The British species of Enicospilus ( Hymenoptera : Ichneumonidae : Ophioninae )

The nine British and Irish species of Enicospilus are revised, mapped and an identifi cation key provided. One species, Enicospilus myricae sp. nov., is described as new; Enicospilus merdarius (Gravenhorst, 1829) is a senior synonym of E. tournieri (Vollenhoven, 1879) syn. nov.; the only available name for E. merdarius auctt. is Enicospilus adustus (Haller, 1885) stat. rev., and a neotype is designated for Ophion adustus Haller, 1885. Enicospilus cerebrator Aubert, 1969 and E. repentinus (Holmgren, 1860) are newly recorded from Britain. Some host data are available for eight of the nine species.


Introduction
Enicospilus Stephens, 1835 is a distinctive genus of primarily nocturnal parasitoids of relatively large Lepidoptera larvae.The genus is immensely species-rich in the tropics (Gauld & Mitchell 1978, 1981;Gauld 1988) but only small numbers of species are found in north temperate regions.However, despite the small number of species in Britain (only five were listed by Fitton et al. 1978), there has been much confusion over the limits and identities of these species.Under the auspices of the first author's nocturnal Ichneumonoidea recording scheme (http://nocturnalichs.myspecies.info/),we have gathered host, distribution and phenology data on the British Enicospilus species and clarified some taxonomic issues.Considering the size of the fauna, now increased to nine species, the taxonomic problems were surprisingly extensive; these issues are summarised in the "Taxonomy of British Enicospilus" section below.
In Europe, Enicospilus species are easily recognised as the only Ophioninae with strongly narrowed mandibles, a large glabrous patch in the fore wing discosubmarginal cell (frequently with detached sclerites) and fore wing vein Rs+2r partly thickened (see Fig. 1).Stauropoctonus bombycivorus (Gravenhorst, 1829), which might be confused with Enicospilus because they share twisted mandibles, lacks the occipital carina and has fore wing vein Rs+2r abruptly angled near its origin on the pterostigma, in both respects unlike Enicospilus.Ophioninae can be recognised by the discosubmarginal cell extending beyond fore wing vein 2m-cu and the presence of a dark line (a "spurious vein") ventrally in the fore wing anal cell (Fig. 1).In common with many other nocturnal ichneumonoids, they are large (or very large), almost always mostly testaceous, with large eyes and ocelli, and long antennae.
Hosts of the British Enicospilus fall into two categories: Lasiocampidae, in the case of the very large E. inflexus (Ratzeburg, 1844) and E. undulatus (Gravenhorst, 1829), and low-feeding noctuids for the remaining species (although E. repentinus (Holmgren, 1860) has not been reared).As with most Ophioninae that have been reared, the host is probably attacked as a late instar larva, and habitually killed as it prepares to pupate; however, few details of the biology of British Enicospilus have been studied.All but one British species seem to be univoltine, with rather narrow periods of flight activity.There are too few data to establish the limits of host ranges, but at least two species (E.merdarius (Gravenhorst, 1829) and E. undulatus) have very restricted host ranges that may reflect limited diversity of hosts in their particular habitats.Habitat specialisation may be important in limiting distribution, rather than absolute host taxon specificity.For example, E. myricae sp.nov.has only been collected in Myrica Linnaeus, 1753 bogs and wet woodland, where it has been reared from a common and widespread host caterpillar which is present in a range of habitats.Interestingly, females of all British species (at least, those with good sample sizes) far outnumber males in light trap samples, whereas both sexes are reared in approximately equal numbers.While this might largely reflect the difference in longevity between the sexes, males may also be rather less nocturnal, as is certainly the case with some other ophionines, such as Eremotylus marginatus (Jurine, 1807) and Ophion ventricosus (Gravenhorst, 1829), males of which can be collected flying around trees in the daytime, whereas females are more strictly nocturnal (pers.obs.), although in these two cases the species are partly patterned with black.

Material and methods
The distribution maps for Britain and Ireland are based on far more records than any preceding maps dealing with the British parasitoid fauna; nevertheless, the ranges shown are very incomplete and suffer from several sources of recording bias.Hopefully these maps, which illustrate broad patterns of distribution, will encourage entomologists to record Enicospilus species in the many blank areas, so that eventually it will be possible to map changes in distributions over time.Maps were plotted using DMAP, developed by Alan Morton (www.dmap.co.uk).The main sources of specimen data are the collections of the Natural History Museum, London (BMNH), and the National Museums of Scotland, Edinburgh (NMS).These collections have been considerably enriched in recent years by donations from many entomologists, particularly moth trappers who have kindly sent their ichneumonoid catches to us.For the geographic spread of their light trap network, the Rothamsted light trap survey (see Woiwod & Harrington 1994;Harrington & Woiwod 2007) was a particularly important source of specimens; we include records from Rothamsted light traps on the Channel Islands, although these are not faunistically a part of Britain.We have also seen specimens from several private collections as well as the collections of British Entomological and Natural History Society, World Museum Liverpool and Cambrige University Museum of Zoology.Following other papers cataloguing the collections of NMS (e.g.Schwarz & Shaw 1998, which explains the rationale), we give the numbers of specimens present in NMS (and BMNH in this case) and list the Vice Counties from which they have been recorded (from all data), as well as cataloguing non-British material in NMS.The full British dataset is available via the National Biodiversity Network Gateway (https://data.nbn.org.uk/), the full dataset via the Natural History Museum's Data Portal (http://data.nhm.ac.uk/) and in Supplementary File."Unsexed" usually refers to specimens which now lack the metasoma.Gauld (1988Gauld ( , 1991)).Sclerites and some wing veins and cells are labelled on Fig. 1; wing length is measured as the greatest distance from the apex of the tegula to the wing tip.Gauld & Mitchell (1978, 1981) and Gauld (1988) employ several wing venation indices but, as these are uninformative in distinguishing closely related British species, they are not detailed here.We include ranges of number of flagellar segments, based on British specimens, rather than the total antennal segments, i.e. we exclude the scape and pedicel from the counts.Photographs were taken using a Canon EOS 450D digital camera attached to a Leica MZ12, with images stacked using Helicon Focus.Whole insect photos were taken by Harry Taylor at the BMNH.

Taxonomy of British Enicospilus
There have been no identification keys to British Enicospilus since Gauld's (1973) key and update (Gauld 1974).Unfortunately, these works contained significant misidentifications and lumped some species together.This is not surprising, as Gauld had access to rather small sample sizes and relied heavily on the number and shape of fore wing sclerites, which are of great use in Enicospilus taxonomy but are, unfortunately, almost identical in five of the British species.There has never been a thorough revision of European Enicospilus species, which is reflected in some frequent misunderstandings regarding species names and limits, although Viktorov's (1957) key is very useful.In Britain, Enicospilus can be divided into three species-groups, based on the sclerites in the fore wing discosubmarginal cell: E. inflexus and E. undulatus entirely lack sclerites (and have been referred to the genus Allocamptus Förster, 1869 by some authors); E. merdarius (= Ophion tournieri Vollenhoven, 1879) and E. repentinus have a welldefined proximal sclerite, with the central sclerite either absent or transparent; and the remaining five species (the ramidulus species-group) have both the proximal and central sclerites pigmented.There has been confusion in each of these species-groups, although it is within the ramidulus complex that species are most morphogically similar and hence have been persistently confused.Gauld (1974) separated the very similar E. inflexus (Ratzeburg, 1844) and E. undulatus (Gravenhorst, 1829), that he had previously (Gauld 1973) confounded under the name E. undulatus; and Viktorov (1957) had already separated E. repentinus and E. tournieri (but see below), which Gauld (1973) had confused by identifying British specimens of E. merdarius (= tournieri) as E. repentinus, whereas the true E. repentinus had not been found in Britain at that time.

Status and taxonomy
As explained above, unfortunately the species generally known as Enicospilus merdarius (citations can be traced through Yu et al. 2012, including the inevitable gross misidentifications) is not conspecific with the lectotype, as designated by Fitton (1984).The next available name and, surprisingly, the only name currently placed in synonymy with E. merdarius, is Ophion adustus Haller, 1885(Horstmann 1997).
Unfortunately, the application of E. adustus is not straightforward either; there is no published type depository for Ophion adustus and it appears that nobody has ever referred to a type, if any existed, since Haller (1885) described the species.This is a fairly widespread but apparently uncommon species, reared from Noctuidae that feed on low vegetation.We have seen only one reared specimen, from an uncertain host.
Flight time (non-reared material) July-September, with one outlying November date, but 90% of specimens are concentrated in July-August.

Host
The only host record is of one specimen labelled as having been reared (M.R. Britton) from either Blepharita adusta (Esper) or Lacanobia oleracea (Linnaeus) (both Noctuidae) (NMS).
Despite the coincidence of the species name, Haller's specimens were not reared but were caught in the daytime, basking on vegetation (Haller 1885).There are no obvious habitat preferences discernible from the collection data.

Remarks
Identification is relatively straightforward but not all material of "E.merdarius" from light traps was retained until it was realised that E. cerebrator had been overlooked in Britain.Enicospilus adustus is a large, testaceous species, lacking dark markings except, sometimes, for discolouration of the metasomal sternites and laterotergites.Morphologically it is very similar to E. combustus and E. ramidulus, which each have distinctive colour characters.The long antennae (58-69 flagellar segments in British specimens, usually in the range of 60-65, modal value 63; 62 flagellar segments in the neotype) with elongate preapical flagellar segments serve to distinguish E. adustus from E. cerebrator, together with the form of the scutellum and the slightly wider temples.Enicospilus myricae sp.nov.differs in several respects (see notes under that species) and the antenna is intermediate in length between E. adustus and E. cerebrator.Some European specimens of E. adustus are noticeably larger, with a more pronounced posterior ridge to the scutellum and there may be additional undescribed species in this complex.In both E. adustus and E. cerebrator the anterior transverse carina of the propodeum varies from complete to largely absent.whose larvae feed in seedheads or on flowers.There are three specimens reared from Hecatera dysodea (Denis & Schiffermüller, 1775), which has a restricted range in England and south Wales (Hill et al. 2010) and was extinct in Britain for many years; it is fairly frequently the case that there are good numbers of parasitoid rearings from rare hosts (e.g.Enicospilus merdarius), which are targeted by entomologists in preference to the more widespread host species (note the paucity of host records for E. adustus and E. combustus).

Flight time (non-reared material)
May-August, with 51% having been collected in July; one specimen is labelled as "xi".From a series in NMS collected at Dungeness (coll.C.W. Plant) from mid-May to late July, it seems that E. cerebrator is plurivoltine (at least bivoltine), in contrast to other British Enicospilus, although May specimens have not been seen from any other locality.

Remarks
Enicospilus cerebrator is a smaller species than E. adustus, with more strongly narrowed temples and a rather distinctive scutellum.The antennal flagellum is shorter than in E. adustus or E. myricae sp.nov.(51-56 flagellar segments, modal value 53), with stouter preapical flagellar segments than in E. adustus; the scutellum appears more parallel-sided, broader posteriorly, bordered posteriorly by a slightly raised ridge and with the sides more abruptly curved posteriorly than in similar species; the surface of the scutellum is more matt than in similar species; the male parameres are square-ended (Fig. 8B) compared to the more tapering parameres of E. adustus and E. myricae sp.nov.

Status
A distinctive species of mainly southern distribution.It has not been found in Scotland.Reared from Melanchra persicariae (Linnaeus, 1761) (Noctuidae).
Flight time (non-reared material) July-October, with the majority in August-September; is on the wing later in the year than E. adustus, E. cerebrator and E. ramidulus.

Remarks
Although very similar in general morphology to E. adustus and E. ramidulus, the colour pattern of E. combustus is distinctive, with the mesosoma extensively black, and the antennae have more flagellar segments (62-70 flagellar segments, modal number 65), especially compared to E. ramidulus, which shares a black-tipped metasoma.Other than colour pattern, though, there seem to be no reliable morphological distinctions from E. adustus, apart from a greater number of antennal segments (but with an overlapping range).Viktorov, 1957 Enicospilus cruciator Viktorov, 1957: 205.

Remarks
This species is very similar to E. merdarius (see notes under E. merdarius) and it is possible that some other continental specimens identified as E. merdarius (or E. tournieri) in BMNH and NMS in fact belong to this species.(Ratzeburg, 1844) Figs 3A, 12, 20A
Flight time (non-reared material) June-September, with the majority in August.

Remarks
Along with E. undulatus, with which it has frequently been confused, E. inflexus belongs to a distinctive group of species (in older literature sometimes referred to as the genus Allocamptus) that lack fore wing sclerites, are very large and have a strongly sinuous fore wing vein Rs+2r.Compared to E. undulatus, E. inflexus has more narrowed temples, giving it a less buccate head, but it is otherwise very similar.B. E. repentinus (Holmgren, 1860).(Gravenhorst, 1829) Figs 1, 2C, 11A, 13A, 19A

Status
As described in the "Taxonomy of British Enicospilus" section above, the lectotype male of Ophion merdarius is a specimen of the species usually called E. tournieri.The (probably non-British) female paralectotype is a specimen of Enicospilus adustus (i.e., the usual interpretation of the name), so the choice of lectotype was unfortunate.We have not examined type material of Ophion tournieri or Enicospilus contributus as these types cannot be located; instead we have followed the synonymies (under tournieri) of Aubert (1962Aubert ( , 1964) ) and Viktorov (1957).The type of E. contributus should be in ZIN but could not be located (A. Khalaim, pers. comm.).The whereabouts of the type male of O. tournieri is a mystery; Townes et al. (1965) report the type depository as the Muséum national d'Histoire naturelle, Paris, but it cannot be found there (A.Touret-Alby, pers.comm.) and it seems unlikely when most of the Vollenhoven's types were deposited in Dutch collections.There is also no trace of a type in Naturalis, Leiden (F.Bakker, pers.comm.), which includes the former Amsterdam collections.The type locality of Switzerland makes it likely that O. tournieri is a synonym of E. merdarius rather than E. cruciator, described from Turkmenistan and apparently more of a species of hot, dry climates (judging by published records and the collections of BMNH).Restricted to a few coastal sites in England and Scotland.Only reared from Agrotis ripae (Hübner, 1823) (Noctuidae) (7 rearings), which inhabits the strandlines of sandy beaches and is very localised.The apparent host specificity of E. merdarius may be a result of the restricted noctuid fauna in its habitat.Gauld (1973) recorded E. repentinus as a British species but, based on his description of the species as being coastal, and the lack of true E. repentinus in the BMNH collections until recently, it seems he was describing E. merdarius; in fact, Sperring (1952) had already published on E. tournieri as a British species, with a host record (specimens in BMNH and BENHS).Additional material in NMS BULGARIA: 6 ♀♀, 1 ♂, Aksakovo (C.W. Plant) (NMS).
The lectotype ♂ was supposedly collected in Netley, Shropshire (Fitton 1984), but this locality has been ascribed to most of the British material sent by F.W. Hope to J.L.C. Gravenhorst and seems very unlikely to be the actual collection locality for this sand dune inhabitant: entomologists of that period seemed often to name their home town, presumably to identify specimens as theirs, on what might otherwise be taken as data labels (which were, to say the least, unfashionable at the time).

Remarks
Most similar in the British fauna to E. repentinus but larger (52-58 flagellar segments, n = 10, modal value 52) and with distinct differences in fore wing sclerites and venation; also the propodeum has rather different sculpture, with the rugosity more raised and thus making it less shiny than in E. repentinus.Unlike in E. repentinus, there are some rather vaguely defined pale yellow patches on the mesosoma (Fig. 13A).The non-British Enicospilus cruciator is similar and the two species may well be confused in collections.Judging by Viktorov's (1957) key and photographs of a female and male of the type series, E. cruciator differs from E. merdarius in the longer, less narrowed temples (in dorsal view of the head) and the larger ocellar-ocular gap.

Diagnosis
Separated from E. adustus and E. cerebrator by the more rounded temples, wider ocellar-ocular space (especially compared to E. cerebrator), more obvious dorsal "dip" on the first tergite and by the distinctly different aedeagus of the male.

Etymology
Named after the association with the distinctive habitat of Myrica gale-dominated bog, from which this species was reared.

Description Female
There is very little variation between specimens; variation is covered in the description.Fore wing length 11-13.5 mm.Antenna with 57-59 flagellar segments (n = 6) (57 in holotype).1 st flagellomere about 5× as long as apically wide, preapical flagellomere 1.5× as long as wide.Head (Fig. 4C) with distinct gap between lateral edge of stemmaticum and edge of eye, lateral ocellus separated from eye by 0.25× maximum length of ocellus; temples in dorsal view curved, rounded immediately behind eye, then more linearly narrowed, measured in straight line from eye margin to lateral margin of occipital carina, c. 0.8× length of greatest eye width; in lateral view, at level of antennal sockets, gena 0.8× width of eye.Mandible strongly bent, slightly twisted, with curved groove containing long setae, from diameter), shiny.Scutellum (Fig. 5C) shiny, regularly punctate, punctures larger than on mesoscutum and further apart (larger and closer anteriorly); lateral carinae complete to near posterior end of scutellum, indicated around posterior end of scutellum as rugosity/carinulae.Fore wing (Fig. 10) as in other species of E. ramidulus group; glabrous area (fenestra) of discosubmarginal cell extending from proximal sclerite to posterior 0.3 of Rs+2r (along thickened area of vein); proximal sclerite entirely pigmented, approximately triangular with rounded anterior angle, more elongate on distal corner; central sclerite roughly "D"-shaped, pigmented distally, fading to unpigmented, transparent proximally; distal sclerite represented by faintly pigmented line along distal-ventral margin of fenestra; fore wing vein Rs+2r sinuous, uniformly widened along anterior 0.7; 3rs-m 0.45× section of M between 1m-cu and 3rs-m; cu-a slightly to distinctly proximal to Rs&M.Metapleuron shiny, closely punctate.Propodeum with weakly defined central section of anterior transverse carina, anterior of this shiny and superficially punctate, posterior to this entirely reticulate-rugose or sculpture much reduced dorso-laterally.First metasomal tergite (Fig. 6C) with shallow dorsal concavity at anterior 0.45.Second metasomal segment with laterotergite narrow, folded under; third tergite with laterotergite not demarked.

Colour
Uniformly testaceous (Fig. 16), varying from dull orange to a darker, reddish-orange (although probably dependent on preservation), except for black mandibular teeth and varying amounts of dark brown infuscation on the venter of the metasoma from 4th tergite onwards, and apical tergites at most weakly infuscate.Antenna darker apically.Generally slightly darker than E. adustus.

Distribution
Austria, England, Scotland, Wales, as detailed in the list of type material (British distribution in Fig. 18D).
The majority of the few known specimens were collected in Scotland but it is a much more widespread species and it may be that it prefers boggy habitats in which few people collect ichneumonids.One paratype was collected in Monks Wood NNR, an ancient deciduous woodland with a rather rich fauna of fen or bog-associated noctuids.Unlike other British Enicospilus there is a distinct sexual dimorphism in antenna length, as males have more flagellar segments, with no overlap in the small sample size available.(Linnaeus, 1758) Figs 2B, 11B, 18E

Status
A common and widespread species, regular in light traps and rather frequently reared from Noctuidae, particularly of the subfamily Hadeninae.
Flight time (non-reared material) June-September, with the majority in July.

Remarks
Amongst the British Enicospilus species with two discrete, pigmented fore wing sclerites, E. ramidulus is distinctive in that the mesosoma is entirely testaceous and the metasoma apically sharply black, from the 5th or 6th tergite onwards.Structurally very similar to E. adustus and E. combustus, but E. ramidulus has shorter antennae (54-60 flagellar segments, modal value 56) and colour patterns are invariable.According to published records this is a very widely distributed species; however, there are other, similar species in various parts of the world that have been misidentified as E. ramidulus.(Holmgren, 1860) Figs 2D, 11B, 13B, 19B

Status
New to Britain.Found in a few localities in southern England, particularly along the eastern end of the Chilterns; all specimens have been collected in the past 30 years, all but one at light.Previous records of E. repentinus in Britain (e.g., Gauld 1973) refer to E. merdarius (= tournieri).We know of no reliable host records.Aldbury (VC 20), 2, 15 Jul. 2008, 8, 20, 21 Jul. 2010, 23, 25 Jul. 2012  The flight time is basically limited to July, other than one specimen collected at the very end of June and one in August.

Remarks
Smaller than E. merdarius (46-49 flagellar segments, modal value 47 in repentinus), with which it has been confused, lacking both the transparent central sclerite in the discosubmarginal cell and the elongate pigmented strip (distal sclerite) along the distal edge of the glabrous patch in the discosubmarginal cell.There are also subtle differences in the propodeal sculpture, which is less raised and shinier in E. repentinus.The two species are found in very different habitats: mainly calcareous grassland or woodland edges in the case of E. repentinus, sandy coasts in E. merdarius.(Gravenhorst, 1829) Figs 2A, 3B, 20B

Status
A very rarely collected species, found on southern coastal heaths where it has been reared from Lasiocampa trifolii (Denis & Schiffermüller, 1775) (Lasiocampidae).The host is now very local and E. undulatus has not been found in Britain since 1971.Additional material in NMS SPAIN: 1 ♀, Zaragoza, Montes de Torrero, 230 m, 24 May 1998 (G.E.King).

Remarks
Enicospilus inflexus has been separated from E. undulatus on the basis of differences in head shape (Gauld 1974) which seem to be consistent, based on the limited material in BMNH.Although there may be a difference in host use, this is based on only two rearings of E. undulatus, from one place and date; although Lasiocampa trifolii is rather smaller than L. quercus, E. inflexus and E. undulatus do not differ significantly in size.(Haller, 1885).B. E. cerebrator Aubert, 1966. C. E. combustus (Gravenhorst, 1829).D. E. myricae sp.nov.E. E. ramidulus (Linnaeus, 1758).

Fig. 22 .
Fig. 22. Phenology of the five British species of the Enicospilus ramidulus species-group, expressed as proportions of the total, excluding reared specimens.

Fig. 21 .
Fig. 21.Frequency distribution of number of antennal flagellar segments in four species of the Enicospilus ramidulus species-group.