Herbivore-induced plant volatiles and tritrophic interactions: from local to landscape scale

Aartsma, Yavanna


Plants need to defend themselves from attack by herbivorous insects. They can do this directly by producing chemical and structural defences such as toxins and spines, but also indirectly by promoting the effectiveness of enemies of the herbivores. This can be accomplished by providing resources such as nectar or shelter for these enemies, but also by providing information to the enemy on the location of their prey. These interactions form a mechanistic basis of tritrophic interactions: Interactions between plants, herbivores and enemies of these herbivores. Plants can provide natural enemies with information on herbivores by releasing chemical compounds upon damage inflicted by these herbivores. These chemical compounds are called herbivore-induced plant volatiles (HIPVs) and can be used by carnivorous animals to find their prey. Some plant species or varieties produce different HIPV blends than others, which can influence carnivore preference such as parasitic wasps or parasitoids.

Parasitoids are insects that lay their eggs in or on other insects. Their offspring feed from the host insect until the parasitoid completes larval development, which usually results in the death of the host. After emerging as adults and mating, female parasitoids need to find new hosts in their environment. These hosts can be spread heterogeneously, which means the parasitoid needs to disperse to locate the hosts. While foraging for hosts, parasitoids can use HIPVs as information on the identity, quality and abundance of hosts. This process has been well characterized at small spatial scales, but little research has been done on how HIPVs attract parasitoids at larger spatial scales. The spatial distribution of HIPVs can be influenced by a range of aspects of the surrounding environment, such as weather conditions, vegetation structure and interference with chemical compounds from other plants. Little is known on how these habitat characteristics affect the foraging behaviour of parasitoids under field conditions. Furthermore, research that combines plant variation in attractiveness to parasitoids with the effects of habitat characteristics is rare.

The aim of this thesis project was to investigate HIPV-mediated interactions between plants and parasitoids from local to landscape scale, and how plant variation in attraction of parasitoids via HIPVs alters parasitoid foraging behaviour at these spatial scales. I used a tritrophic system of white cabbage, Pieris brassicae (large cabbage white butterfly) and the parasitic wasp Cotesia glomerata. Two cabbage accessions were used, which were a priori known to differ in attractiveness to parasitoids under laboratory conditions and in small-scale field experiments. The accessions Christmas Drumhead is preferred over Badger Shipper.

In Chapter 2, the current state of the literature regarding HIPV-mediated interactions across spatial scales is reviewed. Key knowledge gaps in the use of HIPVs as a long-distance cue by parasitoids are the distance from which they can be perceived, how HIPVs from surrounding vegetation alter the ability of a parasitoid to find their hosts and how parasitoids use HIPVs on the landscape scale.

The spatial scale of parasitoid attraction by two plant accessions that differ in attractiveness was studied in Chapter 3. In an open field experiment, I released parasitoids in an experimental set-up with cabbage plants infested with caterpillars either at a spacing of 10 m or of 20 m. Parasitoids which were released in set-ups with the accession Christmas Drumhead parasitized caterpillars at a similar rate in set-ups of 10 m and 20 m. In set-ups with the less attractive accession, Badger Shipper, parasitism rates decreased dramatically when distance between plants was increased from 10 m to 20 m. Additionally, detailed parasitoid behaviour was studied on a smaller scale (up to 8 m) in a semi-field set-up (tent). Similarly, I found that parasitoids are less able to find plants at larger distances, but that the more attractive accession Christmas Drumhead was found more frequently than the less attractive accession Badger Shipper at larger distances. The experiments show that parasitoids can be attracted to herbivore-infested plants over distances between 10 and 20 m, and that a more attractive host-infested variety is found by parasitoids over longer distances.

Habitat characteristics can influence parasitoid foraging behaviour. More specifically, the number of plants in a patch might affect the apparency of this patch to parasitoids, but also to herbivores, and the presence of another plant species might influence the ability of parasitoids and herbivores to find their food plant. In another field experiment (Chapter 4), I established cabbage plots of the two accessions which varied in plot size (small or large) and which had either no border, or a border of black mustard plants, Brassica nigra, a close relative of cabbage. Throughout the season, I investigated whether experimentally introduced Pieris brassicae caterpillars were parasitized and counted naturally occurring Pieris spp caterpillars in the plots. Abundance of the caterpillar Pieris rapae was not affected by cabbage accession or plot size, and only later in the season fewer caterpillars were found in bordered plots than non-bordered plots. Parasitism rates of experimentally introduced caterpillars were also not affected by plot size. The border only affected parasitism rates on the less attractive accession Badger Shipper, where fewer caterpillars were parasitized. The more attractive accession Christmas Drumhead had equal parasitism rates in plots with or without a border. This accession also had higher incidence of superparasitism (where the same or multiple Cotesia glomerata parasitized a caterpillar multiple times). Results show that herbivores and parasitoids responded differently to variation in habitat characteristics and plant accession, which might alter the outcomes of tritrophic interactions on longer timescales. Accessions less apparent to parasitoids might provide herbivores with a refuge space, where parasitism risk is lower. For parasitoids, more apparent plants may be easier to find in complex vegetation structure.

Parasitoid foraging behaviour may be affected when other insects, which are not hosts (non-hosts), are feeding from the same or neighbouring plants and alter HIPV emissions from these plants. In Chapter 5, I studied parasitoid foraging behaviour in environments with different distributions of plant and herbivores on these plants. In a semi-field tent set-up, plants of both accessions were mixed and infested with hosts or hosts and non-hosts. In most combinations of host or non-host caterpillars on the plants, the more attractive accession Christmas Drumhead had higher parasitism rates than Badger Shipper. However, when both hosts and non-hosts were present on Badger Shipper and only hosts on Christmas Drumhead, overall parasitism rates in the tent decreased. In a wind tunnel experiment, parasitoid preference was studied in more detail. The accession Christmas Drumhead was overall the preferred accession, but some combinations of host and non-host infestation led to disappearance of this preference. In a third experiment, blends of volatile organic compounds were collected from the plants through headspace sampling and chemically analysed. HIPV blends different between accessions. However, these differences did not fully explain the findings of the complex semi-field experiment. Nevertheless, I identified that plant trait variation in HIPVs intricately interacts with non-host presence in its effect on parasitoid host-location efficiency.

Finally, in Chapter 6, I investigated how parasitism rates are affected by landscape context and how plants which differ in attraction of parasitoids are affected differently by these aspects of the landscape. In a field experiment, I placed cabbage plants of the two accessions in 19 different landscapes in the vicinity of Wageningen, the Netherlands. On these cabbage plants, I assessed parasitism rates of caterpillars by the naturally occurring parasitoid population. Additionally, I measured landscape characteristics such as the area of arable land, pastures, forest and non-woody seminatural area. Furthermore, a more functional landscape characteristic was quantified, the cover of plants from the family Brassicaceae (the food plant family of the host herbivore, P. brassicae). Parasitism rates were positively associated with the area of (mostly organic) arable land and brassicaceous plant cover, but this effect was larger for the more attractive accession Christmas Drumhead than for the less attractive accession Badger Shipper. The area of forest in the landscape was negatively associated with parasitism rates, which can be explained by the forest’s relatively low abundance of brassicaceous plants. I conclude that a more attractive accession is able to attract more parasitoids when there are sufficiently sized parasitoid populations nearby in the landscape by having a larger attraction radius. Additionally, for tritrophic interactions in which specialist insects are involved, functional characteristics of the landscape such as cover of host plants of the herbivore hosts of the parasitoid, can be more useful to explain parasitism rates than land use classes.

This thesis makes a contribution to the fundamental knowledge of foraging behaviour in complex field situations and, therefore, the relevance of HIPVs in mediating tritrophic interactions in natural and agricultural systems. In Chapter 7, I discuss my findings in a broader context. HIPVs are important long-distance cues for parasitoids to find their hosts in complex situations, where plants with hosts are patchily distributed with varying distance, embedded in vegetation structure. The connection between different plant patches can be strengthened by HIPVs. However, whether a higher attractiveness to parasitoids via HIPVs leads to higher parasitism rates can depend on characteristics of the plant’s habitat. This is interesting for HIPVs as a resistance or defence trait. The production of HIPVs can be ecologically costly when they also ‘advertise’ the plant to other herbivores or to hyperparasitoids which parasitize parasitoids. In situations where a higher HIPV attractiveness does not lead to higher parasitism rates, producing these HIPV blends does not give an advantage over plants which do not produce them. However, the relevance of HIPVs for plant fitness of these plants in different situations remains to be investigated. I also discuss the use of plant varieties with increased attractiveness to parasitoids in agriculture. Plants could be specifically bred for increased attractiveness over larger distance and in more complex situations. However, the landscape context is an important factor influencing natural enemy populations from which these natural enemies can be attracted to crop fields. It is, therefore, important to also consider resource needs of natural enemies and connectivity from stable populations to fields where enemies are needed to suppress pest populations. Because in highly simplified landscapes fragments with stable enemy populations are scattered and have low connectivity with other fragments or agricultural fields were enemies are needed, I conclude that conservation biological control measures should be implemented on a regional scale. Additionally, for more attractive varieties to be marketable to (organic) farmers, they need to produce a higher yield than less attractive varieties, something which yet has to be studied. Finally, how precisely HIPVs from different plants disperse through the environment has yet to be determined, which is methodologically challenging. Also, different parasitoid species may respond differently to the habitat characteristics used in this study to investigate parasitoid behaviour. Studying parasitoid traits in relation to its response to volatiles might give a better understanding of the mechanisms of foraging behaviour via HIPVs in the field. In conclusion, this thesis provides important insights in the role of HIPVs at spatial scales ranging from the local scale to the landscape scale. Such information is important for developing sustainable crop protection under field conditions.