The West Palaearctic genera of Nematinae (Hymenoptera, Tenthredinidae)

Abstract Keys to adults and larvae of the genera of West Palaearctic nematine sawflies are presented. Species of some of the smaller genera are keyed, and their taxonomy, distribution, and host plants reviewed, with a geographic focus on north-western Europe, particularly Sweden. Dinematus Lacourt, 2006 is a new junior subjective synonym of Pristiphora Latreille, 1810, resulting in the new combination Pristiphora krausi (Lacourt, 2006) for the type species of Dinematus. Hemichroa monticola Ermolenko, 1960 is a new junior subjective synonym of Hemichroa australis (Serville, 1823). Lectotypes are designated for Tenthredo opaca Fabricius, 1775, Mesoneura opaca var. nigerrima Enslin, 1914, Mesoneura opaca var. obscuriventris Enslin, 1914, Nematus hypogastricus Hartig, 1837, Nematus alnivorus Hartig, 1840, Leptopus rufipes Förster, 1854, Nematus protensus Förster, 1854, and Platycampus luridiventris var. pleuritica Enslin, 1915. A phylogenetic analysis based on four genes (mitochondrial COI and nuclear NaK, POL2, and TPI) supports the current generic classification.


Introduction
In 2012 a project funded by the Swedish Taxonomy Initiative was launched, with the main objective of improving our knowledge of the taxonomy and distribution of nematine sawflies in Fennoscandia, and Sweden in particular (STI Nematinae Group 2013). As a first step, the generic classification of the world Nematinae was revised by Prous et al. (2014), and the genera keyed. Here, we present a condensed version of that key, covering only the West Palaearctic genera, with which it should be possible to identify most specimens more easily. Included are treatments of the species of some smaller genera: Hemichroa, Mesoneura, Neodineura, Platycampus, and Stauronematus. The species of the other genera were either covered by Prous et al. (2017) and Liston et al. (2017Liston et al. ( , 2019a, or are to be dealt with in works currently in preparation. Geographic scope of the taxonomic treatments at genus / species group level varies between coverage of the whole West Palaearctic, to consideration only of the species which are known from Fennoscandia, or potentially present there. The differences in the size of regions covered for each genus / species group arise through the amount of material available for study, including fresh specimens suitable for genetic sequencing, and the perceived complexity of species-level taxonomy in the group. The present work thus represents an overview of all Nematinae known to occur in Fennoscandia, and in conjunction with the publications covering the remaining genera is intended to enable determination to species level of specimens of all nematine genera from north-west Europe.

Materials and methods
The Swedish Malaise Trap Project is abbreviated to SMTP. Abbreviations for the names of collections referred to in the text are as follows:
In the specimen data the dates are given as dd.mm.yyyy, and coordinates as positive (north or east) or negative (south or west) decimal degrees latitude and longitude.
Morphological terminology mostly follows Viitasaari (2002), but sawtooth is used instead of serrula (see Malagón-Aldana et al. 2017), and the large, ventrally situated, more or less triangular flange above each sawtooth is called a spurette (following Ross 1943;see Figs 108, 112 arrows). Images of complete imagines and morphological details were made at the SDEI with Leica cameras attached to a variety of microscopes. Composite images with an extended depth of field were created from stacks of images using the software CombineZP, and finally arranged and partly enhanced with Ulead PhotoImpact X3. Some of the figures were first published by Prous et al. (2014). Unless otherwise stated, photos of adults and larvae were made by AL, MP, HS, and AT.
First drafts of the key to larvae were based mainly on Lorenz and Kraus (1957), and subsequently modified to include the results of more recently published studies, and the examination of specimens available to us. The tree species known as Mountain Birch, which dominates large areas of vegetation in northern Fennoscandia, is referred to as Betula pubescens var. pumila (Zanoni ex Murray) Govaerts, following Plants of the World online (2017), which treats the formerly widely-used names B. czerepanovii N. I. Orlova and B. tortuosa Ledeb. as its synonyms.
DNA was extracted and purified with an EZNA Tissue DNA Kit (Omega Biotek) according to the manufacturer's protocol and stored at -20 °C for later use. Typically, one or two legs were used for DNA extraction, but for males the whole genital capsule was often additionally used to increase DNA yield and to free penis valves from muscles before photography. In some cases, the whole specimen was used for extraction. One mitochondrial and four nuclear regions were used in the phylogenetic analyses, although not all of these genes were obtained for all species. Primers used for amplification and sequencing are listed in Table 1. The mitochondrial region used is a large fragment (1078-1087 bp depending on the primer set) of the cytochrome oxidase subunit I gene (COI). The fragment includes the entire standard barcode region (658 bp) of the animal kingdom (Hebert et al. 2003). The nuclear markers used are fragments of sodium/potassium-transporting ATPase subunit alpha (NaK), triose-phosphate isomerase (TPI), DNA dependent RNA polymerase II subunit RPB1 (POL2), and transformation/transcription domain-associated protein (TRRAP). The NaK fragment used is a nearly complete sequence of its longest exon, 1654 bp. The TPI fragment used is the nearly complete gene region, containing 676 bp of three exons and two short introns (each around 50-100 bp) in Nematinae, altogether 788-842 bp. The POL2 fragment used is composed of two partial exons (together 2407-2623 bp depending on the primer set) and one short intron (67-86 bp). The TRRAP fragment used is a 3379 bp fragment of its longest exon (sequenced only for Hoplocampa and Monocellicampa). New POL2 and TRRAP primers were designed mainly based on four sawfly genomes (accessions AOFN02000108, AOFN02000124 [Athalia rosae], LGIB01000723, LGIB01000528 [Neodiprion lecontei], AMWH01002735, Table 1. Primers used for PCR and sequencing (preferred primers in bold), with information provided on respective gene fragment, primer name, direction (forward, F or reverse, R), primer sequence, standard PCR annealing temperature, utilization (PCR/ sequencing), and reference. Primer annealing temperatures used for sequencing at Macrogen were usually 50 °C (47-50 °C).  (Misof et al. 2014, Peters et al. 2017) available in Gen-Bank. Numbers in the new POL2 and TRRAP primer names refer to the binding position of the 3' end of each primer in the coding region of Athalia rosae mRNA (accessions XM_012395805 and XM_012406083). PCR reactions were carried out in a total volume of 15-35 μl containing 1.0-2.5 μl of extracted DNA, 1.5-3.5 μl (5.0-15 pmol) of primers and 7.5-17.5 μl of 2× Multiplex PCR Plus Master mix (QIAGEN). The PCR protocol consisted of an initial DNA polymerase (HotStar Taq) activation step at 95 °C for 5 min, followed by 38-40 cycles of 30 s at 95 °C, 90-120 s at 49-60 °C (depending on the primer set used), and 70-180 s (depending on the amplicon size) at 72 °C; the last cycle was followed by a final 30 min extension step at 68 °C. COI (primers symF4 [or symF1] + A2590), NaK (NaK_263F + 1918R) and TPI (TPI_29Fi + TPI706R) were in most cases amplified in one fragment, POL2 in one to three fragments, and TRRAP in two fragments (TRRAP_833F + 3046R and TRRAP_2648Fi + 4213Ri). Three μl of PCR product was visualised on a 1.4% agarose gel and the remaining product was then purified with FastAP and Exonuclease I (Thermo Scientific). 1.0-2.2 U of both enzymes were added to 12-32 μl of PCR solution and incubated for 15 min at 37 °C, followed by 15 min at 85 °C. 2-5 μl of purified PCR product per primer in a total volume of 10 μl (5-8 μl of sequencing primer at concentration 5 pmol/μl) were sent to Macrogen Europe (Netherlands) for sequencing. Both sense and antisense strands were sequenced using the primers listed in Table 1. Ambiguous positions (i.e., double peaks in chromatograms of both strands) due to heterozygosity were coded using IUPAC symbols. Sequences reported here have been deposited in the GenBank (NCBI) database (accession numbers MK624656-MK624923 and MK720818-MK720821), although not all of them are analysed here (covered in further publications on some of the genera not treated here). Some of the sequences analysed here were originally published by Schmidt et al. (2017) and Prous et al. (2016Prous et al. ( , 2017. Alignment of COI, NaK, and TRRAP sequences was straightforward because of the lack of indels (insertions or deletions). Alignment of POL2 and TPI was also straightforward without introns, but these were retained in some analyses published elsewhere (Liston et al. 2019a) and aligned manually. To concatenate separate gene alignments, we used R (R Core Team 2018) package apex (Jombart et al. 2017). For phylogenetic analyses we used the maximum likelihood method (ML) implemented in IQ-TREE 1.5.6 (http://www.iqtree.org/) (Nguyen et al. 2015). By default, IQ-TREE runs ModelFinder (Kalyaanamoorthy et al. 2017) to find the best-fit substitution model and then reconstructs the tree using the model selected according to Bayesian information criterion (BIC). We complemented this default option with SH-like approximate likelihood ratio (SH-aLRT) test (Guindon et al. 2010) and ultrafast bootstrap (Hoang et al. 2017) with 1000 replicates to estimate robustness of reconstructed splits. Minimal p-distances between and maximal distances within BIN (Barcode Index Number) clusters were taken from BOLD (http://www. boldsystems.org/) BIN database. Some of the COI barcode sequences used here were obtained from BOLD (http://www.boldsystems.org/). In this case, DNA extraction, PCR amplification, and sequencing were conducted at the Canadian Centre for DNA Barcoding (CCDB) in Guelph, Canada, using standardised high-throughput protocols (Ivanova et al. 2006, deWaard et al. 2008, available online under www.ccdb.ca/ resources.php. DNA aliquots of SDEI vouchers are deposited in the DNA storage facility of the SDEI (including those that were originally extracted in CCDB).

Results
Previous taxonomic publications have mostly recognised several tribes within the Nematinae. For example, Vikberg (1982) allocated the North European genera to six tribes, of which his Nematini was further divided into three sub-tribes. Subsequently, additional tribes were erected, often for species-poor lineages with more or less distinctive morphological and biological characters, e.g., Pristicampini (Zinovjev 1993), Stauronematini, and Bacconematini (Lacourt 1998). The circumscription of the tribes, and even of the Nematinae itself, has varied considerably between authors. Lacourt (1998), for example, removed Cladius, Hoplocampa, and Susana from the Nematinae, and treated each of these as a separate subfamily of Tenthredinidae. A clearer and more objective assessment of suprageneric classification was first achieved with the application of genetic data by Nyman et al. (2006). A second analysis in Prous et al. (2014), based on extended taxon sampling and more genes, yielded essentially similar results. A further refinement based on mitochondrial COI and three nuclear genes (NaK, POL2, TPI), with stronger support for some clades, is presented in Fig. 1. Noteworthy is that Nyman et al. (2006), Prous et al. (2014), and Malm and Nyman (2015) all recovered the Nematinae as monophyletic and indicated that Cladius (missing in Malm and Nyman 2015), Hoplocampa, and Susana do belong to the subfamily. Because monophyly of Nematinae is unambiguously supported based on previous analyses using the same genes, we did not test this here further. Our analyses of the subfamily without outgroups supports the previous generic classification as proposed in Prous et al. (2014). Because of limited sampling, Prous et al. (2014) were unable to state whether the three subgenera of Cladius are monophyletic, but based on expanded sampling, we now find that the largest subgenus Priophorus is not (Fig. 1). Because the delimitation of the subgenera of Cladius is problematic also morphologically, we propose here to abandon subgeneric classification until better evidence justifies it. Whether the various tribal names which have been proposed for single genera have much practical value is questionable. Hoplocampa, Stauronematus, and Susana, for example, although apparently phylogenetically isolated from other genera, are more clearly referred to by using their generic names. This will remain so at least until genetic data become available for a number of morphologically distinctive genus-series taxa. In the West Palaearctic, genetic data are still lacking for Armenocampus, Neodineura, and Nescianeura. On the other hand, to simplify discussions on phylogeny and biodiversity, use of the tribal names Nematini (equivalent to the "higher Nematinae" of Prous et al. 2014), Dineurini, and Pseudodineurini seems justified and useful. Support for Nematini and Dineurini (Pseudodineurini could not be tested because of the lack of sampling) in our molecular phylogeny is unambiguous (Fig. 1)  Only specimens sequenced for all four genes were included. Short introns from POL2 and TPI were excluded. The best-fit model chosen according to Bayesian information criterion was GTR+R4. Numbers at branches show SH-aLRT support (%) / ultrafast bootstrap support (%) values. Support values for weakly supported branches (<90) are not shown. Letters "f " and "m" stand for "female" and "male", and are not given for larvae. Numbers at the end of the tip labels refer to the length of the sequence and the number of ambiguous positions (e.g., heterozygosities). The number of ambiguous positions given for two males are due to variation in mitochondrial COI because of possible heteroplasmy. The tree was rooted as in Prous et al. (2014). The scale bar shows the number of estimated substitutions per nucleotide position.

Key to the West Palaearctic genera and selected species of Nematinae (larvae)
Numbers of setae on dorsal annulets are for only one side of the body, as in Lorenz and Kraus (1957). The best results should be possible with full-grown larvae, but before these undertake a final "extra moult", in the groups where this applies. Presence or absence of the extra moult is a useful additional taxonomic and identification character in itself (Kontuniemi 1965), but can usually only be scored if the larvae are reared. Larvae of many species which perform an extra moult differ greatly in appearance after this moult from preceding instars: colour pattern and ground-colour frequently change, and setation can be much reduced. Even in species which have no extra moult, pronounced colour differences between instars are often noticeable. Larvae of the monotypic genera Armenocampus, Neodineura, and Nescianeura are unknown, as well as the larvae of many species of Euura and Pristiphora, particularly the northern species. Even in the less speciose genera, larvae of some species are undescribed, while several others are insufficiently described, or existing descriptions are partly contradictory, e.g., for Cladius compressicornis and brullei. Because high interspecific morphological variability is already evident in Euura larvae, it would not be surprising if larvae were found which have combinations of characters not included in the key. Only the two species of the Nematus wahlbergi group known in Sweden are included. Descriptions of larvae of some of the other species of this group may be found in Zinovjev (1979). We have seen no specimens or images of larvae of Nematus brischkei: the characters used below to distinguish it are taken from the descriptions by Zaddach (1876) and Chambers (1950). In view of the incomplete and imperfect nature of the available data, the key is highly provisional. Unless otherwise stated, the larvae are exophytic, and feed mostly on leaves. The numbers of species refer to Fennoscandia.   (Fig. 89); b A row of dark flecks above the abdominal prolegs (Fig. 89)

Taxon commentaries
Synonymy of genus-group names was given by Prous et al. (2014) and is not repeated here, except for Euura and Nematus, where the synonymy proposed in the former work is extensive, and probably not yet familiar to many users. The known nomina nuda and names for aberrations (unavailable names following International Commission on Zoological Nomenclature (1999)) for the listed species were given by Taeger et al. (2010). Taxa are dealt with in alphabetical order.

Anoplonyx Marlatt, 1896
No reliable key or species treatments are available to date.

Armenocampus Zinovjev, 2000
This genus was erected for a single species, Armenocampus necopinus (Zhelochovtsev, 1941), originally described as Caulocampus necopinus, known only from the small type series of both sexes collected in Armenia. Nothing is known about its biology.

Cladius Illiger, 1807
No reliable key or species treatments are available to date.

Euura Newman, 1837
Prous et al. (2014) treated a large number of genus-group names as synonyms of Euura. A complete list of these is contained therein. The synonyms listed below have been recently used as valid for West Palaearctic taxa. Nearly all species formerly included in these genera, and the majority of species previously placed by many authors in Nematus, now belong to Euura. The north-west European gall-making species of Euura were recently revised by Liston et al. (2017).

Pontania Costa, 1852
Amauronematus Konow, 1890Pachynematus Konow, 1890Pteronidea Rohwer, 1911Pontopristia Malaise, 1921(Malaise 1921a Brachycoluma Strand, 1929Decanematus Malaise, 1931(Malaise 1931a) Pikonema Ross, 1937Phyllocolpa Benson, 1960(Benson 1960a Eitelius Kontuniemi, 1966 Gemmura E.L. Smith, 1968Eupontania Zinovjev, 1985Larinematus Zhelochovtsev, 1988Polynematus Zhelochovtsev, 1988Bacconematus Zhelochovtsev, 1988Alpinematus Lacourt, 1996Epicenematus Lacourt, 1998Kontuniemiana Lacourt, 1998Lindqvistia Lacourt, 1998  a Abdomen yellow or orange except for black valvula 3 and more or less tergum 1 (Figs 98, 100); b Upper mesepisternum yellow, lower part black (Fig. 100) (Fig. 107); b Parts of abdominal terga and sterna sometimes pale (Fig. 102) (Figs 104-106); bb Abdomen entirely black, except for harpes and more or less distal edge of sternum 9 (Fig. 103)  Only a single paratype of monticola was available for examination, but we also examined four females (HNHM) which have the combination of colour characters described for monticola and were collected at subalpine levels in the Ukrainian Carpathians, as was the type series of monticola. We did not observe any significant difference in the depth of the clypeal emargination between Carpathian specimens and australis from other parts of Europe. The other characters used to distinguish monticola are either extremely weak, such as the slightly darkened tips of the cerci and the degree of curvature of the lower edge of valvula 3, or are variable among studied australis females, such as the length of the hind wing anal cell and the presence or absence of denticles on the more basal serrulae of the lancet (Figs 108-111). The shape of sawteeth and the number of serrulae can even vary between the left and right lancets of the same individual (Figs 108-109), possibly as a result of wear (see Schmidt and Walter 1995). Ermolenko considered H. monticola to be a neo-endemic element of the Carpathian subalpine fauna, associated with Alnus viridis, but several of the characters which he gave as distinguishing it from australis occur apparently independently of each other in the australis females which we have examined from many parts of the West Palaearctic. For example, tergum 9 mainly pale, but whole wing-membrane blackish from base of fore wing up to approximately the level of the pterostigma [Germany, Berlin], or antennae entirely black, and wing membrane nearly entirely hyaline, but 9 th tergum black [Sweden, Lapland]. In our opinion, Ermolenko underestimated the range of variability in australis, and monticola falls within this range. Therefore, we treat the taxa as conspecific. Nevertheless, comparison of relevant genetic data should still be undertaken.
Previously published descriptions of the male of Hemichroa australis, and the colour characters which are claimed to distinguish it from that of crocea, are partly contradictory, and may not be reliable. Enslin (1915: 317) wrote [translated from German]: "According to Cameron, the male of H. crocea Geoffr. is just like that of H. alni [australis]; Cameron (Monograph Brit. Phyt. Hym. II p. 7) saw some males of crocea reared by Fletcher and could not distinguish them from H. alni. Because nothing further on this subject is reported in the literature and it was not possible for us to obtain males of H. crocea for examination, the separation of the males of these species must remain unresolved until a later date". Benson (1958) stated that the male of australis "Differs from crocea ♂ in that the antenna is at least red below [crocea: antenna entirely black] and the stigma of the wing is piceous [crocea: pterostigma brown in the middle] ". , in his key to World Hemichroa species, wrote that he did not know the male of australis, and repeated the characters given by Benson (1958). But in the text under H. crocea,  wrote "It may be separated from other species by the presence of the radial crossvein [2r-rs] in the fore wing and characters of the genitalia (figs 3, 4)". The first character state was surely mentioned in error: all Hemichroa species usually possess vein 2r-rs, except for the taxon treated by  as H. militaris (Cresson, 1880), which is currently placed in Dineura (Fig. 1, Prous et al. 2014). See below under crocea for additional discussion of diagnostic characters of males of australis and crocea.
Description. Body length: female 6.5-8.5 mm, male 6.0-6.5 mm. Wing colour highly variable in both sexes, from nearly entirely hyaline, to entire hind wing and basal fore wing up to about pterostigma conspicuously darkened. Female (Figs 99, 101): Black. Red are head, except more or less for labrum and antenna; pronotum, tegula, mesoscutum, more or less mesoscutellar appendage; more or less the apex of abdomen. Legs black, except for more or less brownish fore legs. Lancet: Figs 106-109. Male (Fig. 103): Head and body entirely black, except more or less for underside of antennae, tegulae, extreme upper posterior edge of pronotum, and subgenital plate. Legs entirely red, except for black coxa and more or less trochanters and trochantelli. One male (DEI-GISHym20617), presumably atypical, has the thorax red and black patterned, exactly as in females. Penis valve: Figs 104-106; note the variability in shape of the distal projections.
Our characterisation of the male of australis is based primarily on three specimens from Germany (BC ZSM HYM 04094), Lapland (DEI-GISHym20618), and Japan (DEI-GISHym84982), with identity confirmed by barcoding. Fore wing basally darkened or mostly subhyaline, the antennae black with reddish undersides (or nearly completely pale in the Japanese specimen), and the stigma uniformly dark. The body is completely black, except for the slightly brown tegulae, harpes, and distal edge of sternum 9; and all tibiae completely pale. One further male from Torne Lappmark in the SDEI, and the long series of males from Ukraine, have the same coloration except for mostly subhyaline fore wing. The latter exhibit little variability, except that the tegulae and upper posterior edges of the pronotum may be completely black, or more or less brown, and the antennae usually extensively reddish, but occasionally nearly completely black. The wing veins of the males from Lapland, including the fore wing pterostigma, are, however, darker than the Ukrainian specimens.
Similar species. See key, and notes on male (above, and under crocea, below). Compared with crocea (Fig. 112), the most obvious differences in the lancet of australis (Figs 108-111) are the greater number and smaller size of ctenidia on the annular sutures, smaller distance between each basal and median sawtooth and its spurette, and its less hooked median sawteeth.
Life history. Host plants (in Europe): Betula pendula, pubescens (Kontuniemi 1960), pubescens var. pumila (see Specimens examined), utilis (Schedl 2010), Alnus glutinosa, incana, and viridis (Kontuniemi 1960, Pschorn-Walcher andAltenhofer 2000), and further Alnus species in the East Palaearctic. Larvae solitary, and cryptic (Fig. 88).  compared the defensive strategy of australis and crocea larvae. Two overlapping generations in the lowlands. Although males of both European Hemichroa species have generally been considered to be rare (e.g., Benson 1958, males of australis are, at least regionally, evidently rather abundant. In a series of 104 specimens collected by Ermolenko in the montane zone of the Ukrainian Carpathians, 92 are males, and 2 of 5 specimens recently collected in the Torne Träsk Region are males. Malaise (1921b) also noted that although males of australis are usually extremely rare, three of six specimens which he collected in the Torne Träsk area were males. Perhaps males are more frequent in areas with a cooler climate, which would represent an interesting departure from the usual pattern in Tenthredinoidea of a higher female to male ratio in warmer areas (Benson 1950: 126).

Distribution. Trans-palaearctic from the British Isles, through north and central
Europe  to Yakutia (Sundukov 2017) and Japan ; see also Specimens examined).
We have only examined one old male specimen (DEI-GISHym31838), without genetic data, which we think belongs to crocea, because of the similarity of its penis valve to that illustrated by  fig. 4) as crocea, and differences in the penis valves of australis identified by us, using sequence data. This crocea male has its abdomen and parts of the mesoscutum extensively yellow, but completely black antennae, as well as darkened metatarsus and metatibia apex. However, the original descriptions of the males of Hemichroa dyari, pallida and washingtonia (Rohwer 1918, Rohwer andMiddleton 1932), all of which are currently treated as synonyms of H. crocea, indicate that body colouration is variable, and can be as dark as in male australis. The metatibia and metatarsus may apparently also be dark or pale, as respectively described by Rohwer (1918) for males of dyari and pallida. On the other hand, the descriptions of North American crocea males suggest that the antennae are completely dark, as described by Benson (1958) (Pschorn-Walcher and Altenhofer 2000). Salix is mentioned repeatedly in various works as a host, but no unambiguous original record of feeding by larvae on Salix has been located. Larvae gregarious, and brightly coloured (Fig. 87).

Mesoneura Hartig, 1837
Only two species are known from the West Palaearctic (Liston 2012), and only M. opaca occurs in north-west Europe. The nominal taxon described as Tenthredo (Selandria) umbrosa Eversmann, 1847 was treated in several works (e.g., Dalla Torre 1894, Konow 1905, Taeger et al. 2010) as a third, valid West Palaearctic Mesoneura species, but examination of the type revealed it to be a male specimen close to Euura clitellata (Serville, 1823). (2012) a Upper side of abdomen mainly black; at least with a continuous black dorsal vitta (Fig. 113); b Lancet with 14-15 annuli; serrulae, particularly basal ones, rather flat (Fig. 115) (Fig. 114); bb Lancet with ca. 20 annuli; serrulae prominent, hooked (Fig. 116)  a Abdominal terga 5-8 with a deep, sharply delimited medial depression edged with a row of long setae (Fig. 117); b All terga mainly black, except for more or less pale extreme apical margins; c Apical margin of sternum 9 medially slightly produced (Fig. 117); d Length 6.5-8.0 mm....... Mesoneura opaca ♂ -aa Abdominal terga 5-8 with at most a shallow, ill-defined medial depression, without row of modified setae along edge (Fig. 118); bb Terga 2-4 entirely yellow-brown (Fig. 118); cc Apical margin of sternum 9 truncate or medially even slightly emarginate (Fig. 118) Description. Body length: female 5.5-9.0 mm, male 6.5-8.0 mm. Female (Fig. 113): head including antenna black, except for white clypeus and labrum, and sometimes brown flecks on interantennal area / just dorsal of toruli / lower outer orbits. Thorax black. In darkest specimens only pronotum and tegula pale. Palest specimens with yellow-brown whole median mesoscutal lobe, parts of lateral lobes, mesoscutellum and appendage, upper mesepisternum, and parts of metanotum. Fore wing pterostigma completely pale, to pale in middle with darkened edges. Legs pale, with coxae, femora and apical tarsomeres more or less darkened. Abdomen from completely black, to completely pale on underside with lateral parts of terga more or less pale, and pale tergum 10 and cerci. Lancet: Fig. 115. Male (only four examined): Black; only ventral parts of clypeus pale, labrum pale to nearly completely dark. Thorax at most with pale edges of pronotum, and more or less tegulae. Leg colour similar to female, but darkest males with apex of metatibia darkened, and palest with tarsi completely pale. Abdomen black except for brownish narrow distal margin of sternum 9 and more or less harpes, and sometimes around the depressed parts of terga 5-8. Penis valve: Liston (2012: fig. 4) [not distinguishable from that of lanigera]. Similar species. In the West Palaearctic, only Mesoneura lanigera Benson, 1954 (south-east Europe, Transcaucasus and Cyprus) could be mistaken for opaca: see key.

Key to West Palaearctic species, based on Liston
Life history. Host plants: Quercus species, including robur (Pschorn-Walcher and Altenhofer 2000), pubescens, and rubra (Liston 2011). Univoltine species. Oviposition in the leaf midrib or side-veins; maximum two eggs per leaf. Larva (Fig. 69) solitary. Normally entirely parthenogenetic in most of central and northern Europe, where males have so far only been found in the Netherlands (Ad Mol, pers. comm.), but males are apparently more frequent in Greece (Liston 2012, Liston et al. 2015.
Distribution. Widespread in central and southern Europe, from the British Isles, north to Finland

Nematinus Rohwer, 1911
No reliable key or species treatments are available to date.

Nematus Panzer, 1801
No reliable key or species treatments are available to date. Prous et al. (2014) radically altered the circumscription of Nematus: see also under Euura, above. The following synonyms of Nematus have been in recent use as valid: Craesus Leach, 1817 [= Croesus, misspelling], Hypolaepus W.F. Kirby, 1882, andParanematus Zinovjev, 1978. Note that most of the species placed in Hypolaepus by Lacourt (1999) are now placed in Euura.
Description. Body stocky, similar to Mesoneura. Fore wing radial cell divided. Radial cross vein (2r-rs) arises near the apex of stigma and meets the cell 1Rs2; basalis (M) and 1 st medial cross vein (1m-cu) strongly converging; M clearly bent only basally; intercostal crossvein (Sc) lying before the junction of M with the Subcosta (Sc+R+Rs); 1 st and 2 nd medial cross vein (1m-cu and 2m-cu) join the 2 nd cubital cell; submedial crossvein (cu-a) meeting medius (Cul) and brachius (lA) almost perpendicularly; anal cell stalked; humeral vein (3A) straight. Hind wing with 2 middle cells, anal cell with long stalk. Inner eye margins slightly converging downwards; distance between the lower eye corners little longer than the maximum eye diameter; clypeus long, shallowly emarginate, in the middle approx. as long as the diameter of a torulus or ca. 1.5 times as long as the distance between the antennal sockets; labrum weakly emarginate on anterior edge; malar space just under half as long as the anterior ocellus; mandibles almost symmetrical, with subapical tooth, in lateral view tapered approximately evenly to the tip. Antenna approx. twice as long as width of head; scape and pedicel distinctly wider than long. Prepectus separated from mesepisternum by a fine line; inner spur of the fore tibia apically divided. Claws bifid, without basal thickening; inner and outer tooth approx. the same thickness, inner tooth slightly shorter.
Description. This is based on a translation of Taeger (1989), augmented with data gained from examination of specimens which have only recently become available. Body length: female 8.0 mm, male 6.5 mm. Female (Fig. 119) and male (Fig. 120) are similar in colour, apart from the mesopleura: upper mesepisternum pale in female, entirely dark in male. Head and antenna black, except for pale palps and labrum. Thorax dorsally black, with pale tegula and more or less pronotum. Legs entirely pale except more or less for tarsomeres. Wing venation entirely pale brown. Abdomen yellow except more or less for tergum 1. Antennomere 3 little shorter than 4. Postocellar field ca. twice as wide as long; ocellus diameter : POL : OOL = 1 : 1.7 : 2.0; frontal field enclosed by indistinct bulges; supra-antennal groove indistinct; head weakly punctured and shiny; frontal field partly finely wrinkled; thorax slightly more strongly punctured than head. Mesepisternum shiny, with indistinct punctures, evenly covered with rather dense, pale pubescence. Legs relatively thick: femora 3.5 times as long as wide, 0.66 times as long as the tibia; tibia 6.5 times as long as wide and 1.2 times as long as the metatarsus; inner spur of the metatibia nearly as long as the apical width of tibia.
Similar species. In the West Palaearctic, Mesoneura opaca and lanigera are superficially similar in habitus to Neodineura arquata.
Life history. Unknown. Distribution. Only known from Germany, Switzerland , the Czech Republic (Beneš and Holuša 2015), and the Russian Caucasus (see below). We are only aware of the existence of four extant collection specimens: three females and one male. Taeger (1989) interpreted the handwritten label data on the only known male (SDEI) as "Sandbg.
[Sandberg] 11.V.91", and thought it likely that the locality was one of several of that name within the then German-speaking territories. Alternatively, it could refer to "Sonderburg" [German name for the Danish island Sønderborg], although the second letter on the label does look more like an "a" than an "o". Konow received many sawfly specimens, some still in the Konow Collection at the SDEI, from W. Wüstnei, who resided at Sonderburg, and collected from around the late 1880's to the early 1900's.

Platycampus Schiødte, 1839
Notes. Two species have been considered to be represented in the West Palaearctic fauna (Taeger et al. 2010): luridiventris (see below), and obscuripes (Konow, 1896). The latter was described from two females collected in the St Gotthard area, Switzerland. Konow (1896) stated in the original description that obscuripes differed from luridiventris in its [translated from German] "much smaller head, the apically more weakly emarginate clypeus, and the somewhat shorter third cubital cell, as well as the dark colour of the body and the legs". Only fragments of one of these specimens now exist. Conde (1937) proposed the synonymy of obscuripes with luridiventris, basing his concept of obscuripes on two female specimens from Piedmont, Italy, leg. Dodero (name of collection not mentioned), and concluded that it is only a dark, alpine form of luridiventris. A further female which may belong to obscuripes, because it has largely black metafemora, was collected in 1954 in Oberstdorf, Bavaria, by E. Enslin (Manfred Kraus Private Collection). Finally, Weiffenbach (1975) stated that he reared a female obscuripes collected on Alnus viridis, from Montafon, western Austria, 1800 m. Normally coloured specimens of luridiventris are known to occur on Alnus viridis, at lower altitudes, in Central Europe (see below). The status of obscuripes requires reassessment, preferably including the use of genetic data.

Platycampus luridiventris (Fallén, 1808)
Tenthredo alnicola Bechstein & Scharfenberg, 1805: 867 (Herbst and Heitland 1994), and that the different hosts correlated with differences in behaviour (Heitland and Pschorn-Walcher 2005), and partly in the morphology of larvae (Heitland and Pschorn-Walcher 1992): setae on the head and body of larvae from glutinosa tended to be shorter than of those from incana, but setae of larvae from viridis usually did not differ from those on glutinosa. Our genetic data based on sequences of four genes contradicts, at least partly, the results of Herbst and Heitland (1994). Although six sequenced larvae collected in three different localities (Lower Austria) from three different Alnus species do segregate based on mitochondrial COI (1078 bp) into three clusters according to the host plant and locality (maximum distance 2.2%), the nuclear sequences (NaK, POL2, TPI: 5017 bp including introns) are practically identical (only four variable / heterozygous positions, giving a maximal pairwise distance of 0.08%), so that the tree structure for P. luridiventris on Fig. 1 is entirely determined by COI. For comparison, nuclear divergence within most other species of Nematinae (based on heterozygous females) is larger, on average 0.2% or up to 1%. In addition, COI sequences of two specimens reared from A. incana from Abisko (DEI-GISHym21133, DEI-GISHym21134) are identical to two larvae collected from A. glutinosa from Lower Austria (DEI-GISHym21496, DEI-GISHym21497). Since different food plant species can affect gene expression differently in feeding larvae (Yu et al. 2016, Orsucci et al. 2018, Okamura et al. 2019, one can speculate that the allozyme analyses by Herbst and Heitland (1994) were influenced more by differences in the expression of the studied proteins (preferential expression of certain alleles or isoforms) than differences in genetics. Morphologically, we noticed conspicuous differences in the overall shape and spacing of the sawteeth, particularly the apical ones, between the reared Swedish specimens (Figs 128-129) and a German specimen belonging to the other barcoding cluster (Fig. 132). However, examination of further specimens revealed wide variability in the shape and spacing of the sawteeth, with several intermediates (e.g., Figs 130-131), so that finally no clear morphological separation of two groups seemed possible. Perhaps this variability is mainly correlated with geographical occurrence, with a tendency in northern specimens to shorter, more projecting teeth: the lancets of two Abisko specimens (Figs 128-129) have the most clearly projecting and shortest sawteeth (with correspondingly long distances between them), while a specimen from southern Sweden (Småland) has long and flat teeth (more closely spaced) (Fig. 131), and a specimen from Central Sweden is intermediate with regard to the shape of the teeth, although they are widely spaced (Fig. 130). In these examples, the differences are not caused by wear of the saw teeth, because the outlines of the teeth are angular and the denticles are clearly differentiated. A highly worn lancet has rounded edges of the teeth, and the denticles are no longer clearly discernible (Fig.  133). Note that apparent differences in the overall curvature of the illustrated lancets are the result of preparation: each annulus of the lamnium can move slightly, relative to its neighbours, and slight differences in the curvature of the whole lamnium are thus mostly artefacts resulting from preparation. In the light of the foregoing considerations, we conclude that although the three segregates could perhaps be considered to be host plant races ["foodplant races"], as already suggested by Heitland and Pschorn-Walcher (2005), they should certainly not be accorded a formal nomenclatural status. Description. Body length: female 5.0-7.0 mm, male 4.5-6.0 mm. Female: head black except for palps, and more or less labrum, underside of antennal flagellum, and sometimes more or less scape and pedicel. Thorax black, except for yellow tegula and more or less posteriodorsal edges of pronotum. Sometimes lateral edges of median mesoscutal lobe, and upper mesepisternum pale. Legs pale (orange), with dark metatarsus and apex of metatibia, and more or less dark bases of coxae. Wing venation mostly brown, with centre of fore wing stigma paler. Cerci pale; rest of abdomen from completely black except for obscurely brown area of hypopygium, to all sterna bright yellow, sometimes also with yellow on downturned lateral edges of terga. One reared female from Abisko has dorsal parts of terga 2-4 pale. Variability in the shape of the teeth of the lancet is considerable (Figs 128-133): see also under Taxonomy above. Male: colour similar to female, but pronotum entirely black. Sternum 9 black to pale. Harpes more or less pale. Similar species. If the nearly complete loop formed by the curved up base of fore wing vein 2A+3A in Platycampus is overlooked, then it might be mistaken for Stauronematus platycerus, which is similarly coloured and also has bifid claws (but with an additional basal lobe not found in Platycampus), or perhaps a Pristiphora species.
Life history. Host plants: Alnus glutinosa, incana, and viridis (Heitland and Pschorn-Walcher 1992). Mentions by Lorenz and Kraus (1957) of Betula, Corylus avellana and Rubus as hosts of luridiventris are likely to have been based on misidentifications (Zinovjev 1986, Heitland andPschorn-Walcher 1992). A strictly univoltine species, although some populations exhibit polymodal emergence patterns. Correlated with its highly distinctive larval morphology (Figs 72-73) compared to other nematine genera (Boevé and Angeli 2010), Platycampus luridiventris has many peculiar behavioural traits, such as the extremely long time, of approximately three months, taken by the larva to mature (Heitland and Pschorn-Walcher 2005). Oviposition is into the leaf petiole or midrib, with a maximum of three eggs per leaf. The larva is crepuscular according to Heitland and Pschorn-Walcher (2005), and feeds only for very short periods, making holes in the leaf blade, and during the day is normally found immobile on the leaf underside, often in an angle between the midrib and a lateral vein. Sex ratio appears to be normal for netted specimens, i.e., males about as abundant as females, but is heavily skewed towards males in material collected with Malaise traps.
Distribution. Widespread in Europe, from the British Isles to the Balkans, and north to Norway and Finland . Earlier published records of luridiventris from the East Palaearctic and Oriental Realms, such as by Benson (1963) from Sichuan, China, probably often refer to other species (Zinovjev 1986 Notes. As already suggested by Prous et al. (2017), Dinematus krausi probably belongs to the Pristiphora depressa species group: see also comments under the species name, below. One of the main reasons for the erection of a genus separate from Pristiphora for krausi, was the presence of vein 2r-rs in the right fore wing of the holotype (this vein absent in the left wing). The presence of this vein in Pristiphora is rather rare but has been observed in at least four other West Palaearctic species: helvetica (Benson 1960b), malaisei, robusta, and staudingeri (Prous et al. 2014. Within Pristiphora, these species are only distantly related. In our opinion, no characters exist which will reliably distinguish Dinematus from Pristiphora, and we therefore propose their synonymy. For further synonymy of genus group names with Pristiphora see Taeger et al. (2010) but note that Stauronematus is now considered to be a separate genus (Prous et al. 2014).
The north-west European species groups and the majority of species of Pristiphora were recently revised by Prous et al. (2016Prous et al. ( , 2017Prous et al. ( , 2018. Pristiphora krausi (Lacourt, 2006)  Notes. Pristiphora krausi is only known from the holotype. Its character combination of bifid claws, in dorsal view short and emarginate valvula 3, and yellow and black colour pattern of head and body, suggest that it may belong to the Pristiphora depressa group ). On the other hand, other currently known female specimens of this group have a mostly dark forewing vein C and pterostigma, whereas these are entirely pale in krausi. Furthermore, the distal sawteeth of krausi are prominently lobed, and markedly flatter in the other species. Pristiphora ifranensis Lacourt, 1973, only known from the male holotype (private collection of Thierry Noblecourt, examined), type locality Ifrane (Morocco, Middle Atlas), resembles krausi strongly in coloration, including its pale forewing vein C and pterostigma. Based on its penis valve morphology, ifranensis has been placed in the depressa group ). If further specimens become available for study, the possibility should be borne in mind that krausi and ifranensis represent the female and male of the same species.
Pristiphora malaisei (Lindqvist, 1952) Notes. A single larva was obtained in northern Sweden by combing through the leaves of an isolated clump of Dryas octopetala, under which an inverted frisbee was held. The plant was growing on an otherwise bare patch of soil at the edge of a road. Gene sequences of the larva are nearly identical to those of Pristiphora malaisei imagines collected in the same area. Although the specimen (Fig. 74) is small (approx. total length 3 mm), and has been conserved in 96% ethanol, it seems to resemble the larva of P. dasiphorae as described by Zinovjev (1993) much more closely than the larva of P. malaisei (see Fig. 86 134); b Abdomen entirely black; c Mesepisternum more densely pubescent above than below but without extensive entirely glabrous area on lower half (Fig.  134); d Hind coxa with at least basal half black (Fig. 134); e Wing membrane hyaline; f Lancet with ca. 19 teeth (Fig. 136); g Penisvalve with ventral margin of paravalva not emarginate (Fig. 138) (Fig. 135); bb Abdomen apically more or less pale: in ♀ at least hypopygial area pale brown, sometimes abdomen medially completely pale (yellow); in ♂ subgenital plate and harpes brown; cc Mesepisternum with an extensive glabrous area on lower half (Fig. 135); dd Hind coxa with only extreme base black (Fig. 135); ee Wing membrane slightly infuscate; ff Lancet with ca. 16 teeth (Fig. 137); gg Penisvalve with ventral margin of paravalva emarginate (Fig. 139) van Vollenhoven, 1858: 191-194, pl. 12. Lectotype ♀, exam-ined, designated by Thomas (1987: 72), in RMNH. Type locality: Leiden (Netherlands). Synonymy with Nematus compressicornis auct. by Cameron (1878: 267 Description. Body length: female 4.5-7.5 mm, male 4.5-6.0 mm. Head black, except for mandibles and palpi. Pronotum completely black, or only extreme upper and rear edges brown. Mesepisternum more densely pubescent above than below but usually without entirely glabrous area on lower half. Hind coxa with at least basal half black. Trochanters and femora completely pale (yellowish). Tibia more whitish: proand mesotibia and pro-and mesobasitarsus entirely pale, with rest of tarsus darkened. Metatibia with approx. apical third black but spurs pale. Metatarsus black. Wing membrane hyaline; venation largely pale except for dark fore wing stigma. Abdomen entirely black. Female: head in dorsal view subparallel behind eyes. Antennae normal; not laterally compressed. Cerci pale to dark. Lancet: Fig. 136. Male: head in dorsal view behind eyes only slightly contracted. Antennae strongly laterally compressed, flagellomeres ventrally somewhat produced; may be reddish. Penis valve: Fig. 138. Similar species. When the shape of the claw is overlooked, Stauronematus adults are frequently misidentified as Pristiphora. The long, thin cerci of female Stauronematus, and the shape of the valvula 3 in dorsal view, are however quite different to any West Palaearctic Pristiphora species.
Life history. Host plants: mainly Populus spp., especially tremula, but also nigra, balsamifera, deltoides, alba, and many cultivated forms (Pschorn-Walcher and Altenhofer 2000, Brischke 1884, Cavalcaselle 1968); less often on Salix purpurea (Pschorn-Walcher and Altenhofer 2000, our own observations). Frequently recorded as bivoltine, but possibly has even three generations in warmer areas. Sex ratio appears to be normal for netted specimens, i.e., males about as abundant as females, but is heavily skewed towards males in material collected with Malaise traps. Oviposition in a double row in the leaf petiole. The larvae eat holes in the leaf blade and surround the feeding site with "palisades" (Fig. 85) made of a dried secretion produced in their mandibular glands.
Distribution. Found through much of continental Europe, from the Iberian Peninsula and Balkans, to Finland and Norway, and also the British mainland ). According to Sundukov (2017) also occurs in Caucasus, Turkey, Iran, Kyrgyzstan, Kazakhstan, China, Korean Peninsula, and Japan.

Discussion
The conclusions on the phylogeny of Nematinae reached by Niu et al. (2019), based mainly on morphological characters, differ substantially from our results, which are based on molecular data. In our opinion the methodology and data analysis on which their results are based are both seriously flawed. Their results are also affected by misinterpretations of previously published work by other researchers, particularly the papers by Nyman et al. (2006) and Prous et al. (2014). Niu et al. (2019) failed to mention that many of the deepest splits within Nematinae were poorly supported (low statistical support and conflicting relationships in different analyses), although this was acknowledged by both Nyman et al. (2006) and Prous et al. (2014). At the same time, monophyly of Nematinae (including "Hoplocampinae") was strongly supported in all analyses. In the absence of clear evidence to the contrary, there is no justification for the proposal of alternative classifications: Niu et al. (2019) have not provided such evidence, because they rely solely on the classification proposed by Wei and Nie (1998). Wei and Nie (1998) claimed that their "cladistic analysis" of "Tenthredinoidea" (i.e., Tenthredinidae as currently understood) was based on a "…huge data matrix", but that "…the complicated analysis process are omitted here for limited space and they will be reported in detail in a separated monograph." We are unaware of any sources or publications which provide these data. Wei and Nie (1998) basically elevated many existing taxa to higher rank (tribes to subfamilies, subfamilies to families etc.) with little or no increase in information content. In the absence of publicly available evidence, we are sceptical that Wei and Nie (1998) managed to create a highly informative morphological data matrix that could be used to propose a well-supported and stable phylogeny of Tenthredinidae. The cladistic analyses by Vilhelmsen (2015), based on 146 morphological characters, demonstrate how difficult it is using such methods to achieve a high level of statistical support and stability for phylogenies within Tenthredinidae. At the same time, the statement by Niu et al. (2019: page [2]) that the results of Prous et al. (2014) were based "only on 400-bp sequences of the barcode region", is simply wrong. As clearly described in Prous et al. (2014: 3) there were two datasets based on four genes (two mitochondrial and two nuclear), one of them (134 specimens) with little missing data (19 specimens missing one gene and seven specimens missing two genes) and the second one (79 specimens) with more missing data (21 specimens missing one gene, eight specimens missing two genes, and 15 specimens missing three genes). This approach was adopted so that type species of some genera for which only one gene was available could be included in the analyses (only one specimen in the second dataset had 422 bp of COI, all others had at least 658 bp of COI). In the end, the new data presented by Niu et al. (2019) are irrelevant to their discussion on the classification of the Nematinae, because of completely inadequate taxon sampling: they analysed only two specimens of Nematinae. Their data are in fact consistent with all previously proposed classifications, not just with Wei and Nie (1998) as they stated.
Although the Nematini and Dineurini both comprise a relatively large number of genera, the large majority of Holarctic nematine species belong to just two genera of Nematini, Euura and Pristiphora. The proportional representation of genera and species in the Oriental Realm is at present unclear, but compared to the Holarctic Realm, existing data point to a lesser number of Euura species, and more Pristiphora, while the number of species belonging to diverse genera of non-Nematini may also be greater (Taeger et al. 2010). At the same time, although the number of still undescribed nematine species inhabiting the mountains of the Oriental Region can only be guessed at, it seems unlikely that Nematinae make up such a high proportion of the Oriental sawfly fauna as of the fauna of northern regions of the Holarctic. Outside the Holarctic and Oriental Realms, the Nematinae is represented naturally only in the northern regions of the Neotropical Realm, by a few species of Pristiphora (Taeger et al. 2010).
As noted above, the striking abundance and species diversity of nematine sawflies in the northern parts of the Palaearctic, including Fennoscandia, results mainly from the presence of numerous species of Euura and Pristiphora. Although several factors probably contribute to this pattern (Bogacheva 1994, Kouki et al. 1994, it has long been apparent that at progressively high latitudes in the northern hemisphere Salix species are of increasing importance over other plant taxa as hosts of sawflies, particularly Nematinae (Malaise 1931b). On the other hand, it is important to remember that many other plant taxa are hosts of sawfly larvae in the north. An example is our indication that Dryas octopetala is a host plant of Pristiphora malaisei in the more northern and upland parts of the range of this sawfly species. Currently, this is only the second sawfly species to have been found on this host, the other being the al-lantine Empria alpina Benson (Prous et al. 2011). However, based partly on our own experiences during field-work, we suspect that the relative difficulty of collecting larvae from low-growing potential hosts such as Dryas, other herbaceous Rosaceae, Polygonaceae, Fabaceae, grasses and sedges, etc. as opposed to shrubby Salix, may have led to at least a slight underestimation of the significance of the former as host plants in the northern nematine fauna. Furthermore, although Betula species are clearly the second most frequently used hosts of Nematinae in northern Fennoscandia, most published observations and data are for the tree-birch Betula pubescens var. pumila (e.g., Tenow 1963), whereas surprisingly little has been published about the sawfly fauna of Betula nana.
As can be seen from the key to larvae, the larvae of Nematinae exhibit a high level of morphological variability. This is expressed, for example, in the number of dorsal annulets of abdomen segments varying between three and six. By contrast, all European Tenthredininae larvae have seven annulets, six in Selandriinae [only Dolerus] or seven, six in each Athaliinae and Allantinae (Lorenz and Kraus 1957). Only among the Blennocampinae is this character similar in variability to the Nematinae: Blennocampinae have 4-6 annulets, excluding the leaf-mining taxa, in which the number is reduced to two. The variability in Nematinae is all the more remarkable because conspicuous differences such as the number of annulets apparently occur even between species which are certainly quite closely related, such as within the Pristiphora malaisei species group. In the Blennocampinae, differences in the number of annulets are usually regarded as generic characters (Lorenz and Kraus 1957).
Although the genera which we have treated in this paper are comparatively speciespoor, cases nevertheless occur of the sort of taxonomic problems which are regularly encountered in the much larger genera Pristiphora and Euura. An interesting example is Platycampus luridiventris, where three different (mitochondrial) genetic lineages exist. Earlier studies on this species concluded that genetic segregation was correlated with differences in host plant use, behaviour, and partly even the length of setae of larvae. Our own genetic data partly conflicts with this conclusion. Perhaps the apparent differences are caused by differential gene expression: a sort of host plant conditioning. At present, there are no compelling reasons to treat the lineages as separate taxonomic entities. A similar situation may occur in several groups of closely related nominal species of Euura, such as the gall-makers of the dolichura group and oblita group (ischnocera complex), which are thought to be highly host specific, but often exhibit neither clear morphological nor genetic differences .