Siricidae (Hymenoptera: Symphyta: Siricoidea) of the Western Hemisphere

CJAI 21, July, 2012
doi: 10.3752/cjai.2012.21

Nathan M. Schiff, Henri Goulet, David R. Smith, Caroline Boudreault, A. Dan Wilson, and Brian E. Scheffler

A phylogenetic reconstruction of extant taxa at generic level is proposed (Fig. C1.14). Principles and methods of cladistic analysis and phylogenetic reconstruction are based mainly on Hennig (1966). For each lineage, an indented list of characters is given. For each character, the derived state is given first, followed, in brackets, by the ancestral state and its distribution within Siricidae and out–groups representing earlier lineages of Symphyta.

The Siricoidea (including only Siricidae) is an isolated lineage in Hymenoptera. It is defined by many unique derived character states. The general relationships among the basal Hymenoptera lineages are best illustrated in Vilhelmsen (2000, 2001). The old concept of Siricoidea is a grade and is paraphyletic. It includes the Anaxyelidae, Siricidae and Xiphydriidae. All members of these three families have a similar sword-like ovipositor in adult females, and their larvae have a rod-like structure on the abdominal apex and develop in wood. Detail studies of external and internal structures redefine the Siricoidea (Vilhelmsen 2001, Schulmeister 2003) and restrict it to one family, the Siricidae. The lineages (out–groups) that precede the Siricoidea starting with the earliest are the Xyeloidea giving rise to the Tenthredinoidea, and then to the Pamphilioidea, followed by the Cephoidea, the Anaxyeloidea and finally the Siricoidea. The Siricoidea are followed by Xiphydrioidea, and all remaining Hymenoptera. Because Anaxyelidae, Siricidae and Xiphydriidae do not share a common ancestor and are part of a transformation series between Xyelidae and the all remaining Hymenoptera, each family was given superfamily status equal to the preceding and following lineages.

1        The Siricidae form a monophyletic lineage, united by the nine following derived character states:

• Pronotum collar-like, the anterior surface vertical and slightly concave with projecting anterolateral angles, and the dorsal and lateral surfaces flat (Fig. C1.3). The unique pronotal shape is probably an adaptation to protect the head during adult emergence from a tree trunk. [In other Hymenoptera the pronotum is not collar-like.]
• Tergum 10 in female and sternum 9 in male extending posteriorly as a horn ending in a small tube-like rod (Figs. A3.9 & A3.10). This structure is probably involved in helping the adult emerge from a tree trunk. [This structure is unique to the Siricidae.]
• Maxillary palp with one palpomere. [Most Hymenoptera have 2–6 palpomeres.]
• Postgenae broadly meeting medially ventral to the foramen magnum (Fig. C1.1) [In other Symphyta the postgenae do not meet medially ventral to foramen magnum.]
• Tergum 9 of female with median basin (Figs. B1.29 & B1.30). [This structure does not occur in other Hymenoptera.]
• Mesoscutum with posterior portion divided on each side by a broad furrow between base of fore wing and scutellum (Fig. C1.4). [This is probably the precursor of the transscutal fissure clearly outlined in Xiphydriidae and many Apocrita. These furrows are absent in earlier lineages.]
• Fore wing with basal portion of vein 2A clearly outlined (Fig. C1.9). [Though considered as part of the ground plan of Hymenoptera, this vein is absent in later lineages of Symphyta and, in our opinion, is a derived character state, not an atavism.]
• Vein 3R1 with apical portion thinly sclerotized and nebulous (Fig. C1.9). [If the cell is close, it is clearly tubular in earlier lineages of Symphyta.]
• Metatibiae and metatarsomeres 1 or 1 and 2 enlarged and wide in male (Fig. A3.2). [In almost all Symphyta, the metatibia and metatarsus are not particularly enlarged and are quite similar to those of the middle leg, but are usually relatively longer.]

The ten extant genera of the Siricidae are included in our phylogenetic reconstruction. The pivotal character is based on the junction of the fore wing vein Rs. This vein originates from vein 1r–rs and its proximal section ends normally at or near the junction of veins Rs and M. In earlier lineages of Symphyta and in one fossil siricid, vein Rs joins M more distally either at vein 1m–cu or on Rs+M.

1a      Sirex, Sirotremex, Urocerus and Xoanon form a monophyletic group, united by the following derived character state:

• Fore wing vein Rs (originating from vein 1r–rs) ends proximally at or very near the junction of veins Rs and M (Fig. A3.30). [In earlier lineages of Symphyta, vein Rs joins M more distally, either at vein 1m–cu or on Rs+M roughly between the junction of vein Rs and M and vein 1m–cu.]

1aa    Sirex and Sirotemex form a monophyletic group, united by the following derived character state:

• Fore wing vein Cu1 present in almost all specimens (Fig. A3.30). [Vein Cu1 is seen in the distantly related Pamphilioidea and is absent in other, earlier lineages of Siricidae. In Siricidae, we do not think that this vein is an atavism, but rather a new vein. In Siricidae veins are commonly added or disappear; we suspect that this vein was an addition.]

1aaa  Sirex forms a monophyletic group, defined by the following derived character states:

• Ovipositor with ventral edge of the 3–5 apical pits before the teeth annuli clearly ridge-like (Fig. C1.15). [Such a structure is not seen in other Siricidae or other Symphyta. In Xeris such a ridge is present only on the last annulus before the teeth annuli, and the pit forms a long trough extending to the dorsal edge of the lancet.]
• Dark areas of body with blue or green metallic reflections (one species of the distantly related Eriotremex and one of Afrotremex have such metallic reflections). [Metallic reflections are uncommon in Siricidae and generally in Symphyta. We assumed that this state is derived.]

1aab  Sirotremex forms a monophyletic group, defined by the following derived character states:

• Fore wing broadly rounded at apex (Fig. C1.16). [The fore wing apex is angularly rounded in most Symphyta.]
• Cell 1Rs2 very short (2r–m and 3r–m each slightly longer than veins Rs2 or M) (Fig. C1.16). [This cell is usually long in other Siricidae and most Symphyta.]
• Cell 3R1 very short (2.2 times as wide as long) (Fig. C1.16). [This cell is longer in other species of Siricidae, though it is still long relative to most Symphyta except the Cephoidea and Anaxyeloidea. However, the cell in general outline is typical of Siricidae.]
• Middle flagellomeres short: about 1.5 times as long as high in lateral view (Fig. C1.17). [The flagellomeres are rather short in Sirotremex compared to those of early lineages within Siricidae. One expects such short flagellomeres in genera associated with deciduous trees (except larch - a conifer). However, the flagellomeres are round in cross section as in genera of Siricidae associated with coniferous trees rather than dorsoventrally flattened, as in Tremex.]
• Head in dorsal view across eyes wider or very slightly narrower than across the genae (behind eyes) (Fig. C1.18). [A common feature of the Siricidae is the marked head enlargement behind the eyes in dorsal view. In a few other species - Xeris tarsalis (Cresson) and males of Teredon cubensis (Cresson), the head is not wider across the eyes than the width across the genae.]

Note. The female is unknown. Its discovery would probably provide significant additional characters to clarify the relationships of Sirotremex with Sirex.

1ab    Urocerus and Xoanon form a monophyletic group, united by the following derived character states:

• Sheath with teeth dorsally in apical third of apical section with one bristle near apex of each tooth (Fig. C1.19). (teeth absent in a few species of Urocerus from China) [In other siricid genera when the teeth are present each tooth has a very fine seta at its base.]
• Male with sterna clearly folded and smooth medially (Fig. C1.20). [Character state scarcely visible (median surface smooth) in other Siricidae and absent in other Symphyta.]
• Ovipositor with pits clearly present in apical 0.25 but still small on last annulus before teeth annuli (Fig. C1.21). [In Siricidae, the lancet pits are large toward apex and are outlined on many or all annuli before the teeth annuli.]

1aba  Urocerus forms a monophyletic group, defined by the following derived character states:

• Gena without or almost without pits in central area (Fig. A3.5). [The gena is moderately to densely pitted centrally in other species of Siricidae except in a few species of Xeris.]
• Female with median basin about twice as wide as long, with long and divergent ridge-like edges (Fig. C1.22). [Except in Sericosoma, the median basin is not so wide in other genera of Siricidae and, except in Xeris, the lateral edges are not so divergent in other genera.]

1abb  Xoanon forms a monophyletic group, defined by the following derived character states:

• Gena very densely pitted (Fig. C1.23). [The widespread state in Siricidae is less densely pitted at center of gena.]
• Terga without pitted microsculpture even mediobasally. [In other Siricidae the pitted microsculpture is at least present medially on terga 2–8.]
• Fore wing vein 2r–m displaced apically and joined to cell 3M (Fig. C1.24). [In all other extant Siricidae and almost all Symphyta, fore wing vein 2r–m intercepts cell 2M.]
• Male with terga 1–8 densely pubescent. [In other Siricidae at least terga 2–4 are mostly bare.]

1b      Xeris, Sericosoma, Teredon, Eriotremex, Afrotremex and Tremex form a monophyletic group, united by the following derived character states:

• Metatibia with one spur (Fig. C1.25). [In earlier lineages of Symphyta, the metatibia has two spurs.]
• Hind wing without anal cell (a single male of Teredon cubensis with cell – it may be an aberration) (Fig. C1.26). [In most Symphyta and earlier lineages of Siricidae, the anal cell is present.]
• Fore wing with vein 1 cu–a joined to vein Cu clearly close to or at junction of veins M and M+Cu (except in Teredon) (Fig. C1.27). [In earlier lineages of Symphyta, vein 1cu–a intercepts cell 1M near its middle.]

           
2a      Xeris is the earliest lineage of clade 1b and forms a monophyletic group, defined by the following derived character states:

• Gena behind eye with a short vertical ridge (Fig. C1.28). [This is probably not the occipital ridge seen in most Symphyta, as it is found only behind each eye and there is no trough in front of it. Typically, the occipital ridge when present is either completely developed or present only near the mandible.]
• Female with anterior edge of median basin clearly narrower than maximum width of basin (Fig. C1.29). [In other Siricidae, the base is relatively wider.]
• Sheath with apical 0.3 of apical section without teeth dorsally (Fig. C1.30). [In Siricidae except Teredon and some species of Urocerus from China the teeth are present. The presence of teeth is considered an ancestral state within the Siricidae.]

Note: Xeris has been associated traditionally with Urocerus (Gauld and Mound 1982, Benson 1943) because of the very long and medially constricted cornus, and the pale spot on the gena. In Urocerus, there are species with the cornus shorter and not constricted, similar to that of Sirex longicauda or Sericosoma tremecoides. Even in X. tarsalis, the first lineage of Xeris the cornus is not constricted. Therefore, we think that the long cornus could have been evolved independently in Xeris because it does not have the more basal constriction found in Urocerus and Xoanon).

2b    Sericosoma, Teredon, Eriotremex, Afrotremex and Tremex form a monophyletic group, united by the following derived character states:

• Minimum distance between eyes 0.7–1.2 times maximum eye height (Fig. C1.31, black arrows). [In most Symphyta the eye height is clearly less than the minimum distance between eyes.]
• Distance between inner edges of antennae much greater (4.0–10.0) than distance from outer edge of antennal socket to nearest edge of eye (Fig. C1.31, white arrow). [In almost all Symphyta, the antennal sockets are closer together.]
• Eye long and narrow (1.7-1.9 times as long as high) (Fig. C1.44). [In most Symphyta and early lineages of Siricidae (except Xoanon), the eye is moderately narrow (1.2–1.7 times as high as long).]
• Flagellomeres flattened dorsoventrally. [In Symphyta, the flagellomeres are either round or flattened laterally.]

3a    Sericosoma consists of one species and forms a monophyletic group, defined by the following derived character states:

• Fore wing vein M between veins M+Cu and Rs+M strongly and abruptly curved (Fig. C1.32). [In Symphyta and other Siricidae, vein M is straight, or uniformly and widely curved.]
• Female with cercus long and wide at base of cornus (Fig. C1.33). [In Symphyta, the cercus is typically long and narrow; in Siricidae it is very short, thumb- or button-like, articulated or not, or even absent.]
• Tergum 9 with setae in median basin on cone-shape tubercles (Fig. C1.34). [In almost all Symphyta and other Siricidae, each seta in medial area is in a pit.]
• Tergum 9 with a wide shallow furrow separating the smooth dorsal surface from the pitted ventrolateral surface (Fig. C1.34). [In other Siricidae, the shallow furrow is absent.]

Note. Sericosoma (from Southeast Asia) and to lesser extent Teredon (from Cuba) have character states seen in early lineages of Siricidae. Adults of Sericosoma share 10 character states traditionally associated with the Siricinae as defined by Benson (1943). They are:

• Labial palp with three palpomeres.
• Both sexes with antenna as long as or longer than fore wing costal cell.
• Distance between lateral ocelli subequal to distance between lateral ocellus and nearest edge of eye.
• Gena near lower edge of eye with posterior edge of pits not elevated as a coarse scale or a low tooth.
• Metatarsomere 1 in dorsal view not very compressed laterally and lateral surfaces not twisted.
• Subcosta clearly outlined in basal 0.4 of costal cell.
• Middle flagellomeres almost twice as long as wide.
• Hind wing veins 1r–m longer or subequal to M.
• Tergum 9 with sharp edges of median basin short (in basal 0.25).
• Tergum 10 with cornus rather long.

Females of Sericosoma and Teredon have an amazingly similar ovipositor (each annulus of the last ten annuli before the teeth annuli consists of two teeth per annulus with a deep trough between them). If this character state is considered as homologous, then numerous derived character states would have been evolved twice. Therefore, for reasons of parsimony we do not consider Sericosoma and Teredon as sister groups. Their ovipositors are considered as convergent.

3b    Teredon, Eriotremex, Afrotremex and Tremex form a monophyletic group, united by the following derived character states:

• Labial palp with 2 palpomeres. [In Symphyta and earlier lineages of the Siricidae, the palp consists of 3-6 palpomeres.]
• Distance between inner edges of lateral ocelli 1.5–2.3 times greater than between lateral ocellus and nearest edge of eye (Fig. C1.35). [In Symphyta, the lateral ocelli are usually not so close to the edges of eyes.]
• Metatarsomere 1 markedly compressed laterally (about 3 times as high as maximum ventral width) and in dorsal view the lateral surfaces twisted. [In almost all Symphyta, metatarsomere 1 is round or slightly compressed laterally, and is not twisted.]
• Hind wing vein M much longer than vein 1r–m and markedly curved (Fig. C1.26). [In most Symphyta, vein M is not so long.]
• Tergum 9 with median basin sharply outlined laterally for most of its length (Fig. C1.39). [In earlier lineages of Siricidae, the lateral edges are sharp on less than 0.5 of their length.]

4a    Teredon consists of one species and the lineage forms a monophyletic group, defined by the following derived character states:

• Flagellum with 3–8 flagellomeres. [In almost all Symphyta except some Pergidae and a very few Tenthredinidae, the antenna has more flagellomeres.]
• Fore wing vein 2r–m basal to vein 2r–rs (Fig. C1.36). [In Symphyta, vein 2r–m is apical to vein 2r–rs.]
• Metatarsomere 1 remarkably compressed laterally and broadly lobed apically (Figs. C1.8, male, and C1.37, female). [In Symphyta, metatarsomere 1 is typically not lobed dorsally.]
• Metatarsomeres 2–4 very short (as long as metatarsomere 5) (Figs. C1.8, male, and C1.37, female). [In Symphyta and other Siricidae the metatarsomeres 2–4 are much longer than metatarsomere 5.]
• Male with metatibia remarkably compressed laterally (Fig. C1.8). [In Symphyta, the metatibia is almost always not or is at most slightly compressed laterally.]
• Female with basal and apical sheath sections fused and apical section without teeth along dorsal margin in apical 0.3 (only in Xeris and some species of Urocerus from China) (Fig. C1.38). [In almost all Symphyta, the sheath is clearly divided into two sections. The lack of teeth in the apical 0.3 is considered derived within Siricidae.]

Note. Teredon still retains some ancestral characters states shared with the early siricid lineages (Siricinae of Benson (1943)), as follows:

• Hind wing with hamuli both basal and apical to junction of veins C and R1.
• Middle flagellomeres clearly longer than wide (1.1-1.4 times as long as wide).
• Fore wing vein 1cu-a joining vein Cu about half way between vein 1m-cu and M.

4b    Tremex, Eriotremex and Afrotremex form a monophyletic group, united by the following derived character states:

• Fore wing vein 2r-m absent (Fig. C1.27). [In most Symphyta, vein 2r-m present.]
• Fore wing vein 1cu-a aligned or almost aligned with vein M (Fig. C1.27). [In most Symphyta, vein 1cu-a is between veins M and 1m-cu.]
• Hind wing with hamuli present apical to junction of veins R1 and C though hamuli may be present for a very short distance basally (Fig. C1.26). [In Pamphilioidea, Anaxyeloidea and earlier lineages of Siricidae, the hamuli are also present basal to the junction of veins R1 and C.]

5a    Eriotremex forms a monophyletic group, defined by the following derived character states:

• Female with pits of median basin dense or scattered, deep or shallow and basin very convex (Fig. C1.39). [In Siricidae, the median basin has few or no pits and is not so convex.]
• Body pubescence long and very dense (Fig. C1.40). [In Siricidae, the pubescence is usually short and sparse.]
• Sheath with basal section about twice as long as apical section (Fig. C1.40). [In Siricidae except for Teredon, the basal section is subequal or shorter than the apical section.]
• Hind wing with hamuli present for a short distance basal to junction of veins R1 and C (Fig. C1.41). [In Pamphilioidea, Anaxyeloidea and earlier lineages of Siricidae, the hamuli are equally developed basal and apical to the junction of veins R1 and C.]

5b    Tremex and Afrotremex form a monophyletic group, united by the following derived character states:

• Cercus absent (Fig. C1.42). [In Symphyta, the cercus is present.]
• Flagellum with 12-14 flagellomeres (rarely this low in other Siricidae and only in very small specimens). [In Pamphilioidea, Cephoidea, Anaxyeloidea and most Siricidae, the antenna has more flagellomeres.]
• Fore wing cells 2R1 at least 0.6 times to about as long as cell 3R1 (Figs. C30.4 and C1.27); vein 2r-rs joining stigma in apical 0.2–0.33; stigma basal to junction with vein 2r-rs parallel and abruptly narrowed apical to junction (Fig. B1.26). [In Siricidae, cell 2R1 is half as long as cell 3R1, vein 2r-rs joins the stigma near the middle, and the stigma gradually narrows apical to junction.]

6a    Afrotremex forms a monophyletic group, defined by the following derived character states:

• Flagellomere 1 about 0.5 times as long as flagellomere 2 and subequal to flagellomere 3 (Fig. C1.43). [In most Symphyta, flagellomere 1 is subequal to flagellomere 2 and is generally as long as or shorter than flagellomere 3.]
• Head except occiput with setae clubbed at apex (Fig. C1.44). [In Symphyta, setae of the vertex, the frons, the face and the clypeus are apically pointed.]

6b    Tremex forms a monophyletic group, defined by the following derived character states:

• Fore wing cells 2R1 and 3R1 subequal (Fig. C1.27). [In other Siricidae, cell 3R1 is clearly longer.]

Note. Fore wing cell 2R1 length relative to cell 3R1 length is the only character state found to define Tremex as a natural lineage but is variable. Since cell 2R1 varies from 0.9-1.5 times as long as length of 3R1 it may perhaps also be shorter, in which case cell 2R1 could be clearly as small as in Afrotremex. We could therefore treat Afrotremex as part of Tremex, as did Benson (1943). The recent tradition is to consider Afrotremex as generically distinct from Tremex, perhaps because this is the only siricid lineage in tropical Africa whereas species of Tremex are from temperate zones of the Palaearctic and Nearctic regions. Benson (1943) may have been right in including the rather derived Afrotremex within Tremex, but we have studied only a small portion of the Tremex species and have a limited knowledge of the main lineages within Tremex. Based on the few species studied, adults of Tremex show a wide variation in many structures. Therefore, until we better understand the phylogeny of Tremex, we follow the status quo in keeping Afrotremex generically distinct from Tremex.