Sequencing results

Phylogenetic analyses of 1078 bp of COI sequences showed with strong statistical support (bootstrap proportion 92%) that the closest relative of the putative Eupareophora exarmata prepupa is the Nearctic E. parca (Fig. 1), from which it nevertheless differs significantly at the sequence level, by 10.8%. The tree is otherwise poorly resolved, because of the limited amount of sequence data used. Closest relatives of the genus Eupareophora on the tree are Cladardis, Monardis, and Periclista, although without statistical support (Fig. 1). A strongly supported clade of the latter three genera was found by Malm and Nyman 2014 using 8 nuclear and one mitochondrial (COI) protein coding genes, suggesting that Eupareophora (which Malm & Nyman 2015 did not sample) might also belong there.

Figure 1.  

Maximum likelihood tree of COI (1078 bp) sequences of Tenthredinidae. Numbers at nodes are bootstrap proportions (%) derived from 500 pseudoreplicates. Nodes receiving bootstrap support less than 50% have been collapsed, except for those closest to Eupareophora.

Hosts

Liston 1995 stated that the hosts of Eupareophora exarmata are Rosa species (Rosaceae) and that the larvae bore in shoots. Although not cited by Liston, this statement was based on a record by Reichert 1933: “mit Larven von Ardis brunniventris eingetragene Rosenzweige ergaben 13.2.18. im geheizten Zimmer 1 #w, det Enslin”. In view of the morphological similarity of adult M. plana (whose hosts are Rosa spp.: Scheibelreiter 1973) and E. exarmata, it seems likely that Enslin misidentified the specimen. Zhelochovtsev [Zhelohovcev] 1988) mentioned under E. exarmata simply “on ash” [translated] (Fraxinus sp., Oleaceae). Probably this information is based on original observations made by Supatashvili et al. 1972 al. (1972) in Georgia, who reared adults from larvae. Adults were examined by Zhelochovtsev and determined as E. exarmata. The larva is very briefly described by Supatashvili et al. 1972 [translated]: “Larva grey coloured, body covered with awl-shaped processes”. These authors also record “ash” as the host, but do not mention which Fraxinus species was involved. Apart from Fraxinus excelsior L., some other ash species occur in Georgia, such as F. angustifolia Vahl subsp. oxycarpa (M. Bieb. ex Willd.) Franco & Rocha Afonso (USDA 2012). The recent German records indicate that F. excelsior L. is a host. As far as we are aware, the publication by Supatashvili et al. 1972 is unique in referring to E. exarmata as a pest. Their observations were made in stands of planted ash. It is noteworthy that the occasional reports of defoliation caused by E. parca in North America involve “planted ash species [..] in urban settings” (Williams 2007), although D. R. Smith (personal communication) states that it is also fairly common throughout the eastern deciduous forests.

Description of mature larva (Figs 2, 3, 4).

Figure 2.  

Eupareophora exarmata, larva, last feeding instar; Germany, Hagen-Hohenlimburg. Photos: J. Bücker.

Figure 3.  

Eupareophora exarmata, larva, last feeding instar.

Figure 4.  

Eupareophora exarmata, larva, last feeding instar; spl=surpedal lobe, ssl=subspiracular lobe.

Terminology follows Viitasaari 2002, with notation of annulets of abdominal segments according to Vikberg and Nuorteva 1997, i.e. annulet 3 bears the spiracle.

Length: approximately 15 mm.

Head completely black except for pale mouthparts. Ground colour of trunk above spiracular line grey; whitish below this, with yellow tinge on abdominal segments 1-8. Cuticular processes (hereafter: spines) above spiracles located on more or less black glandubae. Above spiracles, on thorax, most spines entirely black; on abdomen all supraspiracular spines blackish above fork, whitish below this; the outermost of each dorsal pair of spines darker. All subspiracular spines paler than more dorsal ones; apically at most pale brown, and if located on glandubae, then these also completely pale.

Antenna with 5 articles. Clypeus with 2 setae. Thoracic leg with 5 articles. The 4 most dorsal and anterior spines on thorax are trifid. Prolegs on abdominal segments 2-8 and 10. Abdominal segments 1-9 with 5 dorsal annulets. Annulet 3 with 2 supraspiracular bifid spines. Annulet 5 with 3 bifid spines: 2 supraspiracular and 1 on spiracular line. Abdominal segment 10 without spine on midline. Subspiracular lobe with two spines; anterior one bifid, other simple. Suprapedal lobe with two simple spines.

Description of prepupa (Fig. 5).

Figure 5.  

Eupareophora exarmata, Germany, Hagen-Hohenlimburg; prepupa. Photo: J. Bücker.

Length: approximately 13 mm.

Head grey above; yellowish on and around mouthparts. Thorax yellow-white. Abdomen largely grey, with yellow patches on and below spiracular line, and yellowish prolegs.

Spines absent, except on abdominal segments 9 and 10, where they are replaced by unbranched, peg-shaped structures. Prothorax dorsally and anteriorly more strongly produced than in the feeding larva, giving it a hooded appearance.

Identification

Other spiny West Palaearctic Blennocampinae larvae belong to the genera Monardis, Periclista, Pareophora, Monophadnoides and Claremontia.  Larvae of all of these, none of which feeds on Fraxinus, have a mainly pale green or yellowish body and are thus easily distinguished from the predominantly grey larva of E. exarmata. Smith 1969 stated that the larva of Eupareophora parca has two bifurcate spines on the subspiracular lobe, and can therefore be distinguished from those of Periclista species in which the anterior of these two spines is bifurcate and the posterior one simple. The larva of E. exarmata in this respect (Fig. 4) is however like Periclista, not E. parca. The coloration of the mature larva of E. parca (illustrated by Williams 2007) and E. exarmata (Figs 2, 3, 4) is similar, although the latter is apparently darker. According to Williams 2007 (fig. 10B), the caudal abdominal terga of the E. parca prepupa entirely lack large cuticular processes, whereas the prepupa of E. exarmata clearly possesses some (Fig. 5).

Behaviour

All adult collection records and observations on larvae (here, and by Supatashvili et al. 1972, suggest that E. exarmata is univoltine, flying soon after bud-break in spring, with larvae developing, according to local climate, between the end of April and start of June. The behaviour of the feeding larvae, although only briefly described by Supatashvili et al. 1972 seems to resemble quite closely that of the Nearctic E. parca as described by Williams 2007.

At Hagen-Hohenlimburg only mature larvae of E. exarmata were found, apparently when they crawled down the trunk of the host in order to reach a spot in which to complete their development. Williams 2007 recovered eight prepupae of E. parca from the soil litter layer, and 2 from branches of the host. Further observations are therefore needed to establish whether E. exarmata always leaves its host before moulting to a prepupa. According to Supatashvili et al. 1972, the mature larvae form cells in the bark of Fraxinus, in which they overwinter. A further apparent peculiarity noted by these authors, is that the freshly moulted larvae lack spines, but that these re-appear within a day. Neither of these phenomena was observed by Williams 2007 in the Nearctic E. parca, who found that the rather flimsy cocoon was usually constructed in the upper layers of the soil.