Spatial ecology of an -associated herbivore community

Ayco TACK

Department of Biosciences Faculty of Biological and Environmental Sciences University of Helsinki Finland

Academic dissertation

To be presented for public examination with the permission of the Faculty of Biological and Environmental Sciences of the University of Helsinki in the Auditorium 1401 in Biocentre 2 (Viikinkaari 5) on May 28th 2010 at 12 o´clock noon. Supervised by: Dr. Tomas Roslin Department of Biosciences and Department of Agricultural Sciences University of Helsinki, Finland

Reviewed by: Dr. Tero Klemola Department of Biology University of Turku, Finland

Prof. Peter Hambäck Department of Botany Stockholm University, Sweden

Examined by: Dr. Marc T. J. Johnson Department of Plant Biology North Carolina State University, USA

Custos:Supervised by: Prof.Dr. Tomas Ulrika Roslin Candolin Department of Biosciences and Department of Agricultural Sciences University of Helsinki, Finland Reviewed by: Dr. Tero Klemola Department of Biology Expert members of the thesis Universityadvisory committee: of Turku, Finland ISBN 978-952-10-6194-3 (paperback) Prof.ISBN Jari 978-952-10-6195-0 Kouki (PDF)Prof. Peter Hambäck Department of Botany Schoolhttp://ethesis.helsinki.fi of Forest Sciences Stockholm University, Sweden University of Eastern Finland, Finland Examined by: Dr. Marc T. J. Johnson YliopistopainoProf. Heikki Roininen Helsinki 2010 Department of Plant Biology Department of Biology North Carolina State University, USA University of Eastern Finland, Finland Custos: Prof. Ulrika Candolin Prof. Pekka Niemela Department of Biosciences Department of Biology University of Helsinki, Finland University of Turku, Finland

ISBN 978-952-10-6194-3 (paperback) ISBN 978-952-10-6195-0 (PDF) http://ethesis.helsinki.fi

Yliopistopaino Helsinki 2010

Contents

CONTRIBUTIONS 5

ABSTRACT 6

SUMMARY 7

1. Introduction ...... 7

2. Oak-based metacommunities as a model system ...... 9 2.1 9 2.2 Gallers, leaf miners and leaf folders 9 2.3 Natural enemies 9

3. Material and Methods ...... 9 3.1 Observational data 10 3.2 Experiments 11

4. Insights ...... 11 4.1 Direct competition among herbivores is of secondary importance (I) 11 4.2 Indirect competition mediated by the host plant plays a minor role (I) 13 4.3 Indirect interactions mediated by shared parasitoids can be positive (II) 14 4.4 Spatial location is more important than host plant genotype in structuring the community (III) 15 4.5 Gene flow among local insect populations affects local adaptation (IV) 16

5. Conclusion ...... 16

6. Future perspectives ...... 17

7. Acknowledgements ...... 18

8. References ...... 19

I competition as a structuring force in leaf miner communities 23

II Can we predict indirect interactions from quantitative food webs? – An experimental approach 35

III Spatial location dominates over host plant genotype in structuring an herbivore community 49

IV overrun by the neighbors: landscape context affects strength and sign of local adaptation 69 The thesis is based on the following articles, which are referred to in the text by their Roman numerals:

I Tack, A. J. M., Ovaskainen, O., Harrison, P. J., and T. Roslin. 2009. Competition as a structuring force in leaf miner communities. Oikos 118: 809-818

II Tack, A. J. M., Gripenberg, S., and T. Roslin. Can we predict indirect interactions from quantitative food webs? – An experimental approach. Manuscript.

III Tack, A. J. M., Ovaskainen, O., Pulkkinen, P., and T. Roslin. 2010. Spatial location dominates over host plant genotype in structuring an herbivore community. Ecology, in press.

IV Tack, A. J. M. and T. Roslin. 2010. Overrun by the neighbors: landscape context affects strength and sign of local adaptation. Ecology, in press.

CONTRIBUTIONS

Contribution table:

I II III IV

Original idea TR, AT AT AT, TR TR

Study design AT, TR AT AT, TR, PP TR, AT

Data collection AT AT, SG AT AT

Analyses AT, OO, TR, AT AT, OO, TR AT PH Manuscript AT, TR AT, TR AT, TR AT, TR preparation

AT = Ayco Tack PP = Pertti Pulkkinen TR = Tomas Roslin OO = Otso Ovaskainen SG = Sofia Gripenberg PH = Phil Harrison

© Ayco Tack (Summary) © Wiley Interscience (Chapter I) © The Authors (Chapter II) © Ecology / ESA journals (Chapters III & IV)

Cover illustration © Jasper Teunissen Layout © Sami Ojanen ABSTRACT

Habitat fragmentation is currently affecting many impact of direct and indirect competition, host plant species throughout the world. As a consequence, genotype and local adaptation (i.e. local factors) to an increasing number of species are structured that of regional processes (as reflected by the spatial as metapopulations, i.e. as local populations context of the local community). As a key approach, connected by dispersal. While excellent studies of I use general theory to generate testable hypotheses, metapopulations have accumulated over the past controlled experiments to establish causal relations, 20 years, the focus has recently shifted from single and observational data to validate the role played by species to studies of multiple species. This has the pinpointed processes in nature. created the concept of metacommunities, where As the central outcome of my thesis, I am able to local communities are connected by the dispersal relegate local forces to a secondary role in structuring of one or several of their member species. To oak-based insect communities. While controlled understand this higher level of organisation, we need experiments show that direct competition does to address not only the properties of single species, occur among both conspecifics and heterospecifics, but also establish the importance of interspecific that indirect interactions can be mediated by both interactions. However, studies of metacommunities the host plant and the parasitoids, and that host plant are so far heavily biased towards laboratory-based genotype may affect local adaptation, the size of these systems, and empirical data from natural systems are effects is much smaller than that of spatial context. urgently needed. Hence, I conclude that dispersal between habitat My thesis focuses on a metacommunity of insect patches plays a prime role in structuring the insect herbivores on the pedunculate oak Quercus robur – community, and that the distribution and abundance a tree species known for its high diversity of host- of the target species can only be understood in a specific . Taking advantage of the amenability spatial framework. By extension, I suggest that the of this system to both observational and experimental majority of herbivore communities are dependent studies, I quantify and compare the importance of on the spatial structure of their landscape – and urge local and regional factors in structuring herbivore fellow ecologists working on other herbivore systems communities. Most importantly, I contrast the to either support or refute my generalization.

6 A metacommunity of herbivores

SUMMARY

Ayco Tack

Metapopulation Research Group, Department of Biosciences, PO Box 65, FI- 00014, University of Helsinki, Finland

1. Introduction

Throughout the world, natural habitats are becoming Hanski 2005, Cronin 2007). Nonetheless, a focus increasingly fragmented (Hanski 2005). As a result, on this group would be justified for several reasons. more and more species are split into metapopulations As plants, plant-feeding insects and parasitoids make (Fig. 1A; Hanski 1999, Hanski and Gaggiotti 2004). up approximately half of the world’s eukaryotic During the last few decades, research has revealed species (Strong et al. 1984, Godfray 1994), they how spatially structured dispersal within such have an enormous impact on ecological systems. metapopulations may critically affect what species Here, an understanding of the natural dynamics of occur where and at what abundances (Thomas plants, herbivores and their parasitoids may then and Kunin 1999). More recently, ecologists also pay off in economic terms, e.g. by yielding an have extended the focus from metapopulations improved understanding of pest control in agro- to metacommunities, i.e. to metapopulations of ecosystems (Losey and Vaughan 2006). Moreover, multiple, potentially interacting species (Fig. 1B; many insect taxa are currently in a state of decline, Leibold et al. 2004, Holyoak et al. 2005). In such and thus conservation efforts require an increasing metacommunities, the abundance and distribution knowledge of insect population dynamics (Samways of species is explained jointly by the interplay 2005, Stewart et al. 2007). Finally, due to their short between direct and indirect interactions within a local generation times and amenability to experimental community, and by the dispersal of its component manipulation, insects can serve as model systems for species between communities (Amarasekare 2008). Yet, the extent to which the structure of local other species groups (Ehrlich and Hanski 2004). communities is mainly affected by random forces This thesis uses herbivorous insects to offer a (Bell 2001, Hubbell 2001), by interactions among new grip on factors structuring the abundance and species (Holt 1977, Denno et al. 1995, Kaplan and distribution of species, and how these result in patterns Denno 2007), by niche differentiation (Urban 2004, in emergent community descriptors like species Freestone and Inouye 2006) or by differences in richness. Drawing on a combination of observations dispersal abilities (Leibold and Miller 2004) remains and experiments, I quantify both local and regional the topic of an ongoing debate (Cottenie 2005, influences on insect community structure, and Driscoll and Lindenmayer 2009, Pandit et al. 2009, compare their relative contribution. Of local factors, Driscoll et al. 2010). I explore direct interactions among herbivores, In the context of this debate, theory roams wild induced defences by the host plant, impacts of host- but empirical data are scarce. Due to the complexity plant genotype, and indirect interactions mediated of natural communities, many studies have been by the host plant and parasitoids (see Fig. 2). Of conducted on laboratory systems ranging from two regional factors, I analyze the effect of spatial context species to a whole suite of species (Huffaker 1958, on the distribution of species. I achieve this through Kendi et al. 2009). But laboratory studies cannot a determined research programme, with each necessarily be extrapolated to the natural world, and chapter of my thesis addressing one key factor that recently researchers have gone back to the field to may structure an herbivore metacommunity, and, study metacommunities in a range of natural study where possible, placing this factor in an explicitly systems, like invertebrates in rock pools (Kolasa spatial context. When combined, my work covers and Romanuk 2005, Vanschoenwinkel et al. 2007), the following series of questions: zooplankton in fresh water (Cottenie and De Meester 2005), invertebrates in moss patches (Gonzalez • What is the importance of direct interactions 2005) and birds in forest patches (Driscoll and among herbivores? (I) Lindenmayer 2009). • What is the importance of indirect interactions Strikingly, very few studies have been conducted among herbivores, mediated by the host plant? on plant-feeding insects (but see van Nouhuys and (I)

7 Summary

Box A. Populations and communities in space

Populations and communities Individuals of a single species often live together in a distinct area, which is referred to as the habitat. Such a group of conspecific individuals, called a population, has for decades, if not for centuries, been the prime research focus for ecologists interested in ‘population dynamics’. However, in many cases it has proven impossible to understand the dynamics of a single species without considering other species, for example its predator or competitor. Hence, classical models of population dynamics were developed in the early decades of the twentieth century by e.g. Alfred Lotka and Vito Volterra to show how prey species, and prey and predator, interact (Lotka 1925, Volterra 1926, Lotka 1932). Later, ecologists incorporated multiple interacting species in their studies, establishing the field of community ecology (e.g. Elton 1927).

Metapopulations and metacommunities A metapopulation consists of a group of spatially separated populations of the same species, which interact by dispersal, and can hence be seen as a ‘population of populations’ (Fig. 1A; Levins 1970, Hanski and Simberloff 1997). This spatial view adds reality to previous models focusing on single locations, by explicitly incorporating the fact that local populations may sometimes go extinct, and empty habitats be colonized. Many species will naturally occur as metapopulations - like beetles that live in tree hollows (Ranius 2000, 2001), or plants adapted to light conditions in forest gaps (Valverde and Silvertown 1997). Other species have for a long time been continuously distributed in space, but their current distribution has now become scattered with the disappearance and fragmentation of their habitat, and the species now occur as metapopulations (Levins 1970, Hanski 1999). Where the metapopulation framework deals with single species, the metacommunity framework focuses on multiple, potentially interacting species (Fig. 1B). Here, the distribution and abundance of species can be explained by both local interactions and by regional processes. Compared to studies of metapopulations, in studies of metacommunities the focus often moves to emergent properties like species richness and to community composition, where the latter refers to a shift in relative abundances among species.

A) B) Metapopulation Metacommunity ´local populations linked by dispersal` ´local communities linked by dispersal`

Fig. 1. Where studies of metapopulations (left) focus on single species in fragmented habitats, studies of metacommunities (right) focus on multiple species. In both cases, the habitat patch (circle) can be occupied or unoccupied by a local population (as depicted by a of a given species).

8 A metacommunity of herbivores

• Do indirect interactions mediated by shared habitat (Janzen 1968). Hence, the oak–herbivore parasitoids link the dynamics of herbivores? (II) system forms an ideal study target for ecologists • Does host plant genotype play a major role in interested in plant-insect interactions in a spatial structuring the insect community? (III) context. • Do local insect populations adapt to individual host plants? (IV) 2.1 Oaks After the last ice age, Finland has been re-colonized 2. Oak-based metacommunities as a by a single oak species, the pedunculate oak Quercus model system robur (Ferris et al. 1998). Most likely, other oak species are not able to withstand the cold Finnish My thesis is based on a specific model system: the winter with its short growing season, and even the oak Quercus robur and its herbivores. Oak trees are pedunculate oak is limited in its distribution to the relatively long-lived, large entities, and in Finland south-western part of Finland (Fig. 4A). oak trees are often patchily distributed. As a result, they form long-lasting, spatially structured habitats. Many of the herbivores feeding on oak in my study 2.2 Gallers, leaf miners and leaf folders area (Finland) are specialists on Quercus robur. Thus, Most insect herbivores on oak are hard to find – for these specialist insects, the landscape will form both larvae and adults can be difficult to spot due to a mosaic, where each oak tree crown offers suitable their tiny body size and concealed habits. Further, as habitat, and is surrounded by a matrix of unsuitable most species are only observed during a small part of the year, a community ecologist would have to search through the oaks multiple times during the season to record the abundance of several species, significantly increasing the amount of work. Luckily, there are exceptions to this rule: the larval stages C C of the leaf miners, gallers and leaf folders are often E very conspicuous, and remain visible during a large part of the season. As a result, one can greatly reduce the number of times one has to search for the different insect species. Another advantage is that A for some species, the fate of the individual larva can be inferred from the larval structure, even when the larva has already died or left the leaf mine or gall (see Box B for more information on the natural history of D gallers, leaf miners and leaf folders). B B 2.3 Natural enemies A wide variety of natural enemies are known to attack leaf miners and gallers, including birds (Itämies and Ojanen 1975, Owen 1975, Burstein and Wool 1992), ants (Sato and Higashi 1987) and vertebrates (Yamazaki and Sugiura 2008). However, the single most dominant natural enemies are parasitoids (Hawkins et al. 1997; see Box B for details on the life-history of parasitoids). Fig. 2. Interactions within a local herbivore community. Interactions may be direct or indirect, and take place within or between trophic levels. (A) Direct competition 3. Material and Methods among herbivores. (B) Direct interactions among the host plant and the herbivore, including the effect of In this thesis, I combine experimental and host plant genotype and of induced plant responses observational data. Importantly, this allows me on the herbivores. (C) Direct interactions among the to establish the mechanisms behind the patterns, herbivore and parasitoid. (D) Indirect interaction and quantify the importance of different factors in among herbivores, mediated by induced responses in structuring metacommunity dynamics. a shared host plant. (E) Indirect interactions among herbivores, mediated by a shared parasitoid. 9 Summary

Box B. Herbivore guilds and their natural enemies

Leaf miners Leaf mining is an intriguing adaptation, which has evolved several times during the last 300 million years (Labandeira 1998). It is currently found in the orders (butterflies and ), Hymenoptera (wasps), Coleoptera (beetles) and Diptera (flies) (Connor and Taverner 1997). In the typical life cycle of a leaf miner, the adult insect oviposits an egg on (or in) a leaf. The larva will then hatch from the egg, and start eating the tissue between the upper and lower leaf epidermis. Many species specialize in different layers of the leaf, like the photosynthetic cells in the upper layer or the cells used for transport of nutrients and photosynthates in the lower layer. In addition, different species make mines of completely different shapes, ranging from blotch-like to linear structures (Fig 3AB). These different mine patterns – besides being examples of a high degree of specialization - make it relatively easy to identify the leaf miner species (or genus).

Gallers Galling insects are highly diverse on the oak trees. Galling insects evolved 300 MY ago, around the same time as leaf miners (Labandeira 1998), and are intricate examples of how insects may manipulate host plant development. The most striking examples are those where the host larva is concealed by plant tissue (‘the gall’), and feeds upon a nutritious tissue layer lining the inner surface of the gall. Oak trees are especially famous for the large number of cynipid galls (Hymenoptera) that can be found on them (Askew 1961), and which are well-known for their fascinating morphologies (Fig. 3CD).

Leaf folder The larvae of leaf folders fold the leaf to create a concealed habitat for feeding and living in (Fig 3E). Only a single leaf folder species, mitterbacheriana, was observed on oak in our study area.

Powdery mildew In addition to insect herbivores, one can often spot white, powder-like blotches on the oak leaf. These are colonies of the powdery mildew Erysiphe alphitoides (syn. Microsphaera alphitoides) (Fig. 3F).

Parasitoids The insect herbivores are attacked by a range of parasitoid species with a fascinating – but arguably abhorrent - lifestyle. The female adult parasitoid (often a wasp or tachinid fly) uses her ovipositor to lay an egg inside (or on top) of the egg, larva or pupa of the insect herbivore (Godfray 1994). The parasitoid larva then hatches from the egg and starts consuming the living body of its host. In contrast to parasites, a parasitoid will always kill the host larva during its own development. This life style made Darwin write I“ cannot persuade myself that a beneficent and omnipotent God would have designedly created the Ichneumonidae (=group of parasitoid wasps).” Parasitoids fulfil an important ecological and economic role by causing a large fraction of the mortality in herbivore populations - especially in leaf miners and gallers (Hawkins et al. 1997) - and are often used in the biological control of herbivore pests (Van Driesche et al. 2008).

3.1 Observational data For this thesis, much of the observational data was collected on the island Wattkast (Fig. 4). As A fundamental understanding of a metacommunity searching the approximately 100 million oak leaves requires detailed knowledge of the distribution and on the island was (far) beyond our reach, we resorted abundance of its component species throughout the to sampling a subset of trees. At the start of my PhD, landscape. When such data is available for multiple I continued two long-term surveys initiated by years, insight is gained into the persistence of local Sofia Gripenberg and Tomas Roslin in 2003 (and communities, the frequency of local extinctions hopefully continued long in the future). In the first and colonizations, and the level of synchronization survey, we search through the leaves of 89 trees among local populations. distributed across the island Wattkast (Fig. 4B), 10 A metacommunity of herbivores

Box B continued

Fig. 3. Examples of herbivore species on oak. A) blotch-mine of Tischeria ekebladella, B) linear leaf mine of Stigmella, C) spherical gall of the sexual generation of Neuroterus quercusbaccarum, D) doughnut-shaped gall of the asexual generation of Neuroterus numismalis, E) leaf folded by Ancylis mitterbacheriana, and F) colonies (white spots) of the powdery mildew Erysiphe alphitoides. Photos by Ayco Tack, Sofia Gripenberg and Riikka Kaartinen. and record the species richness and abundance of densities on trees in a laboratory setting (Fig. 7A); each herbivore species (I, III). These data illustrate pinpointed parasitoid-mediated indirect interactions both spatial patterns in species richness (Fig. 5A) among herbivores by manipulating host densities in and, when compared among years, colonization- the field (Fig. 7BC); used common gardens at two extinction dynamics (Fig. 5B). We further sample spatial scales to detect the relative importance of host a set of twenty trees growing in a relatively small plant genotype and spatial location (Fig. 7D); and area (ca. 5 hectares; Chapter II). On each of these used grafted trees in reciprocal common gardens to trees, we annually look for herbivores on the same detect local adaptation of insects to individual host five branches. This hierarchical sampling design trees (Fig. 7E). Together, these experiments provide gives additional insight into the level of variation in a versatile view of factors affecting community herbivore densities both within and among trees, structure, and offer stringent tests of hypotheses and whether variation observable at one point in generated by field observations. time persists through time. Interestingly, it seems that some insect species have consistently higher 4. Insights densities on the same trees during multiple years, while others show more or less random abundance 4.1 Direct competition among herbivores is of patterns across trees and years (Fig. 6). secondary importance (I)

3.2 Experiments In the fifties and sixties, resource competition was shown to be a key process in structuring While describing spatial patterns in herbivore communities (MacArthur 1958, Connell 1961). dynamics is fascinating, only experimental work However, Hairston and colleagues (1960) argued can establish the mechanisms behind such patterns. that since the world is demonstrably green, some For this, a series of experiments was conducted, other factor is keeping herbivore densities far below ranging from the manipulation of insect densities levels required to fully exploit the plant resources. in the laboratory to the use of grafted oak trees. Consequently, Hairston and colleagues denounced More specifically, I explored direct and indirect resource competition as a structuring force in host-mediated competition by manipulating host herbivore communities. Since then the popularity

11 Summary

A)

Southwest Finland

Läyliäinen

Wattkast

50 km

B) Wattkast

500 m

Fig. 4. Distribution of oak trees. Panel A shows the natural distribution of oak in Finland. The lower-left arrow marks the location of the island Wattkast, which is shown in more detail in panel B. The upper right arrow shows the location of the tree breeding station in Läyliäinen (Metla), where the experiment on indirect, plant-mediated competition was conducted (I; Fig. 7A). To analyze the importance of host plant genotype (III), we collected host plant genotypes form six locations (circles) and planted them in two common gardens (stars; Fig. 7D). Panel B shows the island Wattkast, where each black dot represents one oak tree (n=1868). Grey dots represent a subset of small trees (1-3 m) which we annually search for the presence/absence of herbivores (since 2003; see Fig. 5 for some of the emerging patterns). The white triangles show 20 medium-sized trees, on which we each year search through the same five branches (see Fig. 6 for herbivore dynamics on these trees).

12 A metacommunity of herbivores

A) B) Species richness 2008 Number of years occupied by T. ekebladella

#Years 0 4 1 5 2 6 3

Fig. 5. Spatial patterns in herbivore species richness and population turnover. Panel A shows the species richness on each of 88 trees in 2008, where the colour of the circles ranges from white (zero species) to red (16 species). The spatial pattern suggests that species richness is higher in dense oak stands as compared to isolated oak individuals. The small black dots each represent one of 1868 oak trees on the island. Panel B shows the number of years that each of these trees was occupied by the leaf mining moth T. ekebladella during the period 2003-2008. The fact that a large fraction of trees was occupied for a few years only indicates the importance of colonization-extinction dynamics. of resource competition has been fluctuating among play a role. At the same time, I show that in the ecologists. In 1966, it was revived when Murdoch absence of parasitoids leaf miner survival is very high (1966, Janzen 1977) stated that while the earth may – and far higher than observed in the field. Hence, appear green and tasty in our eyes, the herbivores I conclude that leaf miners are not like “the ancient will see it as “colored lectin, tannin, cyanide, caffeine, Mariner surrounded by undrinkable water” (Morris aflatoxin, and canavanine”, and experience plants as et al. 2005), but rather live within a juicy and “prickly and tasting bad”. In a world like that, one nutritious leaf. Hence, as hypothesized by Hairston might posit that herbivores aggregate onto the few and colleagues, parasitoids are more likely to play a high-quality resources, and compete fiercely for key role in suppressing leaf miner densities. those. In Chapter I, I assess the occurrence of direct 4.2 Indirect competition mediated by the host competition among herbivores sharing the same leaf. plant plays a minor role (I) Drawing on patterns of co-occurrence in nature, I show that conspecific and heterospecific leaf miners In the eighties, ecologists realized that competition (i.e. individuals of the same or different species) can also be indirect, with one herbivore inducing often aggregate on the same trees I( ). Within trees, a response by the host plant, which in turn affects conspecific individuals show further aggregation on the survival or growth of another herbivore species the same branches, and on the same leaves within (reviewed in Karban and Baldwin 1997). The branches. However, I also show direct competition, induced response often consists of an increase in defined as the mortality due to sharing the leaf with secondary chemicals - but it can, among other things, conspecifics, to be relatively weak – and as a result, also involve a change in the physical structure of the intraspecific competition will only be important leaves or the concentration of nutrients (Karban and when densities are unnaturally high. While some Baldwin 1997). In the mid-nineties, Denno, Ott and competition among heterospecifics may take place, McClure declared the “resurrection of competition”, it will be even rarer than intraspecific competition, at least partly based on a rising number of studies as heterospecifics co-occur less often on the same showing indirect competition (Denno et al, 1995). leaf than do conspecifics. Overall, my results then Over the years, this view has gained added support, support the generalization that direct competition is as shown by a recent meta-analysis by Kaplan & rarely important in natural herbivore communities, Denno (2007). This summary of the status quo as densities are far below carrying capacity. Only at shows convincingly that many interactions among extreme and rare densities may competition then herbivores are indirect and asymmetric. Much

13 Summary

A) Phyllonorycter B) T. ekebladella

Tree 901 Tree 901 Tree 902 Tree 902 Tree 903 Tree 903 Tree 904 Tree 904 Tree 905 Tree 905 Tree 906 Tree 906 Tree 907 Tree 907 Tree 908 Tree 908 Tree 909 Tree 909 Tree 910 Tree 910

/ leaf

/ leaf Tree 911 Tree 911 Tree912 Tree912 Tree 913 Tree 913 Tree 916 Tree 916 Tree 920 Tree 920 Tree 1096 Tree 1096 Tree 1163 Tree 1163 Tree 1341 Tree 1341 Tree 1371 Tree 1371 Tree 1467 Tree 1467

T. ekebladella T.

Phyllopnorycter

Density

0.2 0.4 0.6 0.8 1.0

Density

0.0

0.00 0.05 0.10 0.15 0.20 0.25 0.30

2003 2004 2005 2006 2007 2008 2003 2004 2005 2006 2007 2008 Year Year Fig. 6. Variation in herbivore abundance across trees (n=20) and years (n=6). Each line connects annual tree-specific densities of A) Phyllonorycter and B) T. ekebladella during the period 2003-2008. Note that Phyllonorycter shows higher abundances on given trees across years, whereas the abundances of T. ekebladella show no consistency across either years or trees. See Fig. 4B for the location of the 20 trees. competition thus takes place among herbivores 4.3 Indirect interactions mediated by shared that are spatially and temporally isolated, where the parasitoids can be positive (II) latter refers to early-season herbivores affecting the Since the early twentieth century, researchers survival of herbivores occurring later in the season have shown both empirically and theoretically on the same plant. that density-dependent parasitism may regulate In my work, I examined plant-mediated herbivore populations (Howard and Fiske 1911, competition among leaf miner individuals by Nicholson 1933, Hassell 1978, 2000). In contrast, conducting a controlled experiment (Fig. 7A). More the importance of indirect interactions mediated specifically, I assessed whether leaf miners had a by shared predators was realized only decades later higher mortality when conspecific or heterospecific (Williamson 1957), and it took till the late 1970’s individuals were present on the same tree, but on before the idea of apparent competition permeated different leaves (hence, excluding direct interactions). into the ecological literature – as stimulated by the In this experiment, I detected signs of negative seminal paper by Holt (1977). indirect interactions among both spatially segregated In my work, I took a novel approach to apparent conspecific and heterospecific herbivores. However, competition by manipulating leaf miner densities instead of simply increasing the list of studies that in a large-scale field experiment. To investigate the have shown indirect interactions mediated by the role of parasitoids in keeping herbivore densities host plant, I refined this observation by quantifying low, and to assess the impact of parasitoids shared the importance of plant-mediated indirect among herbivores, I first constructed a quantitative interactions using data on field densities. As a result, food web – that is, a description of who eats who, and I was able to conclude that the effect size of indirect in what numbers (Memmott and Godfray 1993). competition was very small, even at densities higher Based on this web, I made explicit predictions on than those encountered in the field. Hence, my work the strength of density-dependent parasitism and shows that induced responses are unlikely to be a key of indirect interactions among herbivores that share structuring force in the leaf miner community on parasitoids. I then tested these predictions in a large- oak. Based on the insight gained by combining effect scale field experiment, and compared the results sizes quantified in the lab with field observations on with herbivore survival and population dynamics in insect abundances, I urge other researchers to take the observational data. Based on the joint evidence, their lab results back into the field, and quantify the I concluded that delayed density-dependent impact of induced responses in a natural setting. parasitism may affect the densities of herbivores, where the parasitism rate in year t+1 is positively correlated with the density of conspecifics in year t.

14 A metacommunity of herbivores

In addition, I detected indirect interactions among intraspecific genetic variation in structuring natural herbivores sharing parasitoids – however, contrary communities (Price 1983, Antonovics 1992, Fritz to what has often been assumed, the interactions 1992). Empirical research in this field took off were mainly positive. Hence, this study shows for real after a stimulating paper by Whitham and that indirect interactions among herbivores with colleagues in a special issue of Ecology (Whitham overlapping parasitoid communities may range from et al. 2003). Already after a few years, common apparent competition to apparent mutualism. I then garden trials conducted in multiple study systems caution that hypotheses derived from quantitative had convincingly demonstrated that host-plant food webs – e.g. predictions about the strength and genotype can affect the associated insect community sign of indirect interactions among herbivores – (Whitham et al. 2006). However, as pointed out by should be rigorously tested in the field. Johnson and Stinchcombe (2007), the next step still remained: go out into the field and test the 4.4 Spatial location is more important than importance of intraspecific genotypic variation as host plant genotype in structuring the insect compared to other ecological factors. community (III) In my thesis work, I am one of the first to meet this challenge. In Chapter III, I compare the In the field of agriculture and forestry, it is well- relative importance of host-plant genotype to that known that some crop varieties suffer less damage of spatial location at two different spatial scales: the from herbivores than others. However, only recently island Wattkast (ca. 5 km2) and the range of oak in have researchers suggested a potential role for Finland (ca. 10,000 km2). I show that the variation

Fig. 7. Photos of the key experiments. Panel A shows the experiment designed to detect host-mediated competition as conducted at the tree breeding station in Läyliäinen (I). To exclude parasitoids, the experiment was conducted in a large cage covered with fabric. The small bags around tree branches are used to introduce adult moths to specific branches on the tree. Panel B shows a tree on which I increased herbivore densities, in order to detect parasitoid-mediated indirect interactions (II). The close-up in panel C reveals that each colored tag in panel B marks a branch with an elevated leaf mine density. Panel D shows the common garden in Parainen, part of the experiment aimed at estimating the relative importance of host plant genotype and spatial location (III). Panel E shows the transport of trees used in a reciprocal common garden experiment to detect local adaptation (IV). Photos by Ayco Tack and Tomas Roslin.

15 Summary in herbivore communities among common gardens distribution of its host plant – and variation in the is much larger than among host-plant genotypes strength of local adaptation may take place at a – thereby providing the first evidence that host- relatively small spatial scale (among individual trees plant genotype may be of secondary importance within a landscape). in structuring the insect community in nature. In addition, I show that host plant traits are mainly 5. Conclusion affected by host plant genotype, and differ little among locations – a pattern precluding a strong My thesis reveals a multitude of interactions with causal link between host plant traits and insect a potential impact on local community structure. distribution. A more likely explanation for variation Overall, it suggests that in my study system, direct in insect community structure among locations is interactions among herbivores are relatively the spatial context, where species richness increases unimportant at insect densities typically encountered with increasing spatial connectivity of the host tree. in the field (I). Moreover, bottom-up processes like Hence, the dispersal ability of the herbivores is likely induced responses by the host plant (I) and effects to play an important role, where species are more of host plant genotype (III) only weakly affect the likely to colonize well-connected trees and may be mortality and distribution of the herbivores. In absent from isolated trees. contrast, parasitoids – a top-down effect – cause a major fraction of the mortality, either through host- 4.5 Gene flow among local insect populations feeding or through parasitism (I, II). When present, affects local adaptation (IV) indirect interactions mediated by the host plant are negative but weak (I), whereas indirect interactions The idea that insect populations may adapt to mediated by shared parasitoids are absent or positive individual host trees was first proposed by Edmunds (II). Both in terms of evolutionary responses (IV) (1973), who studied a non-dispersive scale insect. and local community structure (I-III), the imprint In later studies, the hypothesis has been referred to of spatial context seems to far override that of local as adaptive deme formation (Edmunds and Alstad forces. This consistent imprint of spatial location 1978, Van Zandt and Mopper 1998). Interestingly, leads me to conclude that a metacommunity view when van Zandt and Mopper (1998) reviewed the will significantly improve our understanding of local literature accumulated in this field by the late nineties, community structure in the focal system. as much as half of the studies actually refuted the As little comparative data exists on plant-herbivore adaptive deme formation hypothesis. Due to these metacommunities, the current study may offer a new mixed results, many reviews have tried (and failed) to baseline for studies to come. While ecologists never pinpoint what makes local adaptation more likely in tire of pointing out exceptions to generalizations, some study systems than in others, as based on traits I would expect that the patterns found in this of either the host plant or the insect (e.g. Boecklen thesis may be successfully extrapolated to a large and Mopper 1998, Van Zandt and Mopper 1998, set of herbivorous insects inhabiting fragmented, Lajeunesse and Forbes 2002, Greischar and Koskella terrestrial habitats. Overall, I predict that direct 2007). and indirect interactions are relatively unimportant In chapter IV, I propose that the adaptive deme in understanding the distribution of herbivores, formation hypothesis should be placed in a spatial whereas spatial processes will play a key role. I hope framework. More specifically, I predict that local that this straightforward prediction will be taken up adaptation is more likely in insect populations that as a challenge by fellow ecologists. receive few immigrants, and that lack of adaptation While my impression is then that many herbivore – or even maladaptation – may occur when a large communities are structured by similar processes as fraction of immigrants enters the local population. described for my study system, I still expect large To estimate the fraction of immigrants on each differences in the strength of forces structuring tree, I use predictions from a spatially explicit, herbivore communities and other terrestrial and empirically-parameterized metapopulation model aquatic communities. For example, in marine of the leaf miner Tischeria ekebladella, as developed communities, direct competition seems to play a by Gripenberg et al (2008). Using a reciprocal much larger role in structuring the local community common garden design, I am then able to show a (Connell 1983). However, recent work suggests the relationship between the strength of local adaptation importance of spatial processes to be paramount and the fraction of immigrants into the population. even in aquatic study systems (Jackson et al. 2001, Hence, local adaptation may depend on the dispersal Cottenie and De Meester 2005). Unfortunately, ability of the herbivore as compared to the spatial many difficulties exist in comparing the factors

16 A metacommunity of herbivores structuring metacommunities, and attempts so far al. 2008). Currently, I hypothesize that density- have been impaired by a lack of data on species- dependent mortality takes place in the leaf litter by specific dispersal abilities (Cadotte and Fukami parasitoids and generalist predators, which aggregate 2005, Jacobson and Peres-Neto 2010). In addition, in areas of high host density. It would be relatively spatial scale may affect metacommunity patterns easy, and fascinating, to look for this often ignored (II, III, IV), and has to be taken into account when life in the undergrowth. comparing the factors structuring different study Third, the abundance and mortality of herbivore systems. The appropriate choice of spatial scale will species may vary in different ways across hierarchical then depend on both the dispersal ability of the scales (e.g. branches and trees), with both ecological species and the structure of the landscape (Thomas and evolutionary implications for herbivore and Kunin 1999). communities (Denno and McClure 1983). As an ecological example, female moths of some species 6. Future perspectives may show a strong oviposition preference for specific branches within a single tree, whereas in other While the chapters in this thesis shed light on some species the female moths may show a preference key ecological and evolutionary questions, many for individual trees (and are largely insensitive to new questions and ideas for follow-up studies heterogeneity within trees). As a result, in the former emerged during the writing-up phase. Part of them species the larval abundance will vary mostly among can be answered by analyzing data already available branches within individual trees, whereas in the on the shelf, whereas others will require additional latter species the larval abundance will vary mainly fieldwork. I am particularly keen on following the among trees. From an evolutionary viewpoint, large next four avenues: differences in larval mortality among branches may First, a combination of modelling and experimental lead to the absence of adaptation, or local adaptation work could reveal how species-specific dispersal to individual branches. In contrast, species that abilities affect the distribution and abundance of experience large differences in mortality among trees species in the landscape. To estimate dispersal are more likely to adapt to individual host trees. That abilities, I have already conducted two experiments: species within a single community indeed respond in the first experiment, I placed potted oak trees at differently to variation between trees is suggested by different distances from an oak stand and scored Figure 6: while the abundance of T. ekebladella seems colonization as the presence of leaf mines and galls to vary randomly across trees and years, the density by the end of the season; in the second experiment of Phyllonorycter is consistently higher on some trees I artificially removed focal herbivore species from than on others. The same notion is also supported ca. 80 trees, and scored herbivore colonization by the pattern detected in chapter IV, where I found in the next year. These experimental data are lower levels of local adaptation in T. ekebladella than currently used to fit a spatially-explicit, empirically- in other species. parameterized metacommunity model to our long- Finally, the reader of this thesis may get the term observational data (Tack, Zheng, Ovaskainen, impression that leaf miners and gallers are the only Gripenberg and Roslin, in preparation). Hopefully, herbivores living on oak leaves. However, in real the results will elucidate how species-specific life the leaf miners and gallers share the oak leaves dispersal abilities, in combination with landscape with a wide variety (but low density) of free-feeding configuration, shape local community structure. insects, and with a specialist powdery mildew Second, in the context of the ‘world is green (Erysiphe alphitoides; see Roslin et al 2007). While hypothesis’, I have so far partly failed to explain why I did not focus on these additional taxa in this thesis the natural densities of herbivores on oak are below (except III), I did not ignore them completely. the carrying capacity set by the plant resources. During the years 2006-2008, I conducted many While I have shown parasitism rate to be density- (relatively) small experiments on the interactions dependent, it does not appear strong enough to among powdery mildew, free-feeding insects, and counter-balance herbivore population growth the leaf miner Tischeria ekebladella. In this context, rates. Hence, other density-dependent factors my preliminary analyses show a complex network of must be evoked to explain the continuously low, interactions. For example, the larvae of T. ekebladella but sometimes heavily fluctuating, densities of the grow faster on mildew-infected leaves, but at the herbivores (Fig. 6). Interestingly, a comparable study same time suffer from increased parasitism. In system of leaf miners on oak in Japan, involving the striking contrast with the increased growth rate same leaf miner genera (but different species) shows of T. ekebladella, free-feeding insects grow slower very similar population dynamics (Nakamura et on mildew-infected leaves. Revealing the exact

17 Summary

interaction network among these guilds, and the proposed; Tommi, Raimo, Piritta and Hiski for all those consequences for colonization-extinction dynamics lovely days in the field and greenhouse, driving the mönkijä, of the insect herbivores, will be part of my post- chilling out at rantasauna, your great care of our beloved doctoral studies. oak trees (which should have been your lowest priority), and always borrowing fieldworkers when my experiments ran out of control; tens (?) of forestry students who 7. Acknowledgements volunteered (or were forced) to help with insect research: Following scientific tradition, I am writing the hope you enjoyed this glimpse into ecology! \Students: acknowledgments in the very last moment – a Liselotte, my great help during those uncertain early days strategic mistake given the knowledge that this of experimentation – you were a great companion with so section will, without any doubt, be the heaviest-read many projects; Elena, while you must sometimes have felt part of the thesis. Many of you may even have gone lost in the deep quietness reigning the archipelago, I am straight to this piece of text, ignoring the summary sure you miss it! \opponent: I am deeply grateful that you itself. Then, by keeping it short, I will give the reader accepted the invitation to act as an opponent - on such some extra time to actually browse through (or even a short time notice - within a split-second \5310: here I read!) some of the chapters. was located in (probably) the best-hidden corner of the fifth floor; with Chris, his great stories and unique ability *This thesis would not have been possible without the to show up and disappear without any regularity; with help of numerous collaborators, colleagues and friends* Kristjan and haphazard daily chats, admirable knowledge of Statistix, and ability to hit the right light switch – be \Tomas Roslin: a superb, superior supervisor; assured your garden will flourish when you return back strikingly, even after four years I do not grasp how you to Finland; Eva with her lovely daily life stories; I am only always find the time to provide such lightning-speed, high- worried that when I leave, the office will suddenly turn quality, stimulating, encouraging, optimistic, and realistic clean and organized \ football table: there is no better comments on my work - I must be blessed with such a way to relax than beat a few fellows in fusball \Ilkka: for guidance \Oak-group: many people are blinded by fluffy being one of those bright lights in ecology – and thereby and shiny butterflies. It then is a great delight to providing one of the best places one could imagine to do have colleagues who perceive the true beauty of nature: its PhD; \coffee breaks: where would science be without those tiny critters that rule the earth \Sofia: the one who the joyful moments during early & late coffee? And will bravely (or naively?) labelled the very first oak tree on science improve with the new coffee machine? \Funding: Wattkast – it is great to see how many papers have come Academy of Finland (grant to Tomas) and Societas about, and will still be published, as a result. Thanks for Entomologica Helsingforsiensis\co-authors: Otso and his guiding me during my first steps in the oak-world, and magic data-handling, Phil who so kindly wrote my R-code, being a friend on the continuing journey through the and hey, I mentioned the others already above \MRG: this wondrous world of plants and insects \Riikka: your wonderful, fun and friendly group of dedicated scientists enthusiasm of higher trophic levels like birds, parasitoids creates an amazing scientific and non-scientific atmosphere; (and good beer?) is unrivalled – and highly infectious. idem for the department; \Luova for the courses, and Anni Thanks for sharing the rustic life in Böhle, the PhD-journey for recent help with bureaucracy and the mini-symposium and those cool conference trips \Böhle: this lovely village \Veijo for his quick and cool handling of dissertation- in the middle of Korpo, with its charming wooden house, formalities and last-minute changes \MRG secretaries: Finland’s best wood-heated sauna, its friendly neighbours, this most-gifted, patient and oh-so-friendly buffer zone in and its water supply that runs out during the most critical between me and university bureaucracy – without you I’d and hectic end-of-the-season \Fieldworkers: over the years probably still be busy finding the right form in alma; Sami about fifteen excellent, cheerful and hardship-enduring for fine-tuning my computer, and the splendid layout of the fieldworkers joined for the autumn survey – and made it current thesis \Jasper: for spending so many hours of his such a pleasant time; in particular Bess, Janne and Elina ‘non-working’-career on my cover design\pre-examiners: greatly helped me out with a zillion other jobs \Wattkast: Tero and Peter, for so graciously scrutinizing my thesis, this idyllic island full of tagged oaks, insects, mushrooms insightful comments and encouraging words; \external and friendly inhabitants: the Roslin-family with its chats, supervisors: Heikki, Jari & Pekka: I am very thankful plums and drinking water; the ‘greenhouse-people’ with that you voluntarily, and with dedication, followed the their fresh vegetables; Henry & Thomas for driving around development of my thesis\friends: obvious \sports: hundreds of oaks and thousands of litres of water around playing ice-hockey in Brahe, throwing frisbees along the the island in the name of science \Haapastensyrjä tree bay, and skiing to work are the best ways to keep the mind breeding station: the great leader Pertti and his fondness of fresh and the body in condition. \Family: my parents for oaks – and his willingness to support any experiment I ever their unconditional and ever-lasting support (and bringing 18 A metacommunity of herbivores the booze for the karonkka); Sanna, my shining light for Cottenie, K. 2005. Integrating environmental and so many years; I am deeply grateful for your ability to run spatial processes in ecological community dynamics. our family affairs, where I may often bring more chaos than Ecology Letters8 :1175-1182. order (but one day I’ll make up for that!), your delicious Cottenie, K. and L. De Meester. 2005. Local interactions food and heavenly cakes, and the secret behind those cute and local dispersal in a zooplankton metacommunity. trees in my presentation; (do you already know what’s a Pages 189-211 in M. Holyoak, M. A. Leibold, and R. D. Holt, editors. Metacommunities: spatial dynamics leaf miner?); Helmi, for making me feel such a proud and ecological communities. University of Chicago father, clearing my head from scientific thoughts with a Press, Chicago, USA. single smile, and for playing together with all those cool Cronin, J. T. 2007. Shared parasitoids in a kids toys; and my cats, one soft and affectionate, taking metacommunity: indirect interactions inhibit the important role of paper-weight (preventing any home- herbivore membership in local communities. Ecology study), while the other gives this lovely anarchist touch to 88:2977-2990. the house. Denno, R. F. and M. S. McClure, editors. 1983. Variable plants and herbivores in natural and managed Most likely I forgot to thank somebody. In case you systems. Academic Press, New York. feel neglected, please visit room 5310 for a personal Denno, R. F., M. S. McClure, and J. R. Ott. 1995. Interspecific interactions in phytophagous insects: acknowledgment. competition reexamined and resurrected. Annual Review of Entomology 40:297-331. 8. References Driscoll, D. A., J. B. Kirkpatrick, P. B. McQuillan, and K. J. Bonham. 2010. Classic metapopulations are Amarasekare, P. 2008. Spatial dynamics of food rare among common beetle species from a naturally webs. Annual Review of Ecology, Evolution, and fragmented landscape. Journal of Animal Ecology Systematics 39:479-500. 79:294-303. Antonovics, J. 1992. Towards community genetics. Driscoll, D. A. and D. B. Lindenmayer. 2009. Empirical Pages 426-449 in R. S. Fritz and E. L. Simms, editors. tests of metacommunity theory using an isolation Plant resistance to herbivores and pathogens: ecology, gradient. Ecological Monographs 79:485-501. evolution and genetics. University of Chicago Press, Edmunds, G. F. 1973. Ecology of black pineleaf Chicago, USA. scale (Homoptera: Diaspididae). Environmental Askew, R. R. 1961. On the biology of the inhabitants Entomology 2:765-777. of oak galls of Cynipidae (Hymenoptera) in Britain. Edmunds, G. F. and D. N. Alstad. 1978. Coevolution Transactions of the Society for British Entomology in insect herbivores and conifers. Science 199:941- 14:237-268. 945. Bell, G. 2001. Neutral macroecology. Science 293:2413- Ehrlich, P. R. and I. Hanski, editors. 2004. On the wings 2418. of checkerspots: a model system for population Boecklen, W. J. and S. Mopper. 1998. Local adaptation biology. Oxford University Press, Oxford, UK. in specialist herbivores: theory and evidence. Pages Elton, C. S. 1927. Animal Ecology. Sidgwick & Jackson, 64-88 in S. Mopper and S. Y. Strauss, editors. Genetic London, UK. structure and local adaptation in natural insect Ferris, C., R. A. King, R. Vainola, and G. M. Hewitt. populations. Chapman & Hall, New York, USA. 1998. Chloroplast DNA recognizes three refugial Burstein, M. and D. Wool. 1992. Great tits exploit aphid sources of European oaks and suggests independent eastern and western immigrations to Finland. galls as a source of food. Ornis Scandinavica 23:107- Heredity 80:584-593. 109. Freestone, A. L. and B. D. Inouye. 2006. Dispersal Cadotte, M. W. and T. Fukami. 2005. Dispersal, limitation and environmental heterogeneity spatial scale, and species diversity in a hierarchically shape scale-dependent diversity patterns in plant structured experimental landscape. Ecology Letters communities. Ecology 87:2425-2432. 8:548-557. Fritz, R. S. 1992. Community structure and species Connell, J. H. 1961. Effects of competition, predation interactions of phytophagous insects on resistant by Thais lapillus, and other factors on natural and susceptible host plants. Pages 240-277 in R. S. populations of the barnacle Balanus balanoides. Fritz and E. L. Simms, editors. Plant resistance to Ecological Monographs 31. herbivores and pathogens: ecology, evolution and Connell, J. H. 1983. On the prevalence and relative genetics. University of Chicago Press, Chicago. importance of interspecific competition - evidence from field experiments. American Naturalist Godfray, H. C. J. 1994. Parasitoids: behavioral and 122:661-696. evolutionary ecology. Princeton University Press, Connor, E. F. and M. P. Taverner. 1997. The evolution Princeton, USA. and adaptive significance of the leaf-mining habit. Gonzalez, A. 2005. Local and regional community Oikos 79:6-25. dynamics in fragmented landscapes: insights from a bryophyte-based natural mircrocosm. Pages 146-

19 Summary

169 in M. Holyoak, M. A. Leibold, and R. D. Holt, factors. Canadian Journal of Fisheries and Aquatic editors. Metacommunities: spatial dynamics and Sciences 58:157-170. ecological communities. University of Chicago Press, Jacobson, B. and P. R. Peres-Neto. 2010. Quantifying Chicago, USA. and disentangling dispersal in metacommunities: Greischar, M. A. and B. Koskella. 2007. A synthesis how close have we come? How far is there to go? of experimental work on parasite local adaptation. Landscape Ecology 25:495-507. Ecology Letters10 :418-434. Janzen, D. H. 1968. Host plants as islands in Gripenberg, S., O. Ovaskainen, E. Morriën, and T. evolutionary and contemporary time. American Roslin. 2008. Spatial population structure of a Naturalist 102:592-595. specialist leaf-mining moth. Journal of Animal Janzen, D. H. 1977. Promising directions of study in Ecology 77:757-767. tropical animal-plant interactions. Annals of the Hanski, I. 1999. Metapopulation ecology. Oxford Missouri Botanical Garden 64:706-736. University Press, Oxford, UK. Johnson, M. T. J. and J. R. Stinchcombe. 2007. An Hanski, I. 2005. The shrinking world: ecological emerging synthesis between community ecology and consequences of habitat loss. International Ecology evolutionary biology. Trends in Ecology & Evolution Institute, Oldendorf/Luhe, Germany. 22:250-257. Hanski, I. and O. Gaggiotti, editors. 2004. Ecology, Kaplan, I. and R. F. Denno. 2007. Interspecific genetics and evolution of metapopulations. Elsevier interactions in phytophagous insects revisited: a Academic Press, Amsterdam. Hanski, I. and D. Simberloff. 1997. The metapopulation quantitative assessment of competition theory. approach, its history, conceptual domain, and Ecology Letters10 :977-994. application to conservation. Pages 5-25 in I. A. Karban, R. and I. T. Baldwin. 1997. Induced responses Hanski and M. E. Gilpin, editors. Metapopulation to herbivory. University of Chicago Press, Chicago, biology: ecology, genetics and evolution. Academic USA. Press, San Diego, USA. Kendi, F. D., H. Marcel, A. P. Kim, A. O. Valerie, and Hassell, M. P. 1978. The dynamics of L. Quenby. 2009. Factors controlling community predator-prey systems. Princeton University Press, structure in heterogeneous metacommunities. Princeton, USA. Journal of Animal Ecology 78:937-944. Hassell, M. P. 2000. The spatial and temporal dynamics Kolasa, J. and T. N. Romanuk, editors. 2005. Assembly of host-parasitoid interactions. Oxford University of unequals in the unequal world of a rock pool Press, Oxford, UK. metacommunity. University of Chicago Press, Hawkins, B. A., H. V. Cornell, and M. E. Hochberg. Chicago, USA. 1997. Predators, parasitoids, and pathogens as Labandeira, C. C. 1998. Early history of arthropod and mortality agents in phytophagous insect populations. vascular plant associations. Annual Review of Earth Ecology 78:2145-2152. and Planetary Sciences 26:329-377. Holt, R. D. 1977. Predation, apparent competition, Lajeunesse, M. J. and M. R. Forbes. 2002. Host range and the structure of prey communities. Theoretical and local parasite adaptation. Proceedings of the Population Biology 12:197-229. Royal Society of London Series B-Biological Sciences Holyoak, M., M. A. Leibold, and R. D. Holt, editors. 269:703-710. 2005. Metacommunities: spatial dynamics and Leibold, M. and T. E. Miller. 2004. From ecological communities. University of Chicago Press, metapopulations to metacommunities. Pages 133- Chicago, USA. 150 in I. Hanski and O. E. Gaggiotti, editors. Ecology, Howard, L. O. and W. F. Fiske. 1911. The importation genetics and evolution of metapopulations. Elsevier into the United States of the parasites of the gypsy Academic Press, London, UK. moth and the brown-tail moth. Bulletin of the Bureau Leibold, M. A., M. Holyoak, N. Mouquet, P. of Entomology, US Department of Agriculture 91:1- Amarasekare, J. M. Chase, M. F. Hoopes, R. D. 312. Holt, J. B. Shurin, R. Law, and D. Tilman. 2004. The Hubbell, S. P. 2001. The unified neutral theory of metacommunity concept: a framework for multi- biodiversity and biogeography. Princeton University scale community ecology. Ecology Letters 7:601- Press, Princeton, USA. 613. Huffaker, C. 1958. Experimental studies on predation: dispersion factors and predator-prey oscillations. Levins, R. 1970. Extinction. Pages 77–107 in M. Hilgardia 27:795-835. Gesternhaber, editor. Some mathematical problems Itämies, J. and M. Ojanen. 1975. Autumn predation of in biology. American Mathematical Society, Parus major and Parus montanus upon two leaf-mining Providence, USA. species of Lithocolletis (Lepidoptera, Lithocolletidae). Losey, J. E. and M. Vaughan. 2006. The economic value Annales Zoologici Fennici 14:235-241. of ecological services provided by insects. Bioscience Jackson, D. A., P. R. Peres-Neto, and J. D. Olden. 56:311-323. 2001. What controls who is where in freshwater fish Lotka, A. J. 1925. Elements of physical biology. Williams communities - the roles of biotic, abiotic, and spatial and Wilkins, Baltimore, MD, USA.

20 A metacommunity of herbivores

Lotka, A. J. 1932. The growth of mixed populations: Strong, D. R., J. H. Lawton, and T. R. E. Southwood. two species competing for a common food supply. 1984. Insects on plants. Harvard University Press, Journal of the Washington Academy of Sciences Cambridge, USA. 22:461-469. Thomas, C. D. and W. E. Kunin. 1999. The spatial MacArthur, R. H. 1958. Population ecology of some structure of populations. Journal of Animal Ecology warblers of northeastern coniferous forests. Ecology 68:647-657. 39:599-619. Urban, M. C. 2004. Disturbance heterogeneity Memmott, J. and H. Godfray. 1993. Parasitoid webs. determines freshwater metacommunity structure. Pages 217-234 in J. LaSalle and I. D. Gauld, editors. Ecology 85:2971-2978. Hymenoptera and biodiversity. CAB International, Valverde, T. and J. Silvertown. 1997. A metapopulation Wallingford, UK. model for Primula vulgaris, a temperate forest Morris, R. J., O. T. Lewis, and H. C. J. Godfray. 2005. understorey herb. Journal of Ecology 85:193-210. Apparent competition and insect community Van Driesche, R., M. Hoddle, and T. Center. 2008. structure: towards a spatial perspective. Annales Control of pests and weeds by natural enemies: Zoologici Fennici 42:449-462. an introduction to biological control. Blackwell Murdoch, W. W. 1966. “Community structure, Publishing, Malden, MA, USA. population control and competition” - a critique. van Nouhuys, S. and I. Hanski. 2005. Metacommunities American Naturalist 100:219-226. of butterflies, their host plants, and their parasitoids. Nakamura, T., K. Hattori, T. A. Ishida, H. Sato, and M. Pages 99-121 in M. Holyoak, M. A. Leibold, and R. T. Kimura. 2008. Population dynamics of leafminers D. Holt, editors. Metacommunities: spatial dynamics on a deciduous oak Quercus dentata. Acta Oecologica and ecological communities. University of Chicago 34:259-265. Press, Chicago, USA. Nicholson, A. J. 1933. The balance of animal populations. Van Zandt, P. A. and S. Mopper. 1998. A meta-analysis Journal of Animal Ecology 2:131-178. of adaptive deme formation in phytophagous insect Owen, D. F. 1975. The efficiency of blue tits Parus populations. American Naturalist 152:595-604. caeruleus preying on larvae of Phytomyza ilicis. Ibis Vanschoenwinkel, B., C. De Vries, M. Seaman, and 117:515-516. L. Brendonck. 2007. The role of metacommunity Pandit, S. N., J. Kolasa, and K. Cottenie. 2009. Contrasts processes in shaping invertebrate rock pool between habitat generalists and specialists: an communities along a dispersal gradient. Oikos empirical extension to the basic metacommunity 116:1255-1266. framework. Ecology 90:2253-2262. Whitham, T. G., J. K. Bailey, J. A. Schweitzer, S. M. Price, P. W. 1983. Hypotheses on organization and Shuster, R. K. Bangert, C. J. Leroy, E. V. Lonsdorf, G. evolution in herbivorous insect communities. Pages J. Allan, S. P. DiFazio, B. M. Potts, D. G. Fischer, C. A. 559-598 R. F. Denno and M. S. McClure, editors. in Gehring, R. L. Lindroth, J. C. Marks, S. C. Hart, G. Variable plants and herbivores in natural and managed M. Wimp, and S. C. Wooley. 2006. A framework for systems. Academic Press, New York, USA. community and ecosystem genetics: from genes to Ranius, T. 2000. Minimum viable metapopulation size of a beetle, Osmoderma eremita, living in tree hollows. ecosystems. Nature Reviews Genetics 7:510-523. Animal Conservation 3:37-43. Whitham, T. G., W. P. Young, G. D. Martinsen, C. Ranius, T. 2001. Constancy and asynchrony of A. Gehring, J. A. Schweitzer, S. M. Shuster, G. M. Osmoderma eremita populations in tree hollows. Wimp, D. G. Fischer, J. K. Bailey, R. L. Lindroth, S. Oecologia 126:208-215. Woolbright, and C. R. Kuske. 2003. Community and Roslin, T., A.-L. Laine, and S. Gripenberg. 2007. Spatial ecosystem genetics: a consequence of the extended population structure in an obligate plant pathogen phenotype. Ecology 84:559-573. colonizing oak Quercus robur. Functional Ecology Williamson, M. H. 1957. Elementary theory of 21:1168-1177. interspecific competition. Nature180 :422-425. Samways, M. J. 2005. Insect diversity conservation. Volterra, V. 1926. Variations and fluctuations of the Cambridge University Press, Cambridge, UK. numbers of individuals in animal species living Sato, H. and S. Higashi. 1987. Bionomics of Phyllonorycter together (reprinted and translated in 1931. In: R.N. (Lepidoptera, Gracillariidae) on Quercus. II. Effects Chapman, Animal Ecology. McGraw Hill, New of ants. Ecological Research 2:53-60. York). Stewart, A. J. A., T. R. New, and O. T. Lewis, editors. Yamazaki, K. and S. Sugiura. 2008. Deer predation on 2007. Insect conservation biology. CABI Publishing, leaf miners via leaf abscission. Naturwissenschaften Wallingford, UK. 95:263-268.

21