Spatial patterns in the interaction between Salix triandra and associated parasites

Lena Niemi

2006

Department of Ecology and Environmental Science Umeå University SE-901 87 Umeå

Sweden

AKADEMISK AVHANDLING som med vederbörligt tillstånd av rektorsämbetet vid Umeå universitet för

erhållande av filosofie doktorsexamen i ekologi kommer att offentligen försvaras fredagen den 9 juni 2006, kl 10.00 i Stora hörsalen, KBC.

Examinator: Professor Lars Ericson, Umeå universitet

Opponent: Professor Keith Clay, Department of Biology, Indiana University, Indiana, USA

ISBN: 91-7264-025-1 © Lena Niemi 2006 Cover drawing: Moa Hjulfors Printed by: Print & Media

ORGANISATION Dept. of Ecology and Environmental Science Umeå University SE-901 87 Umeå

DOCUMENT NAME Doctoral Dissertation

DATE OF ISSUE June 2006

AUTHOR: Lena Niemi TITLE: Spatial patterns in the interaction between Salix triandra and associated parasites ABSTRACT This thesis focuses on mechanisms and processes underlying spatial patterns of resistance and virulence and on local adaptations in plant–parasite interactions. The model system used comprises the plant host Salix triandra, the pathogenic rust fungus Melampsora amygdalinae, the leaf beetle Gonioctena linnaeana, and the galler Pontania triandrae. In this work, I (1) emphasize the most important factors determining the outcome of a plant–pathogen interaction, and the types of systems in which local adaptations can be expected, (2) examine the resistance structures of different populations of S. triandra, and whether the leaf beetle G. linnaeana responds to the local conditions of the populations of S. triandra in Sweden, and (3) address whether the distribution of parasites on S. triandra can be explained by the plant content of secondary metabolites. A review of several studies of the subject leads to the conclusion that adaptation of pathogens to their local hosts is more likely to be found in systems in which the pathogen is host-specific, non-systemic, and has a larger dispersal range and evolutionary potential than its host does. Furthermore, the scale of the study must be adjusted to that of the pathogen’s local population distribution. In addition, the temporary nature of host–pathogen interactions influences the importance of sample size, and too-small sample sizes can lower the chance of finding local adaptations, even though they may have evolved in a given system. The results of an inoculation experiment using material from physically isolated natural populations of S. triandra and M. amygdalinae confirm the importance of previous conclusions. Spatial variation in the resistance structure of S. triandra also has effects on the insect herbivore G. linnaeana, which has responded by adapting to the local hosts. However, local differences in secondary chemistry affect different parasites in different ways, and while P. triandrae is attracted by high levels of phenolic compounds, including tannins, M. amygdalinae and G. linnaeana are more rarely found on plant individuals with high concentrations of tannins. In addition, brood deposition by adult females of G. linnaeana and the performance of larvae are positively affected by luteolin-7-glucoside and an additional unidentified flavonoid, whereas they are negatively affected by the presence of (+)-catechin and high levels of tannins.

Our results also show that plants traits that provide resistance to one type of parasite do not necessarily provide resistance to others. This indicates that different natural enemies potentially assert divergent selection pressure on S. triandra phenotypes which can be important for maintaining phenotypic variation in plant species KEYWORDS: Gonioctena linnaeana, host-parasite interactions, leaf beetle, local adaptation, Melampsora amygdalinae, phenolics, pathogens, plant host, secondary chemistry LANGUAGE: English ISBN: 91-7264-025-1 NUMBER OF PAGES: 31 + 6 papers SIGNATURE: DATE: 18 April 2006

TILL SIMON, MOA & PONTUS

LIST OF PAPERS This thesis is a summary and discussion of the following papers, which will be referred to in the text by their Roman numerals. I Niemi, L. & Wennström, A. Can adaptation of parasitic fungi to their

local host plants be a general expectation? Submitted manuscript. II Niemi, L., Wennström, A., Hjältén, J., Waldmann, P. & Ericson, L.

Spatial variation in resistance and virulence in the host–pathogen system, Salix triandra–Melampsora amygdalinae. Submitted manuscript.

III Niemi, L., Wennström, A., & Ericson, L. 2005. Insect feeding preferences

and plant phenolic glucosides in the system Gonioctena linnaeana – Salix triandra. Entomologia Experimentalis et Applicata 115: 61–66.

IV Niemi, L., Wennström, A., & Ericson, L. Preference and performance of

the leaf-eating beetle Gonioctena linnaeana on sympatric and allopatric populations of Salix triandra. Submitted manuscript.

V Niemi, L., Wennström, A., Hjältén, J. & Ericson, L. Mother really knows

best: Host choice of adult phytophagous insect females reflects a within host variation in suitability as larval food. Submitted manuscript.

VI Hjältén, J., Niemi, L., Wennström, A., Ericson, L., Roininen, H. &

Julkunen-Tiitto, R. Variable responses of natural enemies to phenotypic variation in Salix triandra: the importance of plant secondary chemistry. Submitted manuscript.

Paper III is reprinted with the kind permission of the publishers.

TABLE OF CONTENTS INTRODUCTION ……………………………………………………………….. 8

Resistance and virulence …………………………………………………. 8 Specialization ……………………………………………………………… 9 Local adaptation ………………………………………………………….. 10

AIMS OF THE THESIS ………………………………………………………… 10 MATERIALS AND METHODS ……………………………………………… 11 Study system ……………………………………………………………… 11 Methods …………………………………………………………………… 13

Description of the studies ………………………………………………… 15 Spatial patterns in host resistance and pathogen virulence – the interaction between S. triandra and the pathogen M. amygdalinae (Paper I, II) …………………………………………………………….. 15

Spatial distribution and response of the leaf beetle G. linnaeana to the host plant S. triandra (Papers III and IV) ………………………. 16 Dealing with local environmental conditions (Paper V and VI) ……….. 17 MAJOR RESULTS AND DISCUSSION……………………………………… 18 Spatial patterns in host resistance and pathogen virulence - the

interaction between S. triandra and the pathogen M. amygdalinae (Paper I, II) ………………………………………………………………. 18 Spatial distribution and response of the leaf beetle G. linnaeana to the host plant S. triandra (Papers III and IV) ……………………….. 20

Dealing with local environmental conditions (Paper V and VI) ………. 22

CONCLUSIONS ………………………………………………………………… 24 ACKNOWLEDGEMENT ……………………………………………………… 25 REFERENCES …………………………………………………………………... 25 TACK TILL… …………………………………………………………………... 31 APPENDICES: PAPER I - VI

INTRODUCTION The spatial pattern of plant distribution is strongly determined by edaphic and climatic factors. Another important, though often neglected, factor is the presence or absence of parasites. Most, if not all, plants have some associated parasites. These parasites are, for example, herbivores feeding on plant tissue, or fungal pathogens that live between the plant cells and obtain nutrients produced by plant photosynthesis. Parasitic attacks may dramatically affect the individual host and may influence survival, allocation patterns, competitive ability, and reproduction. As a consequence, the effects of parasites may go beyond the individual hosts and affect both the genetic composition and distribution of the host population, as well as the composition and structure of entire plant communities and interactions on several trophic levels (Maschinski & Whitham 1989, Dobson & Crawely 1994). Resistance and virulence If parasites negatively affect plant fitness, there will be a selection pressure on the plants for resistance to those parasites. Resistance in plants includes all characteristics and mechanisms that act to retard the activity of a parasite, in order to decrease the detrimental effects of parasite attack. Consequently, a resistant host plant selects for associated parasites to evolve an ability to overcome the resistance, that is, virulence. Virulence is a term that often causes confusion, because it may be defined as either the ability of a parasite to overcome plant resistance or as a measure of parasite aggressiveness, i.e., the amount of damage caused by a parasite attack. Throughout this thesis virulence is used in the sense of a parasite’s ability to overcome host resistance, or more precisely, the ability of a pathogen to sporulate on the host, or insect recognition and acceptance of a plant as a resource. The reciprocal selective responses of the interacting host and parasite may result in an “arms race” in which any change in resistance structure in the host population is followed by an adaptive response in the virulence structure of the parasite population. This kind of antagonistic coevolution has been documented in several host–parasite systems; however, the interactions commonly differ across both spatial and temporal scales, with some points displaying reciprocal selection, while in others selection is directed towards only one species or is even absent (Thompson 1999). Plant resistance is maintained by morphological characteristics, such as hairs, spines, surface waxes, and cortexes, in combination with the production of chemical substances that act as deterrents and toxins, or have antimicrobial effects (Ehrlich & Raven 1964, Feeny 1992, Schoonhoven et al. 1998). These chemicals are primarily secondary compounds, and are generally considered to be of no nutritional value for the herbivore (Schoonhoven et al. 1998). Resistance to different types of enemies is not completely independent, and both physical structures and secondary compounds may function against both pathogens and herbivores (Krischik et al. 1991). However, since the functions of the components involved in resistance to different organisms may be quite different, they can interfere with each other. It has even

8

been suggested that the production of one substance may inhibit production of another, and also that some substances inhibit the function of others (Levin 1976, Maleck & Dietrich 1999). Resistance in plants is either non-specific or specific. Non-specific resistance is governed by many genes conferring resistance to a broad range of parasites. It acts by delaying disease development, thus decreasing the damage. In specific resistance, a few genes in the host act to produce physiological responses or extremely toxic and sometimes tissue-specific compounds that specifically protect a host plant against attack or infection by certain types of parasites (Levin 1976, Burdon 1987). Gene-for-gene resistance is a type of specific resistance known in several host–pathogen interactions in which the pathogen does not have the option of switching to alternative hosts within the local community. This type of resistance involves a few specific resistance genes in the host; these interact directly with the genes controlling pathogen virulence, conferring protection against attack by specific races or pathotypes of the pathogen (Burdon et al. 1996). Specialization To be able to parasitize a plant, a parasite must find a way to get around plant resistance. The plant kingdom produces several hundred thousand chemical substances, and the mechanisms of defense vary among plant families, species, and even tissues within a plant (Levin 1976, Schoonhoven et al. 1998). Both pathogens and herbivores evolve in response to their plant hosts and in order to increase their efficiency and ability to attack, most parasites have limited their range of hosts. Therefore can fungal pathogens that live the whole life span on or within the tissue of plants, and insect herbivores with the ability to either tolerate or detoxify detrimental substances in plant defence, often be quite specific in what host species they attack (Brattsten 1992, Städler 1992, Agrios 1997). Furthermore, the secondary chemistry properties of a plant are not only restricted to defense, but also provide strong cues used by insect herbivores for recognition and acceptance of plants as possible resources (Schoonhoven et al. 1998, Bernays 2001). The great variety of secondary substances in plants provides a challenging range of stimuli for herbivores to evaluate and respond to, and a strategy that requires fewer receptors and smaller memory capacity should lead to more fixed responses and faster decisions (Bernays 2001). Such a strategy would be specialization on a specific plant species or on a range of plants with similar chemistry. Specialization is common among phytophagous insects, and specialized herbivores have evolved physiological and/or behavioral adaptations to the specific chemistry of their host plants and the recognition and acceptability of a plant is commonly triggered by certain host-specific stimulants in plant chemistry (Tahvanainen et al. 1985, Hemming & Lindroth 1995, Kolehmainen et al. 1995, Schoonhoven et al. 1998, Tikkanen et al. 2000, Renwick 2001, Bernays 2001, Del Campo et al. 2003).

9

Local adaptation Even though most parasites are more or less specialized, this does not mean that they are equally capable of infecting or attacking all individuals within their host range. Most species are patchily distributed in the landscape, with local populations that vary in size as well as in degree of connectedness, and this may have large consequences for shaping the host–parasite interaction (Thompson 1994, 1999, Burdon & Thrall 1999). The genotypic composition of a local population is basically determined by founder effects, genetic drift, colonization, and extinction events, but it is also shaped by present and historical environmental events, which include local abiotic conditions such as resource availability and climate, and biotic conditions such as associated parasites. These environmental conditions exert selection pressure on the populations, and if the environment varies spatially it will impose a spatially varying selection pressure on the populations that inhabit it. As a result, plants and parasites are distributed in groups of genetically differentiated populations, and the outcome of the interaction between a host and an associated parasite may therefore not only vary dramatically between different species interactions, but also between localities and populations within a single system (Thrall et al. 2002). Thus, the interaction between plant hosts and their parasites both shapes and is shaped by spatial patterns, and if local populations differ in resistance and virulence, both hosts and parasites would benefit from evolving adaptations to their local conditions. AIMS OF THE THESIS The purpose of this thesis was to gain a better understanding of the processes underlying spatial patterns of resistance and virulence and of local adaptations in plant–parasite interaction. The main questions addressed in this thesis are as follows:

• What are the most important factors determining the outcome of a plant–pathogen interaction, and are we able to make predictions as to where we might find patterns of local adaptations? (Papers I, II)

• Do different populations of S. triandra have different resistance structures,

and does the leaf beetle G. linnaeana respond to the local conditions of the populations of S. triandra in Sweden? (Papers III, IV)

• Can the distribution of parasites on S. triandra be explained by the plant

content of secondary metabolites? (Papers V, VI)

10

MATERIALS AND METHODS Study system The study system comprises the host plant Salix triandra L (Salicaceae) and three associated parasites: the pathogen Melampsora amygdalinae Kleb. (Uredinales: Basidiomycota), the herbivore Gonioctena linnaeana Shrank (Coleoptera: Chrysomelidae), and the leaf galler Pontania triandrae Benson (Hymenoptera: Tenthredinidae). THE HOST PLANT. Salix triandra L (Salicaceae) is a willow species. It is a dioecious plant that reproduces both sexually, via wind-dispersed seeds, and asexually via broken-off twigs or branches that are easily rooted and may be transported downstream by water. Salix triandra has a large distribution range extending from Japan to Western Europe. The main distribution range is in Russia, where S. triandra forms large continuous stands along rivers over vast areas (Skvortsov 1999). The western limit of its continuous distribution is the Karelian Isthmus (Russia), the Baltic states, and eastern Poland. Further west, S. triandra is rare and occurs scattered in more or less isolated populations. Historically, S. triandra has probably been part of the Scandinavian flora since the most recent glaciation, the Weichsel glaciation, which started to withdraw approximately 10,000 years ago. Sweden was completely free of ice approximately 6,500 years ago (Lundqvist & Robertsson 1994). Salix triandra probably colonized Sweden from the east and established itself on river banks near what was then the coast line. Since then, land uplift has raised this former coastal region to higher altitudes, and S. triandra has spread downstream along the rivers. The highest elevation where S. triandra is found in Swedish river valleys corresponds well with the coast line 6,000−7,000 years ago (Hultén & Fries 1986, Mossberg et al. 1992, Lundqvist & Robertsson 1994). In northern Sweden the land is still rising by approximately 0.9 cm per year. In Sweden, S. triandra now inhabits sandy inundated riverbeds and is native to six river valleys, where it reaches high densities. From south to north, these are the Klarälven, Dalälven, Indalsälven, Moälven, Öreälven, and Torneälven river valleys. Salix triandra produces numerous seeds, but seedling establishment seems to be rare and most dispersal is therefore vegetative (pers. obs.). In addition, since S. triandra is insect pollinated (e.g., by bumblebees and syrphids), the chance of pollen transfer between river systems are limited. THE PATHOGEN. Salix triandra is attacked by a rust fungus, Melampsora amygdalinae Kleb. (Uredinales: Basidiomycota) (Wilson & Henderson 1966, Gjaerum 1974). This is a non-systemic annual rust that is autoecious and host specific to S. triandra, being known throughout the distribution range of the host. The life cycle of M. amygdalinae is macrocyclic and includes a sexual stage in spring, when overwintering teliospores germinate and give rise to basidiospores that

11

infect the host. The haploid basidiospores germinate to produce mycelia that form spermogonia. After dikaryotization by transfer of spermatia to receiving structures of compatible mating types, aecia are formed. From these, aecidiospores are dispersed, which, after germination, form uredia on infected host tissue. Symptoms of disease, seen during summer, are pustules with orange uredospores on leaves and shoots. The uredospores are produced asexually and several generations appear in one season. The spores are wind dispersed and can infect the same or different individuals. In years when conditions for infection are good, the current annual shoots can be covered with pustules, which may result in premature leaf shedding (pers. obs.). In autumn, teliospores are produced. M. amygdalinae is also able to overwinter as mycelium in S. triandra buds (Gäumann 1959, Wilson & Henderson 1966). THE HERBIVORE. Gonioctena linnaeana Shrank (Coleoptera: Chrysomelidae) has a large distribution range, which extends from northern Spain and Norway to Mongolia and Sakhalin (Warchalowski 2003). The leaf beetle is reported as oligophagous on willow species, and has been found on Salix purpurea, S. alba, and S. triandra in central Europe (Koch 1992). In Russia it has occasionally been found on S. viminalis, but more rarely in localities where S. triandra is likely to be found (pers. com. B. Koryatrev). In Sweden there are no natural occurrences of Salix purpurea and S. alba, and G. linnaeana seems to have specialized on S. triandra. We have only been able to find it in two of the three Swedish river populations included in our experiments, Öreälven and Torneälven, and only in a few of the S. triandra subpopulations along each river. In the host populations where G. linnaeana is found, it almost completely defoliates the attacked host individuals. Gonioctena linnaeana is ovoviviparous, which means that the developing eggs are kept inside the female until they are ready to hatch. The beetles overwinter as adults, and mating and brood deposition take place in the early spring (at the time of leaf burst). Each larva is covered with a thin, transparent membrane which soon after deposition ruptures, and the larvae start to feed on the youngest leaves of the host plant. Soon thereafter, the adults die. The larvae go through four instars and pupate on the ground. After a few days, the adults emerge. They feed for a short period of time before they hibernate until the next spring. THE LEAF GALLER. The sawfly Pontania triandrae Benson (Hymenoptera: Tenthredinidae) is monophagous on S. triandra. The adult females emerge in spring and oviposit in developing leaves, causing a gall to form. It produces multiple galls on both sides of the midrib of the leaves. The larvae feed inside the gall until late autumn, when they leave the gall and descend to the ground to pupate. P. triandrae are often patchily distributed among S. triandra clones, but can sometimes reach high densities.

12

Methods SAMPLING. We used plants from three Swedish and one Russian host population: the Dalälven, Öreälven, and Torneälven river populations from Sweden, and the Oredezh river population from Russia (see map, Fig. 1). All populations are well separated in space, limiting gene flow. The distance between the Dalälven, Öreälven, and Torneälven populations is approximately 450 km, while the Russian population is located approximately 750 km from Dalälven – the closest Swedish population. The plant material was collected from the natural populations as cuttings prior to leaf burst, a few years before our experiments; thereafter, the plants were grown together in a common garden at the Umeå University campus. All plants therefore experienced the same environmental conditions up to the beginning of the experiments. Three pathogen populations originating from the three Swedish host populations were used. Uredia of 3 mm in diameter were collected early in the season when individual pustules could be easily separated, and stored separately at room temperature for 24 h before being used. The individual pustules are hereafter referred to as “rust isolates.” The herbivore was only found in two Swedish populations and we used both. For a conditioning experiment, we collected adults and used the larvae that emerged in the lab; however, for all other experiments with the herbivore we used first-instar larvae collected in the field.

Figure 1. Distribution of Salix triandra in Northern Europe and Russia (after Hultén and Fries 1986 and Mossberg et al. 1997). D = Dalälven, Ö = Öreälven, T = Torneälven and R = Oredezh River, Russia.

INOCULATION EXPERIMENTS (I). The inoculations were made on cuttings of S. triandra with one or two leaves attached. Three cuttings from each of the host

13

clones used in the experiment were randomly inserted into a perforated piece of Styrofoam and placed floating on water in a container. The rust isolate was cut out of the leaf and put in a glass vial together with 10 ml of water. The suspended spores were sprayed onto the leaves of the cuttings and the container was covered with transparent plastic to ensure high humidity. This was repeated for all isolates, each in a different container. The containers were kept in a laboratory at a temperature of 20˚C and subject to a natural light regime. The results of the inoculations were registered after 14 days. A plant was considered susceptible if pustules occurred on the leaves of any of the three replicates. PREFERENCE EXPERIMENTS (III, IV). In all our experiments with the leaf beetle (III, IV, V) we used larvae and not adults. This was because the adults of this species do not feed either very much or for very long, and also because larval preference and conditions during larval stages may be important for adult performance (Nylin & Gotthard 1998, Phillips 2004, De Block & Stoks 2005). Preference of larvae was tested in multiple-choice cafeteria trials in the lab. Using a punch tool, leaf discs of 1 cm in diameter were cut from a random sample of four hosts from each of the three studied host populations. The leaf discs were placed on moistened filter paper in Petri dishes, and one larva was released in each dish and allowed to feed for 48 h. The consumed area of the various leaf discs, in percent, was estimated visually. CONDITIONING EXPERIMENT (IV). Since we used first-instar larvae collected in the field, we wanted to exclude the possibility that larvae preference may be determined by their initial feeding experiences (conditioning). Therefore, we collected 20 adults from the field and placed them together in a transparent plastic box, together with S. triandra leaves. As soon as the females deposited their broods, the newly emerged larvae were collected and placed in a box with leaves from one of the three host populations, i.e., the Öreälven, Torneälven, or Russian population. After feeding for 48 h, 15 larvae from each box were randomly selected and used in a multiple-choice cafeteria trial in which they were allowed to choose between leaves from the three different host populations. We could not detect any conditioning effect, since all larvae showed a significant preference for leaves from Öreälven, regardless as to what they had experienced before. PERFORMANCE EXPERIMENT (IV, V). Our studies of larval performance when larvae were reared on leaves from different clones or host populations were conducted in a laboratory. The larvae were kept in transparent plastic boxes and provided with leaves of different origins. The larvae had free access to fresh leaves for the duration of the experiment. Performance was measured as survival, pupal mass, and development time from first instar to adult.

14

CHEMICAL ANALYSIS (III, V, VI ). Ten leaves from current annual shoots of the different individuals and populations of Salix triandra were analyzed for content of phenolic glucosides and condensed tannins. The analyses were performed at the laboratory of Riitta Julkunen-Tiitto at the University of Joensuu, where methods have been developed for determining the content of willow phenolics (Julkunen-Tiitto 1989, Julkunen-Tiitto et al. 1996). Description of the studies Spatial patterns in host resistance and pathogen virulence – the interaction between S. triandra and the pathogen M. amygdalinae (Papers I and II) Spatial variability in the forces of selection, and the obvious benefits for a local population of evolving traits that provide an advantage under local conditions, have caused local adaptation to be a general expectation in host–parasite interactions. However, studies of various host–pathogen interactions have produced contradictory results, some studies finding local adaptations and others not (Parker 1985, Jarosz & Burdon 1991, Davelos et al. 1996, Roy 1998, Kaltz et al. 1999, Thrall et al. 2002). In response, recent studies have suggested several specific characteristics of the species involved that may be important to consider when studying local adaptation (see, e.g., Gandon et al. 1996, Kaltz & Shykoff 1998, Burdon & Thrall 1999, Gandon 2002, Gandon & Michalakis 2002, Kawecki & Ebert 2004). Here we have brought these characteristics together, and since they mainly concern three different aspects of local adaptation, we have placed them into the following categories (I):

i) First, there are species-specific characteristics that influence the dynamics of the host–pathogen interaction and the evolutionary potential of both species. Such characteristics include factors contributing to genotypic variation in the populations (e.g., mode of reproduction, migration rates, population size, and mutation rates), the spatial pattern of the distribution and dispersal abilities of both species, and life strategies (e.g., whether the host is perennial/annual and whether the pathogen is systemic/non-systemic). In addition, the degree of host specificity has important implications for host availability, and a pathogen that is dependent on a single host for survival and reproduction can increase the likelihood of future successful infections through adapting to the local resistance structure.

ii) Second, there are factors that are important determinants of which organism

is likely to adapt to another. In the interaction between a plant host and a pathogen, either the host adapts to pathogen virulence or the pathogen evolves virulence compatible with host resistance. It is generally considered that the pathogen should be ahead in this race, and if not, this is called maladaptation. Whether or not a pathogen becomes adapted or maladapted depends on the relative evolutionary potential (including mode of

15

reproduction, population size, and generation time) of the involved species and also on relative migration rates because migration provides new genetic input to the receiving population, and it should be the population that migrates the most that should adapt to the other.

iii) Third, there are factors that influence the likelihood of finding local

adaptation. Even though a host–pathogen system may have the potential to evolve local adaptations, all populations may not display this pattern at all times. As a consequence of local population dynamics, the outcome of the interaction will vary in both time and space; therefore, local populations must be identified for both species, and several local populations must be included in the study.

In paper I we examined a number of studies of host–pathogen systems to see whether we could pinpoint some of the main factors and characteristics of both the species involved and the study design that was important for the outcome of the interactions and for the results of the study. In paper II we specifically investigated the host–pathogen system of S. triandra–M. amygdalinae, which fulfilled criteria that, according to previous suggestions and the findings of paper I, should promote both the evolution of local adaptation and increase the likelihood of finding it. Our aim was to investigate spatial variation in host resistance and pathogen virulence, and to determine whether the local resistance patterns in the host populations resulted in the local adaptation of the pathogen. Spatial distribution and response of the leaf beetle G. linnaeana to the host plant S. triandra (Papers III and IV) The distribution of herbivores on plants is largely influenced by plant resistance patterns, which in turn are largely determined by the presence of plant secondary substances (Edmunds & Alstad 1978, Jaenike 1990, Berenbaum & Zangerl 1992, Schoonhoven et al. 1998). The classes or types of compounds that form such defenses vary between specific ranges of hosts, and members of the Salicaceae family are dominated by phenolic compounds, namely, flavonoids, phenolic glucosides, and tannins. However, both the composition and concentration of the various compounds vary between different species in the family. Salix triandra is characterized by a relatively high concentration of condensed tannins, and differs from other willows by containing mostly non-salicylate phenolic glucosides (Julkunen-Tiitto 1989). This plant species also produces salidroside, a compound rarely found in other willows, whereas the otherwise common compound salicortin not has been found in S. triandra (Tahvanainen et al. 1985, Julkunen-Tiitto 1989). The production of secondary substances in a plant is the result of genotype or environment, or both. Therefore, plant chemical composition may not only vary between plant species but also between populations or individuals within a species (Edmunds & Alstad 1978, Strauss 1990, Fujiyama & Katakura 1997, Schoonhoven

16

et al. 1998, Glynn et al. 2004). If a host plant is dominated by chemicals or structures that make it unattractive to generalist herbivores, this may provide an enemy-free space for a specialist herbivore that is able to evolve the ability to tolerate or make efficient use of that host (Oppenheim & Gould 2002). Non-random patterns in the presence or absence of herbivores on their host species may thus indicate either different levels of host defense or the development of adaptations to the chemistry of specific hosts. In a few Swedish populations, the herbivore G. linnaeana has specialized on S. triandra, but is unevenly distributed both between and within populations. In paper III we wanted to investigate whether the uneven distribution of G. linnaeana could be explained by different population levels of defensive secondary substances, while in paper IV we examined whether the herbivore evolves in response to its local hosts. If so, the insect should have the ability to distinguish between host individuals from allopatric and sympatric host populations, and a preference for local hosts should be accompanied by superior performance on these hosts. Dealing with local environmental conditions (Papers V and VI) The local environment, including associated parasites, is important in shaping the genetic structure of plant populations; it thus has important implications for the local resistance patterns of the host, and thereby also for the distribution of parasites within the distribution range of their hosts. However, plants often face a multitude of natural enemies that may vary in their response to different resistance traits (Tahvanainen et al. 1985, Matsuki & MacLean 1994, Ikonen et al. 2002). Differential selection by natural enemies is therefore believed to be an important selective force for maintaining the variation in resistance traits that occurs in natural plant populations (Thompson 1994, Stinchcombe & Rausher 2002, Strauss et al. 2005). Paper III revealed that there is significant variation in secondary chemistry, between both S. triandra populations and host clones. Paper IV found that the larvae of G. linnaeana, in laboratory conditions, discriminate between host clones and prefer their sympatric hosts over allopatric hosts. During our work with S. triandra and G. linnaeana we have noticed a non-random pattern of larvae distribution, not only between host populations but also within populations. For several consecutive years we observed that the same host clones in a population were severely damaged by larvae, whereas others harbored hardly any larvae at all. In G. linnaeana, as in many other phytophagous insects, it is the adult female that is the most responsible for selecting the plants on which the larvae will feed and develop. Since offspring performance is closely connected to adult fitness, egg-laying females can evolve behavior that is beneficial for larval performance; they should, according to the preference–performance hypothesis, oviposit on plants that maximize offspring performance (Thompson 1988). However, different types of parasites are likely to respond to different characteristics of a specific host, and one type may also be influenced by the presence of another. In paper V we sought to

17

determine whether the selective behavior of the adult females of G. linnaeana affects offspring performance in a population of S. triandra, and also whether the level of chemical defense can explain host choices. Paper VI concerns three different types of enemies with different distribution patterns within a single population of S. triandra; in the paper we sought to explain spatial patterns, e.g., phenotype preference of a galling sawfly in a natural willow population. Furthermore, we evaluated whether traits that confer resistance to one type of natural enemy, i.e., the galling sawfly, also confer resistance to others. MAJOR RESULTS AND DISCUSSION Spatial patterns in host resistance and pathogen virulence – the interaction between S. triandra and the pathogen M. amygdalinae (Papers I and II) The literature survey compiled in paper I and our experimental results presented in paper II confirm the importance of the previously suggested characteristics of involved species for the evolution of local adaptations of pathogens to their host plants. However, paper I reveals that some factors seem to be more important than others for the likelihood of local adaptations to evolve and also for the finding of locally adapted pathogen populations when they have evolved. These factors are host specificity and the proper identification of the spatial scale of the local pathogen population. Host specificity implies a narrow host range, which is likely to reduce the number of available hosts compared to situations in which several different host species are acceptable. To ensure the survival of future generations, it will therefore be more important for a specialized pathogen to maintain the ability to track changes in its particular hosts and a locally adapted pathogen should have increased its ability to track the resistance of the most-frequently encountered hosts, which in turn should have positive effects on pathogen fitness. Identification of the specific local pathogen population is crucial, as this has implications for the spatial scale at which the results of the interaction can be detected. Local adaptation of a pathogen occurs at the scale of differentiation of the pathogen populations, which is determined by the rate and range of gene flow (Gandon et al. 1996, Thrall et al. 2002). The selection processes in local populations only differ from the processes shaping the interaction as a whole, at a scale at which the selection occurs without the homogenization of a constant gene flow. Fungal pathogens do often occur throughout the distribution range of their hosts. However, the spatial distributions of local populations of host and pathogen do not necessarily coincide, especially if their dispersal abilities differ. Fungi are the most dispersal-effective organisms on earth, due to their small, light, and numerous spores, and one local pathogen population may therefore encompass several local host populations

18

(Nagarajan & Singh 1990, Aylor 2003). If no consideration is given to pathogen dispersal abilities and if a study is not adjusted to the scale of the local pathogen population, local adaptations of the pathogen to the host may not be detected. The system of the clonal willow S. triandra and the rust fungus M. amygdalinae fulfils the criteria for when local adaptation of a pathogen to the host can be expected. The results presented in paper II indicate that all host clones had unique patterns of resistance and that all rust isolates had different infection profiles. There was also considerable variation between populations in the different inoculation combinations, suggesting that physical isolation of populations allows the interaction between host and pathogen to work independently in the different populations. As a result, the populations differ in resistance and virulence structure, and the pathogen population has, in all three Swedish S. triandra populations studied, adapted to its sympatric hosts. Spatial variability in an interaction that is caused by local differences in selection does not only result in local differences in the resistance and virulence structure of the populations. The selection for resistance and virulence may be reciprocal in some populations, and asymmetric or even absent in others (Thompson 1999). This may contribute to a pattern in which the levels of resistance and virulence differ between populations, and it has been suggested that high levels of host resistance may select for high levels of virulence in the associated pathogen population, or vice versa (Thrall et al. 2002, Thrall & Burdon 2003). Our results in paper II indicate that this may be the case in the S. triandra–M. amygdalinae system. The rust population from Torneälven was the most virulent of the Swedish rust populations, and in addition, the host population from Torneälven displayed high resistance to allopatric rust isolates. In contrast, lower levels of resistance found in the other two Swedish host populations were associated with lower levels of virulence. The finding of local adaptation in the S. triandra–M. amygdalinae system is most likely an effect of the sampling of physically isolated populations. The importance of the spatial structure for the evolution of local adaptation, and of the scale of the study for the ability to detect local adaptation can be exemplified by a study by Thrall et al. (2002). They examined local adaptation at two scales, regional and local. Three regions, each containing two populations, were sampled; the authors found that the pathogen was significantly more successful at inoculating the sympatric than the allopatric hosts in three of the populations. Thus, in this system there were selective forces favoring the evolution of local adaptation, but they were only found in half of the studied populations. However, at the regional level there was clear evidence of local adaptation. Had fewer populations been examined, and only within the level of a single region, patterns of local adaptation may not have been found. Sample size is often proposed as one of the most important explanations as to why local adaptation is not found in a given study (Kaltz et al. 1999, Kawecki & Ebert 2004). One reason for this is that local conditions are not static. The dynamics of

19

local populations are influenced by seasonal or stochastic events that may cause local extinctions and ensuing recolonizations. This occurs independently in different populations, and the genotype frequency may therefore vary both between populations and within populations over time. The result is a spatial mosaic where selection varies in intensity in different locations at different times, and as a consequence, local adaptation becomes a temporal phenomenon. To increase the likelihood of sampling a population in which local adaptations have developed, several populations must be included in a study, and failure to find local adaptation does not necessarily mean that it does not exist in a given interaction; rather, the failure may simply result from the examination of too few samples. However, both the studies presented in paper I and the study in paper II indicate that the number of populations that needs to be included in a study is highly unpredictable. The studies examined in the literature survey were generally quite small, mostly examining one to three populations – as did our study. Despite this small sample size, local adaptations were detected in several studies. Spatial distribution and response of the leaf beetle G. linnaeana to the host plant S. triandra (Papers III and IV) The distribution of an insect species is determined by its host-use characteristics, which in turn are the result of past ecological and evolutionary processes. All the historical interactions between plants and parasites and past and present abiotic environmental conditions have together shaped the chemical structure and quality of the host plants. The suitability of a plant as food for an associated herbivore is determined by the nutrient content in relation to the composition and amount of detrimental secondary compounds (Schoonhoven et al. 1998). Specialization of an insect herbivore on a plant species may include physiological adaptations in its digestion or detoxifying mechanisms. Such adaptations influence insect performance, and specialized insect herbivores may also evolve behavioral adaptations that result in preferences for specific hosts (Berenbaum 1980, Mopper et al. 1995, Stiling & Rossi 1998, Labeyrie & Dobler 2004). Behavioral adaptations and preferences may have consequences for insect performance if they affect the ability to recognize and accept alternative hosts. Gonioctena linnaeana has, in two Swedish river valleys, evolved an ability to make efficient use of the willow species S. triandra. Since no other leaf beetles have S. triandra as their major host plant in Sweden, G. linnaeana has eliminated much of the competition from other herbivores. In Sweden, Salix triandra is distributed in spatially isolated populations, and results from paper III indicate that the concentration of the different phenolic compounds varies among host individuals and that some of these differences can be detected at the population level. The most obvious differences are found between clones from Dalälven and those from Torneälven, and also between Swedish S. triandra clones and those originating from

20

the population located within the major distribution area in Russia. Furthermore, the clones from Torneälven seem to share more similarities with Russian clones than with most of the Swedish clones. The results from papers III and IV also reveal that the Swedish populations of G. linnaeana do not have the ability to recognize or accept the Russian S. triandra plants as a resource. Instead, the larvae of G. linnaeana from both Swedish populations prefer Swedish (III) or even their sympatric hosts over allopatric hosts (IV). Furthermore, both larval populations perform better when reared on sympatric than on allopatric hosts (IV). This combination of preference for particular hosts and enhanced performance on preferred hosts is strong evidence for the evolution of adaptations. Several factors are important for the local adaptation of this herbivore to its host in Sweden. First, G. linnaeana is also known to use S. alba and S. purpurea as hosts in central Europe (Koch 1992); however, none of these willows occurs naturally in Sweden, so the Swedish G. linnaeana populations have instead specialized on a single host species. Second, the host is a clonal, long-lived perennial plant, a fact that allows this univoltine insect herbivore to spend several generations on the same host genotypes, increasing the probability of adaptations to evolve. Third, the populations have a limited distribution in the area studied and are isolated by large distances, which limits gene flow that otherwise could swamp any pattern of adaptation. Fourth, S. triandra has a limited number of associated herbivorous insects in these populations, and G. linnaeana is thereby liberated from interspecific competition. Phytophagous insects often have low mobility in their immature stages, and have thus a limited opportunity to influence the location of their development. It has therefore been argued that the larvae of holometabolous herbivorous insects are rarely forced to choose between different substrates, and have consequently not been subject to selection for behavioral adaptations (Roininen and Tahvanainen 1989). In both papers III and IV we could see that palatability varies between clones, and that our larvae are likely to have the ability to rank them as more or less suitable resources. A very common behavior of the larvae was first to crawl around and nibble on most or all of the leaf discs present, before deciding which one to feed on (pers. obs.). This suggests that there may be selection pressure on the larvae of this leaf beetle for an ability to identify suitable host individuals. This can be explained by the behavior of the larvae in dropping to the ground when disturbed (pers. obs.). This is most likely a defensive behavior, but it has the consequence that the larvae have to find and crawl back up onto a host plant to resume feeding. In the two-choice cafeteria trials presented in paper III, the larvae did not show any preference for sympatric hosts. Host suitability was tested in the performance test presented in paper IV, and even though it was obvious that the local hosts were generally more suitable as food for the larvae, the results from paper III show that although there are individual differences between hosts, the acceptability of a host

21

may be dependent on the actual combination of clones available to choose between. However, adaptation to the resistance structure of local hosts should make it more likely that sympatric hosts would be more suitable than allopatric hosts. It is not clear what the basis is of recognition and acceptability of G. linnaeana in this system. However, the results in paper IV emphasize how strongly influenced G. linnaeana is by the variability of the host plant S. triandra, and this may explain the spatial distribution of this leaf beetle in Sweden. Dealing with local environmental conditions (Papers V and VI) The distribution of herbivores in a host population may be determined, for example, by competition, predation, or individual differences in host defense levels. Gonioctena linnaeana meets with no competition from other leaf beetles on Salix triandra in these populations, but the gall-maker Pontania triandrae is quite common in some populations. In addition, the pathogenic rust Melampsora amygdalinae is present in all populations. We have no indications of predation on the larvae of G. linnaeana and there are no parasitoids associated with this species in the surveyed populations. Therefore, we assume that the distribution of parasites in these populations is mainly caused by individual variations in secondary chemistry. A joint plot of the PCA in paper V places the preferred and non-preferred hosts in two separate clusters, and it is two flavonoid compounds – luteolin-7-glucoside and the unidentified flavonoid 2 – that are important for this separation. Luteolin-7-glucoside has earlier been identified both as toxic to a non-adapted coleopteran species (Regnault-Roger et al. 2004) and as a feeding stimulant to several species of leaf beetles (Matsuda & Matsuo 1985). Other differences, not, however, that obvious in the joint plot, are that preferred clones do not contain (+)-catechin whereas non-preferred hosts do, and that the concentration of condensed tannins is higher in the non-preferred clones (V). Tannic compounds may thus be of importance for host selection by adult females of G. linnaeana and also for the performance of the larvae. This is in line with the results of some studies (Ikonen et al. 2002, Donaldson & Lindroth 2004, Stenberg et al. 2006) but in contrast to those of other studies that have failed to detect any negative effects of (+)-catechin on the feeding of adults or on the performance of the larvae of oligophagous leaf beetles (Kelly and Curry 1991, Ikonen et al. 2002). It has also been suggested that condensed tannins have no function as feeding deterrents or toxins to insects, especially not to insects that are specialized on plants characterized by high concentrations of these compounds (Kelly and Curry 1991, Ayres et al. 1997). The contrasting results of earlier studies and the results of paper V indicate that the response to tannins may differ greatly between insect species, a finding supported by the distribution of the gall maker, P. triandrae (VI). Pontania triandrae seems to be attracted by plants with high concentrations of tannins, since gall density was higher

22

on such clones. This is probably explained by the fact that high concentrations of phenolics often indicate vigorous growth, and galling sawflies are generally assumed to prefer fast-growing plants (Price et al. 1989, Price 1993, Kolehmainen et al. 1994). Paper V reports a significantly higher concentration of salidroside in the non-preferred than in the preferred plants. Salidroside is a compound previously suggested as a possible stimulant of phytophagous insects associated with S. triandra (Tahvanainen et al. 1985). Although salidroside may be involved in host recognition by insects specialized on S. triandra, our results suggest no clear correlation between salidroside concentration and acceptability in G. linnaeana, since the non-preferred plants had a significantly higher concentration of salidroside than the preferred plants did (V). Furthermore, the leaf damage caused by G. linnaeana in the population surveyed in paper VI displayed no relationship with the concentration of salidroside. However, mechanisms involved in host recognition in G. linnaeana may differ from those determining the acceptability and suitability of a specific plant individual. The presence of specific compounds, such as salidroside, may be important for the recognition of S. triandra as a host plant, but threshold values of the concentrations may affect the acceptability and suitability of the plant, and this step may also include, for example, nitrogen and water content. The range of host plants that is acceptable for females of G. linnaeana would thus be narrower than that of the recognizable hosts (Tahvanainen et al. 1985, Kelly & Curry 1991, Kolehmainen et al. 1995, van der Meijden 1996, Rank et al. 1998, Leimu et al. 2005). For the gall-maker P. triandrae, however, it seems likely that high concentrations of salidroside may function as an attractant, because clones with high gall densities were also the clones with the highest concentrations of salidroside (VI). The two different populations of S. triandra surveyed in papers V and VI differ slightly in chemical composition and also in the concentrations of shared compounds. The production of secondary chemicals is not only affected by genotype but also by environmental factors (Haukioja & Neuvonen 1985, Karban 1987). Biotic factors, such as herbivory, exert a very strong selection pressure for increased resistance in a population (Leimu et al. 2005). In addition, abiotic environmental factors, such as resource availability and mechanical damage, affect plant allocation patterns and may therefore affect plant resistance. Most empirical evidence is related to damage caused by herbivory, and the chemical differences between the two host populations examined in papers V and VI may well result from differential selection by the different insect species (Thompson 1994, Stinchcombe & Rausher 2002, Strauss et al. 2005 and references therein). Gonioctena linnaeana is the dominant herbivore in the population surveyed in paper V and P. triandrae is dominant in the population surveyed in paper VI. Still, resistance traits that provide resistance to one type of enemy do not necessarily confer resistance to others. Low-gall clones were, in contrast to the pattern for gallers, more susceptible to rust infections. The damage caused by leaf-beetles, on the other hand, did not differ between clone types.

23

However, in the case of the population examined in paper V, there is an additional abiotic environmental factor that may be important for the differences between the preferred and non-preferred plants. We remarked that some plants in the population displayed damage caused by ice break-up and flooding in the spring. When we collected the larvae in the field, we noticed that these plants tended to harbor more larvae of G. linnaeana than did plants without such damage. Environmental factors, including those causing damage, have been shown to affect individual trade-offs in plant metabolism and may cause allocation patterns to be altered (Stamp & Bowers 1994, Keinänen et al. 1999, Osier & Lindroth 2004). If an annual loss of biomass affects the production of secondary compounds in S. triandra, it may be that not only plant genotype but also environment may be responsible for the distribution of insect herbivores in this population. CONCLUSIONS In this thesis I have demonstrated that the interaction between S. triandra and the associated parasites M. amygdalinae and G. linnaeana has resulted in local adaptation of the parasites to the host. These interactions occur in populations isolated by large distances, which may have important implications for the outcome of plant–parasite interactions. Locally adapted parasites can be expected in systems in which the parasite is host specific, has great evolutionary potential, and the interaction is allowed to work without homogenization by a large gene flow. The distance between our sampled populations not only promoted the development of local adaptations, but probably also increased the likelihood of finding them, since it is important to adjust the scale of the study to include several local parasite populations. The secondary chemistry differs both among populations and between individuals within them. Salidroside may be important for insect recognition, but it seems as though the precise combination and concentration of all compounds is important; in particular, it seems as though luteolin-7-glucoside, an unidentified flavonoid, and the concentration of tannins determine the acceptability and suitability of plant individuals. Our results also indicate that plant traits that provide resistance to one type of herbivore do not necessarily provide resistance to others. This suggests that different natural enemies could potentially exert divergent selection pressure on S. triandra phenotypes; this could be important for maintaining phenotypic variation in plant species, and our results provide some support for this view. Furthermore, local adaptation is an important mechanism for maintaining genetic variation in a population as a whole, and it is also an important step towards speciation (Huyse et al. 2005). However, in our case the host is a clonal plant, and the adaptation of parasites to clonal hosts may have implications for the longevity of the interaction. Vegetative reproduction does not generate genetic variation in the

24

population, and clonal hosts thus have less potential to respond to selection. If the parasites have severe detrimental effects on the host, the interaction may not last long enough for speciation processes to take place. In our studied populations there are still host individuals that resist attack from the parasites, but since the populations are isolated at the margin of the distribution range and seedling establishment is rare, new resistance genotypes may not appear sufficient rapidly and the number of parasites with the ability to evolve adaptations to S. triandra is likely to increase with time. With an increasing parasitic load and limited seedling recruitment, the persistence of S. triandra in Öreälven and Torneälven may be threatened. However, if abiotic environmental conditions significantly affect the resistance level, this may counteract the negative trend and support more persistent interactions. ACKNOWLEDGEMENTS I would like to thank Anders Wennström and Lars Ericson for their helpful comments on earlier versions of this thesis. Financial support for the research presented in this thesis was kindly provided by the Gunnar and Ruth Björkman Foundation, the J.C. Kempes Minnes Akademiska fond, and the Helge Ax:son Johnson Foundation to LN and by grants from the Oscar and Lili Lamm Foundation and the Swedish Research Council for Environment, Agricultural Sciences and Spatial Planning (FORMAS) to AW. REFERENCES AGRIOS, G.N. (1997) Plant pathology. 4th ed. Academic Press, San Diego, USA. AYRES, M.P., CLAUSEN, T.P., MACLEAN, JR., S.F. REDMAN, A.M. & REICHARDT,

P.B. (1997) Diversity of structure and herbivore activity in condensed tannins. Ecology 78(6): 1696-1712.

BERENBAUM, M. (1980) Adaptive significance of midgut pH in larval Lepidoptera. American Naturalist 115:138-146

BERENBAUM, M.R. & ZANGERL, A.R. (1992) Genetics of behavioural and physiological resistance to host furanocoumarins in the parsnip webworm. Evolution 46: 1373-1384

BERNAYS, E.A. (2001) Neural limitations in phytophagous insects: Implications for diet bredth and evolution of host affiliation. Annual Reviews of Entomology 46: 703-727.

BRATTSTEN, L.B. (1992) Metabolic defenses against plant allelochemicals. In: Rosenthal, G.A. and Berenbaum, M.R. (eds.), Herbivores. Their interactions

25

with secondary plant metabolites. Vol. II: Ecological and evolutionary processes, pp 439-475. 2nd ed. Academic press, inc, USA.

BURDON, J.J. (1987) Diseases and plant population biology. Cambridge University Press, Cambridge, UK.

BURDON, J.J. & THRALL, P.H. (1999) Spatial and temporal patterns in coevolving plant and pathogen associations. American Naturalist 153: S15-S33.

BURDON, J.J., WENNSTRÖM, A. & ELMQVIST, T. (1996) The role of race specific resistance in natural plant populations. Oikos 76(2): 411-416.

DAVELOS, A.L., ALEXANDER, H.M. & SLADE, N.A. (1996) Ecological genetic inter- actions between a clonal host plant (Spartina pectinata) and associated rust fungi (Puccinia seymouriana and Puccinia sparganioides). Oecologia 105: 205-213.

DE BLOCK, M. & STOKS, R. (2005) Fitness effects from egg to reproduction: Bridging the life history transition. Ecology 86: 185-197.

DEL CAMPO, M.L., VIA, S. & CAILLAUD, M.C. (2003) Recognition of host-specific chemical stimulants in two sympatric host races of the pea aphid Acyrthosiphon pisum. Ecological Entomology 28: 405-412.

DOBSON, A. CRAWLEY, M. (1994) Pathogens and the structure of plant communities. Trends in Ecology & Evolution 9(10): 393-399.

DONALDSON, J.R. & LINDROTH, R.L. (2004) Cottonwood leaf beetle (Coleoptera: Chrysomelidae) performance in relation to variable phytochemistry in juvenile aspen (Populus tremuloides Michx.). Environmental Entomology 33: 1505- 1511.

EDMUNDS, G.F. & ALSTAD, D.N. (1978) Coevolution in insect herbivores and conifers. Science 199: 941-945.

EHRLICH, P.R. & RAVEN, P.H. (1964) Butter- flies and plants: a study in coevolution. Evolution 18: 586-608.

FEENY, P. (1992) The evolution of chemical ecology: contributions from the study of herbivorous insects. In: Rosenthal, G.A. and Berenbaum, M.R. (eds.), Herbivores. Their interactions with secondary plant metabolites. Vol. II: Ecological and evolutionary processes, pp 1-44. 2nd ed. Academic press, inc, USA.

FUJIYAMA, N. & KATAKURA, H. (1997) Individual variation in two host plants of the ladybird beetle, Epilachna pustulosa (Coleoptera: Coccinellidae). Ecological Research 12: 257-264.

FUTUYMA, D.J & KEESE, M.C. (1992) Evolution and coevolution of plants and phytophagous arthropods. In: Rosenthal, G.A. and Berenbaum, M.R. (eds.), Herbivores. Their interactions with secondary plant metabolites. Vol. II: Ecological and evolutionary processes, pp 439-475. 2nd ed. Academic press, inc, USA.

GANDON, S. & MICHALAKIS, Y. (2002) Local adaptation, evolutionary potential and host–parasite coevolution: interactions between migration, mutation, population size and generation time. Journal of Evolutionary Biology 15: 451- 462.

GANDON, S. (2002) Local adaptation and the geometry of host–parasite coevolution. Ecology Letters 5: 246-256.

26

GANDON, S., CAPOWIEZ, Y., DUBOIS, Y., MICHALAKIS, Y. & OLIVIERI, I. (1996) Local adaptation and gene-for-gene coevolution in a metapopulation model. Proceedings of the Royal Society of London B 263: 1003-1009.

GÄUMANN, E. (1959) Die Rostpilze Mitteleuropas mit besonderer Berücksichtigung der Schweiz. Bern. Switzerland.

GJAERUM, H.B. (1974) Nordens rustsopper. Fungiflora. Oslo. GLYNN, C., RÖNNBERG-WÄSTLJUNG, A. C., JULKUNEN-TIITTO, R. & WEIH, M.

(2004) Willow genotype, but not drought treatment, affects foliar phenolic concentrations and leaf-beetle resistance. Entomologia Experimentalis et Applicata 113: 1-14.

HAUKIOJA, E. & NEUVONEN, S. (1985) Induced long-term resistance of birch foliage against defoliators: Defensive or incidental? Ecology 66(4): 1303-1308.

HEMMING, J.D.C. & LINDROTH, R.L. (1995) Intraspecific variation in aspen phytochemistry: effects on performance of gypsy moths and forest tent caterpillars. Oecologia 103: 79-88.

HULTÉN, E. & FRIES, M. (1986) Atlas of North European Vascular Plants. North of the Tropic of Cancer I. Map 591. Koeltz Scientific Books, Königstein. Germany.

HUYSE, T. POULIN, R. & THÉRON, A. (2005) Speciation in parasites: a population genetics approach. Trends in Parasitology 21(10): 469-475).

IKONEN, A., TAHVANAINEN, J. & ROININEN, H. (2002) Chlorogenic acid as an antiherbivore defence of willows against leaf beetles. Entomologia Experimentalis et Applicata 99: 47-54.

IKONEN, A., TAHVANAINEN, J. & ROININEN, H. (2002) Phenolic secondary compounds as determinants of the host plant preferences of the leaf beetle, Agelastica alni. Chemoecology 12: 125-131.

JAENIKE, J. (1990) Host specialization in phytophagous insects. Annual Review of Ecology and Systematics 21: 243-273.

JAROSZ, A.M. & BURDON, J.J. (1991) Host–pathogen interactions in natural populations of Linum marginale and Melampsora lini: II. Local and regional variation in patterns of resistance and racial structure. Evolution 45: 1618- 1627.

JULKUNEN-TIITTO, R. (1989) Distribution of certain phenolics in Salix species Salicaceae. PHd Thesis. University of Joensuu publications in Science. No: 15, Finland.

JULKUNEN-TIITTO, R., ROUSI, M., BRYANT, J. P., SORSA, S., KEINÄNEN, M. & SIKANEN, H. (1996) Chemical diversity of several Betulaceae species: comparison of phenolics and terpenoids in Northern birch stems. Trees 11: 16- 22

KALTZ, O. & SHYKOFF, J.A. (1998) Local adaptation in host–parasite systems. Heredity 81: 361-370.

KALTZ, O., GANDON, S., MICHALAKIS, Y. & SHYKOFF, J.A. (1999) Local maladaptation in the anther-smut fungus Microbotryum violaceum to its host plant Silene latifolia: evidence from a cross-inoculation experiment. Evolution 53: 395-407.

27

KARBAN, R. (1987) Environmental conditions affecting the strength of induced resistance against mites in cotton. Oecologia 73(3): 414 – 419

KAWECKI, T.J. & EBERT, D. (2004) Conceptual issues in local adaptation. Ecology Letters 7: 1225-1241.

KEINÄNEN, M. JULKUNEN-TIITTO, R., ROUSI, M. & TAHVANAINEN, J. (1999) Taxonomic implications of phenolic variation in leaves of birch (Betula L.) species. Biochemical Systematics and Ecology 27: 243-254.

KELLY, M.T. & CURRY, J. P. (1991) The influence of phenolic compounds on the suitability of three Salix species as hosts for the willow beetle Phratora vulgatissima. Entomologia Experimentalis et Applicata 61: 25-32.

KOCH, K. (1992). Die Käfer Mitteleuropas, Ökologie Band 3. Goecke &Evers Verlag, Krefeld, Germany.

KOLEHMAINEN, J., ROININEN, H., JULKUNEN-TIITTO, R. & TAHVANAINEN, J. (1994) Importance of phenolic glucosides in host selection of shoot galling sawfly, Euura amerinae on Salix pentandra. Journal of Chemical Ecology 20: 2455- 2466.

KRISCHIK, V.A., GOTH, R.W. & BARBOSA, P. (1991) Generalized plant defense: effects of multiple species. Oecologia 85: 562-571.

LABEYRIE, E. & DOBLER, S. (2004) Molecular adaptation of Chrysochus leaf beetles to toxic compounds in their food plants. Molecular Biology and Evolution 212: 218-221.

LEIMU, R. RIIPI, M. & STAERK, D. (2005) Food preference and performance of the larvae of a specialist herbivore: variation among and within host–plant populations. Acta Oecologica 28: 325-330.

LEVIN, D.A. (1976) The chemical defenses of plants to pathogens and herbivores. Annual Review of Ecology and Systematics 7: 121-59.

LUNDQVIST, J. & ROBERTSSON, A-M. (1994) Inlandsisens avsmältning and Vegetationens utveckling. Berg och jord. Sveriges nationalatlas. (Eds U. Arnberg, & C. Fredén,), pp. 124-137. Bokförlaget Bra Böcker, Höganäs. Sweden.

MALECK, K. & DIETRICH, R.A. (1999) Defense on multiple fronts: how do plants cope with diverse enemies? Trends in Plant Science 4(6): 215-219.

MASCHINSKI, J. & WHITHAM T.G. (1989) The continuum of plant responses to herbivory: The influence of plant association, nutrient availability, and timing. American Naturalist 134(1): 1-19.

MATSUDA, K. & MATSUO, H. (1985) A flavonoid, luteolin-7-glucoside, as well as salicin and populin, stimulating the feeding of leaf beetles attacking salicaceous plants. Applied Entomology and Zoology 20(3): 305-313.

MOPPER, S., BECK, M., SIMBERLOFF, D. & STILING, P. (1995) Local adaptation and agents of selection in a mobile insect. Evolution 49: 810-815.

MOSSBERG, B., STENBERG, L. & ERICSSON, S. (1992) Den nordiska floran. Wahlström & Widstrand, Stockholm. Sweden.

NYLIN, S. & GOTTHARD, K. (1998) Plasticity in life-history traits. Annual Review of Entomology 43: 63-83.

OPPENHEIM, S.J. & GOULD, F. (2002) Behavioural adaptations increase the value of enemy-free space for Heliothis subflexa, a specialist herbivore. Evolution

28

56(4): 679-689. OSIER, T.L. & LINDROTH, R.L. (2004) Long-term effects of defoliation on quaking

aspen in relation to genotype and nutrient availability: plant growth, phytochemistry and insect performance. Oecologia 139: 55-65.

PARKER, M.A. (1985) Local population differentiation for compatibility in an annual legume and its host-specific fungal pathogen. Evolution 39: 713-723.

PHILLIPS, N.E. (2004) Variable timing of larval food has consequences for early juvenile performance in a marine mussel. Ecology 85: 2341-2346.

RANK, N.E., KÖPF, A., JULKUNEN-TIITTO, R. & TAHVANAINEN, J. (1998) Host preference and larval performance of the salicylate-using leaf beetle Phratora vitellinae. Ecology 79(2): 618-631.

REGNAULT-ROGER, C., RIBODEAU, M., HAMRAOUI, A., BAREAU, I., BLANCHARD, P., GIL-MUNOZ, M.-I. & BARBERAN, F.T. (2004) Polyphenolic compounds of Mediterranean Lamiaceae and investigation of orientational effects on Acanthoscelides obtectus (Say). Journal of Stored Products Research 40(4): 395-408.

RENWICK, J.A.A. (2001) Variable diets and changing taste in plant-insect relationships. Journal of Chemical Ecology 27: 1063-1076.

ROININEN, H. & TAHVANAINEN, J. (1989) Host selection and larval performance of two willow-feeding sawflies. Ecology 70(1): 129-136.

ROY, B.A. (1998) Differentiating the effects of origin and frequency in reciprocal transplant experiments used to test negative frequency-dependent selection hypotheses. Oikos 115: 73-83.

SCHOONHOVEN, L.M., JERMY, T. & VAN LOON, J.J.A. (1998) Insect – plant biology, from physiology to evolution. Chapman and Hall, London.

SKVORTSOV, A.K. (1999) Willows of Russia and adjacent countries. Taxonomial and Geographical review. Univ. Joensuu Fac. Mathem. and Natur. Sci. Rept. Ser. 39.

STAMP, N.E. & BOWERS, M.D. (1994) Effects of cages, plant age and mechanical damage on plantain chemistry. Oecologia 99: 66-71.

STENBERG, J.A., WITZELL, J. & ERICSON, L. (2006) Tall herb herbivory resistance reflects historic exposure to leaf beetles in a boreal archipelago age-gradient. Oecologia, in press.

STILING, P. & ROSSI, A.M. (1998) Deme formation in a dispersive gall-forming midge. In: S. Mopper & S.Y. Strauss (Eds), Genetic structure and local adaptation in natural insect populations: effects of ecology, life history, and behaviour, pp 22-36. Chapman & Hall, New York.

STINCHCOMBE, J.R. &. RAUSHER, M.D (2002) Evolution of tolerance to deer herbivory: modifications caused by the presence or absence of insects. Proceedings of the Royal Society of London, Ser. B 269: 1241-1246.

STRAUSS, S.Y. (1990). The role of plant genotype, environment and gender in resistance to a specialist chrysomelid herbivore. Oecologia 84: 111-116.

STRAUSS, S.Y., SAHLI, H. & CONNER, J.K. (2005) Towards a more trait-centered approach to diffuse (co) evolution. New Phytologist 165: 81-90.

STÄDLER, E. (1992) Behavioural responses of insects to plant secondary compounds. In: Rosenthal, G.A. and Berenbaum, M.R. (eds.), Herbivores. Their interactions

29

with secondary plant metabolites. Vol. II: Ecological and evolutionary processes, pp 439-475. 2nd ed. Academic press, inc, USA.

TAHVANAINEN, J., JULKUNEN-TIITTO, R. & KETTUNEN, J. (1985) Phenolic glycosides govern the food selection pattern of willow feeding leaf beetles. Oecologia 67: 52-56.

THOMPSON, J.N. (1988) Evolutionary ecology of the relationship between oviposition preference and performance of offspring in phytophagous insects. Entomologia Experimentalis et Applicata 47: 3-14.

THOMPSON, J.N. (1994) The coevolutionary process. The University of Chicago press, Chicago, USA.

THOMPSON, J.N. (1999) Specific hypothesis on the geographic mosaic of coevolution. American Naturalist 153: S1-S14.

THRALL, P.H., BURDON, J.J. & BEVER, J.D. (2002) Local adaptation in the Linum marginale-Melampsora lini host–pathogen interaction. Evolution 56: 1340- 1351.

THRALL, P.H. & BURDON, J.J. (2003) Evolution of virulence in a plant host– pathogen metapopulation. Science 299: 1735-1737.

TIKKANEN, O-P., NIEMELÄ, P. & KERÄNEN, J. (2000) Growth and development of a generalist insect herbivore, Operophtera brumata, on original and alternative host plants. Oecologia 122: 529-536.

VAN DER MEIJDEN E. (1996) Plant defense, an evolutionary dilemma: contrasting effects of (specialist and generalist) herbivores and natural enemies. Entomologia Experimentalis et Applicata 80: 307-310.

WARCHALOWSKI, A. (2003) Chrysomelidae: the leaf-beetles of Europe and the Mediterranean area. Warszawa: Natura optima dux Foundation.

WILSON, M. & HENDERSON, D.M. (1966) British rust fungi. Cambridge University Press, Cambridge, UK.

30

TACK TILL… Anders. Då är jag slutligen där. Vid vägs ände. Ärlig talat så har jag haft mina tvivel om att jag skulle få uppleva detta ögonblick och en sak är säker, utan ditt stöd hade jag aldrig nått fram (åtminstone inte till våra lokaler vid Dalälven…). Tack för en rolig och stimulerande tid, för ditt imponerande tålamod, för all uppmuntran och för din förmåga att ta ner mig på jorden när det ville braka iväg. Lasse E, för att du trodde på mig. Utan dig hade jag aldrig börjat och du är verkligen en klippa med din outsinliga kunskap, din entusiasm och din förmåga att alltid ta dig tid att svara på alla upptänkliga frågor. Ulla, tack för att du finns. Din ofelbart positiva inställning och uppenbara tillit har faktiskt fått mig att tro att jag kan. Åsa H, du är en fantastisk inspiratör, mina svampkunskaper skulle inte ha varit ens i närheten av vad de är idag utan dig. Min rumskamrat Patrik. Vad kan man säga? Tack för din oklanderliga koll på äppleskrottarna, för trevligt sällskap, muntra tillrop och fina bildspel. Jag antar att det även vore på sin plats med en ursäkt för mitt dåliga inflytande på dig. Det börjar se bra stökigt ut på ditt skrivbord! Viktoria, min ständiga reskamrat. Vem annars skulle vara enveten nog att rusa igenom Berlin Zoo på 90 minuter, i 35gradig värme och med fågelskit i håret? Och du, Yorkshire kommer alltid att minnas den vinröda tröjan… Åsa G, Jörgen, Ante, Maria, Zlatko, Emma, Veronica, Carola, Ulrika och…. Ja, tack till alla er doktorander som fortfarande jobbar på det, och givetvis också till er som redan hunnit lägga doktorstiteln till ert namn. Ni har alla, var och en på sitt sätt bidragit till att förgylla tillvaron för mig under den här tiden. Tack även till alla er andra som är en del av EMG. Tack för alla öppna dörrar och den glada och trivsamma gemenskap som råder. Eftersom det trots allt finns ett liv utanför EMG så är det några fler som förtjänar ett tack för att de låtit mig vara en del av deras liv: Cia, som tog hand om mig då jag var ny, ensam och vilsen i Umeå. Och även senare då jag inte var fullt så ny eller ensam men kanske fortfarande lite vilsen… Barbro Sjöberg, egentligen borde avhandlingen ha tillägnats dig eftersom det faktiskt var du som övertygade mig om att jag borde söka in på Biologprogrammet, där allt började. Carina, den bästa vän någon kan önska. Alltid där, alltid med glimten i ögat, alltid sen…. Tack för alla pratstunder, fikor, goda middagar, utflykter. Mamma och pappa för att ni alltid funnits vid min sida, och alltid stöttat mig vad än jag fått för mig att jag skulle göra. Storasyster Marie för att du varit och är en sådan god vän och lillebror Lasse för att du är en sådan fantastisk bror, alltid beredd att ställa upp. Till sist har jag sparat de viktigaste och mest betydelsefulla: Ulf, tack för att du härdat ut, vad vore jag utan dig? Simon, Moa, Pontus, mitt hjärta är ert.

31