Euura triandrae ( Benson, 1941 )

Liston, Andrew D., Heibo, Erik, Prous, Marko, Vårdal, Hege, Nyman, Tommi & Vikberg, Veli, 2017, North European gall-inducing Euura sawflies (Hymenoptera, Tenthredinidae, Nematinae), Zootaxa 4302 (1), pp. 1-115 : 85-86

Euura triandrae ( Benson, 1941)

Pontania triandrae Benson, 1941: 131 –132. Described: ♀, gall, recorded host: Salix triandra . Holotype, ♀, BMNH [not examined]. Type locality: England, N. Somerset, Bristol.

Euura triandrae: Beneš (2015b) .

Notes on types and taxonomy. Carleton (1939) first conducted oviposition experiments on this taxon, which showed that females reared from Salix triandra would not readily oviposit on S. × fragilis , and that when they did, galls were undersized, larval mortality high, and the few adults reared were undersized and weak, and would not lay eggs on the "wrong" host. Furthermore, she found phenological differences between sawfly samples reared from these two hosts. Her results led to the tentative conclusion that proxima and triandrae might represent "biological races" of the same species, with the caveat that further investigation was required. Kopelke (2003a, 2005) performed similar trials with both E. triandrae and E. proxima on a larger number of Salix species, and similarly recorded that only the "normal" host species was used by the respective females.

Benson (1941) stated that proxima and triandrae were distinguishable by the length of the ovipositor sheath relative to the length of the metatibia, and the length of valvula 3 relative to the combined length of metatarsomeres 1+2. In both cases, the index was stated to be larger in proxima . In seven females reared from S. triandra and six from S. × fragilis we obtained indices of ovipositor sheath length / head width of respectively 0.98–1.14 (mean 1.03) and 0.93–1.27 (mean 1.07). It was not possible to measure the leg parts suggested by Benson, because these could not be seen clearly in most of the available specimens. Nevertheless, we suspect that ovipositor length is not a suitable character for separating these species. Kopelke (2005: 87) in his identification key mentioned the following diagnostic characters for distinguishing proxima from triandrae (character state of the latter in parentheses): sawsheath in dorsal view triangular (sawsheath in dorsal view almost evenly rounded); saw slightly curved only in the basal part (saw slightly S-shaped); ctenidea starting at the 2nd annulus (ctenidea starting at the 1st annulus); saw consisting of 23–24 annuli (saw consisting of 24–25 annuli); head and mesonotum with slight microsculpture, mostly not dotted (head and mesonotum with slight microsculpture, normally finely dotted). In the specimens available for examination, we could not detect differences between the two segregates for any of these characters, except for the number of annuli of the lancet. However, the lancets of proxima (two specimens from Finland) had 24 and 25 annuli, while triandrae (one specimen from England) had 22 annuli. Ctenidia were present from annulus 2 in all three specimens. We conclude that these two taxa are in practice not morphologically separable.

Carleton (1939) and Benson (1941, 1958) observed that the appearance of the galls of these nominal species differ. Such differences may be a result of physiological and chemical differences in the respective host species ( Beneš 1968a) and are therefore not necessarily of taxonomic value with respect to the gall-maker. Carleton (1939) undertook no-choice oviposition trials, in which females reared from S. triandra were offered S. fragilis for oviposition, and vice versa. Zinovjev (1999: 207), commenting on the results of Carleton's oviposition tests, wrote that "All galls that developed on S. triandra [induced by females reared from S. fragilis ] retained shape and colour typical for galls on S. fragilis , although they were undersized". Carleton (1939) did not however provide such a clear statement of differences: she merely noted (p. 595) that "[galls induced by females reared from S. fragilis ] were pale greenish-yellow and never became red on the upper surface as in normal galls".

Vikberg (1970) has already proposed the formal synonymy of triandrae with proxima . However, the differences in host choice certainly speak for some degree of genetic segregation. Whether these nominal taxa should be regarded as host plant races, or separate species, cannot at present be decided. For the moment, it seems desirable to retain the name triandrae , because of the rather large number of biological studies that have previously treated these segregates as separate species.

Variability. Female: Body length: 2.9–4.5mm. Lateral vertex with obscure brown fleck, to black. Male: none examined. Total number of specimens examined: 13.

Genetic data. See above, under E. proxima .

Similar species. Morphologically not clearly distinguishable from E. proxima (see there).

Bionomics. Host plant: Salix triandra ( Benson 1941, Kopelke 2005). Biology: Carleton (1939; as P. proxima on S. triandra ), Hjältén et al. (2007), Leitch (1994; as P. proxima on S. triandra ), Magnus (1914; as P. proxima on S. amygdalina ), Niemi (2006), Price (2003), Rey (1967, 1968; as P. proxima on S. triandra ).

Distribution. Central and North Europe north to N. Trøndelag in Norway ( Kopelke 2005). Occurrence in Sweden: published records; Skåne ( Coulianos & Holmåsen 1991), Norrbotten, Torne älv ( Wahlgren 1929). Material examined: Dalarna.