J. Awl. en/. Sol,.. 1980. 19: 131-137 131 THE LARVAL AND EGG STAGES OF THE PRIMITIVE NOTHOM YRMECIA MACROPS CLARK (: FORMICIDAE) GEORGEC. WHEELER,JEANETTE WHEELER and ROBERTW. TAYLOR Deseri Reseurch Insrirure, Uniwrsity qf' Nevudu Sys/em. Reno. Nevudu 89506. U.S.A Division o/ Eniomology, CSIRO. Cunberru. A.C.T. 2601.

Abstract The egg, and mature, young and very young larval stages 01' Norlioniyrmc,c~iumucrup.s (subfamily Nothomyrmeciinae) are described and illustrated with line drawings and scanninyelectron micrographs. All stages closely resemble those of the Myrmeciu F. (subfamily ), supporting the evidence from studies of adults that Norliomyrmecia Clark and Myrmeciu share many primitive formicid attributes, though neither supporting nor challenging the hypothesis that they represent generalised survivors of the earliest diverging stocks of two separate lineages important in ant evolution.

Introduction Ever since publication of the first description of Norhomj-rmeciu mucrops Clark, which was then known only from two damaged worker specimens (Clark 1934), myrmecologists have eagerly awaited the collection of additional material of this ; more especially since Brown and Wilson (1959) suggested that it could be the most primitive living ant. It was not until October 1977, 46 years after its first collection, that Nothomyrmwiu was rediscovered by Taylor and a group of his colleagues attached to the Australian National Collection (ANIC), Canberra (Taylor 1978). We had hoped that the larvae would prove markedly different from those of the genus F., several species of which were described by Wheeler and Wheeler (1 97 I). Mj?rmrciahas been widely acknowledged as a very primitive ant, likely related to Nothomyrmecia (Clark 1934, Brown 1954), though differing from it in some possibly fundamental features, especially in abdominal morphology (Taylor 1978). However, our expectations have proved unfounded, since the larvae of Nothomyrmc>ciuare very similar to those of Myrmecia. This is not without significant implications. G. C. W. and J. W. have compared Nothomyrmecia larvae separately with material of each of 30 species of Myrmrciu in their collection and have been able to discern only a few inconspicuous (but apparently constant) dinerences between the two genera. This paper summarises their findings, illustrated in part with micrographs prepared by R. W. T. using a J.E.O.L. JSM U3 scanning electron microscope (Figs 9- 24). A brief description of the egg (with micrographs, Figs 25, 26) and a photograph depicting N. macrops brood in an observation nest are also presented (Fig. 8). The specimens illustrated in the micrographs were taken from storage in 70% ethanol, dehydrated, and critical point dried from carbon dioxide using amyl acetate as the intermediate fluid. They were then sputter coated with gold for study. Despite great care during preparation, the bodies of the larvae usually displayed some cuticular swelling or shrivelling, and the eggs, with one exception, crumpled badly.

Nothomyrmecia Clark Diagnosis qf larva* Body shape myrmecioid (i.e. elongate and rather slender, curved ventrally, without a differentiated neck, diameter diminishing only slightly from AV to anterior end), lateral longitudinal welts well developed. Body hairs of two types: (I) rather stout, slightly curved, with minute denticles near the apex, on all somites; (2) short, sharply bent and denticulate, on venter of anterior somites. Head small, cranium subcircular in frontal view. Head hairs moderately numerous and moderately long, simple. Clypeus feebly granufose. Labrum of moderate size, short. Mandible large, pogonomyrmecoid (i.e. subtriangular, with three conspicuous medial teeth, which are in approximately the same plane), heavily sclerotised. Maxillary spinules rather large and isolated apically, minute and in short rows basally; palp equal to galea in height and stouter, with 12 sensilla. Labium with spinules large and isolated apically and laterally. small and in short rows basally; with a dorsal, transverse, densely and coarsely spinulose welt; each palp a slightly elevated knob with about seven sensilla; opening of sericteries wide and salient. Hypopharynx with a few minute spinules. * For explanation of terminology etc., see Wheeler and Wheeler (1976). I32 G.C. WHEELER, J. WHEELER and R. W. TAYLOR

Ficis 1-7- No//tonl~.r.riic,rrtr/rrtrc~r.o~~~.\.larviic: (I) malure larva in hide vicw x 12:(2) head in Irontill vicw x.54: (3) leli mandible in frontal view x107: (4, 5) types of body hairs x358: (6)very young larca in side LICU (hair\ omitted) x12: (7) young larva in sidc view (hairs omitted) XI?.

Nothomyrmecia macrops Clark Mulurc kurvu (Figs 1-5, 8-20) Length through spiracles about I I mm. Body diameter greatest at AV and AVI, tapering to slightly rounded posterior end and gradually to slender thorax. Leg, wing and gonopod vestiges present. Anus

FIG.8.- Norhomyrmc~iumuc'rops in laboratory culture. showing worker and male adults, an egg (lower Mt), large larvae and pupal cocoons (upper left). N07'HOM ). RMEC'IA LARVAE AND EGGS I23 ventral, with a small posterior lip. Head apical. All somites distinct. Ten pairs of spiracles. peritreme about 17 pn in diameter, atrial opening about 6 pm. outside diameter 01' atrium ahout 28 pm. tracheal opening about 9 pm. atrial wall heavily sclerotised and without spinules. Integument with minute spinules in short transverse rows on venter of anterior somites and dorsal surface of posterior somites. an integumentary pigmented area of unknown structure and function posteroventral to each abdominal spiracle. Body hairs numerous. those ottype I 25-1 75 pm long; those of type 2 short. about 25 pm long. Antennae slightly below

FIGS9-14 Norhoniiwm,c,iu mac'rop. mature larva (length through spiracles about 10 mm): (9) mouthparts in frontal view x120; (10) labrum in frontal view x230; (11) left maxilla in frontal view x275: (12) labium in frontal view x255; (13) mouth opening in oblique frontal (buccal) view x230: (14) opening of sericteries in buccal view x600. I34 G.C. WHEELER, J. WHEELER and R. W. TAYLOR midlength ofcranium, minute: each with three sensilla. each of which bears a minute spinule. Head hairs 35- 70 pm long. Labrum bilobed. each lobe with two minute hairs about 3 Itm long and 14 sensilla near lateral and ventral borders of anterior surface; ventral and ventrolateral borders spinulose. the spinulcs rather long and isolated; ventral surface with three contiguous sensilla medially; entire posterior surrace spinulose. the spinules large and isolated ventrally and laterally. much smaller and in short rows medially, with about eight sensilla near middle ofeach half. Maxilla with apex narrowly paraboloidal; palp a skewed frustum with three

FIGS.15-20---Nothomyrmrc.iu mucrop. mature larva. mouthpart details. sccn tisu:111~with hcid 111 diIcci frontal or oblique frontal (buccal) view: (IS) left maxillary palp in 1.1-onial \WH ~1290:(16) ape\ ut right maxillary palp in buccal view ~1870;(17) apex ol'lel't galca in I'rontal view x IYS: (18) apiciil scnsillii 01 riglit galea x2200; (19) right labial palp in buccal view ~1600:(20) lel'i in l'roiital biew x 1260. NOTHOMYRMECIA LARVAE AND EGGS 135 apical and nine subapical to lateral sensilla, each bearing a spinule; galea more slender than palp. with two apical sensilla. each bearing a minute spinule. Labium small. the small basal spinules in short rows oftwo or three. each palpal sensilla bearing a minute spinule; an isolated sensillurn between each palp and the opening of the sericteries. Young lurvu (Figs 7, 2 1-24) Length through spiracles about 6.3 mm. Similar to mature larva. except as follows. Head relatively larger. Integument with spinules more distinct and more numerous. integumentary structures near spiracles

FIGS21-26 t\or/~ou~I~I~YUI u1m r0/1.7. immature stages: (21-24) young larva (length through spiraclesabout 0.5 mm). cephalic details in frontal view: (21) head x65: (22) mouthparts x109: (23) labrurn and labium x210; (24) left maxilla x790: (25, 26) egg: (25) general view x58: (26) detail of surface ~1240. I36 G. C. WHEELER, J. WHEELER and R. W. TAYLOR lacking. Mandible with teeth more sharply pointed. Maxillary galea more slender. Labium with smaller spinules. opening of sericteries less prominent.

Very young larva (Fig. 6) Length through spiracles about 2.8 mm. Body relatively stout, widest at AV and tapering more rapidly anteriorly and posteriorly. Spiracles varying in diameter: T2 smallest. T3 largest: abdominal spiracles medium-sized, all of same diameter. Body hairs 13-90pm long, fewer crooked hairs. Head moderately large. sides ofcranium less curved. Labrum lacking ventral and lateral spinules, each half of posterior surface with three or four sensilla near middle, ventromedial surface with threc or four sensilla laterally and minute isolated spinules in middle. posterior surface with a few short rows ofminute spinules basally. Mandible with apical half moderately sclerotised. apical tooth longer, all teeth very sharp-pointed. Maxilla with spinules smaller and palp and galea shorter. Labium with palp shorter. opening of sericteries inconspicuous. No spinules on hypopharynx. Otherwise similar to mature larva. Marerial srudird-Nineteen larvae from Eyre Peninsula, SOUIHAUSIRAL.IA, in G. C. & J. Wheeler collection; voucher specimens in ANIC. including those prepared for scanning microscopy. Egg (Figs 25, 26) The possible phylogenetic and taxonomic implications ofegg shape, surface structure and adhesiveness among have never been comprehensively surveyed. However. the very broadly oval. almost subspherical shape illustrated in Fig. 25 might be characteristic of Norliom~~rmcrirr~~~iuand Mjrniwia eggs, judging from a brief survey of the eggs of many ponerine and pseudomyrmecine species stored in spirit in ANIC. Norhornj,rnirc,iu,like Mj.mic,cia. transports its eggs individually. not clumped. and there is no evidence that the eggs are inherently adhesive. or subject to salivary or other gluing.

Discussion In the Wheeler and Wheeler (1976) key to the genera of ants based on mature larvae, the section dealing with larvae having"Profi1e 5. Myrmecioid" (p. 75) should be modified as follows: la. Mandibles ectatommoid ...... Myopopone (in ) 1 b. Mandibles pogonomyrmecoid ...... 7 7a. Hairs on venter of anterior somites very short and sharply bent; palp on maxilla with 12 and on labium with about 7 sensilla each; maxillary palp similar in length to galea ...... (Nothomyrmeciinae) Nothomyrmecia 7b. Hairs not as above; palps with 5 sensilla each; maxillary palp shorter than galea ...... (Myrmeciinae) Myrmecia Wheeler and Wheeler (1976: 87-88) found the larvae of Mjwwcirr more specialised than they had expected. In discussing "specialization", they stated: "We realize that primitive is not equivalent to generalized, but we did not expect the degree of specialization we found in Myrmecia". Their specialisation index for a hypothetical generalised ant larva was given as 0, while for their hypothetical ancestral larva it was 15. For Mymirciu, the specialisation index was given as 22, but following further analysis this should now be changed to 24. For Noth~,ni!~rmcc~irr,the specialisation index calculates to 24.5, but since fractions are inappropriate in such a subjective procedure we cite the index as 24 for both Mq'rmeciu and Notlioni!.rmcJc,iu. The similarities between the larvae of Mj'mieciu and Nothotn~rmci.iuin no way challenge the hypothesis of Taylor (1978) that these genera should be classified in the separate subfamilies Myrmeciinae and Nothomyrmeciinae [following Clark (195 1 ), but not Brown (1954)], and that they might well represent generalised survivors ofthe basal stocks of two major lines of ant evolution, the poneroid and formicoid phylads. Neither do these findings support Taylor's phylogenetic hypothesis over that previously suggested by Brown (1954). Brown classified Notho~~j'r.mec,r(iand Mjwwcia together in the subfamily Myrmeciinae, which he placed near the base ofthe hypothetical phylad represented by his "Myrmecioid Complex" of subfamilies. A second phylad included the subfamilies of his "Poneroid Complex". Taylor transferred Myrmeriu, but not Norhomj,rmecia, to the Poneroid Complex, and reinstated Clark's subfamily Nothomyrmeciinae to contain Notkomj~micr,icr.This process necessitated a new name for the erstwhile Myrmecioid Complcx sutw Mj-rmeciu, which he named the Formicoid Complex. Taylor's conclusions were based largely on aspects of adult abdominal exoskeletal structure and their possible NOTHOM YRMEC‘IA LARVAE AND EGGS 137 implications. A logically more appealing interpretation of the data on abdominal structure in ants must be provided in any challenge to his model. Obviously, the details of larval structure reviewed here can have little influence on these alternative hypotheses, especially considering the undoubted logic of Taylor’s statement (1978) that “these ants would be expected to retain primitive characteristics in common, even if exemplifying different basal lineages which we, in hindsight, see realised in modern ants”. This statement can logically be read as applying to the similarities between Myrmecia and Nothomyrmecia larvae and eggs. References BROWN,W. L. (1954).-Remarks on the internal phylogeny and subfamily classification of the family Formicidae. Insec1r.s soc. 1: 21-31. BROWN,W. L. and WILSON,E. 0. (1959).-The search for Norhomjrmecia. Wesr. Ausr. Nu!. 2 25-30. CLARK,J. (1934).-Notes on Australian ants, with descriptions of new species and a new genus. Mew. natn. MUS.Vic!. 8: 5-20. CLARK,J. (1951).-”The Formicidae of Australia, (I) Subfamily Myrmeciinae”. Commonwealth Scientific and Industrial Research Organisation: Melbourne. TAYLOR.R. W. (1978).-No!hom.~rmecia mucrops: a living-fossil ant rediscovered. Science, N. Y. 201: 979- 985. WHEELER.G. C. and WHEELER,J. (1971).-Ant larvae of the subfamily Myrmeciinae. Pan-Pacif. Ent. 47: 245-256. WHEELER,G. C. and WHEELER,J. (1976).-Ant larvae: review and synthesis. Mem. en!. Soc. Wash. 7: 1-108. [Munuscript recrired 23 November 1979.1