J. J-J YM RES. V nl. 11 (2), ?OO?, pp. 3"12- 325

Phylogeny of the Genera of Ticoplinae (Hymenoptera: Mutillidae)

ANDREW MITCl-:I ELL AND D ENIS J. BROTH ERS

School of Botany and Zoology, University of Natal Piete rma ritzburg, Pri va te Bag XOl, Scottsville, 3209 South Africa (AM now: School of Molecular and Cellular Biosciences);

AM em ail: mitchella@nu .ac.za DJB email: brothers@nu.ac.za

Abstmct.- The subfamily Ticoplinae Nagy is one of the more basa l taxa in Mutillidae. Cladistic ana lyses us ing ?1 characte rs have demons trated that recognition of two tribes is supported. Ti­coplini Nagy includes NniiOJJilltillo And re ( = Ticopla Nagy syn. nov.) and Areotilla Bischoff; Smi­cromyrmi ll ini Argaman includes SllliCmlllyrlllilla Suarcz (and possibly Cantcro11illn Lelej a nd Krom­bein, Eoslllicrolllynuillo Lelej and Kron1bein, and Hi11dustanilla Lelej and Krombein, should these be consid ered valid). The subfamily a nd tribes are rev.icwed and characterized, a key to tribes and genera is provided, and both sexes of typical members of the three main genera are illus trated. Na nonllltilln 11ndnc Arga man 1988 is selected as the correct spelling for the species also spelled N. nadn in its original description, and Arcotilln fcrmgillcn Mitchell and Brothers 1998 for the species also spelled Arcotillo ferruginatn in its o rig ina 1 description.

Ticoplinae Nagy 1970 is one of the rel­a tively basal subfamilies of Mutillidae, as shown by Brothers (1975, 1999) (Fig. 1) and by Lelej and Ne1nkov (1997) whose analyses differed from Brothers ' in several respects. It was established as a subfamily of Nagy's Heterogynidae (properly Het­e rogynaidae; International Commission on Zoological NmTtenclature 1987), to ac­commodate the genus Ticoplo Nagy 1970, known only fron1 1nale speci1nens. Broth­ers (1975) transferred Ticoplinae to Mutil­lidae, and Day (1984) placed Heterogyno Nagy in Sphecidae s.l.; it is now consid­ered to cmnprise a distinct fam ily, Heter­ogynaidae (Brothers 1999, Melo 1999). The genera placed in Ticoplinae by Brothers (1975) were Areotilln Bischoff 1920, Nnno­mutilln And re J 900, Snlicrontyrmil/a Suarez J 965 and Ticoplo. Independently, Suarez (J975) proposed a new s ubfamily, Nano­mutillinae, to contain Nnnomutillo, and placed Smicro/1/ynni//a in Myn11illinae but he did not rea li se that Ticopln or Arcotil/o were of re leva nce. Brothers' (1975) study settled much of the controve rsy over the classification of these genera by showing

that they belong in a single s ubfamily, the valid name of which is Ticoplinae (and would re.main so even if Ticopln were con­sidered a junior synonym of Nanomutilln; International Cmn1nission on Zoological Nomenclatu re 1999: Article 40.1). He also concluded that the relationships an1ongst the component genera were such as to preclude the recognition of tribal divi­sions. However, Argama n (1988) pro­posed such divisions: Ticoplini, including Ticopln and Nn nonllltilln, which he consid­ered dis tinct, and Smicron1yrmillini, in­cluding Smicromyrmillo. He d id not ex­amine Arcotillo .

The clad is tic study presented here elu­cida tes the phylogeny of the genera of Ti­coplinae, enabling objective assessmen t of Argaman's tribal d iv isions. The n1orpho­logica1 terms used are those of Gauld and Bolton (1988). Specimens exan1ined are in the collection of one of the authors (DJB) or were borrowed frorn numerous insti ­tutio ns (particularly the N atural History Museum (London) and Museum National d'Histoire Natu re lle (Paris)) over many yea rs.

VOLUM E 11, UMBL:.R 2, 2002

-- Myrmosinae

-- Pseudophotopsidinae

-- Ticoplinae

-- Rhopalomutillinae

-- Sphaeropthalminae

-- Mutillinae

fig. 1. Phylogeny of subfam ilies of Mutillidae (sim­pli fied from Brothers 1975, 1999).

GENERA OF TICOPLINAE

N anonwtilla Andre (Figs. 2-9)

Mutilln (NanoJJlutilln) Andre 1900: 130. Type species: Mutilln vnucheri Tournjer 1895, Morocco, by subsequent designation of Ash­mead (1903).

Ticopln Nagy 1970: 85. Type species: Ticopln yoca Nagy 1970, Jordan, by or iginal designation. Syn. nov.

The first description of Nnnonwtilln ap­peared in a key to subgenera of Mutilla published in April1900, without any n1en­tion of included species. Later, in the san1e work, Andre (190l b: 223) p resented a for­n1al description based on a single species known fron1 fen1ales on ly (Mutilln vnuclz­eri) but also provided a description of the 1nale in a footnote (p. 224), based on a sec­on d species su pposedly known from both sexes (Muti!ln lllicrosOIII17) which he had re­cently described; it nllls t thus be conclud­ed that there were two originally-included s pecies, al though subsequent authors h ave considered Nanomutilln to ha ve b een a n1onotypic genus at establjshment. An­dre (1901a) described both sexes of Mutilln (Nnnoni/Ltilln) IJiicrosomn from South Africa, being under the im pression that the male and fe1nale specimens had been collected in the same area, Willowmore (" ... j'ai trouve une autre espece du meme sous­genre ... accompagnee d 'un male rencon­tre dans les mem es parages"). Thjs, along

313

with their similarity in size, convinced h in1 that the two specimens were consp e­cific, despite tbe fa ct that they were not captured in copula. We have exa mined the type series of M. (N.) 111 icrosomn in the Transvaal Museum and found that the fe­m ale specin1en was actually collected at Bothaville, in the Free State, while the male was collected near Willowmore, in the Eas tern Cape, approximately 700 km away. They a re not conspeci fie, nor even congeneric; Nonvciller (1973) concluded that the male really belongs in Smicrolllyr­milln. To add to the confusion, Bischoff (1921) described s ix new species of Nnno­nllltilln without examining the type spe­cies; all of his n.ew sp ecies would later be recognised as belonging in Smicrolllyrlllifla (Nonveiller 1973). Not surprisin gly, Ar­n old (1946, 1960) made the same error in describing another two species. Nonveiller (1973) transferred to Smicronn;rnzilln all

'

species of Nnnonlulilln, except for N. vnuclr-eri and the female of N. microsomn, and d e­limited both genera, although he had also not exan1ined the type specimen (s) of N. microsonw. We confirm Nonveiller's con­clusions as correct. A rgam an (1988) de­scribed a third species, N. JTadnc from Spain, again known on ly fron1 female specimens. (Although the name is mostly spelled "noda" in that paper, it is " nndne" in the key; there is a s tatem ent that the species is named after M rs Nom·eHler, us­ing her nickna1nc [which is ada, DJB p ers. obs.], so the fcnainine geniti\'e form is preferable, and the commoner spelling is probably an inadvertent error.)

Many statements by o ther authors re­ferring to Nauollllllilla hc.1vc been based on a presumption that they app ly to the ty pe spec ies, N. z•ouchcri. It is now clea r to us, however, that the ~pccimens identified and illustrated a~ N. z•nuc/1cri by Nonvei ll­er (1973), Argamc~n ( 19H8), c~nd poss ibl y Suarc7 (1975), were mi~idcntjfied, a l­though they v\'ere abo collected in Moroc­co (the type locality of N. 1..'17/lchcri being Tangier). When comp<l red with the o rigi-

}PU~NA I . or H YMt.:NCli'TbRA RFsrARCH

2

4

6

8

Figs. 2- 9. Nnnonrutilln spp., dorsa l and l,1tera l \' iL'ws. 2-5, N. ua uclrcri (Tou rnicr), 2 , length = 2.0 mm (Gi­braltar, compared w ith holoty pc). 6- 9, N . sp., d , length = -± .-+ mm (Zimbabwe). Sca les - 1.0 mm.

nal description (Tournier 1895) and the fuller description and illustration by An­d re (1901 b), some discrep ancies are obv i­ous . Both Tournier and Andre referred to a m edi an longitudina l ca rina on the me­sosom a (thi s is shown in Andn?s illus tra­tion as ending in a fin e tooth pos teriorly) and also s tated (and illustrated) tha t the mcsosoma was twice as long as wide. The illustrat ions given by Nonveiller (1973)

and Argaman (1988) s how the m esosoma as 111 uch less s lender, without a c01nplete longitudinal ca rina and without a n1edian posterior tooth; Suarez (1975) expressed puzzlement a t the lack of s uch a ca rina in spccin1ens he id entified as N. vnuchcri but p rovided no illustrat ions. O ne of us (DJ B) has examined the holotype of N. vnuc!Jcri (coll ected at T<:rngie r by Vaucher, with Tou rnicr's determination label referring to

VOLUME' l l, NUMBER 2, 2002

the publication of the n an1e, labelled as from the Tournier Collection an d housed in the Geneva Museum). Unfortunately, it has been glued dorsal-side down to a card so tha t the dorsal surface of the mesosoma is ahnos t entirely obscured . There is, how­ever, a clearly consp ecific specimen in the san1e collection, also collected a t Tangier (in 1896), which is essentially identica l to the holo type (although with the tibiae very slightly p aler); the mesosomal d orsal suTface is clearly visible and shows an al­most complete very fine median longitu­dinal ca rina ending in a very s1nall p os­terior tubercle, and the mesosmn a is rela­tively n1ore elongate than in the speci-1nens illus tra ted by N on veiller (1973) and Argam an (1988). Another specimen, fron1 Gibraltar and housed in the N atural His­tory Museum, London (illustra ted here, Figs. 2-5), is also clearly consp ecific al­thou gh the appendages are slightly paler than in the holotyp e; it has lost tl1e scat­tered lon g erect setae on the mesosmna and m ost o f the decumbent pubescence, but shows the ea rina and tubercle n1ore clearly as a result. The carina is extremely fine and smnewh at irregular, nonnally concealed under fairly dense dia gonally oriented decumbent pubescence tha t gives the appearance of a mid-d orsal line in un­w orn sp ecimens, and even w hen visible needs careful illu1nina tion; Suarez (1 975) m ay thus have overlooked it, altho ugh he did comment on the pubescent line. Both species of Nnno111utilln illustra ted and d is­cussed by Argan1an (1988) are different from the true N. ·vnuchcri in a U of the £ea­hires listed by him as important in species differenti a tion, and , s ince the specimen he considered to be N. vnuclu:ri was obta ined from Nonvelller, it is d ear tha t Nonveiller (1973) also misid entified the species. (Of four specimens now in the Paris M useum identified as N. vnuc!Jeri by And re, on ly one (from Gibralta r, obtained from Saun­ders and thus aln1ost certainl y collected at the same time as the specimen in the Lon­don Museum) is correctly identified; the

315

others, one from Algeria and hvo from Sy­ria, represent two different species.)

Ticopla was described for two new spe­cies collected in the Jordan region and known only frmn males (Nagy 1970). Brothers (1975) suggested that one of these may be the male of N. vnuchcri, suppos­edly known from the sa1ne area but only fron1 females (based on specimens so identified by Andre, see above). Argaman (= Nagy) (1988) countered this by describ­ing fen1ales of both species of Ticopln. Nonetheless, he stated that "the resem­blance between NonOilllltilln and Ticopla fe­males is so ren1arkable, and the difference so delica te, [that it is ] entirely m1derstand­able" tha t Andre had identified a speci­n1en from Syria (tha t Argaman called a Ti­copln) as N. vnuchcri; i.e., Argan1an (1988) s tated that these genera are so similar as to be easily confused.

We h ave examined five specimens (three fem ales frOin Syria and two males fron1 Amman, Jordan) tha t a re unques­tion ably Ticopln based on Argaman's (1988) crite ria, and find no consistent dif­ferences between them and fen1ale speci­Inens of 6 fur ther species from Gibraltar and Morocco (N. vouc!Jcri) and southern Africa (including N. lllicrosomn), and males of 17 species fron1 Kenya and Angola to southern Africa. The differences in the sculp tur ing of the mesosoma 1 dorsum of both sexes, the 1nain character used by Ar­gama n in distinguishing these two genera, are by no means as distinct as he s up­p osed, since in termed ia te forms occur. O ther differences given by Argaman, such as the shape of the flagcllomeres, depend on the angle at w hich the specimen is viewed. Loss of the second submarginal ce ll (lS), thought ch c.1 racteris tic of Ticopln by Brothers (1975), is c1lso not significa nt since different degrees of reduction in wing venc1tinn arc evident. We thus have no hesitation in regarding Nano/11/{til/a zmd Ticopla <1~ synonymo us. Lclej and Krom ­bein (2.001) abo regarded Ticopln as a syn­onym of Nono11111fillo, implying that they -

3 lb

were following A rgaman (1988) in this (which was incorrect since Argaman clear­ly stated that he considered them distinct, see above); it is probable that they were reaJly following the suggestion made by Mitchell and Brothers (1998). The genus has an extensi ve distribution in the Afro­tropical and southern Palaearctic (Iberian peninsula, Morocco, Algeria, Levant) Re­gions, most species being as yet unde­scribed.

Areotilla Bischoff (Figs. 10- 17)

Arcotilln Bischoff 1920: 25, 174. Type species: A reot ilia n reo/a tn Bischoff J 920, Transvaal, by original designation.

This genus was based on the type spe­cies and A. nwrshnlh (Andre 1903). It is the s1nallest genus in the subfamily in terms of species numbers, and has been revised by Mitchell and Brothers (1998). It com­prises e ight species known from males and two known from fen1ales, all fron1 southern Africa. (It should be noted tha t the correct spelling of the nan1e for the new species referred to as A. fermgincn (in the text) and A. f ermginotn (in a figure cap­tion) by Mitchell and Brothers (1998) should be A . ferruginen; the lopsus is re­gretted.)

Smicromynni/la Suarez (Figs. 18-25)

Snlicronllf/'1/Jil!tJ Suarez 1965: 570 . •

Type species: Mutilln nrinsi Andre 1896, Spoin, by origiJ.1al des ignation.

This genus was described for a single species (and single fen1ale specimen), a l­though a second species from Spain, SJJJi­crOIIllfrJili!ln 111irnndn Nonveiller and Gras

' 1996, based on a single n1ale specimen, has since been described; these are the only speci1nens recorded from tha t cowl­try, and may thus be conspecific. Never­theless, many other sp ecies o f SmicroJ/lyr­lllilla have been described and even 1nore <l wai t description; they were being revised

j OURNA l. 01-' H YMI-'NOPTI-RA R ESEARCH

by Nonveiller (pers. comtn.) before his re­cent death. The genus exhibits consider­able morphological variation (it is the only ticopline genus in which brachypterous and apterous males are known) and is widely distributed throughout the Afro­tropicat southern Palaearctic (Spain, North Africa) and Oriental Regions.

After this paper had been accepted for publication, Lelej and Krombein (2001) de­scribed three new genera of Smicromyr­n1illini (Can1eronilln, Eosmicron11;rmilla and

'

Hindustanilla) fron1 the Oriental Region and prov ided a key for their recognition. For our s tudy we had examined an apter­ous male of one of these genera (Hindus­tani/ln) and considered it to be a Smicro­mynnilln, al though a somewhat anoma­lous one. We had also exan1ined several Afrotropical species with characteristics different frmn those Lelej and Krmnbein considered limited to Sll1icronl1Jrlllilla, but

'

again did not consider them as generically distinct. We recognised that Smicromyrmil­/n, as we conceived it, was quite variable, but saw independent variation in several of the characters used by Lelej and Krom­bein (2001) to distinguish their new gen­e ra, with many intermediates making rec­ognition of new putative genera question­able. For this reason, we do not distin­guish between Smicrol1Iyrmilla and the new genera proposed by Lelej and Krom­bein (2001), but do not wish to synony­mise them forn1ally. As far as we can as­certain, those genera agree with 511licro-11I lfr11 lilln in all of the characters we have

•

used in this analysis. (We also s uspect, however, that Cnmerouilla may not actu­al ly be a ticopline. Lelej and Krombein based their con cl us ions entirely on the rathe r inadequate original description and figure of the female of M!lfilln oedipus Cameron 1897 in placing it in this subfam­ily, citing the presence of a median and late ral spines on the propodeum. Those characteri s tics would not preclude its placen1ent in Myrmillinae, however,

VOLU1\ l E 11, UMBI::R 2, 2002 317

10

\

12

14

16

figs. 10-17. Areotilla s pp. 10- 13, A. ji:rm:;-Juco Mitcl1cll and Brothers, 'I', length = ..t-.Y mm (para typc, South Africa, KwaZulu- ata l). 1-l, 16, A. nmrc;Jllllli (And re), o, length - 7.7 mm (h1)loLypc, Soulh A ~rica, Norlhcrn Province). 15, 17, A. vulgaris Mitche ll a nd Brothers, 6 , lcnglh = 8.9 mm (p<H,1tvpc, South Africa, Eastcm Cape). Scales = 1.0 mlTl.

1 ~ lL R'\1 \L 01 H \ \IL0.0PTLRA Rr~1 ,\RLI I

18

20

I

22

24

Fig..,. I~ 2S. ..,1111< rollllfmlllla "PP· l R-21, S. ~p., + , leng th ~A mrn (Sllulh t\fricc1, ivlpumai<Jnga) . 22-25, <;.

1.0 mm . ..,p., ', ll'nglh b.S mm (South Africa, C,lutcng). Sc,11L·~

which pl<lccmcnt is a lso suggested by its enlarged qu adr<1 tc head.)

M /\TERI/\LS AND METHODS

Spccin1l'Jb of ,1 11 ,lvililable species ofTi ­coplinile (1\rl'oli//o: R species based on m,1lc!'>, 2 on fcrn,llc !:>; NanOIII IIfilln: 18 on male">, H on femcllc~; S111icrOIIIlfnllilla: 21 on

•

rnZJlL·s, R on females) were surveyed. Of the more th .:111 -W mo rphological charZJcte rs ex a m incd, 2 1 (/\ ppend ix I) had appropri ­cl lc levels of varicltion (i.e. they were found to be vari~1blc an1ong but not 'Within gcn­crcl) c1nd were used in cladis hc anal ses. Chilr,lCtcr polarity was es tablished by ou t­group comp.1rison. When there was va ri -

VOLUMF 11, UMBER 2, 2002 319

Table 1. Da ta matrix for analys is of genera of Ticoplinac using 21 ch<tracters L)f Appendi>-. l.

1 .., 3 .j 5 b 7 8 1.) 10 -

Ances tor 0 0 0 0 0 0 0 0 0 0 A reotilln 0 0 0 1 0 1 I 1 0 1 Nn11011111filln 0 0 0 l 0 0 0 0 0 0 51/1 ic YOIIIIJ I'll/ i//11

' l 1 1 0 1 0 1 () 1 'I

ation an1ong out-group taxa, ad lzoc parsi­Inony analysis was used to detennine the plesiomorphic s tate; this is discussed, w here applicable, below. MaxiJnum-par­s imony analysis was carried out using the software package H e1u1ig86 version 1.5 (Farris 1988) (c01nmand ie*), and analysis using in1plied weights (Goloboff 1993) was done using Pee-Wee version 2.1 (Go­loboff 1994) (con1n1ands hold* search = hold / 20 n1ult*15). A h ypothetical ances­tral taxon was included, wi tb al l cha racter s tates coded "0", to root the tree . Trees were an alysed using Clados version 1.6.1 (Nixon 1994).

Out-group selection presented some dif­ficulties, as the sis ter group of the T icopli­nae consists of the Rhopalon1utilhnae, Sphaeropthalminae, Mutillinae and Myr­m illi nae (see Fig. 1), i.e., mos t of the di­versity of tbe family. In addition, these subfan1ilies all tend to show a g rea ter pro­portion of apomorphic characteristics than does the Ticoplinae, and their usefuln ess is therefore lessened, particularly in the case of the Rhopal01nutillinae, which has a comparatively large proportion of d e­rived characte ris ti cs. W e expected that

6

1 l 1~ n I ~ l 'i lh 17 IX 19 ~t) ~I

0 0 l) u 0 0 0 0 0 0 0 ] 0 I (} I I 0 1 1 0 0 ') ] I 0 I n l 1 I 0 0 -0 0 0 I () 0 ') 0 1 l -

1nore accurate determination of character polarity would be obtained by including more relatively basal groups in the out group; thus the out group comprised four genera: Mynnosn Latreille (about 5 species examined) and Mynnosuln Brad ley (about 2 species) (both Myrmosinae), Pscudoplw­topsis And re (about 10 species) (Pseudo­photops id in ae) a nd Dns ylnhris Rad osz­kowski (abou t 15 species) (Sphaeropthal­Ininae). The plesiomorphic sta te is tha t state found in all four out-group genera, unless otherwise s tated . Note tha t Pscu­doplwtopsis also has a large proportion of apon1orphic character s tates and quite of­ten has the derived state of a character when the other three out-group have the plesion1orphic state.

RESULTS

o·enera b

Table 1 shows the dis tribution of char­acter states among the genera of Ticopli­nae. The single most-parsimonious tree found (Fig. 26) has Arcotilla and Nonolll/1-

tilln as sister groups; the same tree was found using implied weighting. This tree has length = 26 s teps, consis tency inde-._ (Cl) = 0.88 and retention index (RI) =

8

Areotilla

7 10 11 12 17

Nanomutilla

1 1 1 I 1 2 1 I Smicromyrmilla

Fig. 26. Single most-pars in1c.miL)US cl,ldog ram of gene ra ot TiLnplin,w (k>nglh 2h, con.._i..,lency indl'x ll.HH, retention index O.n2). C haracte r numbe rs ab(we, s tutc~ bclm\ h,,..,h rnMb H,l..,h-marh. -,hading. bi,Kh. uniqul'

d erivation, g rey = con\'crgent derivation, white re\·cr!->al.

320

0.62. Under fast (accelerated trans form.a­tion) optin1isation, the 23 derived s tates comprise 14 autapomorphies, 5 unique synapon1orphies for Areotilln and NmiO­mutilln, 1 unique synapomorphy for all three genera, and 3 h01noplasies ( charac­ters 7M, 10M and 16M) . An alternati ve tree with Arcotilln and SmicrO/Illfrlllillo as

•

sister groups (supported by 3 synapon1or-phies) is two s tep s longer and has a much lower Rl (length = 28, Cl = 0.82, RI =

0.37); the other alternative tree with Nn-1/0IIIUtil/n and Smicro nzyrmilln as sister groups is not supported by any synapo­morphies and is even longer (1ength = 31, Cl = 0.74, RI = 0.00).

Fron1 the results, it is evident that Ar­cotilln and Ntmomutilln are the most closely related cladistically, and a re distinct from Sl/licrol/lyrl1lilln in several respects. The de­gree of difference seen between the two groups is s imilar to, if not grea ter than, that between the tribes of Sphaeroptha l­n1inae or Mutillinae (Brothers 1975) or those of Mynnosinae (Brothers 1999). Tl1is contras ts with the opi nion o f Brothers (1975) w ho had seen far fewer represen­ta tives of these genera and who felt tha t there were no n1arked groupings between them. We thus consider that recognHion of two tribes, as proposed by Argan1an (1988), is warranted despite the fact that both contain rela tively few species w hen compared wi th m ost other tribes of Mu­tillid ae.

CHARACTERISTICS OF SUBFAMILY AND TRIBES

Previous descriptions or diagnoses of the subfamily and tribes, such as those by Brothers (1975, 1993), Sua rez (1975) and Argan1an (1988), are incon1plete or inac­curate, n1ainly because those authors had access to far fewer species than we were able to exam in e. The following descrip­tions are followed by comments indi ca ting differences from prev io us attempts.

Ticoplinae Nagy 1970 ( = Nano tTluti 11 i­nae Smirez 1975).-No felt line on second

)OUl<NAL or H YMLN O I' I LRA RI·S I- A RCH

metasomal tergum .. Macropterous males with fully articulating meso-metapleural su ture (i.e ., no ventral bridge-like fusion between meso- and metapleuron), poste­riorly convex n1esopleural 1nargin, petio­late second subtnarginal cell in the fore­wing, and volsella lacking digitus (i.e., only cuspis present). Females and microp­te rous/ apterous males with m.esosoma widest posteriorly (seldmn with sides n1ore or less parallel) with one or more weak to s trong tee th or spines at postero­late ral angle, posterolateral m argin of pronotun1 indistinguishably fused with mesopleuron (except in males with pron­otum articulating w jth mesothorax), and distance from humeral angle to pronotal spiracle a t leas t as long as that between pro nota I and propodea 1 spiracles (except in 1nales w ith articulating prono tum where it may be shorter).

Ticoplini Nagy 1970 ( = Nanomutillini Sua.rez 1975).-Eye strongly pubescent; ante1u1al tubercles closely approxi1nated basally but separate, not joined by a s traight transverse ridge, scarcely protrud­ing; pronotun1 smoothly and evenly con­vex over anterior declivity, without a transverse carina; propodeum with disc and declivity dis tinct; second n1etasomal s ternum w ithout fel t line. Males w ith no­tauli usually dis tinct (often faint and som etimes absent in Nnnomutilln); scute1-lun1 not apically produced; propodeum with three or five large fields covering en­tire surface of disc and defined by well de­veloped ca rinae; me tasmnal s ternun1 2 w ith a sho rt m edian lo ng itudinal carina basally; penis va lve > 0.75 X as long as para1nere. Fen1ales with at most one short spine on posterola tera l ang le of propo­deun1 a t apex of lateral oblique transverse carina; 'auricle' a t base of firs t n1etasomal tergun1 merely a stnall rounded protuber­ance; no defined pygidial area.

SmicrOinyrmillini Argan1an 1988.- Eye pubescence absent although pores and/or very spurse minute setae may be present; an tenna] tubercles fused basa lly, joined by

Vol l \11 11, :-JL \IBCR :2, ?0(}:2

a s ma ll .... tra ight transver..,e ridge, distinctly protruding; pronota l dor..,um .... harply sep­ara ted from anterior decl ivity, w ith junc­tion nng u la r and n1arked by a transverse carin CI; pro podeum with disc and decl ivity evenl y merg ing, no t dis tinct; ~econd me­tasoma l ~ternum with well developed la t­e ra l fe lt line. MacropterOLh males lacking no tauli; scutellum apical!} produced O\'er metanotum; propodeum with three poorly defined anterior fie lds and m Ci ny reticu­la tio ns forming 1nini fi e lds over posterio r half; mctasomal s ternurn 2 lacking n1edian long itudinal carina; penis \'a lve < 0.60>' as long clS paramere. Fema le~ and microp­te rou..,; apte rous males wi th a t least two s pines on posterolatera l ang le o f propo­deum, lacking la teral ob I iq ue carina to base of spine; 'auricle' a t base of firs t me­tasoma l tergum forming a s tron g lam el­late o r spinose protuberance; glabrous py­gidia l Circa \Vell defined .

SuareL (1975) included o nl v Nano111util/a -(female<.;) in h is subfa mil v anom utilli--nae. Va rio us of the fea tures tha t he hi gh-

lighted as being cha rac te ri ~ ti c of the group (as compared w ith S111 icroll/yrlllifla w hich he pl aced in the Myrmillinae) a re thus re­stricted to tha t genus, and in pa rti cular to a species w hich h e con~idcred to be N. Pt7llc!Jcri. o n-di ffe renti ated pubescence on the bod y \-Vas thought characteristic, as compa red with the varied p ubescence genera lly forming patte rns in other Niu til ­lidae; a lth ough this is pa rti cul arl y true of Nanon/1/ t il/a, it is approached in Areotilla but is not particula rly significant since it occurs e l ewhere in the IVlu tillidae a I so. The peculia r 'bethyloid' o r 'proctotrupoid' body fo rm was also h igh lighted, but thi'-> is not true of Areotilla.

Argaman (1988) al..,o included onlv \Ja--nOllfl' til/a (and 'Tiwpln'), but both sexe~, in the f icoplini. He thought that the flagcl­lomcrcs were different in shape and ~true­tu rL' from those in Sn1 icromyrmillini, that the pronotum (m the female) differed in the number of lateral pits and that the de­gree of production of the ape' of the pro­podeum at the (.uticul,ltion \.\ ith the mL'­tc"''..,oma differed. 'vVL• h,we tound that tla­gell omerc shape \ ,lril''-1 clCro<.,s both tribe.., and also according to 'ievvpoint, that tlw develop1nent of pit.., on the pronotum \ aries considerabl_y and that the produc­tion of the propodeal lobe abo 'ariec.., and i.., not significantly different from the con­dition in most other \ 1utillidae. Argaman remarked on the fact that the second me­ta~ornal tergum i'-> longL'r than wide in fe­male Ticoplini, but thi'-> ic; true only of some Nmzo111utillo and not of Areotillo. For male Ticoplini, Argam,"''n noted the pre<.,­cnce of a single complete ridge on the scape, but this i<:> true of 1\ 'anollluti/la only (c..,ee our character 7, beiO\\ ).

either Suare7 (1lJT1) nor Argaman ( 1988) mentioned the pubescence of the eye in female Ticoplini, presumably be­cau~e it is difficult to ..,ee in ~a11omuttllt1

c..,pecin1ens, which arL' \er~ c..,mall. Both clll­thor~ noted the ab .... enu.• of 'auricle...,' (Brothers 197:J) at the ba .... e of the meta..,u­ma, but they are actually pre .... ent although . . I I"''COnSplCUOUS.

For the Smicrom\ rmillini, Argamclll (1988) made much of the carinate anterior mclrgin of the lllL'...,thllltum in the lll<llL', con .... idering thi.; cl lll"''lqUL' chclracter in Hy­Jlll'lloptl'ra; -..udl a L·,uin,l b certainly prL'­o...,cnt in ..:;ome -..pL'dl'-.. ot ~ ·n icronll!rmilla but m,"''n\ other ::,pL•( il'..., h.WL' no trace of it.

KL'Y TO TRIBES AI\.0 <._,f'\,1 R \ Or TICCWII:\ t\ F

['\lote· \'n• cnn...,tder Eosnll i Hlllllfl"lllllla ,1nd fi111d11.., f m "1 dnub ttulh dJ-.tiJKI trom '>11111 rollllfl"llllll.l; Camcro/lllla j...,

prob,1bl) rm~placcd in TicoplinaL', .... cc abll\ L', and '" tlwn"'tnrc lm1tlted twm th1" kc\ .)

1 ( il) (b)

Wing.., well den~loped (mall') Wing.., absent or rudimentar)

. . . . . . .. . . . . . . . . . . . . Uemale, rardv m.1le) . . . -

? . . . . . . . . . . . . . . . . . . . . . . . -. • • • . • . • • • • • . . . . . . . . . . . . ..1-

~'.2 jOURNAL OF H Yf\ l i.::Nl)PTFRA R ESEARCH

2 (a) Eye pubescence d is tinct; scutellum apex not overhanging metanotum; propodeal disc covered by 3 or 5 large fields each surrounded by well developed carinae; metasomal sternLIIn 2 without fe lt line (Ticoplini) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

(b) Eye pubescence absent; scutellun1 apex produced and overhanging metanotum; pro­podeum with 3 very weakly defined anterior fields and many reticulations forming mini fie lds over posterior ha If; metasom.a I s ternum 2 with wel l developed la teral fe lt line (Smicromyrnti llini) .. Smicromyrmilla (including Eoslllicrolllymri/la and Hinduslmzilln)

3 (a) Tegula elongate and reniform, > 0.75 X as long as mesoscutum; propodeal disc with 5 fjclds; paramere apex very strongly curved ventrally ...................... Areotilla

(b) Tegula ova l, < 0.60 X as long as mesoscutu m; propodeal disc wi th 3 fields; paramere almost straight .................................................. Nanomu tilla

-l (a) Eye pubescence distinct; metasomal sternllln 2 without fe lt line; propodeum with disc and declivity distinct, separi'lted laterally by an oblique transverse carina ending in a single small pos terolateral tooth; no distinct pygidiaJ area (Ticoplini) . . . . . . . . . . . . . . . 5

(b) Eye pubescence absent; metason1c1l s ternum 2 w ith lateral fe lt line; propodeun1 wi th disc and declivity smooth ly merging, without any lateral tra nsverse ca rinae and with 2 or more pos terolatera l teeth or spines; female with distinct glabrous pygidial area (Smi­cromyrmillini) . . . . . . . . . . Smicromymt illn (including Eosnricronrymri!la and Hindustnni/la)

5 (a) Body length > 4 mm; eye lc1rge relative to hccld (rat io of eye height to head height > 0.6); eye with > -lOO small omtDatidia; second meta.somal tcrgum abou t 1.5 X as wide as long ........................................................... Areotilla

(b) Body length < 3 mm; eye small relative to head (ratio of eye height to head height < 0.5); eye w ith < I 00 large om n1atidia; second metasomal tergum about as wide as long ............................................................... NaJtOin tttilla

GEOGRAPHTCA L DTSTR I BUTION

The subfamily occurs in the sou thern Palaearctic, Afrotropical and Orienta l Regions. The tribes differ in distribution only in that Ticoplini have not yet been found in the Oriental Region. Brothers (1975) proposed that the Ticoplinae arose in eastern Central Africa, and from there spread northwards to the Mediter­ranean region, southwards into southern Africa and eastwards to the Ind ian plate, while it was sti ll in con tact with Africa or very close to it, i.e. at least about 80 million yea rs ago (Smith, Hurley and Briden J 981 ). However, this does not ac­count for the apparent absence of Tico­plinae from Madagascar, as reflected by at least two recent collec ting expeditions there (fron1 the Natura I History Museum (London) and the University of Kansas) which have failed to co me up with any specimens, despite their emphasis on Hymenoptera.

ACKNOWLEDGMENTS

This s tud y formed pJrt of a disse rtation s ubmitted by AM in partia l fulfilment of the requirements for the M.Sc. at the Univcrsitv of Nata l. Financi.:~l 0ssis-., ta nce to DJB from the University of N,1 tal Research Commi ttee is gratefully acknowledged. The curators of the m,1ny institution~ vvhich loaned spec imens are thu n ked for thei r ass is tuncc, as a rc the technicians at the Electron Microscopy Centre of the University of N<1tal (Pietermaritzburg).

LITERATURE CITED

And re, E. HNo. otes pour serv ir .'1 la connaissance des l\tlu tillides palearctiq ues e t description de qu­dqucs cspcccs J1l1uvcllcs, de u>.icmc partie. Mt'­llloircs de In Societe Zoologil711t' de Frrmcc 9: :261-277.

Andre, E. 190ll. 2nll' CL·nrc.-l\tluti lb, Linnc. In: An­dre, E. Monographic des Mutillide:-. d'Europe e t d' A lge ri e. Sp,;cics des Hynrcnopth"cs d'Eurol'c et d'l\l:;;t'ric R: 1?5- 135.

Andre, E. l90 1a. Materiau'\ pour se rvir .'! la connaiss­ancc des Mutill idcs d' J\friquc. ZcitsclirUi Ji"ir sys­tcnwl isc!IC I lynrcnoplcrolosic und Vrptcrologie I: 3Wi- 35?.

J\ndre, E. llJOib. 6mc Sous-Genre.- N,lnomutill,l nov. ::. ubg. In : J\ndr<.\ r:: . MorwgrZ~phie des Mu­ti 11 idc::. d' Europe et d 'A lger ie. Sp(\· ic~ des 1-Tynn'n­oplcrcs d'Europc cl d'Aigcric R: 223-226, pJ. X.

Vot L\1 1 11 , l\.L\lB~R '>, 2002

t-\ndn:•, I . 1903. ~>/lutillid e.., d 't\fnque nou,·eau' ou impMfaitement connu~. /cthtln~fl tin· :-IJ5h'llltlft~­clic llyntmoptcrologtc 1111d OijJ/ crolngtc 3: 81 - KH, I ;\7- I..J...J., 232- 239.

Arg.1m,1n, Q. llJH8. Description of the lema le of Ti­l'OJ'IIl. \\'ith bio log ical and t,1\ori0mic notes. Fmg ­IIJCttfa Ralmuim 1-l: 33--+6.

Arnold , C. llJ-!6 .• ew ~pcciL•-.. nf African I Jymennp­le ra, '\.o. 6. Occa.;;ioual rapa ... t>( I lie ,\tlf/011111 ,\ l u scl/111 of Sout!Jcm Rlwdc ... ta '1: -I ll-SS.

Arnokl , C. I 9n0. t\:e\'\' specie~ ot \ fricc~r1 H~ mcnop­tera , \In. Is. Occasit>lllll Paper~ ot lite NatJOual i\lu­

..;cunt of Sou them Rlwdcsul 2-1-: -i-S7-+h I.

Ashn11.~.1d, W . H. 1903. Clc1s~ific,1lion n t the fossorial, p red ,1ceou:-. .:tnd pJ ra::-i tic w .~~ ~)-., nr the supcr­f,lmil v VL'::-poidca. Ctmoduw f'IIIOttlt>logi~t 35: 323-332.

Bi-..chnH, 11. 19'>0- JQ'>J. \lonographrc der \lutillidL·n 1\ trrka-... Arclll'i' fiir Nafttr..:c ... tludtit' {i\) H6: 1-8.10.

Brother-.., D. J. 1975. Phy logem ,1nd cl,,..,~ i fication ot tlw aculeate H ymenopterc1, '' ith ...,pcc ial refe re nce to Muti llidae. UuiZ'crStlrJ of "llll~th Sctcucc Bul/e/t/1 SO: ..J.H1- 6-iH.

BrollK•r-.., D. 1. 1993. Familv Muti llid ,w. In: Coule t, ll. •

• 1nd lluber, 1. T. cd~, HyntCIIOJiit'l'tl c~f tltc I Vt>r/d: 1\11 ldcllft{lcatioll Cwdc to fanuln·.... Agriculture C111,1d,, Rc...,L·arch Branch, Ottawa. pp. 187-201 .

Brotlwr-.., D. I. 1499. Phylogcn~ and e\ olution of \\'•"'P"'' ant-.. and bees { 11 ymenoptera, Chry~idoi­

clL-,1, Ve-..poidea and Apoidca) . /oologim Scripta 28: 231-2""-9.

C1 mcron , 1~. 1897. Hymenopte r,l Oricnta li,1, or Con­tribution.., tn a knowl edge of the ll ymenoptcra o f thL• Oriental Zoologica l l~egion . l'.1rt V. Alaucltc..;­lcl Alclti0/1:' -!1 (-!): 1- 1-l-.J., pi. 1--+.

Da\ , \I. C 1 98-L The enigm.1 trc genu~ Hctc.'nl~lflltl

'\.ag\ (11} menoptera : Sphccrdae: Heterog) ni ­n,w). Svstcmattc EllfOIIIt>logv I.J: 2tn-i07.

1-arrb, ). ~- llJHH. Hc111ug86. ('t'I"•IOII /. 1. [Computer '->O flwarc .l!1d m c1 11llc1 11 . 1-,ort ldfer::.on Stati o n, New York

c .• u ld , I. <l lld B. Bolton (elh). ILJHH . 1'/te lly/1/CIIOpfcm. (),ford Uni\'crsity Pre~s, 0'\ftlrd

Colobol I, P. A . l lJq3. E<;timatrng cht~r,Ktcr weight-.. durrng tree -..c,~rch. Cladbllc.., LJ: ~n tJl.

Golobotl, P ,\ . 199-l-. Pcc-1\ cc. 1'cr-.JOit 2 I 'l [Computer -..otl\\,Hl' and manual! . fuLum.ln.

lntcrn,ll ron,,l Commi~~ron on / ooltl)!,K·' ' '\lnmencla­tu I'L'. ILJH7. Opinion I-I-..J.5 H clcrogy n 1d ,1c Rambur, IH66 (ln..,cct,l, LepidoplL' r,1) .md ll l'tc rngynid,,L' N.1gv, ILJ(1lJ (Insecta, HymL·noplc r,l): ,, ru ling to remm c the homonymy. Bul/1'1111 cl/ /oologiwl No

llll'tldalurc -!4: 1 S0-151. 1n lern,1 t ron.11 Corn m i~~ion on / tltllt >grt .1! '-mnentl,1 ·

turL' ll)l)!.J lntcmaftonal Ct>dc of /oolngrml \tt/1/c 11

doll/re. -llh edition. lntern.1tron.1l I ru-..t tor Zno-lngiL.11 nrnenclature, l ondPn

I clt:' l· 1\ . ~- .1 nd K. V . Krombcin . 2!l01 R L'\ IL'\\ tt f the

< )rit•nt,ll mutlllid \\ ·"'P"' of Hw ...,ubfamih Trul­plln,w (l-I) nwnoptl'r,,, \lutrlhdae). Far fa..,fc'lll I IIIPtltPittg,._t 99 I I K.

I c lq, \ . "l .1 nd P.C. f\Jt•mkm. llJLJ7. Ph) logL·ny, L'\ u ­lul JPil and ci .N-.ili('<l tion nt Mutillid .1e (ll vnw--noptu ,, ). fa1 ftr-lc'tll I nltllll<>losht ""-6: 1-2-t

\lclo, C \ R. JlJLJLJ. Ph\ lngt•ndk relatJOn~hrp-.. and d,,..,..,,trcatinn of tlw m.11or lineage~ nf ApntdL',l (I I) tnL'J1l)pkr,,), with t•mph.,..,l', on the uabn1111d '''•l"P" St"inlltflc PaJic 'r-., .\atural Ht::-torrt ,\ ltl ~<'llllt,

J

T ltt' Ll tll'l'cr.:;iftt of "till .,,.., 1-k I-SS.

~ lr ttlw l l, ,\ and D. I Brotlwr-... llJlJK. RI.?\ r-..ron and cl,1dJc.t rc analy~i-.. 111 tlw .\lwtrupica l gen u .., A1 cnttlltl (Hymenoptc t\1: Mutd lid,w). A_ti·lct111 [ IIl O

11/tl/tl,~ If 6: ] LJ1-2J-i-.

Nagv, C. C. llJ70. 1 urtlwr rm L" .. tigation-.. on the lwt­crog) no id wa-..p-... f 11/<llltolc>sl'dtc \ lltft'tfiiiiSt'll tlll..,

de m LMlo.;:istl1cn A ltt..,c'/1111 llttlllltur" -l: 83-~6 '-1'\nn, h.. C. 199-L Cladc>:-, •'c'l'..,/cl/1 Ill 1 [Compukr

"lift\\' are and manu.1lj lth,lC,l, '\. ew 1 ork omcilkr, C. 1973. RL•clwrdw-.. -..ur IL·~ 1\lutillrc..k-..tk

I'A friquc (1\ lu tillrd.w, I h ·mennptera). Ill. Rc..• ­m ,1rqucs conccrn,mt IL• gl'I1 1'L' NtlltOIII utJfltl 1\ ndrv IIN9. i\llnnlc~ de lo lot 11/lc eft·.., Sncnccs n11 Cttlllc'

I'PIIII 1 S-16: 63-73 . \lom ciller, C. ,1nd F Cnh. I tJtJh. De-..cripcion de ~m i ­

t rollllfl'lltilla mimnda n. -.p. (I l ) menopter,l, \lutJI ­Itdae) de la Penm-..u(,, lbL'f'll,, ( \luhlid o-.. p.11t>.ir­tl lll~ \JJ). Bollcfnt de la ...,t,tlt'lat d'fl~:-touo \atural de le" Ralcars 39: SLJ-h ~

~mith , A. G., A. 1\1. HurlL·\ .md I. C Briden. llJK I Pllili/CI'O:OJC ralcocollllllc'/1/111 \Vt>r/d Alaps. (..,)Jl) ~

bridge Uni\·e r~it\ l'rL'""' l arnbridge. SuML'L., r. I. 1465. Dato-.. ~)rc l imin,li'L'" a l e~tudJO dl'

lo-.. mutilidos ibL•nu1-.. ( I h nwnnptera). l c>" -Ul SCllJ-"ih6.

~u,1 re ,r, I I ILJ7:;. l.tlm mt•nl.HI\1-.. "l'bre Sllllt'l'tllltttr

11111111 SuareL ~ \.t7llcllllllllflt1 \ ndre (H) melll)plL't-.l, \lutillidae). ' \rd//,'ct" de'! flhltlufp dt· Atlnllilftlttt!ll

20: 11 N-1 19. 'I nurn rer, H . lH9S. Sur 1\ltillo"J'tlllnt.., tlu>[lt1llft Dt•..,b r

(Ct.>IL•optere~) et -..ur dL'll\ ~lutdl c::.. nouvcllL•..., du i'v ltli'Pc (HvmennptL·n·-..) Hul/1'1111 de lo Sooc'lt' t'll ·

IOIIIOic•slqllt' dt 11 t111t c' I ~NS: ~ 7-...J.ll.

.\PPI "\DI\ l

l h,lt\lLk'r-. u ... t•d t111 .. l.hii-.tr~.· .lll.ll) "''"' t't gt'nl't .. , nl lr Lnp lin,w. ~uth'\L'"> : B clpplic,lbll' lP bnth ... L''-L'"> I·

1l'm,1k nnl\, \I rn.tll' 11111\ l'nmili\ L' -..t,llL'" tPdvd ,, .., ll, dt•ri\'cd -.. t,llL'" '"' I 111 '1 \11 .. h.1J\ldt•r.., nlll-.. rd ­L'l'l'd ,Hiditi\ L'.

I H. F\'L' pubL'"tPrlL'L'. !l Prt.•...,t.•nl, ''"ibk' ,1 [ :w m,1gniftt.lllon. I \b..,l'nl ,11lhnugh ptll'l'"'

and/lH· \ L'r~ -..par-..l' .... hnrl "l'l,w ma\ be cl! .... trn gui:..h.1biL' under high m.1gn1trt..1tinn. (._,tall' 0 i:.. tound in mchl l iphiid,w (Brother..., I LJ7=i),

Ft.•d l!:->C11l' n kiii1.lL' ( s,) p yg id .w) I t\ 1 y /'1110 .... 11 c1 nd

Myrnwsula; since almost all Pseudophotopsi­dinae and all other mutillids lack eye pubes­cence, ste1te J has most likely evolved inde­pendently within the Ticopl inae.)

2B. Antenna! tubercles. 0 = Separate a ltho ug h closely approximated, not joined by a straight transverse ridge, scarcely protruding. 1 =

Fused medially, joined by a straight trans­verse ridge, distinctly protruding.

38. Pronotum, dorsal and anterior faces. 0 =

Smoothly and evenly merging, without a transverse carina. 1 = Sharply separnted by a distinct transverse carina, ctt least laterally. (Of the l)Ut-group genera, Pscudopl!olopsis has state 0, Dnsylnhris has state 1, while Mym1osn and M!Jrlllosulo appear variable. As none of the out-group taxa have s tate 1 developed as strongly as in the in group, this s tate is con­sidered apomorphic.)

48. Propodeum, disc and declivity distinction. 0 = Smoothly merging, not distinct. l = Dis­tinct, in different planes.

58. Felt line on mctasomal sternum 2. 0 = Ab­sent. J = Present. (Brothers (1975) stated that " ... the tendency toward developn1cnt of lterga l] felt lines is considered to have been established after the divergence of the Myr­mos inEle" which have neither tergal nor ster­na] felt lir1es (like the Rhopalomutillinae, in which traces of tergal felt lines are present in only a few moles). Referring to the phylogeny of the Mutillidae (Fig. 1) this indicates that the actua l development of tergal felt lines has opparently occurred on two occasions, once in Psl.'udop!wtopsis and again on internode 4-5. Sim ilarly, when consid ering sterna! felt lines, which are present in Pscudopliolopsis, Slllicrolllynnilla and sporad ica lly w ithin taxa derived above Rhopalomutillinae, it is most parsimonious to consider felt lines to have been developed independ ently on several oc­casions. Thus, absence of felt liliL'S is p les iom­orphic for the Ticnplinae.)

6M. Eye, inner margin shape. 0 - Sha llowly emarginatc at or below mid height. 1 Strongly notched above mid height.

7M. Scape, ventral longitudinal carin,le. 0 =

One (lateral). ·1 = Two (mesal and 1,1tera l) . (Pri mitively, there is tmly one longitudinal ca­rina on the sca pc, or none. A ltho ugh Dasylo­ln·is has two carinae, this appec:Hs to have been derived sepa rately in many higher taxa.)

jOURNAl O J· 1-l YMFNOPTFRA R ESEARCH

BM. Ratio of tegula length to mesoscutum length. 0 = < 0,60. 1 = > 0,75.

9M. Scutellum, posterior margin. 0 = A butting metanotum. 1 = More or less lomella te and overhanging mctanotum.

lOM. Scu tellum and dorsellun1, profile. 0 = On essentially the same plane. 1 = On two dis­tinct planes. (State 0 is found in a ll out-group taxa e.>..cept for Dnsyln/Jris.)

llM. Propodeum, fields. 0 = Three sma l1 fields on anter ior half defined by weakly developed car inae, many mini fie lds o n posterior half. 1 = Five large fields defined by very well de­veloped carinae. 2 = Three very large fields defined by well developed carinae. (This character was treated a::. addi tive because the states are comp lex, wi th state 1 appearing to be intermediate between 0 and 2.)

12M. Propodeum, extent of disc and declivity. 0 = Disc about as long as declivity height. ·1 = Disc at least 1.5 X as long as declivity height.

J3M. Metasomal sternum 2, short median lon­gitudinal basal carina. 0 = Absent. 1 = Pre­sent.

14M. Hypopygium, apical margin. 0 = Shal­lowly emargina.te or notched. 1 = With deep narrow n1edian split. (StClte 0 is the more sim­ilar to the conditions in all the out-group taxa; state ·1 is unique in Mutillidae.)

15M. Penis valve, relative length. 0 = < 0.60 X as long as paramere. 1 = > 0.75 X as long as paran1ere. (State 0 is found tn Myrnwsn and Pseudopl10topsis, despite the high ly derived, spinose state of the penis valve in the latter subfamily; whi le Oasylobris has s ta te 1, this has probably been separately derived.)

16M. Paramere curva ture. 0 = Straight. 1 =

Ape.>.. strongly curved ventrally. 17F. Eye size. 0 = Large (eye height > 0.60X

head height) with > 400 small om matidia. 1 = Small (eye height < 0.50 X head height) with < 100 large ommatidia.

·18F. Propodeum, posterolateral spines. 0 = None. 1 = One. 2 = At leas t two. ( one of the out-group genera has spines on the de­clivity; the development is postulated as pro­gressive, the character thus being regarded as additive.)

19F. Propodeum, lateral corinae. 0 = o d is­tinct carinc1e. 1 = Distinct obliquely trans­verse carinae extending posterolaterally.

20F. '1\uri clc' (Brothers ·1975) at b<~se of first me-

Vo1 L 1\ll 11 , UI\IB[R 2, 2002

tasomal tcrgum. 0 = Forming ~light rounded protuberance. 1 = Forming prominent lamel­late or s pinose protubercHKC. (These struc­tures arc absent in females of Myrmosinac, fairl y well developed in Pscudophotops idi ­nae and well developed elsewhere a lthough gcncrCll ly not so prominently as in Smicro­JIIIfl"lll ilia fcma les .)

•

211 1\·gtdtum. 0 = o defined pygidial area or 1-.late. 1 = Di~tinct glabrous pygidial area prc .... ent. (There is no pygid ia l a rea in Myr­mu~inclC, but such cln clrl'cl is present in mosl other Mutillidcw, including the other out­group representatives, cllthough it varies con­..,idl'rclbl) m form. The .... uggc'>ted polarity \\'cl'-. thu .... considered the nwrc lik.cly to be correct.)