• Tweet this page
• Like this page
• Email this page
• Share this page

• Home»
• Publications »
• Books & monographs »
• Fauna of New Zealand Series »
• Extracts »
• FNZ45: Nemonychidae, Belidae, Brentidae »
• Appendix 1

FNZ 45 - Appendix 1: Phylogenetic relationships of the genera of Belinae (Kuschel & Leschen)

Kuschel, G 2003. Nemonychidae, Belidae, Brentidae (Insecta: Coleoptera: Curculionoidea). Fauna of New Zealand 45, 100 pages.
( ISSN 0111-5383 (print), ; no. 45. ISBN 0-478-09348-9 (print), ). Published 28 Apr 2003
ZooBank: http://zoobank.org/References/9143DAB5-942F-4E3B-8E68-D47BB28498D5

Appendix 1. Phylogenetic relationships of the genera of Belinae

Guillermo Kuschel
7 Tropicana Drive, Mt Roskill
Auckland 1004, New Zealand
g.kuschel@xtra.co.nz

and

Richard A. B. Leschen
Landcare Research, Private Bag 92170
Auckland, New Zealand
LeschenR@LandcareResearch.co.nz

Introduction

The current classification of Belidae follows that of May (1994), Kuschel (1995), and Lawrence & Newton (1995) where three subfamilies are recognised: Belinae Schoenherr, Oxycoryninae Schoenherr, and Aglycyderinae Wollaston (see also Alonso-Zarazaga & Lyal 1999). We focus our study on the subfamily Belinae because it is the principal group present in New Zealand and Chile and is relatively well known taxonomically.

Belidae is a monophyletic family and its three subfamilies have been definedby Kuschel (1995). Two tribes were recognised in Belinae by Kuschel (1959) -Belini Schoenherr and Pachyurini Kuschel - which were based mainly on mandibular (C9, see list of characters below) and aedeagal characters (C43-45). The tribes were elevated to subfamily rank by Vanin (1976). Zimmerman (1994) added a third tribe, Agnesiotidini Zimmerman, to separate two groups from Pachyurini based onthe presence or absence of sulci on the frons (C6 and C7). This character has not been described with sufficient precision in Zimmerman (1994) as it consists of more than one character state. Moreover, the placement of some genera in these two tribes is rendered rather doubtful, because of the variability found in the sulci by Zimmerman (1994).

There are also several problems with the taxonomic limits of some genera as recognised at present. The South American Atractuchus and Dicordylus seems to be very closely related because they share two unique features, the presence of teeth on the aedeagal pedon and the presence of asubmedian dorsal spot on each elytron, while differing in the shape of the prothorax, a character considered of little phylogenetic significance. Because these two genera share two strong synapomorphies and differ only in a weak character, Atractuchus is here synonymised with Dicordylus (new synonymy), although these genera were placed in different tribes byZimmerman (Table 1). The genus Pachyura was determined to be paraphyleticby Kuschel (1959) who proposed Pachyurinus for the Pachyura species described by Broun (1880, 1893, 1915) from New Zealand. In this treatment, three species of Pachyurinus are placed in the new genus Rhicnobelus. The monotypic Australian genus Pachyura is listed by Zimmerman (1994) in Agnesiotidini, as well as in Pachyurini in the keys because frontal sulci may be present in some specimens and absent in others. The genus Agathinus Broun of New Zealand was synonymised by Zimmerman (1994) withthe Australian genus Cyrotyphus (3 species) though the former differs from Cyrotyphus in a number of characters (C3, 4, 7, 12, 19). The speciesof Agathinus and Cyrotyphus are treated as terminals in the cladistic analysis.

The genus Rhinotia from Australia and Papua New Guinea contains well over 80 species in Zimmerman (1994). It is remarkably diverse and could easily turn out polyphyletic. We include two rather disparate species in the data matrix, among them the type-species R. haemoptera.

Methods

A total of 28 species in 22 genera were examined (Table 1) for the data matrix (Table 2). The sample includes all of the New Zealand and Chilean species. The taxa selected repre-sent the full spectrum of variation of the subfamily of Australasia and South America. Only the following four genera were left out of consideration because no specimens were available for examination:

Habrobelus Zimmerman (Pachyurini, 1 sp., Queensland). Distinguishedby the following jointly occurring characters of an elongated prosternum, long temples, and no concave curvature between head and rostrum. Isacanthodes Zimmerman (Belini, 2 described and 5 undescribed spp., widespread in Australia). This genus has the facies of Isacantha, but differs considerably from all other Belini in femoral, tibial, and elytral features, particularly in having long hind femora that extend beyond ventrite 2.

Pachybelus Zimmerman (Pachyurini, 1 sp., Queensland and New SouthWales). The presence of elytral tubercles and tufts render this genus unique amongst Pachyurini. Rhinotiodes Zimmerman (Belini, 1 sp., Queensland and New SouthWales). Said to be rather similar to Rhinotia haemoptera, but has a very inflated metasternum, an elytral suture that ends in a prominent tooth, and dentate femora.

To root the trees we chose the basal chrysomeloid Palophagus (Megalopodidae) because this genus has the ancestral features considered basal for Curculionoidea (Kuschel and May 1990, 1996a,b). A total of 48 characters from adults of both sexes were scored for a cladistic analysis.

The data were coded and entered into MacClade version 3 (Maddison andMaddison 1992) and analysed using PAUP* Version 4.0b8 for Macintosh (Swofford2001). The settings used in PAUP* for heuristic tree searches included a 100 random addition sequence with stepwise addition (holding 25 trees at each step) set with steepest descent; character states were treated as unordered; TBR branch swapping on all trees; and zero-length branches were collapsed. Polarity of character states was determined after the trees were produced. To further resolve relationships and find consistency for character congruence in these trees, successive approximations character weighting (SAW; Farris 1969;Carpenter 1988, 1994), an a posteriori weighting method, was performed on the resulting trees (weights were set at 10 and based on the consistency index with analyses truncated). Ambiguous characters were plotted onto trees using standard ACCTRAN and DELTRAN optimisations (Maddison et al. 1984) and in the discussion below, characters that reverse to other states are indicated by the superscript Cr.

Table 1. Terminal taxa selected for phylogenetic analysis of Belinae (Belidae). Placement according to Alonso-Zarazaga & Lyal(1999).

Agnesiotidini

Agathinus tridens (Fabricius)

Agathobelus bivittatus Zimmerman

Agnesiotis pilosula Pascoe

Apagobelus brevirostris (Lea)

Basiliobelus flavovittatus Zimmerman

Cyrotyphus blandus (Faust)

Cyrotyphus fascicularis Pascoe

Cyrotyphus variegatus Lea

Dicordylus binotatus (Philippi)

Dicordylus marmoratus (Philippi)

Macrobelus insignis Lea

Belini

Araiobelus filum (Jekel)

Homalocerus lyciformis (Germar)

Isacantha dermestiventris (Boisduval)

Rhinotia bidentata (Donovan)

Rhinotia haemoptera Kirby

Stenobelus tibialis (Blackburn)

Trichophthalmus miltomerus (Blanchard)

Pachyurini

Arhinobelus agathophagus Zimmerman

Atractuchus annulifer (Philippi)

Callirhynchinus exquisitus (Fairmaire & Germain)

Hadrobelus undulatus (Zimmerman)

Pachyura australis Hope

Pachyurinus sticticus (Broun)

Rhicnobelus aenescens (Broun)

Rhicnobelus metallicus (Broun)

Rhicnobelus rubicundus (Broun)

Sphinctobelus cinereus (Blanchard)

Palophaginae (Megalopodidae)

Palophagus bunyae Kuschel

Character list and argumentation

1. Shape of temples in dorsal view. Converging caudad (0);parallel-sided or broadening caudad (1). This character is to be observed withthe head in normal position, not extruded.
2. Length of temples (t) to length of eyes (e). t = or < e (0); t> e (1). A usually unequivocal character, but is marginally distinct for Cyrotyphus and the Atractuchus/Dicordylus complex.
3. Breadth of frons in relation to apex of rostrum. Equal or narrower(0); wider (1). The frons is the whole area between the eyes when viewed fromabove. The breadth of the frons is the minimum distance between the eyes. Variations may be encountered, particularly in Cyrotyphus and the Atractuchus/Dicordylus complex, rendering a safe scoring somewhat difficult.
4. Surface of frons in lateral view. Flat or convex (0); depressed(1). A depression at the front and base of rostrum is ignored for coding this character. It is primitive if the frons above the middle is level or convex in relation to the upper margin of the eyes.
5. Frontal carina. Absent (0); present (1). This character refers only to the presence of elongate, high carinae; consequently a small, inconspicuous carina sometimes present is not considered. This feature is exclusive of Hadrobelus and Pachyura.
6. Sides of frons near eyes. Lacking depressions (0); with superficialor well defined grooves (1). A superfical groove, longitudinal depression, or deep groove may be present on many beline taxa.
7. Form of groove of frons near eyes. Shallow (0); deep (1);inapplicable for taxa coded as 6-0. A shallow superficial groove is a longitudinal, rather broad depression without sharp sides and with microsculpture at the bottom; a deeply engraved or incised groove is a narrow sulcus with sharply defined sides and a smooth, impunctate bottom. Shallow impressions in some species are often covered by setae or scales, and the latter state is a rather distinctive feature given prominent status by Zimmerman (1994)when proposing his tribe Agnesiotidini.
8. Apex of epistome. Truncate or bilobed (0); bidentate (1). There often is a slight emargination or fovea on either side of a median emarginationor excision of the epistome that induces a pointed or blunt tooth. The difference between the two states is obvious except in Sphinctobelus which has a minute, punctiform fovea not producing a distinct bidentate epistome.
9. Apical tooth of mandibles in dorsal view. Exposed in front ofmedian tooth (0); hidden, tucked under the median tooth (1). Closed mandibles ofBelini and Pachyurini are noticeably different if viewed from the side, the front and above. Some practice with viewing this character is required to assess it safely.
10. Labial palp. 3-segmented (0); 1-segmented (1). It is advisable to dissect the mouthparts to count the palpal segments with confidence. The derived state is found only in the South American Homalocerus and Trichophthalmus.
11. Ommatidia. Coarse (0); fine (1). The varying coarseness of eyes is the result of the time of prevalent activity; fine eyes for diurnal species and coarse ones for nocturnal species. Careful observations are required to distinguish character states. If, for instance, the eyes of Rhicnobelus and Rhinotia are checked against those of Agathinus and Cyrotyphus, the difference is immediately apparent. If, on the other hand, the eyes of Pachyura are examined, it may not be easy to decide its state. Colour appears to also vary with development of the ommatidia. In general, if the integument of the species has a metallic sheen and its tergitesare heavily pigmented, that would indicate a diurnal habit.
12. Protrusion of eyes. Moderately convex (0); hemispherical (1). Theeyes of all Belinae protrude to the sides, but the convexity varies. This character is of limited value; the extreme, spherical state is found only in the New Zealand Agathinus and the Chilean Dicordylus binotatus.
13. Insertion of antennae. Antemedian or median (0); postmedian (1).The measurements are done from the suprainsertional knob to the epistomal apex and from the middle of the knob to the anterior margin of the eyes.
14. Antennomere reaching eyes. 1 or 2 (0); 3 (1) These states areeasily determined if antennae are folded back and/or the first two segments,excluding the condyle of the scape, are measured to verify whether their combined length would reach the eyes. The first segment reaches the eyes in Homalocerus and Trichophthalmus. The character is of limited value in certain taxa, because the two segments pass the front margin in Atractuchus, only slightly in D. marmoratus, and not at all in D. binotata.
15. Length of antennal segment 3 in relation to 1+2 combined. Equal or shorter (0); longer (1). The 3rd antennal segment is distinctly longer than thefirst two combined in Agnesiotis and Atractuchus, and is only marginally longer in Dicordylus.
16. Pronotal impressions on either side of median line. Absent (0);present (1). Impressions on the pronotum are present in some Agnesiotidini and Pachyurini.
17. Groove at midline of pronotum. Absent (0); present (1). A median groove is usually obvious as a sulcus or a median scaly stripe; it is rather weak in Rhicnobelus and Sphinctobelus.
18. Sides of prothorax at apex in dorsal view. Not or hardly constricted (0); strongly constricted to form a collar (1). If the constriction is distinct enough, it shows as a collar on the sides and sometimes also on the dorsal surface.
19. Widest breadth of prothorax. Near middle (0); at basal third (1); at basal angle (2). The lateral outline of the prothorax can vary individually a good deal in some species. The derived states are sometimes difficult to separate if the widths of the sides and hind angles are about the same.
20. Surface of scutellum. Flat (0); impressed (1); convex (2).
21. Anterior declivity of scutellum. Sloping (0); vertical (1); proclinate (2). A sloping condition is always evident, but the prothorax andelytra must be slightly apart to see whether the anterior declivity of the scutellum is vertical or overhanging, i.e., inclined forward or proclinate.
22. Base of elytra and prothorax in dorsal view. Elytra with weak orno lobe, prothorax subtruncate (0); elytra with large lobe, prothorax strongly bisinuous (1). The shape of the base of the prothorax is a direct response tothe degree of protrusion and overhang of the elytral lobes, the two features being directly interdependent.
23. Base of elytra in profile relative to pronotum. Level or lower(0); higher (1). This character may define two distinct groups in Pachyurini. Elytra that are present at a lower level than the pronotum are found in a number of Belini.
24. Elytral declivity at apex in profile. Pronounced (0); absent orgentle (1). State 1 is characteristic for most Rhinotia species, their elytral apex with some frequency somewhat upturned. Macrobelus is the sole pachyurine genus approaching this feature which otherwise is common inBelini.
25. Transverse rugae on elytra. Absent (0); present (1). Anautapomorphy for Rhicnobelus.
26. Tubercles and/or tufts on elytra. Absent (0); present (1). State 1 is present only in Agathinus, Cyrotyphus, and D. binotatus.
27. Integumental sheen. Non-metallic (0); metallic (1). Among thes elected terminal taxa, Callirhynchinus, Macrobelus, and Rhicnobelus have a metallic, usually bronze sheen.
28. Shape of radial cell window of hind wing. Trigonal (0); pentagonalor tetragonal (1). For trigonal state see Fig. 46-49, 54-56, for the other state see Fig. 50-53.
29. Extension of Cu in relation to hind margin of wing. Extending to the hind margin or very nearly so (0); ending some distance from it (1). Samples of an abbreviated Cu are shown in Fig. 50-53 and 56, a feature of Agnesiotodini and the South American Belini.
30. Crossvein a2-a3. Present (0); absent (1). Individual variations inthe wing venation occur and discrepancies even in the two wings of the same specimen may be noticed on occasions. The cross vein a2-a3 is usually present,though the length may vary (see next character) and it may be absent where veins a2 and a3 are fused at their bases (Fig. 49).
31. Crossvein a2-a3. Present as a relatively long vein (0);rudimentary (1); inapplicable for taxa coded as 30-1. Samples of a long a2-a3 are shown in Fig. 46-48, 59-53. No figure is provided for across vein of state (1), but it is considered rudimentary if the crossvein isonly about as long as wide.
32. Shared track of A1+A2 apicad of a2-a3 or junction of the shared trackwith A3. Long (0); short (1). This character separates Pachyurini andBelini. Vanin's (1976) Figure 23 for Trichophthalmus, showing a long shared track, is at odds with our observations and Fig. 56.
33. Vein A1. Complete (0); interrupted or absent (1). This vein is consistently interrupted in Agnesiotidini (Fig. 50-53) and two species of NewZealand Pachyurini (Fig. 49) and absent only among the taxa surveyed in Rhicnobelus metallicus.
34. Direction and distance of a3-a4 in relation to cu-a. Oblique and distant (0); transverse and close (1). There is no figure here provided to illustrate this character. State 1 is peculiar to Apagobelus and Basiliobelus.
35. Position of procoxae. Slightly postmedian (0); slightly antemedian(1). The position of the procoxae relative to the front and hind margin of the prothorax distinguishes the two tribes Belini and Pachyurini.
36. Sides of pro- and mesocoxae. Partially or entirely open (0); completely closed (1). The coxal cavities are completely closed if the lobes that enclose the coxae show not even an indentation at the suture. In state 0 the lobes leave at least a triangular gap right next to the coxae. Closed coxal cavities may constitute a strong apomorphy for Agnesiotidini.
37. Length of hind femora in relation to hind margin of ventrite 2. Passing it (0); at most reaching it (1). This character may be useful for separating the tribes Belini and Pachyurini.
38. Crenulated ridge on laterodorsal surface of tibiae. Absent (0);present (1). Tibiae must be seen at the right light and turn to be certain of the presence or absence of the ridge. A crenulated ridge is present in all taxa of Belini and Agnesiotidini.
39. Tibial formula of spurs. 2-2-2 (0); 1-2-2 (1). Spurs are always distinct and hardly ever different in size on the middle and hind tibiae. If the protibiae have two spurs, these may be small and of uneven length, and could be missing from a tibia because of breakage.
40. Denticulation on lower edge of tibiae. Absent (0); present (1).Presence of denticulation on the protibiae is observed in all Belini and in overhalf of the terminal taxa of Pachyurini selected, but the denticulation is sometimes rather fine (e.g., R. metallicus).
41. Mucro on fore and/or middle tibia in male. Absent (0); present(1). Among the taxa considered, a mucro is absent only in Agathinus, Cyrotyphus, and Pachyurinus.
42. Spines on lower edge on one or more femora. Absent (0); one spine present (1); two or more spines present (2). Having femoral teeth is a feature of the Agnesiotidini (except in odd males of Atractuchus annulifer argus) and the vast majority of Belini.
43. Pedon and tectum of aedeagus. Free throughout (0); at least partially fused to a tubular structure (1). This character invariably separates the two recognised tribes.
44. Basal and guide sclerites in internal sac. Absent (0); present (1). A basal sclerite is absent from Macrobelus, Pachyura, and Sphinctobelus, the remainder of Pachyurini have a rigid basal sclerite, apart from other small sclerites.
45. Form of basal and guide sclerites in internal sac. Basal sclerite present as a rigid 'flagellum' (0); basal and a long sclerite with a snare-like structure when viewed in profile (1). Belini have a very strange sclerite illustrated by Zimmerman's (1994) Figure 216a with a conspicuous 'snare'. The South American genera Homalocerus and Trichophthalmus have a weakly sclerotised guide-sclerite much like that of Stenobelus illustrated in Figure 235 of Zimmerman (1994).
46. Length of parameral sector relative to apodeme. Shorter or equal (0); distinctly longer (1). A long parameral sector is a salient feature for Agnesiotidini.
47. Length of spermathecal duct relative to hemisternites. As long as or longer than hemisternites (0); very short, sessile, or subsessile (1). All Belinae examined have a thick and short spermathecal duct, but some among these may be as long as or even a little longer than the hemisternites.
48. Median or postmedian spots of elytra. Absent (0); present (0). Rhinotia haemoptera has spots, but these are more laterally placed on the elytra and are considered not homologous to those present in Atractuchus annulifer, Dicordylus binotatus, D. marmoratus, and Rhinotia bidentata.

Results

Support for the tribes and generic groups we discuss here are provided in Table 3. An initial analysis resulted in 309 equally parsimonious trees (Treelength (TL) = 125; Consistency Index (CI) = 0.40; Retention Index (RI) = 0.77), and a consensus of these trees is shown in Fig. 188. In this tree only Belini (Bin Fig. 188) are monophyletic, as well as a group that includes Agathinus, Agnesiotis, Atractuchus, Cyrotyphus, and Dicordylus (we call this the Agathinus-group, A in Fig. 188). Successive weighting of these trees resulted in 12 trees by 2 iterations (the consenus of these is shown in Fig. 189). In this tree the Agathinus-group ismonophyletic and is sister group to the remaining Belinae with a monophyletictribe Belini.

The relatively poor resolution in the initial parsimonious trees suggests that some of the characters we recognise may be either too variable, intrinsically difficult to score due to ambiguous character states, ormorphological features that are simply not useful for recognising monophyletic groups. We decided to a priori weight characters 9, 23, 28, 30, 36, 43,and 46 because these were recognised as critical characters in the classification of the group. We applied a base weight of 10 to the aforementioned characters and 8 trees were produced in this analysis (TL = 191;CI = 0.59; RI = 0.90), a consensus of these trees is shown in Fig. 190. In these trees, Belini and the Agathinus-group are monophyletic, but in this case Belini is sister taxa to the remaining members of Belinae wich includes amonophyletic Agathinus-group. SAW analysis of these trees resulted in the same consensus tree (of 6 total trees).

Since adult characters have not been completely useful to resolve the beline tree, we decided to determine if larval data could provide useful phylogeneticsignal to resolve relationships in Belinae, and we scored 6 larval characters to this matrix based on the larval work of May (1993, 1994). The larvae that have been described thus far are Agathinus, Cyrotyphus blandus, Hadrobelus, Rhicnobelus rubicundus, Pachyurinus sticticus, Sphinctobelus niger Zimmerman, Rhinotia bidentata, and R.haemoptera as well as the outgroup taxon Palophagus (Kuschel &May 1990, 1996). Note that a broad selection of larval Agnesiotidini is not available and only represent taxa from the Agathinus-group. The characters are as follows (data matrix in Table 4):

49. Mandible. Bidentate (0); tridentate (1); adentate (2). The number of teeth present on the mandible varies from one to two, and in Rhinotia,the mandibles lack teeth altogether (May 1994).
50. Tormae. Present (0); absent (1).
51. Frontal horn. Absent (0); present (1).
52. Ocelli. Six (0); one (1); two (2).
53. Occipital foramen. Closed (0); open (1).
54. Median suture. Present (0); absent (1).
55. Form of spiracles. Annular (0); bicameral (1).
56. Placement of anus. Terminal or subdorsal (0); ventral orsubterminal (1).

Before combining larval data with the original data matrix we ran the larval data separately with the terminals represented in the data matrix (Table 4) which resulted in 12 trees (TL = 17, CI = 0.59, RI = 0.63). In these trees (a consenus is shown in Fig. 191) Belini is monophyletic and an SAW analysis resulting in 9 trees (by two iterations, consensus tree shown in Fig. 192) resulted in a monophyletic Belini which is sister group to the remaining taxa which includes a monophyletic Pachyurini. Using these exemplar taxa, we added the adult characters which resulted in 1 tree (TL = 83, CI excludinguninformative characters = 0.63, RI = 0.6824) shown in Fig. 193. This tree shows a monophyletic Agnesiotidini which is sister taxon to the remaining taxa which includes a monophyletic Belini.

A completely combined analysis (all taxa and characters) resulted in 536 trees (TL = 144, CI = 0.42, RI = 0.76) with the same relationships shown in Fig.188 (the SAW analysis resulted in the same trees as shown in the consensus tree in Fig. 189). A priori weighting of characters 9, 23, 28, 30, 36, 43, and 46 resulted in 168 trees (TL = 209, CI = 0.59, RI = 0.89) that are similar to those seen with in Fig. 190 but with poorer resolution (consenus tree in Fig. 194). A consenus of 12 trees found by SAW is shown in Fig. 195.

Agnesiotidini

To this group we add Atractuchus from Pachyurini and formally remove Agathobelus, Apagobelus, Basiliobelus, and Macrobelus, which are transferred to Pachyurini. The Agnesiotidini ismonophyletic based on the following characters (reference tree in Fig. 189 andunique characters indicated by an *): shape of temples in dorsal view parallel-sided or broadening caudad (1-1, present also in Belini), sides offrons near eyes with superficial or well defined grooves (6-1), third antennomere reaching eyes (14-1r, reverses in Atractuchus and Dicordylus marmoratus), surface of scutellum convex (20-2*), base of elytra in profile relative to pronotum higher (23-1*), shape of radial cellwindow of hind wing pentagonal or tetragonal (28-1*), extension of Cu in relation to hind margin of wing ending some distance from it (29-1, presentoutside this group in Trichophthalmus), sides of pro- and mesocoxae completely closed (36-1*), crenulated ridge on laterodorsal surface of tibiae present (38-1, present outside this group in all Belini), one spine present on lower edge on one or more femora (42-1, present outside this group in some Belini), length of parameral sector relative to apodeme distinctly longer (46-1*). The following characters are concordant with these, including larval characters which are coded as missing for most terminals: sides of prothorax atapex in dorsal view strongly constricted to form a collar (18-1, equivocalancestral reconstruction), median suture absent (54-1), and spiracles bicameral (55-1). If Agnesiotidini are a derived group in Pachyurini (Fig. 190), the following characters support its monophyly: apex of epistome truncate or bilobed (8-0), ommatidia coarse (11-0 r, reverses in Agnesiotis), (20-2*), (22-0*), (23-1*), (28-1*), (29-1), crossvein a2-a3 present as a relatively long vein (31-0, DELTRAN), vein A1 interrupted or absent (33-1), (36-1*), (38-1), (42-1), and (46-1*).

Dicordylus marmoratus is more closely related to Atractuchus annulifer than it is to D. binotatus, according to characters 4-0, 12-0, 19-2, and 26-0, which would suggest that these two species could be merged under one genus or that an additional genus would have to be proposed for D.marmoratus. We adhere to the synonymy of the two genera as stated previously as their differences are too slight to warrant independent status, althoug happearing in the trees as separate taxa. Though Agathinus groups assister taxon to Cyrotyphus in one case it should not be included in this genus (see Remarks in taxonomic section. p. 28).

Belini

Belini is a monophyletic group based on (reference tree in Fig. 190): shape of temples in dorsal view parallel-sided or broadening caudad (1-1, present alsoin Agnesiotidini), shared track of A1+A2 apicad of a2-a3 or junction of the shared track with A3 short (32-1*), position of procoxae slightly antemedian (35-1*), length of hind femora in relation to hind margin of ventrite 2 not or just reaching it (37-1*), crenulated ridge on laterodorsal surface of tibiae present (38-1, present outside this group in all Agnesiotidini), aedeagus with pedon and tectum at least partially fused to a tubular structure (43-1*), insertion of antennae postmedian (13-1, equivocal ancestral reconstruction), and ventral or subterminal placement of anus (56-1, larval character that is coded as missing for most terminals) are concordant with these. If Belini is derived within Pachyurini (Fig. 189), then the following characters support its monophyly: (1-1), apical tooth of mandibles in dorsal view exposed in front of median tooth (9-0*), elytral declivity at apex in profile absent or gentle (24-1r, reverses in Homalocerus, Isacantha, and Trichophthalmus), (32-1*), (35-1*), (37-1*), (38-1), (43-1*), mandible adentate (49-2, unique and probably specialised in Rhinotia), tormae absent (50-1, unique and probably specialised in Rhinotia), occipital foramen closed (53-0, shared with Agnesiotidini).

Pachyurini

As stated above to this group we add Agathobelus, Apagobelus, Basiliobelus, and Macrobelus from Agnesiotidini, and formally transfer Atractuchus to Agnesiotidini. By fixing Pachyurini as amonophyletic group in MacClade, this group is supported by the absence of a crenulated ridge on laterodorsal surface of tibiae (38-0, primitive) and two larval characters confined to this group: mandible tridentate (49-1) and occipital foramen open (53-1). The genera Rhicnobelus, Pachyurinus, and Pachyura are separate taxa and support the generic separation of these taxa. Although, Pachyurinus is consistently placed a sister taxon to Rhicnobelus and it would seem that the later name is unnecessary.   

Discussion

Based on the analyses it is clear that some genera must be shifted around between the paraphyletic tribes Agnesiotidini (sensu lato) and Pachyurini, while the placement of the monophyletic tribes Agnesiotidini (sensu stricto) andBelini is uncertain as to whether they are basal or included within a paraphyletic Pachyurini. While there are several excellent adult characters that define the monophyly of Agnesiotidini (sensu stricto) and Belini, there are two larval characters that may actually be evidence for the monophyly of Pachyurini. Unfortunately larvae are known for only a few terminals, so additional larval material and a detailed study of chaetotaxy and other characters would be useful towards understanding the monophyly of the groups.

We propose to maintain three tribes, whilst recognising the fact that there is little evidence for maintaining these groups because of problems with paraphyly. However, Agnesiotidini (sensu stricto) and Belini are demonstrably monophyletic, though their placements vary among the trees, and Pachyurini is clearly a paraphyletic group recognised only for convenience.

Purchase this publication

• Purchasing details