FNZ 68 - Simuliidae (Insecta: Diptera) - Cladistic analysis based on morphology
Craig DA, Craig REG, Crosby TK 2012. Simuliidae (Insecta: Diptera). Fauna of New Zealand 68, 336 pages.
(
ISSN 0111-5383 (print),
ISSN 1179-7193 (online)
;
no.
68.
ISBN 978-0-478-34734-0 (print),
ISBN 978-0-478-34735-7 (online)
).
Published 29 June 2012
ZooBank: http://zoobank.org/References/9C478D54-FEB2-45E8-B61C-A3A06D4EB45D
Cladistic analysis based on morphology
Outgroup
Morphologically Paraustrosimulium anthracinum (Bigot) is considered closely related to Austrosimulium (e.g., Wygodzinsky & Coscarón 1962) and while Dumbleton (1973) assigned it subgeneric status in Austrosimulium, others since have not. Moulton (1997, 2000, 2003) using molecular analysis, showed that P. anthracinum was sister to the then Austrosimulium colboi Davies & Györkös (Davies & Györkös 1988). Examination of modern material of A. colboi (Moulton 2000, DAC pers. obs.), in particular the male genitalia, indicates that it is not Austrosimulium (by possessing a well developed paramere, which is absent in Austrosimulium) and should perhaps be considered as the Australian representative of Paraustrosimulium. Indeed, Davies & Györkös (1988) noted that their placement was provisional. However, redescription and taxonomic transfer has not yet been done, so we refer to this entity as ?Austrosimulium colboi. Moulton (2000) noted that placement of other Australian simuliids not in Austrosimulium or Simulium was problematic. These are the “Cnephia” of various authors. He makes the point that even placement in Cnephia is inappropriate. This segregate of some 10 species was assigned to Paracnephia by Crosskey & Howard (1997), but not to any subgenus. Hence, Cnephia pilfreyi Davies & Györkös is referred to here as ?Paracnephia pilfreyi. Moulton (2003) showed ?P. pilfreyi is sister to P. anthracinum + ?A. colboi. Placement of the above taxa in relation to the subgenera Novaustrosimulium and Austrosimulium was a number of nodes removed in Moulton’s (2003) phylogeny.
Wygodzinsky & Coscarón (1973) when redescribing Cnesiamima atroparva (Edwards) noted that it and P. anthracinum form a genus pair. In a preliminary morphological cladistic analysis of the phylogeny of Simuliidae of the Southern Hemisphere, Gil-Azevedo & Maia-Herzog (2007), while not including ?A. colboi in their analysis, show Cnesiamima to be sister to Paraustrosimulium + Austrosimulium. So, we also included Cnesiamima in our outgroup. Therefore, the outgroup that we use here for cladistic analysis of the subgenus Austrosimulium is comprised of Paraustrosimulium anthracinum, ?Austrosimulium colboi, ?Paracnephia pilfreyi, and Cnesiamima atroparva.
Character scoring of these outgroup taxa is uneven. A full series of stages was available for P. anthracinum and ?A. colboi courtesy of S. Coscarón and J. K. Moulton, respectively. Of note is that the type material of ?A. colboi and ?P. pilfreyi was listed as missing by Bugledich (1999), but has, in part, been recovered (Craig 2011). The other outgroup species were scored using literature (Davies & Györkös 1988, Coscarón & Coscarón-Arias 2007); information for some characters is lacking.
Ingroup
For New Zealand Austrosimulium we consider 18 species. We exclude A. vailavoense because only females are known. Relationships of A. vailavoense are discussed under the species section and below under Molecular Analysis. Inclusion of A. dumbletoni is arguable since characters for the male were taken from gynandromorphs (Craig & Crosby 2008) and earlier stages are yet unknown. For the Australian members of the subgenus Austrosimulium we have chosen exemplars (Yeates 1995) from the species-groups. Choice was dictated by availability of material and level of description (Mackerras & Mackerras 1949 et seq.). Material available originally was of poor quality and badly bleached, so scoring of characters not provided in the literature was, at times, questionable. The Australian taxa used included A. (A.) cornutum Tonnoir (ungulatum species-group) and A. (A.) montanum Mackerras & Mackerras (mirabile species-group). New material, however, collected in 2011 allowed full characterisation of A. montanum. Similar problems bedeviled the Australian Novaustrosimulium where again we used a single exemplar from each of the two species-groups designated by Dumbleton (1973), namely A. (N.) furiosum (Skuse) of the furiosum species-group and A. (N.) pestilens Mackerras & Mackerras from the bancrofti species-group.
Given the taxon gap for the Australian material –– 4 species used versus 12 known (Adler & Crosskey 2012), this cladistic analysis should be considered preliminary.
Analysis
The data matrix in Table 3 was entered into MacClade 4.0™ with character states coded as “0”, “1”, “2”, as required (Maddison & Maddison 2001), and analysed using PAUP* 4.0b8 (Swofford 1998). All characters were equally weighted and unordered. A heuristic search was executed using the default PAUP settings. That resulted in 10 680 Equally Parsimonious Trees (EPT). We comment on the resultant Strict Consensus tree (Fig. 505). The successive weighting method (Carpenter 1988) was applied to the EPTs using the Rescaled Consistency Index (RCI) at a base weight of 10. This procedure was done once, when the number of EPTs stabilised at 815. We comment on the 50% Majority Rule Consensus from that analysis (Fig. 506). That tree was fully resolved in MacClade™ using the Resolve Polytomies function and is shown in Fig. 507 with an overlay of distribution and general ecological requirements of larvae.
Another heuristic analysis was done using the Goloboff fit at k=2. This is another method for weighting characters (Goloboff 1993). The numbers of EPTs were reduced to 816 and the Strict Consensus tree produced was the same as the 50% Majority Rule Tree above. Reweighting the characters made no difference.
Characters
Craig & Currie (1999) used 37 characters for cladistic analysis of Polynesian Inseliellum. That was expanded to 45 by Craig et al. (2001). Gil-Azevedo & Maia-Herzog (2007) used 119 for their analysis of Southern Hemisphere simuliids. While we use some of their characters, we restrict ours to 38 (see below & Table 3). Characters are fairly equally distributed among adults (characters 1–13), pupae (characters 14–26), and larvae (characters 27–38). They are listed below with the same numbers as those in the data matrix (Table 3). For each character we show the Consistency Index (CI); a measure of how well the character fits on a phylogenetic tree. Simply, if not completely compatible with the tree a character’s CI will be less than unity.
Austrosimulium Character States
Adult structures
1. Number of antennal flagellomeres (CI 1.0)
0. Nine
1. Eight
Strictly, the basal 2 divisions (scape and pedicel) of the antenna are segments, containing musculature. The remaining divisions lack muscles and should be referred to as articles, or flagellomeres (Fig. 18). Character 1 of Gil-Azevedo & Maia-Herzog (2007: 42), character 5 of Py-Daniel (1982: 298). Nine flagellomeres is considered plesiomorphic for Simuliidae; lower numbers are derived. Seven in A. (N.) bancrofti is autapomorphic for Austrosimulium.
2. Vein R1 (CI 1.0)
0. Spiniform setae and hairs
1. Spiniform setae absent, hairs only (Fig. 3)
Character 25 of Gil-Azevedo & Maia-Herzog (2007: 47), 10 of Py-Daniel (1989: 298).
3. Female mandible (CI 1.0)
0. Teeth present on both inner and outer edges at apex
1. Teeth only on inner edge at apex (Fig. 17)
Teeth on both sides of the mandible is plesiomorphic for simuliids. Teeth only on the inner edge of the apex is considered derived. Character 7 of Gil-Azevedo & Maia-Herzog (2007: 43), 18 of Py-Daniel (1982: 298), and 38 of Pinto-Sanchéz et al. (2005: 21). State (1) is synapomorphic for Austrosimulium. Of interest here is that female gynandromorphs of A. dumbletoni show a reversion to state (0) (Craig & Crosby 2008).
4. Female hind basitarsus (CI 0.25)
0. Row of stout setae present (Fig. 44)
1. Row of stout setae absent (Fig. 48)
A row of stout setae parallel to the comb of the hind basitarsus of females is widespread in basal simuliids; poorly formed in Parasimulium, but distinct elsewhere. Stout setae are absent in the ungulatum species-group of Austrosimulium and considered a synapomorphy for that group. Absence in A. australense is considered autapomorphic.
5. Male hind basitarsus (CI 0.17)
0. Row of stout setae present (Fig. 43)
1. Row of stout setae absent (Fig. 49)
State (0) is plesiomorphic for simuliids; state (1) is derived and occurs in some species-groups in Austrosimulium.
6. Female tarsal claw (CI 0.25)
0. Basal tooth, medium–large (claw distinctly bifid) (Fig. 62)
1. Basal tooth, absent–small (less than 1/6 length of claw) (Fig. 59)
A claw tooth is considered a plesiomorphic condition in simuliids (Currie & Grimaldi 2000; Adler et al. 2004) and has been clearly associated with bird biting. Character 17 of Gil-Azevedo & Maia-Herzog (2007: 44), 13 of Coscarón & Miranda-Esquival (1998:163), 142 of Adler et al. (2004: 152), and 48 of Pinto-Sanchéz et al. (2005: 22). Absence and reduction within Austrosimulium is considered derived.
7. Female genital fork (CI 1.0)
0. Anterior arm narrow
1. Anterior arm broad (Fig. 108)
Character 34 of Gil-Azevedo & Maia-Herzog (2007: 48) who considered state (0) plesiomorphic. State (1) is synapomorphic to Paraustrosimulium + Austrosimulium.
8. Female genital fork apodeme (CI 1.0)
0. Lateral plate apodeme not markedly developed
1. Apodeme developed (Fig. 108)
This apodeme is situated at the junction of the lateral plate and lateral arm of the genital fork and is not homologous to character 35 of Gil-Azevedo & Maia-Herzog (2007: 48) and 145 of Adler et al. (2004: 152) — vis-à-vis the extensive apodeme in Gigantodax Enderlein + Pedrowygomyia Coscarón & Miranda-Esquivel that arises from the lateral arm and extends anteriorly. Character 137 of Adler et al. (2004: 151) who are of the opinion that an apodeme is synapomorphic for Simuliini.
9. Female genital fork lateral lobe (CI 1.0)
0. Large (Fig. 110)
1. Small (Fig. 108)
State (1) is synapomorphic for the australense-subgroup in New Zealand.
10. Male gonostylus shape (CI 0.25)
0. Quadratic to subtriangular–broad (Fig. 126–128)
1. Narrower and tapered (Fig. 136–142)
The outgroup and all the australense species-group possess state (0), with the exception of A. (A.) albovelatum. State (1) is synapomorphic for the ungulatum species-group in New Zealand and for one of the Australian exemplars.
11. Gonostylus terminal spine number (CI 0.25)
0. Two (Fig. 130)
1. Three (Fig. 126)
2. Four
Modified character 48 of Gil-Azevedo & Maia-Herzog (2007: 50), who considered the plesiomorphic condition for Simuliini to be two spines; more spines is considered derived.
12. Paramere development (CI 0.5)
0. Present and well developed (plate-like)
1. Only basal connection to gonocoxa present (Fig. 128, 140)
2. Poorly developed (sometimes present as thin twisted rod) (Fig. 131)
Character 52 of Gil-Azevedo & Maia-Herzog (2007: 50), 26 of Py-Daniel (1982: 299). In part character 136 of Adler et al. (2004:151). Present in basal simuliids, and markedly variable in development elsewhere. Here, the poor development in Austrosimulium is considered derived.
13. Paramere spines (CI 0.33)
0. Present
1. Absent
Character 25 of Currie & Grimaldi (2000: 479) who considered presence of spines as synapomorphic for Simuliini. Adler et al. (2004), however, considered lack of spines as plesiomorphic in Simuliini and their presence derived. Character 54 of Gil-Azevedo & Maia-Herzog (2007: 50), 182 of Adler et al. (2004: 155). Absence in Austrosimulium can be considered derived.
Pupal structures
14. Cocoon fabric (CI 0.40)
0. Fibrous (Fig. 178)
1. Alveolate (Fig. 187)
2. Alveolate and gelatinous (Fig. 191)
A fibrous (0) cocoon is plesiomorphic for simuliids and occurs here in the outgroup. The alveolate (1) and gelatinous (2) states are here considered derived. State (1) is particularly developed in A. alveolatum and A. laticorne.
15. Anterior projections on cocoon (CI 0.5)
0. Absent (Fig. 160)
1. One (Fig. 177)
2. Two (Fig. 175)
A simple anterior opening with or without a definite rim and no projections appears plesiomorphic for simuliids. Anterior projections are, here, considered derived.
16. Cephalic plate depression (CI 1.0)
0. Not concave dorsally (Fig. 202)
1. Concave dorsally (Fig. 205)
A non-depressed cephalic plate appears plesiomorphic in simuliids. The marked concavity in the tillyardianum species-group of New Zealand Austrosimulium is considered synapomorphic.
17. Cephalic plate shape in female (CI 0.5)
0. Width of frons base distinctly more than 1/2 the height of plate (Fig. 212–216)
1. Width of frons base subequal to 1/2, or less than 1/2, height of plate (Fig. 202–215)
State (0) is considered plesiomorphic. All members of the australense species-group possess state (1), those of the ungulatum species-group state (0).
18. Cephalic plate facial setae (CI 0.5)
0. Present (Fig. 214)
1. Absent (Fig. 205)
State (0) appears plesiomorphic in simuliids. State (1) is synapomorphic to the australense and tillyardianum species-group (with the exception of A. albovelatum).
19. Ocular setae (CI 1.0)
0. Absent (Fig. 203)
1. Hair-like (Fig. 204)
2. Spine-like (Fig. 214–216)
These are sensilla as shown by the dendritic sheath that is sometimes apparent (Fig. 196). State (2) is synapomorphic for the ungulatum species-group. State (1) in A. albovelatum of the tillyardianum-subgroup (otherwise all with state (0)) is something of a conundrum, similar to other character states of that species which tend to emulate those of the ungulatum species-group.
20. Thoracic cuticle (CI 0.67)
0. Tuberculate (Fig. 240, 253)
1. Nontuberculate (Fig. 238, 239)
State (0) is widespread in the taxa under consideration, although in some it is poorly developed. Hence tuberculation is considered plesiomorphic. State (1) in the australense-subgroup appears to be synapomorphic. Within the tillyardianum-subgroup, state (1) in A. fiordense is autapomorphic.
21. Tubercle pattern (CI 0.67)
0. Tubercles randomly arranged (Fig. 252)
1. Tubercles grouped in patterns (Fig. 240)
2. Tubercles absent
State (0) appears plesiomorphic in the taxa considered. State (1) is synapomorphic for the tillyardianum-subgroup.
22. Thoracic dorsal sensilla (CI 0.50)
0. Posterior sensilla pair stiff, spine-like (Fig. 250)
1. All sensilla fine and trichoid (Fig. 240)
Often referred to as “trichomes”, the subdorsal sensilla on the scutum are state (0) in the outgroup and one clade of New Zealand Austrosimulium. The trichoid state (1) appears to be the derived condition, and such sensilla are often broken off.
23. Gill base (CI 0.33)
0. Horn-like (Fig. 270–273)
1. Not horn-like (Fig. 280–282)
State (1) is synapomorphic for the ungulatum-subgroup and autapomorphic for A. longicorne in the australense-subgroup. Scoring of this character is probably suboptimal.
24. Gill filament surface (CI 0.6)
0. Annulated (Fig. 292–293)
1. Reticulated (Fig. 297)
2. Pseudoannulated (Fig. 296)
3. Nonfilamentous gill
State (0) is plesiomorphic for the taxa considered. State (1) is autapomorphic for A. tillyardianum. State (2) is synapomorphic for the ungulatum-subgroup and autapomorphic for A. longicorne.
25. Spine combs on posterior abdomen (CI 0.6)
0. Present
1. Absent (Fig. 194)
State (1) is synapomorphic for Austrosimulium.
26. Grapnel hooks on posterior abdomen (CI 1.0)
0. Present (Fig. 195)
1. Absent
State (1) is synapomorphic for the tillyardianum-subgroup.
Larval structures
27. Antenna apical article (CI 1.0)
0. Distinctly longer than median plus basal article (Fig. 367–383)
1. Equal in length to median plus basal article
Modified character 99 of Gil-Azevedo & Maia-Herzog (2007: 56). State (1) appears synapomorphic for Novaustrosimulium.
28. Mandible sensillum and serrations (CI 1.0)
0. Single serration and sensillum
1. Serration and sensillum complex (Fig. 421)
Basically character 101 of Gil-Azevedo & Maia-Herzog (2007: 56). State (1) is synapomorphic for Paraustrosimulium + Austrosimulium.
29. Mandible spinous teeth number (CI 1.0)
0. Three or four
1. More than four (Fig. 421)
State (1) is synapomorphic for Austrosimulium.
30. Shape of pharate pupal gill (CI 0.5)
0. Not L-shaped (Fig. 351–360)
1. L-shaped (Fig. 361–366)
State (1) is synapomorphic for the New Zealand ungulatum species-group within subgenus Austrosimulium.
31. Rectal scales (CI 0.0)
0. Present
1. Absent (Fig. 437–442)|
State (1) is synapomorphic for Austrosimulium.
32. Anal sclerite central region (CI 0.33)
0. Not robustly developed (Fig. 441)
1. Robustly developed (Fig. 437–440)
State (0) appears plesiomorphic for simuliids. New Zealand Austrosimulium with one exception (A. dugdalei) possess state (1). In the subgenus Inseliellum, in Polynesia, a robust central region appears correlated with the presence of a semicircular sclerite (Craig & Currie 1999: Figure 5).
33. Anal sclerite arm lengths (CI 0.33)
0. Ventral arm distinctly longer than dorsal arm (Fig. 437, 442)
1. Dorsal and ventral arms subequal in length
Character 114 of Gil-Azevedo & Maia-Herzog (2007: 59). State (0) is synapomorphic to Paraustrosimulium + Austrosimulium (with exceptions in Australia).
34. Anal sclerite anterior arms expression (CI 0.5)
0. Not flared or emarginated
1. Flared and emarginated (Fig. 447)
With the exception of ?A. colboi, state (0) occurs in the outgroup.State (1) occurs in ?A. colboi + Austrosimulium. Flared anterior arms, which are muscle apodemes, appear correlated to presence of a semicircular sclerite.
35. Anal sclerite interarm strut (CI 1.0)
0. Absent
1. Present (Fig. 347-440)
Character 117 Gil-Azevedo & Maia-Herzog (2007: 59), 149 of Adler et al. (2004: 153). State (1) is synapomorphic for Paraustrosimulium + Austrosimulium.
36. Accessory sclerite (CI 0.5)
0. Absent (Fig. 451–453)
1. Poorly developed (Fig. 448–450)
2. Well developed (Fig. 437–442)
State (0) occurs in the unicorne-subgroup, state (1) in the ungulatum-subgroup, and state (2) in the australense species-group.
37. Semicircular sclerite (CI 1.0)
0. Absent
1. Present (Fig. 437–442)
The semicircular sclerite considered here is not continuous with the ventral arms of the anal sclerite. Modified character 118 of Gil-Azevedo & Maia-Herzog (2007: 59). Character 148 of Adler et al. (2004: 53) who make the strong point that this sclerite is not homologous to the condition seen in Gigantodax. It is similarly homoplasious in Simulium (Gomphostilbia) palauense Stone (Takaoka & Craig 1999), S. (Inseliellum) cataractarum Craig (Craig 1987), S. (I.) concludium Craig (Craig 1997), and Crozetia spp. (Craig et al. 2003). State (1) occurs in subgenus Austrosimulium, and state (0) in Novaustrosimulium. Of note is that in subgenus Austrosimulium the semicircular sclerite is continuous around the circlet of hooks, but in A. cornutum (Australian ungulatum species-group) it is not continued mid-ventrally. See Terms section (p. 42) for more explanation.
38. Number of hooks in anal proleg circlet (CI 0.33)
0. More than 1 000 and up to 1 800
1. More than 2 000
2. Less than 999
Palmer & Craig (2000) showed a strong correlation between the number of the hooks in the posterior circlet and the velocity of water experienced by larvae of particular species. This shows particularly in larvae of A. longicorne which inhabit flows of markedly low velocity and possess the smallest number of hooks.
Results and discussion
Because of the considerable taxon data gap for Australian Austrosimulium species, in particular that of subgenus Novaustrosimulium, results from this morphological cladistic analyses should be considered preliminary even though there is especially strong support (Fig. 506) for the backbone of the phylogeny.
Consensus Tree
Analysis of the data matrix (Table 3) produced 10 680 equally parsimonious trees (TL 99, CI 0.5). The strict consensus tree is shown in Fig. 505. The outgroup is fully resolved with Cnesiamima atroparva + ?Paracnephia pilfreyi and Paraustrosimulium anthracinum + ?Austrosimulium colboi as sister clades; not at variance to relationships suggested by Moulton (2003) and Gil-Azevedo & Maia-Herzog (2007). Within Austrosimulium the exemplars of Novaustrosimulium (Australian only), A. (N.) furiosum (furiosum species-group) and A. (N.) pestilens (bancrofti species-group) are in polytomy with subgenus Austrosimulium (Australia and New Zealand). This is perhaps an indication that Novaustrosimulium needs taxonomic revision. Within the australense species-group, Dumbleton’s australense- and tillyardianum-subgroups resolve, but, the terminal taxa in the latter subgroup do so poorly. For the ungulatum species-group a clade of the two Australian exemplars, A. (A.) cornutum (ungulatum species-group) + A. (A.) montanum (mirabile species-group) is unresolved with respect to the remainder of the ungulatum species-group. The unicorne-subgroup is a terminally unresolved clade. The relationship of the Australian exemplars is slightly at variance to that proposed by Dumbleton, where he placed the mirabile species-group sister to the ungulatum species-group (Australia + New Zealand). Within the New Zealand ungulatum species-group, the constituents of the ungulatum-subgroup are in a polytomy with the unicorne-subgroup. The overall strong concordance of the backbone of this tree to that of Dumbleton’s species-groupings (his Fig. 252) was expected since he used synapomorphies for his relationships within Austrosimulium.
50% Majority Rule Tree
A single reweighting of the original 10 680 trees, using the rescaled consistency index at a base weight of 10, resulted in 815 trees, with no further reduction. The 50% Majority Rule tree of those trees is shown in Fig 506. Extremely high to full support shows for the backbone of the tree, and more of the terminal taxa were fully resolved. We show, simply, just branch support with no other statistics, since it is known (Craig & Currie 1999, Craig et al. 2001) that such basic analyses produce robust trees for Simuliidae.
For the outgroup taxa, Paraustrosimulium anthracinum + ?Austrosimulium colboi are sister to Cnesiamima atroparva + ?Paracnephia pilfreyi, with full support for both clades. There is no disagreement with the relationships suggested by Moulton (2003) and Gil-Azevedo & Maia-Herzog (2007). For Austrosimulium, the Novaustrosimulium exemplars, A. furiosum and A. pestilens are now resolved as sister taxa (Fig. 505).
There is full support, again, for Dumbleton’s australense and ungulatum species-groups. In the former, the A. australense + A. longicorne clade that comprises the australense-subgroup is fully supported. Similarly supported is the tillyardianum-subgroup, but A. albovelatum, A. extendorum, A. stewartense, A. laticorne are unresolved in relation to the clades ((A. alveolatum + A. tillyardianum) and (A. dugdalei +A. multicorne)) with A. fiordense sister. The latter two clades are only moderately well supported. Austrosimulium fiordense is sister because of the unique lack of tuberculation on the pupal thoracic cuticle. That A. dugdalei and A. multicorne are sister species was expected.
In the ungulatum species-group the two Australian exemplars are again sister to the remainder, and both clades have full support. For the New Zealand species the two currently recognised subgroups, ungulatum- and unicorne- of Dumbleton, resolve as sister clades; the former with little support, the latter with full support. Austrosimulium unicorne (unicorne-subgroup) is sister to A. bicorne + A. tonnoiri, the latter clade with poor support. The ungulatum-subgroup is poorly supported and largely unresolved, with A. dumbletoni sister to the other three species, and that clade too has poor support.
Shortest Possible Tree
Resolving all terminal taxa using the Resolve Polytomies function in MacClade, resulted in the tree shown in Fig. 507. We have drafted the figure to be in concordance with that of Dumbleton’s Figure 252 and have overlain geographic and ecological data.
There is little difference from the Majority Rule Tree (Fig. 506). The ungulatum species-group resolves as for Dumbleton, with A. unicorne sister to A. bicorne + A. tonnoiri, and that clade is strikingly similar to that derived from molecular data (Fig. 514). Within the tillyardianum-subgroup of the australense species-group, A. fiordense is again sister to all the remaining species. The previously fully supported, but terminally partially unresolved clade is now resolved with A. tillyardianum + A. alveolatum as sister taxa with the sister clade of Austrosimulium dugdalei +A. multicorne, the latter expected. Sister to those is A. albovelatum. The latter’s placement was expected as A. albovelatum possesses a series of unusual character states reminiscent of the ungulatum-subgroup. Austrosimulium extendorum + A. stewartense are sister to A. laticorne.
The geographic overlay shows that both lineages of the outgroup have sister taxa in Australia and South America, and indicates a Gondwanan provenance for that outgroup clade. That the subgenus Novaustrosimulium does not resolve into a monophyletic clade is problematic and indicates that the subgenus needs taxonomic revision. Austrosimulium (A.) cornutum + A. (A.) montanum (Australia) remain resolved as sister species, but taxonomically they are in different species-groups. Placement as sister to the New Zealand clade of the ungulatum species-group suggests dispersal of that clade from Australia, independent of the New Zealand australense species-group. The majority of Austrosimulium species are restricted to the South Island, Stewart Island, plus the subantarctic islands. The North Island has only one endemic species, A. dugdalei, which is sister to A. multicorne (see species descriptions). Otherwise, both A. australense and A. tillyardianum are widely spread in the North Island, with the former to the northern tip at Cape Reinga.
Of importance phylogenetically is that the South American outgroup species (Cnesiamima atroparva, Paraustrosimulium anthracinum) have cool to very cold water requirements for the immature stages and occur at higher altitudes, similar to immatures of the unicorne-subgroup in New Zealand. Both the Australian representatives of the outgroup (?Paracnephia pilfreyi, ?Austrosimulium colboi), while tending to have less cold requirements, are found in cool temperate conditions and at moderate altitudes (Davies & Györkös 1988). We are of the opinion that this is indicative of a basic ecological template for this segregate of simuliids that includes Austrosimulium. Also, the normal substrate for the outgroup species is vegetation of various sorts, as it is for some members of the ungulatum group. That preference also occurs in the australense species-group, namely for A. australense. The typical habitat of the tillyardianum-subgroup is, in general, hard substrate in small to open sunlit streams, with a wide range of altitudes. The phylogeny indicates this is a derived preference.
The ecological requirements of the immature stages of New Zealand simuliid species are for the most part quite distinct. For this reason Dumbleton’s original ecological keys for Austrosimulium spp., and which we have reprised here, work quite well. Therefore it behooves aquatic ecologists to identify to species any simuliid involved in their work – something that currently is generally not done.
Of the Australian representatives involved here, A. pestilens occurs mainly in Queensland and also farther west of other species and can be considered a warm temperature species. Austrosimulium furiosum is widely distributed from Queensland south into Tasmania. However, apart from the northern localities, the majority are at higher altitudes and cooler temperatures, particularly in Tasmania. Similarly, A. montanum and A. cornutum occur at higher altitudes along the eastern coast, with the latter also being found in Tasmania.