Seasonal head dimorphism and taxonomy of some European species of Aprosthema

Die Kopfform wurde bislang als wichtiges Merkmal fur die Unterscheidung der Arten des Sterictiphorinen-Genus Aprosthema Konow (Argidae) angesehen. Eiablage-Experimente mit sudfinnischen Weibchen von Aprosthema melanurum (Klug, 1814) an Lathyrus pratensis L. (Fabaceae) zeigten, dass uberwinterte Weibchen (Fruhlingsgeneration) einen langen Kopf mit parallelen oder erweiterten Schlafen und ein kurzeres Flagellum haben, wahrend die Weibchen (und Mannchen) der Sommergeneration stark verengte Schlafen und ein langeres Flagellum aufweisen. Die kraftigen Winterkokons befinden sich in der Erde. Die dunnen, schwach gebauten Sommerkokons hingegen werden an der Futterpflanze oder einer nebenstehenden anderen Pflanze gebildet. Zum Offnen dieser Kokons benotigen die Imagines keine starken Muskeln. Ein ahnlicher Dimorphismus der Kopfform wurde auch bei den zwei Generationen des an Vicia cracca L. (Fabaceae) lebenden Aprosthema fusicorne (Thomson, 1871) sowie dem an Lathyrus vernus (L.) Bernh. lebenden A. intermedium (Zaddach, 1864) [= A. hyalinopterum Conde, 1934, syn. nov.)] nachgewiesen. Es handelt sich hierbei um die ersten Nachweise von saisonalem Kopfdimorphismus bei Pflanzenwespen. Lectotypen wurden fur Hylotoma melanura Klug und Aprosthema hyalinopterum Conde festgelegt. Ein Bestimmungsschlussel fur die 5 untersuchten Aprosthema-Arten wird vorgelegt. 3 Aprosthema-Arten werden fur England nachgewiesen.

Seasonal head dimorphism and taxonomy of some European species of Aprosthema (H y m e n o p te ra : S y m p h y ta : A r g id a e ) V e l i V i i c b e r g With 11 figures S u m m ary Head shape has been regarded as an important character distinguishing species in the sterictiphorine genus Aprosthema KONOW (Argidae).Oviposition experiments on southern Finnish females of A. melanurumlKLUC, 1814) on Lathyrus pratensis L. (Fabaceae) showed that the overwintered female (spring generation) has a long head with temples parallel or broadening posteriorly and a shorter flagellum, whereas the female (and male) of the summer generation has a short head with strongly narrowing temples and a longer flagellum.The thick winter cocoons are in the ground.The thin soft summer cocoons are formed on the food plant or some other nearby plant, and the emerging adult does not need strong muscles to open the cocoon.A similar head shape dimorphism is found in the two generations of Aprosthemapusicorne (THOMSON, 1871) feeding on Vicia cracca L. (Fabaceae), and of A. intermedium (ZADDACH, 1864) = A. hyalinopteruen CONDE, 1934, syn.nov.feeding on Lathyrus vernus (L.) BERNH.These are the first observations of seasonal head dimorphism in sawflies.Lectotypes are designated for Hylotoma melanura KLUG and Aprosthema hyalinopterum CONDE.A key is presented to the 5 species of Aprosthema examined.Three species of Aprosthema are reported from England.

Introduction
The species of Aprosthema are mostly rare and seldom collected (ENSLIN 1917), and many of them have been described on the basis of one or few specimens.The biology of rather few species is known but all reared species have larvae feeding on leguminous plants (Fabaceae).The head form has been regarded as an important character to distinguish between the species of the genus (THOMSON 1871, KONOW 1892, ENSLIN 1917, CONDE 1934a, GUSSAKOVSKIJ 1935, ZHELOCHOVTSEV 1988, and VlITASAARI 1990).However, this paper aims to show that the head form is different between the specimens of the spring or overwintered and summer generations of the same species in three European species.Notes on the taxonomy of some European species of Aprosthema are given.
Schispcerafusicornis THOMSON, 1871:46, 9. Syntypes: one female from Oland, Sweden found by Prof. Boheman.(The male mentioned in a Swedish note from Jemtland is not a syntype because its identity remained unclear to the author).Holotype female was found among the specimens in Naturhistoriska Riksmuseet, Stockholm, and it is labelled: 1) Oel; 2) Bhn; 3) Schispcera fusicornis [in Thomson's handwriting]; 4) Holotype 9 Aprosthema fusicorne (THOMSON, 1871), spring form; V. VlKBERG 2001; 5) Naturhistoriska Riksmuseet, Stock holm Loan no 22/99.CONDE, 1934a: 22 (key), 25-28, 9 < 5 larva.Adult syntypes: Latvia, North of Viesite, in the state forest near Gerkani, 1 d 1 9 28.5.1933, and one further fe male reared by letting the above female lay eggs on the leaflets of ILathyrus vernus (L.) BERNH., the female emerged on 4 Feb, 1934.The reared female (in coll.DEI, Eberswalde) was examined: it had been eaten by dermestid larva, and only the left fore wing, tegula, pronotum and most of the mesonotum are left.As the only available syntype it is designated hereby as the lectotype of Aprosthema hyalinopterum Another ovipositing experiment (8b/94): one male and 4 females were reared from larvae on Lathyrus on 9.7.-10.7.All were put into large vessel with Lathyrus.On 10.7.11.7.females laid eggs through the margin of leaflet.One egg was laid singly, after that the female flew away from the food plant.Copula was not observed.The larvae grew well until 19.7 indoors, when fourth instar larvae were put into a net on a Lathyrus bush outside.On 25.7 they were inspected when I returned from Lapland.21 full or almost full grown larvae were found but workers of Lasius niger L. had attacked them and injured them by biting.The larvae were put into alcohol.

Aprosthema hyalinopterum
The ovipositing experiments injanakkala: on 15.5.2000 a female was found on Lathyrus pratensis.She laid eggs on leaflets of Lathyrus, laying one egg took 55 seconds.25-30 larvae emerged on 19.5.-20.5, they eat on the leaflet margin keeping body s-shaped.The second instar larvae got some acute disease and all died.They were briefly boiled and put into alcohol.
Next year I visited the same place on 27 May, and swept simultaneously two males and two females of A. melanurum on Lathyrus pratensis.The females with strong head form were used for ovipositing experiments (no.6/01 and 7/01).On 27.5.-28.5.Vida cracca-mA Lathyrus pratensis were offered to each of them.
Both laid eggs only on margins of leaflets and stipules of Lathyrus.Larvae hatched on 31 May, they moulted four times and part of larvae five times.On 9 June first small cocoons were found on the food plant.Males emerged from these on 16-17 June.Altogether 27 and 8 cocoons were found on 15 June.Female no.6 gave 5 males and 20 females and female no.7 gave 6 females, all progeny had the short weak head form.
Specimens of two generations of Aprosthe?nafusicorne(THOMSON) from Finland: Palkane (Ta) (mostly 680:35) were studied and measured.The biology of the species has been studied there for many years by Jaakko K. Kangas and his collaborators (KANGAS 1985; as A. melanura and At.fusicornis).Specimens with both strong and weak head form of the spe cies have been reared from larvae on Vicia cracca L. (Fabaceae).The specimens with a strong head form occur earlier in the season in the same places where the specimens with a weak head form are found later.All available specimens from Palkane were studied and measured.Flight periods are as follows: Females with stronger head (n=l 1) were captured in June, except one on 27 May, 29 May, 3 July and 14 July each.Males with strong head (n=7) were captured on 1.6.-17.6.Females with weaker head were found in the last half of July, except one on 14 July and one on 2 August.The males with weaker head were found on 15-24 July.Only one ovipositing experiment (no.3/86 JK) has been carried out: the female with strong head found on 12.6.1986laid eggs singly on the leaflets of Vida cracca (she did not lay eggs on the offered Lathyrus pratensis), the prepupae overwintered and two males emerged in March 1987 (these males have a furrow on lateral vertex, but rather short, narrowing temples).Although there is no direct confirmation by an ovipositing experi Beitr.Ent.54 ( 2004) 1 ment that the spring and summer generations belong to the same species, all indirect evidence strongly suggests it, especially when compared with the similar proven situation in Aprosthema melanurum.Further evidence is presented in Results (the similar structure of lancets and penis valves).
The sawfly specimens and dry prepared larvae were studied under a Leitz stereomicroscope at magnifications of 50x and lOOx; the light source was a halogen lamp of 12V 20W with a light beam width of 10 degrees.The light was diffused by placing a piece of thick (0.2 mm) tracing acetate near the specimen.Measurements and drawings were made using a grid of squares (50x50, side 0.20 mm) in one eyepiece.Some larvae in alcohol were treated with KOH, and dissected body parts were put into polyvtol 17 under a cover glass on a slide.The slide was examined with a compound microscope at magnifications up to 450x; this made it possible to see very small setae (microsetae) and glandubae.Body part nomenclature is according to HUBER & Sharkey (1993) and V iitasaari (2002).The length of the body was measured from the anterior irons to the apex of the abdomen, including the sawsheath of the female or hypopygium of the male.The length of the fore wing was measured from the humeral plate at the margin of the tegula to the apex of the wing (the apical part of the wing was raised with an insect pin to the same plane if needed), and the length of the costa from the same origin to the apex of the costa.The head width was measured as the distance between the lateral margins of the compound eyes.The head height was measured as the distance between the lower margin of the clypeus and the uppermost hind margin of the postocellar area.The head width at the temples was measured in dorsal view as the maximum width behind the compound eyes.The length of the female flagellum was measured as the maximum length (the basal neck-like constriction was not included), and the maximum width of flagellum was measured in lateral view, perpendicular to the longitudinal axis of the segment.The length of the male flagellum was measured as the maximum length from the basal arch of the anterolateral projection to the apex of the flagellum.

R esults
The results of the measurements of Aprosthema melanurum 2.rAA.fuskorne are presented in Tables 1-3.Table 1 shows measurements indicative of the size of the specimens.The body length and fore wing length are most commonly used in previous published studies.Two additional measurements are presented: length of costa of fore wing and the head width at eyes.They can be measured accurately and easily.The body length is the most inaccurate measurement of the four (position of head varies and extension of abdomen varies).Apical part of wing is often bowed below at the base of pterostigma.Costa is a strong straight vein.Head width can be measured most accurately.This is shown in percentual variation observed (mean =100 %), e.g. in A.fuskorne females, summer genera tion: body from -18.4 to + 9.8, wing from -7.6 to + 10.6, costa from -5.8 to + 8.0, and head width from -2.9 to + 7.6.If head width values of different samples are compared, follow ing results are obvious: spring generations have a slightly larger head than summer genera tions, females have a slightly larger head than males, and in A.fusicorne the mean and maxi mum are a little larger than in A. melanurum.The summer generation samples contained more or consisted entirely of reared specimens which are usually smaller than those cap tured in the wild; this could explain the size difference between the generations.The head shape characters are presented in Table 2.The spring generation has a stronger head than the summer generation: in dorsal view the temples are longer and broader be hind the eyes, they broaden posteriad or are parallel or narrow only slightly.As an expres sion of the stronger muscles of the head the postocellar area is also longer and more convex than in summer generation, and there is a shallow furrow from near the lateral ocellus to the compound eye of the same side.In the summer generation the lateral vertex and upper lateral irons are entirely even, without any furrow.The longer postocellar area is reflected in longer head height values and larger head height/head width index values in the spring generations.The index values do not overlap or overlap only slightly.In spring generations the head width at the temples/head width at eyes index has slighdy larger values than in corresponding summer generations.The heads of several specimens are figured in dorsal view to show presence or absence of furrow on lateral vertex and form of temples (Figs 1-2).If possible, the heads of a mother sawfly and her offspring are figured.The characters of flagella are presented in Table 3.The female flagellum is undivided, more or less spindle shaped, covered all over with rather short hairs.Anterior margin is slightly keeled.The flagellum of the spring generation is shorter than that of the summer generation, and in A.fusicorne the flagellum is shorter than in the corresponding generation of A. melanurum.This difference is most distinct in flagellum/head width index.The flagellum of the spring generation is thicker than that of the summer generation.
Especially thick and short is the flagellum of the spring generation of A. fusicorne, its small length/width index does not overlap with that of the spring generation of A. melanurum.
The male flagellum is highly modified: it consists of a long, narrow flagellum which has dorsally short adherent hairs to the apex and of a basally attached long anterior or ventral projection.This is apically parallel with the true flagellum but differs in having short adherent hairs only basally.Both the flagellum and its projection are covered with long curved hairs on anterior surface, the lateral hairs are long and medial hairs short.The maximum length of male flagellum is longer in summer generations than in spring generations; the difference is more obvious in flagellum/head width index.The lamnia of the lancets of the saws of Finnish species of Aprosthema were figured by VlITASAARI (1990: Fig. 3 a-f).The general shape of lamnia is very similar.What is w aned.Aprosthema melanurum (KLUG, 1814) in this paper, consisted of two species in VlITASAARI (1990): the spring generation was ca lled A. tardum (KLUG ), and the summer generation A. brevkorne (FALLEN).An examination of the two female specimens of Hylotoma tarda in K l u g 's collection (not syntypes but fitting the original description), and the lectotype female of Hylotoma brevkornis FALLEN revealed that these species are not conspecific w ith Aprosthema melanurum (see Key later).The species which is identified as A. fusicorne (THOMSON, 1871) in this paper was named as two species in VlITASAARI (1990): the spring generation was correctly called A. fusicorne (THOMSON) and the summer generation^, bifidum.There is no specimen of Hylotoma bifida (KLUG, 1834: 244, d ) in KLUG's collection at present, and a neotype is needed to settle its identity.It is now most often regarded as a synonym of A. melanurum which seems a good solution to me.VlITASAARI (1990)  Beitr.Ent. 54 (2004) 1 115 (1990).I studied the radix-parts of one female of each of two generations of Aprosthema melanurum from Liperi and A .fusicom e from Palkane.The base is modified in Aprosthema, and the most basal parts lie along the inside of the hypopygium and thus form a right angle with the tractium.Both generations of A. melanurum have a three-branched (trifurcate) basal part (Fig. 3a) whereas both generations of A. fusicorne have a bifurcate basal part (Fig. 3b).
The penis valves of Finnish species of Aprosthema were figured in VlITASAARI (1 9 9 0 : fusicorne are similar to each other, but clearly different from those of A. melanurum.Thus the shape of penis valve does not correlate with the head form type at all which shows that the head form type (strong or weak) is not a character distinguishing between species.I have drawn the apical part of the penis of both species in situ (A.melanurum-.Fig. 4a and A. fusicorne-.Fig. 4b).
The body colour of the spring and summer forms of Aprosthema melanurum may show big difference.Females of the summer form are often paler than females of the spring form.Hind femur is mostly black in the spring, but its base can be only slightly black or it is entirely yellow in the summer.Terga 2 and 9 lose often their blackish colour in the summer generation.The apex of mesoscutellum and axillae may show reddish brown colour in summer females.The summer males of A. melanurum often have a pale costa and their scape and pedicel can be reddish.Similar differences in body colour can be found be tween the specimens of two generations of A. fusicorne.
The larva of Aprosthema intermedium (Z ADDACH), under the name A. hyalinopterum CONDE, was described in many details by CONDE (1934a).He also briefly described the larva of A. melanurum (KLUG).The chaetotaxy of the third abdominal segment of A. melanurum from Liperi is shown in Fig. 5.The anal tergum and the subanal lobe is shown in Fig. 6a-b.
The larva of A. fusicorne (THOMSON) is described here for the first time.According to KANGAS (1985), and cited by VlITASAARI (1990), itlives in colonies.This is incorrect: itis a solitary feeder, like the larva of A. melanurum.The wrong conclusion was based on the observation that the summer cocoons were found in small clusters on the host plant.I have examined 5 larvae from Palkane, prepared by Erkki O. Peltonen by blowing air into an empty skin in an oven.Two larvae were taken on Vicia cracca by J. Kangas, andprepared on 17.7.1962 and7.7.1971; three larvae are from the ovipositing experiment (3/86 JK).
Description of full grown larva of A.fusicome: Head width 1.48-1.59mm, frons width 0.62 0.67 mm, length of body 16.6-18.9mm (inflated larva).Head pale, brownish (on vertex composed of small brown maculae), coronal stripe darker brownish.Trunk green, cuticle covered with small dark granulae; longitudinal dorsomedial stripe on thoracic segments 1 -2 and abdominal segments 6-9 composed of larger dark granulae.Spiracles brownish black, brown-winged.Head in anterior view roundish, height from anterior margin of clypeus to topmost vertex slightly smaller (0.97) than head width.Key to species of Aprosthema treated in this paper (all species have a spring form with stronger head with shallow furrow on lateral vertex, and shorter flagellum, and a summer form with weaker head having entirely flat lateral vertex, longer flagellum and often paler colouration of the body): 1 (6) Mesoscutellum with rather erect hairs, apical and lateral margins with a hairless zone (Fig. 8).Metascutellum with hairs mosdy on anterior surface, in dorsal view the hairs do not project behind the sclerite (Fig. 8).
2 (3) Female: apical 0.6 of hind tibia infúscate, hind tarsus, except base of basitarsus, infúscate.Terga 1-2 and 9 often blackish.Male: apex of penis valve with a small incisura and lateral margin strongly projecting (Fig. 4a).Larva on LatbyruspratensisV, anal tergum of full grown larva with short medial projections.... A. melanurum (KLUG) 3 (2) Female: at most apical 0.15 of hind tibia and apical 0.20 of hind basitarsomere infúscate, hind tarsomeres 2-3 with pale base.Terga 2 and 9 without blackish colour.Male: apex of penis valve with deep incisura, and lateral margin rounded (Fig. 4b).Anal tergum of full grown larva with long, sharp medial projections.and Turkey there are probably only two species: Aprosthema tarda and A. melanura.This is not true at all.Also the real Aprosthema melanurum (KXUG) has been found in England: I have examined a male (in coll.NHM, London) of the spring generation taken in Oxford(shire), Marston, Cherwell Meadows on 4. VlKBERG, V.: Seasonal head dimorphism in Aprostema found small differences in the details of lamnia of the spring and the summer generations, e.g. the teeth are better developed in summer generation of A. fusicorne (Fig. b as A. bifida) than in spring generation (Fig. f: almost apical half smooth, without teeth).The base of lancet (radix) was not figured by VlITASAARI

Fig. 5 a
-f).The penis valves of the spring generation (Fig. 5 d: as A. tardum) and the summer generation (Fig. 5 a: as A. brevicorne-, also 5 b and 5c: as A. frieser, these are only paler conspecific specimens) of A. melanurum are rather similar.The penis valves of the spring generation (Fig. 5 f) and the summer generation (Fig. 5 e: as A. bifidum) of A.
Trunk cylindrical, maximal width on On anterolateral surface of proleg 3-4 setae.Caudal margin of anal tergum shown in Fig.7a.Subanal lobe with projections shown in Fig.7b.Notes.I communicated about my first rearing and ovipositing observations in Liperi at the monthly meeting of the Entomological Society of Finland on 16 September 1994.VlITASAARI et al. (1998: 142)cited this communication but the human memory can be unreliable: the species which I observed was not A. fusicorm or A. bifidum.The latter name is wrongly mentioned also byVlITASAARI (2002: 172).Later I could examine the reported two Estonian species with the help of Mikk Heidemaa: the species which was named Aprosthema bifidum (K l u g ) is A. fusicorm , and A. melanurum was misidentified as Al.brevicorne (FALLÉN), DOI: 10.21248/contrib.entomol.54.1.107-125 6.1941 by P. M. Miles.It has been studied and labelled by several entomologists: Sterictophora [!] furcata (Villers) < S det.R. B. Benson 1948; Aprosthema sp.det.E Koch 87; Aprosthema fusicorne (Thomson, 1871) det.L. Clemons 1993; Aprosthema d cf.tarda Klug det.C. v. Achterberg 1993.Three more specimens of A. fusicorne (THOMSON) have been found in England: a female of the spring form in Souldrop, Bedfordshire on 24.6.1956 on Vida cracca, leg.V. H. Chambers; and a female and a male of the summer generation in Chippenham Fen, Cambridgeshire, TL 650693, Malaise trap: carr at reed bed edge,6.7.-20.7.1985  (female), 20.7.-1.8.1984 (male), leg.J. Field (all three in coll.RSMNH, Edinburgh).I have studied much material of Aprosthema from Europe and Asia, including most of the types.The seasonal head dimorphism appears to be a regular phenomenon in the genus.This will change the taxonomy of the genus greatly.It is a pity that the rearing experiments of the species are very few.This has prevented us from a better understanding of this interesting genus.The thin, weak summer cocoons constructed on the food plant or other plant nearby appear to be characteristic for Aprosthema.To open them no strong muscles are needed.Obviously this is the reason why during evolution the head form with weak muscles (short postocellar area and short weak temples) of the summer generation has developed.The winter cocoon is in the ground and it is much thicker.Therefore strong muscles are needed when the adult emerges in the spring time.Sometimes the summer generation is passed over; this favours the view that the strong head form is the basic form.This study would not have been possible to complete without the help of the curators of several entomo logical collections.I wish to thank for loan of specimens Anders Albrecht, Alexander Antropov, Stephan Blank, RoyDanielsson, Erich Diller, Fritz Gusenleitner, Bert Gustafsson, Mikk Heidemaa, Lars-AkeJanzon,  Frank Koch, Martti Koponen, Manfred Kraus, Michael Maidl, Rudolf Meier, Matti Nuorteva, Martin Schwarz,  Stefan Schodl, Mark Shaw, Akihiko Shinohara, David Smith, Andreas Taeger, Christine Taylor, Jukka Vauras,  Matti Viitasaari, Alexey Zinovjev, and LajosZombori.My thanks no longer reach Jaakko K. Kangas who died on 4 March, 2003 at the age of 88 years; I dedicate this paper to the memory of my friend since 1959.Andrew Liston and Andreas Taeger critically commented the manuscript and the pictures, and Andrew Liston improved my English language.
A ckno w ledgem en ts