Introduction
EntedonDalman, Reference Dalman1820 (Eulophidae, Entedoninae) comprises small to middle-sized (1.8–8.8 mm long) parasitoid wasps, which are endoparasitoid koinobionts of beetles, chiefly Curculionidae (Bouček & Askew, Reference Bouček and Askew1968; Graham, Reference Graham1971; Schauff, Reference Schauff1988; Askew & Kopelke, Reference Askew and Kopelke1989; Rasplus, Reference Rasplus1990). Some species of the genus, e.g. E. ergias Walker, an egg-larval parasioid of bark beetles, have been used in biological control programmes (Peck, Reference Peck1963; Beaver, Reference Beaver1966). Many other species may be considered as potential biocontrol agents as a result of their associations with economically important beetles, e.g. parasitoids of bean beetles (Chrysomelidae: Bruchinae) in the Afrotropics (Rasplus, Reference Rasplus1990), some unidentified Oriental species associated with pest weevils (Abedin & Quayum, Reference Abedin and Quayum1972), or the bark beetle parasitoids from the E. cioni species group (Schauff, Reference Schauff1988; Askew, Reference Askew1991).
Poor representation in biocontrol projects probably underlies the relatively poor current knowledge of the biology and morphology of the preimaginal stages of Entedon, in comparison with other Chalcidoidea. The most detailed reviews of biology, supplemented by descriptions and drawings of immature stages, were given by Beaver (Reference Beaver1966) for E. ergias and by Fisher (Reference Fisher1970) for E. rumicis Graham, E. pharnus Walker and E. aff. philiscus Walker. Recently Gumovsky (Reference Gumovsky2006) presented a similar review and provided scanning electronic micrographs of larvae of E. sylvestris Szelényi. This revealed many peculiar details of these larvae, necessitating a search for comparative data on other related species.
The E. sparetus species group was initially formalized as the subgenus Megalentedon by Erdös (Reference Erdös1944). Then it was treated as a group of species related to E. sparetus Walker by Graham (Reference Graham1963, Reference Graham1971), which was accepted thereafter (Askew, Reference Askew1992; Gumovsky, Reference Gumovsky1997a). The taxonomic part of this paper continues the series of revisionary reviews of Entedon in various areas of the Palearctic region started by Erdös (Reference Erdös1944) and Graham (Reference Graham1963, Reference Graham1971) and then followed by other authors (Askew, Reference Askew1991; Gumovsky, Reference Gumovsky1998a,Reference Gumovskyb, Reference Gumovsky1999a,Reference Gumovskyb; Gumovsky & Boyadzhiev, Reference Gumovsky and Boyadzhiev2003). The biological observations described below are the result of the author's long term interest in relationships between weevils and their parasitoids.
Materials and methods
Collections
Abbreviations for depositories are as following: BMNH, The Natural History Museum London (UK); CIRAD, Centre de Coopération Internationale en Recherche Agronomique pour le Développement (Montpellier, France); INRA, Centre de Biologie et de Gestion des Populations (Montpellier, France); LUZM, Zoological Museum of Lund University (Lund, Sweden); MNHN, Muséum National d'Histoire Naturelle (Paris, France); PU, the collection the University of Plovdiv ‘Paisii Hilendarski’ (Plovdiv, Bulgaria); RMNH, Nationaal Natuurhistorisch Museum (Leiden, Netherlands); SIZK, Schmalhausen Institute of Zoology of National Academy of Sciences (Kiev, Ukraine); TMA, Természettudományi Múzeum Álattára (Budapest, Hungary); UCRC, Entomology Research Museum, University of California (Riverside, USA); ZISP, Zoological Institute of Russian Academy of Sciences (St Petersburg, Russia).
Terminology used in this paper is taken from Graham (Reference Graham1971) and Gibson (Reference Gibson, Gibson, Huber and Woolley1997) ; main body measurements and abbreviations are explained by Gumovsky & Boyadzhiev (Reference Gumovsky2003). The term ‘metasoma’ is used for a compound of the gaster and the petiole (sensuGibson, Reference Gibson, Gibson, Huber and Woolley1997).
Field studies
Adult parasitoids were observed and collected in the field on their host's food plants during early June to early July in 2003–2005 in Kiev (50°28'N; 30°32'E) (associates of Verbascum spp.) and Kherson oblast of Ukraine (v. Lazurnoe) (46°04'N; 32°29'E) (associates of Onopordum acanthium). Fragments of hosts' food plant stems where oviposition of the parasitoids was recorded were covered immediately after oviposition to prevent parasitism by other parasitoids and confusion of the preimaginal stages. Adult females of parasitoids were captured on the hosts' food plant stems and transferred into temporary transparent containers containing the hosts' food plant parts.
Laboratory studies
Oviposition was observed in the field, and the sites where oviposition took place were marked with black ink. Female parasitoids were also collected in the field and fed with honey:water (1:1) solution in the laboratory. Fragments of stems of Verbascum sp. and Onopordum acanthium with oviposition marks of Lixus cardui and Gymnetron asellus (Gravenhorst), respectively, were exposed to the females of Entedon species kept in Petri dishes. Oviposition was observed at different intervals: some females oviposited soon after capture and being placed in Petri dishes, but occasionally it happened only after one day in captivity.
To study the morphology of the different larval instars of Entedon, weevil eggs and larvae were removed from plant tissue at regular intervals and dissected. Endoparasitoid larvae found were fixed in Bouin's fixing fluid (picric acid (saturated), 37% formaldehyde solution, acetic acid) to keep their original shape, and subsequently washed out in 96–98% ethanol. For comparison of the parasitoid larvae at different stages of their development, some parasitized hosts were kept in a refrigerator (at about +4°C), which postponed development of the larvae.
Imaging
All fixed larvae were kept in 100% (but not molecular-sieved) ethanol for one day and then in 100% molecular-sieved ethanol for maximal dehydration. After absolute ethanol the specimens were dried in a critical point drying apparatus (Polaron, E3000 Series), involving replacement of the ethanol by liquid carbon dioxide (CO2) at temperatures below 10°C followed by heating to 37°C under raised chamber pressure, taking CO2 through its critical point. The minute parasitoid larvae were put into pipette tips of various diameters sealed on the sides with cotton wool plugs, to avoid their loss during drying. The dried parasitoid larvae were transferred to SEM stubs on metallic pins, using static electrical charges to avoid damaging their extremely fragile integuments. Finally the specimens were coated with gold and observed using a scanning electronic microscope LEO 1530VP in the Max-Planck Institute for Metal Research, Stuttgart (MPI).
Field and laboratory video recordings were made using 8 mm VP-A800 Pal Samsung Video Camera, 4.0 Mega Pixels Samsung, Sony DCR-HC30E digital cameras, and Leica IC A Videomodule integrated in a Leica MZ 125 stereomicroscope, using the video grabbing option conducted in Adobe Photoshop 6.0 program through the use of the Falcon\Eagle Frame Grabber. Light microscopy was carried out using an Olympus system microscope CX-41 connected with Olympus digital camera C-4040ZOOM, under Olympus DP-Soft programme (Version 3.2).
Taxonomy
Species group sparetus of the genus Entedon Dalman
Both sexes
Body generally relatively large (up to 8 mm, figs 1, 2A,B); fore tibia with single dorsal pale stripe (fig. 4E,F); other tibiae mostly darkened (just knees and distal ends pale), trochanters darkened; occipital margin sharp, but not carinated; frontal suture (sulcus) absent; anterior margin of clypeus truncate, or very weakly produced (fig. 2C,D); antenna with 3- or 4-segmented funicle (fig. 5); marginal vein of fore wing longer than costal cell; subcosta of submarginal vein of fore wing with 2–4 setae on its dorsal surface; speculum open in all known species (fig. 3); propodeum with spiracular elevations delimited only laterally, their projections dull; lateral propodeal sulcus incomplete; submedian areas flat, often with weak to distinct reticulation (fig. 2F); metasoma generally longer than head plus mesosoma, petiole reduced to a narrow band (figs 1, 2A,B).
Entedon zerovae, female, habitus, dorsal view.
External morphology: A, Entedon tobiasi, female; B–F, E. sparetus, female. A, B, habitus, lateral view; C, head, frontal view; D, face, frontal view; E, distal part of fore tibia and fore basitarsus; F, posterior part of mesosoma; cly, clypeus, ms, malar space; white arrow, lamellate projection on anterior surface of fore tibia; black arrow, propodeal spiracular tubercle.
Fore wing. A, Entedon sparetus; B, E. thomsonianus; C, E. lucasi.
A–D, apical margin of fore wing: A, B, Entedon sparetus, female; C, D, E. thomsonianus, female. E, F, fore leg of E. sparetus, female; black arrow, setae along the apical margin of fore wing; white arrow, longitudinal pale stripe on anterior surface of fore tibia.
Antennae: A–F. females; G–J, males. A, G, Entedon sparetus; B, E. tobiasi; C, I, E. cardui; CU, specimens reared from flowerheads of Cirsium ucrainicum in Ukraine; P, specimens reared from Pinus branches in France; D, E. lucasi; E, J, E. thomsonianus; F, H, E. zerovae.
Males
Gaster without a pale subbasal spot.
Comparative results
The species of the sparetus group are similar to the species of the kerteszi group in having a single longitudinal pale stripe on the dorsal edge of fore tibia (fig. 4E,F), but differ in having the anterior margin of clypeus (fig. 2C,D) truncate (produced in species of the kerteszi group). Other species of the genus that have the clypeus with truncate anterior margin (assigned to the hercyna, squamosus and other groups) differ from the members of the sparetus group in that their tibiae are either nearly evenly darkened, or bear two longitudinal stripes.
Biology and hosts
Species of this group for which the biology is known (see below) are egg-larval endoparasitoids of weevils (genera Lixus, Gymnetron). The association of the species of this group with weevils of the genera Mecinus and Larinus are also trustworthy, whereas other host records (e.g. beetles from the families Buprestidae and Ernobiidae) require confirmation.
Key to the species of sparetus group of Entedon
Entedon sparetus Walker (figs 2B–F, 3A, 4A,B,E,F, 5A,G)
Entedon sparetusWalker, Reference Walker1846: 182.
Entedon (Megalentedon) insignisErdös, Reference Erdös1944: 29; syn. by Gumovsky & Boyadzhiev, Reference Gumovsky2003.
Entedon lixiErdös, Reference Erdös1951: 220; syn. by Gumovsky & Boyadzhiev, Reference Gumovsky and Boyadzhiev2003.
Entedon meciniAskew, Reference Askew1992: 122; syn. by Gumovsky & Boyadzhiev, Reference Gumovsky and Boyadzhiev2003.
Diagnosis
Apical margin of fore wing with a fringe of setae; metasoma of female 1.1–1.4 times as long as head plus mesosoma, 1.5–3.7 times as long as broad; eye height 2.1–2.4 times as long as malar space; fore wing 2.1–2.3 times as long as broad; hind tarsus 0.7–0.8 times shorter than hind tibia, fore and mid tarsi about 0.8 times as long as their tibiae; female: antennal scape 5.8–6.6 times as long as broad; pedicel about twice as long as broad, F1 2.3–3.0, F2 1.4–1.7, F3 1.0–1.2 times as long as broad, clava two-segmented, 1.8–2.1 times as long as broad; male: antennal scape 2.5–3.7 times as long as broad; pedicel about 1.1–1.7 as long as broad, F1 2.6–3.3, F2 1.3–1.6, F3 1.0–1.5, F4 1.0–1.8, F5 2.0–3.0 times as long as broad.
Distribution
Throughout in Europe (Bouček & Askew, Reference Bouček and Askew1968; Gumovsky & Boyadzhiev, Reference Gumovsky and Boyadzhiev2003), new for Ukraine, Georgia, Morocco, Kyrgyzstan, Russian Far East.
Hosts
Weevils (Gymnetron spp., Larinus spp., Lixus spp., etc.) in stems of species of Plantago, Cirsium, Carduus (Erdös, Reference Erdös1951; Bouček & Askew, Reference Bouček and Askew1968; Gaham, 1971), Verbascum (here) (Curculionidae).
Material examined
Small variety (3.5–4.5 mm): 6♀, France, Hte Corse, PN Scandola, 25.VI.90 (Rasplus) (INRA); ♀, Morocco, Tanger, Route Cap Spartel, 25.V.1992 (Delvare) (CIRAD).
Mid variety (5.0–6.0 mm): 5♀, 8♂, Moldova, Chisinau, airport, reared from stems of Verbascum sp., 4.04.1984 (Fursov); 3♀, 2♂, Ukraine, Kiev, Trukhaniv Island, reared from larvae of Gymnetron asellus in stems of Verbascum sp., coll. 06.04.1996, reared 19.04.1996 (Gumovsky); 5♀, 4♂, Ukraine, Kiev, Park “Syretsky”, reared from larvae of Gymnetron asellus in stems of Verbascum sp., coll. 03.04.1994, reared 05.05.1996 (Gumovsky) (SIZK).
Larger variety (6.0 mm and more): ♀, “♀”, “1618”, Schmiedeknecht “E. squamosus Thomson”; ♀, Czech Republic, Bohemia c.:údolí, Neutonice-Kováry, 6.VI.1953 (Bouček) (SIZK); ♂, France, Alpes, Maritimes, 1900 m, Entraunes, Esteng, 19.VII.1992 (Delvare); ♂, France, Hérault, St Guilhem Desert, 31.V.1990 (Delvare); ♂, France, Cahors, 8.VI.1989 (Tussac) (CIRAD); ♀ France, Hte Corse, PN Scandola, 25.VI.90 (Rasplus); ♀, France, Aveyron, Nant, 12.07.1991 (Rasplus); ♂, France, Htes Pyr., Vallée Marcadau, 07.08.1991 (Rasplus); ♂, France, Yvelines, Bazainville, 2.VI.1991 (Rasplus) (INRA); ♂, Georgia, Lagodekhi, 27.IV.1951 (Bagvadze) (ZISP); ♀, Hungary, Mecsek h., 36.6.1951 (Erdös) (SIZK); ♀, Kyrgyzstan, N 48°48.7', E 71°08.4', Chandalash ridge h=1860 m, Chun-Kurchak lakes, 28.7 km from Jangy-Bazar, 30.06.1998 (Korneyev & Kameneva) (SIZK); 3♀, Russia, Volga region, “Sarepta, Bekker”; ♀, ibid., Sarepta, 16.V.1917, collection of Kuznetsov-Ugamsky; ♀, Russia, Volga region, “Lukash.” 1.V.1911; ♀, Russia, Altai, Leb'azh'e (Yegorov), 11.VI.1847; ♂, Russia, St. Petersburg vicinity, 24.V.1998; ♀, Russia, vicinity of Penza, 1929 (Filina); ♂, Ukraine, Lugans'ka oblast, Lugansk vicinity, 23–28.V.1982 (Talitzky) (ZISP); ♂, Spain, Gatova, 17.V.1989 (Martinez, Rasplus) (INRA). Also more than 160 specimens from England, Bulgaria, Serbia, Greece (Gumovsky, Boyadzhiev, Reference Gumovsky2003).
Discussion
Gumovsky & Boyadzhiev (Reference Gumovsky2003) demonstrated wide morphological variety of this species. In particular, there is a peculiar variety, which occurs in the Balkan peninsula and is characterized by notably smaller body size; also, the gaster and flagellar segments of the female are somewhat shorter in this smaller variety than in ‘typical’ E. sparetus. However, there are also some intermediate-sized varieties filling at least the size hiatus between the ‘typical’ E. sparetus and the ‘smaller variety’. Many measurements were employed for the testing of possible distinction between the ‘large’, ‘mid’ and ‘small’ varieties but unexpectedly, only one character (the length of the terminal flagellar segment of the male) suggested a separation between these forms. However, this measurement overlapped in ‘large’ and ‘mid’ varieties, and the difference between the ‘small variety’ and other ones, is probably merely size-dependent. The assumption of the conspecificy of these forms was supported by molecular studies (in particular D2 28S rDNA gene) (Gumovsky & Boyadzhiev, Reference Gumovsky and Boyadzhiev2003).
Apart from the forms discussed above, the females of E. sparetus are easily confused with the females of E. cardui, which differs only by the shorter malar space (eye height 1.8–2.5 times as long as malar space in E. sparetus, but about 3.0 times in E. cardui).
Entedon cardui Askew (fig. 5C,I)
Entedon carduiAskew, Reference Askew, Blasco-Zumeta and Pujade-Villare2001: 67.
? Entedon longiventrosusDalla Torre, Reference Dalla Torre1898: 40 (replacement name for Entedon longiventrisThomson, Reference Thomson1878 nec Ratzeburg, 1848).
Diagnosis
Apical margin of fore wing with a fringe of setae; metasoma of female about as long as head plus mesosoma, 1.8 times as long as broad; eye height 3.0–3.3 times as long as malar space; fore wing 2.1–2.2 times as long as broad; hind tarsus 0.7 times as long as hind tibia, fore and mid tarsi about 0.8 times as long as their tibiae; fore wing 2.1–2.2 times as long as broad; eye height about 3.0 times as long as malar space; female: antennal scape 6.2–6.4 times as long as broad; pedicel 2.0–2.2 times as long as broad; F1 2.5, F2 2.1–2.2, F3 1.3 times as long as broad, clava slightly more than twice longer than broad; male: mouth opening 2.0–2.7 times as long as broad; antennal scape 3 times as long as broad; pedicel 1.7 times as long as broad; F1–F3 separated by distinct peduncles, degree of constriction between F4 and F5 varies, but anyway they are closely attached to each other, fig. 5I (likewise in E. thomsonianus); F1 2.8–3.0, F2 1.6, F3 1.3 times as long as broad, F4 as long as broad and F5 1.33 times as long as broad.
Distribution
Spain (Askew et al., Reference Askew, Blasco-Zumeta and Pujade-Villare2001), Bulgaria, Greece (Gumovsky & Boyadzhiev, Reference Gumovsky and Boyadzhiev2003), new for Italy, France and Ukraine.
Hosts
No biological observations are available for this species. The reported association of E. cardui with the pine bark anobiid Ernobius mollis (Linnaeus) on Pinus in France (MNHN label) seems doubtful and requires confirmation, whereas the weevil host records (SIZK, UCRC labels) are more reliable. Entedon cardui is obviously associated with the seed-eating weevil larvae in flowerheads of plants of Asteraceae (Sylibum marianum, Cirsium ucrainicum). The reported association of E. cardui with the musk thistle weevil Rhinocyllus conicus (Froelich) is remarkable, because this weevil was included in biocontrol programmes against thistles in North America (Kok & Surles, Reference Kok and Surles1975). Rhinocyllus conicus attacks many thistle genera (e.g. Carduus, Cirsium, Silybum and Onopordum), but host races of this weevil are associated with separate food plants (Zwölfer & Preiss, Reference Zwölfer and Preiss1983). The female R. conicus lays her eggs singly or in small clusters on bracts on the undersides of mature flower buds of various thistles and then covers the laid eggs with a cap of chewed thistle tissues (Zwölfer & Harris, Reference Zwölfer and Harris1984), similarly to Lixus cardui, which is discussed below. De Santis et al. (Reference De Santis, Monetti and Briano1987) studied the parasitoids of R. conicus, but no Entedon species were recorded from this host. Harris (Reference Harris2005) reported that the caps covering the weevil's eggs are tended by ants, which in Europe may discourage specialisedegg parasites. If E. cardui proves to be an egg-larval parasitoid likewise other members of the sparetus group, it is expected to manage attacking the eggs of the musk thistle weevil.
Material examined
♂, Bulgaria, Rhodope Mts, v. Mandritsa, 100 m, 19.05.1981 (Donev, Boyadzhiev) (PU); ♂, Greece, Sparta, 180 m, 10.V.1987 (Angelov); ♂, ibid., Tripolis, 700 m, 10.V.1987 (Angelov) (TMA); 2♀, 5♂, Ukraine, Zaporiz'ka oblast, Melitopol'sky region, v. Sosnovka, reared from flowerheads of Cirsium ucrainicum, coll. 2.IX.82, reared 18–19.IV.1983 (Volovnik) (SIZK); ♀, 2♂, France, Perrusson (I.L.), Méq., eclosion, éclose de branche de Pinus avec Ernobius mollis (ex coll. DeGaulle) (MNHN); 7♀, France, Hérault, Vélieux, 500 m, 13.VI.1997 (Delvare); ♂, ibid., St Martin Londres, Frouzet, 16.VI.1990 (Delvare); ♀, France, Alpes, Maritimes, 1900 m, Entraunes, Esteng, 19.VII.1992 (Delvare); ♀, 5♂, Spain, Algésiras, 29.IV.1989 (Tussac); ♂, Spain, Cambrils, 23–24.V.1990 (Tussac) (CIRAD); ♂, France, Drome, Suze la Rousse, 1.5.1990 (Rasplus); ♂, France, Pyren.Or, Banyuls, 15.4.1990 (Rasplus); ♂, Spain, Valencia, Masanasa, 4.5.1989 (Luna) (INRA); 2♂, France, Toulon, Var, 26.4.1952 (Barbier) [identified as “Entedon longiventrosus” by J. Erdös]; ♂, ibid., 28.5.1954; ♀, ♂, France, Nîmes, “capitules la Carlina corymbosa” 11.VI.1918 (Cabanis) (Coll. F. Picard (coll. Lichtenstein) Mus. Paris 1939) (MNHN); 7♀, 2♂, Italy, Rome, Lazio, Via Appin Antien, in mature head of Sylibum marianum Gaetrn. as Rhinocyllus conicus parasite, 06–22.VII.1971 (Goeden) (UCRC).
Discussion
The males of this species are easily recognizable by the combination of the closely attached two terminal antennal flagellomeres and the possession of a fringe of setae on the apical margin of the fore wing (distinguishing them from the males of E. thomsonianus having the same antennal formula). Askew et al. (Reference Askew, Blasco-Zumeta and Pujade-Villare2001), and then Gumovsky & Boyadzhiev (Reference Gumovsky2003) failed to provide any reliable characters for separation of females of this species from the females of E. sparetus. However, here a new distinguishing character, the ratio of eye height to the length of malar space, is reported. This ratio is 2.1–2.4 in females of E. sparetus and about 2.7–3.3 in females of E. cardui. The possession of the latter ratio by the holotype female of E. longiventris Thomson (studied, LUZM) may suggest that this species and E. cardui are one and the same species. However, I postpone the final decision on synonymy and name priority until more straightforward morphological markers and other data (e.g. biological, molecular) are available.
Entedon thomsonianus Erdös (figs 3B, 4C,D, 5E,J)
Entedon (Megalentedon) thomsonianusErdös, Reference Erdös1944: 27.
Entedon thomsonianusErdös; Graham, Reference Graham1971: 342.
Diagnosis
Apical margin of fore wing without apical fringe; metasoma of female 1.3 times as long as head plus mesosoma, 2.8–3.0 times as long as broad; eye height 2.5 times as long as malar space; fore wing 2.3 times as long as broad; hind tarsus about 0.7 times shorter than hind tibia, fore and mid tarsi 0.8 times shorter than their tibiae; female: antennal scape 5.6 times as long as broad; pedicel 2.7 times as long as broad, F1 3.0, F2 1.8, F3 about as long as broad, clava two-segmented, 1.6 times as long as broad; male: antennal scape 3 times as long as broad; pedicel about 2.7–2.8 times as long as broad, F1 2.8, F2 about twice, F3 1.3, F4 1.3, F5 1.75 times as long as broad, F4 and F5 closely attached to each other.
Distribution
Widely in Europe (Bouček & Askew, Reference Bouček and Askew1968; Askew et al., Reference Askew, Blasco-Zumeta and Pujade-Villare2001; Gumovsky & Boyadzhiev, Reference Gumovsky and Boyadzhiev2003), new for Ukraine, Georgia, Turkmenistan.
Host
Lixus cardui Olivier (Curculionidae) (Erdös, 1949, Reference Fisher1951; Bouček & Askew, Reference Bouček and Askew1968; here). The reported association of E. thomsonianus with a buprestid Meliboeus amethystinus in France (MNHN label) requires confirmation.
Material examined
Lectotype: ♀, Hungary, Kerepes, 6.VI.1917 (Horváth) (selected by Szelényi, 1976, designated by Thuróczy, 1992). Paratypes: 12♀, № 4885–4893, 8213 (all in TMA). Other material: 8♀, Azerbaijan, Giandzha (Elisavetpol), 3.VI.1912 (Babadzhanidi) (ZISP); ♀, Bulgaria, v. Cherna Gora, near Chirpan, 170 m, 07.VI.1967 (Germanov); ♀, 150 m, v. Aprilovo, 19.V.1989 (Boyadzhiev); ♀, Bulgaria, v. Lozenets, 30 m, 06.VII.1969 (Angelov); ♂, Bulgaria, v. Boljarovo, Strandzha Mt., 200 m, 09.V.1971 (Angelov); ♀, Bulgaria, W. of Tsarevo, 50 m, swept on Carduus sp. 05.VI.2001 (Boyadzhiev) (PU); ♀, Bulgaria, v. Pastra, Rila Mt., 850 m, 11–31.V.1998 (Achterberg, de Vries, Atanassova) (RMNH); Georgia, Lagodekhi, 1910 (ZISP); 3♀, 2♂, France, St Martin de Grau, ex Meliboeus amethystinus in Onopordum acanthium, VI–VII.1987 (Liskenne) (MNHN); ♀, Russia, Taganrog vicinity, 24.V.-27.VI.1927 (Anger); ♀, Turkmenistan, Kopetdag, 6–8.V.1913; 2♀, Russia, Dagestan, Derbent (Bekker); ♀, Ukraine, Crimea, Sudak, 28.V.1938 (Ljubischev) (ZISP); 4♀, 2♂, Ukraine, Kherson oblast, Black-sea nature reserve, Ivano-Rybalchansky plot, ex stem Carduus sp., 16.V.1971 (Zerova); ♀, ♂, Ukraine, Kherson oblast, Lazurnoe vicinity, collected on shoots of Onopordum acanthium, 10.VI.2003 (Gumovsky); 5♀, Ukraine, Crimea, Belogorsk region, Belaya Skala range, on O. acanthium, 18.V.1983 (Volovnik); ♀, ♂, Ukraine, Crimea, Saki vicinity, reared from stem of O. acanthium, 07.VIII.1994 (Gumovsky) (SIZK).
Entedon zerovae Gumovsky (figs 1, 5F,H)
Entedon zerovaeGumovsky, Reference Gumovsky1995: 45.
Diagnosis
Apical margin of fore wing with a fringe of setae; metasoma of female nearly twice as long as head plus mesosoma, about 5 times as long as broad; eye height nearly twice as long as malar space; fore wing 2.4 times as long as broad; hind tarsus about as long as hind tibia, fore and mid tarsi just slightly shorter than their tibiae (about 0.9 times as long); female: antennal scape 7.4 times as long as broad; pedicel 2.2 times as long as broad; F1 4.8, F2 2.7, F3 1.6, F4 1.6, F5 slightly more than twice as long as broad, F4 and F5 separated; male: antennal scape slightly more than 2.5 times as long as broad; pedicel about twice as long as broad, F1 2.8–3.0, F2 and F3 about twice, F4 about 2.6, F5 about 5 times as long as broad.
Distribution
Tadjikistan (Gumovsky, Reference Gumovsky1995), new for Turkmenistan, Uzbekistan.
Host
Lixus strangulatus Faust (Curculionidae) (Gumovsky, Reference Gumovsky1995).
Material examined
Holotype: ♀, Tadjikistan, Ghissar mountain range, Kondara canyon, ex Lixus strangulatus Faust in stem of Inula grandis Schrank, 24.III.1981 (Zerova). Paratypes: 3♀, 1♂, ibid (SIZK). Other material: ♂, Turkmenistan, ex flowerheads of Cousinia eringioides, 29.III.1980, reared 17.IV.1980 (Zerova) (SIZK); ♀, Uzbekistan, Chatkalsky Nature Reserve, 1300 m, 14.V.1980 (Kasparyan); ♂, Turkmenistan, Kyrk-Kichik-Tau mountains, 27.V.1925 (Dobzhansky) (ZISP).
Entedon tobiasi Gumovsky (figs 2A, 5B)
Entedon tobiasiGumovsky, Reference Gumovsky2004: 165.
Diagnosis
Apical margin of fore wing with a fringe of setae; metasoma of female 2.1–2.2 times as long as head plus mesosoma, about 6 times as long as broad, its syntergum at least 2.5 times longer than broad, about one-quarter to one-fifth of length of the metasoma; antennal scape of female 5.0–6.2 times as long as broad; pedicel about 1.1–1.5 times as long as broad; F1 4.4–5.0, F2 2.5–2.6, F3 1.8–1.9 times as long as broad, clava two-segmented, 2.3–2.6 times as long as broad; eye height 1.4–1.8 times as long as malar space; fore wing 2.4–2.5 times as long as broad, hind tarsus nearly as long as hind tibia, fore and mid tarsi just slightly shorter than their tibiae.
Male
Unknown.
Distribution
Uzbekistan.
Host
Unknown.
Material examined
Holotype ♀, Uzbekistan, Zhamansaj range, Kyzylkum desert, 22.IV.1976 (Falkovich). Paratypes: 1 ♀, ibid; 1♀, Kyzylkum desert, Ayakgudiumdy, sweeping, 07.IV.1966 (ZISP).
Entedon lucasi Gumovsky sp. n. (figs 3C, 5D)
Diagnosis
Metasoma 1.8–1.9 times as long as head plus mesosoma, about 4.3 times as long as broad; syntergum twice longer than broad, about one-seventh of length of the metasoma; antennal scape of female 7.2 times as long as broad; pedicel about 2.2 times as broad; F1 about 5.0, F2 2.5, F3 1.3–1.4 times as long as broad, clava two-segmented, about twice longer than broad, slightly more than 1.5 times longer than the preceding segment; eye height 1.5 times as long as malar space; breadth of mouth opening 1.4 times as long as malar space; fore wing 2.6 times as long as broad, apical margin with a fringe of very short setae, which half as long as width of marginal vein at its narrowest part; hind tibia 1.4 times as long as its tarsus, fore and mid tibiae about 1.3 times as long as their tarsi.
Female
Length 6.7–7.3 mm. Body metallic green, frons with weak bronze tint. Entire antennae dark. Legs dark, except knees, extreme distal ends of tibiae and first tarsomere of all legs, which are pale. Dorsal pale longitudinal stripe on fore tibia discernible along entire tibia. Oval membranaceous areas adjacent to petiolar insertion point on first metasomal tergite brown to dark orange. Head in dorsal view 2.75 times as broad as long; POL:OOL:MDO:OCL=25:14:6:1 in holotype. Occipital margin moderately sharp. Eye with short sparse setae, eye height 1.5 times as long as malar space. Head in facial view 1.6 times as broad as long. Interocular distance 2.8 times as long as eye breadth. Malar sulcus replaced by a row of finer alveoli. Breadth of mouth opening 1.4 times as long as malar space. Clypeus reticulate, its anterior margin weakly produced. Antennae inserted at the level of ventral eye margins. Antennal scape of female 7.2 times as long as broad; pedicel about 2.2 times as long as broad; F1 about 5.0, F2 2.5, F3 1.3–1.4 times as long as broad, clava two-segmented, with short terminal spine, about 1.8–2.0 times as long as broad, slightly more than 1.5 times longer than the preceding segment; eye height 1.5 times as long as malar space. Mesosoma 1.5–1.6 times as long as broad. Pronotal collar hardly traceable, postero-lateral corners of pronotum evenly rounded. Mesoscutum 1.5–1.6 times as broad as long, notauli traced anteriorly as very fine sutures, posteriorly as shallow depressions; scutellum slightly longer than broad and slightly longer than mesoscutum. Propodeal surface light reticulate, median carina complete, lateral sulcus incomplete; supracoxal flange moderate; spiracular elevation with blunt projection below, propodeal callus with 1 long and 8–9 shorter setae. Metapleuron with comparatively blunt protrusion. Hind coxa reticulate dorsally. Fore femur about 4.3 times as long as broad, fore tibia 10.0 times as long as broad, as long as its femur; mid femur 4.3 times as long as broad; mid tibia 13 times as long as broad, spur of mid tibia slightly longer than breadth of tibia, half as long as dorsal margin of mid basitarsus; hind femur about 4.0 times as long as broad, hind tibia about 10 times as long as broad, spur of hind tibia about 0.5 times as long as breadth of its tibia, about quarter as long as dorsal margin of hind basitarsus. Hind tarsus 0.6–0.7 times as long as its tibia, mid tarsus 0.8 times as long as its tibia. Ratio of tibiae and tarsi of holotype are as follows: fore tibia:tarsus 70:60; fore tarsomeres: 10:19:10:10 (+ pretarsus 10); mid tibia:tarsus 100:76; mid tarsi: 18–25 (dorsal – ventral edge of basitarsus): 24–26:13:11 (+ pretarsus 8); hind tibia:tarsus 202:135; hind tarsi: 30–24:31–28:15:16 (+ pretarsus 7). Fore wing 2.6 times as long as broad; costal cell bare, comparatively wide, 6.5 times as long as broad, slightly longer than marginal vein; subcosta of submarginal vein with 2 dorsal setae, postmarginal vein slightly shorter than stigmal; speculum open; apical margin with very short fringe, setae of which half as long as width of marginal vein at its narrowest part. Petiole reduced, strongly transverse. Metasoma 1.8–1.9 times as long as head plus mesosoma, about 4.3 times as long as broad; syntergum twice longer than broad, about one-seventh of length of the metasoma.
Male
Unknown.
Distribution
Algeria.
Host
Unknown.
Type material
Holotype ♀, Algeria (H. Lucas) (BMNH); paratypes: 1 ♀, ibid. (BMNH); 2 ♀, ibid. (MNHN).
Biology
Entedon thomsonianus
The host, Lixus cardui
Adults of Lixus cardui (fig. 6A) feed on shoots and leaves of plants of the family Asteraceae, in particular the scotch thistle, Onopordum acanthium L. In the south of Ukraine (Kherson oblast) oviposition starts at the end of May, lasts during most of June, and occasionally continues till the beginning of July, but not later because of desiccation of the hosts' food plant in mid–end of July. The oviposition and mating behaviour of the weevils are remarkable. During copulation the female starts preparation of a ‘fibre groove’ on the stem of the hosts' food plant (fig. 6A). She deeply inserts her rostrum into the plant tissues (up to the base of the rostrum) and chews during most of the term of the copulation, which generally lasts for some hours. The groove is bordered by the peculiar gnawed fibrous tissues of the host-plant (fig. 6E) and its bottom possesses a row of pits where the eggs will be laid later. Soon after copulation the female lays relatively large oval eggs (1.1–1.5 mm long, 0.9–1.0 mm wide) into these pits, generally 2–6 per groove (fig. 6D,E). The eggs are then covered with chewed plant fibres.
A, mating pair of the weevil Lixus cardui; the female burying her rostrum (arrowed) into a stem of Onopordum acanthium; B, female of Entedon thomsonianus discovers the mating weevils; C, group of females of E. thomsonianus is attracted by mating weevils; D, female of E. thomsonianus compared with the egg of L. cardui; E, lateral section of the stem of O. acanthium under the ‘fibre groove’ (egg set of L. cardui is shown); F, female of E. thomsonianus searching for the eggs of L. cardui in the dissected stem of O. acanthium (laboratory experiment); black arrow, ‘fibre groove’ and the place where female's rostrum is inserted there. Scale bars: A–C, 10.0 mm; D–F, 1.0 mm.
The egg develops during 5–7 days, the embryo is visible from 3–4 days. Soon after hatching, the young larva leaves the ‘fibre groove’, bores deeper into the stem, forming an individual mine, and begins feeding on the structural tissues of the plant. The first instar larva moults and the second instar larva continues moving within the host-plant stem. One plant shoot typically contains larvae in various instars. Larval development of L. cardui lasts 3–4 weeks, and pupation occurs inside the hosts' food plant. Adult beetles emerge the following spring. Lixus cardui was included in biocontrol programmes against the scotch thistle in North America and Australia; in the latter region it was reported as a contributor to biological control of the weed (Briese, Reference Briese1996; Huwer et al., Reference Huwer, Briese, Dowling, Kemp, Lonsdale, Michalk, Neave, Sheppard and Woodburn2004).
Host searching and oviposition by E. thomsonianus
The females of E. thomsonianus can be found in the field in the south of Ukraine (Kherson oblast) from late May until early July. The females may often be found on shoots and leaves of O. acanthium searching for the ‘fibre grooves’ or mating host weevils (fig. 6B). Occasionally, a few to several females gathered around one mating couple of L. cardui (fig. 6C), waiting for the freshly laid eggs of the host. These female parasitoids occasionally attack each other, or even try to probe the ‘fibre groove’ being prepared by the female weevil, with their ovipositors. Once one female gets access to the ‘fibre groove’, she drums it with her antennae (fig. 6F), then bends her gaster downwards and begins probing the groove, attempting to locate the host eggs. The weevil's egg is about one-quarter of the body length of the adult female parasitoid (fig. 6D). The detection of the exposed weevil eggs by the female of E. thomsonianus in the laboratory is much quicker: as soon as the egg is found, the female probes the egg briefly with her ovipositor sheaths, and then fixes the end of the ovipositor's tip onto the egg chorion, but does not penetrate the egg yet. Then she releases the gaster so that it strengthens in a position perpendicular to the ovipositor and carries out rhythmic, twisting movements which finally lead to the penetration of the egg. Oviposition lasts about 45–60 seconds and during all this time the female carries out twisting movements. When the female withdraws the ovipositor from the host egg, the egg keeps its shape.
In the laboratory one female occasionally oviposited in the same host egg twice. In another experiment the female laid her eggs into the mature host eggs, containing moving young embryonic larvae.
Parasitoid–host relations
Dissections of the oviposition sites of E. thomsonianus in field and laboratory experiments revealed that this species is an egg-larval parasitoid of L. cardui.
Eggs and first instar larvae of E. thomsonianus were found inside the eggs of L. cardui: both collected in field, and observed in the laboratory after oviposition. The parasitized eggs occasionally bear melanized marks (fig. 7A), which correspond probably to the marks resulting from the penetrations of the parasitoid's ovipositor. Normally, only a single parasitoid larva was found per host egg (fig. 7B,C); however, in seven cases two, in two cases three, and once four larvae (fig. 7D) were found in the same host egg. In these cases, either one (when three larvae were found) or two (when four larvae were found) first instar parasitoid larva were alive and moving, whereas other ones were motionless, with contents of the mid gut extruded through the body wall, and thus probably dead (fig. 7D), suggesting that these are cases of superparasitism, when one larva kills other ones. It is unclear whether the larva from the first or the last laid egg kills the rival larvae. Occasionally, motion of the parasitoid larvae may be observed through the transparent chorion of the host egg. These movements resemble crawling to some extent, but the larva is floating inside the egg, among the germplasm.
A, egg of the weevil Lixus cardui parasitized by Entedon thomsonianus. B–G, first instar larvae of E. thomsonianus in eggs and larvae of L. cardui; B, C, first instar parasitoid larva isolated from the egg figured in A; D, four first instar parasitoid larvae in the same host egg; E, first instar parasitoid larva isolated from newly hatched host larva; F, G, first instar parasitoid larva isolated from actively feeding first instar host larva. cf, caudal filaments; fr, ‘caudal formation’; grey arrow, melanized dots on the parasitized host egg; black arrow, live larva; white arrow, dead larva. Scale bars: D, 0.35 mm; E, 10.0 mm.
Dissections of the first instar larvae of the host weevil produced the same first instar larvae of E. thomsonianus (fig. 7E), recognizable by its peculiar body shape (see below), but more swollen and bearing the caudal filaments (fig. 7F,G). In two cases, two first instar larvae were found inside the weevil larva's body, but in such cases one of the two was motionless and probably dead. Second instar larvae of the parasitoid were found in the second instar larvae of the weevil (fig. 11A,B) and in all other instars of the weevil, except its final instar. Parasitized and unparasitized host larvae do not differ, neither visually, nor in behaviour, i.e. they actively move within the plant shoot and try to protect themselves with their mandibles, when disturbed. Unlike the earlier instars, the parasitized final instar larvae of L. cardui demonstrated rather slow movement, and later became motionless. Then the body of the final instar weevil larva appeared filled entirely by the body of the final instar larva of the parasitoid. Entendon thomsonianus always pupates inside the hosts' food plant, near the remains of the host larva. Pupae were found from mid-late July, but adults emerged the following spring.
Immature stages of E. thomsonianus
Egg
The eggs found in the host eggs do not differ notably from those isolated from ovaries, transparent, without discernible sculpture. Length, about 450 μm, width about 120 μm.
First instar larva
Habitus. The first instar larva of E. thomsonianus is hymenopteriform, pale (nearly transparent), has 13 body segments and a cranium (fig. 8). The larvae isolated from the host eggs are 460–510 μm long and about 120–140 μm wide. The terminal, XIIIth, segment bears sharp triangular tubercles along its dorsal margin, arranged in groups of two or three, or in tufts, and in the general shape of a semi-circular crown (fig. 10). Segments I–III (thoracic) are bare, whereas segments IV–XII (abdominal) bear distinct dorsal semicircular serrations along their anterior margins, which consist of small teeth. Occasionally the serration of the IV (first abdominal) segment is missing or obscure (fig. 8A,B). No spiracles were found on the body of the larvae. The larvae isolated from the first- and second instar host larvae are more ‘swollen’, about 360 μm wide (fig. 7F), their ‘caudal crown’ is less clearly visible (it appears contracted, as in fig. 10F), and their segments are more spread (fig. 8B). However, these larvae possess non-sclerotized caudal filaments, which are visible in live specimens as a compact bunch of transparent fibres (fig. 7F,G, cf), but are destroyed under fixation and drying. These caudal filaments are emanating from a dark ‘caudal formation’ (fig. 7G, fr) that is of uncertain nature. This organ is clearly visible in live specimens under light microscopy, and, judging from its darker coloration (in comparison with the rest of the larval body), it may represent a dense tissue aggregation.
First instar larvae of Entedon thomsonianus, habitus. A, C, the larva is isolated from the host egg; B, the larva is isolated from the host first instar larva; A, B, lateral view; C, ventral view.
First instar larva of Entedon thomsonianus, details of head morphology (see text for abbreviations). A, lateral view; B, latero-ventral view; C, ventral view.
First instar larva of Entedon thomsonianus, cauda close-ups. B, C, close-ups of A; E, close-up of D. A–E, postero-lateral view; F, posterior view.
Head capsule (cranium), fig. 9. The head capsule is weakly sclerotized, narrowing ventrally, with a characteristic ‘beak’-shaped end, which is formed by the protruding palpi of the labrum, which covers thin mandibles. Under light microscopy, the labrum is seen to adopt an active backward and forward motion. Antennae (ant) are indicated as small but broad swellings on the upper part of the head capsule. The sensorial structures of the head are well-developed and arranged in a relatively fixed position. The lateral area of the cranium bears three pairs of cranial palpi: the upper (cp1), the lower (cp2) and the posterior (cp3) palpi. There is a pair of enlarged pleurostomal palpi (plp) just above the labrum. The labrum bears two pairs of labral palpi (grouped as one from each side): the upper lateral labral palpi (ulrp) and the lower lateral labral palpi (llrp). A large palpus is located near each maxilla (mxp) and behind them there is a pair of smaller labial palpi (lbp). A pair of comparatively long pharyngeal setae (phs) is situated on their palpi behind the labial palpi. Occasionally one pharyngeal palpus may bear two setae (fig. 9B).
Second instar larva
Morphological peculiarities of the instar which is believed to be the second remain largely obscure. The body segmentation of this instar is more distinct in live specimens (fig. 11A–F), but hardly discernible in critical point dried specimens (fig. 12) The second instar larva of E. thomsonianus is also hymenopteriform and pale, body segmentation is difficult to estimate. The larva of this instar, which is recently moulted has a body about 750 μm long and about 300–350 μm wide (figs 11B, 12A). In comparison with the first instar the second instar larva is more robust (fig. 11A–F, H), none of the body segments has serrations, and the terminal segment bears no visible ‘caudal crown’. The mature second-instar larva is notably more elongate, body about 2300 μm long, 500–550 μm wide (figs 11D–F, H, 12D), its body is milky and the head is more spherical, the sensorial structures were not detectable (fig. 12A–C). The caudal filaments of the young second instar larva are less distinct than in the first instar larva, but have a similar, cone-like, shape (fig. 11B, cf). In the mature second-instar larva the caudal end bears a nearly transparent spherical appendage (fig. 11E,F, ca), with indistinct structure. These appendages were smashed in critical point dried specimens (fig. 12E). Again no spiracles were found.
Larval instars of Entedon thomsonianus. A, B, young second instar parasitoid larva isolated from second instar host larva; C–F, mature second instar parasitoid larva isolated from older host larva; G, final instar larva isolated from final instar host larva; H, all larval instars of E. thomsonianus isolated from their hosts at different instars: cf, caudal filaments; ca, caudal appendage; I(e), first instar larvae, isolated from eggs; I(L), fed first instar larva (caudal filaments are traceable), isolated from the host larva; II, second instar larvae (young and mature); f, final instar larva. Scale bar, 1.0 mm.
Final instar larva (figs 11G, H, 13, 14)
The final instar larva of E. thomsonianus is of distinct habitus, being characteristic for the final instar larvae of Hymenoptera: body swollen, with distinct segmentation and clearly outlined head having elongate mandibles (fig. 13). The larva is 3.4–3.5 mm long, its body has 13 distinct segments, and the anal segment is the smallest (fig. 13B, white arrow). The larva has distinct labrum (fig. 14A, lr), labium (fig. 14A, lb) and large (about 100 μm long) and heavily sclerotized (dark brown in colour, fig. 15B) mandibles with robust bases (fig. 14, bam) and curved blades (figs 14, blm; 15B). The sensory structures of the head are less distinct than in the first instar larva, but some are recognizable enough: the antennae (ant) are large and situated on depressed areas, the cranial palpi (cp), the pleurostomal palpi (plp), the upper (ulrp) and the lower (llrp) labral palpi and the maxillar palpi (mxp). The spiracular openings were not recognized on the critical point dried specimens, but studies on the larval skins detached from the caudal ends of the pupae suggest the presence of one pair of thoracic (probably on the second thoracic segment) and 6 pairs of abdominal spiracles (fig. 15A,C).
Final instar larva of Entedon thomsonianus, habitus. as, white arrow: anal sclerite.
Final instar larva of Entedon thomsonianus. A, head close-up; B, mouth area, close-up (see text for abbreviations).
Skin of final instar larva of Entedon thomsonianus. A, general view; B, head capsule; C, spiracle; black arrows, spiracles.
Pupa
Pupation takes place inside the hosts' food plant, within the host's stem mine, at various sites of the plant stem, and near the remnants of the final instar larva of the host. The pupa is black, obtect, with distinct outlines of head, mesosoma, metasoma, wings, legs and antennae (similar to E. sparetus, see below). The average length is 4.5–6.3 mm. The skin of the final instar larva (fig. 15) is often attached to the caudal end of the pupa.
Entedon sparetus
As was stated above, the forms similar to E. sparetus were studied by Gumovsky & Boyadzhiev (Reference Gumovsky2003), but no obvious differences were found based on morphology, and 28S D2 rDNA sequences were the same in large- and small-sized forms. This was considered to support the conspecific nature of these forms by the authors and is accepted here. However, further studies are needed to establish the true status of the forms considered here as E. sparetus, and it remains possible that the host strains of what is here considered E. sparetus are indeed separate species.
The host, Gymnetron asellus
Adults of Gymnetron asellus feed on shoots and leaves of various species of mullein (Verbascum spp.). In the north of the Ukraine (Kiev) oviposition starts at the beginning of June, before the blooming of the hosts' food plant. The females lay their eggs in various places of the host-plant shoots, but the oviposition sites are mostly concentrated at the top of the plant (fig. 16A,D,E). The female prepares an opening in the host-plant tissues with her rostrum and then she lays eggs there. The oviposition sites are visible as black marks, resulting from the host-plant's sap (fig. 16A,D,E).
A, D, E, shoots of Verbascum sp. damaged by the weevil Gymnetron asellus (swollen parts with dark pits caused by insertion of the rostrum of the female weevil are enlarged in A (inset) and E). B, dissected stem of Verbascum sp. with exposed egg of G. asellus (enlarged in C). F, first instar larva of G. asellus in its egg (the same magnification as in C). White arrow, partition in the egg. Scale bars in A, B, D, E, 10.0 mm.
The egg of G. asellus (fig. 16B,C,F) is about 260–270 μm long, of a peculiar ‘mushroom’ shape: wider anteriorly (90–94 μm) and narrowed posteriorly (60–63 μm). Its posterior part (the ‘neck’) is separated from the anterior part containing the yolk and then the embryo (the ‘yolk chamber’) by a transverse membranous partition (fig. 16F). The egg develops during about 7 days; the embryo is visible on day 4–5. Immediately after it hatches, the young larva bores into the host-plant stem. Weevil larvae in various instars may be encountered in one and the same plant shoot. Finally the mature larva of the weevil prepares a pupation cell situated just beneath the outer layer of the stem. Larval development of G. asellus lasts for about 3 weeks, and the pupae may be found inside the hosts' food plant from late June.
Although mulleins are counted among the invasive weeds (Martin, Reference Martin1987; Haragan, Reference Haragan1991) and some insects have been tested as agents against these plants in North America (Maw, Reference Maw1980), G. asellus has not yet been considered for biocontrol of Verbascum species.
Host searching and oviposition by E. sparetus
The oviposition behavior of E. sparetus was observed in early June. It is similar to that described above for E. thomsonianus, except that the parasitoid female more accurately searches for the host, drumming along the stem of mullein with her antennae.
Parasitoid–host relations
Dissections of mullein shoots in places where the oviposition of E. sparetus was registered revealed that this species is also an egg-larval parasitoid.
Eggs and first instar larvae of E. sparetus were found inside the eggs of G. asellus. Only a single parasitoid larva was found per host egg. Since the weevil's egg is relatively small and transparent, the motion of the parasitoid larvae may be observed inside the egg (fig. 17). These movements also resemble crawling to some extent, and occasionally it is possible to trace how the larva penetrates the host embryo during the crawling movements (fig. 17A–C). Occasionally, the parasitoid larva may be found within the ‘neck’ of the host egg (fig. 17D,E). In this case the larva tosses and turns within this ‘neck’, scratching the partition between the ‘neck’ and ‘yolk chamber’ in turn with its caudal (fig. 17D) and cranial (fig. 17E) ends. These movements may last for about one day, and the next day the larva penetrates the partition entering the ‘yolk chamber’; occasionally with its caudal end (fig. 17F).
First instar larva of Entedon sparetus in egg of Gymnetron asellus. White arrow, partition in the host egg; black arrow, parasitoid larva.
The dissections of the first instar larvae of the host weevil produced the first instar parasitoid larvae which are either identical to ones isolated from the host eggs (fig. 18A) or slightly thicker and larger (fig. 18D), but still recognizable by its peculiar body shape and the ‘caudal crown’ (likewise in E. thomsonianus). Only a single first instar larvae was found per weevil larva. The second instar larvae of the parasitoid were found in the middle and final instars of the host weevil larvae (fig. 22A), and the final instar parasitoid larvae were found only in the final instar host larvae (fig. 22B). The parasitized host larvae did not differ from the unparasitized ones and actively moved within the plant shoots. Unlike the earlier instars, the parasitized final instar larvae of G. asellus demonstrated rather slow movements, and later became motionless. Finally, the body of the final instar weevil larva appeared filled entirely by the body of the final instar larva of the parasitoid (fig. 22C–E). The final instar larvae of E. sparetus always pupate inside the hosts' food plant shoots (fig. 22F), near the remains of the host larva. Pupae were found from late June, but adults emerged in the following spring.
First instar larva of Entedon sparetus. A, parasitoid larva isolated from the egg of G. asellus; B, parasitoid larva inside the body of newly hatched larva of G. asellus; C, parasitoid larva isolated from newly hatched host larva; D, parasitoid larva isolated from actively feeding first instar host larva; E, F, caudal end of the larva figured in D. fr, ‘caudal formation’ transiting into transparent (and thus hardly visible) caudal filaments; black arrow, parasitoid larva; white arrow, ‘caudal crown’.
First instar larvae of Entedon sparetus isolated from the host egg (A) and host first instar larvae (B–F). A–C, habitus; D, cauda close-up; E, F, head close-up.
Second instar larva of Entedon sparetus. A, habitus; B, E, cranial end (lateral (B) and ventral (E) views); C, caudal end; D, body fragment; F, head enlarged, ventral view. cf, caudal filaments; fr, ‘caudal formation’; mg, mid gut; md, mandible; oes, oesophagus; black arrows, vesicles of tracheal system.
Second instar larva of Entedon sparetus, the moment of moulting from the first instar, SEM. A, habitus; B, cranial end; C, first instar skin and head capsule at caudal end (white arrow in A).
Second (A) and final (B–F) instars of larvae of Entedon sparetus. A, B, parasitoid larvae isolated from the final instar host larva; C, final instar parasitoid larva inside the intact skin of the final instar host larva (entire body of the host is consumed and substituted by the parasitoid); D, final instar parasitoid larva emerging from the caudal end of the final instar host larva; E, final instar parasitoid larva emerging from the cranial end of the final instar host larva; F, final instar larvae of E. sparetus in the dissected stem of Verbascum sp. damaged by Gymnetron asellus. White arrow, parasitoid larva; black arrow, host's larval skin. Scale bar, 1.0 mm.
Immature stages of E. sparetus
Egg
The egg is similar to that of the previous species: length about 250 μm, width about 90 μm.
First instar larva
The first instar larva of E. sparetus isolated from the host egg (figs 18A, 19A) is similar to that of E. thomsonianus, being nearly transparent, about 240–300 μm long and about 80–100 μm wide, with a peculiar cranium with sharp sickle-shaped mandibles, and 13 body segments, of which IV–XII bear dorsal serrations and the terminal (XIII) bears a ‘caudal crown’ (fig. 19D). The larvae isolated from relatively freshly hatched first instar host larvae are similar to those isolated from the host eggs (fig. 19B–F). However, the parasitoid larvae isolated from older first instar host larvae, are slightly larger, about 500 μm long, and more ‘swollen’, about 190–200 μm wide (fig. 18D), with the ‘caudal crown’ hidden below wider caudal segments (fig. 18F, arrowed), and the caudal end bearing the transparent (thus hardly visible under light microscopy) caudal filaments originating from the dark ‘caudal formation’ (fig. 18E, F, fr). Occasionaly, the larva may be seen when inside the host's body (fig. 18B). No spiracular openings were found, but a reticulum of vesicles was found in its body (similar to the next instar, see below).
Second instar larva
The second instar larva of E. sparetus is also hymenopteriform and pale (fig. 22A), and similar to that of E. thomsonianus. This stage may be identified as the second instar through the location of the skin of the first instar larva (fig. 21A). The second instar larva is about 0.7–0.8 mm long, 0.2–0.4 mm wide, with 13 body segments discernible under SEM, without serrations on segments, and with a widened head bearing poorly sclerotized mandibles which are about 0.02–0.03 μm long (fig. 20). The caudal end of the larva also possesses the dark ‘caudal formation’ continued as transparent (thus discernible only during their motion) caudal filaments (fig. 20C, fr, cf). No spiracular openings were found in this instar, but a reticulum of vesicles is discernible in its body under light microscopy (fig. 20D,E, arrowed). These vesicles may probably be regarded as a vestigial tracheal system.
Final instar larva (figs 22B–F, 23)
The final instar larva of E. sparetus is nearly identical to that of E. thomsonianus: 3.4–3.7 mm, with swollen body consisting of 13 distinct segments and a peculiar head with robust mandibles and traceable sensoria, similar to those of the larva of E. thomsonianus. The epistome is recognizable in the final instar skin sliding from the pupa during the moult (fig. 24A,B).
Final instar larva of Entedon sparetus. A, B, habitus (A, lateral view; B, ventral view); C, D, head close-up; E, F, mouth area, close-up (see text for abbreviations).
Pupa of Entedon sparetus. A, B, the moment of moulting from the prepupa, ventral view: A, habitus; B, head area with larval skin displaying mouthparts (white arrow); C, habitus; D, head close-up. Black arrow, pupal cuticle.
Pupa
Entedon sparetus also pupates inside the hosts' food plant, within the host's stem mine, near the remnants of the final instar larva of the host. The pupa is black, obtect, with distinct outlines of head, mesosoma, metasoma, wings, legs, antennae and mouthparts (fig. 24C,D). The average length of the entire pupa is 3.2 mm, of head 1.0 mm, of mesosoma 1.4 mm, of metasoma 1.6 mm.
Discussion
Species concepts in the sparetus group
Identification of chalcidoid wasps is generally considered a difficult field of expertise because of the minute body size of these insects. However, larger body size does not always facilitate solving taxonomic problems in Chalcidoidea. The sparetus-group of Entedon is such a case. Specific identity may be determined with high probability only for E. thomsonianus, E. zerovae, E. tobiasi and E. lucasi owing to the possession of distinct morphological markers (antennal and wing peculiarities) in adults of these species. However, E. sparetus and E. cardui, according to the concepts provided above, may represent complexes of sibling species, as suggested by differences in size and host associations of their morphotypes. However, in view of the gradual character of the morphological differences between these morphotypes and the identity of the ‘large’ and ‘small’ morphs of E. sparetus on molecular data demonstrated earlier (Gumovsky & Boyadzhiev, Reference Gumovsky and Boyadzhiev2003), I avoid affording them species ranks.
Egg-larval parasitism may also constitute a background for morphological variability in species of the sparetus-group. If the host records are confirmed, the species of this group are associated with various beetles in different hosts' food plants, and therefore the size of adult may vary not only proportionally to the normal size of the final instar larva of the host, but also resultant from the different dietary peculiarities of the larvae of the same host in different regions. Naturally, the size of the final instar larva (and thus the size and body proportions of the adult parasitoid) is more dependent on environmental conditions than the size of the host egg that is attacked by the parasitoids.
However, these considerations are provisional and require experimentally–based verification: it is hoped that the present research will encourage further studies in this field.
Egg-larval parasitism in Entedon
Egg-larval parasitism is quite peculiar if compared with other types of larval endoparasitism, because it involves a transition between two developmental stages of the host. This also suggests an incorporation of the parasitoid larva into the host's embryo, with minimal impact to the latter and then co-existence with the developing host, under its immune response. In the braconid genus Chelonus the host's immune system is apparently not suppressed by the egg-larval parasitoid, but encapsulation of the parasitoid larvae is prevented by the injected venom containing symbiotic viruses (Lanzrein et al., Reference Lanzrein, Pfister-Wilhelm, Wyler, Trenczek and Stettler1998), which, however, are not known to occur in chalcidoid parasitoids.
Egg-larval parasitism has been considered as a rare or exceptional biology in Entedon (mostly E. ergias has been mentioned as an example), whereas larval parasitism is assumed as the ground-plan biology of the genus (Bouček, Reference Bouček1988; Schauff, Reference Schauff1988, Reference Schauff1991). However, at least three other species (namely E. rumicis Graham, E. pharnus Walker and E. subfumatus Erdös/E. philiscus Walker) were recorded by Fisher (Reference Fisher1970) as egg-larval parasitoids. Apart from the above-mentioned species, oviposition behaviour has been reported for four species of Entedon (E. cioni Thomson, E. cionobius Thomson, E. zanara Walker (Gumovsky, Reference Gumovsky1997b) and E. sylvestris Szelényi (Gumovsky, Reference Gumovsky and Boyadzhiev2006), which are larval parasitoids. So, the assumption of the wide representation of larval parasitism was likely based on the host stage from which the final instar parasitoid larva emerges, but not the stage attacked. Up to now, among the 10 species of Entedon for which the parasitized host stage is known, six are egg-larval and four are larval parasitoids (table 1).
Entedon species, for which attacked host stage is known.
One of benefits of egg-larval parasitism for Entedon species is the possibility of attacking the most unprotected developmental stage of their hosts, which otherwise are nearly inaccessible for the parasitoids. The relatively short ovipositor (1.5–2.5 mm long) does not allow the females of Entedon to reach the weevil larvae, especially when they are in the inner layers of the hosts' food plant. Apart from the remote location, the young larvae of Lixus and Gymnetron weevils actively resist attacker and crawl along their stem-mines when disturbed.
Also, parasitism on relatively large hosts allows the final instar larvae of the parasitoids to accumulate considerable biomass to produce relatively large-sized adults: the sparetus-group includes the largest Entedon species (up to 8.8 mm, whereas the average size is about 3.5 mm in the genus). Since insects require flight muscles constituting at least 12–16% of their total mass and flight performance increases proportionally to this percentage (Marden, Reference Marden2000), a larger body size may directly facilitate the capacity for flight. Apart from muscular mass, the wing size may correlate with flexural stiffness of the wing (Combes & Daniel, Reference Combes and Daniel2003) and improved flight efficiency, in its turn, probably facilitates host-searching (timing and distribution capacities), because insect flight speeds at least tend to scale positively with body mass (Dudley, Reference Dudley2000).
Another benefit of the egg-larval strategy is the decreasing risk of multiparasitism. Fisher (Reference Fisher1970) mentioned that some weevil larvae consumed by ectoparasitoid larvae of Eurytoma curculionum Mayr (Eurytomidae) in stems of Rumex spp., contained Entedon larvae inside. Success of oviposition by E. curculionum depended on the possibility of reaching the host larva. Successful oviposition was observed when the only barrier between the adult parasitoid and the weevil larva was the thin cap of tissue at the end of the weevil's exit tunnel. However, when the weevil larva was situated deeper in the pith of the hosts' food plant stem, females of E. curculionum failed to oviposit (Fisher, Reference Fisher1970).
Similar relationships were found by the author in the insect guild on mullein, where E. curculionum attacks the same host (G. asellus) as E. sparetus. However, when the most active period of host searching by E. curculionum was observed (mid–end of June), E. sparetus had already reached its final instar larval or pupal stage, which occured in the central pith of the host-plant stem (fig. 22F). This suggests that generally the final instar larvae of E. sparetus kill their hosts and emerge before the mature weevil larvae commence their exit tunnels, and thus are hardly reachable by larval ectoparasitoids (e.g. E. curculionum). Therefore early pupation inside the hosts' food plants with relatively thick and rigid stems may apparently decrease the risk of rivalry with and a loss of progeny from various multiparasitoids and hyperparasitoids. And, taking into account that the young host larvae are isolated from the females of Entedon by thick stem walls, the attack of the host eggs is possibly the only way to the following early pupation for these parasitoids.
Larval morphology of Entedon
The morphology of the larval stages of E. sparetus and E. thomsonianus are similar to that of E. sylvestris, the larval endoparasitoid of Ceutorhynchus sisymbrii Dieckman (Gumovsky, Reference Gumovsky and Boyadzhiev2006). The descriptions and diagrammatic illustrations of the first instar larvae of other Entedon species, for which the larvae have been studied (all egg-larval parasitoids), also suggest similar morphology (e.g. the ‘caudal crown’ and peculiar cranium of the first instar; Beaver, Reference Beaver1966; Fisher, Reference Fisher1970). The descriptions of the first instar larvae of other Entedoninae (e.g. of Mestocharis bimacularis (Dalman) (Jackson, Reference Jackson1964) or Edovum puttleri Grissell, (Laudonia & Viggiani, Reference Laudonia and Viggiani1986) do not refer to any structures similar to the ‘caudal crown’ of Entedon, so this spinuous edge of the terminal abdominal segment may be considered as a putative synapomorphy for the genus.
It is hard to establish the main function of the ‘caudal crown’, but at least it is used by the larva of E. sparetus for destroying the partition in the egg of G. asellus and further penetration of the host's embryo. The first instar larvae of some other egg-larval parasitoids (e.g. the braconid Chelonus inunctus) also have caudal appendages facilitating penetration of the host's embryo (Kaeslin et al., Reference Kaeslin, Wehrle, Grossniklaus-Burgin, Wyler, Guggisberg, Schittny and Lanzrein2005). The proposed function of the ‘caudal crown’ as a penetration device in egg-larval parasitoids and its possible uselessness in larval parasitoids (e.g. E. sylvestris) may support a preposition of egg-larval parasitism as the ancestral biology of Entedon. Another possible function of the caudal ‘crown’ is the destruction of the rival conspecific larvae in cases of superparasitism. Beaver (Reference Beaver1966) proposed that superparasitic larvae are mechanically eliminated (e.g. by mandibles). However, Fisher (Reference Erdös1970) suggested that they are eliminated by physiological suppression (e.g. lack of oxygen), in particular because of the lack of obvious scars or melanized areas on dead supernumerary larvae. In the present studies, the dead rival larvae of E. thomsonianus were evidently damaged (fig. 7D), which suggests mechanical elimination of superparasitic larvae in this species.
The head sensoria of Entedon larvae are also topographically similar in all studied species, and are most distinct in the first instar larva. However, unlike many ectoparasitoid first instar larvae (e.g. Eurytoma spp. (Tormos et al., Reference Tormos, Asís, Gayubo and Martín2004), these sensoria are represented by palpi, not setae (except a pair of pharyngeal setae). The mandibles of the first instar larva are acute, sickle-shaped, and the larva was observed to use the mandibles when crawling onto or within the host embryo (either for nutritional or mechanical purposes, fig. 17C), and it also probably kills the rivals with its mandibles.
The second larval instar is less well sclerotized, which causes difficulties in studying its external morphology with SEM (figs 12, 21). The only sclerotized body parts which were recognized in the second instar of E. sparetus are the mandibles, which are nearly transparent, subtriangular (fig. 20F), only slightly mobile, and look vestigial in comparison with the mandibles of the first and final instars. The final instar larva of Entedon looks typical for Hymenoptera, better sclerotized than the two previous instars, head sensoria are not as clear as in the first instar, but are nevertheless quite recognizable, and mandibles are strong and massive.
The non-sclerotized caudal appendages and associated structures of the larvae of Entedon are rather peculiar. These appendages are represented in the first and second instars by the transparent filaments aggregated into a knob-like or spherical (the caudal bladder) formation, which emanates from a dense dark formation in the caudal end of the body of the larva. Beaver (Reference Beaver1966) mentioned the possession of ‘a series of papillae’ in the second instar larva of E. ergias, which are probably the same as the above-mentioned ‘caudal filaments’. The first and second instar larvae of Entedon are probably apneustic, albeit at least the second instar has a tracheal system represented by gas-filled vesicles (fig. 20D,E). One of the probable functions of the enigmatic ‘caudal formation’ is that it accumulates the oxygen absorbed by the caudal filaments and then transfers it as a gas into the concealed tracheal reticulum.
Respiratory caudal appendages (the filaments and the bladder) have not yet been recorded for Chalcidoidea, but are known in endoparasitic larvae of some Braconidae (Chapman, Reference Chapman1982). In such larvae these caudal appendages represent a protrusion of the hind gut. The possession of these formations in Chalcidoidea and their apparent respiratory function suggest obvious parallel adaptations of different groups of parasitic Hymenoptera to endoparasitism.
Summarizing the above, it is concluded that the larvae of Entedon have peculiar habitus, easily recognizable and hardly different morphologically between larval and egg-larval parasitoids. The immature development of the studied egg-larval parasitoids of the genus may be generalized as follows. The female lays her egg into the host weevil egg and the newly hatched parasitoid larva either is passively incorporated into the developing host embryo, or penetrates the developed embryo using its cranial (e.g. mandibles) and caudal (e.g. the ‘crown’ of spines) sclerotized appendages The larva moves actively inside the developing embryo, kills conspecific rivals (if any), and keeps its original shape until the first instar host larva begins active feeding. Then the first instar parasitoid larva also begins feeding: it enlarges to nearly twice its original size, and the caudal ‘crown’ becomes hardly visible; meanwhile the caudal filaments (? everted hind gut) connected with the ‘caudal formation’ appears. The caudal filaments are retained by the poorly sclerotized second instar, which stays in the stable environment of the host's haemocel, in contrast with the more sclerotized first and the final instars, which change media (egg yolk/larval haemolymph and host haemolymph/open surroundings). The parasitoid larva stays at its second instar until the host reaches its final instar, and just then moults to its final instar. The tracheal system of the parasitoid larva develops probably in yet earlier instars, though they appear apneustic. The tracheal system of the final instar larva is fully formed, bearing seven pairs of spiracles. Probably the locking mechanism of the spiracles provide its isolation from the liquid environment at earlier moments of the development of the final instar larva, while it is still in the host's haemolymph.
Proper affiliation of the first instar Entedon larvae to any types proposed by earlier authors (Parker, Reference Parker1924; Parker & Thomson, Reference Parker and Thompson1925), is problematic, because these groups are rather vaguely defined. In general, the larvae discussed above fit the concept of the group of apneustic ‘endophagous forms which float freely in the body cavity of the host egg, larva, or pupa’ and ‘possess … caudal appendage or tail’ (Parker & Thomson, Reference Parker and Thompson1925) or the ‘Groupe V’ of Parker (Reference Parker1924). However, the sclerotized caudal appendage (the ‘crown’) in Entedon larvae is different from the appendages illustrated by these authors for the representatives of the proposed larval groups. Also, the head sensoria are mentioned as ‘incospicuous and few in number’ (Parker & Thomson, Reference Parker and Thompson1925), but this may be an impression resulting from light microscopy, whereas SEM reveals full set of head sensoria at least for Entedon larvae.
Acknowledgements
This paper represents a part of the revisionary research on the genus Entedon by the author, supported by the European Commission's FPVII SYNTHESYS Project (GB-TAF-535) and by the SFFR (State Fund of Fundamental Research, Ukraine, grant 05.07/00078), and in part by the SCOPES Institutional Partnership project ‘Emphasising Classical and Conservation Biological Control in Research and Teaching’ (7 IP 65648) and DAAD (Deutcher Akademischer Austauschdienst – German Academic Exchange Service, grant 322 – A/04/15867). The author appreciates the help of J. Noyes (BMNH), G. Delvare (CIRAD), J.-Y. Rasplus (INRA), R. Danielsson (LUZM), C. Villemant (MNHN), P. Boyadzhiev (PU), C. v. Achterberg (RMNH), L. Zombori, J. Papp (TMA), S.Tryapitsyn (UCRC), V.A. Trjapitzin, S.A. Belokobylsky, V.I. Tobias, D.R. Kasparyan (ZISP) for their help in obtaining the materials for this study and is also grateful to Drs S. Gorb, U. Wegst, S. Kühnemann and J. Siewert (MPI) for their help in conducting the SEM work, and Dr. V. Fursov (Schamalhausen institute of Zoology) for his assistance in obtaining some of the digital photos. The author is also much indebted to Dr M. Shaw (National Museums of Scotland) for reviewing the manuscript.
