Introduction
Ants (Formicidae) are among the ecologically most important insects, occurring in many terrestrial habitats in all zoogeographic regions, with their highest diversity in tropical forest habitats (Hölldobler and Wilson, Reference Hölldobler and Wilson1990; Grimaldi and Engel, Reference Grimaldi and Engel2005). The family currently is composed of 471 genera and almost 14,000 species placed in 16 extant and four extinct subfamilies (Bolton, Reference Bolton2014). The fossil record indicates that ants evolved in the Early Cretaceous between 110 and 130 Ma, and that a diversification of ant subfamilies representing most of the extant genera and their rise to ecological dominance took place in the Eocene (Grimaldi and Agosti, Reference Grimaldi and Agosti2000; Dlussky and Rasnitsyn, Reference Dlussky and Rasnitsyn2007; LaPolla et al.Reference LaPolla, Dlussky and Perrichot, Reference LaPolla, Dlussky and Perrichot2013). The Cenozoic record is extensive (except for the Paleocene) and includes three-dimensionally preserved ants from various amber deposits, such as Eocene Baltic (Dlussky and Rasnitsyn, Reference Dlussky and Rasnitsyn2009) and Indian amber (Rust et al., Reference Rust2010), and Miocene Mexican (Solórzano Kraemer, Reference Solórzano Kraemer2007) and Dominican amber (summarized in Arillo and Ortuno, Reference Arillo and Ortuno2005). In addition, there are Cenozoic compression and impression fossils from various localities, such as the Okanagan Highlands in British Columbia, Canada, and Washington State, USA (Ypresian; Archibald et al., Reference Archibald, Cover and Moreau2006); the Green River Formation, USA (Ypresian; Dlussky and Rasnitsyn, Reference Dlussky and Rasnitsyn2002), the Fur Formation in Denmark (Ypresian; Rust and Andersen, Reference Rust and Andersen1999; Archibald et al., Reference Archibald, Cover and Moreau2006); Florissant, Colorado, USA (Eocene/Oligocene; Carpenter, Reference Carpenter1930); and others (overview in LaPolla et al., Reference LaPolla, Dlussky and PerrichotReference LaPolla, Dlussky and Perrichot2013). However, all of these localities are in the Northern Hemisphere and comparatively little is known about the Cenozoic myrmecofauna in the Southern Hemisphere, making new finds such as those described here from New Zealand important for reconstructing the past diversity and geographic range of Formicidae.
The Southern Hemisphere record of fossil ants is very patchy. The South American record includes the myrmeciines Archimyrmex piatnitzkyi (Viana and Rossi, Reference Viana and Rossi1957) and A. smekali (Rossi de Garcia, Reference Rossi de Garcia1983) from the Paleocene–Eocene of Argentina (Dlussky and Perfilieva, Reference Dlussky and Perfilieva2003) and an undescribed ant fossil in mid-Miocene amber from Peru (mentioned in Perrichot et al., Reference Perrichot, Antoine, Salas-Gismondi, Flynn and Engel2014). The few sub-Saharan Africa fossils include the formicine Oecophylla leakeyi Wilson and Taylor, Reference Wilson and Taylor1964 from the Miocene of Kenya, and a single ant from the Eocene Mahenge maar in Tanzania (Harrison et al., Reference Harrison, Msuya, Murray, Jacobs, Báez, Mundil and Ludwig2001). The record for Australasia, a region with a generally highly diverse extant ant fauna, is similarly poor, with a few published records from mainland Australia: Hypoponera (Ponera) scitula (Clark, Reference Clark1934a) and Iridomyrmex Mayr, Reference Mayr1862 from Plio–Pleistocene copal at Allendale, Victoria (Hills, Reference Hills1957; Oke, Reference Oke1957); several formicid genera from the Miocene ″Upper Site″ locality at Riversleigh (Archer et al., Reference Archer, Hand and Godthelp1991); and 14 as-yet undescribed ants from the fossiliferous Cape York amber (Hand et al., Reference Hand, Archer, Bickel, Creaser, Dettmann, Godthelp, Jones and Wicks2010), for which a recent study suggested a origin in Southeastern Asia rather than Australia, and a maximum Eocene age (Sonibare et al., Reference Sonibare, Agbaje, Jacob, Faithfull, Hoffmann and Foley2014).The sole record from New Zealand is a member of the Amblyoponinae recently reported from the early Miocene Foulden Maar, Otago (Kaulfuss et al., Reference Kaulfuss, Harris, Conran and Lee2014a). In this paper, we describe a diverse fossil myrmecofauna, preserved as compression fossils, from Foulden Maar (Fig. 1), which allows a first evaluation of early Miocene ant diversity in New Zealand and the wider Australasian region. It enables comparisons with the extant fauna in Australasia and with the Northern Hemisphere Cenozoic record.
Map showing the location of the Foulden Maar fossil lagerstätte in southern New Zealand.
The Recent New Zealand ant fauna currently comprises 37 established species in 23 genera, placed in the subfamilies Amblyoponinae Forel, Proceratiinae Emery, Heteroponerinae Bolton, Ponerinae Lepeletier de Saint-Fargeau, Ectatomminae Emery, Myrmicinae Lepeletier de Saint-Fargeau, Dolichoderinae Forel, and Formicinae Latreille (with the Ectatomminae and Dolichoderinae comprising only introduced species). More than two-thirds of the established species are exotic or introduced, mainly from Australia and the Pacific islands, and only one genus, Huberia Forel, Reference Forel1890, and 11 species are considered endemic: Stigmatomma saundersi (Forel, Reference Forel1892), Discothyrea antarctica Emery, Reference Emery1895a, Heteroponera brounii (Forel, Reference Forel1892), Austroponera castanea (Mayr, Reference Mayr1865), A. castaneicolor (Dalla Torre, Reference Dalla Torre1893), Huberia brounii Forel, Reference Forel1895, H. striata (F. Smith, Reference Smith1876), Monomorium antarcticum (F. Smith, Reference Smith1858), M. antipodum Forel, Reference Forel1901, M. smithii Forel, Reference Forel1892, and Prolasius advenus (F. Smith, Reference Smith1862) (Don, Reference Don2007).
The small stock of both native and introduced ants in New Zealand has been referred to as depauperate or meagre in comparison to species numbers in Australia and the southwest Pacific region (e.g., Brown, Reference Brown1958a; Harris and Berry, Reference Harris and Berry2001; Gibbs, Reference Gibbs2006). For example, 1275 ant species are known from Australia (Shattuck, Reference Shattuck1999) and 1222 species from the Melanesian region, with ∼800 species in New Guinea (Janda et al., Reference Janda, Alpert and Borowiec2008), 187 species on Fiji (Sarnat and Economo, Reference Sarnat and Economo2012), and 125 species in the New Caledonian fauna (Ward, Reference Ward1985). However, Don (Reference Don2007) considered 11 endemic species in New Zealand as a realistic number with respect to its temperate climate, its isolated position in the south Pacific, and a general decline of ant diversity with increasing latitude. It has been recognized that the native ant fauna in New Zealand might be the remnant of a more diverse fauna in the past (Watt, Reference Watt1975; Gibbs, Reference Gibbs2006; Don, Reference Don2007) and the early Miocene fauna described herein from Foulden Maar suggests that this was indeed the case.
Material and methods
Photomicrographs of specimens were taken with a Nikon Coolpix 4500 camera attached to a Leica Wild MZ8 binocular microscope. Wetting the specimens with ethanol enhanced the visibility of anatomical structures and specimen images are composites of close-up photographs. For the preparation of drawings, we traced enlarged photographs, while checking anatomical structures of each specimen visually under the microscope. The resulting draft drawings were then scanned and digitalized. Solid, thick lines in the drawings indicate boundaries of visible sclerites, dashed lines trace supposed sclerite boundaries, thin solid lines show structures other than sclerite boundaries, e.g. deformation folds and elements of sculpture, and areas highlighted in grey denote the preserved, even if chemically modified, original sclerites. The drawings are based on the more complete parts and were finalized with morphological details preserved on the counterparts (where available). Measurements of anatomical structures were taken with an ocular micrometer and are given in millimeters as preserved, unless otherwise stated. Our nomenclature of the wing venation (Fig. 2) follows Rasnitsyn (Reference Rasnitsyn1980) and Dlussky (Reference Dlussky2009); other terminology is adopted from Bolton (Reference Bolton1994) and classification and synonymy from Bolton (2003) and Bolton (Reference Bolton2014). Fossils described herein are stored in the Department of Geology, University of Otago (OU); catalog numbers for each specimen are provided below.
Illustration of the wing nomenclature used by Dlussky (Reference Dlussky2009). Wings of a gyne of Gnamptogenys europaea (Mayr, Reference Mayr1868), neotype SIZK, no. UA-822, late Eocene, Rovno amber. Designations: C, R, RS, M, Cu, A: longitudinal veins; 1RS, RS + M, 2M, etc.: longitudinal vein sections; r-rs, r-m, m-cu, etc.: cross-veins; 1r + 2r, rm, 3r, mcu, cua: cells.
Locality and age
Foulden Maar is a maar-type fossil lagerstätte in the early to mid-Miocene Waipiata Volcanic Field in Otago, southern New Zealand (NZ Map Grid Infomap I43 reference 929166: 45.5271° S, 170.2218° E) (Fig. 1). Fossils at the site are preserved in finely laminated diatomite deposited in a ~1000 m diameter maar lake (Lindqvist and Lee, Reference Lindqvist and Lee2009) and include mainly micro- and macrofloral remains such as pollen, spores, seeds, fruits, flowers, leaves, fern fronds. The flora is indicative of an evergreen Lauraceae-dominated rainforest growing around the paleolake under a warm-temperate to subtropical climate, including some forest openings near the lake margin (Pole, Reference Pole1996; Bannister et al., Reference Bannister, Lee and Raine2005, Reference Bannister, Conran and Lee2012; Lee et al., Reference Lee, Conran, Lindqvist, Bannister and Mildenhall2012, Mildenhall et al., Reference Mildenhall, Kennedy, Lee, Kaulfuss, Bannister, Fox and Conran2014). Arthropod fossils found at the site include representatives of the Araneae, Plecoptera, Isoptera, Hemiptera, Coleoptera, Hymenoptera, Trichoptera and Diptera, making Foulden Maar the most informative site for terrestrial fossil arthropods in New Zealand (Kaulfuss et al., Reference Kaulfuss, Lee, Barratt, Leschen, Larivière, Dlussky, Henderson and Harris2014b). Insect taxa previously described in detail include diaspid scale insects (Harris et al., Reference Harris, Bannister and Lee2007), the termite Stolotermes kupe Kaulfuss, Harris and Lee, Reference Kaulfuss, Harris and Lee2010, the flat-bug Aneurus sp. (Aradidae) (Kaulfuss et al., Reference Kaulfuss, Wappler, Heiss and Larivière2011) and a winged ant in the subfamily Amblyoponinae (Kaulfuss et al., Reference Kaulfuss, Harris, Conran and Lee2014a).
The earliest Miocene age of Foulden Maar is based on radiometric dating of volcanics (Ar40/Ar39 age of 23.2 Ma for the maar eruption; Lindqvist and Lee, Reference Lindqvist and Lee2009) and on the palynological assemblage within the diatomite, which represents the uppermost Rhoipites waimumuensis Zone to lower early Proteacidites isopogiformis Zone (New Zealand local stages: late Waitakian–early Otaian) (Mildenhall et al., Reference Mildenhall, Kennedy, Lee, Kaulfuss, Bannister, Fox and Conran2014). The Foulden Maar diatomite represents a Southern Hemisphere mid-latitude lowland ecosystem under an ocean-influenced, seasonal warm-temperate to subtropical climate, following a period of major marine transgression in the Oligocene.
Systematic paleontology
Family Formicidae Latreille, 1809
Subfamily Amblyoponinae Forel, 1893
Genus and Species indet.
Amblyoponinae gen. et sp. indet., gyne, OU44898; (1) photomicrograph; (2) line drawing.
Description
Part and counterpart of gyne in dorsal view, head not preserved; body length (without head) 8.7 mm, estimated total body length about 10 mm, mesosoma length 3.3 mm; mesosoma width 1.4 mm, petiole length 1.37 mm, petiole width 1.05 mm.
Mesosoma elongate, 2.3 times longer than wide, with nearly parallel sides. Legs comparatively short and thick. Middle tibia with two spurs; first and second segments of middle tarsus both with a pair of long setae. Petiole a little longer than wide, with short anterior peduncle, without free posterior surface, broadly attached to first gastral segment, separated from gaster only by constriction. Second gastral segment a little wider and about as long as first segment. Tip of gaster acute, indicating the probable presence of a sting. Wings fragmentary preserved; wing venation indistinct.
Material
Specimen OU44898; gyne, dorsoventrally compressed, head not preserved; deposited in the Department of Geology, University of Otago.
Occurrence
Foulden Maar diatomite, Waipiata Volcanic Field, Otago, New Zealand; early Miocene.
Remarks
The assignment of the fossil as an amblyoponine ant is based on the construction of the petiole and the general habitus; two features that have been regarded traditionally as reliable for the subfamily Amblyoponinae (Brown, Reference Brown1960). Genera of Malagasy Amblyoponinae were revised by Yoshimura and Fisher (Reference Yoshimura and Fisher2012), who reassigned species of Amblyopone sensu lato into Xymmer Santschi, Reference Santschi1914, Amblyopone Erichson, Reference Erichson1842 sensu stricto and Stigmatomma Roger, Reference Roger1859, based on male morphological characters, which are not applicable to the gyne from Foulden Maar, as are morphological characters of workers (e.g. Brown, Reference Brown1960; Taylor, Reference Taylor1987). Two extant members of Amblyoponinae are present in the New Zealand fauna: the endemic S. saundersi (Forel, Reference Forel1892) and the introduced A. australis Erichson, Reference Erichson1842; only gynes of the latter species approach the size of the fossil. In the absence of morphological characters such as the structure of mandibles and wing venation, we assign the fossil to tribe Amblyoponini without placement into an extinct or extant genus.
Published fossil Amblyoponinae include a winged male without generic placement from Foulden Maar (Kaulfuss et al., Reference Kaulfuss, Harris, Conran and Lee2014a), Casaleia inversa (Dlussky, Reference Dlussky1981) (Kyrgyzstan, Middle Miocene), C. longiventris (Heer, Reference Heer1849) (early Miocene, Croatia), Myopopone sinensis Zhang, Reference Zhang1989 (Shanwang, China, early Miocene) and, from the Eocene Messel pit in Germany Stigmatomma groehni (Dlussky, Reference Dlussky2009), S. electrinum (Dlussky, Reference Dlussky2009), and C. eocenica (Dlussky and Wedmann, Reference Dlussky and Wedmann2012). Two of these species, C. eocenica and C. longiventris, were based on gynes, which can by readily distinguished from the gyne described here by a smaller body size and a much shorter and wider gaster (C. eocenica) and a wider petiole and first gastral segment, differing proportions of the gastral segments (in dorsal aspect) and a less pointed tip of the gaster (C. longiventris).
Subfamily Ectatomminae Emery, 1895
Genus Rhytidoponera Mayr, 1862
Rhytidoponera waipiata new species
Rhytidoponera waipiata n. sp., gyne, holotype, OU44899; (1) photomicrograph; (2) line drawing of the holotype; (3) reconstruction of forewing.
Diagnosis
Gyne: Body length about 6 mm; petiole 1.8 times longer than high, with short peduncle and low node, somewhat angulated in side view; forewing with closed cells 1+2r, 3r, rm and mcu; cross-vein cu-a meets M+Cu near cell mcu, proximal to section 1M, at a distance less than cu-a length. May be distinguished from R. kirghizorum Dlussky, Reference Dlussky1981 (Kyrgyzstan, Miocene) by body size (3.8 mm in R. kirghizorum) and form of the petiole (rounded in side view in R. kirghizorum).
Description
Part and counterpart of winged gyne, partly preserved in lateral view; length of preserved compression 4.7 mm, estimated total body length about 6 mm. Outlines of head and mouth parts obscure, head apparently longer then wide; mandibles rather short and apparently triangulate. Scape 1.2 mm long, longer than head (in living ant extending more than one-third of its length beyond the posterior corners of the head), funicular joints longer than thick. Mesosoma elongate, 2.5 times longer than high (2.8 mm/1.1 mm); dorsum of propodeum gradually rounded in lateral view. Legs rather long, tibial spur visible only on fore tibia. Pretarsal claws (visible on a fragment of middle leg) with median tooth. Petiole 1.9 times longer than high (0.77 mm/0.41 mm), with short peduncle and low node, somewhat angulate in side view. Helcium projects from about mid-height of anterior face of first gastral segment; dorsum of first gastral tergite gradually rounded above the helcium. First gastral sternite anteriorly with sharp angle. Head, mesosoma and petiole with coarse sculpture, visible as rows of grooves along sutures (partially filled by sediment).
Wing with veins and knots partly preserved. Cells 1+2r, rm and mcu closed, cells 1r and cua probably closed. Cell mcu pentagonal, but vein section 2M very short. Cross-veins 2r-rs and r-m meeting RS almost at one point; cross-vein cu-a close to cell mcu: vein section 2M+Cu shorter than cu-a.
Etymology
The species name refers to the Waipiata Volcanic Field, New Zealand, where the fossil was found.
Type
Holotype OU44899 (part and counterpart); a laterally compressed gyne; deposited in the Department of Geology, University of Otago.
Occurrence
Foulden Maar diatomite, Waipiata Volcanic Field, Otago, New Zealand; early Miocene.
Remarks
The distal position of the cross-vein cu-a is a character found in Myrmeciinae, Dorylinae and poneromorph subfamilies. The new species cannot belong to Myrmeciinae since they have longer mandibles (0.7–1.2 times as long as head). Dorylinae have wingless gynes (except most ants in the former subfamily Cerapachyinae) and differ in general wing venation. The poneromorph subfamily-group includes Amblyoponinae, Ectatomminae, Heteroponerinae, Paraponerinae, Ponerinae and Proceratiinae. Paraponerinae include the sole genus Paraponera F. Smith, Reference Smith1858, with one extant and one extinct species from the Neotropical Region. The new species differs from Paraponera in many habitual characters. In Amblyoponinae, the helcium projects from a very high position on the anterior face of abdominal segment III and in Ponerinae the helcium projects from a very low position on the anterior face of abdominal segment III, the latter with a high vertical anterior face above the helcium. Proceratiinae have an extremely large abdominal segment III. These characters are not found in the new species, which may therefore belong to Ectatomminae or Heteroponerinae. These subfamilies include the genera Ectatomma F. Smith, Reference Smith1858; Gnamptogenys Roger, Reference Roger1863; Rhytidoponera Mayr, Reference Mayr1862; Typhlomyrmex Mayr, Reference Mayr1862; Acanthoponera Mayr, Reference Mayr1862; Aulacopone Arnoldi, Reference Arnoldi1930; and Heteroponera Mayr, Reference Mayr1887. The new species can be excluded from Ectatomma, Typhlomyrmex, and Acanthoponera because these genera are only known from the Neotropical Region. The known species of Gnamptogenys, Aulacopone and Heteroponera differ from the new species in having rather short legs and scapes, a more compact mesosoma, and often an angulated propodeum. Only some Australian Rhytidoponera have relatively long legs and a long scape, elongated mesosoma, and a long propodeum with a gradually rounded dorsum. We therefore assign the new species to the genus Rhytidoponera, subfamily Ectatomminae.
The only described extinct species of Rhytidoponera is R. kirghizorum Dlussky, Reference Dlussky1981 reported from the Miocene of Chon-Tuz, Kotchkorka District, Kyrgyzstan. It differs from R. waipiata n. sp. by its smaller size (3.8 mm) and by the form of the petiole (rounded in side view).
Rhytidoponera gibsoni new species
Rhytidoponera gibsoni n. sp., gyne, holotype, OU44900; (1) photomicrograph; (2) line drawing.
Diagnosis
May be distinguished from gynes of R. waipiata n. sp. by the more robust mesosoma, shorter appendages, and the shape of the petiole with concave anterior and posterior and flat dorsal faces of petiole.
Description
Gyne in lateral view; total body length 11.6 mm, gaster enormously long (probably swollen due to water drift before burial). Head 1.75 mm long, 1.3 times longer than wide. Mesosoma 3.3 mm long, robust, less than twice as long as high. Dorsal and declivous sides of propodeum form rounded obtuse angle. Declivous side several times longer than dorsal side of propodeum. Legs rather short and thick. Petiole 1.1 times longer than high (1.12 mm/1.03 mm), with concave anterior and posterior sides and flat dorsal side. Tip of gaster acute with short sting.
Etymology
The species name honors the Gibson family on whose property the Foulden Maar lagerstätte is located.
Type
Holotype OU44900; a laterally compressed gyne; deposited in the Department of Geology, University of Otago.
Occurrence
Foulden Maar diatomite, Waipiata Volcanic Field, Otago, New Zealand; early Miocene.
Remarks
Although most key characters of Rhytidoponera are not preserved in the fossil, we assign the new species to this genus because its general habitus and particularly the distinctive petiole is very similar to some extant species of Rhytidoponera. Rhytidoponera gibsoni n. sp. differs from R. waipiata n. sp., described above, by the more robust mesosoma, shorter appendages, and by the shape of the petiole. Gynes of Gnamptogenys europaea (Mayr, Reference Mayr1868) in Baltic amber resemble R. gibsoni in the position of the helcium and by the robust mesosoma. However, their propodeum is distinctly angulated and the petiolar node is rounded in side view.
Subfamily Ponerinae Lepeletier de Saint-Fargeau, 1835
Tribe Ponerini Lepeletier de Saint-Fargeau, 1835
Genus Austroponera Schmidt and Shattuck, 2014
Austroponera schneideri new species
Austroponera schneideri n. sp., worker, holotype, OU44901; (1) photomicrograph; (2) line drawing.
Diagnosis
Worker: A ponerine ant morphologically very similar to extant Austroponera, in particular A. castanea (Mayr, Reference Mayr1865) but may be distinguished from the latter by a more slender constitution and the slightly more posterior position of the eyes.
Description
Worker, laterally compressed, middle and hind legs and most parts of antennae absent. Body length 5.8 mm, body without coarse sculpture. Mesosoma length 1.7 mm, petiole length 0.47 mm, petiole height 0.55 mm. Head length including mandibles 1.2 mm (1.0 mm without mandibles). Eyes nearly round, small, situated laterally, at about mid-length of head. Gena twice as long as maximum diameter of eye. Scape protruding slightly beyond the occipital margin of head. Middle segments of funiculus about as long as thick. Promesonotal dorsum slightly convex in side view, without distinct promesonotal groove. Mesopleural suture absent. Metanotal groove distinct. Propodeum angular in side view, with slightly convex dorsum, rounded angle, and nearly straight declivity; propodeal declivity nearly as long as propodeal dorsum. Petiole nearly triangular in side view, 1.2 times higher than long, dorsal tip rounded. Mesosoma and petiole without specialized structures such as spines, teeth, tubercles, etc. Helcium projecting from a very low position on the anterior face of first gastral (III abdominal) segment, the latter having a high vertical anterior face above the helcium.
Etymology
In honor of Jörg W. Schneider, emeritus professor at the Freiberg University of Mining and Technology, Germany.
Type
Holotype OU44901; a laterally compressed worker; deposited in the Department of Geology, University of Otago.
Occurrence
Foulden Maar diatomite, Waipiata Volcanic Field, Otago, New Zealand; early Miocene.
Remarks
The position of the helcium and the form of the first gastral segment undoubtedly indicate that this species belongs to the tribe Ponerini. Its habitus is very similar to Austroponera Schmidt and Shattuck, Reference Schmidt and Shattuck2014, particularly to the New Zealand endemic A. castanea (Mayr, Reference Mayr1865) (formerly Pachycondyla castanea) (Brown, Reference Brown1958a, fig. 1; Schmidt and Shattuck, Reference Schmidt and Shattuck2014, fig. 41). We therefore attribute the new species to Austroponera, although key characters of the genus (features of the clypeal margin and the mandibles, shape of the propodeal spiracle, etc.) are not discernable in the fossil. The new fossil species differs from A. castanea by a more slender constitution and the slightly more posterior position of the eyes. No other fossil species of Austroponera are known.
Subfamily Formicinae Latreille, 1809
Genus Myrmecorhynchus André, 1896
Myrmecorhynchus novaeseelandiae new species
Myrmecorhynchus novaeseelandiae n. sp., male, holotype, OU44569; (1) photomicrograph; (2) line drawing; (3) line drawing of left antenna.
Diagnosis
Male: waist with one segment (petiole). Antennae 13-segmented, geniculate, scape long, protrudes beyond the occipital margin of the head, all funicular segments longer than thick. Propodeum gradually rounded in side view. Forewing with closed cells 1+2r, 3r and mcu; closed cell rm absent. Vein sections 5RS and 4M with joint start (rs-m lost); cross-vein cu-a displaced to the base of wing.
Description
Part and counterpart of laterally compressed male. Body length about 6 mm, mesosoma length 2.2 mm, head length about 0.75 mm, head width about 0.85 mm. Head rounded, a little wider than long, mandibles presumably long, more than one-third of head length (outline of mandibles not visible). Antennae 13-segmented, geniculate, scape long, protrudes beyond the occipital margin of the head, all funicular segments longer than thick. Scutum overhangs pronotum. Propodeum gradually rounded in side view. Petiole indistinct. Genital parameres nearly as long as wide, with rounded tip.
Forewing 4.8 mm long, with closed cells 1+2r, 3r and mcu; closed rm absent. Pterostigma well developed. Vein sections 5RS and 4M with jointed start (rs-m lost). Cell 3r 4 times longer than wide. Cell 1+2r nearly three times longer than wide. Cell mcu trapezoid (RS+M appreciably shorter than 1Cu). Vein 1RS shorter than 1M. Cross-vein cu-a displaced to the base of wing; vein section 2M+Cu longer than 1Cu and three times as long as cu-a.
Etymology
Novaeseelandiae refers to New Zealand, where this species was once native.
Type
Holotype OU44569 (part and counterpart); a laterally compressed winged male; deposited in the Department of Geology, University of Otago.
Occurrence
Foulden Maar diatomite, Waipiata Volcanic Field, Otago, New Zealand; early Miocene.
Remarks
The wing venation of the new species is usual among genera of Formicinae. Some members of Dolichoderinae and Myrmicinae have a similar venation, but male representatives of these subfamilies possess antennae with a short scape. The six extant formicine species known from New Zealand are the endemics Prolasius advenus (F. Smith, Reference Smith1862) and Camponotus (Colobopsis) newsealanicus Donisthorpe, Reference Donisthorpe1940 and the introduced Anoplolepis gracilipes (F. Smith, Reference Smith1857) and three species of Paratrechina Motschoulsky, Reference Motschoulsky1863 (Brown, Reference Brown1958a; Taylor, Reference Taylor1987; Don, Reference Don2007). All species of these genera have forewings without a closed cell mcu, so the new species cannot belong to these taxa.
Native species of the following 18 formicine genera are known in the extant Australasian fauna: Calomyrmex Emery, Reference Emery1895a; Camponotus Mayr, Reference Mayr1861; Echinopla F. Smith, Reference Smith1857; Opisthopsis Dalla Torre, Reference Dalla Torre1893; Polyrhachis F. Smith, Reference Smith1857 (tribe Camponotini Forel); Acropyga Roger, Reference Roger1862; Prolasius Forel, Reference Forel1892; Stigmacros Forel, Reference Forel1905; Teratomyrmex McAreavey, Reference McAreavey1957 (Lasiini Ashmead); Melophorus Lubbock, Reference Lubbock1883 (Melophorini Forel); Myrmecorhynchus Andrè, Reference André1896; Notoncus Emery, Reference Emery1895b; Pseudonotoncus Clark, Reference Clark1934b (Myrmecorhynchini Wheeler); Notostigma Emery, Reference Emery1920 (Notostigmatini Bolton); Oecophylla F. Smith, Reference Smith1860 (Oecophyllini Emery); Paratrechina Motschoulsky, Reference Motschoulsky1863; Plagiolepis Mayr, Reference Mayr1861; and Pseudolasius Emery, Reference Emery1887 (Plagiolepidini Forel) (Taylor, Reference Taylor1987). Winged Teratomyrmex and Pseudonotoncus are not known. Amongst the remaining genera, the cell mcu is present only in Opisthopsis, Myrmecorhynchus and Notoncus, while wings of representatives of other genera are lacking this cell (Emery, Reference Emery1925). In Opisthopsis, the cell mcu is either triangulate (O. haddoni Emery, Reference Emery1893, O. major Forel, Reference Forel1902) or absent (male of O. respiciens moestus Wheeler, Reference Wheeler1918). In Notoncus, the cell mcu is present in gynes but absent in males. Only in Myrmecorhynchus both gynes and males have a trapezoid cell mcu (Clark, Reference Clark1934a). Similar wing venations occur in Formicini Latreille (Formica Linnaeus, Reference Linnaeus1758; Cataglyphis Förster, Reference Förster1850, etc.) and some Lasiini (Lasius Fabricius, Reference Fabricus1804; Acanthomyops Mayr, Reference Mayr1862; Myrmecocystus Wesmael, Reference Wesmael1838). However, all representatives of these genera inhabit the Northern Hemisphere, and there are no reasons to suppose that they were present in the Australasian region in the early Miocene. Similar wing venations are also present in Gesomyrmex Mayr, Reference Mayr1868 and Drymomyrmex Wheeler, Reference Wheeler1915. The native range of extant Gesomyrmex extends from the highlands of southern Borneo north into western India. In the Eocene, they also lived on the territory of modern Europe (Dlussky et al., Reference Dlussky, Wappler and Wedmann2009). Drymomyrmex is known exclusively from late Eocene Baltic amber (Wheeler, Reference Wheeler1915). The new species can be excluded from these genera because of the following differences: males of Gesomyrmex have much larger eyes and very short 8–11 segmented antennae. Only the gynes of Drymomyrmex are known and they possess 11-segmented antennae. Usually, in all known ants whose females have less than 12 antenna segments, males have less than 13 segments. We therefore consider the new species as a member of the genus Myrmecorhynchus. An additional argument in favor of Myrmecorhynchus is the similar length of mandibles in the new species, whereas males of the majority of other ants exhibit shorter mandibles.
Formicidae incertae sedis
Formicidae species A
Formicidae species A, isolated forewing, OU44903; (1) photomicrograph; (2) line drawing.
Description
Forewing, visible length 3.7 mm. Closed cells 1+2r, 3r and mcu; closed rm absent. Pterostigma well developed. Vein sections 5RS and 4M with separate start (rs-m present). Cell 3r 3.8 times longer than wide. Cell 1+2r nearly 2.5 times as long as wide. Cell mcu trapezoid (RS+M appreciably shorter than 1Cu). Vein section 1RS nearly as long as 1M. Cross-vein cu-a displaced to the base of wing: vein section 2M+Cu a little longer than 1Cu, and 2.2 times as long as cu-a.
Material
Specimen OU44903; a partly preserved forewing; deposited in the Department of Geology, University of Otago.
Occurrence
Foulden Maar diatomite, Waipiata Volcanic Field, Otago, New Zealand; early Miocene.
Remarks
The cross-vein cu-a displaced to the base of the forewing is a character found in Formicinae, Dolichoderinae, and Myrmicinae (Dlussky, Reference Dlussky1981). However, in Formicinae the vein sections 5RS and 4M initiate jointed (cross-vein rs-m absent) and most Myrmicinae possess an open cell 3r. Thus, the fossil wing could belong to a representative of Dolichoderinae. It is very similar to forewings of the extant Tapinoma ambiguum Emery, Reference Emery1925, but as all species of the Tapinoma erraticum group, including T. ambiguum, inhabit the Palearctic Region, this similarity is likely to represent convergence.
Formicidae Species B
Formicidae species B, isolated forewing, OU44904; (1) photomicrograph; (2) line drawing; (3) reconstruction of the wing. The wing has been torn and opened up; the lower part in (1) and (2) is the hind portion of the wing that has been displaced downward and to the left.
Description
Forewing, partly preserved in two fragments. Pterostigma well developed. Cells 1+2r, 3r, rm and mcu closed. Cell 3r 4.4 times longer than wide. Cross-vein rs-m more distal than r-rs, so cell rm pentagonal. Cell mcu pentagonal, 1.8 times longer than wide. Section 2M+Cu nearly as long as cross-vein cu-a.
Material
Specimen OU44904; a partly preserved forewing; deposited in the Department of Geology, University of Otago.
Occurrence
Foulden Maar diatomite, Waipiata Volcanic Field, Otago, New Zealand; early Miocene.
Remarks
The position of cross-vein cu-a in this wing is typical for members of the poneromorph subfamilies. The only fossil poneromorph ant with preserved wing venation from New Zealand is Rhytidoponera waipiata, described above. The described wing cannot belong to R. waipiata since wings of this species have a shorter cell rm (this character is not quite authentic since the cell rm was reconstructed in both specimens) and cross-veins r-rs and rs-m starting from RS at the same point while in Formicidae species B rs-m initiates distinctly more distally than r-rs.
Formicidae Species C
Formicidae species C, isolated forewing, OU44902; (1) photomicrograph; (2) line drawing.
Description
Forewing, 7.7 mm long, with closed cells 1+2r, 3r and mcu; closed rm absent. Pterostigma well developed. Vein sections 5RS and 4M with joint start (rs-m lost). Cell 3r 3.3 times longer than wide. Cell 1+2r nearly 2.6 times longer than wide. Cell mcu trapezoid (RS+M appreciably shorter than 1Cu). Vein 1RS nearly as long as 1M. Cross-vein cu-a displaced to the base of wing: vein section 2M+Cu a little shorter than 1Cu and nearly twice as long as cu-a.
Material
Specimen OU44902; an isolated forewing; deposited in the Department of Geology, University of Otago.
Occurrence
Foulden Maar diatomite, Waipiata Volcanic Field, Otago, New Zealand; early Miocene.
Remarks
The wing venation of the fossil is usual in the subfamily Formicinae. The wing differs from those of Myrmecorhynchus novaeseelandiae n. sp., described above, by its larger size, the proportions of cells 1+2r and 3r, and the comparatively short vein section 2M+Cu. The two native genera of extant Formicinae in New Zealand, Camponotus and Prolasius, both have forewings without a closed cell mcu.
Discussion
Comparison with the fossil record of Formicidae
The ants described here represent the first fossil myrmecofauna from New Zealand and provide a preliminary account of early Miocene (23 Ma) ant diversity in New Zealand and Australasia. Until now, the published record of fossil Formicidae from Australasia comprised the subfamilies Ponerinae and Dolichoderinae in Plio-Pleistocene copal from Allendale, Victoria, Australia (Hills, Reference Hills1957; Oke, Reference Oke1957). The few pre-Pleistocene ant fossils have not yet been described (Archer et al., Reference Archer, Hand and Godthelp1991; Hand et al., Reference Hand, Archer, Bickel, Creaser, Dettmann, Godthelp, Jones and Wicks2010) and those from the Cape York amber may represent a fauna from southeastern Asia (Sonibare et al., Reference Sonibare, Agbaje, Jacob, Faithfull, Hoffmann and Foley2014). Including the recently published record of an Amblyopone-like ant (Kaulfuss et al., Reference Kaulfuss, Harris, Conran and Lee2014a), the taxa from Foulden Maar now establish the subfamilies Amblyoponinae, Ectatomminae, Formicinae and Ponerinae in early Miocene (23 Ma) Australasia, with a presence in warm-temperate to subtropical New Zealand (approx. 45°S), shortly after a period of major marine transgression of the Zealandian subcontinent at ca. 25 Ma (New Zealand paleogeography and paleoclimate summarized in Lee et al., Reference Lee, Lee and Mortimer2001; Buckley et al., Reference Buckley, Krosch and Leschen2014, and references therein). Members of the subfamily Dolichoderinae were probably also present in the early Miocene fauna, as indicated by the isolated wing of Formicidae sp. A, but this needs to be confirmed by more complete fossils. Fossil records of Amblyoponinae from Croatia, Kyrgyzstan, China, and Foulden Maar in southern New Zealand indicate a wide geographic distribution of this subfamily during the Miocene epoch (Fig. 11), whereas in the Eocene amblyoponines are only known from Germany.
World map showing the occurrence of published fossil members of subfamily Amblyoponinae (Ambl.) and extinct species of Rhytidoponera (R), Myrmecorhynchus (M) and Austroponera (A) in one Eocene and four Miocene localities. (1) Messel Maar, Germany, Eocene (Lutetian): Amblyoponinae: Stigmatomma groehni, S. electrinum and Casaleia eocenica; (2) Foulden Maar, New Zealand, early Miocene: Amblyoponinae gen. et. sp. indet., Rhytidoponera waipiata n. sp., Rhytidoponera gibsoni n. sp., Myrmecorhynchus novaeseelandiae n. sp., Austroponera schneideri n. sp.; (3) Radoboj, Croatia, early Miocene: Amblyoponinae: Casaleia longiventris; (4) Shanwang Formation, China, middle Miocene: Amblyoponinae: Myopopone sinensis; (5) Kotchkorka District, Kyrgyzstan, middle Miocene: Amblyoponinae: Casaleia inversa, Rhytidoponera: R. kirghizorum.
On the genus level, the formicine Myrmecorhynchus novaeseelandiae n. sp. and the ponerine Austroponera schneideri n. sp. are the first fossil records globally, thus extending the fossil age of these two genera back to the early Miocene. Myrmecorhynchus is here reported from outside Australia for the first time, demonstrating a wider distribution of this genus within the Australasian region in the past. Rhytidoponera is a large genus containing 104 extant species native to New Guinea, the Solomon Islands, New Caledonia, Australia, and eastern Indonesia (Bolton, Reference Bolton2014; Shattuck, Reference Shattuck1999). The only previously known fossil of Rhytidoponera, a poorly preserved gyne described as R. kirghizorum from the middle–upper Miocene of Kyrgyzstan (Dlussky, Reference Dlussky1981; stratigraphic position in Prokofiev, Reference Prokofiev2007) indicates that the genus was formerly present in the Palearctic region. Rhytidoponera waipiata n. sp. and R. gibsoni n. sp. from Foulden Maar extend the fossil age of this genus back to ~23 Ma and indicate that the geographic range of Rhytidoponera included southern New Zealand in the early Miocene. Ectatomma (Rhytidoponera) europaeum (Mayr, Reference Mayr1868) described from Baltic amber was allocated to Gnamptogenys europaea within Ponerinae (Brown, Reference Brown1958b).
Comparison with the Recent Australasian ant fauna
Amblyoponinae, Ectatomminae, Dolichoderinae, Formicinae, and Ponerinae are diverse and widely distributed in warm temperate to tropical regions in modern Australasia (Ward, Reference Ward1985; Janda et al., Reference Janda, Alpert and Borowiec2008; Bolton, Reference Bolton2014) but they are low-diverse or absent in cooler climates on the isolated New Zealand land mass in the Southern Pacific. For instance, there are no native species of Ectatomminae and Dolichoderinae in New Zealand, although introduced species are now established or doubtfully established (Don, Reference Don2007). Subfamilies Amblyoponinae and Formicinae are represented in New Zealand by only one native species each, and Ponerinae include merely two species. It has been suggested that the few ants native to New Zealand are the relict of a more diverse fauna in the past (Watt, Reference Watt1975; Gibbs, Reference Gibbs2006; Don, Reference Don2007) but fossil evidence was missing until now. The small fossil myrmecofauna from Foulden Maar demonstrates for the first time that the early Miocene fauna indeed contained taxa that are now extinct in New Zealand, including Rhytidoponera and subfamily Ectatomminae in general, the formicine Myrmecorhynchus, now confined to Australia, as well as possibly subfamily Dolichoderinae. Similarly, Amblyoponinae seem to have been more diverse in the early Miocene fauna, with two fossil taxa at Foulden Maar but only one native species, Stigmatomma saundersi, in the modern fauna. This confirms previous suggestions of a long establishment of Amblyoponinae in New Zealand, which were based on the wide distribution of the extant S. saundersi on the North and South Islands and offshore islands such Three Kings and Chatham Islands (Brown, Reference Brown1958a, Reference Brown1973; Don, Reference Don2007).
Only one genus, the ponerine Austroponera Schmidt and Shattuck Reference Schmidt and Shattuck2014, is present in the taphocoenosis at Foulden Maar and in the modern New Zealand fauna. It is a small genus containing two species in northern parts of New Zealand (A. castanea and A. castaneicolor) and one species, A. rufonigra (Clark, Reference Clark1934a), in Australia. Austroponera schneideri n. sp. described herein provides rare fossil evidence for the antiquity dating back to at least 23 Ma of some New Zealand insect genera, as has previously been shown for the bibionid Dilophus Meigen, Reference Meigen1803 from the Eocene (Harris, Reference Harris1983), the termite Stolotermes and the aradid Aneurus Curtis, Reference Curtis1825 from the early Miocene (Kaulfuss et al., Reference Kaulfuss, Harris and Lee2010, Reference Kaulfuss, Wappler, Heiss and Larivière2011) and at least two Miocene beetle genera yet to be described in detail (Kaulfuss et al., Reference Kaulfuss, Lee, Barratt, Leschen, Larivière, Dlussky, Henderson and Harris2014b).
In the absence of post-early Miocene Formicidae fossils in New Zealand, the timing of extinction of the above mentioned ant subfamilies and genera is currently not assessable. A major diversity decline and extinction due to late Miocene to Pliocene climate cooling and Pleistocene glacial cycles has been shown for many subtropical floral elements such as eucalypts, Araucaria, the Brassospora group of Nothofagus and Casuarinaceae, accompanied by a simultaneous appearance of temperate and alpine plants in New Zealand (McGowran et al., Reference McGowran, Archer, Bock, Darragh, Godthelp, Hageman and ….Warne2000; Lee et al., Reference Lee, Lee and Mortimer2001, Reference Lee, Conran, Lindqvist, Bannister and Mildenhall2012; Gibbs, Reference Gibbs2006). For invertebrates, molecular evidence indicates radiation of many groups, such as weta, cicadas, grasshoppers, cockroaches, and snails, in response to newly formed alpine habitats during the Pliocene, and local extinction of many groups during Pleistocene glacial cycles (Watt, Reference Watt1975; Gibbs, Reference Gibbs2006; Goldberg et al., Reference Goldberg, Trewick and Paterson2008 and references therein). Climate variables such as temperature, rainfall, and humidity are key factors in ant distribution (Hölldobler and Wilson, Reference Hölldobler and Wilson1990; Kaspari et al., Reference Kaspari, Alonso and O’Donnell2000). Thus, it is very likely that climate cooling since the late Miocene was also a main factor for the complete loss (Ectatomminae, Dolichoderinae, Myrmecorhynchus) or local extinction (north shift of Austroponera) of the early Miocene, warm-temperate to subtropical ant fauna in New Zealand.
The effect of major late Oligocene (ca. 25 Ma) marine transgression on the New Zealand ant fauna (and other terrestrial arthropods) is very poorly known (Buckley et al., Reference Buckley, Krosch and Leschen2014), mainly due to the lack of fossils. Our results show that some ant genera from Foulden Maar are no longer present in New Zealand, which suggests that the early Miocene myrmecofauna at 23 Ma, shortly after the maximum marine transgression, was taxonomically different and possibly more diverse. We thus conclude that the previous assumption that the small extant New Zealand myrmecofauna might reflect the failure of ants to re-colonize New Zealand after the Oligocene marine transgression (Cranston, Reference Cranston2009) is no longer valid.
Acknowledgments
Our thanks go to the Gibson family and Featherston Resources Ltd. for allowing us access to the fossil site at Foulden Maar. We are grateful to B. Barratt, A.C. Harris, M. Solórzano Kraemer, S. Martin, R. Debbie, and M. Archer for helpful information and literature on living and fossil ants; K. Perfilieva for help in the interpretation of wing venations; the Queensland Museum for providing photographs of Australian ant specimens; and the Department of Botany, Otago University for providing photographic equipment. D.E. and W.G. Lee and Vincent Perrichot provided helpful and constructive comments on the manuscript. Our study has been funded by the Russian Foundation for Basic Research (grant nos. 08-04-00701 and 11-04-00421), and the University of Otago, Division of Science Scholarship, and the Marsden Fund by the Royal Society of New Zealand (awarded to D.E. Lee, University of Otago).
