IOBC wprs Bulletin Vol 22(9), 1999

Chemical ecology of the European apple sawfly, Hoplocampa testudinea

Jean-Luc Boevé

Département d'Entomologie, Institut Royal des Sciences Naturelles de Belgique, Bruxelles, Belgium

Abstract - The European apple sawfly can be a pest in apple orchards. Their larvae feed within apple fruitlets. The chemical ecology of the sawfly larvae was studied by field observations, laboratory bio-assays and chemical analyses (GLC, GC-MS) at two levels. Firstly, a glandular secretion is emitted by full-grown larvae. It is mainly composed of acetogenins. For migrating larvae which pupate in the soil, the secretion should act as an allomone, defending the larvae against epigaeic predators such as ants. For younger larvae, a mechanical protection is offered by the fruitlet. Secondly, fruitlets emit mainly terpenoids. Among them, (E,E)-a-farnesene, (E)-b-ocimene and another terpenoid are emitted in significantly larger amounts when the fruitlets are infested by the sawfly larvae, compared to healthy fruitlets. This increased emission of volatiles could act as a synomone, attracting specialised parasitoids. The ichneumonid Lathrolestes ensator, for instance, is highly specialised and lays eggs when young sawfly larvae occur in the field. On the other hand, preliminary results on volatiles emitted by adults of the sawfly are given; the possible occurrence of pheromones being evoked. It is concluded that an integrated control of the European apple sawfly in orchards probably will not be achieved by increasing a predatory pressure on the sawfly larvae. It should rather consist of favouring the parasitoids.

Key words - chemical ecology, orchard, Hoplocampa testudinea, Tenthredinidae, Hymenoptera, defensive allomones, synomones, pheromones, parasitoids, integrated control


European apple sawfly

The genus Hoplocampa (Hymenoptera, Tenthredinidae) is well-known in orchards. Each one of their species feed at larval stage within the fruits of one genus of the Rosaceae, such as H. testudinea, the European apple sawfly, on apple. The economic impact and related aspects of this sawfly species is abundantly documented (e.g., Alford 1984).

Predators and sawfly defensive allomones

If one opens an apple infested by a full-grown sawfly larva, it is easy to perceive a strong odour. This odour is due to a series of epidermal glands located in the abdomen. The glands can be devaginated when a larva is disturbed. Glands were dissected from larvae, extracted in pentane, and analysed by GLC and GC-MS. A series of chemical compounds could be identified from the extracts (Boevé et al. 1997). Major compounds are acetogenins (i.e., biosynthesised by accumulation of acetate units). They can possess groups such as an alcohol, acid or aldehyde. They can be esterified and possess one or more double bonds.

The abdominal glands are present at all larval instars. However, they vary in relative size during these instars. L5 possess over-proportional large glands (Boevé et al. 1997). It is also this instar where a secretion is detected by head-space analysis of larvae alive (Boevé et al. 1996). Larvae were disturbed during volatile collection. In the same conditions, no compound is detected in L3-4. The lack of an evident function for glands of young instars was shown by laboratory bio-assays. Single sawfly larvae were placed with a potential predator, namely workers of the ant Myrmica rubra. L3 and L4 were attacked, whereas L5 not at all. Thus, defensive allomones emitted by abdominal glands are present in the sawfly, but obvious at last larval instar only.

In the field, L5 will leave the apple in order to pupate in the soil. During this migration, the larva risks to encounter and to be attacked by epigaeic arthropods, birds, etc. The raison d'être of defensive allomones, only present at L5 (see above), could be that the environment changes concurrently for this larval instar. In other words, the larva would be protected mechanically by the apple at L1-4, whereas chemically by a glandular secretion at L5 (Boevé et al. 1997).

In an orchard, field observations on sawfly larvae brought some nuances to the predator-prey relationship which is supposed to exist with ants as predators of sawfly larvae (Boevé et al. 1997). Workers of Lasius niger were observed commonly, feeding on frass (as did also other insects such as elaterid adults). Indeed, sawfly larvae generally accumulate their frass outside the apple, by depositing it all around the opening previously self-made by the larva and by which it entered the fruit. The "crater" of frass is obviously attractive to the ants. Most probably, ants profit from the "open door" created by sawfly larvae, having access to apple pulp as a food resource. An attack or aggressive behaviour of the ants was never observed. A sawfly larva was even observed once, half outside an apple with two ant workers feeding on its body surface covered by fruit liquid. However, effective predator-prey relationships might occur, but especially during larval migration (see above). As far as known from the literature, there are no specialist predator species that would be adapted to prey upon larvae of the European apple sawfly.

Parasitoids and potential synomones from apple

The ichneumonid Lathrolestes ensator is a specialised endoparasitoid of the sawfly larva. Are chemical cues used by the parasitoid female when searching a host for egg laying? It is not likely that the larval secretion of the sawfly (see above) could constitute such a chemical cue. On the one hand, by comparing the phenology of the parasitoid to the one of the sawfly, it is clear that parasitoid females occur when sawfly eggs and young larvae are present (Zijp & Blommers 1993). On the other hand, a larval secretion is easily detected only at L5 (see above). Thus, in natural conditions, a glandular secretion is simply not available as a chemical cue to the parasitoid for host finding (at a medium distance range). It is not excluded, however, that minute amounts of glandular secretions from L1-2 might play a role in host recognition (at a very short distance range).

A discovery during the last decade is that host finding might occur indirectly. A parasitoid can be attracted to volatiles emitted by the plant, this emission being induced by the feeding phytophagous insect (Turlings et al. 1990). The hypothesis of such an induction in apples was studied. In the laboratory and in the field, the chemical composition of apple volatiles was analysed by searching for possible differences between sawfly infested and healthy apples. The volatiles show a quantitative change for three terpenoids, (E,E)-a-farnesene, (E)-b-ocimene and an unidentified terpenoid; these three terpenoids being emitted in larger amounts by infested apples (Boevé et al. 1996). There are at least two arguments indicating the vegetal origin of the three terpenoids. Firstly, healthy apples already emit them. Secondly, the sawfly larval secretion can not be at the source of them, since not detected by the analysis of isolated glands (Boevé et al. 1997). Thus, the plant actually responds by emitting given volatiles in larger amounts, as induced by the feeding behaviour of a sawfly larva. The plant response is not believed to be systemic, since pairs of infested and healthy apples were picked on single trees (Boevé et al. 1996).

The plant response described above could constitute a synomonal signal by which the parasitoid E. ensator would find more efficiently its host. Such a hypothesis remains to be verified. In preliminary bio-assays using parasitoids in a wind tunnel, we were not able to obtain satisfying results (D. Babendreier, J.L.M. Steidle, J.-L. Boevé).

Sawfly pheromones?

The chemical ecology of the European apple sawfly, to be presented more thoroughly, should not be restricted to the larvae. But much less is known on the subject for other instars. Adults occur generally in (the vicinity of) flowers where they feed on nectar and pollen, reproduce and lay eggs. After the oviposition into a blossom, plant exudates seem to be recognised by females which will avoid further egg layings into that blossom (Roitberg & Prokopy 1984). The consequence of this phenomenon should be a more homogeneous exploitation of food resources for the larvae.

No sexual pheromone is known for the European apple sawfly. Some of their behaviours suggest, however, the presence of such a pheromone. Miles (1932) describes for adults maintained experimentally on caged apple seedlings: "... the males ran rapidly up the trunk, dropped to the soil and run round waving their antennae continually and occasionally exserting the genitalia. This was repeated several times while the female rested more or less motionless on the soil beneath the tree. Gradually she became active, running for short distances with rapidly vibrating antennae. When in close proximity to the female the male became exceedingly excited, running up on her back and beating her head with his antennae, and at the same time exserting and retracting the genitalia."


Table 1. Relative amounts of components detected in the head-space of adults of the European apple sawfly.





2.7 ± 1.1

1.8 ± 0.1

14.3 ± 0.2


10.6 ± 0.8
Pentacosadiene ?

11.4 ± 1.4

11.5 ± 2.9

39.2 ± 1.4


33.6 ± 1.7

4.6 ± 0.0

4.4 ± 0.3
Heptacosadiene ?

3.0 ± 0.3

4.1 ± 1.0

10.9 ± 1.0


13.2 ± 2.2

1.9 ± 0.3

2.1 ± 0.4

2.0 ± 0.4


3.5 ± 0.1

1.6 ± 0.5


3.6 ± 0.3

As a first attempt to analyse the chemical composition of volatiles emitted by the sawfly adults, I analysed by gas chromatography - mass spectrometry (GC-MS) the head-space of males and females (head-space performed by P. Witzgall). Three pairs (i.e., from both sexes) of samples were analysed on an apolar capillary column (DB-5, 30 m) with following temperature program: 2 min at 60 °C, then up to 280 °C with 3 °C min-1, and followed by 30 min at 280°C. Compounds detected were in majority saturated and unsaturated hydrocarbons with 23, 25 and 27 carbon atoms (Table 1). The relative amount of each component in a sample is rather constant. By considering such relative amounts, there is a trend for some volatiles to show quantitative differences between volatiles emitted by both sexes, but these differences are weak. Thus, these first results do not constitute a clear indication of the presence of a sexual pheromone. In the case, for instance, of the yellowheaded spruce sawfly, the pheromonally active components are alkadienes which are 100 times more abundant in females than in males (Bartelt et al. 1984). A more thoroughly research on the volatiles emitted by adults of the European apple sawfly is in progress in our laboratory, with external collaborations.

The amount of a component was calculated as a percentage of the total amount of detected components in the respective sample. Components are given here as an average value ± of these percentages (N=3, two times) and when representing > 1%. They are listed in their elution order. (C.A.) Compared abundance between sexes: any of the three percentages for females was (>) higher or (<) lower than any of the three percentages for males. (?) Identification not sure. Position of double bonds are still not determined.

Some perspectives on an integrated control

On the base of the knowledge gathered on the chemical ecology of the sawfly, is it possible to propose some recommendations about the control of this orchard pest? It seems that at larval stage, the introduction in an orchard of predators would probably have any effect on the sawfly population. Indeed, no specialist predator is known to attack larvae. Thus, potential predators will encounter rarely young larvae, due to the mechanical protection offered by the apple, as well as the full-grown larvae, due to the chemical defence offered by their glandular secretion. In contrast, the specialised parasitoid L. ensator might be an efficient control agent. It has been used in Canada in the frame of a biological control program (Babendreier & Kuhlmann 1998). A biological control by using pathogens (i.e., fungi and nematodes) is promising as well (e.g., Vincent & Bélair 1992). Finally, I would like to remember here the ecological inadequacy of spraying insecticides against sawfly adults. Since sawfly adults occur during the period of apple flowering, insecticides will have profound negative effects on bees and other pollinating insects which are, of course beneficial, or better said, essential in the life of an orchard.


P. Witzgall provided head-space samples of sawfly adults. These samples were analysed in the laboratory of Prof. M. Hilker to whom I am indebted.



Alford DV (1984) A colour Atlas of Fruit Pests. Wolfe Publishers, London

Babendreier D, Kuhlmann U (1998) European apple sawfly (Hoplocampa testudinea Klug), Annual Report 1997-1998. CABI Bioscience Centre, Switzerland

Bartelt RJ, Krick TP, Jones RL (1984) Cuticular hydrocarbons of the yellowheaded spruce sawfly, Pikonema alaskensis. Ins Biochem 14, 209-213

Boevé J-L, Lengwiler U, Tollsten L, Dorn S, Turlings T (1996) Volatiles emitted by apple fruitlets infested by the European apple sawfly. Phytochem 42, 373-381

Boevé J-L, Gfeller H, Schlunegger UP, Francke W (1997) The secretion of the ventral glands in Hoplocampa sawfly larvae. Bioch Syst Ecol 25, 195-201

Miles HW (1932) On the biology of the apple sawfly, Hoplocampa testudinea Klug. Ann appl Biol 19, 420-431

Turlings TCJ, Tumlinson JH, Lewis WJ (1990) Exploitation of herbivore-induced plant odors by host seeking parasitic wasps. Science 250, 1251-1253

Vincent C, Bélair G (1992) Biocontrol of the apple sawfly, Hoplocampa testudinea, with entomogenous nematodes. Entomophaga 37, 575-582

Zijp JP, Blommers L (1993) Lathrolestes ensator, a parasitoid of the apple sawfly. Proc Exper Appl Entomol, Nederl Entomol Veren Amsterdam 4, 237-242

table of contents