Introduction

The early to mid-Miocene Manuherikia Group in Otago, southern New Zealand comprises a non-marine sedimentary sequence that accumulated in a temporally (19–16 Ma) and spatially (about 5600 km²) extensive paleolake and associated fluvial plain environments (Douglas 1986; Mildenhall & Pocknall 1989). Fluvial conglomerates and sandstones and palustrine lignites in the lower part of the Manuherikia Group (Dunstan Formation) are overlain by typically non-carbonaceous, fine-grained lacustrine sediments of the Bannockburn Formation (Douglas 1986). The rich fossil record from the Manuherikia Group includes diverse and differentiated floral communities and various aquatic and terrestrial vertebrates, such as fish, amphibians, reptiles, birds and mammals (Pole 1993; Pole et al. 2003; McDowall & Lee 2005; Tennyson 2010; Schwarzhans et al. 2012; Worthy et al. 2017; Hand et al. 2018 and references therein). By contrast, the published invertebrate record is comparatively sparse, comprising two genera of ostracods, Paralimnocythere sp. and Gomphocythere sp. (Lindqvist 1994), the bivalve Echyridella sp. (Douglas 1986), land and freshwater gastropods (Marshall 2011; Marshall & Worthy 2017) and the freshwater crayfish Paranephrops fordycei (Feldmann & Pole 1994). To date, no insect fossils have been documented, although insect-damaged leaves and the presence of insectivorous bats (Hand et al. 2015) have provided indirect evidence for insect life in habitats associated with Lake Manuherikia. Given the diversity of insects in modern lacustrine and fluviatile environments (e.g. Winterbourn et al. 2006), and the abundance of fine-grained Manuherikia sediments suitable for the preservation of small and delicate organisms, the previous absence of insect fossils probably reflects incomplete sampling.

Between 2016 and 2018, several insect fossils were collected by the first author from plant-bearing claystones of the Dunstan Formation at Bannockburn Inlet near Cromwell (Figure 1). Here, we describe three isolated wings from this collection as the first insects to be reported from the Manuherikia Group. The three taxa included in this preliminary account are from the insect orders Hemiptera, Coleoptera and Trichoptera, and represent the first fossil records of their families from the Australasian region.

First insects from the Manuherikia Group, early Miocene, New Zealand

All authors

Uwe Kaulfuss, Samuel D. J. Brown, Ian M. Henderson, Jacek Szwedo & Daphne E. Lee

https://doi.org/10.1080/03036758.2018.1477054

Published online:

14 June 2018

Figure 1. Map of locality F41/f249 at the Bannockburn Inlet near Cromwell.

Display full size

Materials and methods

The three insect fossils included in this study are held in the Geology Museum (OU), University of Otago, Dunedin, New Zealand; collection numbers for specimens are provided below. Our terminology for wing venation in Hemiptera follows Bourgoin et al. (2015), in Coleoptera Lawrence & Ślipiński (2013) and in Trichoptera Holzenthal et al. (2007). Measurements are given in mm as preserved (unless stated otherwise).

Locality description

The fossil site is located at Bannockburn Inlet near Cromwell, Central Otago (−45.0835E, 169.1717N), where fine-grained, fossiliferous sediments are exposed in a steep-sided gully. The locality is registered as F41/f249 in the New Zealand Fossil Record Electronic Database (Pole 1992). A steeply dipping sedimentary sequence at the site includes a c. 1 m thick package of carbonaceous mudshales and silty claystones with rich floral assemblages (Figure 2). The insects described herein were collected from two claystone horizons which are rich in organic detritus, diverse fern fronds, numerous angiosperm leaves and monocots resembling arecoid palms. Many of the angiosperm leaves are insect-damaged, with damage types related to hole feeding, margin feeding and skeletonisation being most common.

First insects from the Manuherikia Group, early Miocene, New Zealand

All authors

Uwe Kaulfuss, Samuel D. J. Brown, Ian M. Henderson, Jacek Szwedo & Daphne E. Lee

https://doi.org/10.1080/03036758.2018.1477054

Published online:

14 June 2018

Figure 2. Fossiliferous section of Dunstan Formation sediments where the insect fossils were found (F41/f249). A, Outcrop situation in August 2016. Parts of the steeply dipping sequence collapsed in winter 2017. Spade (circled) for scale; B, graphical representation of the fossiliferous sequence with the two insect-bearing claystone horizons.

Display full size

Overlying the carbonaceous lithologies is a coarsening-upward sequence of homogeneous silt to sandstone with few leaves. The lower third of the sequence consists of structureless fine to medium sandstones and laminated mud and siltstones. Sand flasers and lenticular medium sand bodies up to 10 cm wide are commonly observed near the base of the sequence, which, except for coalified organic detritus and rare freshwater bivalves, appears to be free of macrofossils. According to facies analysis in Douglas (1986), the sequence records a transition from well-drained lake margin and interdistributary bay facies to poorly drained interdistributary bay and swamp facies and then back to lake margin facies. Stratigraphically, these sediments belong to the Cromwell sub-member of the Dunstan Formation (Douglas 1986). Climatically, the early Miocene in southern New Zealand was characterised by mesothermal, ‘warm-temperate to subtropical’ conditions (Mildenhall et al. 2014; Pole 2014; Reichgelt et al. 2015).

Systematic paleontology

Hemiptera

Family Tropiduchidae Stål, 1866

Material examined. OU46665, part and counterpart of the mid-portion of an isolated, possibly slightly deformed tegmen (Figure 3)

First insects from the Manuherikia Group, early Miocene, New Zealand

All authors

Uwe Kaulfuss, Samuel D. J. Brown, Ian M. Henderson, Jacek Szwedo & Daphne E. Lee

https://doi.org/10.1080/03036758.2018.1477054

Published online:

14 June 2018

Figure 3. Tropiduchidae wing fragment, OU46665. A, Part; B, counterpart; C, interpretational drawing based on part and counterpart. CuA—cubitus anterior; MP—media posterior, MP_1 + 2 and MP_3 + 4—branches of MP; ScP + R—subcosta + radius. Scale bar = 1 mm.

Display full size

Description. A fulgoroid tegmen fragment, 4.1 mm long, 3.2 mm wide, hyaline, without apparent colour markings. The tegmen appears to have been rather short (but macropterous according to the recent definition by Bourgoin et al. 2015). Costal area is likely absent or, if present, very short and developed near the base of the tegmen only. Stem subcosta + radius (ScP + R) may have been short as indicated by the distinct curvature of stem media posterior (MP) towards the costal margin, suggesting the presence of a rather wide basal cell (not preserved). Towards the tegmen base, stem cubitus anterior (CuA) also appears to be slightly curved towards the costal margin. No distinct nodal line is decipherable although a functional nodal line could be traced across the wing. Branching of ScP + R as typical for Fulgoroidea; MP branched early, closer to the branching of ScP + R than to that of CuA, then with regular further branching (although MP₃₊₄ possibly slightly displaced towards the postclaval margin [tornus]). The stigmal area is narrow.

Remarks. The general course of veins in the fossil wing differs from that in the Fulgoroidea families Delphacidae, Achilidae and Derbidae, and the branching pattern of the main stems and lack of transverse veinlets in the stigmal area distinguish it from species of Dictyopharidae. It can also be excluded from Cixiidae (based on lack of a stigma), Flatidae (based on the lack of costal area) and Ricaniidae (Australasian Ricaniidae have fuscous stigmata and costal and distal wing margins). Tegmina of members of the Tropiduchidae vary in length and structure, with a costal area either developed or absent and with or without transverse veinlets. Early branching of MP is a typical character of Tropiduchidae s.str. but with the recent inclusion of taxa from other families (e.g. Fletcher 2001; Gnezdilov 2007) this is no longer a reliable diagnostic character. The fossil bears characters that are present in Tropiduchidae and in some Lophopidae (e.g. absence of costal area and the sequence of branching of ScP + R , MP and CuA); considering the sum of characters present we tentatively assign it to the Tropiduchidae. Within this family, 20 extant and four extinct tribes are currently recognised (Szwedo & Stroiński 2017). The fossil taxon from the Manuherikia Group represents a new, extinct tribe, based on the branching pattern of CuA and the absence of a distinct nodal line, characters that are not found in any other tropiduchid tribe.

Coleoptera

Family Brentidae Billberg, 1820

Genus Perroudia Damoiseau, 1962

Perroudia Damoiseau, 1962, p. 286. Type species Trachelizus victoris Perroud in Perroud and Montrouzier, 1864; the sole extant species, New Caledonia.

Perroudia manuherikia n. sp. Brown and Kaulfuss

Figure 4

First insects from the Manuherikia Group, early Miocene, New Zealand

All authors

Uwe Kaulfuss, Samuel D. J. Brown, Ian M. Henderson, Jacek Szwedo & Daphne E. Lee

https://doi.org/10.1080/03036758.2018.1477054

Published online:

14 June 2018

Figure 4. Perroudia manuherikia n. sp., holotype OU46666. A, Part; B, interpretational drawing. hp—humeral protuberance; sac—subapical constriction. Scale bars = 1 mm.

Display full size

urn:lsid:zoobank.org:act:E4C0F14C-BD7E-4CD3-B36A-BD27C949678C

Type material. Holotype OU46666, part and counterpart of left elytron embedded in claystone; Bannockburn Inlet, Central Otago, New Zealand (−45.0835E, 169.1717N).

Stratigraphy. Dunstan Formation, Manuherikia Group, early Miocene (19–16 Ma).

Description. Left elytron. 2.86 mm long, 0.64 mm wide at anterior margin, 0.75 mm wide in middle, 0.34 mm wide at apex. Subparallel to 3/4 length of elytron, where it rapidly constricts (subapical constriction, Figure 3: sac) before tapering posteriorly in the apical 1/8 to the truncate apex. Apex without posterior processes. Humeral angle with small protuberance. Seven striae visible; strial punctures large, circular, 0.056 mm in diameter, similar in size and shape along length of elytron; interstriae evenly spaced; interstria 3 apparently carinate towards apex. Scutellary striole absent.

Etymology. Named after the Miocene lake system Manuherikia in Otago where the new species was found; to be treated as a noun in apposition.

Remarks. Placement of this specimen into Brentinae is supported by the subapical constriction of the elytron, the truncate apex and the lack of a scutellary striole. Many species placed in Trachelizini have elytra with strongly punctured striae (Zimmerman 1991, 1994; Beaver et al. 2009); however, the tribal classification of Brentinae remains highly artificial and the Trachelizini in particular is probably a polyphyletic assemblage (Sforzi et al. 2014). Assignment of this taxon to Perroudia Damoiseau is based on the strong similarity between the fossil and P. victoris Perroud, 1864, the punctation and spacing of striae, having the third interstria carinate towards the apex and the presence of a humeral protuberance (Figure 5).

First insects from the Manuherikia Group, early Miocene, New Zealand

All authors

Uwe Kaulfuss, Samuel D. J. Brown, Ian M. Henderson, Jacek Szwedo & Daphne E. Lee

https://doi.org/10.1080/03036758.2018.1477054

Published online:

14 June 2018

Figure 5. Perroudia victoris (Perroud) for comparison: female, New Caledonia, 17 km S.W. La Crouen, 15 Mar 1961, J. Sedlacek, determined by G. Kuschel in 2005. A, Dorsal habitus; B, close up of elytra. Scale bars = 1 mm.

Display full size

A number of other beetle taxa with similarly strongly punctured striae were considered as the identity of this specimen, but were rejected for various reasons. Species of Pycnomerus Erichson, 1842 (Zopheridae) have elongate strial punctures, frequently connected by a groove. They also do not have the subapical constriction. The New Zealand weevil Exomesites optimus Broun, 1886 has large, circular punctures anterior of the elytral declivity, as in the fossil, but interstriae 3, 5 and 7 are wider than 2, 4 and 6. Additionally, posterior of the elytral declivity the punctures collapse into a narrow groove. Various genera in the Cossoninae (e.g. Oxydema Wollaston, Eutornopsis Broun, Touropsis Broun) have the striae strongly punctured and the elytra tapering in the apical ¼, but they do not have a pronounced subapical constriction, nor is the apex truncate as in the fossil specimen.

Trichoptera

Family Polycentropodidae Ulmer, 1903

Material examined. OU46667, part and counterpart of a nearly complete forewing (Figure 6).

First insects from the Manuherikia Group, early Miocene, New Zealand

All authors

Uwe Kaulfuss, Samuel D. J. Brown, Ian M. Henderson, Jacek Szwedo & Daphne E. Lee

https://doi.org/10.1080/03036758.2018.1477054

Published online:

14 June 2018

Figure 6. Polycentropodidae wing fragment, OU46667. A, Part; B, interpretational drawing based on part and counterpart. I–V—apical forks 1 to 5; C—costa; Cu_1a–2—branches of the cubitus; DC—discoidal cell; M_1–4—branches of the medius; MC—median cell; R_1–5—branches of the radius; r, s, r-m, m, m-cu—crossveins according to Holzenthal et al. (2007); Sc—subcostal; TC—thyridial cell. Scale bars = 1 mm.

Display full size

Description. Forewing, moderately broad, length 5.0 mm, maximum width 2.4 mm, estimated total length 5.7–6.1 mm (based on comparison with forewings of 11 extant species of Australasian Polycentropodidae illustrated in Neboiss (1986a) and Ward (1995)). Discoidal cell and median cell closed, similar in shape, both rather long and thin. Thyridial cell terminates at the base of the median cell. Apical forks I, II, III, IV and V present; fork V expanded near the fork. Sectorial crossveins s and r transverse, crossveins r-m and m slightly curved.

Remarks. This wing fragment is placed into Polycentropodidae based on the general similarity of the venation to that in extant species (e.g. Neboiss 1986a), specifically the similar, rather long and thin discoidal and medial cells (which are a characteristic in Australian and New Zealand Polycentropodidae) and the expansion of apical fork V near its base. A closed thyridial cell terminating at the origin of the median cell is also present in other families of net-spinning Trichoptera (Philopotamidae, Stenopsychidae, Ecnomidae and some Hydropsychidae), but in these families the discoidal cell is 20%–50% shorter than the median cell (see Neboiss 1986a). A short fork I compared to fork II, and branching well beyond the cross vein closing the discoidal cell, is also present in most Polycentropodidae, but also in Hydropsychidae and some Ecnomidae (but in those families the discoidal cell is short and triangular). The venation and size of the fossil wing most closely resembles Plectrocnemia Stephens, 1836, which is represented by two endemic species in New Zealand (Macfarlane et al. 2010). The range of forewing lengths of P. maclachlani reported by Neboiss (1986a) is 5.5–10 mm. The second New Zealand polycentropodid genus is Polyplectropus Ulmer, 1905 with five endemic, slightly larger species (forewing lengths 7–14 mm). Polyplectropus is also strongly represented in New Caledonia (Johanson & Ward 2009). Australasian Plectrocnemia and Polyplectropus are distinguished by venational characters of the hindwings (Mosely & Kimmins 1953), and the almost identical forewings in both genera (and the Australian Neureclipsis napaea 1986a, 1986b) prevent further identification of this fossil.

Discussion

Fossil evidence for the evolutionary history of insects in New Zealand is still extremely scarce, which makes new finds, such as the fossils reported herein, important for reconstructing the past diversity and diversification of insects on this long-isolated archipelago (Buckley et al. 2015). To date, by far the most diverse insect record comes from early to mid-Miocene strata, with faunas reported from Foulden Maar (Kaulfuss et al. 2015; Kaulfuss & Dlussky 2015; Engel & Kaulfuss 2017), Hindon Maar (Kaulfuss & Moulds 2015; Kaulfuss et al. 2018), amber from three Miocene sites and one late Oligocene site in Otago and Southland (Schmidt et al. 2018) and now from the Manuherikia Group at Bannockburn. Tropiduchidae planthoppers, brentine weevils and polycentropodid caddisflies described here from the latter site are the first fossil insects to be reported from the Manuherikia Group, and the first fossil records of these groups from the Australasian region. These taxa are not recorded in the faunas from Foulden Maar (at least 18 reported insect families), Hindon Maar (> 20 families) and in late Oligocene to Miocene ambers (10 insect families). A more detailed study of insects from these sites may help to elucidate whether this reflects different paleohabitats and/or depositional environments (fluvial large-lake margin habitat at Bannockburn; small, enclosed, forested volcanic lakes at Foulden and Hindon; forest habitats for amber), temporal shifts in community compositions and/or incomplete sampling. Although presently very small, the insect fauna from the Manuherikia Group provides further evidence for post-early Miocene extinction of some insect lineages and, by contrast, a long history of at least 16–19 Ma for other extant insect lineages in New Zealand.

The presence of tropiduchid planthoppers under mesothermal conditions in the early Miocene of New Zealand concurs with the distribution of most extant Tropiduchidae species in tropical and subtropical regions (Bourgoin 2018). This family is well represented in forest habitats (rarely in mangroves or arid areas) in Australia, including several taxa recently transferred from the family Nogodinidae (Fletcher 2001). More than 180 genera and 650 species of Tropiduchidae have been described globally (Szwedo & Stroiński 2017; Bourgoin 2018) but none are represented in the modern New Zealand fauna. Our results demonstrate that this family was present in New Zealand in the Miocene, indicating that Tropiduchidae is another Miocene insect group that has become extinct in New Zealand, probably as a result of post-Miocene climate cooling, as has been documented for ants (Kaulfuss & Dlussky 2015), termites (Engel & Kaulfuss 2017) and tettigarctid cicadas (Kaulfuss & Moulds 2015). The diversity and ecological traits of Australasian Tropiduchidae indicate that complex non-biological and biological events shaped the recent fauna of this region, as exemplified by Montrouzierana (Wang et al. 2014), leading to the presence of a few distinctive taxa in New Caledonia and the absence of this family in New Zealand. The fossil record of Tropiduchidae (Szwedo & Stroiński 2017; Szwedo 2018) dates back to the Turonian; the specimen from the Miocene Manuherikia Group is the first definitive fossil representative for the Australasian region. A tropiduchid-like planthopper nymph recently discovered in Oligocene New Zealand amber (Schmidt et al. 2018) has yet to be described in detail.

The brentine weevil Perroudia manuherikia n. sp. from Bannockburn provides a valuable addition to our knowledge of the evolution of the Brentidae. The Brentidae (sensu Oberprieler 2014) is a family made up of six distinct subfamilies that have been considered as full families by various workers (e.g. Alonso-Zarazaga & Lyal 1999). Recent phylogenetic studies have confirmed that these seemingly disparate taxa do indeed form a monophyletic clade (McKenna et al. 2009; Gunter et al. 2016). The fossil record of the nominate subfamily Brentinae is poorly known. A species has been described from Miocene Dominican amber, Dominibrentus leptus Poinar, 2009, which was placed in the extant tribe Cyphagogini (Poinar 2009). The oldest specimens of the subfamily, which have not been formally named, have been reported from mid-Eocene deposits in Germany as part of the Messel and Eckfeld Maar faunas (Tröster 1993; Wappler 2003). The oldest fossils of the Brentidae as a whole are from the Cretaceous of Orapa, Botswana and Santana, Brazil, which have been assigned to the Eurhynchinae (Kuschel et al. 1994; Zherikin & Gratshev 2004). There are 30 extant genera of Brentinae in Australia (Lawrence & Ślipiński 2013) and six genera and seven species in New Caledonia (Kuschel 2008); in contrast only one genus and species is present in New Zealand, namely the endemic giraffe weevil Lasiorhynchus barbicornis (Fabricius, 1775) (Kuschel 2003). As with the tropiduchid planthopper mentioned above, Perroudia manuherikia n. sp. indicates that some taxa present in the early Miocene fauna are now extinct in New Zealand, suggesting that the current paucity of New Zealand brentine weevils is the result of extinction of a formerly more diverse fauna. The sole extant species of Perroudia, P. victoris is restricted to New Caledonia (Alonso-Zarazaga & Lyal 1999) where it inhabits rainforest and savanna habitats around Canala on the East Coast of Grande Terre.

Polycentropodidae caddisflies and the two genera in New Zealand, Plectrocnemia and Polyplectropus, are cosmopolitan. The Polycentropodidae fossil from Bannockburn indicates that this family has been present in New Zealand since at least the early Miocene. Larvae of the six extant New Zealand species construct loose silk webs for capturing prey, usually between or under stones in flowing water, although some species are also found on exposed shores in large lakes (Timms 1982; James et al. 1998; Winterbourn et al. 2006). Although they could be expected in shore environments of Lake Manuherikia, the presence of an isolated wing of an adult (no larvae have yet been found) in interdistributary bay and swamp sediments suggests transport from an upstream position, which also applies to the other two isolated insect wings from the site. The earliest Polycentropodidae are known from the Barremian of Mongolia (Sukacheva 1993) and fossils have been described from various Cretaceous and Cenozoic New World and Old World localities. However, the fossil from Bannockburn is the first fossil record of this family from the Australasian region.

Lastly, this discovery of fossil insects in the Dunstan Formation at Bannockburn highlights the possibility of further finds from the same strata or from fine-grained sediments elsewhere in the Manuherikia Group in the future.