Axiagastus rosmarus Dallas, 1851

Kment, Petr, Lemaître, Valérie A., Webb, Michael D. & Roca-Cusachs, Marcos, 2025, Revision of the Austro-Oriental shield-bug genus Axiagastus (Hemiptera: Heteroptera: Pentatomidae), with the description of five new species, and taxonomic notes on related genera, Zootaxa 5603 (1), pp. 1-82 : 14-22

Axiagastus rosmarus Dallas, 1851

( Figs 9–43)

Axiagastus Rosmarus Dallas, 1851: 222 , pl. VIII: figs 5, 5a, 5b, 5c (original description and illustrations). Syntype (s): ♂, Philippines (BMNH).

Axiagastus Rosmarus : Walker (1867): 268 –269 (morphological variability, distribution; partim); Stål (1876): 94 (catalogue, distribution); Lethierry & Severin (1893): 170 (catalogue, distribution; partim); Breddin (1905): 210 (distribution; partim).

Axiagastus rosmarus : Dohrn (1859): 14 (checklist); Atkinson (1888): 129 (catalogue, redescription, distribution; partim); Distant (1901): 586 (comparison with A. mitescens ); Kirkaldy (1909): 126, 381 (catalogue, distribution; partim); Hoffmann (1935a): 149 (catalogue, distribution; partim); Black (1968): 563 (distribution, partim; record from Balabac); Takara & Azuma (1972): 102 (distribution; partim); Hua (2000): 171 (checklist, distribution; partim); Cassis & Gross (2002): 458 – 459 (catalogue, distribution; partim); Rider et al. (2002): 137 (checklist, distribution; partim); Rider (2006): 260 (catalogue, distribution; partim); Hayashi et al. (2016): 496 (distribution; partim).

Axiagastus rosmaus (incorrect subsequent spelling; misidentification): Zhang (1985): 85 (diagnosis, distribution).

Axiagastus rosmanus (incorrect subsequent spelling; misidentification): Zhang (1995): 42 (comparison with A. mitescens ).

Axiagastus rosmarius (incorrect subsequent spelling, misidentification): Arnold (2011): 43 (variability, distribution; Indonesia: Sumatra, Taiwan, Thailand).

Axiagastus Rosmarus (misidentification): Breddin (1901): 12 (distribution); Breddin (1905): 210 (distribution, partim; Sulawesi, Sumatra); Hoffmann (1935b): 59 (catalogue, distribution).

Axiagastus rosmarus (misidentification): Jensen-Haarup (1937): 322 (comparison with A. dubius ); Takara & Hidaka (1960): 183 –184, fig. 2 (distribution; Japan: Ryukyus ); Miyamoto (1965): 228 (distribution; Taiwan); Takara & Azuma (1972): 102 (distribution; Japan: Ryukyus ); Takara &Azuma (1973): 162 (checklist); Hsiao & Zheng (1977): 132 [footnote] (distribution; Taiwan); Zhang et al. (1980): 25 (distribution; China: Jiangxi); Zhang (1985): pl. XLVIII: fig. 176 (habitus drawing); Zhang et al. (1985): 10 (distribution; China: Hainan); Zhang & Lin (1986): 61 (distribution; China: Jiangxi); Miyamoto & Yasunaga (1989): 184 (checklist; Japan); Zhang & Lin (1990): 2 (distribution; China: Yunnan); Lin & Zhang (1993): 123, fig. 34 (redescription, habitus illustration, distribution; China: Fujian); Yasunaga et al. (1993): 228, pl. 110: figs 304a,b (diagnosis, habitus photos of adult and last instar larva, plant association, distribution); Zhang (1994): 33 (distribution; China: Jiangxi); Bu et al. (1995): 122 (distribution; China: Zhejiang); Zhang (1995): 42 (comparison with A. mitescens ); Hayashi (2002): 147 (checklist; Japan: Ryukyus ); Chandra & Rajan (2004): 31 (distribution; India: South Andaman); Arnold (2012): 55 (distribution; Indonesia: Sumatra, Malaysia: Sabah); Hoàng & Ðặng (2013): 789 (distribution; Vietnam); Aoyagi (2014): 65 –66 (plant association, distribution); Hayashi et al. (2016): 496 (checklist; Japan: Ryukyus ).

Axagastus rosmorus (incorrect subsequent spelling, misidentification): Su et al. (1993): 282 (distribution; China: Guangdong).

Aeschrocoris rosmarus (incorrect subsequent spelling; misidentification): Hua (1989): 43 (list).

Type locality. Philippines .

Type material. Lectotype (here designated): ♂ ( Figs 9–20), “ Phil / Isla [p, obverse]; 42 22 [hw, reverse] // Axiagastus [hw] / Rhaphigaster [p, hw crossing] / rosmarus [hw] / Walker’s catal. [p] // SYN- /TYPE [p, blue-margined disc] // ♂ [p] // [QR code] / NHMUK 013589093 [p] // LECTOTYPUS / AXIAGASTUS / ROSMARUS / Dallas, 1851 / des. SALINI et al. 2024 [p, red label]” ( BMNH). The specimen is pinned through the scutellum, left antennomeres IIb–IV, right antennomeres III–IV, left pro- and mesotarsomeres II–III, and complete left hind leg are missing; the dissected male genitalia are preserved in a glass microvial with a drop of glycerin, attached to the same pin.

Additional material examined. PHILIPPINES: Mindanao: Central Mindanao University Campus, 27.vii.2019, 7♂, 11♀ (1♀, dissected female genitalia preserved in a glass microvial with glycerol attached to the same pin, dissected and illustrated by S. Salini; 1♂, dissected male genitalia preserved in a glass microvial with glycerol attached to the same pin), Roca-Cusachs lgt. [collected on palm, Roca-Cusachs com. pers.], dissected and illustrated by Salini. S. (2♂, 2♀ NIM; 4♂, 6♀ MRCC; 2♂, 2♀ NMPC); Misamis Oriental, 2.5 km SW of Civoleg, 0.5 km S Haze Kaffe, 8°42.0′N 125°0.9′E, 1210 m a.s.l., clearing in broadleaved forest, at light, 31.v.2022, 1♀, J. Hájek, L. Sekerka & D. Vondráček lgt., P. Kment det. ( NMPC) .

Redescription. Colour, integument and vestiture. Body above pale olivaceous green with more or less well-expressed markings as follows: lateral margins of head, two narrow, longitudinal lines along lateral margins of clypeus (extending to posterior margin of head), lateral and posterior margins of pronotum, four small spots placed equidistant to one another on the anterior pronotal disc behind anterior pronotal margin, 1+1 small irregular spot on humeri, anterior margin of scutellum with small spot on each basal angle and 1+1 small spot submedially and, two oblique spots anteromedially to frenal margin, a moderately broad, transverse band outlining the large, pale levigate scutellar apical spot. Coria each with one or two small pale callose spots laterally near middle of endocorium and a moderately large, irregular black spot towards posterior end of corium or with additional black spot in anterolateral angle of corium. Antennae with scape pale yellow with lateral margin black, basipedicellite pale yellow or blackish, remaining antennomeres black with bases narrowly paler; connexivum with anterior and posterior fourth of laterotergites black and median half pale levigate; membrane fuscous. Body punctures black, less coarse, sparsely and uniformly distributed on head, pronotum, and anteromedian portion of scutellum, and coarser and denser posterolaterally on scutellum and corium. Body beneath pale, yellow; head on ventral surface with two narrow, longitudinal stripes on either side of bucculae and two oblique posteriorly diverging stripes on gula black; thoracic sternites with several transverse black stripes alternating with yellow ventral ground colour. Labium concolourous with ventral body colour except labiomere IV black. Ventral surface of abdomen with intersegmental sutures black. Ventrites III–VII laterally with black, transverse muscle scars, posteriad of spiracles and anteriad to a pair of trichobothria; a row of small spots laterally on ventrites III–VII, spiracular outline, moderately large spot coalescing with spiracle anteriorly, small round spot laterad of spiracle, longitudinal band at middle of ventrite VII, broad longitudinal bands outlining inner margins of valvifers VIII, black. Legs concolourous with ventral body colour with coarse, round punctures densely distributed on femora; punctures on tibiae fine, densely distributed; anterior apex of tarsomeres I–III including apical half of claws, black.

Structure. Labium reaching or surpassing posterior margin of ventrite IV. Other characters as in generic redescription.

Male genitalia ( Figs 13–20, 23–40). Genital capsule ( Figs 13–15, 23–26, 30‒33) subquadrangular; dorsal rim with deeper incision than ventral rim; lateral wall of dorsal rim slightly concave, with indistinct serrations (sr), ending in moderately angular projection contiguous with moderately deep, transverse emargination (te) at middle of dorsal rim; dorsal sinus of posterior aperture broadly U-shaped (ds), ventral margin of posterior aperture semicircular; posterolateral lobes with an angulation (a) in lateral view; ventral rim broadly V-shaped at middle (vr), infoldings of ventral rim deeply impressed on either side of moderately developed distension (dn) at middle, distension (dn) slightly emarginated at middle, and visible on ventral side of genital capsule as narrow roughly U-shaped concavity; infoldings of ventral rim laterally (inner to posterolateral lobes) with a narrow sclerotized, black ridge with sawtooth-like projections ending in a moderately developed, blunt denticle (dt). Paramere ( Figs 16–17, 27–29, 34‒36) simple, crown with an upright finger-like process (flp) forming an acute angle with a laminate disc (ld), dorsal margin (dm) convex, but with broad angulation, and abruptly narrowed towards apex, appearing as acute apex in lateral view ( Figs 17, 27, 34); numerous fine, elongate setae on finger-like crown and along periphery of laminate disc (ld); stem (st) moderately elongate, apodeme (am) disc-like. Phallus ( Figs 18–20, 38–40). Articulatory apparatus (ap) as in Fig. 37; phallotheca nearly as long as endophallus, less sclerotized, distal part transparent with ventral convex projection (vcp) in close proximity with inner margin of processes of aedeagus (pa); a pair of conjunctival processes (cp), sclerotized, elongate, ribbon-like with rounded apex, located dorsally; aedeagus (ad) short, deflected dorsad, apex swollen, drop-like, transparent, with embedded phallotreme and a pair of elongate processes of aedeagus as in Figs 20, 40.

Female genitalia ( Figs 41–43). Terminalia ( Figs 41, 42). Valvifers VIII (vlf 8) large, nearly quadrate with inner (mesial) margins straight, inner posterolateral angles truncate; valvifers IX (vlf 9) short, narrow, transverse plate with anterior margin biconcave and angulate at middle, posterior margin slightly concave; laterotergites IX (lt 9) elongate, apically rounded, reaching caudal margin of abdomen; laterotergites VIII (lt 8) short, subtriangular, outer margins convex, caudal margins smooth without denticle. Gynatrium. Orifice of spermathecal duct surrounded by narrow, inverted U-shaped sclerite (us); ring sclerites (rs) rounded with outline slightly pigmented brown. Spermatheca ( Fig. 43). Spermathecal dilation long, regularly fluted; proximal spermathecal duct tubular, distal spermathecal duct (dsd) gradually widened towards proximal flange; proximal flange slightly shorter than distal flange; apical receptacle orbicular with three ductules; a pair of short, subequal ones, and third one longest with apex twisted hook-like.

Differential diagnosis. Axiagastus rosmarus can usually be recognized by the luteous to olivaceous colouration with four small black spots nearly equidistant to one another on the anterior disc of the pronotum, and the characteristic structure of the genital capsule and parameres (parameral crown with abruptly narrowed apex and convex dorsal margin).

Etymology. Neither the etymology nor the gender of the name Rosmarus (originally spelled with an initial capital letter) was explicitly given in the original description. It is presumably a New Latin word, latinized from the Old Norse rosmhvalr, meaning “walrus” ( Anonymus 2024a,b), viz. Odobenus rosmarus (Linnaeus, 1758) ; noun in apposition. The name evidently refers to the long, “tusk-like” spines that originate from the anterior angles of the bucculae of Axiagastus , resembling the long tusks (elongated canines), which are present in both sexes but sexually dimorphic in walruses.

Plant association. The specimens from Central Mindanao University Campus were collected on an unknown palm species (M. Roca-Cusachs, pers. observ.).

Distribution. At present, the species is known only from the Philippines ( Dallas 1851, Walker 1867): Mindanao (this paper).

The following published records refer to the new species described in the current work: Indonesia: Sulawesi ( Walker 1867), Indonesia: Sumatra (Payakumbuh) and Malaysia: Sabah ( Arnold 2012) to A. prathapani sp. nov.; Australia: Queensland ( Cassis & Gross 2002) and New Guinea ( Walker 1867) probably to A. votypkai sp. nov.; China (Zhang 1985), China: Fujian ( Lin & Zhang 1993), Guangdong ( Fan 2011) and Jiangxi ( Fan 2011), India: Meghalaya ( Distant 1879, 1902) and Assam ( Atkinson 1888), Japan: Ryukyus ( Yasunaga et al. 1993, Aoyagi 2014), Taiwan ( Arnold 2011, as A. rosmarius ; Arnold 2012, as A. rosmarus ), and Thailand ( Walker 1867) to A. yeshwanthi sp. nov.

Specimens based on the following other records require further study to ascertain their identity: China: Guangdong ( Su et al. 1993, Lin & Zhang 1993, Hua 2000, Rider et al. 2002), Guangxi ( Fan 2011, no exact record), Hainan ( Zhang et al. 1985), Jiangxi ( Zhang et al. 1980, Zhang & Lin 1986, Lin & Zhang 1993, Zhang 1994, Hua 2000, Rider et al. 2002), Yunnan ( Zhang & Lin 1990, Hua 2000), Zhejiang ( Bu et al. 1995, Rider et al. 2002, Fan 2011); India: South Andaman ( Chandra & Rajan 2004); Indonesia: Sulawesi ( Breddin 1901, 1905, as A. rosmarus ), Sumatra ( Breddin 1905, as A. rosmarus ; Arnold 2011, as A. rosmarius ; Arnold 2012, as A. rosmarus ); Philippines: Balabac ( Black 1968); Thailand: Bankau ( Arnold 2011, as A. rosmarius ); Vietnam ( Hoàng & Ðặng 2013).

Remarks. Based on the examination of a diverse array of specimens in BMNH (many of which were assigned to A. rosmarus by various researchers), it was determined that the true A. rosmarus was found only from the Philippines (the type locality). Other specimens previously identified as A. rosmarus at BMNH, originating from various other localities, were found to belong to other species, which will be described herein, including a misidentified type of A. rosmarus (see A. yeshwanthi sp. nov.).

Axiagastus rosmarus was originally described from an unspecified number of male specimens (syntypic) with the following data: “a. Philippine Islands. From Mr. Cuming’s Collection”. The use of the letter “a” in Dallas (1851) indicates a provenance rather than a specimen (see below). A single specimen “a” with the type data was listed in Walker’s (1867) catalogue together with four other specimens from Siam, Celebes and New Guinea (“b–e”).A single specimen from the Philippines was found in the BMNH collection (presumed to be the one listed by Walker above) with the following (unusually small and easily overlooked) label: “Phil Isla ” [printed] and “42 22” [handwritten] on reverse (see Fig. 12). The number “[18]42-22”, in the Entomology acquisitions registers of the BMNH corresponds to a collection of 282 Hemiptera from the Philippines, which was purchased from Mr. Cuming. This specimen, when found, was not labelled as type, but another specimen from “ Siam ” in the same tray, presumably the specimen “d” listed above by Walker (see the Additional material examined section for A. yeshwanthi sp. nov.), was mislabelled with a green Walker’s type disc and a long printed name label for A. rosmarus ( Fig. 230). Similarly to other types attributed to Dallas and Walker in the BMNH, the printed name label was cut from the proof sheet by Walker and initially placed beneath the type series. It was later affixed to the type specimen by W.L. Distant (following his appointment in 1899), together with a red-margined disc for the types of Dallas and those of other authors and a green-margined one for Walker’s types; the word “ Type ” is printed at the centre of these early discs. Only one specimen of the species was identified as a type, regardless of the presence of a type series. Finally, a printed label, bearing the name of the species and the phrase “Walker’s catal.” was affixed to all specimens listed in Walker’s catalogue (see Fig. 12). For further information, see Smith (1893:7–8) and Roell et al. (2023: 17). It appears that, concerning the type of A. rosmarus, Distant (unaccountably) not only selected the incorrect type specimen but also affixed the wrong green-margined disc to it, mistakenly treating the specimen as the type of a species described by Walker.

In the present study, when assessing the type status of specimens used by William S. Dallas and Francis Walker to describe new species, the following statement by Roell et al. (2023) was taken into account: “Assessing type status—Information from original descriptions. Early authors did not (or only erratically) mention the number of individuals on which they were basing their descriptions. The type status of the species described by F. Walker in his various catalogues, however, is easily defined as he indicated the different listed specimens of each species with letters (a, b, c …), followed by the localities and the donors/sellers, when applicable (see Gray 1867a, 1867b, 1868). Although this pattern had previously been suggested for W. S. Dallas’s (1851) Catalogue (see Gray 1851), Dallas did not follow it (except, apparently, at the very beginning, see Solenosthedium attenuatum in Dallas (1851: 7)), and used the letters (a, b, c …) to indicate the full provenance (both locality and donor/seller, see Early type labels (last paragraph below). It is, thus, usually not possible to know how many specimens Dallas had in front of him when describing his new species.”

The possibility that the original description of A. rosmarus was based on a single specimen, the holotype fixed by monotypy ( ICZN 1999: Art. 73.1.2), was considered. However, although the original description (including figures) presents only a single value for the body length, notes that the two distal antennal segments are missing and states that the studied material was sourced from a single collection (Mr. Cuming’s collection) in a single locality ( Philippine Islands), this in itself does not preclude the possibility that there were more than one specimen, all of the same length, with the two distal antennal segments missing, and originating from the same provenance. Even the fact that Walker (1867) listed one specimen of A. rosmarus with the original data does not mean that Dallas had only one. Here are some examples to illustrate the above four points (only references to Walker’s catalogue are added as the reference to Dallas’s is indicated there): 1) Dallas mentions only one provenance (a), while Walker lists two, three, or even five specimens from it (e.g., Aelia americana , Homoeocerus chinensis , Pentatoma basalis and Mictis gallina ; Walker 1867: 271, 300 and Walker 1871: 22, 93, respectively); 2) A single value for the body length does not necessarily indicate the presence of only one specimen, as Dallas states that he has both male and female, and Walker lists two specimens with the same provenance (e.g., Nematopus affinis , as Notobitus affinis in Walker 1871: 79, and Metapodius granulosus, Walker 1871: 49 ); 3) Antennae (or any part thereof, or any structure) “wanting” does not necessarily indicate the presence of only one specimen as Dallas, even so, states that he has both male and female, while Walker lists two specimens (e.g., Pachylis tenuicornis, Walker 1871: 53 ); 4) Specimens went missing between Dallas’s and Walker’s re-curations of the collection (either due to deterioration or because Walker determined them otherwise; for a possible example of the former, see the entry for Arma cornuta and for an example of the latter, see the entry for Asopus leprosus , both in Roell et al. (2023). Therefore, the number of specimens that Dallas had when describing species remains unknown unless he clearly specified it (which he occasionally did, as indicated in the entry for Canthecona grandis in Roell et al. 2023). In contrast, the specimens that Walker had are always clearly documented; however, this does not necessarily reflect what Dallas had. Consequently, all of Dallas’s types are best assessed as syntypes, even if only one is found in the BMNH collection. For the above reasons, the single syntype specimen from the Philippines in the BMNH, NHMUK 013589093, is designated here as the lectotype (see ICZN 1999: Recommendation 73A).