Species of Adialytus Förster, 1862 (Hymenoptera, Braconidae, Aphidiinae) in Iran: taxonomic notes and tritrophic associations

Abstract The species of Adialytus Förster in Iran are taxonomically studied and new data on distribution and host associations are presented. The existence of a species complex, in the case of Adialytus ambiguus (Haliday), and the morphological variability in commonly used taxonomic characters has been discussed. In total, four valid species belonging to the genus Adialytus including Adialytus ambiguus (Haliday), Adialytus salicaphis (Fitch), Adialytus thelaxis (Starý) and Adialytus veronicaecola (Starý) have been identified and recorded from Iran. Also, we recognized two additional phenotypes: “Adialytus arvicola” (Starý) and “Adialytus cf. ambiguus” (Haliday). These phenotypes and Adialytus veronicaecola are newly recorded from Iran in association with Sipha and Aphis species, respectively. An illustrated key for identification of the species and two variable phenotypes is presented.

Here we review the species of Adialytus in Iran, together with new data on their host associations and distribution. In addition, the possible existence of species complexes and morphological variability within genus are discussed.

Material and methods
Samples of different host plants including wild and cultivated trees, shrubs and herbs bearing the aphid colonies were gently cut off and placed inside the semi-transparent plastic boxes. The collected material were subsequently transferred to the laboratory and kept under controlled conditions with temperature range of 24-28°C and RH: 60±5%, for 2-3 weeks until the emergence of the adult parasitoids. The rearing boxes were inspected daily to prevent the activity of emerging hyperparasitoids. Once detected, they were immediately removed from the rearing boxes. The emerged parasitoids were also carefully collected using an aspirator and dropped into 75% ethanol for further examination. A few specimens from each sample were carefully dissected and mounted in slides using a Hoyer medium. The ratio measurements were based on these slide-mounted specimens using an ocular micrometer. Additional material from European and central Asian countries were also used for comparison of the morphological variation. The characters of flagellar segments, clypeus, fore wing, first metasomal tergit (=petiole) and female genitalia were used for comparison and differentiation of the species, as well as to find the reliable characters for identification key. The external morphology was studied using a NIKON Eclips E200 microscope equipped with a SONY DSC digital camera.
The morphological terminology for parasitoids used in this paper follows Sharkey and Wharton (1997) and for the aphids Remaudière and Remaudière (1997), respec-tively. The nomenclature of host plants was based on Flora of Iran (Ghahreman 1978(Ghahreman -2006. The specimens were deposited in the collection of the first author. Abbreviations of the names of provinces (Fig 1)

Results
Four valid species of the genus Adialytus, as well as two additional phenotypes: "A. arvicola" (Starý) and "Adialytus cf. ambiguus" (Haliday) (Table 1) were collected and identified in association with 14 aphid species on 15 host plants. Many specimens of A. ambiguus (Haliday) were inconsistently different from examined specimens which originated in other countries. We categorized these specimens as "Adialytus cf. ambiguus". Adialytus veronicaecola (Starý) and "A. arvicola" (Starý) are newly recorded from Iran. The latter species was reared from Sipha aphids which were also the specific hosts for A. ambiguus. We found significant differences between the A. ambiguus and "A. arvicola" phenotype, based on the characters of fore wing, flagellar segments, hind legs, petiole (Table 2) and coloration. Additionally, a comparison with type specimens of A. arvicola from the Czech Republic (Starý 1961a) clearly confirmed the existence of strong differences.

Aphis craccivora Koch
Adialytus veronicaecola (Starý) Aphis gossypii Glover Aphis sp.   5D). Flagellar segments covered with prevalently semi-erect setae which are equal to diameter of segment. Flagellomere 1 bearing 3-4 longitudinal placodes (Fig 2D). Hind femur covered with prevalently semierected setae (Fig 4D)  First metasomal tergite (petiole) short, 1.9-2.1X as long as wide at spiracles ( Fig 5F). Flagellar segments covered with adpressed setae which are distinctly shorter than diameter of segment. Flagellomere 1 with 0-1 longitudinal placode ( Fig 2F). Hind femur covered with short adpressed setae (Fig. 4F)  Comments: This species is closely related to other parasitoids of Sipha aphids, in its elongated ovipositor sheath ( Fig 6A) and triangular shape of petiole which bears anterior and spiracular tubercles (Fig 5A). It can be differentiated from other species in having an extremely long vein R1 (= metacarpus) (Fig 3A). The hind femur and tibia are covered with both short and prevalently erect long setae (Fig. 4A). Comments. The specimens normally run to A. ambiguus according to the general characters of the first metasomal tergite (Fig 5B), ovipositor sheath (Fig 6B), the flagellomeres ( Fig 2B) and the setae on the hind femur ( Fig 4B). It can be differentiated from A. ambiguus by having the shorter vein R1 that is 0.9-1.1 × as long as pterostigma that does not reach the apex of the fore wing (Fig 3B). It can be separated from A. arvicola (Fig 3C), by its longer vein R1. Comments. Generally this species can be confused with other Adialytus species on Sipha aphids, but it is immediately distinguishable by its very short vein R1 (0.7-0.8 × as long as pterostigma) (Fig 3C). Also, its petiole has much stronger anterior and spiracular tubercles (Fig 5C). Most of the metasoma is yellowish, while in other Adialytus species it is uniformly brown to dark brown.

Comments.
A. salicaphis differs from other congeners in having very elongated first metasomal tergite (petiole) (Fig 5D), and short and dense marginal setae of the fore wing (Fig 3D). It can also be differentiated from A. arvicola by the number of longitudinal placodes on flagellomere 1 (3-5 in A. salicaphis vs. 0-1 in A. arvicola). The specimens of A salicaphis associated with Salix spp., especially those reared from Chaitophorus salijaponicus niger on Salix alba, were slightly different from the specimens that reared from Chaitophorus spp. on Populus. The major differences were the lesser number of setae on the clypeus (4-5 vs. 8-10), lesser longitudinal placodes on the first flagellomere (1-2 vs. 3-5) and predominantly adpressed and short setae on the flagellomeres and hind femur compared with the long semi-erect to erect setae among the short setae in specimens from Populus. (Starý, 1961) http://species-id.net/wiki/Adialytus_thelaxis Figs  Comments. This species can be easily separated from other congeners by having mainly erect long setae on the flagellomeres (Fig 2E) and the hind femur (Fig 4E). The setae on the postero-dorsal aspect of petiole are similar (Fig 5E). Additionally, A. thelaxis is the only species with a sharply pointed ovipositor sheath (Fig 6E). (Starý, 1978) http://species-id.net/wiki/Adialytus_veronicaecola Figs 2F, 3F, 4F, 5F, 6F

Adialytus veronicaecola
Lysiphlebus veronicaecola Starý, 1978: 528-529. Comments. This species is unique in that it was reared from Aphis species. According to the general characters of the fore wing (Fig 3F), petiole or first metasomal tergite ( Fig 5F) and the ovipositor sheath ( Fig 6F) it is closely related to A. salicaphis from which it can be immediately distinguished in having prevalently short and adpressed setae on the flagellomeres (Fig 2F) and hind femur (Fig 4F). It can also be differenti- ated from A. salicaphis by having lesser longitudinal placodes on flagellomeres 1 and 2 (0-1 in A. veronicaecola vis 3-5 in A. salicaphis). In addition, A. veronicaecola differs from the other species in having a stout ovipositor sheath with a strongly convex postero-dorsal outline (Fig 6F).

Discussion
In a biological aspect, the host range pattern of Adialytus species can be used as an appropriate criterion supporting its generic status as separate from, but closely related to the genus Lysiphlebus Förster. Species of the genus Lysiphlebus are mostly parasitoids of the genera Aphis and Brachycaudus (Starý 1999, 2006, Starý et al. 1998) but, exceptionally, include some other aphid groups such as Metopeurum (Macrosiphini) in the case of Lysiphlebus hirticornis Mackauer (Mackauer 1960, Starý 1961b). On the other hand, about half of the Adialytus species are associated with different aphid subfamilies consisting of Thelaxinae and Chaitophorinae, while others attack Aphis Juchnevič 1978, Pike et al. 2000) and Dysaphis (Starý and Rakauskas 1979). It can be suggested here that the members of the latter group are biologically more closely related to the genus Lysiphlebus. The Nearctic species, A. fuscicornis (Ashmead), a parasitoid of Aphis species (Pike et al. 2000) tends also to resemble morphologically the Lysiphlebus species except for its more reduced wing venation. Among the recorded species, A. veronicaecola manifests two major diagnostic characters including the stout ovipositor sheath and prevalently adpressed setae on the flagellar segments and hind legs. Other species have a more elongated ovipositor sheath and different types of chaetotaxy bearing both semi- erect and erect setae. In contrast, A. balticus Starý has erect and perpendicular setae on the flagellomeres. The habitat and host associations of this species on the root collar of Anthriscus sp. (Starý and Rakauskas 1979) might be the reason for having perpendicular setae on the flagellomeres as well as the reduction in length of the segments (Starý et al. 1998). So, we lack clear diagnostic characters for separation of these groups given the present state of knowledge.
Adialytus veronicaecola was originally described from Kazakhstan Juchnevič 1978, Starý 1979). The new evidence also supports the original distribution of this species in central Asia, as well as host specificity on Aphis species. Three other Aphis species are added to the list of its host, of which A. craccivora and A. gossypii are of economic importance. "Adialytus arvicola" phenotype is also newly recorded from Iran, but the earlier records are most probably cited under the synonymy with A. ambiguus. While it can be considered as the first evidence of the existence of a species complex in the case of A. ambiguus, it sounds to be a rather specific parasitoid of Sipha aphids of various subgenera including Atheroides Haliday, Chaetosiphoniella and Sipha Passerini (Mackauer 1965), "A. arvicola" seems to be restricted to the later subgenus (Starý 1961a, b). On the other hand, the separation of these two species, as well as the intermediate "Adialytus cf. ambiguus", cannot be clearly justified without molecular analyses, since they were collected from almost the same host aphids at the studied area. Generally, A. ambiguus seems to be a very rare species in Iran, and it might be replaced by the geographical species/subspecies manifesting significant morphological differences. The most important diagnostic character is in the pattern of the venation of the fore wing.
It is yet unclear which "phenotype" of A. ambiguus was used for the phylogenetic analyses (Kambhampati et al. 2000, Sanchis et al. 2000 but, nominally, the genus Adialytus was classified as a paraphyletic group due to the arrangement of A. ambiguus inside the genus Lysiphlebus (Sanchis et al. 2000). On the other hand, "A. arvicola" was grouped with the other Adialytus species, separated from Lysiphlebus spp. (Kambhampati et al. 2000). Differences among the specimens of A. salicaphis associated with Salix and Populus seem to be an intra-specific variation together with some other characters including the length/width ratio of petiole and carination of the propodeum (see Takada 1979). Shujauddin (1978) also found the same difference between the Indian and European specimens. These variations should be considered in further taxonomical studies.

Conclusion
In general, identification of the Adialytus species merely based on the morphological characters is rather difficult, since they are very similar and even these characters may be contributed to intraspecific variation. Nevertheless, the host range patterns which are mostly specific can be greatly useful for separation of most species, excluding taxa in the A. ambiguus species complex, which have almost the same host range. Further investigations based on the geometric morphometric analysis, as well as suitable mo-lecular markers might reveal the exact identity of the above-mentioned taxa and status "A. arvicola" and "Adialytus cf. ambiguus". Furthermore, a re-classification at a tribal level is necessary to reconstruct the relationships between two groups of Adialytus species and their position compared to the genus Lysiphlebus.