Most life forms, including plants, are more or less susceptible to infection by pathogens. Whereas plant-pathogen interactions are well studied in human-managed systems they are less known in natural ecosystems. In this thesis the interactions between the sterilising anther-smut fungus Microbotryum silenes-dioicae and the host plant Silene dioica were studied in natural habitats in the Skeppsvik archipelago in northern Sweden. The purpose was to study some population characteristics that can potentially influence the colonization of anther-smut (Microbotryum silenes-dioicae) in young Silene dioica populations. Specifically, it was tested whether the population age, floral-  and total density, population size and distances to nearest diseased populations affect incidence of disease (presence of disease) and prevalence of disease (proportion of diseased individuals in populations) in younger host populations. It was found that incidence and prevalence of disease increased with host population age and size. In addition, it was seen that distances to diseased populations could not explain incidence or prevalence of disease. These results can potentially be used to predict disease risk in similar plant-pathogen systems.

The aim of this thesis project was to elucidate patterns and processes associated with the biotic interactions in a natural plant–herbivore–parasitoid food web characterized by spatial and temporal heterogeneity with regard to species composition. The system examined is based on island populations of the perennial herb Meadowsweet (Filipendula ulmaria, Rosaceae), located in the Skeppsvik Archipelago. The area is subject to isostatic rebound, amounting to 0.85 cm per year; this makes it possible to calculate the age of the rising islands. Meadowsweet colonizes new islands when they are about 100 years old. Meadowsweet is consumed by two major herbivores in the study area: Galerucella tenella and Altica engstroemi (Coleoptera: Chrysomelidae). Both herbivores overwinter in the topsoil and successful colonization occurs when the islands reach a height that prevents the beetles from being removed or killed as a result of wave wash during the winter. I found that both herbivores significantly reduced individual plant fitness and population growth rate. A “cafeteria experiment” with Galerucella showed that this beetle discriminated between plants from different islands, avoiding plants from old islands which contained high concentrations of putative defence compounds, while readily accepting plants from younger islands which contained lower concentrations of these chemicals. Further, the plant species exhibited a trade-off between growth and production of the putative defence compounds. Taken together, these results were interpreted as providing evidence of herbivore-driven evolution of resistance in Meadowsweet. Further, laboratory studies suggested that Galerucella gradually includes a less preferred host plant (Rubus arcticus, Rosaceae) in its diet as Meadowsweet resistance increases. This implies that Galerucella drives its own host-breadth enlargement by selectively inducing a ‘rent rise’ in the original host, Meadowsweet. In a number of field studies I showed that the oligophagous parasitoid Asecodes mento (Hymenoptera: Eulophidae) has a strong positive effect on Meadowsweet seed set by removing large numbers of G. tenella larvae. This top-down effect is, however, altered by the presence of a close relative of G. tenella, namely G. calmariensis, which is monophagous on Purple loosestrife (Lythrum salicaria, Lythraceae). G. tenella experiences associational susceptibility when coexisting with G. calmariensis since the latter supports a higher and more fit pool of shared parasitoids and because Meadowsweet attracts a higher proportion of the shared parasitoid females than Purple loosestrife. This implies that G. tenella densities are very low in coexisting populations and that Meadowsweet experiences associational resistance and produces more seeds when co-occurring with Purple loosestrife. Thus, selection for increased resistance in Meadowsweet is likely to be relaxed in populations mixed with Purple loosestrife. I conclude that the evolution of plant resistance is likely to depend on the length of time and intensity of selection. When Meadowsweet colonizes new islands it experiences a period of enemy-free space; followed by a midlife and ageing with selection by herbivores. The intensity of this selection does, however, depend on the presence of additional plant and herbivore species.

About 6% of all angiosperms are dioecious. This separation of sexual function to male and female individuals, and the fundamentally different patterns of reproductive resource allocation that follows that separation, are thought to have important ecological and evolutionary consequences for plant enemy interactions.  I have studied whether intersexual differences in susceptibility to natural enemies can be explained by intersexual differences in resource allocation. In cases when sexual dimorphic traits form the target resource of a particular enemy I expected the enemy to select the best resource. The study system is the perennial dioecious herb, Silene dioica (Caryophyllaceae) and three of its specialist natural enemies, two insect herbivores the fly Delia criniventris (Anthomyiidae) and the twirler moth Caryocolum viscariella (Gelechiidae) and one systemic anther smut fungus Microbotryum violaceum. All three share the same food recourse, the floral stems, of their host plant. I studied the interaction on nine islands in a rising Bothinan archipelago over seven consecutive years. Both herbivores attacked female plants more than male plants (D. criniventris, 32.8% females, 30.7% males; C. viscariella, 4% females, 2% males). This attack pattern was consistent over years and islands and also correlated with a number of sexually dimorphic traits suggesting that females offer the better resource. Herbivore attack had no effect on plant survival but a significant effect on re-flowering the following year. Non-attacked females had an estimated mean re-flowering rate of 30.2%, and non-attacked males of 31%. Herbivore-attacked females had an estimated re-flowering rate of 46% compared with 38.4% for males. Females showed a stronger compensatory response to attack and tended to re-flower more often than males. Attack rates differed markedly in the different stages of primary succession. They were consistently higher in the youngest zone and decreased in parallel to progressing succession. This zonal pattern of decreasing attack rates correlated with several plant attributes, a decrease in plant size and nitrogen content, and an increased content of secondary compounds, but not to host plant density. We failed to come up with a simple explanation for the spatial structure with chronic high attack rates in the younger zones. However, the consistent patterns in attack rate suggest that a suite of abiotic and biotic factors interact and reinforce the strength and direction of selection. In general females were more frequently diseased by the anther smut Microbotryum violaceum than males with two exceptions. Disease frequencies were male biased on islands with low disease levels and in one of the seven study years. The change in disease frequencies from male to female bias confirm earlier studies suggesting that the relative contribution of the two components of infection risk, disease encounter and per contact infection probability can vary with population disease level. The change in the proportions of diseased males and females that was observed in one of the study years, followed a year of extreme weather conditions (prolonged drought). Both sexes showed a similar decline in flowering but diseased females decreased more than diseased males. This difference in response can be explained if considering that disease is more resource demanding in females than in males. Except for resources needed for mycelial growth and spore production, in females resources are also needed to restructure their sex expression and produce anthers. My study shows (i) that in dioecious species traits that are sexually dimorphic are of great importance for understanding the outcome of interactions with natural enemies, (ii) that the strengths and directions of enemy-host plant interactions are strongly shaped by both biotic and abiotic conditions.

Selection and genetic drift are often competing forces in shaping genetic structure in populations. Genetic drift will often effectively cancel out the effect of selection when population sizes are small, such as in colonizing island populations. On a small island in the Skeppsvik Archipelago in northern Sweden, a newly founded population of Silene dioica has been monitored since it first established around 1993. Though inhabiting an area of merely 173 m2, the population has been shown to exhibit a genetically differentiated patch structure where closely related individuals are tightly grouped, distanced from other family groups. In this study, the effect of selection was evaluated as compared to that of genetic drift. Variation in phenotypic traits in flowers, leaves and stalks were compared to that of neutral markers, in the form of PST and FST measures, to assess a measure of what proportion of differentiation among patches in phenotypic traits could not be attributed to genetic drift. Males and females were analysed separately to obtain measures of sex specific selection. Signs of divergent and stabilizing selection were found in several traits in both males and females despite the small spatial scale and short time since colonization. Further analysis is needed to assess explanations for trait divergence among patches and direction of selection.