Malagidris

Bolton, Barry & Fisher, Brian, 2014, The Madagascan endemic myrmicine ants related to Eutetramorium (Hymenoptera: Formicidae): taxonomy of the genera Eutetramorium Emery, Malagidris nom. n., Myrmisaraka gen. n., Royidris gen. n., and Vitsika gen. n .., Zootaxa 3791 (1), pp. 1-99 : 16-19

Genus Malagidris

(Figs 10-29, Maps 126-131)

Malagidris Bolton & Fisher nom. n.

Brunella Forel , 1917: 234. Type-species: Aphaenogaster belti Forel , 1895: 248, by monotypy. (Junior synonym of Aphaenogaster Mayr, 1853: 107: Bolton, 1982: 341; revived from synonymy here.) [Junior homonym of Brunella Smith , G.W. 1909: 87 (Crustacea).]

Diagnosis of worker

Monomorphic myrmicine ants.

Mandible triangular; masticatory margin with 8-13 teeth, longer than the basal margin.

Palp formula 5,3.

Stipes of maxilla with a weak transverse crest.

Clypeus posteriorly moderately broadly inserted between the frontal lobes (width of clypeus between the lobes as great as or greater than width of one of the lobes); median portion of clypeus longitudinally costulate, not bicarinate.

Clypeus with a stout unpaired seta at the midpoint of the convex anterior margin.

Frontal carinae short, restricted to well defined but narrow frontal lobes.

Antennal scrobes absent.

Antenna with 12 segments, with an apical club of 3 segments that may be very slender.

Torulus with upper lobe visible in full-face view; maximum width of torulus lobe is just posterior to the point of maximum width of the frontal lobes.

Eyes present, located in front of the midlength of the head capsule in full-face view.

Head capsule without a median, longitudinal carina; occipital carina conspicuous.

Promesonotal suture absent; pronotal humeri weakly angulate.

Pronotum plus anterior mesonotum swollen and distinctly convex in profile; dorsalmost point of promesonotum at a considerably higher level than the long propodeal dorsum.

Propodeum bispinose; propodeal lobes rounded.

Propodeal spiracle behind midlength of sclerite, at about midheight of side of propodeum and far in front of the margin of the declivity; spiracle separated from apex of metapleural gland bulla by at least one spiracle diameter.

Metasternal process obsolete, at most a narrow crest on each side of the metasternal pit, each crest sometimes extended posteriorly as a narrow carina.

Legs long and slender.

Tibial spurs: mesotibia 1; metatibia 1, both simple.

Abdominal segment 2 (petiole) with a long, slender anterior peduncle; spiracle slightly in front of the midlength of the peduncle. Abdominal segment 2 node short, rounded to subconical in profile.

Subpetiolar process an anteroventral denticle.

Abdominal segment 3 (postpetiole) elongate, not dorsoventrally flattened, not markedly broader than high; ventral surface in profile flat to shallowly convex.

Stridulitrum present on pretergite of abdominal segment 4.

Abdominal tergite 4 (first gastral) does not broadly overlap the sternite on the ventral gaster; gastral shoulders absent.

Sting strongly developed, simple.

Main pilosity of dorsal head and body: simple, often sparse, usually absent from propodeal dorsum.

Diagnosis of queen (gyne)

This caste is alate in belti , ergatoid in alperti , jugum and sofina , and remains unknown in dulcis and galokoa . The alate queen matches the description of the worker except for its conspicuous ocelli and the structure of the enlarged mesosoma, where there is a full complement of flight sclerites and a broad sulcus across the mesopleuron. The ergatoid forms match the definition of the worker but with ocelli variably developed. In alperti and the single known ergatoid of jugum ocelli are absent, although a vague pit-like depression in one alperti ergatoid may indicate the last vestige of the median ocellus. Among the known ergatoids of sofina some retain three distinct ocelli, while others exhibit only a reduced median ocellus. The mesonotum is somewhat larger and better defined than in the worker, with a fully fused but more obvious promesonotal suture across the dorsum. The mesopleuron shows a reduced or absent transverse sulcus. The propodeal spines are shorter and stouter than in the respective workers, the postpetiole is more swollen and the gaster is more voluminous. Venation: see under male.

Diagnosis of male

Males taken in association with workers are known for alperti , belti , jugum and sofina . In addition, males of two more species that appear to belong in this genus are known. One of them is tentatively associated with dulcis here, but this remains to be confirmed; the other does not associate with any known worker but appears morphologically closest to jugum , it is discussed there. Slightly smaller than conspecific worker. Mandible triangular and strongly dentate, with 6-11 sharp teeth. Palp formula 5,3. Stipital crest present on maxilla. Antenna with 13 segments, long and filiform. SI 42-79. First funicular segment short, not globular, about one quarter to one half the length of the second funicular segment. In full-face view eye located in front of midlength of head capsule. Ocelli conspicuous. Occipital carina sharp, forming a distinct crest. Mesotibia and metatibia each with a single simple spur. Notauli very reduced or absent. Mesopleuron with a marked transverse sulcus. Mesoscutum convex in profile, the mesoscutum and mesoscutellum elevated, much higher than the propodeal dorsum, which is depressed and slopes downward posteriorly. Propodeum unarmed, the spiracle high on the side and at about the midlength, or slightly in front of the midlength, of the sclerite; propodeal lobes conspicuous, rounded. Petiole with a long anterior peduncle and a low node, the spiracle at or behind the midlength of the peduncle, but in front of the level of the node. Subpetiolar process present, small. Parameres large. Cerci present.

Forewing venation (based on males as available belti queen specimens are dealate). Rs f4-5 does not meet Rf3 on anterior margin of wing (= marginal cell open). 2rs-m absent. 1m-cu present. Fusion of Rs+M extended distally, so that 1m-cu arises from Rs+M, not from M. Rsf3 absent (Rs+M divides into Rsf4-5 and Mf3-4 at or distal of the junction with 2r-rs). Af2 a mere stub distal of cu-a; the latter is retracted and arises from M+Cu, proximal of the point where it divides into Mf1 and Cuf1-2.

Comments on Malagidris

The type-species of this genus, belti (Forel, 1895), has had a moderately varied taxonomic history. It was originally described in the genus Aphaenogaster Mayr (1853), even though Forel remarked on its 3-segmented antennal club, rather than 4 as is usual in that genus. This character, coupled with the presence of angulate humeri and a queen with a depressed mesosoma, caused Emery (1915a: 68) to exclud belti from Aphaenogaster and transfer it to Atopula Emery (1912). Forel (1917: 234) decided that Atopula was artificial, "composed of disparate species," and established the genus Brunella to include only belti . The component species of Atopula were later dispersed to other genera by Bolton (1976: 362), who retained genus Brunella as its "affinities are unclear." Later however, Bolton (1982: 341) synonymised Brunella under Aphaenogaster , thus returning belti to its original generic combination.

Extensive recent sampling of the Madagascan ant fauna by Brian Fisher and associates has made it clear that this synonymy was incorrect. The discovery of several species referable to Forel’s Brunella has allowed the diagnosis of a distinct group of Madagascan endemics, which are convergent in some characters with Aphaenogaster but certainly not congeneric with it. The final act of this history has been the realisation that Brunella Forel (1917) is the junior homonym of a crustacean genus Brunella G.W. Smith (1909), from Tasmania. No replacement name was essential for Brunella Forel while it was a junior synonym, but now that it is revived from synonymy a replacement name is necessary: Malagidris nom . n.

The larger, more gracile species (e.g. alperti , galokoa , jugum , sofina ) of Malagidris are remarkably convergent on the widely distributed genus Aphaenogaster . However, all species of Malagidris have two critical features never exhibited by Aphaenogaster species. First, the midpoint of the anterior clypeal margin of Malagidris has a single, stout, unpaired seta. In Aphaenogaster there is always a conspicuous pair of setae, one on each side of the midpoint of the anterior clypeal margin. Second, Aphaenogaster species do not have the characteristic structure of the anteroventral peduncle of the petiole, as described under the definition of Eutetramorium group. In addition to these, Malagidris always has the following: a transverse crest present on the stipes of the maxilla; a 3-segmented antennal club; a subpetiolar process present; a strongly developed sting; the anterior clypeal margin convex at the midpoint. In Aphaenogaster , by contrast, the stipes usually lacks a crest (a crest is incompletely and weaky developed in a few species, strong only in A. relicta Wheeler & Mann , from Haiti, which may not be properly referable to Aphaenogaster ), usually has a 4-segmented antennal club (5-segmented to gradually incrassate in some species but never 3-segmented), lacks a subpetiolar process, has a very weakly developed or vestigial sting, and usually (but not always) has the midpoint of the anterior clypeal margin concave or indented.

Synopsis of Malagidris species alperti Bolton & Fisher sp. n.

belti (Forel, 1895)

dulcis Bolton & Fisher sp. n.

galokoa Bolton & Fisher sp. n.

jugum Bolton & Fisher sp. n.

sofina Bolton & Fisher sp. n.

Key to Malagidris species (workers)

1 Head relatively broad (CI 81-84). Scape relatively short (SI 97-108); when scape laid straight back on head, in full-face view, its apex reaches, or fractionally exceeds, the posterior margin (Fig. 14). Metafemur relatively short, MfL/HW 1.25-1.35. Subpetiolar process subtended by a narrow cuticular crest that extends posteriorly beyond the level of the spiracle......... belti

- Head relatively narrow (CI 56-74). Scape relatively long to extremely long (SI 145-231); when scape laid straight back on head, in full-face view, its apex very conspicuously surpasses the posterior margin. Metafemur relatively long, MfL/HW 1.70-3.13. Subpetiolar process not subtended by a cuticular crest................................................2

2 More densely setose species: in full-face view 4-5 setae project from side of head behind eye, and in addition 1-2 project out from below the eye (Fig. 17); a pair of setae present in the metanotal groove, and at least 1 pair of setae on propodeal dorsum (Fig. 16). Metafemur relatively shorter, MfL/HW 1.70-1.83, in anterior or posterior view relatively more stout, MfL/MfH 4.60-5.50........................................................................................ dulcis

- Less densely setose species: in full-face view 0-3 setae project from side of head behind eye, setae never project out from below the eye; without setae in the metanotal groove, and propodeal dorsum without setae. Metafemur relatively longer, MfL/ HW 1.95-3.20, in anterior or posterior view relatively slender, MfL/MfH 5.77-12.29................................3

3 Scape extremely elongated, SI 201-231; head relatively narrow, CI 56-62 (Fig. 26); metafemur relatively very long and slender, MfL/HW 2.81-3.20, MfL/MfH 10.24-12.29. Pronotal dorsum predominantly very densely reticulate-punctulate (Fig. 27), also with minute, low, weak disorganised rugulae, but the rugulose sculpture is secondary to the reticulate-punctulate component............................................................................................. sofina

- Scape shorter, SI 160-192; head relatively broader, CI 65-73; metafemur relatively shorter and deeper, MfL/HW 1.95-2.65, MfL/MfH 5.77-8.92. Pronotal dorsum predominantly reticulate-rugose; spaces between the rugae smooth, or with punctulate ground sculpture, any punctulae present are usually distinctly secondary to the rugose component...................... 4

4 Propodeal spines very broad basally, extremely elevated, and so strongly recurved that their apices are directed anterodorsally (Fig. 19). Smaller species, HW 0.72-0.78, SL 1.20-1.30, MfL 1.42-1.54 ................................... galokoa

- Propodeal spines straight to feebly sinuate, never recurved so that their apices point anterodorsally. Larger species, HW 0.82-1.12, SL 1.45-1.98, MfL 1.86-2.76...................................................................5

5 Median clypeal seta arises from a prominent, raised, false anterior margin (Fig. 22), below which is a vertical or reflexed section of the clypeus that descends to the true anterior margin; outline of clypeus in profile is roughly an anteriorly-directed V-shape. Pronotal dorsum usually with 1-2 pairs of setae present. Side of head in front of eye with a single projecting seta (rarely absent), side behind eye with 1-3 projecting setae (Fig. 23). With head in profile the dorsum above the eye flat or very nearly flat, without a distinct concave impression or indentation.................................................. jugum

- Median clypeal seta does not arise from a raised false anterior margin (Fig. 10); outline of clypeus in profile is more or less evenly convex. Pronotal dorsum usually without setae. Side of head, both in front of and behind the eye, without projecting setae (Fig. 11). With head in profile the dorsum above the eye distinctly indented and concave, the outline of the impression rising posteriorly to the highest point of the vertex....................................................... alperti

Provisional key to males of Malagidris species

1 With petiole in dorsal view the peduncle with a pair of laterally projecting tubercles, upon which the spiracles are borne; width across the tubercles is distinctly greater than the maximum width across the node ................................... 2

- With petiole in dorsal view the peduncle with a slight convexity at the site of the spiracles, without tubercles; width across the convexity is less than the maximum width across the node ..................................................... 3

2 Postpetiole in dorsal view and in profile longer and more slender, length/width ca 1.90 or more, length/height> 2.00. Head longer and narrower, CI <80. Larger species, HW> 0.80, MfL> 1.80........................................ jugum

- Postpetiole in dorsal view and in profile shorter and more stout, length/width ca 1.25, length/height ca 1.43. Head shorter and broader, CI 84. Smaller species, HW 0.72, MfL ca 1.10....................unassociated male (discussed under jugum )

3 With head in profile the posterodorsal outline, behind the ocelli, vertical or very nearly vertical. Head relatively broader, CI 80-84 ...............................................................................................4

- With head in profile the posterodorsal outline, behind the ocelli, conspicuously sloped to the posterior margin, not vertical. Head relatively narrower, CI 67-74........................................................................5

4 With head in profile the vertical posterior surface terminates ventrobasally in a posteriorly projecting lobe, which has the occipital carina at its apex. Eye relatively slightly larger, EL/HW 0.58-0.61.................................... belti

- With head in profile the vertical posterior surface descends directly to the occipital carina, without a posteriorly projecting lobe. Eye relatively slightly smaller, EL/HW 0.50................................................ dulcis (putative)

5 Scape relatively short, SI 45-48. First funicular segment of antenna relatively long, length of first segment/length of second segment 0.38-0.40. Smaller species, HW <0.70, MfL <1.50, maximum width of mesoscutum <0.75.............. alperti

- Scape relatively long, SI 77-79. First funicular segment of antenna relatively very short, length of first segment/length of second segment 0.23-0.26. Larger species, HW> 0.70, MfL> 1.90, maximum width of mesoscutum> 0.85............ sofina