Battle between insect eggs and host plants: Ecology and evolution of pierid egg-induced responses in Brassicaceae

In this thesis, I explore egg-induced responses of B. nigra and other brassicaceous plant species to Pieris brassicae and other pierid eggs. I focus on HR-like necrosis, OIPVs and how egg induction changes insect behaviour and caterpillar performance. Furthermore, I study how different plant species and plant accessions of the same species differ in their expression of HR-like necrosis and OIPVs.

The aim of this thesis is to shed light on the ecology and evolution of lepidopteran egg-induced responses in Brassicaceae plants by comparing responses between and within plant species. I show how egg induction affects egg survival, caterpillar performance and the behaviour of conspecifics and their parasitoids, and how it may vary between different plant species and plant accessions. Additionally, I aim to provide evidence that HR-like necrosis in the Brassicaceae is specifically elicited by pierid butterfly eggs, and may be understood as a counteradaptation to these specialist herbivores.

In chapter 2, I made use of the bitrophic B. nigra-P. brassicae system to study genotypic variation in induction of HR-like necrosis. I tested different plant accessions for their expression of HR-like necrosis and how it affected survival of clustered butterfly eggs. Both in greenhouse and field assays, some plant accessions expressed HR-like necrosis more frequently and more severely. Interestingly, even severe HR-like necrosis had no effect on egg survival.

I then tested whether or not the clustering of eggs provided protection against the egg-killing effect of HR-like necrosis. Indeed, egg mortality on HR-like necrosis expressing plants was higher when P. brassicae females were forced to lay single (non-touching) eggs instead of equally sized egg clutches. I hypothesized that singly laid eggs are more likely to die from desiccation, as more surface area is exposed to the surrounding air. To test this hypothesis, I conducted an experiment were I showed that under some humidity conditions survival of single eggs was lower than of clustered eggs.

Having proven variation in the expression of HR-like necrosis induced by clustered P. brassicae eggs within B. nigra accessions, I continued to research variation in another egg-induced trait, OIPVs. In chapter 3, I studied intraspecific variation between indirect egg-induced responses of B. nigra and consequences for P. brassicae and its parasitoids. In this tritrophic system, I collected and analysed plant volatiles from both control and P. brassicae egg-induced plants of two different plant accessions. Furthermore, I tested the behavioural response of gravid butterflies (P. brassicae), egg parasitoids (Trichogramma evanescens) and larval parasitoids (Cotesia glomerata) to control and egg-induced plants of both accessions. I found that the emission of five individual compounds was induced or reduced by egg deposition. Additionally, two more volatiles were released in differing quantities by the different plant accessions. Both the preferences for OIPVs by egg-laying butterflies and egg parasitoids were mainly influenced by egg induction rather than plant accessions. For C. glomerata, I showed that larval parasitism rates were different between control plants and plants previously exposed to eggs as well as between plant accessions.

In chapter 4, I additionally introduced a solitary butterfly species (P. rapae) and seven more brassicaceous plant species. I tested if the preference-performance hypothesis (PPH), according to which mothers will prefer to deposit eggs on plants where their offspring has the best performance, holds for the two butterfly species. I correlated oviposition preferences with the survival of eggs, and the tested the performance of the feeding caterpillars by comparing control plants with plants previously induced by conspecific eggs. Only P. rapae showed a clear preference for some plant species over others, while P. brassicae did not show a preference. While HR-like necrosis lowered the survival of P. rapae eggs, females preferred to lay eggs on plant species more prone to express HR-like necrosis. After hatching and three days of feeding, caterpillar performance was not different between plant species, for neither butterfly species. However, P. brassicae caterpillars performed significantly worse on plants previously induced by eggs than on control plants, while P. rapae caterpillars showed no such difference. In contrast, if they fed on plants expressing HR-like necrosis to previously laid eggs, P. rapae caterpillars performed better than on plants that did not express HR. I did not find such a difference for P. brassicae caterpillars. After seven days of feeding, the performance of P. brassicae caterpillars correlated with oviposition preferences, but only on egg-induced plants. This difference between control and egg-induced plants was not observed for P. rapae caterpillars. In summary, I showed that egg-laying preference and offspring performance differed between the solitary and gregarious butterfly species. More importantly, egg induction and plant responses to eggs were necessary to fully understand preferences and performances of the two butterfly species.

Lastly, I was interested in prevalence and specificity of the HR-like necrosis trait within the Brassicaceae. In chapter 5, I used a phylogenetic approach to unravel the potential co-adaptive relationship between pierid butterflies and Brassicaceae, in context of the evolved HR-like necrosis as an egg-killing trait. I made a wash of P. brassicae eggs to test 28 Brassicaceae and three Cleomaceae plant species for their ability to express HR-like necrosis upon induction. Only species from the tribe Brassiceae and a plant species of the Aethionemeae expressed HR-like necrosis. Conversely, I tested egg wash and eggs of eight pierid and one nymphalid butterflies, as well as the cabbage-feeding moths Plutella xylostella and Mamestra brassicae for their ability to induce HR-like necrosis in B. nigra plants. Only eggs of Pieridae species adapted to feed on Brassicaceae plants (belonging to the subfamily Pierinae) induced necrosis. PR1 expression was strongly induced by Pieris spp. and Anthocharis cardamines, while other Pieridae and the moths did not elicit PR1 expression or only a minor one. I also showed that HR-like necrosis lowered survival of single Pieris eggs in the field and the greenhouse, for four out of five plant species.