In the Veluwe, larvae of A. limacodes have, at least in recent years, occurred commonly on Quercus robur growing under a wide range of conditions, from quite dense woodland (where it is also common on Fagus sylvatica) to open heathland (Fig. 1), with 2004, 2005, 2008 and 2014 being years of particularly high abundance. Larvae, which exhibit some colour polymorphism not connected with parasitism (Figs 2, 3; final instar), were present in the field for a long period, from late May until early October, in variable stages of growth. By mid-September many A. limacodes larvae had become fully grown and spun their cocoons, and virtually all remaining were in their final instars. Attack from other parasitoids (in the area the campoplegine ichneumonid Phobocampe alticollis (Thomson), the microgastrine braconid Dolichogenidea lacteicolor (Viereck), and the rogadine braconid Triraphis tricolor (Wesmael)) was by then finished, so the A. limacodes larvae that remained were unparasitized. From late September onwards host larvae were increasingly found bearing eggs of Sphinctus serotinus (the first one seen on 25.ix.2005), but almost exclusively on relatively isolated Q. robur trees in open sunny heathland situations — in particular at Edese Heide (Fig. 1) where, as late September and the first half of October wore on, the majority of A. limacodes larvae were found to bear one or more eggs (Figs 2, 3) of this strictly solitary parasitoid. Good numbers of the host occured also in more shaded woodland in the Veluwe, on both Q. robur and F. sylvatica, but the parasitoid was essentially absent from this situation and none were found on hosts feeding on the latter tree (which was lacking in the open, xerothermic, areas favoured by the parasitoid).

Figure 1. 

Biotope at Edese Heide with Quercus robur.

Figures 2, 3. 

Final instar larva of Apoda limacodes with 2 one egg 3 four eggs of Sphinctus serotinus.

In the wild eggs of S. serotinus were seen on a large number of hosts, and appeared to be positioned more or less at random on the exposed surface of the host’s body, including positions well down at the sides and near both anterior and posterior ends, though more central and dorsal positions were commonest. Many instances of two, three and even four eggs on a single host were seen (Fig. 3).

The adult female is about 14 mm long and adopts a fully-resting stance (i. e. when asleep; Figs 4, 5) that is rather unusual in Ichneumonidae, with wings spread away from the metasoma, but which clearly accentuates its mimicry of vespid wasps (perhaps Eumenes in particular). This exposure of the metasoma is maintained while the parasitoid is actively searching on foot (Figs 6–8), though at such times the antennae are directed forwards and sideways and the wing tips are not flattened to the substrate. The experimental female fed avidly on dilute honey, but showed no tendency to host feed. Even when deprived of hosts, eggs were neither dumped nor carried externally at the apex of the ovipositor.

Figures 4–7. 

Sphinctus serotinus, 4, 5 asleep 6, 7 in motion.

The female was first offered hosts six days after emergence, when she started to show weak interest in host traces (chewed leaf edges, and the slime trails by which the host moves). This interest became stronger on subsequent days. When a host was encountered (Figs 8, 9) it was often antennated with one antenna along each side of the host’s long axis, which often (not always) caused the host to rock rhythmically from side to side (a behaviour that is provoked by various kinds of disturbance, including from another A. limacodes larva, and did not unduly delay the female parasitoid although a slight deterrent effect was evident). The position of the antennae suggested that the host was being measured: all hosts offered were final instar, but it was clear from all interactions and ovipositions that large (14 mm or more, female) hosts were much preferred to smaller ones (9–12 mm, male), though some ovipositions on the smaller size range did occur. Except that the antennae were curled downwards during actual oviposition and in the subsequent brief period of inspection (but usually not in contact with anything), they were not used for further investigation or manipulation. Altogether 15 ovipositions were obtained, interspersed with rejections that were sometimes later reversed, on days 6 (2 ovipositions), 8 (1), 9 (1), 13 (7) and 14 (4) of this single female’s adult life. When the female crawled, often rather slowly, onto the host (Fig. 10) she sometimes did so transversely and kept the hind legs on the substrate (Fig. 11), but usually she approached longitudinally and then used all six legs to perch upon it (Fig. 12). This difference accounts for the variable egg positions seen (respectively lateral and dorsal). The ovipositor was then brought forward between her legs to pierce the host integument, with the sheaths supporting the ovipositor. The main body of the egg issued from the genital opening caudal end first (Fig. 12) and slid down the outside of the ovipositor, with its anchor travelling down the shaft internally (Fig. 13) until oviposition was completed, leaving the egg on the surface of the host but its stalk implanted (Fig. 14). This process took about 10–15 (exceptionally 20) seconds, during which the host was usually completely quiescent, but appeared not to have suffered temporary paralysis. The female usually paused, apparently reviewing her achievement, for a few seconds after oviposition then, invariably, left purposefully — usually by flying away. A host already bearing an egg was accepted without any sign of hesitation or inhibition, so this prompt post-oviposition reflex departure seems to be the only way in which self-superparasitism is avoided. Normally the female fed avidly on honey:water between ovipositions, and there was sometimes a clear cocking and jerking motion at the apex of the metasoma suggestive of a new egg being moved down the oviduct to a position in the genital opening ready for oviposition as a host was approached (Fig. 15). The anchored egg is usually light brown at the time of oviposition and remains so until it hatches, but if more than two ovipositions occured in close succession the subsequent eggs were very pale yellowish green (Figs 16, 17), darkening to the usual light brown over a few hours. After the available hosts had run out the female was denied access to food, and died (undoubtedly of starvation) nine days later.

Figures 8–11. 

Sphinctus serotinus, approaching and accepting Apoda limacodes larva.

Figures 12–14. 

Sphinctus serotinus, oviposition sequence.

Figure 15. 

Sphinctus serotinus, priming for oviposition.

Figures 16–17. 

Sphinctus serotinus, laying an immature egg.

The matt, rather roughened egg of S. serotinus is 1.1 mm long and anchored into the host integument (Figs 18–20), as is typical for Tryphoninae (Bennett 2015). In this case the egg is fastened by a clear flange-like expansion at the distal end of its stalk. The anchor extends only into the already secreted cuticle which is abnormally thick in Apoda(stained pink in Figs 19 and 20), and not through it into the deeper cell layers involved in secretion of new cuticle. The stalk of the egg appears to stain in a different way from the other collagens involved in the structure of the egg, but the stains used do not adequately characterise it (C. Gielis, pers. comm.). Normally the egg of S. serotinus does not hatch until the host has made its cocoon, but under artificially high humidity (closed plastic box) it may do so. The chorion splits (Fig. 21) at the anterior end (furthest from the anchor) but the setose first instar larva remains partly (caudally) in the eggshell as it starts to feed in the late autumn to early winter, as evidenced by a blackened feeding lesion (Figs 22, 23), with presumed urate patches already evident. The larva remains in this position, feeding slowly on the cocooned host prepupa, through the winter but by the second week of May a larva in a cocoon opened for observation had left the eggshell behind, grown considerably and moulted at least once, now moving freely across the body of the still living host prepupa and making new feeding lesions. The parasitoid larva was still setose and now very liberally dotted with presumed urate deposits. (In the milder early spring of 2007 the parasitoid larva in a cocoon opened on 19 March in the Netherlands had already reached this state (Fig. 24).) It is unclear at what point the supernumaries (in cases of superparasitism) are eliminated, or how, but presumably this cannot take place while the larvae are still confined by their eggshells. Growth continued slowly through the spring and early part of summer (in Edinburgh in 2006 the weather was particularly cold, with daytime temperatures in the rearing shed seldom exceeding 10 °C during May). The larva had moulted at least twice, probably three or possibly even four times, leaving the rather robust cast skins split along a dorsal line (Fig. 25), before dying about half grown around 11.vi.2006 — probably largely a result of excessive disturbance, but no doubt exacerbated by cold weather.

Figures 18–20. 

Sphinctus serotinus, egg on host 18 as laid 19, 20 sectioned.

Figure 21. 

Sphinctus serotinus, egg chorion splitting.

Figures 22–23. 

Sphinctus serotinus, first instar larva feeding.

Figures 24, 25. 

Sphinctus serotinus, intermediate instars 25 showing exuviae.

A further host cocoon (ca 8.5 mm long) containing S. serotinus was opened on 25.vii.2006, and by this date the parasitoid had completed its feeding and was prepupal. (The Edinburgh weather had been much warmer during late June and July, with temperatures regularly in the low 20s°C.) The host cocoon now contained an area of coarse white silk, at its midlength, isolating the host remains between the host cocoon wall and a frail white diaphanous cocoon enclosing the S. serotinus prepupa (Figs 26–28). The isolating wall of silk was positioned against the venter of the parasitoid’s cocoon, within which faeces were visible. The Sphinctus cocoon was easily removed (Fig. 28) and opened, but the prepupa (whose eye spots were clearly evident) could not be withdrawn without damage because it was anchored firmly, via its anus, to its faecal material by a tough viscous/elastic strand of brown material capable of supporting much more than its weight (Fig. 29). This individual was kept under observation and pupated on 1.viii.2006. Unlike the prepupa, the fresh pupa is easily removed from its cocoon (Fig. 30). One pupa was brought indoors on 25.viii.2006 after which it gradually developed colour (Figs 31–33), the mesoscutum being the first area to darken and the legs among the last.

Figures 26–29. 

Sphinctus serotinus, prepupa 26, 27 in situ 27, 29 removed from host cocoon.

Figures 30–33. 

Sphinctus serotinus, pupa 30 fresh 31–33 progressively developing adult colour.

On one host that had received an egg the parasitoid failed to develop beyond its first instar, and the host then progressed to the (pharate) adult stage, indicating that no venom component that interferes with the host’s endocrine system had been injected to control host development.