Polyozus Eyles and Schuh, 2003

WEIRAUCH, CHRISTIANE, 2007, Revision and Cladistic Analysis of the Polyozus Group of Australian Phylini (Heteroptera: Miridae: Phylinae), American Museum Novitates 3590, pp. 1-64 : 1-6

Polyozus Eyles and Schuh, 2003 was described as a monotypic genus from New Zealand with the type species Polyozus galbanus Eyles and Schuh, 2003 . Sorting and dissecting large numbers of Phylinae from

Australia, which were mainly collected over the past 10 years by Randall T. Schuh and Gerasimos Cassis, but also material on loan from several institutions revealed that the genus Polyozus is by no means restricted to New Zealand, but has at least seven undescribed species in Australia. Furthermore,

1 Division of Invertebrate Zoology, American Museum of Natural History ([e-mail]).

Carvalho’s description and genitalic illustrations of Orthotylus australianus ( Carvalho, 1965) indicate that this species, which was placed in the Orthotylinae , is in fact another species of the genus Polyozus . Apart from the nine species of Polyozus , an additional seven species, which have several characters in common with species of the genus Polyozus , are treated in this revision and cladistic analysis of the group.

The so-called Polyozus group appears to be distinct among Australian Phylini . A cladistic analysis shows that it is well supported by several synapomorphies, among them a short labium, the combination of simple/suberect and flattened/subadpressed setation, the Jshaped vesica with a process arising close to the secondary gonopore, the dorsal apical process bent in an right angle, which renders the apex of the vesica virtually anchor-shaped, and characters of the female genitalia. The genus Polyozus is supported as a monophyletic group. The seven species of the ingroup, which are not part of Polyozus , fall into two distinct clades and are here placed in the genera Ancoraphylus , n.gen. (four species) and Exocarpocoris , n.gen. (three species).

Host records for the 16 species treated in this paper are given, optimized on the cladistic hypothesis, and discussed. The distributions of the plant bug species are documented and discussed.

METHODS

About 1370 specimens were examined. Matrix code labels were attached to all specimens, which uniquely identify them, and are thus called unique specimen identifiers (USIs). The USI numbers (e.g., AMNH_PBI 00094810), which comprise an institution and a project code (AMNH_PBI) and a unique number (00094810), are provided for all specimens.

Habitus photographs (taken on a Microptics-USA photographic system equipped with an Infinity Photo-Optical K-2 lens system) and illustrations of male genitalia are provided for all taxa. Drawings of female genitalia, including a dorsal view of the bursa copulatrix and associated structures, as well as a ventral view of the posterior wall, are provided for one representative of each genus. Scanning electron micrographs (SEMs) are given for at least one representative of each genus, documenting the habitus in lateral view, vestiture, pretarsi, and the evaporatory areas associated with the metathoracic scent gland. Additional SEMs for selected representatives show details of the male genitalia, including the pygophore, paramere, and vesica.

Unless otherwise stated, the right paramere is drawn in dorsolateral view, the left paramere is drawn in anterolateral view, the lateral right side of the phallotheca is shown, and the vesica is drawn in lateral view from the left side. See table 1 for measurements.

Specimens were deposited in and/or borrowed from the following institutions. Abbreviations for institutions follow the suggestions by Arnett et al. (1993). The institutional abbreviations listed are used throughout this paper:

AM Australian Museum, Sydney

AMNH American Museum of Natural History , New York

CNC Canadian National Collection of Insects, Agriculture and Agri-Food Canada, Ottawa

MAGD Museum of the Northern Territory, Darwin

MNRJ Museu Nacional, Rio de Janeiro

SAMA South Australian Museum , Adelaide

USNM [ United States] National Museum of Natural History , Smithsonian Institution, Washington, DC

WAMP Western Australian Museum, Perth

ZISP Russian Academy of Sciences , Zoological Institute, St. Petersburg

Host plants collected during the R.T. Schuh and Gerasimos Cassis collecting trips were deposited at the National Herbarium of New South Wales (Sydney) or the Western Australian Herbarium (Perth). The numbers provided with the plant identifications are vouch- er numbers of these institutions.

The distribution of host plants mentioned in the discussions and host plant sections are derived from the website of the Australia’s Virtual Herbarium (AVH) (http://www.cpbr. gov.au/avh/).

The distribution maps for Phylinae are drawn in a Mercator projection. The Dis-

TABLE 1 Measurements of Species

TABLE 1 (Continued) TABLE 1 (Continued)

cover Life homepage (http://www.discoverlife. org/) gives online maps of the species as well as images of plant bugs and host plants.

CLADISTIC ANALYSIS: Nineteen taxa comprising 16 ingroup species and 3 outgroup species were analyzed using 43 morphological characters. The characters are listed in table 2. Most characters are illustrated and character and character states pointed out in the illustrations (e.g., 2-0 in fig. 3E refers to character 2 and state 0 in table 2). The matrix is provided as table 3. Sixteen characters were coded as multistate, eight of which were coded as nonadditive (11, 17, 22, 24, 29, 36, 37, 41). The ingroup species comprise the three species of Exocarpocoris , four species of Ancoraphylus , and 9 species of Polyozus , all of which are revised herein. The three outgroup species of Phylini used, which are deposited at the AMNH, are an unidentified species of Campylomma from Australia, Opuna annulatus (Knight, 1935) , with an Indopacific distribution, and an undescribed Australian species close to or congeneric with Xiphoides Eyles and Schuh, 2003 (referred to as Xiphoides for the purposes of this project).

The heuristic search under equal weights was run through WinClada in NONA (hold10000; hold/100; mult*100;) ( fig. 18 ). The relative degree of support for each node was calculated with NONA using up to 10 extra steps (h10000; bsupport10;) and is shown on figure 18 above the branches in square boxes. Pee-Wee was used for analysis with implied weighting (hold10000; hold/100; mult*100; jump50;), with a k value of N 5 3 ( fig. 19 ). The character distributions shown in figures 18 and 19 are unambiguous optimizations (homoplasy setting: any additional step treated as homoplastic), but the consequences of fast and slow character optimization as implemented in WinClada are discussed in the character discussion. Numbers (in gray circles) were assigned to nodes and clades and are referred to in the discussion.

To gain insights into the evolution of insect–host plant relationships in this group of Phylini , the distributions of host plants were analyzed on the cladogram. Following the approach of Carpenter (1989), the host plant genus was coded as a character state, deactivated during analysis, and then optimized on the unweighted cladogram.

THE POLYOZUS GROUP OF AUSTRALIAN PHYLINI

The Polyozus group, as here defined, is characterized by several features of the male and female genitalia, but also by the vestiture with a mixture of simple, suberect and subadpressed, flattened setae ( fig. 4A–F ). Among the characteristic features of the male genitalia are the anchor-shaped (or almost so) vesica of the male, with the dorsal apical process usually projecting at a right angle from the body of the vesica, and a process projecting from the vesica that arises from the area of the secondary gonopore ( figs. 12 , 13 ). Among the diagnostic features of the female genitalia are areas beset with microtrichia in the posterior region of the dorsal labiate plate and the ornamentation of the posterior wall, which comprises paired areas of microtrichia and paired processes on the posterior margin ( fig. 14 ).

KEY TO GENERA OF THE POLYOZUS GROUP

1.

Figures

Fig. 3. Habitus in lateral view and evaporatorium of A, B, Ancoraphylus arctous (AMNH_PBI 00087680); C, D, Exocarpocoris tantulus (AMNH_PBI 00135796); E, F, Polyozus kojonup (AMNH_PBI 00136470).

Fig. 18. Phylogenetic relationships of the 16 species of the Polyozus group: The strict consensus of the eight equally most parsimonious trees (L 5 91; CI 5 67; RI 5 83) derived from the character matrix in table 3. Solid circles on the branches indicate uniquely derived apomorphic character states; empty circles indicate homoplastic characters under unambiguous character optimization. Figures above the bars indicate number of character as shown in table 2 and appendix 1; figures below bars indicate the character state. Large figures in shaded circles denominate clades. Branch support values (Bremer support) are given above branches in square boxes.

Fig. 19. Strict consensus of the two most parsimonious trees obtained with the analysis using implied weights. Characters and clades are shown as explained in figure 18.

Fig. 4. Setae comprising hemelytral vestiture. A, Ancoraphylus arctous (AMNH_PBI 00087680); B, Exocarpocoris tantulus (AMNH_PBI 00135796); C, Polyozus bulita (AMNH_PBI 00168436); D, P. furcilla (AMNH_PBI 00099437); E, P. galbanus (AMNH_PBI 00139062); F, P. kojonup (AMNH_PBI 00136470). Scale bar: 50 mm.

Fig. 12. Vesica of species of Ancoraphylus, Exocarpocoris, and Polyozus.

Fig. 13. Vesica of species of Polyozus.

Fig. 14. Female genitalic structures in Ancoraphylus, Exocarpocoris, and Polyozus, comprising bursa copulatrix in dorsal view and posterior wall in ventral view.