University Press Oxford United Kingdom

Schluter D 2001 Ecology and the origin of the speciesTrends Ecol Evol 16 372ETH380

Sokal R R and F J Rohlf 1981 Biometry 2nd ed W HFreeman San Francisco CA

Solıs A and B Kohlmann 2002 El genero Canthon (Co-leoptera Scarabaeidae) en Costa Rica G Ital Entomol10 11ETH35

Villalobos F J ADiaz andMEFavila 1998 Two speciesof Canthon Hoffmannsegg (Coleoptera Scarabaeidae)feed on dead and live invertebrates Coleopt Bull 52101ETH104

Vines T H and D Schluter 2006 Strong assortative mat-ing between allopatric sticklebacks as a by-product ofadaptation to different environments Proc Soc B 273911ETH916

Received 7 December 2005 accepted 2 August 2006

1256 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 99 no 6

Reproductive Success among Populations Fecun-dity was different among populations (Fig 2 H 2177 df 5 P 005) The fecundity of femalesfrom Gomez Farotildeas and that of the Los Tuxtlas-GomezFarotildeas combination was signiTHORNcantly lower than thoseof Los Tuxtlas Palma Sola and Papantla (Fig 2Q testP 005 in all three cases) Female fertility was alsosigniTHORNcantly different among the populations (Fig 3H 1976 df 5 P 0001) The fertility of femalesfrom Los Tuxtlas was higher than fertility from GomezFarotildeas and Palma Sola (Fig 3 Q 36 P 005)Although the fecundity of females belonging to the

Los Tuxtlas-Gomez Farotildeas combination was the lowest(maximum four brood balls per female) their fertilitywas high and not signiTHORNcantly different from that ofthe other populations (80 offspring survival)Chemical Analysis The southern population of Los

Tuxtlas had 48 cuticular compounds Twenty-onecompounds were exclusive to females but only twocompounds were exclusive to males The other 25compounds were shared by individuals of both sexesThe intermediate population (Papantla) had 41 com-pounds nine exclusive to females and three exclusiveto males Both sexes shared 29 compounds The north-

Fig 2 Brood balls produced by Canthon c cyanellus females from THORNve populations located along a latitudinal gradienton the east coast of Mexico (see Fig 1) and brood ball production by females from Gomez Farotildeas mated with males fromLos Tuxtlas (GF-TX) The other abbreviations are Papantla (PP) Palma Sola (PS) and Tancoco (TN) Brood ball productionby females fromtheGFpopulationand fromtheGFETHTXcombinationwas signiTHORNcantly lower than thatof theotherpopulations(DunnOtildes method P 005)

Fig 3 Survival of eclosed beetles from THORNve populations located along a latitudinal gradient on the east coast of Mexicoand beetles eclosed from brood balls produced by females from Gomez Farotildeas mated with males from Los Tuxtlas (GF-TX)Survival was higher in the Los Tuxtlas population than in the Gomez Farotildeas and Palma Sola populations (DunnOtildes methodP 005) Abbreviations as in Fig 2

1252 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 99 no 6

Table 2 Cuticular compounds found in the elytra of Canthon c cyanellus males and females from Los Tuxtlas Papantla and GoacutemezFariacuteas

Retentiontime

Tuxtlasfemale

Tuxtlasmale

Papantlafemale

Papantlamale

Gomez Farotildeasfemale

Gomez Farotildeasmale

Sharedcompounds

1041 014 (086) 006 (015) 004 (080) 51061 003 (028) 046 (333) 017 (018ETH025) 008 (080) 11128 052 (478) 015 (033) 006 (090) 51199 007 (040) 056 (348) 013 (011ETH022) 009 (040) 11229 006 (058) 047 (332)1258 21 (1868)1272 005 (027)1279 004 (042) 004 (023)1309 01 (052) 027 (087)1321 032 (197) 2 (12) 055 (399) 033 (194) 006 (031) 23135 005 (029) 006 (062) 3145 015 (075) 004 (023)1464 042 (423) 116 (695) 055 (058) 021 (212) 002 (190) 04 (130) 121482 028 (177) 019 (095) 006 (056) 61507 008 (034)1529 083 (342) 116 (256) 045 (176) 043 (233) 046 (068) 016 (048) 121567 013 (129) 039 (252)1585 093 (274) 029 (173) 02 (069) 009 (040) 371595 157 (535) 009 (055)1658 007 (074) 008 (030)1679 009 (067)1702 038 (191) 082 (42)1715 581 (074ETH1372) 332 (028ETH1667) 235 (602) 263 (629) 014 (030) 05 (146) 121733 007 (065)1769 141 (051ETH296) 124 (040ETH457) 062 (108) 054 (101) 019 (077) 231789 16 (634)1803 011 (160)1815 263 (679) 053 (320) 036 (254) 152 (445) 009 (050) 231830 55 (136ETH1275 344 (053ETH1358 505 (1729 393 (933) 038 (135) 076 (024ETH217) 121851 323 (838) 061 (368) 076 (305) 148 (511) 007 (034) 231865 06 (299) 144 (566) 001 (060) 81879 184 (545) 092 (441) 31885 256 (1572) 424 (1711) 011 (052) 81892 13 (753) 05 (299)1900 052 (129) 048 (110) 075 (343) 003 (032) 005 (026) 03 (136) 121905 047 (254) 093 (666) 16 (529) 31924 604 (2447) 077 (469) 22 (884) 344 (1547) 361943 054 (227)196 351 (1128) 062 (373) 006 (030)) 104 (017ETH41) 471967 103 (419) 039 (234) 1472 (9678) 1848 (4854) 361979 073 (279) 029 (084) 359 (1128) 174 (473) 016 (079) 037 (172) 122007 84 (2859) 155 (928) 201 (593) 372 (957) 02 (098) 232025 016 (098) 019 (19) 04 (32) 62047 09 (438) 213 (695)2056 029 (287) 039 (233) 034 (092) 09 (349) 472062 97 (3886) 1684 (380)2074 021 (10) 163 (1075) 106 (45) 039 (145) 382084 309 (862) 05 (298) 359 (1396) 782 (1839) 362117 396 (1353)2129 016 (162)2146 102 (553) 179 (529)2167 782 (2483) 151 (978) 056 (247) 72173 67 (3218) 1085 (3574)2183 104 (1044)219 006 (030) 019 (093)2235 459 (1964) 082 (494) 118 (347) 259 (875) 362273 166 (1662)2278 176 (665)2306 061 (312) 162 (669)2323 013 (312) 009 (093)2348 028 (277) 098 (376)2442 1527 (7818) 46 (2760)2576 123 (529)2641 4 (529)

Mean concentration in micrograms per milliliter (minimum and maximum value) Minimum value is 0 except where another value isindicated

Italics indicate compounds exclusive to sex for a given population as follows 1 females from all the populations 2 males from all thepopulations 3 females from Los Tuxtlas and Papantla 4 females from Los Tuxtlas and Gomez Farotildeas 5 females from Papantla and GomezFarotildeas 6 males from Los Tuxtlas and Papantla 7 Males from Los Tuxtlas and Gomez Farotildeas 8 males from Papantla and Gomez Farotildeas

November 2006 ORTIZ-DOMINGUEZ ET AL MATE RECOGNITION IN ALLOPATRIC SCARABS 1253

ern population of Gomez Farotildeas had 24 compoundsNo compounds were exclusive to females whereas 11compounds were exclusive to males Thirteen com-pounds were shared (Table 2)

The comparison among populations showed thatindividuals of Los Tuxtlas had 18 compounds notpresent in individuals of Papantla or Gomez FarotildeasIndividuals of Gomez Farotildeas had only one com-pound shared by both sexes that was not present inthe Los Tuxtlas or Papantla populations Individualsfrom Papantla had 11 compounds not present in theGomez Farotildeas or Los Tuxtlas populations In generalboth sexes from the three populations shared six cu-ticular compounds The chemical analysis showed noquantitative differences in shared compounds be-tween the populations or between the sexes of eachpopulation (Table 2)

Two groups were formed with the Simple MatchingCoefTHORNcient (Fig 4) One group has females from LosTuxtlas The other group has two subgroups one withmales and females from Papantla and the other withmales and females from Gomez Farotildeas that were morerelated to males from Los Tuxtlas (Fig 4)

Discussion

In C c cyanellus the behavior during sexual en-counter and food ball rolling by intrapopulation pairswas mostly cooperative However although most ofthe pairs of individuals belonging to different localitiesalso were able to roll cooperatively noncooperativebehaviors increased signiTHORNcantly with distance sug-gesting that individuals prefer similar mates to matesfrom other populations This was very clear in maleETHfemale pairs of individuals from 600 km apart thatwere unwilling to roll a food ball with a partner eitherkeeping it away from the partner or THORNghting insteadof rolling the food ball cooperatively for nestingThese noncooperative behaviors are similar to those

observed between individuals of the same sex (Favila1988a) but in this case they act as behavioral precop-ulatory barriers The increased conszligict during sexualencounters between males and females from distantpopulations suggests that an incipient process of spe-ciation is occurring in C c cyanellus

Reproductive success was low in females fromGomez Farotildeas and in Gomez Farotildeas-Los Tuxtlas com-binations Pairs were fed THORNsh because this type offood has been successfully used in our C c cyanellusbreeding program (Favila 1993) However when pairsfrom Gomez Farotildeas and those from the Gomez Farotildeas-Los Tuxtlas combinations were feeding we observedthat after two to eight brood balls had been made thepair did not use the food for nesting thus reducingtheir fecundity In general C c cyanellus breed usingsmall vertebrate corpses (Halffter et al 1983 Favila andDotildeaz 1996) However Villalobos et al (1998) foundthat individuals from Gomez Farotildeas are able to eat andbreed by using carcasses ofOrthoporus ornatosGirard(Spirostreptida Spirostreptidae) a common diplopodin this region This behavior together with our resultssuggests that individuals from Gomez Farotildeas may havebecome very specialized in eating and nesting withDiplopoda as a local adaptation However althoughfecundity was low for the females from Gomez Farotildeasand the Gomez Farotildeas-Los Tuxtlas combinations off-spring survival was high so there is no evidence ofpostmating isolation in the interpopulation crossesUnfortunately the low number of eclosed beetles ob-tained in these combinations did not allow us to crossthe F2 generation

According to the ecological theory of speciationreproductive isolation ultimately evolves as a conse-quence of divergent natural selection on traits be-tween environments (Schluter 2001) If intrademicmate recognition systems are similar to species rec-ognition systems divergence in signaling systems canlead to premating isolation when geographically iso-

Fig 4 Tree representing cluster analysis for Canthon c cyanellus males and females in three populations from the eastcoast of Mexico on the basis of cuticular compounds TX Tuxtlas PP Papantla GF Gomez Farotildeas M male F female

1254 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 99 no 6

lated populations have secondary contact (Blows andAllan 1998 Greenberg et al 2003 Nosil 2004) Forexample hydrocarbons are involved in the mate rec-ognition system of Drosophila mojavensis Paterson(Diptera Drosophilidae) (Etges 1998) but this spe-cies shows regional variation associated with switch-ing hosts This has led to premating isolation amonggeographically isolated populations (Etges and Ahrens2001) In C c cyanellus the most important signalingsystem during sexual recognition is chemical via cu-ticular compounds (Ortiz-Domotildenguez et al 2006)Males and females from Gomez Farotildeas had the lowestnumber of cuticular compounds and were most ag-gressive with partners from the most distant popula-tion (Los Tuxtlas) which had the highest number ofcompounds The reduction in compounds could limitthe beetleOtildes ability to recognize individuals of theopposite sex from Los Tuxtlas which have the mostcomplex blend of cuticular compounds We hypoth-esize that switching in food preferences from car-casses of small vertebrates to Diplopoda carcasses inindividuals from the Gomez Farotildeas population couldmodify the quality and quantity of their cuticular com-pounds and produce the difTHORNculties observed in ac-cepting individuals from Los Tuxtlas and Palma SolaBecause Canthon c cyanellusmales recognize femalesby odor but females recognize individuals of the op-posite sex by other unknown behavioral cues as well(Ortotildez-Dominguez et al 2006) we do not discard thepossibility that another system of signals also could beinvolved in sexual recognition and these signals mayhave affected our results

Solotildes and Kohlmann (2002) suggest that allCanthoncyanellus subspecies from South America to Texasshould disappear However our results show that forthe subspecies C c cyanellus prezygotic isolationcould occur between distant populations when thereis secondary contact Owing to the ample distributionof C cyanellus in the Americas it is possible that thesouthern populations will have greater difTHORNculties insexual recognition if they come into contact withnorthern populations So we think that at presentit is not convenient to eliminate the subspecies ofC cyanellus but rather to reanalyze them in the con-text of Avise and Ball (1990) who consider subspeciesas members who share a unique geographic locale aset of phylogenetically concordant phenotypic char-acters and a unique natural history relative to othersubdivisions of the species but that are reproductivelycompatible with other such groups Phylogenetic dis-tinction must be based on the distributions of severalindependent genetically based traits (Miththapala etal 1996) not only on morphological traits as has beendone with C cyanellus

In summary the experiments here presented sug-gest that premating isolation is result of local adapta-tion related to a switch in food preferences If that isthe case future studies should analyze the factorsrelated with this switch Comparison with other south-ern populations will be crucial to determiningwhether this species is still cohesive or is in a processof THORNssion or whether there are even a complex of

cryptic species that are difTHORNcult to distinguish mor-phologically

Acknowledgments

We are grateful to Ingrid Marquez and Bianca Delfosse forlinguistic corrections and also to Carlos Fragoso for helpfulcomments and suggestions on the manuscript This study wassupported by Consejo Nacional de Ciencia y Tecnologia-Mexico Grants 4161P-N9607 35125-V and 49472-Q

References Cited

Avise J C and R M Ball 1990 Principles of genealogicalconcordance in species concepts and biological taxon-omy Oxford Surv Evol Biol 7 45ETH67

Blows M W and R A Allan 1998 Levels of mate recog-nition within and between two Drosophila species andtheir hybrids Am Nat 152 826ETH837

Boake CRB D K Price and D K Andreadis 1998 In-heritance of behavioral differences between two infer-tile sympatric species Drosophila silvestris and D het-eroneura Heredity 80 642ETH650

Del Campo M L S Via and M C Caillaud 2003 Rec-ognition of host-speciTHORNc chemical stimulants in two sym-patric host races of the pea aphid Acyrthosiphon pisumEcol Entomol 28 397ETH404

Etges W J 1998 Premating isolation is determined by lar-val-rearing substrates in cactophilic Drosophila mojaven-sis IV Correlated response in behavioral isolation toartiTHORNcial selection on a life-history trait Am Nat 152129ETH144

Etges W J 2002 Divergence in mate choice systems doesevolution play by rules Genetica 116 151ETH166

Etges W J and M A Ahrens 2001 Premating isolation isdetermined by larval-rearing substrates in cactophilicDrosophila mojavensis V Deep geographic variation inepicuticular hydrocarbons among isolated populationsAm Nat 158 585ETH598

Favila M E 1988a Comportamiento durante el periodode maduracion gonadica en un escarabajo rodador(Coleoptera Scarabaeidae Scarabaeinae) Folia EntomolMex 76 55ETH64

Favila M E 1988b Un metodo sencillo para marcar escar-abajos Folia Entomol Mex 75 117ETH118

Favila M E 1993 Some ecological factors affecting thelife-style of Canthon cyanellus cyanellus (ColeopteraScarabaeidae) an experimental approach Ethol EcolEvol 5 319ETH328

FavilaM E andADıaz 1996 Canthon cyanellus cyanellusLeConte (Coleoptera Scarabaeidae) makes a nest in theTHORNeld with several brood balls Coleopt Bull 50 52ETH60

GreenbergA J J RMoran J ACoyne andC IWu 2003Ecological adaptation during incipient speciation re-vealed by precise gene replacement Science (WashDC) 302 1754ETH1757

Halffter G 1961 Monografotildea de las especies norteameri-canas del genero Canthon Hoffsg (Coleopt Scarab)Ciencia 20 225ETH320

Halffter G M E Favila and V Halffter 1992 A compar-ative study of the structure of the scarab guild in Mexicantropical rain forest and derived ecosystems Folia Ento-mol Mex 84 131ETH156

Halffter G V Halffter and C Huerta 1983 Comporte-ment sexual et nidiTHORNcation chez Canthon cyanellus cya-nellus LeConte Bull Soc Entomol Mex 88 585ETH594

November 2006 ORTIZ-DOMINGUEZ ET AL MATE RECOGNITION IN ALLOPATRIC SCARABS 1255

Mendelson T C 2003 Sexual isolation evolves faster thanhybrid inviability in a diverse and sexually domorphicgenus of THORNsh (Percidae Etheostoma) Evolution 57 317ETH327

Michalak P JGrzesik and J Rafinski 1997 Test for sexualincompatibility between two new species Triturus vul-garis and Triturus montandoni no choice mating designEvolution 51 2045ETH2050

MillerGLGE StrattonPRMiller andEHebets 1998Geographical variation in male courtship behavior andsexual isolation in wolf spiders of the genus SchizocosaAnim Behav 56 937ETH951

Miththapala S J Seidensticker and S J OrsquoBrien 1996Phylogeographic subspecies recognition in leopards(Panthera pardus) molecular genetic variation ConservBiol 10 1115ETH1132

Noor MAF M A Williams D Alvarez and M Ruiz-Garcıa 2000 Lack of evolution divergence in courtshipsongs of Drosophila pseudoobscura subspecies J InsectBehav 13 255ETH262

Nosil P 2004 Reproductive isolation caused by visual pre-dation on migrants between divergent environmentsProc R Soc Lond B 271 1521ETH1528

Ortız-Domınguez M M E Favila M R Mendoza-LopezO Garcıa-Barradas and J Cruz-Sanchez 2006 The re-lationship between epicuticular compounds behavioralcues and sexual recognition in the ball roller scarabCanthon c cyanellus Entomol Exp Appl 119 23ETH27

Paillette M N Bizat and D Joly 1997 Differentiation ofdialects and courtship strategies in allopatric populationsof Drosophila teissieri J Insect Physiol 43 809ETH814

Panhuis T M R Butlin M Zuk and T Tregenza 2001Sexual selection and speciation Trends Ecol Evol 16364ETH371

Ritchie M G E J Halsey and M G Gleason 1999 Dro-sophila song as a species-speciTHORNc mating signal and thebehavioral importance of Kyriacou and Hall cycles inD melanogaster song Anim Behav 58 649ETH657

Rybak F G Sureau and T Aubin 2002 Functional cou-pling of acoustic and chemical signals in the courtshipbehavior of the male Drosophila melanogaster Proc RSoc Lond B 269 695ETH701

Schluter D 2000 The ecology of adaptive radiationOxford University Press Oxford United Kingdom

Schluter D 2001 Ecology and the origin of the speciesTrends Ecol Evol 16 372ETH380

Sokal R R and F J Rohlf 1981 Biometry 2nd ed W HFreeman San Francisco CA

Solıs A and B Kohlmann 2002 El genero Canthon (Co-leoptera Scarabaeidae) en Costa Rica G Ital Entomol10 11ETH35

Villalobos F J ADiaz andMEFavila 1998 Two speciesof Canthon Hoffmannsegg (Coleoptera Scarabaeidae)feed on dead and live invertebrates Coleopt Bull 52101ETH104

Vines T H and D Schluter 2006 Strong assortative mat-ing between allopatric sticklebacks as a by-product ofadaptation to different environments Proc Soc B 273911ETH916

Received 7 December 2005 accepted 2 August 2006

1256 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 99 no 6

Table 2 Cuticular compounds found in the elytra of Canthon c cyanellus males and females from Los Tuxtlas Papantla and GoacutemezFariacuteas

Retentiontime

Tuxtlasfemale

Tuxtlasmale

Papantlafemale

Papantlamale

Gomez Farotildeasfemale

Gomez Farotildeasmale

Sharedcompounds

1041 014 (086) 006 (015) 004 (080) 51061 003 (028) 046 (333) 017 (018ETH025) 008 (080) 11128 052 (478) 015 (033) 006 (090) 51199 007 (040) 056 (348) 013 (011ETH022) 009 (040) 11229 006 (058) 047 (332)1258 21 (1868)1272 005 (027)1279 004 (042) 004 (023)1309 01 (052) 027 (087)1321 032 (197) 2 (12) 055 (399) 033 (194) 006 (031) 23135 005 (029) 006 (062) 3145 015 (075) 004 (023)1464 042 (423) 116 (695) 055 (058) 021 (212) 002 (190) 04 (130) 121482 028 (177) 019 (095) 006 (056) 61507 008 (034)1529 083 (342) 116 (256) 045 (176) 043 (233) 046 (068) 016 (048) 121567 013 (129) 039 (252)1585 093 (274) 029 (173) 02 (069) 009 (040) 371595 157 (535) 009 (055)1658 007 (074) 008 (030)1679 009 (067)1702 038 (191) 082 (42)1715 581 (074ETH1372) 332 (028ETH1667) 235 (602) 263 (629) 014 (030) 05 (146) 121733 007 (065)1769 141 (051ETH296) 124 (040ETH457) 062 (108) 054 (101) 019 (077) 231789 16 (634)1803 011 (160)1815 263 (679) 053 (320) 036 (254) 152 (445) 009 (050) 231830 55 (136ETH1275 344 (053ETH1358 505 (1729 393 (933) 038 (135) 076 (024ETH217) 121851 323 (838) 061 (368) 076 (305) 148 (511) 007 (034) 231865 06 (299) 144 (566) 001 (060) 81879 184 (545) 092 (441) 31885 256 (1572) 424 (1711) 011 (052) 81892 13 (753) 05 (299)1900 052 (129) 048 (110) 075 (343) 003 (032) 005 (026) 03 (136) 121905 047 (254) 093 (666) 16 (529) 31924 604 (2447) 077 (469) 22 (884) 344 (1547) 361943 054 (227)196 351 (1128) 062 (373) 006 (030)) 104 (017ETH41) 471967 103 (419) 039 (234) 1472 (9678) 1848 (4854) 361979 073 (279) 029 (084) 359 (1128) 174 (473) 016 (079) 037 (172) 122007 84 (2859) 155 (928) 201 (593) 372 (957) 02 (098) 232025 016 (098) 019 (19) 04 (32) 62047 09 (438) 213 (695)2056 029 (287) 039 (233) 034 (092) 09 (349) 472062 97 (3886) 1684 (380)2074 021 (10) 163 (1075) 106 (45) 039 (145) 382084 309 (862) 05 (298) 359 (1396) 782 (1839) 362117 396 (1353)2129 016 (162)2146 102 (553) 179 (529)2167 782 (2483) 151 (978) 056 (247) 72173 67 (3218) 1085 (3574)2183 104 (1044)219 006 (030) 019 (093)2235 459 (1964) 082 (494) 118 (347) 259 (875) 362273 166 (1662)2278 176 (665)2306 061 (312) 162 (669)2323 013 (312) 009 (093)2348 028 (277) 098 (376)2442 1527 (7818) 46 (2760)2576 123 (529)2641 4 (529)

Mean concentration in micrograms per milliliter (minimum and maximum value) Minimum value is 0 except where another value isindicated

Italics indicate compounds exclusive to sex for a given population as follows 1 females from all the populations 2 males from all thepopulations 3 females from Los Tuxtlas and Papantla 4 females from Los Tuxtlas and Gomez Farotildeas 5 females from Papantla and GomezFarotildeas 6 males from Los Tuxtlas and Papantla 7 Males from Los Tuxtlas and Gomez Farotildeas 8 males from Papantla and Gomez Farotildeas

November 2006 ORTIZ-DOMINGUEZ ET AL MATE RECOGNITION IN ALLOPATRIC SCARABS 1253

ern population of Gomez Farotildeas had 24 compoundsNo compounds were exclusive to females whereas 11compounds were exclusive to males Thirteen com-pounds were shared (Table 2)

The comparison among populations showed thatindividuals of Los Tuxtlas had 18 compounds notpresent in individuals of Papantla or Gomez FarotildeasIndividuals of Gomez Farotildeas had only one com-pound shared by both sexes that was not present inthe Los Tuxtlas or Papantla populations Individualsfrom Papantla had 11 compounds not present in theGomez Farotildeas or Los Tuxtlas populations In generalboth sexes from the three populations shared six cu-ticular compounds The chemical analysis showed noquantitative differences in shared compounds be-tween the populations or between the sexes of eachpopulation (Table 2)

Two groups were formed with the Simple MatchingCoefTHORNcient (Fig 4) One group has females from LosTuxtlas The other group has two subgroups one withmales and females from Papantla and the other withmales and females from Gomez Farotildeas that were morerelated to males from Los Tuxtlas (Fig 4)

Discussion

In C c cyanellus the behavior during sexual en-counter and food ball rolling by intrapopulation pairswas mostly cooperative However although most ofthe pairs of individuals belonging to different localitiesalso were able to roll cooperatively noncooperativebehaviors increased signiTHORNcantly with distance sug-gesting that individuals prefer similar mates to matesfrom other populations This was very clear in maleETHfemale pairs of individuals from 600 km apart thatwere unwilling to roll a food ball with a partner eitherkeeping it away from the partner or THORNghting insteadof rolling the food ball cooperatively for nestingThese noncooperative behaviors are similar to those

observed between individuals of the same sex (Favila1988a) but in this case they act as behavioral precop-ulatory barriers The increased conszligict during sexualencounters between males and females from distantpopulations suggests that an incipient process of spe-ciation is occurring in C c cyanellus

Reproductive success was low in females fromGomez Farotildeas and in Gomez Farotildeas-Los Tuxtlas com-binations Pairs were fed THORNsh because this type offood has been successfully used in our C c cyanellusbreeding program (Favila 1993) However when pairsfrom Gomez Farotildeas and those from the Gomez Farotildeas-Los Tuxtlas combinations were feeding we observedthat after two to eight brood balls had been made thepair did not use the food for nesting thus reducingtheir fecundity In general C c cyanellus breed usingsmall vertebrate corpses (Halffter et al 1983 Favila andDotildeaz 1996) However Villalobos et al (1998) foundthat individuals from Gomez Farotildeas are able to eat andbreed by using carcasses ofOrthoporus ornatosGirard(Spirostreptida Spirostreptidae) a common diplopodin this region This behavior together with our resultssuggests that individuals from Gomez Farotildeas may havebecome very specialized in eating and nesting withDiplopoda as a local adaptation However althoughfecundity was low for the females from Gomez Farotildeasand the Gomez Farotildeas-Los Tuxtlas combinations off-spring survival was high so there is no evidence ofpostmating isolation in the interpopulation crossesUnfortunately the low number of eclosed beetles ob-tained in these combinations did not allow us to crossthe F2 generation

According to the ecological theory of speciationreproductive isolation ultimately evolves as a conse-quence of divergent natural selection on traits be-tween environments (Schluter 2001) If intrademicmate recognition systems are similar to species rec-ognition systems divergence in signaling systems canlead to premating isolation when geographically iso-

Fig 4 Tree representing cluster analysis for Canthon c cyanellus males and females in three populations from the eastcoast of Mexico on the basis of cuticular compounds TX Tuxtlas PP Papantla GF Gomez Farotildeas M male F female

1254 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 99 no 6

lated populations have secondary contact (Blows andAllan 1998 Greenberg et al 2003 Nosil 2004) Forexample hydrocarbons are involved in the mate rec-ognition system of Drosophila mojavensis Paterson(Diptera Drosophilidae) (Etges 1998) but this spe-cies shows regional variation associated with switch-ing hosts This has led to premating isolation amonggeographically isolated populations (Etges and Ahrens2001) In C c cyanellus the most important signalingsystem during sexual recognition is chemical via cu-ticular compounds (Ortiz-Domotildenguez et al 2006)Males and females from Gomez Farotildeas had the lowestnumber of cuticular compounds and were most ag-gressive with partners from the most distant popula-tion (Los Tuxtlas) which had the highest number ofcompounds The reduction in compounds could limitthe beetleOtildes ability to recognize individuals of theopposite sex from Los Tuxtlas which have the mostcomplex blend of cuticular compounds We hypoth-esize that switching in food preferences from car-casses of small vertebrates to Diplopoda carcasses inindividuals from the Gomez Farotildeas population couldmodify the quality and quantity of their cuticular com-pounds and produce the difTHORNculties observed in ac-cepting individuals from Los Tuxtlas and Palma SolaBecause Canthon c cyanellusmales recognize femalesby odor but females recognize individuals of the op-posite sex by other unknown behavioral cues as well(Ortotildez-Dominguez et al 2006) we do not discard thepossibility that another system of signals also could beinvolved in sexual recognition and these signals mayhave affected our results

Solotildes and Kohlmann (2002) suggest that allCanthoncyanellus subspecies from South America to Texasshould disappear However our results show that forthe subspecies C c cyanellus prezygotic isolationcould occur between distant populations when thereis secondary contact Owing to the ample distributionof C cyanellus in the Americas it is possible that thesouthern populations will have greater difTHORNculties insexual recognition if they come into contact withnorthern populations So we think that at presentit is not convenient to eliminate the subspecies ofC cyanellus but rather to reanalyze them in the con-text of Avise and Ball (1990) who consider subspeciesas members who share a unique geographic locale aset of phylogenetically concordant phenotypic char-acters and a unique natural history relative to othersubdivisions of the species but that are reproductivelycompatible with other such groups Phylogenetic dis-tinction must be based on the distributions of severalindependent genetically based traits (Miththapala etal 1996) not only on morphological traits as has beendone with C cyanellus

In summary the experiments here presented sug-gest that premating isolation is result of local adapta-tion related to a switch in food preferences If that isthe case future studies should analyze the factorsrelated with this switch Comparison with other south-ern populations will be crucial to determiningwhether this species is still cohesive or is in a processof THORNssion or whether there are even a complex of

cryptic species that are difTHORNcult to distinguish mor-phologically

Acknowledgments

We are grateful to Ingrid Marquez and Bianca Delfosse forlinguistic corrections and also to Carlos Fragoso for helpfulcomments and suggestions on the manuscript This study wassupported by Consejo Nacional de Ciencia y Tecnologia-Mexico Grants 4161P-N9607 35125-V and 49472-Q

References Cited

Avise J C and R M Ball 1990 Principles of genealogicalconcordance in species concepts and biological taxon-omy Oxford Surv Evol Biol 7 45ETH67

Blows M W and R A Allan 1998 Levels of mate recog-nition within and between two Drosophila species andtheir hybrids Am Nat 152 826ETH837

Boake CRB D K Price and D K Andreadis 1998 In-heritance of behavioral differences between two infer-tile sympatric species Drosophila silvestris and D het-eroneura Heredity 80 642ETH650

Del Campo M L S Via and M C Caillaud 2003 Rec-ognition of host-speciTHORNc chemical stimulants in two sym-patric host races of the pea aphid Acyrthosiphon pisumEcol Entomol 28 397ETH404

Etges W J 1998 Premating isolation is determined by lar-val-rearing substrates in cactophilic Drosophila mojaven-sis IV Correlated response in behavioral isolation toartiTHORNcial selection on a life-history trait Am Nat 152129ETH144

Etges W J 2002 Divergence in mate choice systems doesevolution play by rules Genetica 116 151ETH166

Etges W J and M A Ahrens 2001 Premating isolation isdetermined by larval-rearing substrates in cactophilicDrosophila mojavensis V Deep geographic variation inepicuticular hydrocarbons among isolated populationsAm Nat 158 585ETH598

Favila M E 1988a Comportamiento durante el periodode maduracion gonadica en un escarabajo rodador(Coleoptera Scarabaeidae Scarabaeinae) Folia EntomolMex 76 55ETH64

Favila M E 1988b Un metodo sencillo para marcar escar-abajos Folia Entomol Mex 75 117ETH118

Favila M E 1993 Some ecological factors affecting thelife-style of Canthon cyanellus cyanellus (ColeopteraScarabaeidae) an experimental approach Ethol EcolEvol 5 319ETH328

FavilaM E andADıaz 1996 Canthon cyanellus cyanellusLeConte (Coleoptera Scarabaeidae) makes a nest in theTHORNeld with several brood balls Coleopt Bull 50 52ETH60

GreenbergA J J RMoran J ACoyne andC IWu 2003Ecological adaptation during incipient speciation re-vealed by precise gene replacement Science (WashDC) 302 1754ETH1757

Halffter G 1961 Monografotildea de las especies norteameri-canas del genero Canthon Hoffsg (Coleopt Scarab)Ciencia 20 225ETH320

Halffter G M E Favila and V Halffter 1992 A compar-ative study of the structure of the scarab guild in Mexicantropical rain forest and derived ecosystems Folia Ento-mol Mex 84 131ETH156

Halffter G V Halffter and C Huerta 1983 Comporte-ment sexual et nidiTHORNcation chez Canthon cyanellus cya-nellus LeConte Bull Soc Entomol Mex 88 585ETH594

November 2006 ORTIZ-DOMINGUEZ ET AL MATE RECOGNITION IN ALLOPATRIC SCARABS 1255

Mendelson T C 2003 Sexual isolation evolves faster thanhybrid inviability in a diverse and sexually domorphicgenus of THORNsh (Percidae Etheostoma) Evolution 57 317ETH327

Michalak P JGrzesik and J Rafinski 1997 Test for sexualincompatibility between two new species Triturus vul-garis and Triturus montandoni no choice mating designEvolution 51 2045ETH2050

MillerGLGE StrattonPRMiller andEHebets 1998Geographical variation in male courtship behavior andsexual isolation in wolf spiders of the genus SchizocosaAnim Behav 56 937ETH951

Miththapala S J Seidensticker and S J OrsquoBrien 1996Phylogeographic subspecies recognition in leopards(Panthera pardus) molecular genetic variation ConservBiol 10 1115ETH1132

Noor MAF M A Williams D Alvarez and M Ruiz-Garcıa 2000 Lack of evolution divergence in courtshipsongs of Drosophila pseudoobscura subspecies J InsectBehav 13 255ETH262

Nosil P 2004 Reproductive isolation caused by visual pre-dation on migrants between divergent environmentsProc R Soc Lond B 271 1521ETH1528

Ortız-Domınguez M M E Favila M R Mendoza-LopezO Garcıa-Barradas and J Cruz-Sanchez 2006 The re-lationship between epicuticular compounds behavioralcues and sexual recognition in the ball roller scarabCanthon c cyanellus Entomol Exp Appl 119 23ETH27

Paillette M N Bizat and D Joly 1997 Differentiation ofdialects and courtship strategies in allopatric populationsof Drosophila teissieri J Insect Physiol 43 809ETH814

Panhuis T M R Butlin M Zuk and T Tregenza 2001Sexual selection and speciation Trends Ecol Evol 16364ETH371

Ritchie M G E J Halsey and M G Gleason 1999 Dro-sophila song as a species-speciTHORNc mating signal and thebehavioral importance of Kyriacou and Hall cycles inD melanogaster song Anim Behav 58 649ETH657

Rybak F G Sureau and T Aubin 2002 Functional cou-pling of acoustic and chemical signals in the courtshipbehavior of the male Drosophila melanogaster Proc RSoc Lond B 269 695ETH701

Schluter D 2000 The ecology of adaptive radiationOxford University Press Oxford United Kingdom

Schluter D 2001 Ecology and the origin of the speciesTrends Ecol Evol 16 372ETH380

Sokal R R and F J Rohlf 1981 Biometry 2nd ed W HFreeman San Francisco CA

Solıs A and B Kohlmann 2002 El genero Canthon (Co-leoptera Scarabaeidae) en Costa Rica G Ital Entomol10 11ETH35

Villalobos F J ADiaz andMEFavila 1998 Two speciesof Canthon Hoffmannsegg (Coleoptera Scarabaeidae)feed on dead and live invertebrates Coleopt Bull 52101ETH104

Vines T H and D Schluter 2006 Strong assortative mat-ing between allopatric sticklebacks as a by-product ofadaptation to different environments Proc Soc B 273911ETH916

Received 7 December 2005 accepted 2 August 2006

1256 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 99 no 6

ern population of Gomez Farotildeas had 24 compoundsNo compounds were exclusive to females whereas 11compounds were exclusive to males Thirteen com-pounds were shared (Table 2)

The comparison among populations showed thatindividuals of Los Tuxtlas had 18 compounds notpresent in individuals of Papantla or Gomez FarotildeasIndividuals of Gomez Farotildeas had only one com-pound shared by both sexes that was not present inthe Los Tuxtlas or Papantla populations Individualsfrom Papantla had 11 compounds not present in theGomez Farotildeas or Los Tuxtlas populations In generalboth sexes from the three populations shared six cu-ticular compounds The chemical analysis showed noquantitative differences in shared compounds be-tween the populations or between the sexes of eachpopulation (Table 2)

Two groups were formed with the Simple MatchingCoefTHORNcient (Fig 4) One group has females from LosTuxtlas The other group has two subgroups one withmales and females from Papantla and the other withmales and females from Gomez Farotildeas that were morerelated to males from Los Tuxtlas (Fig 4)

Discussion

In C c cyanellus the behavior during sexual en-counter and food ball rolling by intrapopulation pairswas mostly cooperative However although most ofthe pairs of individuals belonging to different localitiesalso were able to roll cooperatively noncooperativebehaviors increased signiTHORNcantly with distance sug-gesting that individuals prefer similar mates to matesfrom other populations This was very clear in maleETHfemale pairs of individuals from 600 km apart thatwere unwilling to roll a food ball with a partner eitherkeeping it away from the partner or THORNghting insteadof rolling the food ball cooperatively for nestingThese noncooperative behaviors are similar to those

observed between individuals of the same sex (Favila1988a) but in this case they act as behavioral precop-ulatory barriers The increased conszligict during sexualencounters between males and females from distantpopulations suggests that an incipient process of spe-ciation is occurring in C c cyanellus

Reproductive success was low in females fromGomez Farotildeas and in Gomez Farotildeas-Los Tuxtlas com-binations Pairs were fed THORNsh because this type offood has been successfully used in our C c cyanellusbreeding program (Favila 1993) However when pairsfrom Gomez Farotildeas and those from the Gomez Farotildeas-Los Tuxtlas combinations were feeding we observedthat after two to eight brood balls had been made thepair did not use the food for nesting thus reducingtheir fecundity In general C c cyanellus breed usingsmall vertebrate corpses (Halffter et al 1983 Favila andDotildeaz 1996) However Villalobos et al (1998) foundthat individuals from Gomez Farotildeas are able to eat andbreed by using carcasses ofOrthoporus ornatosGirard(Spirostreptida Spirostreptidae) a common diplopodin this region This behavior together with our resultssuggests that individuals from Gomez Farotildeas may havebecome very specialized in eating and nesting withDiplopoda as a local adaptation However althoughfecundity was low for the females from Gomez Farotildeasand the Gomez Farotildeas-Los Tuxtlas combinations off-spring survival was high so there is no evidence ofpostmating isolation in the interpopulation crossesUnfortunately the low number of eclosed beetles ob-tained in these combinations did not allow us to crossthe F2 generation

According to the ecological theory of speciationreproductive isolation ultimately evolves as a conse-quence of divergent natural selection on traits be-tween environments (Schluter 2001) If intrademicmate recognition systems are similar to species rec-ognition systems divergence in signaling systems canlead to premating isolation when geographically iso-

Fig 4 Tree representing cluster analysis for Canthon c cyanellus males and females in three populations from the eastcoast of Mexico on the basis of cuticular compounds TX Tuxtlas PP Papantla GF Gomez Farotildeas M male F female

1254 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 99 no 6

lated populations have secondary contact (Blows andAllan 1998 Greenberg et al 2003 Nosil 2004) Forexample hydrocarbons are involved in the mate rec-ognition system of Drosophila mojavensis Paterson(Diptera Drosophilidae) (Etges 1998) but this spe-cies shows regional variation associated with switch-ing hosts This has led to premating isolation amonggeographically isolated populations (Etges and Ahrens2001) In C c cyanellus the most important signalingsystem during sexual recognition is chemical via cu-ticular compounds (Ortiz-Domotildenguez et al 2006)Males and females from Gomez Farotildeas had the lowestnumber of cuticular compounds and were most ag-gressive with partners from the most distant popula-tion (Los Tuxtlas) which had the highest number ofcompounds The reduction in compounds could limitthe beetleOtildes ability to recognize individuals of theopposite sex from Los Tuxtlas which have the mostcomplex blend of cuticular compounds We hypoth-esize that switching in food preferences from car-casses of small vertebrates to Diplopoda carcasses inindividuals from the Gomez Farotildeas population couldmodify the quality and quantity of their cuticular com-pounds and produce the difTHORNculties observed in ac-cepting individuals from Los Tuxtlas and Palma SolaBecause Canthon c cyanellusmales recognize femalesby odor but females recognize individuals of the op-posite sex by other unknown behavioral cues as well(Ortotildez-Dominguez et al 2006) we do not discard thepossibility that another system of signals also could beinvolved in sexual recognition and these signals mayhave affected our results

Solotildes and Kohlmann (2002) suggest that allCanthoncyanellus subspecies from South America to Texasshould disappear However our results show that forthe subspecies C c cyanellus prezygotic isolationcould occur between distant populations when thereis secondary contact Owing to the ample distributionof C cyanellus in the Americas it is possible that thesouthern populations will have greater difTHORNculties insexual recognition if they come into contact withnorthern populations So we think that at presentit is not convenient to eliminate the subspecies ofC cyanellus but rather to reanalyze them in the con-text of Avise and Ball (1990) who consider subspeciesas members who share a unique geographic locale aset of phylogenetically concordant phenotypic char-acters and a unique natural history relative to othersubdivisions of the species but that are reproductivelycompatible with other such groups Phylogenetic dis-tinction must be based on the distributions of severalindependent genetically based traits (Miththapala etal 1996) not only on morphological traits as has beendone with C cyanellus

In summary the experiments here presented sug-gest that premating isolation is result of local adapta-tion related to a switch in food preferences If that isthe case future studies should analyze the factorsrelated with this switch Comparison with other south-ern populations will be crucial to determiningwhether this species is still cohesive or is in a processof THORNssion or whether there are even a complex of

cryptic species that are difTHORNcult to distinguish mor-phologically

Acknowledgments

We are grateful to Ingrid Marquez and Bianca Delfosse forlinguistic corrections and also to Carlos Fragoso for helpfulcomments and suggestions on the manuscript This study wassupported by Consejo Nacional de Ciencia y Tecnologia-Mexico Grants 4161P-N9607 35125-V and 49472-Q

References Cited

Avise J C and R M Ball 1990 Principles of genealogicalconcordance in species concepts and biological taxon-omy Oxford Surv Evol Biol 7 45ETH67

Blows M W and R A Allan 1998 Levels of mate recog-nition within and between two Drosophila species andtheir hybrids Am Nat 152 826ETH837

Boake CRB D K Price and D K Andreadis 1998 In-heritance of behavioral differences between two infer-tile sympatric species Drosophila silvestris and D het-eroneura Heredity 80 642ETH650

Del Campo M L S Via and M C Caillaud 2003 Rec-ognition of host-speciTHORNc chemical stimulants in two sym-patric host races of the pea aphid Acyrthosiphon pisumEcol Entomol 28 397ETH404

Etges W J 1998 Premating isolation is determined by lar-val-rearing substrates in cactophilic Drosophila mojaven-sis IV Correlated response in behavioral isolation toartiTHORNcial selection on a life-history trait Am Nat 152129ETH144

Etges W J 2002 Divergence in mate choice systems doesevolution play by rules Genetica 116 151ETH166

Etges W J and M A Ahrens 2001 Premating isolation isdetermined by larval-rearing substrates in cactophilicDrosophila mojavensis V Deep geographic variation inepicuticular hydrocarbons among isolated populationsAm Nat 158 585ETH598

Favila M E 1988a Comportamiento durante el periodode maduracion gonadica en un escarabajo rodador(Coleoptera Scarabaeidae Scarabaeinae) Folia EntomolMex 76 55ETH64

Favila M E 1988b Un metodo sencillo para marcar escar-abajos Folia Entomol Mex 75 117ETH118

Favila M E 1993 Some ecological factors affecting thelife-style of Canthon cyanellus cyanellus (ColeopteraScarabaeidae) an experimental approach Ethol EcolEvol 5 319ETH328

FavilaM E andADıaz 1996 Canthon cyanellus cyanellusLeConte (Coleoptera Scarabaeidae) makes a nest in theTHORNeld with several brood balls Coleopt Bull 50 52ETH60

GreenbergA J J RMoran J ACoyne andC IWu 2003Ecological adaptation during incipient speciation re-vealed by precise gene replacement Science (WashDC) 302 1754ETH1757

Halffter G 1961 Monografotildea de las especies norteameri-canas del genero Canthon Hoffsg (Coleopt Scarab)Ciencia 20 225ETH320

Halffter G M E Favila and V Halffter 1992 A compar-ative study of the structure of the scarab guild in Mexicantropical rain forest and derived ecosystems Folia Ento-mol Mex 84 131ETH156

Halffter G V Halffter and C Huerta 1983 Comporte-ment sexual et nidiTHORNcation chez Canthon cyanellus cya-nellus LeConte Bull Soc Entomol Mex 88 585ETH594

November 2006 ORTIZ-DOMINGUEZ ET AL MATE RECOGNITION IN ALLOPATRIC SCARABS 1255

Mendelson T C 2003 Sexual isolation evolves faster thanhybrid inviability in a diverse and sexually domorphicgenus of THORNsh (Percidae Etheostoma) Evolution 57 317ETH327

Michalak P JGrzesik and J Rafinski 1997 Test for sexualincompatibility between two new species Triturus vul-garis and Triturus montandoni no choice mating designEvolution 51 2045ETH2050

MillerGLGE StrattonPRMiller andEHebets 1998Geographical variation in male courtship behavior andsexual isolation in wolf spiders of the genus SchizocosaAnim Behav 56 937ETH951

Miththapala S J Seidensticker and S J OrsquoBrien 1996Phylogeographic subspecies recognition in leopards(Panthera pardus) molecular genetic variation ConservBiol 10 1115ETH1132

Noor MAF M A Williams D Alvarez and M Ruiz-Garcıa 2000 Lack of evolution divergence in courtshipsongs of Drosophila pseudoobscura subspecies J InsectBehav 13 255ETH262

Nosil P 2004 Reproductive isolation caused by visual pre-dation on migrants between divergent environmentsProc R Soc Lond B 271 1521ETH1528

Ortız-Domınguez M M E Favila M R Mendoza-LopezO Garcıa-Barradas and J Cruz-Sanchez 2006 The re-lationship between epicuticular compounds behavioralcues and sexual recognition in the ball roller scarabCanthon c cyanellus Entomol Exp Appl 119 23ETH27

Paillette M N Bizat and D Joly 1997 Differentiation ofdialects and courtship strategies in allopatric populationsof Drosophila teissieri J Insect Physiol 43 809ETH814

Panhuis T M R Butlin M Zuk and T Tregenza 2001Sexual selection and speciation Trends Ecol Evol 16364ETH371

Ritchie M G E J Halsey and M G Gleason 1999 Dro-sophila song as a species-speciTHORNc mating signal and thebehavioral importance of Kyriacou and Hall cycles inD melanogaster song Anim Behav 58 649ETH657

Rybak F G Sureau and T Aubin 2002 Functional cou-pling of acoustic and chemical signals in the courtshipbehavior of the male Drosophila melanogaster Proc RSoc Lond B 269 695ETH701

Schluter D 2000 The ecology of adaptive radiationOxford University Press Oxford United Kingdom

Schluter D 2001 Ecology and the origin of the speciesTrends Ecol Evol 16 372ETH380

Sokal R R and F J Rohlf 1981 Biometry 2nd ed W HFreeman San Francisco CA

Solıs A and B Kohlmann 2002 El genero Canthon (Co-leoptera Scarabaeidae) en Costa Rica G Ital Entomol10 11ETH35

Villalobos F J ADiaz andMEFavila 1998 Two speciesof Canthon Hoffmannsegg (Coleoptera Scarabaeidae)feed on dead and live invertebrates Coleopt Bull 52101ETH104

Vines T H and D Schluter 2006 Strong assortative mat-ing between allopatric sticklebacks as a by-product ofadaptation to different environments Proc Soc B 273911ETH916

Received 7 December 2005 accepted 2 August 2006

1256 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 99 no 6

lated populations have secondary contact (Blows andAllan 1998 Greenberg et al 2003 Nosil 2004) Forexample hydrocarbons are involved in the mate rec-ognition system of Drosophila mojavensis Paterson(Diptera Drosophilidae) (Etges 1998) but this spe-cies shows regional variation associated with switch-ing hosts This has led to premating isolation amonggeographically isolated populations (Etges and Ahrens2001) In C c cyanellus the most important signalingsystem during sexual recognition is chemical via cu-ticular compounds (Ortiz-Domotildenguez et al 2006)Males and females from Gomez Farotildeas had the lowestnumber of cuticular compounds and were most ag-gressive with partners from the most distant popula-tion (Los Tuxtlas) which had the highest number ofcompounds The reduction in compounds could limitthe beetleOtildes ability to recognize individuals of theopposite sex from Los Tuxtlas which have the mostcomplex blend of cuticular compounds We hypoth-esize that switching in food preferences from car-casses of small vertebrates to Diplopoda carcasses inindividuals from the Gomez Farotildeas population couldmodify the quality and quantity of their cuticular com-pounds and produce the difTHORNculties observed in ac-cepting individuals from Los Tuxtlas and Palma SolaBecause Canthon c cyanellusmales recognize femalesby odor but females recognize individuals of the op-posite sex by other unknown behavioral cues as well(Ortotildez-Dominguez et al 2006) we do not discard thepossibility that another system of signals also could beinvolved in sexual recognition and these signals mayhave affected our results

Solotildes and Kohlmann (2002) suggest that allCanthoncyanellus subspecies from South America to Texasshould disappear However our results show that forthe subspecies C c cyanellus prezygotic isolationcould occur between distant populations when thereis secondary contact Owing to the ample distributionof C cyanellus in the Americas it is possible that thesouthern populations will have greater difTHORNculties insexual recognition if they come into contact withnorthern populations So we think that at presentit is not convenient to eliminate the subspecies ofC cyanellus but rather to reanalyze them in the con-text of Avise and Ball (1990) who consider subspeciesas members who share a unique geographic locale aset of phylogenetically concordant phenotypic char-acters and a unique natural history relative to othersubdivisions of the species but that are reproductivelycompatible with other such groups Phylogenetic dis-tinction must be based on the distributions of severalindependent genetically based traits (Miththapala etal 1996) not only on morphological traits as has beendone with C cyanellus

In summary the experiments here presented sug-gest that premating isolation is result of local adapta-tion related to a switch in food preferences If that isthe case future studies should analyze the factorsrelated with this switch Comparison with other south-ern populations will be crucial to determiningwhether this species is still cohesive or is in a processof THORNssion or whether there are even a complex of

cryptic species that are difTHORNcult to distinguish mor-phologically

Acknowledgments

We are grateful to Ingrid Marquez and Bianca Delfosse forlinguistic corrections and also to Carlos Fragoso for helpfulcomments and suggestions on the manuscript This study wassupported by Consejo Nacional de Ciencia y Tecnologia-Mexico Grants 4161P-N9607 35125-V and 49472-Q

References Cited

Avise J C and R M Ball 1990 Principles of genealogicalconcordance in species concepts and biological taxon-omy Oxford Surv Evol Biol 7 45ETH67

Blows M W and R A Allan 1998 Levels of mate recog-nition within and between two Drosophila species andtheir hybrids Am Nat 152 826ETH837

Boake CRB D K Price and D K Andreadis 1998 In-heritance of behavioral differences between two infer-tile sympatric species Drosophila silvestris and D het-eroneura Heredity 80 642ETH650

Del Campo M L S Via and M C Caillaud 2003 Rec-ognition of host-speciTHORNc chemical stimulants in two sym-patric host races of the pea aphid Acyrthosiphon pisumEcol Entomol 28 397ETH404

Etges W J 1998 Premating isolation is determined by lar-val-rearing substrates in cactophilic Drosophila mojaven-sis IV Correlated response in behavioral isolation toartiTHORNcial selection on a life-history trait Am Nat 152129ETH144

Etges W J 2002 Divergence in mate choice systems doesevolution play by rules Genetica 116 151ETH166

Etges W J and M A Ahrens 2001 Premating isolation isdetermined by larval-rearing substrates in cactophilicDrosophila mojavensis V Deep geographic variation inepicuticular hydrocarbons among isolated populationsAm Nat 158 585ETH598

Favila M E 1988a Comportamiento durante el periodode maduracion gonadica en un escarabajo rodador(Coleoptera Scarabaeidae Scarabaeinae) Folia EntomolMex 76 55ETH64

Favila M E 1988b Un metodo sencillo para marcar escar-abajos Folia Entomol Mex 75 117ETH118

Favila M E 1993 Some ecological factors affecting thelife-style of Canthon cyanellus cyanellus (ColeopteraScarabaeidae) an experimental approach Ethol EcolEvol 5 319ETH328

FavilaM E andADıaz 1996 Canthon cyanellus cyanellusLeConte (Coleoptera Scarabaeidae) makes a nest in theTHORNeld with several brood balls Coleopt Bull 50 52ETH60

GreenbergA J J RMoran J ACoyne andC IWu 2003Ecological adaptation during incipient speciation re-vealed by precise gene replacement Science (WashDC) 302 1754ETH1757

Halffter G 1961 Monografotildea de las especies norteameri-canas del genero Canthon Hoffsg (Coleopt Scarab)Ciencia 20 225ETH320

Halffter G M E Favila and V Halffter 1992 A compar-ative study of the structure of the scarab guild in Mexicantropical rain forest and derived ecosystems Folia Ento-mol Mex 84 131ETH156

Halffter G V Halffter and C Huerta 1983 Comporte-ment sexual et nidiTHORNcation chez Canthon cyanellus cya-nellus LeConte Bull Soc Entomol Mex 88 585ETH594

November 2006 ORTIZ-DOMINGUEZ ET AL MATE RECOGNITION IN ALLOPATRIC SCARABS 1255

Mendelson T C 2003 Sexual isolation evolves faster thanhybrid inviability in a diverse and sexually domorphicgenus of THORNsh (Percidae Etheostoma) Evolution 57 317ETH327

Michalak P JGrzesik and J Rafinski 1997 Test for sexualincompatibility between two new species Triturus vul-garis and Triturus montandoni no choice mating designEvolution 51 2045ETH2050

MillerGLGE StrattonPRMiller andEHebets 1998Geographical variation in male courtship behavior andsexual isolation in wolf spiders of the genus SchizocosaAnim Behav 56 937ETH951

Miththapala S J Seidensticker and S J OrsquoBrien 1996Phylogeographic subspecies recognition in leopards(Panthera pardus) molecular genetic variation ConservBiol 10 1115ETH1132

Noor MAF M A Williams D Alvarez and M Ruiz-Garcıa 2000 Lack of evolution divergence in courtshipsongs of Drosophila pseudoobscura subspecies J InsectBehav 13 255ETH262

Nosil P 2004 Reproductive isolation caused by visual pre-dation on migrants between divergent environmentsProc R Soc Lond B 271 1521ETH1528

Ortız-Domınguez M M E Favila M R Mendoza-LopezO Garcıa-Barradas and J Cruz-Sanchez 2006 The re-lationship between epicuticular compounds behavioralcues and sexual recognition in the ball roller scarabCanthon c cyanellus Entomol Exp Appl 119 23ETH27

Paillette M N Bizat and D Joly 1997 Differentiation ofdialects and courtship strategies in allopatric populationsof Drosophila teissieri J Insect Physiol 43 809ETH814

Panhuis T M R Butlin M Zuk and T Tregenza 2001Sexual selection and speciation Trends Ecol Evol 16364ETH371

Ritchie M G E J Halsey and M G Gleason 1999 Dro-sophila song as a species-speciTHORNc mating signal and thebehavioral importance of Kyriacou and Hall cycles inD melanogaster song Anim Behav 58 649ETH657

Rybak F G Sureau and T Aubin 2002 Functional cou-pling of acoustic and chemical signals in the courtshipbehavior of the male Drosophila melanogaster Proc RSoc Lond B 269 695ETH701

Schluter D 2000 The ecology of adaptive radiationOxford University Press Oxford United Kingdom

Schluter D 2001 Ecology and the origin of the speciesTrends Ecol Evol 16 372ETH380

Sokal R R and F J Rohlf 1981 Biometry 2nd ed W HFreeman San Francisco CA

Solıs A and B Kohlmann 2002 El genero Canthon (Co-leoptera Scarabaeidae) en Costa Rica G Ital Entomol10 11ETH35

Villalobos F J ADiaz andMEFavila 1998 Two speciesof Canthon Hoffmannsegg (Coleoptera Scarabaeidae)feed on dead and live invertebrates Coleopt Bull 52101ETH104

Vines T H and D Schluter 2006 Strong assortative mat-ing between allopatric sticklebacks as a by-product ofadaptation to different environments Proc Soc B 273911ETH916

Received 7 December 2005 accepted 2 August 2006

1256 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 99 no 6

Mendelson T C 2003 Sexual isolation evolves faster thanhybrid inviability in a diverse and sexually domorphicgenus of THORNsh (Percidae Etheostoma) Evolution 57 317ETH327

Michalak P JGrzesik and J Rafinski 1997 Test for sexualincompatibility between two new species Triturus vul-garis and Triturus montandoni no choice mating designEvolution 51 2045ETH2050

MillerGLGE StrattonPRMiller andEHebets 1998Geographical variation in male courtship behavior andsexual isolation in wolf spiders of the genus SchizocosaAnim Behav 56 937ETH951

Miththapala S J Seidensticker and S J OrsquoBrien 1996Phylogeographic subspecies recognition in leopards(Panthera pardus) molecular genetic variation ConservBiol 10 1115ETH1132

Noor MAF M A Williams D Alvarez and M Ruiz-Garcıa 2000 Lack of evolution divergence in courtshipsongs of Drosophila pseudoobscura subspecies J InsectBehav 13 255ETH262

Nosil P 2004 Reproductive isolation caused by visual pre-dation on migrants between divergent environmentsProc R Soc Lond B 271 1521ETH1528

Ortız-Domınguez M M E Favila M R Mendoza-LopezO Garcıa-Barradas and J Cruz-Sanchez 2006 The re-lationship between epicuticular compounds behavioralcues and sexual recognition in the ball roller scarabCanthon c cyanellus Entomol Exp Appl 119 23ETH27

Paillette M N Bizat and D Joly 1997 Differentiation ofdialects and courtship strategies in allopatric populationsof Drosophila teissieri J Insect Physiol 43 809ETH814

Panhuis T M R Butlin M Zuk and T Tregenza 2001Sexual selection and speciation Trends Ecol Evol 16364ETH371

Ritchie M G E J Halsey and M G Gleason 1999 Dro-sophila song as a species-speciTHORNc mating signal and thebehavioral importance of Kyriacou and Hall cycles inD melanogaster song Anim Behav 58 649ETH657

Rybak F G Sureau and T Aubin 2002 Functional cou-pling of acoustic and chemical signals in the courtshipbehavior of the male Drosophila melanogaster Proc RSoc Lond B 269 695ETH701

Schluter D 2000 The ecology of adaptive radiationOxford University Press Oxford United Kingdom

Schluter D 2001 Ecology and the origin of the speciesTrends Ecol Evol 16 372ETH380

Sokal R R and F J Rohlf 1981 Biometry 2nd ed W HFreeman San Francisco CA

Solıs A and B Kohlmann 2002 El genero Canthon (Co-leoptera Scarabaeidae) en Costa Rica G Ital Entomol10 11ETH35

Villalobos F J ADiaz andMEFavila 1998 Two speciesof Canthon Hoffmannsegg (Coleoptera Scarabaeidae)feed on dead and live invertebrates Coleopt Bull 52101ETH104

Vines T H and D Schluter 2006 Strong assortative mat-ing between allopatric sticklebacks as a by-product ofadaptation to different environments Proc Soc B 273911ETH916

Received 7 December 2005 accepted 2 August 2006

1256 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 99 no 6