Trail marking and following by larvae of the small ermine moth Yponomeuta cagnagellus

The importance of chemical cues in insect behaviour is well established (Bell & Cardé, 1984). The best known examples include the sex pheromones of butterflies and moths, and the aggregation pheromones of bark beetles. In eusocial insects (bees, wasps, ants, and termites) pheromones are widely used to maintain the organization of the colony. Many of these species produce chemical markers (trail pheromones) and deposit them on terrestrial trails that lead to food sources or nesting sites. Trail pheromones may also serve as cues in home range orientation and can facilitate migration of colonies (Attygalle & Morgan, 1985). However, trail following is not confined to eusocial species and is, for instance, also found in the Lepidoptera. Fabre (1922) already described the striking following behaviour of the procession caterpillar Thaumetopoea pityocampa (Denis & Schiffermüller). To explain his observations, he stressed the importance of tactile stimuli from the silken treads that these (and other) caterpillars produce, and that can be followed. Although this argument still holds today (Chapter 2), it has become clear that, in addition to silk, chemical trail markers may also be important in the social behaviour of caterpillars (Fitzgerald & Peterson, 1988). The best documented examples are found in the Lasiocampidae. In Eriogaster lanestris trail marking was demonstrated by Weyh & Maschwitz (1978), and in the genus Malacosoma chemical trails convey information about the quality of a feeding site, and recruit other larvae to these places (Fitzgerald & Peterson, 1983: Peterson, 1988). In spite of these thoroughly studied examples, knowledge about chemical communication in caterpillars is limited, and mainly restricted to the Lasiocampidae. To gain more insight in trail following and trail marking in the Lepidoptera it is necessary to study this behaviour in other families.This thesis focuses on caterpillars of small ermine moths, members of the genus Yponomeuta. This group has been studied in the context of a long term multi-disciplinary research program on speciation, and host plant selection is thought to be an important element in the speciation process (Wiebes, 1976). Food preferences of larvae are related to host preferences of female moths. This makes it interesting to see whether speciation is accompanied by interspecific differences in larval trails to feeding sites.There are additional reasons to investigate trail marking and following in the Lepidoptera. Caterpillars have been advocated as model systems in the study of feeding behaviour (Schultz, 1983; Schoonhoven 1987), in part because their behaviour is relatively simple. A caterpillars primary function, gathering as much food as possible, is not complicated by tasks such as mate finding or taking care of offspring. In addition the sensory system is limited. Only about 90 chemosensory cells function in translating chemical messages from the environment into signals for the central nervous system (Albert, 1980; Devitt & Smith, 1982; Schoonhoven,1987). In spite of this restricted number of input channels, caterpillars can display striking food preferences, and will often die from starvation, rather than accept a non-host plant. Such behaviour, together with the possibility of tracing sensory connections into the central nervous system (Kent & Hildebrand, 1987) make caterpillars a good choice for studying the relationship between neurophysiology and behaviour.A further point of interest is the possible integration of sensory information. The receptors for sex pheromones are in general separated from those that perceive stimuli associated with food. Trail pheromones of caterpillars bear a close relation to food finding. Therefore these insects may have an integrated receptor system that responds to both food and trail pheromone stimuli.The objectives of this study were (1) to determine whether trail pheromones are employed by Yponomeuta and, if so, whether they differ in different species, (2) To identify receptors responsible for pheromone detection, and (3) to determine whether these receptors operate in an integrated way with receptors for food perception. However, as a first step in the analysis of this system basic knowledge must be gained about trail following behaviour itself, the chemicals involved, the senses used and the oecological context in which it functions. These questions form the main topics of this thesis.Most experiments were performed on larvae of Y. cagnagellus (Hübner) (Fig. 1). This species is common in the Netherlands and suitable for behavioural studies. In addition, the caterpillars are gregarious throughout their development, suggesting that they may use a trail marker. Malacosoma caterpillars were used in some experiments to allow comparison to a species with well defined trail following behaviour, and an identified trail pheromone.Outline of the thesisTrail following in Y. cagnagellusThe study begins by asking whether Y. cagnagellus in fact exhibits trail following behaviour (chapter 2). Two-choice tests on filter paper Y-mazes show clearly that this is the case. In addition it is demonstrated that a tactile component of the trail (the silk) can be used as a cue. Y. cagnagellus does not discriminate between its own trails and those of 5 other Yponomeuta species, but does prefer its own trails over those of M. neustria . This lack of species specificity within the genus is, in contrast to sex pheromones, not uncommon for trail pheromones, possibly because the relationship between mating success and the signal is indirect.A chemical markerThe existence of trail following behaviour does not by itself prove that a chemical marker is involved. Evidence for the presence of a chemical signal is presented in chapter 4. The marker appears to be water soluble, and highly stable under laboratory conditions. Behavioural responses to extracts from several glands and body parts show that the marker is present in the labial glands (the silk gland) only. Therefore, the marker is probably secreted with the silk.The receptors involved in trail following behaviourChapter 3 describes the sensory organ used for the perception of the trail. In this chapter a comparison is made with the American tent caterpillar Malacosoma americanum, a known trail follower (Fitzgerald, 1976). Chemoreceptors in caterpillars are located on the antennae, the maxillary palps, the galea and on the inner side of the labrum (Fig. 1). Systematic removal of various relevant structures shows that the maxillary palps are necessary for the detection of the trail in both M.americanum and Y.cagnagellus. Since the source of the trail marker, as well as its chemical composition, differs between the two species, this is most likely an example of convergence of chemoreceptor function.Although the maxillary palps contain olfactory as well as gustatory receptors, the trail markers seem to be perceived only by contact chemoreception. This follows from the observation that trails covered with fine nylon mesh do not elicit any response. Moreover the long lifetime and stability of the markers, suggest that they have a low volatility.Electrophysiology of the maxillary palpAlthough the palps house a considerable fraction of the sensory equipment of a caterpillar (30-40 cells, more than 1/3 of the total), only very little is known about these organs. Therefore an electrophysiological survey of the chemoreceptors was conducted (Chapter 5). Because the sensilla are too small for tip recording, gustatory stimuli were applied to the whole tip of the palp. Electrical activity of only a few cells at a time was recorded with a glass microcapillary electrode. To aid analysis, a computer program was developed (Chapter 7). Following the ideas of van Drongelen et al. (1980), the program was designed to be highly interactive and to function as both as a display- and manipulation tool.Plant volatiles were used as olfactory stimuli (terpenoids and C6 fatty acid derivatives or 'green odours', Visser & Avé, 1978). These were chosen in part on the basis of the results from a dynamic headspace analysis (Cole, 1980) of Euonymus europaeus , the host of Y. cagnagellus . Silk extracts and salt solutions were employed as gustatory stimuli. Evidence was found for the existence of two groups of olfactory receptor cells, sensitive to (E)-2-hexenal and hexanal (aldehydes) or to (Z)-3-hexen-1-ol and 1-hexanol (alcohols). Receptors responsive to the silk extracts (and probably to the trail pheromone) were also identified. These cells do not show the degree of specificity typical of cells specialized for lepidopteran sex pheromones but, rather, resemble the generally more broadly tuned receptors for food components.The results from this and the preceding chapters strongly suggest the existence of a chemical trail marker in Y. cagnagellus , secreted with the silk and detected by contact-chemosensory neurons housed in the maxillary palps.Oecological and evolutionary aspectsChapter 6 addresses the the oecological and evolutionary relevance of such a signal. In eusocial insects as well as in Malacosoma , trail pheromones are often used to recruit siblings to high quality feeding sites (Peterson, 1988). In Y. cagnagellus this does not happen, but field observations have shown that a groups of caterpillars moves its silken nest over considerable distance, on average four times during development. The trail marker could help to maintain gregariousness during these migrations. Thus, it is of interest to ask whether gregariousness is advantageous. While many authors have discussed the benefits of larval aggregation (e.g. Tsubaki, 1981; Fitzgerald & Peterson, 1988; Weaver, 1988), gregariousness may also be associated with distinct disadvantages, for instance those arising from competition for food (Charnov et al., 1976). In chapter 6 a simple evolutionary model is presented to analyze the influence of these conflicting parameters on the evolutionary stability of gregarious behaviour. One result from this study is that it would be informative to classify larval behaviour in terms of the time of which larvae switch from gregariousness to solitary food searching